EPA-600/2-76-212
ERDA 47
AUGUST 1976
Environmental Protection Technology Series
SYMPOSIUM ON ENVIRONMENT AND
ENERGY CONSERVATION
(November 1975, Denver, Colorado)
Energy Research and Development Administration
Office of the Assistant Administrator
for Conservation
Washington, D.C.
Industrial Environmental
Research Laboratory
Research Triangle Park, NC 27711
-------
EPA-600/2-76-212
ERDA 47
August 1976
SYMPOSIUM ON
ENVIRONMENT AND ENERGY CONSERVATION
(NOVEMBER 1975, DENVER, COLORADO)
Franklin A. Ayer (Compiler)
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, NC 27709
Contract No. 68-02-1325, Task 29
ROAP No. 21BBZ-012
Program Element No. 1AB013
and
Purchase Order No. WA-76-3777
EPA Project Officer: W. B. Steen
Industrial Environmental
Research Laboratory
Research Triangle Park, NC 27711
ERDA Project Officer: K. D. DeVine
Office of the Assistant
Administrator for Conservation
Washington, DC 20545
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
and
U.S. ENERGY RESEARCH AND DEVELOPMENT ADMINISTRATION
Office of the Assistant Administrator for Conservation
Washington, DC 20545
-------
This report has been reviewed by the Environmental Protection Agency
(EPA) and the Energy Research and Development Administration
(ERDA) and approved for publication. Approval does not signify that
the contents necessarily reflect the views and policies of EPA or ERDA,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
ii
-------
FOREWORD
The proceedings for the symposium on "Environment and Energy Conserva-
tion" is the final report submitted to the I ndustrial Environmental Research
Laboratory for the Environmental Protection Agency, Research Triangle
Park, North Carolina (Contract No. 68-02-1325), and the Energy Research
and Development Administration, Office of the Assistant Administrator for
Conservation, Washington, D.C. (Purchase Order No. WA-76-3777). The
symposium was held at the Denver Marriott Hotel, Denver, Colorado, on 3-6
November 1975.
The principle objective of this symposium was to identify the environmental
benefits and threats of alternative energy conservation systems and to com-
pare the environmental impacts of energy conservation strategies. It was also
the objective of the symposium to anticipate and publicize environmental
impacts so as to: (1) facilitate proper consideration of these impacts in the
assignment of funding priorities for energy conservation programs under the
national energy R&D program and (2) begin work in a timely manner to solve
future environmental problems.
I
Mr. Walter B. Steen, Chemical Engineer, Energy Assessment and Control
Division, Industrial Environmental Research Laboratory, Environmental Pro-
tection Agency, Research Triangle Park, North Carolina, and Mr. Kelly D.
DeVine, Environment and Safety Specialist, Office of the Assistant Adminis-
trator for Conservation, Energy Research and Development Administration,
Washington, D.C., were Co-General Chairmen of the symposium.
Mr. Franklin A. Ayer, Manager, Technology and Resource Management
Department, Center for Technology Applications, Research Triangle Insti-
tute, Research Triangle Park, North Carolina, was the Symposium Coordina-
tor and Compiler of the proceedings.
iii
-------
Table of Contents
(*indicates speaker)
Page
3 November 1975
Introduction and Opening Remarks
P. P. Turner
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Energy Conservation Policy and Overview
Dennis W. Bakke
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4
The ERDA Conservation Program: The Technology
of Improved Efficiency. . . . . . . . . . . . . " . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
James S. Kane, Ph.D.
Environmental Implications of Energy Conservation. . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Stephen J. Gage, Ph.D.
Session I: NATIONAL ENERGY CONSERVATION STRATEGIES. . . . . . . . . . . . . .. . . . . 21
Rene R. Bertrand, Ph.D., Session Chairman
Environmental Implications of High Energy Use
Mark D. Levine, Ph.D.
. . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Implications of a National Conservation Strategy
Kenneth R. Woodcock
(presented by Joan Davenport)
. . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Legislative Programs in Energy Conservation
Maxine L. Savitz, Ph.D.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Highlights of Energy Conservation Programs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Howard Hagler* and
Harvey M. Bernstein
The Effect of Energy Components on System
Costs and Efficiency. . . . . . . . . . . . .
T. J. Thomas, Ph.D.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
ERDA's Conservation Objectives in the Transportation Sector. . . . . . . . . . . . . . . . . . . . 69
John J. Brogan
Session II: NONTECHNOLOGICAL METHODS TO CONSERVE ENERGY. . . . . . . . . . . . . . 73
David R. Berg, Session Chairman
Environmental Implications of Nontechnological
Methods to Conserve Energy. . . . . . . . . .
John H. Gibbons
. . . . . . . . . . . . . . . . . . . . . . . . . . . 75
v
-------
Table of Contents (can.)
Page
Federal Initiatives to Save Energy in
Lighting and Appliances. . . . . .
Kur1 W. Riegel, Ph.D.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
4 Novemb"r 1975
The De~iign of Electricity Tariffs
Charles J. Cicchetti, Ph.D.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
OpportlJnities in Electrical Load Management
Dou!llas C. Bauer, Ph.D.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Policies to Reduce Transportation Fuel Use
Eric Hirst, Ph.D.
. . . . . . . . . . . . . . . . . . . . . . . . . . . .. .111
Session III RETROFITTING OUR PRESENT-DAY ENERGY SYSTEMS. . . . . . . . . . . . . ..125
Roger S. Carlsmith, Session Chairman
Near-Te-m Potential for Improved Automobile
Fuel Economy. . . . . . . . . . . . . . . .
Karl '-I. Hellman, Ph.D.,* and
John P. DeKany
. . . . . .
. . . . . . . . . . . . . . . . . . . . . .127
Thermodynamic Analysis of Industrial Energy
Conservation Opportunities. . . . . . . . . .
Elton H. Hall, Ph.D.
. . . . . . . . . . . . . . . . . . . . . . . . . . . .131
LUNCHEON ADDRESS
Energy a nd Environmental Resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145
John A. Green
Session III Icon.): RETROFITTING OUR PRESENT-DAY ENERGY SYSTEMS
Roger S. Carlsmith. Session Chairman
. . . . . . . . . . .151
Energy Conservation Techniques Applicable to the
Iron Foundry's Cupola. . . . . . . . . . . . . .
Dennis J. Martin
. . . . . . . . . . . . . . . . . . . . . . . .
. .153
Energy Lise, Efficiency, and Conservation in Industry
Robert H. Essenhigh, Ph.D.
. . . . . . . . . . . . . . . . . . . . . . . .161
Session IV: CENTRAL POWER STATIONS
Fred L. Robson, Ph.D., Session Chairman
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189
Waste Heat Utilization/Reduction
A. G.:hristianson and
D. J. Cannon*
. . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .191
vi
-------
Table of Contents (con.)
Page
Staged Combustion in Combined-Cycle Supplementary
Fired Boilers. . . . . . . . . . . . . . . . . . . . . .
D. R. Bartz, *
S. C. Hunter, and
J. K. Arand
. . . . . . . . . . : . . . . . . . . . . . .196
5 November 1975
Electric Power Transmission Modes, Means, and Style
Robert W. Flugum
. . . . . . . . . . . . . . . . . . . . . . . .212
Geothermal Energy Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .217
Paul Kruger
Panel: ENVIRONMENTAL AND CONSERVATION EFFECTS OF
ADVANCED POWER SYSTEMS
Conservation and Environmental Implications of
Open-Cycle MHD . . . . . . . . . . . . . . . .
Finn A. Hals
. . . . . . . . . . . . . . . . . . . . . . . . . . .230
Energy Conservation Through the Use of
Combined-Cycle Power Systems
Albert J. Giramonti* and
William A. Blecher
. . . . . - . . . . . . . . . . . . . . . . . . . . . . . . .249
Gas Turbine HTGR for Economical Dry Cooling
or High-Efficiency Combined Cycle. . . . . . .
John M. Krase
. . . . . . . . . . . . . . . . . . . . . . . . . . .268
Environmental and Safety Considerations for a Fossil
Fuel-Fired Potassium-Steam Binary Vapo~ Cycle. . . . . . . . . . . . . . . . . . . . . . . . . . .279
A. P. Fraas and
Robert S. Holcomb *
A Topping Cycle for Coal-Fueled Electric Power
Plants Using the Ceramic Helical Expander. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .297
B. Myers,*
R. Landingham,
P. Mohr, and
R. Taylor
Balance-of-Plant Power Requirements for
Advanced Power Systems. . . . . . . .
Irving L Chait
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .311
Session V: RESIDENTIAL AND COMMERCIAL ENERGY SYSTEMS
Robert Roscnber~l, Ph.D., Session Chairman
. . . . . . . . . . . . . . . .321
vii
-------
Table of Contents (con.)
Page
The Potential Role of Solar Energy in Conservation for
Buiiding Heating and Cooling. . . . . . . . . . . . .
Frank R. Biancardi * and
Maurice D. Meader
. . . . . . . . . . . . . . . . . . . . . . . .323
Eva uation of Engine-Driven Heat Pump Systems
of Small Capacities. . . . . . . . . . . . . . .
Jaroslav Wurm* and
(iopal P. K. Panikker
. . . . . . . . . . . . . . . . . . . . . . . . . . .333
Development and Application of an Optimum Distillate
Oil 3urner Head for Residential Furnaces. . . . . . . .
l.. P. Combs, *
W. H. Nurick, and
JI. S. Okuda
. . . . . . . . . . , . . . . . . . . . . . .347
Eva.uation of the Potential Impact of Alcohol Fuels on Pollutant
Emissions and Energy Requirements of Area Sources. . . . . . .
G. Blair Martin
. . . . . . . . . . . . . . . . . .363
Assl'ssment of the Applicability of Catalytic Oxidation
of Hydrocarbon and Other Fuels for Control of NOx
and Other Pollutants From Area Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .373
J. P. Kesselring, Ph.D.,*
Ci. B. Martin,
F:. A. Brown, and
C. B. Moyer, Ph.D.
Hea': Pipe Appliances. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .390
J3rnes F. Rice and
Edward F. Searight *
6 Novenber 1975
Session VI: NEW INDUSTRIAL PROCESSES
c: . Ray Smithson, Session Chairman
, . . . . . . , . . . . . . . . . . . . . . . , . . . . .403
Pro!=erties of Industrial Fuel Gases Manufactured From
Coal Using Commercially Proven Technology. . . . . .
James I. Joubert, Ph.D.,* and
Daniel Bienstock
. . . . . . . . . . . . . . . . . . . . . . .405
The EPA R&D Program in Wastes-as-Fuel: An Overview Focusing
on Process Environmental/Energy Impacts. . . . . . . . . . . .
George L. Huffman
. . . . . . . . . . . . . . . . . .422
Preliminary Assessment of the Role of Energy Storage and Implicated
Tec~ nologies for Ener!IY Conservation in Industry. . . . . . . . . .
Donald R. Glenn
. . . . . . . . . . . . . . . .435
viii
-------
Table of Contents (con.)
Page
Energy Conservation by the Renovation of High-Temperature
Textile Wastewater. . . . . . . . . . . . . . . . . . . . . .
Craig A. Brandon, Ph.D.
. . . . . . . . . . . . . . . . . . . .476
New Processes and Energy Conservation in the Primary
Metals Industry. . . . . . . . . . . . . . . . . . . .
R. H. Cherry, Jr., Ph.D.
. . . . . . . . . . . . . . . . . . . . . . . .485
Energy Conserving Industrial Changes and Their
Environmental Impact . . . . . . . . .
Herbert S. Skovronek, Ph.D.
. . . . . . . . . . . . . . . . . . . . . . . . . . .505
ix
-------
3 November 1975
INTRODUCTION AND OPENING REMARKS
P. P. Turner*
*Chief, Advanced Process Branch, Energy Assessment and Control Division, Industrial Environmental Research
Laboratory, Environmental Protection Agency, Research Triangle Park, North Carolina.
-------
2
-------
INTRODUCTION AND OPENING REMARKS
P. p, Turner*
Good morning, I am Pic Turner of EPA and I am
delighted that you can be here. Our symposium on
"Environment and Energy Conservation" is sponsored
jointly by EPA and ERDA.
Dr. John K. Burchard, Director of EPA Industrial
Environmental Research Laboratory of Research Tri-
angle Park, North Carolina, was originally scheduled to
make the welcoming address. Dr. Wilson Talley of EPA
headquarters had last minute demands on John's time
and it is with deep regret that he cannot be here. Speak-
ing for the Environmental Protection Agency, I extend a
hearty welcome to the Energy Research and Develop-
ment Administration, which joins us as cosponsors of
this symposium on energy conservation and the environ-
ment. Since the two issues--energy conservation and the
environment-are inseparable, this cosponsorship is quite
appropriate in our search for an environmentally clean
solution to our country's energy problems. Everyone is
aware of these problems. They range from where they
are most evident, in the not-so-broad American pocket-
.Chief, Advanced Process Branch, Energy Assessment and
Control Division, Industrial Environmental Research Laboratory,
Environmental Protection Agency, Research Triangle Park,
North Carolina.
book, to where they are not quite so evident, but
certainly more important, the not-so-happy American
body.
The overall program represents an international
balance of expertise from industrial, academic, and
governmel,1t communities. The purposes of this sympo-
sium are: to identify the environmental benefits and
threats of alternative energy conservation systems; to ex-
amine our energy conservation strategies and their impli-
cations; to look at several nontechnical methods of con-
serving energy; to consider the implications of retro-
f i tt i ng present-day energy systems, central power
stations, and residential and commercial energy systems;
and to examine the environmental impact of certain in-
dustrial processes.
As a spokesman for both the EPA and ERDA, I
extend a hearty welcome to the speakers scheduled for
this timely and important meeting. Each speaker is fully
qualified in his or her own particular facet of this com-
plex arena of research, development, demonstration,
legislation, and enforcement activities. And speaking
both for the cosponsors and for the speakers, I extend
the heartiest welcome to you who are usually referred to
at these meetings as attendees. Your presence is proof of
the interest, time, energy, and effort being extended by
many talented and productive people to help meet the
energy and environmental conservation challenge. Once
again, welcome to Denver.
3
-------
ENERGY CONSERVATION POLICY AND OVERVIEW
Dennis W. Bakke*
Abstract
Consen'ation is the cost-effective substitution of
capital, tim~, or labor for energy consumption. Viewed
from this perspective, energy conservation becomes an
equal partner with efforts to increase energy supplies.
Quite literally, a barrel of oil saved is equal to a barrel of
oil producec ,.
The ovorall potential for energy conservation, using
existing tec,'mology, is enormous. Because conservation
is in the economic self-interest of all energy users, a large
portion of the potential will be achieved without further
Federal intt'rvention. Nevertheless, the Federal Energy
. Administrat:on (FEA) estimates that additional savings
of about 7.5 million barrels per day are achievable by
1985. Withcut the enactment of pending legislation and
the expansion of existing programs, however, only about
half of the potential savings in 1985 is likely to be
realized.
By comparing the cost of such conservation meas-
ures as ins railing insulation and storm windows or
manufacturi.7g more fuel-efficient automobiles to the
cost of incMBsing energy supplies, it is clear that con-
servation is our cheapest future "source" of energy.
There howe been numerous proposals to accelerate
the Federal ,mergy conservation effort, which has up to
now been ccnstrained by limited funding and authority.
But, unfortunately, the Administration and Congress
remain dead,'ocked on many key issues. Hopefully, these
issues can soon be resolved, enabling us to stop talking
about the poblem and begin concentrating on solving it.
It is no"" generally accepted that energy conserva-
tion should be a major objective of any national energy
policy. But there is still much disagreement on what
specific steps should be taken and, I believe, much
misunderstar ding of the real meaning of energy con-
servation.
The Cor:gress and the Administration remain dead-
locked over many key issues, and the ensuing confusion
has encouraged public indifference. Nevertheless, I feel
there is sti II ! ome hope that the major issues can soon be
resolved, perllitting us to stop talking about the prob-
lem and begin concentrating on solving it.
. Deputy II,ssistant Administrator for Energy Conservation
and Environmnnt, Federal Energy Administration, Washington,
D.C.
Before discussing my own views on energy conserva-
tion and summarizing the status of Federal programs and
legislation, I would like to list the basic components of
our national dilemma on which there is general agree-
ment. .
1. Domestic reserves of oil and natural gas are
limited-and declining at an alarming rate;
2. A heavy reliance on imported petroleum poses a
serious threat to both our economy and national
security;
3. Expanding domestic supplies of energy, whether
they be from conventional or exotic sources, will
take many years, and these new supplies are likely
to be increasingly costly;
4. Energy supply, conversion, and use often pose
serious threats to environmental quality.
Some time ago, largely because of a growing recog-
nition that these conflicting problems could not be
easily resolved, many persons began seriously to examine
the potential for reducing energy demand. It could not
be disputed that the rapid growth in the demand for
energy during the past two decades was largely respon-
sible for our heavy reliance on imported oil. Further-
more, the prolifigate waste of energy seemed to be self-
evident. But was it? Critics of conservation asserted that
economic growth was dependent upon increasing energy
consumption, and, at first glance, this conclusion
appeared to be supported by similarities between the
growth in GNP and energy demand and by a simple
comparison of the energy use per capita of the United
States and that of any underdeveloped country. In addi-
tion, it was apparent that turning off lights, eliminating
electric toothbrushes, and driving slower certainly would
not solve the problem.
Fortunately, these criticisms did not deter thl!
advocates of conservation. The list of opportunities fO!
saving energy, together with our perception of the over-
all potential, rapidly expanded. It soon became evident
that conservation did not necessarily pose a threat to
economic growth. On the contrary, saving energy could
lead to higher productivity without any loss in comfort,
services, or goods. In fact, because energy costs are likely
to continue to rise in the future (fig. 1), it is essential
that we use energy more efficiently if we are to maintain
our quality of life and continue economic growth.
Out of these initial explorations, there emerged a
better understanding of what "energy conservation"
really meant-in both economic and social terms.
4
-------
COST PER BARREL($)
49.00
50.00
51.00
D BTU'S OF
HOME HEATING
~ BTU'S OF FUEL
~ BURNED AT RETAIL
40.75
35.50
25.00
16.75
16.25
16.00
11.25
7.50
COAL
($70 TON
RETAIL)
PRE-EMBARGO REGULATED
MIDDLE EAST NATURAL
FUEL OIL GAS
($2.00 BBL) ($2.00 MCF
RETAIL)
FUEL
OIL
($9.00 BBL)
22.00
23.00
FUEL
OIL
FROM
SHALE
($15.00 BBL)
SOLAR
HEATING
(SOLAR
INSTITUTE)
GAS
FROM
COAL
($16.00 BBL)
NUCLEAR
BASE
ELECTRICllY
($.03 KWH)
Figure 1. Rising costs of all energy sources pose a threat to the continued
economic welfare of this and succeeding generations.
The term "conservation" by itself leaves consider-
able room for interpretation. The following categories
encompass those actions which we at the Federal Energy
Administration have come to accept as "energy con-
servation":
1. Elimination of wasted energy: such as extinguishing
lights in unoccupied buildings at night or removing
excessive bulbs in existing commercial lighting fix-
tures.
2. Cost effective substitutions of human time or effort
for energy use: such as improving maintenance of
energy-consuming devices, using carpools or mass
transit instead of a private auto for commuting, or
consolidating shipments and personal travel.
3. Cost effective substitutions of capital investment for
energy use: such as increasing the use of insulation
in buildings, investing in heat pumps instead of
electrical resistance heating, or installing equipment
to capture the waste heat from boilers for a useful
purpose.
4. Substitutions of alternative lower energy activities
for higher energy activities which have a similar
effect: such as using telecommunications in place
of long-distance travel or dressing differently to
permit thermostats to be raised in the summer and
lowered somewhat in the winter.
This definition does not assume that energy con-
servation requires any dramatic changes in the basic
preferences of individuals. But it does suggest that some
changes in attitudes or lifestyles will occur as a greater
awareness of energy and its costs emerge. As may be
seen, each of the categories represents, at least in a broad
sense, a way to use energy more efficiently. Each
individual makes choices regarding energy-use conserva-
tion measures by weighing the importance of many
different factors, including energy costs, time value,
comfort, convenience, and force of habit. The cost of
energy, long disregarded by most consumers and busi-
nesses, has recently become a key factor guiding these
individual choices.
It is important to distinguish between the type of
efforts that clearly fall into one of the categories men-
tioned above and those steps that might be required in
the event of an emergency shortage of energy. A host of
mandatory curtailments, including gasoline rationing,
curbing exports, and Sunday closing of gasoline stations
could be instituted to cope with serious disruptions of
energy supplies. But such immediate emergency meas-
ures are viewed as distinctly different from the longer-
term measures, which I believe truly constitute the
5
-------
"C()lIservilli(JI1" ot IJlJeIUY. It is il1lportlllll lIot 10 lasH this
distinction, )t)cause the concerns, priorities, and the
governmenta roles shift dramatically.
Viewed from this perspective-that is, that conserva.
tion means improved energy efficiency and not simply
reduced con:iumption-energy conservation becomes an
equal partner with efforts to increase supplies. Quite
literally, a barrel of oil saved through the adoption of
cost-effectivE energy conservation measures is equal to a
barrel of oil aroduced through an expansion of domestic
production. =01' example, if a homeowner installs insula-
tion, he will save perhaps 6 to 10 barrels of oil during
the winter months. The oil saved represents a reduction
in total U.S. energy use, and as a result, our dependence
on imported petroleum is reduced.
This type of analysis can be used to determine the
"cost per barrel saved" of conservation actions. For
example, ceil ing insulation can be installed in the average
home for IE ss than $300. This amount of insulation
would save, on the average, 40 million 8tu's of fuel
annually, w~ ieh is equivalent to about seven barrels of
oil. AssuminlJ that the insulation has a life expectancy of
at least 20 years and that the cost of money is 10
percent, the equivalent of 60 barrels of oil would be
saved. By dividing $300 by 60 barrels, we cOllclud(! that
$5 worth of insulation is needed to save one barrel of
oil. This figure of $5 is less than virtually every other
form of energy available to the homeowner. For
example, at current prices, fuel oil costs over $16 per
barrel. Solar heating and electricity from nuclear power
plants would also be about four times more expensive
than saving fuel by installing insulation.
Similar calculations have been performed for storm
windows, which cost about $9 per barrel of oil saved,
and improved auto efficiency, which costs about $2.50
per barrel saved, or 6 cents per gallon. And just to show
we are honest and that not all conservation options are
cheap, we estimate that installing Washington's urban
rapid transit system to be the equivalent of buying oil at
$287 a barrel (fig. 2). Of course, this same approach can
be applied to numerous other conservation measures in
industry, transportation, utilities, and buildings. Without
going into these measures in detail, let it suffice to say
that most energy conservation measures turn out to be
far cheaper than the available supply alternatives. This
conclusion has relevance for individuals as well as for
policymakers.
As fuel becomes more expensive, substitutes for fuel
2.50
(HEATING)
COST PER BARREL ($)
50.00
( GASOLINE)
. 20.00
11. 2~
L~
CEILING
INSULATiON
. STORM
WINDOWS
SOLAR,
ELECTRICllY,
SYNTHETICS
AUTO
EFFICIENCY
SHALE
GASOLINE
(lOIt GAL)
WASH.
METRO
MASS
TRANSIT
Figure 2. Since unconstrained demand for energy is greater than
"business as usual" supplies:
.. The market will seek supply/demand equilibrium by choosing
the most cost-effective options from among both supply and
demand opportunities;
. One way to measure cost effectiveness is cost per barrel (sav'ed
or produced).
6
-------
become more cost-effective. It now pays to spend the
money to make cars more efficient, to insulate buildings,
to adjust thermostat settings at night, to schedule planes
so they are not half empty, to improve railroad freight
handling, and to modernize our factories to improve
their use of energy.
By defining energy conservation measures as cost-
effective substitutes for energy consumption, we can
begin'to assess the ultimate potential for energy savings.
It is clear that as a Nation we tolerate huge amounts of
energy waste. Sweden and Germany, two countries with
per capita incomes approximating our own, consume
only half as much energy per capita. Certainly part of
the difference can be accounted for by the larger homes
and greater use of transportation in the United States.
But the major reason is that Germany and Sweden have
responded to many years of high energy prices by
improving the efficiency of energy use. For example,
average new car fuel economy in Germany is 21 mpg
versus 14 mpg in the United States. Furthermore, the
manufac,turing of paper requires only 37 percent and
chemicals require only 57 percent of the energy used to
produce the same amount of these products in the U.S.
We should not necessarily attempt to duplicate
European energy-use patterns, but they are indicative of
the gross potential for energy savings. FEA estimates, for
example, that by applying existing technology we could
evenwal/y save more than 50 percent of the energy
consumed in the United States (fig. 3). To fully achieve
even these savings, however, will require the replacement
of much of our existing capital stock with more energy-
efficient equipment, vehicles, and buildings. Because we
cannot afford simply to scrap existing buildings and
industries, even though they may be grossly inefficient,
improvements in energy efficiency will have to coincide,
at least partially, with the gradual replacement of exist-
ing facilities and equipment. Nonetheless, substantial
savings can be achieved in the short term through the
modification of existing facilities and practices.
Improving the energy efficiency of both new and
existing buildings, vehicles, and equipment is now justi-
fied by cost savings alone, especially when compared
with the cost of increasing energy supplies. It is
abundantly clear that conservation is our cheapest
source of energy. Because energy conservation is in the
economic self-interest of individuals and businesses, we
expect that most of the potential savings will be
a chi e ved without further government intervention.
These "price-induced" savings are depicted by the mid-
dle line on figure 3, labeled "Business As Usual." But
today's higher energy prices will not achieve the full
potential for conservation alone. We believe there is the
100
1100% OF
75 1974
MilliONS CONSUMPTION
OF
50 " WITH MAXIMUM CONSERVATION
BARRELS
PER
DAY 25
75
80
85
90
2000
2010
YEAR
Figure 3. The potential: A composite demand picture shows that
cost effective conservation has the theoretical potential to reduce
demand by 16 percent in 1985 and 29 percent in 2010.
7
-------
additional potential, for example, to save about 5 mil-.
lion barrels per day of oil equivalent by 1980, 7.5 mil-
lion barrels per day by 1985, and more than 15 million'
barrels per day by the year 2010.
Most of this potential can be realized by energy
conservation measures, which fall into one of six general
categories:
In transoortation:
1. Improving the fuel economy of motor vehicles.
2. Reducing vehicles miles travel~d.
In resid,mtial and commercial buildings:
3. De! i!)ning and building more efficient new
stn'ctures and appliances.
4. Improving the energy efficiency of existing
buidings and homes.
In utilit/es:
6. Rec ucing the amount of scarce fuel required to
genl~rate each kWh of electricity.
If indivduals and businesses were totally committed
to conservin} energy and aware of the opportunities to
save, we wo Jld not have nearly as serious a problem as
we do today. But, for a variety of reasons, the response
to the calls for energy conservation has not been as
strong nor CiS enduring as we might have hoped. Cer-
tainly the dramatic increases in the price of fuel and the
increased awareness of the need for conservation have
resulted in s:>me significant savings. Energy demand so
far this year has been somewhat less than 1973 levels.
Considering that energy demand grew at about 4.3
percent annually from 1964 to 1973, no growth in
demand for nearly 2 years is certainly an important
accomplishmmt. Although the recession accounts for a
significant portion of this reduction, FEA believes that
conservation has reduced petroleum consumption on the
order of 5 percent from what it would have been other-
wise. This is tl,e equivalent of at least 1 million barrels of
oil per day. T he slackening of the rate of energy demand
growth is a significant achievement, but, as figure 3
indicates, them remain opportunities for substantially
greater saving::.
Because there is an overriding national interest in
reducing our dependence on imported petroleum, the
Federal Gover nment has actively sought ways to increase
the conservat on of energy. Unfortunately, the Federal
effort, up to now, has been constrained by limited
authority and .inadequate funding. Current Federal
activity, with the exception of the 55 mph speed limit,
falls into four general areas:
1. Information transfer. This includes the distri-
bution of pamphlets and other materials on
conserving energy, publ ic service advertising,
and several programs that involve direct
cont3cts with building managers, industry
groups, and individuals. I
2. Goal setting. This includes Federal programs
directed at improving automobile fuel economy
(40 percent by 19801. appliance efficiency (20
percent by 1980), and industrial energy effi-
ciency (10 to 15 percent by 19801. as well as
programs in the development stage, which will
involve utilities and individual States.
3. Research and development. This includes all
efforts directed at the development of more
energy-efficient technologies. With' the
establishment of the Energy Research and
Development Administration, Federal energy
conservation R&D efforts have expanded and
become more cohesive.
4. Import tariffs. Although the President has
continually urged Congress to adopt measures
that would replace the current $2 per barrel
tariff and has pledged to remove the tariff if the
price of "old oil" were completely decon-
trolled, it remains the only measure now in
effect that encourages conservation by artifi-
cially raising the price of energy.
F EA estimates that the combination of these exist-
ing programs (excepting the oil import tariff) will result
if) savings of approximately 2.5 million barrels per day
by 1980 and 3 million barrels per day by 1985. This
means that, without further Federal action, only half of
the economically justified potential for energy savings
will be achieved during these years and even less in the
future.
If, as I contend, energy conservation is in the eco-
nomic self interest of individuals and businesses, why is
only a portion of the overall savings potential being
achieved? Although we still do not have all the answers,
we know that many factors exist which prevent the
marketplace from providing the appropriate signals or
which prevent individuals from taking totally rationa.
actions with regard to energy use.
First, Federal regulations have kept oil and gas
prices below world market prices. This factor has
continued to hamper our response to the radical changes
brought on by the 1973 oil embargo and the dramatic
increases in the price of oil on the world market.
Secondly, energy-efficient homes, buildings, and prod-
ucts are generally more expensive than less efficient
homes, buildings, or products. This is because more
efficient products require more careful engineering and
design, material inputs with better (more expensive)
thermal characteristics, or other special energy-conserv-
ing features. Because of a lack of information and long-
standing consumer habits, purchases are made on the
basis of initial cost rather than lifecycle costs-which
8
-------
include the cost of the energy needs of the product. This
market factor encourages the design, construction, and
sale of relatively inefficient products.
Another genera"1 category of market imperfection is
the lack of information available to individuals, firms,
and units of government regarding opportunities for
conservation. This ranges from a lack of information on
the savings achievable through adjustments in thermo-
stats to an inadequate technical understanding of heat-
recovery systems.
A final but important factor that prevents a quick
response to rising energy prices is that a large portion of
the American economy no longer responds rapidly or
completely to traditional market forces. The largest
1,000 firms in the United States produce approximately
half of the total goods and services, with the other 12
million firms (and farms) producing the remaining half.
These 1,000 firms have relatively greater control over
the prices they charge than do the rest of the firms in
the economy, and they, therefore, have greater ability to
pass on the effect of increasing energy costs. The rela-
tively large market shares and advertising budgets of
these 1,000 firms help them maintain demand for their
products while simultaneously increasing the prices of
their products in response to the higher energy costs.
This factor reduces the incentive of these large firms to
respond to the market signals of rising energy prices
through energy conservation. In a similar manner, the
Nation's utilities can also pass on their cost increases to
the consumer, making this category of firms slow to
respond to new market conditions.
Federal energy conservation efforts are justified to
the extent that they alleviate market imperfections in
order to achieve a more cost-effective allocation of
scarce national and world energy resources. Justifiable
Federal energy conservation efforts are not attempts to
replace the market mechanism, but instead to reinforce
the market and help it operate in a more nearly perfect
manner.
What more can the Federal Government do to
achieve conservation? I believe that much can be done,
both through the enactment of pending legislation and
through the expansion of existing voluntary programs.
In January 1975, the President proposed a wide-
ranging program both to increase domestic supplies of
energy and to reduce consumption. The program
included excise taxes on both oil and natural gas, price
decontrol coupled with an excess profits tax, and provi-
sions for the return of these increased Federal revenues
to individuals and businesses, as well as to State and
local governments. It was recognized that increasing
energy prices posed some threat to the economy by
adding to the general rate of inflation, and the rebate
provisions were proposed to help counteract these
impacts. Our estimates of the resulting energy savings
were based on relatively conservative assessments of the
demand elasticity for petroleum. We assumed that every
10-percent increase in price would result in about a
lo9-percent decrease in demand within 2 years. Never-
theless, in view of the available alternatives, the price
mechanism seemed to be the most effective and least
disruptive means of achieving savings in the short run.
Unfortunately, the issues surrounding the decontrol
of domestic oil prices, gasoline taxes, crude oil excise
taxes, import quotas, and the import tariffs have
dominated the media and obscured the host of other
conservation proposals and programs initiated by both
the President and the Congress.
For example, the President has proposed legislation
to establish mandatory energy conservation standards
for new residential and commercial buildings.
(N ote: This mandatory conservation proposal was
recently made voluntary by the House.) He also has
proposed tax credits for individuals who install insula-
tion or make other energy-conserving modifications to
their homes; mandatory appliance and motor vehicle
labeling; and. legislation to authorize and fund a program
to winterize the homes of low-income persons. As
mentioned earlier, the Administration has established
voluntary programs to improve automobile fuel
economy and appliance efficiency. In these two in-
stances, the President has pledged to seek mandatory
legislation if progress toward the established goals is
insufficient. FEA, in cooperation with other Federal
agencies, has also established programs to encourage
greater energy conservation in industry, buildings, and
electric utilities, and has just recently instituted a pro-
gram to improve truck and bus fuel economy.
Although Congress has yet to enact any energy
conservation legislation, it has considered many pro-
posals to mandate some of the existing voluntary
conservation programs. For example, both the House
and the Senate have passed legislation mandating in-
creased automobile fuel economy. The House has also
passed a bill that would authorize the establishment of
mandatory appliance efficiency., standards. On the other
hand, the House rejected a proposal to require major
industries to submit energy conservation plans and
report energy use data to the Federal Energy Adminis-
tration.
It now appears that the Congress will eventually
enact legislation mandating automobile fuel economy
improvements and energy efficiency labeling for appli-
ances and autos; directing the development (and perhaps
mandating the adoption) of minimum performance
standards for new buildings and guidelines for appliance
9
-------
efficiency; and establishing tax incentives for the pur-
chase of irsulation, storm windows and doors, solar-
heating ami -cooling equipment, and perhaps a wide
range of other "energy-conserving" products. Unfor-
tunately, olily the tax credits for insulation and storm
windows alld doors are likely to have a significant
impact on Imergy demand in the near future, and even
the savings resulting from such incentives will only
scratch the !;urface of the overall potential.
Thus, I'Ve are left with only two alternatives to
stimulate immediate conservation: (1) increasing energy
prices thro Jgh rapid decontrol or exci~e taxes on
domestic 01 and natural gas; and/or (2) expanded
Federal effclrts to inform and motivate individuals and
businesses.
It now ;eems unlikely that Congress will enact legis-
lation that would permit a substantial rise in energy
prices withi1 the next 2 years. As a result, FEA has
begun to e:
-------
D MAXIMUM
CONSERVATION
m NEW
EFFORTS
~ EXISTING
EFFORTS
~ /.
equipment and processes and hopes to eventually
expand the program to include 90 percent of all indus-
trial plants. Such an expansion might result in savings as
high as 450,000 barrels per day.
Electric utilities account for nearly 25 percent of
total U.S. energy use, and yet opportunities for improv-
ing load factors and reducing utility consumption of
scarce fuels have only recently received serious atten-
tion. FEA is now sponsoring six load-management
demonstration projects, which include such conservation
measures as peak load pricing, ripple load controls,
power pooling, and others. We are also intervening in
State regulatory hearings concerning public utilities to
voice Federal energy policy and to advocate the adop-
tion of peak load pricing. This is another area where we
hope to expand our efforts in the future. An expanded
program might result in savings of 1.3 million barrels per
day of oil in 1985.
Recently, FEA and the National Governors'
Conference launched a State/Federal Energy Conserva-
tion Program. This program will establish a framework
within which individual States can collect data on energy
supply and use, project their own energy needs, and
establish goals for energy conservation. FEA will provide
15
MILLIONS
OF
BARRELS 10
PER
OAY
5
1977
1980
information and assistance to those States interested in
establishing specific conservation programs. FEA is also
investigating the possibility of providing direct funding
to the States to help support conservation programs.
Finally, we are convinced that we must greatly
expand our public education efforts to help increase
public understanding of the Nation's energy dilemma
and the opportunities for energy conservation available
to each individual.
With expanded Federal programs in each of these
areas, and with the enactment of the conservation
legislation now pending before the Congress, I am
confident that we can achieve the bulk of the potential
for energy savings in 1985. Figure 4 depicts the savings
we anticipate from these expanded energy conservation
programs. Nevertheless, FEA continues to explore other
ways in which the Federal Government might help
encourage greater savings. They include possible changes
in Federal regulatory policies, mandating individual
metering in multifamily house, establishing energy effi-
ciency standards for selected industrial equipment, and
providing financial and technical assistance to State
regulatory commissions and utilities in their efforts to
revamp rate structures and encourage conservation.
1985
1990
1995
2000
Figure 4. If we do all this, we can substantially improve the short-term
results and create the consensus needed for long term results: 675,000
barrels per day by 1977, all at a total cost far less than supply options
and with the American public behind the problem.
11
-------
Looking beyond 1985, the opportunities for energy
conservatio., increase. Perhaps new technologies, unfore-
seen today, will enable further increases in energy effi-
ciency. Certainly, much of our current R&D program
will only then begin to have an impact on energy use.
Eventually we will have to direct more attention to the,
ways in which the Federal Government could accelerate
the adoption of new energy-conserving technologies and
also to the broader issues of land use, community
development, and alternative transpo~tation systems.
12
-------
THE ERDA CONSERVATION PROGRAM: THE TECHNOLOGY OF
, IMPROVED EFFICIENCY
James S. Kane, Ph.D. *
Abstract
This country's industry, its buildings, and its trans-.
portation systems have been constructed at a time when
energy was extremely cheap. As a result, our facilities,
our institutions, and our lifestyle have tended to favor
the use of energy over capital and labor.
The uncertain availability and the not-too-distant
exhaustion of low-priced energy sources require that we
now examine all the ways in which our society uses
energy, with the object of improving efficiencies through
technological improvements. Since the use of energy
pervades every aspect of our society, the task is enor-
mous.
The total job to be done, however, is far more than
just technical development. Conservation only occurs
when and if the necessary measures are adopted by
literally millions of decisionmakers. It is therefore su-
premely important that the entire process, from tech-
nical concepts through commercialization, be free of
impediments, whether they be technical, institutional,
legal, or economic.
The E R DA program will examine efficiency
improvements in buildings-both commercial and resi-
dential, in industry, in the transportation sector, and in
our electrical energy systems. In all the work, the goal is
to accelerate the introduction of energy-saving tech-
nology and practices.
More efficient use of energy should be viewed as an
alternative to increased energy production. If the econ-
omic considerations of production versus conservation
are equal, the choice should favor conservation because
of its more benign environmental and sociological
impacts.
When E R DA was formed last January, it marked the
first time that a government agency had been given the
responsibility for a comprehensive R&D program on the
technical aspects of energy conservation. I n the time
since then, we've learned a lot about this new business-
including an appreciation of how much there is still to
learn. Many of the subsequent speakers at this meeting
are conservation veterans, meaning they have been at it
for more than a year, and I am sure much of what I will
* Deputy Assistant Administrator for Conservation, U.S.
Energy Research and Development Administration, Washington,
D.C.
say will not be news to them. But, they will have to
tolerate my talk today, because~ as is true for so many
converts, I cannot resist preaching the faith.
Let me start by distinguishing between two aspects
of conservation that are often confused: curtailment and
efficiency improvement. The former always implies a
reduction in the benefits, or at least the perceived bene-
fits to the consumer. Curtailment can be either volun-
tary, e.g., reduction in thermostat settings or mileage
driven, or involuntary, e.g., reduced voltage, gas ration-
ing, and speed limits. Curtailment procedures can be
invoked almost instantly, which is good, but their ef-
fects, both wanted and unwanted, also come instantly,
which can be bad. I am not going to spend any more
time on curtailment, other than to say it will undoubt-
edly be necessary in the coming years, and that we
should be extremely selective in the topics we choose.
Incidentally, my favorite example of ideal curtailment is
to turn out most of the billboard I ights in Las Vegas!
The kind of conservation that ERDA is concerned
with is increased efficiency, usually through technical
improvements. It has been aptly observed that our
country, as it exists today, is a monument to cheap
energy. It has nearly always been economically
advantageous to waste energy and conserve capital or
labor. Now we are faced with reversing this situation,
with all the difficulties that usually go along with major
cultural changes. The transition will involve much more
than technical developments.
How, then, do we go about identifying, developing,
and bringing into being the technology of improved effi-
ciency? In other words, how do we actually accomplish
an increase in the efficiency with which the country uses
energy?
Perhaps at this time I should observe that the "we"
I have just used so grandly does not mean ERDA, far
from it, the job will involve us all. This country has only
a short time to mend its ways, and the job that must be
done will deeply involve all aspects of our society. All
we in E RDA can hope to do is assist in the process. Even
this helping job, however, is an enormous undertaking.
What, then, do we foresee that we, as a Nation,
must do? I certainly cannot claim to know all the
answers to this question but, unti I we get smarter, I
suggest the following six-step process as a testing pro-
cedure:
1. First, we must identify the ways and processes
by which the country wastes energy. Much of this work
13
-------
has already been done, but it is amazing how we con-
tinue to fine new wasteful habits.
2. Sec:>nd, we must ask the question: Is the waste
due to foolishness, indulgence, or poor technology?
3. Thi'd, what would be the impediments to the
introduction of improved technology? If capability cur-
rently exist!, why is it not being used? Why have the
improvemen ts not come about from market forces?
4. What would be the impact of widespread in-
troduction of the energy-saving technology? What are
the time sca es and the size of the benefits? What are the
direct impacts? What are the impacts that cannot be
expressed i:I economic terms-sociological, environ-
mental, etc.7
5. Fifth, how can the barriers that prevent or
impede usin] the new technology be overcome? How
can we accelerate the process? What changes-legal,
regulatory, illstitutional-should be made?
6. Fin;llIy, with the technology available, and the
barriers and the means of overcoming them identified,
how can we most rapidly move the technology into com-
mercial and public use?
As YOl can see, the business of conservation
through improved technology involves far more than just
the technoll)~IY itself. The problem can only be dealt
with holistically; if one link in the chain is ignored, the
end results VIi II not be achieved.
Now th.1t I have outlined the approach, let me give
you a very brief description of the program that we are
now develoJing in ERDA. Later in this symposium,
other ERDA speakers will describe their individual
research are. s in greater detail.
Buildings
The bui ding sector in many ways is the most diffi-
cult of all, ::01' it involves by far the largest number of
decisionmak ~rs. I n most cases, the technology is rather
straightforw,lrd, but the nontechnological obstacles are
many and d.fficult. Ultimately, each individual who has
a choice invJlving conservation must clearly perceive a
benefit in a:!opting conservation measures and, having
decided to do so, be able to carry them through to suc-
cessful concl Jsion.
. Although I have emphasized the difficulties, the
benefits of widespread conservation in buildings are also
very great. -rhe existing systems are grossly inefficient;
very large s
-------
Although personal automobiles are by far the largest
factor in transportation energy consumption, other
forms of transportation must also be considered, with an
eye to improved economy and multifuel capability.
Simply put, we must prepare for a future in which petro-
leum fuels are increasingly scarce and costly, and
replacements, such as coal-derived fuels or shale prod-
ucts, are apt to be expensive even in terms of today's
cartel prices.
The two options we foresee for highway vehicles are
new, highly efficient engines capable of multifuel use,
and vehicles using stored energy originally derived from
coal or nuclear. Our transportation program is designed
to aid in the development of this technology.
Transportation includes modes other than highvyay
vehicles. Some areas such as pipelines do not have the
capability for improvement that highway systems do,
but there are significant energy saving developments
possible.
Electrical Energy
The other major area within conservation is elec-
trical energy systems-chiefly the transmission, distribu-
tion, and storage of electricity. Most predictions of the
future involve an increased trend toward electrification,
as we utilize currently available sources, such as coal and
nuclear, and learn to use new sources such as solar and
fusion energy.
O~r current system for transmitting and distributing
electricity is about '90 percent efficient, but further
improvements are possible. Higher voltage for both a.c.
and d.c. is needed. Very large shipments of energy, as
from power parks, would be more feasible if new devel-
opments in superconducting lines were possible.
Another aspect in which our current system is defi-
cient is in its use factor. .The "grid" has an average load
that is only about 55 percent of its peak capacity; a
great waste of installed capacity in a capital-intensive
system. Better management of the system, including
load leveling through technological (storage) and non-
technological (time-of-day pricing) would give additional'
capacity with a minimum of additional investment.
Until now, I have tried to give you a list of technical
tasks our Nation must accomplish if we are to use energy
more sensibly. For my last few minutes, I want to make
a few comments on the merits of conservation that
explain why conservation and environmental protection
share common goals, as well as the sponsorship of this
symposium.
I have often heard it said that "conservation is only
a means of buying time," or "conservation will tide us
over until we develop new, cheap forms of energy." I
think it is high time we conservationists rebelled against
being such stepchildren. I want to spend my final few
minutes tell ing you why I believe conservation is not just
a good way to go-it is the best way to go.
In my opinion, it is highly unlikely that we will ever
again see such cheap energy as our domestic natural gas,
or Mideast oil. Coal is bound to be expensive, once it is
made to pay its way in terms of occupational safety and
environmental responsibility. Its production as well as its
use, whether via direct combustion to produce elec-
tricity or conversion to synthetic fuels, is highly capital
intensive. Nuclear sources, fission and fusion, are poten-
tially very hazardous, and suffer from being even more
capital intensive than coal, especially when total fuel-
cycle costs are included. Solar energy, while perhaps the
least damaging to the environment, is almost sure to
remain extremely, perhaps prohibitively, expensive,
except as a source of low-grade heat. So the promise of
future supplies of cheap and abundant energy seems
dubious, to say the least.
I therefore think it highly likely that conservation is
here to stay. As a society, we are only now starting to
. question the dogma of exponential energy growth. Yet
to anyone who has given the question any thought, it is
obvious that someday energy production must cease to
grow. In my opinion, the most complex question facing
the next generation is how to get off this curve of
exponential demand for energy without catastrophe.
There are advocates who claim that GNP, well being, and
energy use are all locked into an immutable relationship.
I would urge them to examine more closely the coeffici-
ents in this equation, since they are the constants that
measure the efficiency with which we transform energy
into per capita national product, or other measures of
well being. It has been pointed out that these constants
are not the same for all countries, and I also believe we
can change them with an intelligent program in conserva-
tion.
Further, I think we in conservation must present
our case more aggressively. To paraphrase Ben Franklin,
a barrel saved may be worth more than a barrel pro-
duced. I n many instances it is far more cost beneficial to
invoke conservation than to produce additional fuel.
And I am sure it is not necessary for me to point out to
an environmentally concerned group such as this that
conservation technology is gentler to the environment
than energy production, which is highly capital intensive
and very low in labor intensity. Almost any other sector
will create more jobs per dollar invested.
I therefore maintain that conservation through
improved efficiency is a subject that must have a great
deal more attention that it has gotten until now. I am
convinced that it will hold its own economically with
increased production and, in fact, will often be more
15
-------
cost effective. Many conservation measures can be
achieved in "elatively s,mall increments, and will thus not
require largn amounts of front.end capital. While not
extremely rapid, the results will be felt sooner than
those of production technology. The environmental can-
sequences of increased efficiency are benign, and the
social effects are favorable. Changes in lifestyles, if any,
will be minimal and gradual. What other area of endeav-
or can make these statements?
16
-------
ENVIRONMENTJ~L IMPLICATIONS OF ENERGY CONSERVATION
Stephen J. Gage, Ph.D.*
Abstract
Implications of energy conservation for protection
of the environment are presented. The interrelationships
among natural resources, energy supplY' and use, environ-
mental quality, and economics are addressed, with
special emphasis on the practical implications of the
Second Law of Thermodynamics. It is concluded that
reduced energy consump tion, with thE' associated reduc-
tion in the flow of other matter, generally leads to
reduced environmental damage. The limitations of classi-
cal economics in describing the impacts of resource
shortages, as well as of nonquantifiable environmental
damages, is discussed. The activities of the Environ.
mental Protection Agency as part of the Federal research
efforts on energy conservation are discussed.
Denver is a wonderful place for a conference on the
en vi ronment and energy conservation. The snowy
Rockies, looming beyond the city, are a national symbol
of an environment free of the fumes and haze we associ.
ate with urbanized, industrialized society. Yet, even
though a view of the mountains is all too often obscured
by a curtain of pollution, Denver ha5, become a seat of
the fight against unquestioned and questionable growth.
Widespread concern in Colorado ov,er the potentially
explosive growth of energy resource exploitation in the
western part of the State has raised a flag of caution to
the rest of the country.
Perhaps Colorado is a forerunner of an America in
which the true value of a breath of fresh air is a known
quantity-a life-sustaining and enriching benefit which
has been weighed against the costs of controlling eco-
nomic growth and protecting the environment. Here, in
Colorado, people have come to act 011 the idea that the
quality of life is founded on the daily experience of a
healthful and satisfying environment, that the physical
amenities of the technological society can be enjoyed
only if the relationships of man to man and man to
nature are not debased by carelessness and waste.
I n this context, conservation of energy resources is
not an option to be considered; it is a necessary way of
life. In a period of rising prices and declining resources,
increasing the efficiency in the way we do things is basic
.Deputy Assistant Administrator for EnerHY, Minerals, and
Industry. Environmental Protection Agency, Washington, D.C.
to both maintaining material sufficiency and preserving
the quality of life.
When the Arab oil embargo hit us totally unpre-
pared in late 1973, energy conservation boiled down to a
straightforward proposition for Americans. Domestic oil
shortages created international dependency, and national
economic and policy independence was threatened.
After the initial jolt of widespread fuel shortages, long
lines at gasoline stations, and lower thermostat settings
came the second jolt: rocketing international oil prices.
More recently has come the third jolt: diminishing oil
and gas production in the lower 48 States. The U.S.
Geological Survey has, in the past year, reduced dramati-
cally its estimate of domestic oil and gas resources. And
although our known coal resources are vast, production
of coal is growing much too slowly to offset the decline
in oil and gas production. New energy technologies are
not now, nor will be soon, available. Synthetic fuels
from coal or shale, breeder reactors, solar electric plants,
and fusion electric plants are years away. Thus it is
imperative that ex isting sources must be stretched and
extended and that existing technologies must be im-
proved and made more efficient. Energy conservation is
essential.
It is, however, as the Ford Foundation Energy
Policy Project has pointed out:
. . . a mistake to regard energy conservation
as an end in itself; that puts the cart before
the horse. Conservation is worthwhile as a
means to alleviate shortages, preserve the en-
vironment, stretch out the supply of finite
resources and project the independence of
U.S. foreign policy.
I n an industrialized society, extraction, processing, and
use of energy and other nonrenewable resources is close.
Iy linked with environmental degradation. Increasing the
efficiency of energy and material use in our society is a
relatively quick, relatively inexpensive, and often an
environmentally favorable alternative to the extraction,
processing, and use of more energy and materials.
Whether the stimulus for conservation is jawboning,
higher fuel prices, or the availability of more efficient
techniques and technologies, the fact remains that if the
flow of energy and materials through the economic cycle
is reduced through more efficient use and increased
recycling, then the environmental degradation associated
with that cycle will generally be reduced. Obvious, too,
is the fact that such a reduction means that there will be
more reserves of raw materials available for future use.
17
-------
The National Materials Policy Commission has
pointed out that:
Deple ~ion of reserves and pollution have the
same cause-failure to manage the flow
materals as a cycle, resulting in a resource
depleting dispersal of energy and materials
into the environment as pollutants. A
Natiollal policy for the management of ener-
gy ami materials is needed to transform this
open-e nded process of wastage into a sub-
stantially closed system. The goals of envi-
ronmental quality require a comprehensive
systerr of resource recovery: the recycling of
materi )Is through the economy with min i-
mallo~;ses in matter and energy.
We ma\" in fact, be entering a new era in under-
standing the relationships among certain types of human
endeavor ard the natural world. Man, it seems, has
generally progressed by uneven steps. As he has success-
ively grasped new perceptions of how nature functions
and how he can use that knowledge to bend nature to
his will, miln has moved stepwise to more advanced
stages of technological achievement. That evolutionary
pattern has heen interpreted as being supportive of wide-
ly varied religio.us and ideological beliefs in the Western
World and, in turn, has been enhanced by purposeful
attempts to )pply those beliefs to increase man's control
over natural processes. Of course, such continual rein-
forcement-~ nown technically and now popularly as
"positive feE dback" -has encompassed man's approach
to social systems as well, resulting unfortunately in some
aberrations in which man has treated his fellow man
most inhumanely and nature most unnaturally.
As we step back to view currently emerging percep-
tions, we are struck by the growing acceptance of certain
thoughts tha t would have been rank heresies but a few
years ago.
We are f.lced with the long-term prospect of paying
much more for energy and probably for many other
materials: i.e., energy is no longer cheap.
We are, .n fact, running out of certain energy and
other material resources. Domestic oil and gas resources
will be seriously depleted before the end of the century.
At the h:gh rates of consumption that are occurring
or are projected to occur, we can foresee serious deple-
tion of even our more abundant critical nonrenewable
resources, although the time horizon may be several
centuries. In other words, we live on a finite planet with
finite resourCE s.
We are I::eginning to have serious second thoughts
about the basic economic premise that the marketplace
wi II, over the long term, adjust for any shortages. Higher
prices for a rr ineral commodity are supposed to lead to
increased production of the mineral, since lower quality~
deposits, which had previously been unworked because
they were uneconomical or technically infeasible, are
exploited because of more favorable economics and/or
technological improvements. However, if the resources
are not in the ground, higher prices will not have any
effect.
We are also beginning to question another basic
assumption generally made by economists: economic
processes are reversible, and it only takes money to re-
verse the flow of economic processes. There is the gen-
eral underlying belief that, if the economics were right,
then we could take most of the products of our society
once they have been used and recycle them, neglecting,
of course, the minor losses due to less than perfect proc-
esses. It is not economic to do much of that now, the
conventional wisdom has it, because virgin resources are
still pretty inexpensive and, since society is fairly dis-
persed, gathering up the byproducts is relatively expen-
sive. But when it becomes economical, the economists
tell us, the marketplace will require such recycling and,
furthermore, the dollars we put into recycling, just like
the dollars are now put into disposing of our refuse, will
all contribute to our GNP.
There is a serious flaw in such conventional econ-
omic thinking, however, which I believe is beginning to
dawn on serious thinkers. In an imaginary world where
there are infinite resources and where technological
processes are perfect machines, the economic principle
of reversibility would hold. But in the real world, where
dollars are only surrogates for tons of steel, barrels of
oil, hours of labor, etc., the flow of economic processes
must obey physical laws as well. When the real world
more closely resembles the imaginary world of econom-
ics-such as America from the 1870's to the 1920's-
then, to a fair approximation, pure economics apply.
Resource depletion can be ignored. Industrial wastes cal
be disposed of in a vast country without adversely af.
fecting, in turn, the economic processes. When energy
and land are cheap and there is always more farther west
and when the waste products of inefficient use can be
"lost," no one worries about reversing the processes.
Let me turn briefly to an analogy in physics. When
man began to understand energy and matter at the
macroscopic level, he could apply the powerful observa-
tions that matter was conserved and that energy was
conserved. A substantial portion of the structure of
classical physics and most of the industrial machines that
we operate today were developed on the basis of this
insight. Einstein's assertion that energy/matter was con-
served and that energy could change into matter and vice
versa provided a new insight and presaged the nuclear
era. As we come to appreciate the implications of under-
18
-------
standing the microscopic structure of our world, we are
now building the fission and fusion energy machines of
the future.
The flaw which I believe exists in conventional
economic thought is that a critical physical law has not
been considered in the formulation o'f economic prin-
ci pies. The Second Law of Therrnodynamics-that
entropy is always increasing, that ener9Y is always mov-
ing from a state of organization to a state of chaos-has
been totally ignored by most economists. As energy
moves from a state of low entropy (for instance, the
chemical energy bound in a molecule of methane gas) to
a state of high entropy (for instance the carbon dioxide,
water vapor, and heat created by the combustion of the
methane), the flow is thermodynamically irreversible.
The methane molecule can be reassembled again, but at
a relatively large expenditure of additional energy: i.e.,
many more molecules of methane must be irreversibly
destroyed to recreate the original molecule. Further,
once the heat of combustion is expended in doing work
or is otherwise lost, the heat is dispersed to a point
where it becomes both thermodynamically and economi-
cally worthless; hence thermal pollution results. Finally,
if there are contaminants such as su Ifur in the methane
or if coal is used, there are many side reactions leading
generally to undesirable byproduct:; such as sulfur
dioxide, which are generally discarded as environmental
pollutants. We well know that it takes additional energy
even to capture these pollutants, let alone attempt to
convert them back to some more benign state.
The industrial ized nations have uniformly been
ignoring the ultimate practical implications of the
Second Law. They have been con!iuming increasing
amounts of nonrenewable energy and material as if
somehow they could get ahead of the Second Law. Now,
with higher energy prices and real fuel shortages, it does
not take a Malthusian doomsday forHcast or a Club of
Rome computer study to raise serious questions about
the way we have been doing things. If we continue to
expand rapidly our use of the finite nonrenewable re-
sources and to disperse the byproduc:ts of their u~e on
our all-too-finite planet, we will find the Second Law of
Thermodynamics overtaking the prindples of econom-
ics.
Energy conservation may well have a much more
important and general role to play in the future. I am
pleased to be here today to address this conference and
to add my voice to the support of conservation of all of
our nonrenewable resources, including energy.
While other Federal agencies have more central roles
to play in promoting energy conservation, the Environ-
mental Protection Agency has an important responsibil-
ity. We must aid and assist in the promotion of energy
conservation measures while providing gu idance to
insure environmental protection as conservation meas-
ures are adopted. To fulfill both of these objectives, we
work closely with other Federal agencies such as the
Federal Energy Administration, the Energy Research
and Development Administration, the Department of
Transportation, the General Services Administration,
etc.
Specifically, our research program on the environ-
mental aspects of energy conservation is comprised of
several parts which I would like to describe briefly for
you today.
First, let me turn to the industrial sector, which uses
the largest fraction of energy. Studies estimate that up
to one-third of the industrial energy consumption in the
United States can be offset through conservation meas-
ures. If it is, by the year 2000, industrial energy de-
mands may be as much as 20 quadrillion Btu's lower
than it might be under "business as usual" practices.
Between today and the year 2000, more than four times
as much energy could be "saved" in the industrial sector
than will be used in all of the United States in 1975.
EPA has underway a major study which is examin-
ing the environmental aspects of energy-conserving proc-
ess changes in 13 major industrial groups. The industries
include glass, phosphorous/phosphoric acid, chlorine/
alkali, olefins, petroleum refining, textiles, steel, copper,
cement, and aluminum/alumina. Leading process options
are being studied for implementability, timing of imple-
mentation, etc., and for the adequacy of existing envi-
ronmental control techniques to abate emissions and
residuals from these processes. Additionally, EPA is
developing environmental control techniques that are
energy conserving, such as cupola off-gas heat recovery
devices which reduce the cost of bag-houses in metal
industries.
Buildings are also large consumers of energy. It
appears that the efficiency of energy use in buildings can
also be improved by about one-third. National engineer-
ing associations have begun to act to encourage realiza-
tion of these savings by issuing energy performance
standards to buildings. The Federal Government and the
Congress are also moving in this direction.
Here, however, EPA has waved a caution flag.
Studies of the relationship of indoor and outdoor air
quality have raised the possibility that unhealthy con-
centrations of pollutants may buildup within tightly
closed buildings unless great care is taken in the design
of energy-conservative measures. It appears that the
standards proposed by the engineering societies, while a
good first attempt to save energy, did not adequately
19
-------
address the rnaintenance of indoor air quality.
We are working with other Federal agencies, includ-
ing ERDA and FEA, to study the potential problem of
indoor air quality. An EPA study about to commence
should provide results useful in the development of
Federal guidelines or standards for building design and
revision of eHisting standards.
Another EPA program is addressing the reuse of
low-grade We ste heat which is rejected in large quantities
by electricity and steam generation. I n fact, about two-
thirds of th,! Btu's consumed in the steam cycle are
rejected as v/aste heat. In numerous European installa-
tions, the reject heat is being used in industrial applica-
tions or dis1rict heating systems. EPA, in cooperation
with agencie~; such as the Tennessee Valley Authority, is
studying approaches to waste heat reuse, which may be
applicable in the United States. Approaches such as heat-
ing fish pond; and greenhouses are being studied.
A fourth and major EPA program is resource recov-
ery. The ef1 iciency of our energy economy can be
increased through the recovery of both Btu's and mate-
rials. By 1980, over 1 billion dry tons of solid waste will
be generated~ach year in the United States; this is equiv-
alent in heat content to over 1 billion barrels of oil per
year, or neatly 3.5 million barrels of oil per day. In
addition to t 1e Btu content, large tonnages of ferrous
and nonferrolJs metals and glass are present in most solid
waste streams.
The Environmental Protection Agency estimates
that, through 1985, energy derived from wastes could
exceed in magnitude energy produced from any syn-
thetic fuel sources or any new energy technology. By
1985, up to .5 million barrels of oil equivalent could be
generated per day from waste recovery.
EPA's research, development, and demonstration
programs for fuel and materials recovery from municipal
wastes are pushing ahead new technologies for realizing
this potential. In St. Louis, a joint EPA/utility project
has brought to near commercial availability the first
system to convert solid waste to electricity by cofiring
refuse and coal in an environmentally acceptable man-
ner. Other projects include other combustion tech-
niques, pyrolysis, and biodigestion.
The programs of the Environmental Protection
Agency relating to energy conservation are aimed at
insuring that the environmental benefits of increased
conservation are realized. The Agency is investigating
efficient environmental control technologies and strate-
gies; analyzing the environmental consequences of con-
servation technologies, methods, and approaches; and
supporting the introduction of environmentally accept-
able conservation opportunities in every sector of our
economy.
In closing, I would like to emphasize again the
important fact that conservation of energy and other
nonrenewable resources may be the most important
thrust of the current day-more important, in fact, in
charting a new course for all of society than in saving
Btu's or tons of ore.
20
-------
3 November 1975
Session I:
NATIONAL ENERGY
CONSERVATION STRATEGIES
Rene R. Bertrand, Ph.D. *
Session Chairman
*Manager, Engineering Research, Exxon Research and Engineering Company, Linden, New Jersey.
21
-------
22
-------
ENVIRONMENTAL IMPLICATIONS OF HIGH ENERGY USE
Mark D. Levine, Ph.D.*
Abstract
The meaning of the term "high energy growth" is
discussed. Analysis of energy projections made between
1972 and 1975 reveal that the OPEC oil embargo mark-
ed a significant change in attitudes about energy growth.
After the embargo, most analyses of future energy
growth rates were 15 to 20 percent lower than projec-
tions made before the embargo.
There follows a discussion of the relationship be-
tvveen energy use and the environment from two points
of view: that of the proponents of energy conservation
and that of the proponents of energy growth. Analysis
of the most significant adverse impacts of energy growth
on the environment lead the author to advance the
axiom:
To the extent that high energy growth requires
rapid commercialization and deployment of new
and advanced energy conversion and processing
technology, the probability and extent of signifi-
cant environmental, health, and safety problems
are greatly increased.
A number of examples are presented in support of this
statement
I am very pleased to be given the opportunity to
speak at this symposium, especially as I applaud its pur-
pose: the reassessment of the relationship betWeen
energy use, energy conservation, and the environment.
The linkage between energy growth and the environment
is a complex one, as will be apparent from my discussion
today, and it is important to reevaluate this linkage in
symposia such as this. Indeed, part of my message today
is the complexity of the relationship betWeen the envi-
ronment and energy. I hope also to indicate ways of
increasing our insight into this complexity, making sense
of the complex system. I am addressing these topics
from the perspective that I have gained, first as a staff
member of the Energy Policy Project and then as a re-
searcher on energy and environmental policy at Stanford
Research I nstitute. I wish to point out that the ideas
expressed in this presentation are my own and do not
necessarily reflect those of any of my employers, past or
present.
'Senior Operations Analyst, Operations Evaluation Depart-
ment, Stanford Research Institute, Menlo Park, California.
Let me summarize the main points of this presenta-
tion. First, I define and clarify the title, "Environmental
Implications of High Energy Use." There follows a dis-
cussion of what I have termed the conventional wisdom
concerning energy use and environmental impacts among
proponents of energy conservation. I then turn to a
review of the conventional wisdom. among energy
growth advocates. Finally, in good dialectical fashion, I
present my own views in an attempt at a synthesis of the
relationship betWeen high energy use and the environ-
ment.
The Meaning of the Term "High Energy Growth"
To discuss the implications of high energy use, we
should first have a clear definition of the meaning of
high energy use in the United States. Projections of
future energy consumption in the United States have
undergone significant change with respect to both their
underlying assumptions and their results in the past 2 or
3 years. In figures 1 through 6 are shown six well-known
energy forecasts and scenarios that have been published
between 1972 and the present. The first three of these
forecasts were made before the 1972-73 OPEC oil
embargo and the rapid increases in oil prices that fol-
lowed the embargo. The second three were completed
after the embargo. Perusal of these six figures, which are
summarized in table 1, and other projections made in
the time frames before and after the embargo leads one
to conclude that:
1. Before the embargo most base-case or high-
growth projections of U.S. energy demand in the year
2000 were on the order of 200 quadrillion Btu's per year
(about three times the energy demand in 1970).
2. After the embargo most high-growth energ',I
scenarios were on the order of 165 quadrillion Btu's per
year in 2000.
3. I n the period after the embargo, increased
attention to energy conservation strategies led to the
development of a number of plausible low growth
energy futures.
What has brought about this significant change in
energy projections? To some extent, the oil embargo
encouraged many to recognize the likelihood of future
energy supply problems. Futhermore, higher prices of
energy in the last 2 years have led many economists to
suggest that long-term demand elasticities will reduce
future energy demand by perhaps as much as 15 to 20
percent of the base case projections for the year 2000.
23
-------
In
~
...
g:j
z
o
40f0lJLLJlD. .":....... ". 'M.' - ""'"
aor. ....
,.- .;:
..
o COAL 011. GAS JlUCioUR
200
180
180
ltO
no
100
80
80
1995 2000
Figure 2. National Petroleum Council, energy forecast, December 1972.
140
:I
III:
~
::>
co
z
OUIII5
11m 2000
lao
Q
z
~
uJ
C)
>-
l!I
'"
I1J
~
100
///'
~
110
-
U')
=>
60
1810
1915
19110
19115
YEAR
1990
Figure 1. Associated Universities, Inc. energy forecast, April 1972.
ENEIIGY SUI'PI,Y. 19H~ ANII ~ooo;
IIIUII GIIOW1lI, CAliE 11
~20
2UO
1110
160
.'1U
I:lO
IIIU
110
6U
19!J5 :1000
Q
z
«
:c
w
Q
>-
CJ 100
'"
I1J
Z
W
GO',\.
GO
z 14U
IIIGIl
1~0
-
V)
=
IIU
t'lO
ltl70
JUKO
1~1;)
I~H:-)
YEAR
1990
24
-------
60' ENERGY SUPPLY, 1985 AND 2000 " 200
'"
~ 40J 180
~
a:a lu
z
0
::I 20J - 160
;;;
~ 140 140
:;)
CZI COAL OIL GAS HUC"';AH
z
I~O E3 1985 120
Q
z a ~ooo
~
UJ
Q 100 100
>-
CI
""
UJ
Z 80 - 80
uJ
./'
~
=> 60 60
1910 1915 1990 19t15 1990 1995 ~ooo
YEAR
Figure 3. Department of Interior (Dupree and West) energy projections,
December 1972.
200
'"
:;)
~
a:a
z
o
-J
-J
1It1"
IJIJIRGY SUPI'I.Y. 1985 AND 2000:
~IT:iJ
COAl. OIL OIL GAS JlUCLlAIl
(Synlh"llco)
160
""
Q
«
:;)
CZI 140
Z
140
Q
~ 120
UJ
Q
~ 100
0:
UJ
Z
uJ
D 1985
mm 2000
D IJlI'OR1'ED OIL
100
.
V'>
=>
- 80
1975
1980
1985
YEAR
1990
1995
60
2000
Figure 4. Energy Policy Project, energy scenarios, 1974.
25
-------
II)
~
....
I:Q
z
o
...J
...J
""
CI
c(
a: 140
z
-
~ I~O
~
...
~
>- 1011
'"
'"
...
:z
W 80
(/)
::::>
200
ENERGY SUPPLY. 1985 AJID 2000
ACCKLIRATiD SUPPLY ($11/bblo11)
401
~ _OL~uJl._DI
COAl. OIL OIL GAS NUCLEAR
180
160
DI1I85
- ~ooo
(Synlhellca)
- 100
80
110
11170
80
~ooo
11175
1980
19115
YEAR
1990
1995
Figure 5. Project Independence Energy Forecasts (extended). November 1974.
II)
~
....
""
z
o
...J
...J
""
CI
c(
~
C2I
z
CI
Z
~
...
Q
>-
C>
""
...
Z
.....
(/)
~
~oo
110
ENERGY SUPPLY. 198:-) AN)) :.moo
ACCELERATE" SYTIlET1CS (II)
::il~_~_nl
COAl. 011. 011. (;~S N\J(;I.t:~1I
(Synlhpl.lc',,)
I~II
tHo
I~O
D I!lH5
. 2000 . -----
D IMI'ORTED OIL ~\
~ ENERGY CONSERVA1'10N (I~
. 100
I~()
100
NO
80
199!;
fiO
20110
11/7~.
11/11.0
19H:i
lH!m
YEAR
Figure 6. ERDA-48 energy scenarios, June 1975.
26
-------
Table 1. High energy growth projections
Projections made before OPEC oil embargo and oil price increases
1.
Associated University, Inc., base case
177 quads
2.
National Petroleum Council, high growth
medium growth
215 quads
200 quads
3.
DuPree and West, Department of Interior
Projections made after oil embargo
1.
Energy Policy Project, high growth
187 quads
148 quads
2.
Project Independence, accelerated supply
165 quads
3.
ERDA-48, base case without conservation
It is important to differentiate between the impacts
of higher energy prices and those of a long-term energy
conservation strategy. Both higher energy prices and
explicit energy conservation strategies reduce the energy
growth rate. Higher prices tend to reduce energy demand
by causing some behavioral changes that induce people
to do without some accustomed activities. The effects of
higher prices are most likely to be felt among the poorer
economic groups, for whom energy costs are a substan-
tial fraction of their total budget. Energy conservation,
to the extent that it is accomplished in economically
efficient ways, should provide savings to all consumers as
well as to the environment. One of the important
research needs not currently addressed adequately is the
establishment of a data base and analysis that adequate-
ly distinguish the degree to which reductions in energy'
growth can occur as a result of economically efficient
energy conservation measures as distinct from higher
energy prices.
In summary, I assume that high energy growth
means approximately 165 quadrillion Btu's per year in
the year 2000. I assume that higher energy prices are
responsible for this consumption being lower than earlier
forecasts. This projection, plus or minus perhaps 10 per-
cent, is a reflection of an energy future without substan-
tial governmental or private sector initiatives to foster an
energy conservation program that significantly reduces
aggregate U.S. energy demand. Later in the discussion I
return to this figure in discussing the environmental
consequences of higher energy growth.
The Conventional Wisdom Among Proponents of Energy
Conservation Concerning the Relationship Between
Energy Use and the Environment.
The Council on Environmental Quality, in a report
entitled "Energy and the Environment: Electric Power,"
in August 1973, expressed succinctly the attitude of
most students of environmental impacts of energy devel-
opment towards energy growth:
Much of the environmental damage from our use
of energy comes from the systems that provide the
energy to the consumer. If the systems for provid-
ing energy were to function more efficiently, then
the adverse environmental effects of energy pro-
d u ction would be reduced. Similarly, if the
consumer were to use energy more efficiently, ...
then both energy production and environmental
damage could be reduced....
The idea here is simple. To the extent that specific
air pollutants and other effluents that degrade the envi-
ronment are released by energy production, conversion,
and end-use, reduction in energy demand will reduce
environmental degradation. Of course, the same state-
ment could be made about any other activity or con-
sumer product. The difference is that energy production
and use is responsible for a large percentage of total
27
-------
emissions int') the environment. As shown in figure 7,30
percent of the particulate emissions, 80 percent of the
502 emissiors, 95 percent of the NOx emissions, and 60
percent of tht~ hydrocarbon emissions from man made
sources are e.:timated to be derived from energy produc-
tion and con:iumption activities. Thus, it is evident that
reductions in energy growth rates cause substantial
reductions in the total emissions of air pollutants in the
United States.
Three generalizations are useful to summarize the
point of vievl that attributes a substantial portion of
increased em'ironmental pollution to the growth in
energy demands:
1. All energy extraction, conversion, and end-use
activities are accompanied by adverse environmental
impacts.
2. Thes,! energy-related environmental impacts
constitute a sllbstantial portion of all adverse effects on
the environment.
3. As a result, high energy growth is likely to lead
to an increasHd magnitude of adverse impacts on the
environment.
To these three generalizations, I would add two
corollaries:
1. Accel ~rated growth in energy consumption
allows less tim~ and fewer opportunities to safeguard the
environment, 35 new technologies are deployed before
100
a::
-I ....
«
-------
1. Energy use is a partial substitute for labor; as
such, it is used to reduce man's toil. Thus, increased
energy consumption is generally a net benefit to society.
2. New technology may be more energy intensive
than the technology that it replaces (e.g., the SST); how-
ever, the use of the new technology both fosters further
technological advances as well as providing direct
benefits in the form of improved services. To the extent
that the new, more energy-intensive technology is
accepted in the marketplace, consu mers have decided
that it is worth additional energy and dollar costs.
3. Environmental control technology consumes
significant quantities of energy, resulting in increased
energy consumption; thus, control technology reduces
environmental degradation whi Ie increasing energy con-
sumption.
4. Many of the energy resources to supply high
growth requirements are located in frontier areas of the
United States with few inhabitants. As such, the effect
of the additional energy development on people is likely
to be limited.
The first two points are not directly relevant to the
discussion of this paper. These beliefs, although not
bearing directly on the relationship between energy
development and the environment, nonetheless provide
the setting for the conflict between the growth and low-
or no-growth advocates regarding environmental issues. I
might mention, parenthetically, that the gap between
the two sides has narrowed somewhat in the past year or
two. It was not too long ago that the words "energy
conservation" evoked a response in some corners of
government and private industry of either disbelief or
great concern. Nonetheless, the point of view that tends
to deemphasize the environmental impacts of energy
activities should be understood as being motivated in
substantial measure by considerations that are more re-
lated to the economy than to the environment.
The third point, made by proponents of growth
regarding the effect on energy consumption of environ-
mental control technology, is in my judgment a red her-
ring. Recent studies-for example, an analysis performed
by University of Michigan staff and reported in "The
Energy Conservation Papers" of the Ford Foundation
Energy Policy Project-have shown that additional
energy consumption associated with control technol-
ogies constitutes a very small percentage of total energy
consumption. The exception to this can occur when
regulations are established that encourage industry to
employ inefficient means of environmental control to
meet standards. If intelligent and farsighted environ-
mental policy decisions are made, there is no reason to
believe that energy consumption will grow in response to
environmental needs.
I defer discussion of the fourth point mape by
growth proponents until later in the presentation, where
it is analyzed in some depth.
Analysis of the Relationship Between Energy Growth
and the Environment
At this point, we would do well to step back and
make an assessment of where our discussion has taken
us. The four points made by the growth advocates can, I
think, be summarized by the statement .that growth
proponents generally place great emphasis on economic
rather than environmental values. This is a complicated
issue that takes us afield of our topic, which is to better
understand the environmental implications of high
energy growth. To get back on track, I consider the
differences between high and low energy growth. For
this purpose, I have chosen to use the two ERDA scen-
arios shown in figure 6: scenario II, the base case with
coal synthetics and oil shale, and scenario I, which in-
cludes substantial energy conservation measures. I could
just as well have chosen a different set of scenarios to
depict high and low energy growth.
Analysis of these two scenarios indicates that many
of the environmental impacts of the high growth case
not occurring (or occurring in substantially lesser degree)
in the lower growth scenario are due to the rapid devel-
opment of a synthetic fuels industry. A large part of this
synthetic fuels industry will probably be located in the
Western States, where oil shale and much of the coast
resources are found. The ERDA scenario projects a
growth rate of 16 percent per year between 1985 and
2000 for coal synthetics. 0 il shale is also projected to
grow rapidly in this time frame. Although skeptics might
question the likelihood of ERDA's high synthetic fuel
growth rate, the exact rate of growth is not important to
my analysis. The key environmental impacts (definrd
broadly to include social impacts) of this rapidly devel-
oping synthetic fuel industry include:
1. Potential urbanization problems, such as those
already experienced at Gillette and Rock Springs,
Wyoming, caused by rapid population influxes into rural
areas not adequately prepared to deal with large num-
bers of incoming construction and industry workers and
their families.
2. Water problems, associated with both the allo-
cation of scarce water supplies-about which the Col-
oradans present are especially well informed-and po-
tential contamination of aquifers and surface waters.
3. Aesthetic problems, especially the degradation
of the presently clean air that most of the West enjoys
and the construction of transportation corridors to the
energy-rich regions.
29
-------
4. Solid waste disposal problems, esp(~dally of oil
shale residue, ,HId rev(~!I(!tation 01 sllrlac:(~ l1Iirwd land.
5. Poter Itial rclt:ases to the environment of toxic
or carcinoger ic substances not now associated with
large-scale indlJstrial or energy activities.
I cite these as critical environmental and social prob-
lems likely to occur as a resu It of a rapidly expanding
synthetic fuel; industry. A key issue, to which I now
turn my atten tion, is that of the conditions under which
these environrnental problems are either more or less
likely to be manageable. I have suggested that one key
environmental difference between the high and low
energy growth scenarios is manifest in the rapidly grow-
ing synthetic fuels industry of the higher energy fu-
ture-there are other important areas of environmental
differences as well. It is my judgment that the diffi-
culties of dealing with many environmental problems are
considerably I!xacerbated under conditions of high
growth. In fact, 1 suggest as an axiom the following
statement:
To the ex tent that high energy growth requires the
rapid commercialization and deployment of new
and advar Iced energy conversion and processing
technologl" the probability and extent of signifi-
cant envilonmental, health, and safety problems
are greatly increased.
To illustrc te this axiom, 1 use as an example an
accelerated program of developing nuclear power in ad.
vance of publi: acceptance and of assurance that long-
term environmental problems have been adequately
dealt with. Under conditions of rapid growth, construc-
tion firms of limited experience will undoubtedly
achieve substantial participation in the construction of
nuclear power Jlants. Such a situation is highly undesira-
ble from the Joint of view of reactor reliability and
safety. Further more, considerable numbers of nuclear
power plants rlay become operational before standards
of safety have been adequately defined. The implemen-
tation of new !afety standards is I ikely to be more diffi-
cult in a climate of rapid nuclear growth, because an
industry beset with constantly changing standards is less
able to meet tl1e desired rapid development schedules.
Needed solutiol15 to the long-term high-level radioactive
waste problem I:ould come after substantial quantities of
waste are already difficult if not impossible to retrieve.
Commitments ".:0 research and development as well as
implementation of suboptimal reactor systems-illus-
trated perhaps bV the liquid metal fast breeder reactor-
might become substantial before the necessary knowl-
edge has been gained and analysis performed to establish
which energy !ystems are most nearly consistent with
our energy and mvironmental objectives.
Lest this e
-------
this time; it is, however, essential that we use the time
productively to reduce the adverse impacts, as a failure
to do so will otherwise serve only to delay the problems
rather than to solve them. Thus, my general conclusion
is that there is an important relationship between a suc-
cessful erwr~IY conservation stratc~JY and tlw protection
of environmental values, if we are able to inject a strong
environmental component into our energy research,
development, and planning activities.
31
-------
IMPLICATIONS OF A NATIONAL CONSERVATION STRATEGY
Kenneth R. Woodcock *
Abstract
Under thl' National Environmental Policy Act, the
Federal Energ:' Administration has the responsibility to
assess the 9n vironmental implications of energy
conservatior' policies, programs, and legislative
proposals. Th..oqgh the use of two analytic systems,
PIES and MEIlES, the environmental impact analysis is
based directly upon the projected energy impacts of pro-
posed prograrr's or policies. Furthermore, the economic
implications of energy strategies are assessed for the
preparation ot inflation impact statements. This paper
describes this methodology and illustrates its use in the
assessment of fllternative energy conservation programs
and proposals. Three types of conservation strategies are
discussed: energy-pricing strategies, sector-specific
strategies, and ,~ombined strategies. Analytical results are
presented and compared from legislative environmental
impact statemlmts prepared for the Energy Independ-
ence Act of 1975 and the Mandatory Oil Import Pro-
gram.
INTRODUCTION
. Under the National Environmental Policy Act, the
Federal Energy Administration has the responsibility to
assess the environmental implications of energy policies,
programs, and !egislative proposals. Accordingly, the
agency has examined individual energy resource develop-
ment and con.:€rvation programs, as well as energy-
pricing policie~ that provide for both reductions in
energy demand and incentives for additional domestic
energy resource development.
The major purpose of this paper is to describe the
analytical proc~ss that has been used to assess the
energy, economic, and environmental implications of
national energy conservation strategies. Some results of
certain energy r,olicy assessments conducted to date are
provided. F inall y, an overview of the implications is pre-
sented, in additon to a description of limitations of the
analytical appro~c;hes that are currently available.
The analytical procedures which I will describe have
been developed to assess the energy, economic, and envi-
ronmental implcations of energy policy proposals. The
environmental i "pact analysis is based upon alternative
. Kenneth R. Woodcock is the Associate Assist"nt Admini-
strator for Environmental Programs, Federal Energy Administra-
tion, Washington, J.e.
energy supply and demand forecasts which are projected
for the short and midterm (i.e., 1980 and 1985). In
general, the environmental analysis is national in scope
with regional resolution when such is possible.
The economic impact analysis presented here
focuses primarily on energy prices and associated con-
sumer cost implications of candidate policies. Further
emphasis is currently being placed on the capital require-
ments of both energy conservation programs and re-
source development programs. Such analysis will facili-
tate the assessment of the effectiveness of capital invest-
ments in energy conservation or energy development
programs.
TYPES OF CONSERVATION STRATEGIES
The analytical approaches utilized in assessing can-
d i da te-conservation strategies vary widely and the
appropriateness of a procedure is highly dependent upon
the energy conservation strategy that is proposed. For
the purpose of this paper, various approaches to energy
conservation have been categorized as follows: (1)
sector-specific strategies; (2) energy-pricing strategies;
and (3) combined strategies. A brief description of the
strategies is provided below.
A. Sector-Specific Strategies
Sector-specific energy conservation strategies strive
to increase the efficiency of energy consumption within
individual economic sectors, such as: transportation,
buildings, industry, and utilities. Sector-specific strate-
gies may be implemented through voluntary or regula-
tory approaches.
Voluntary programs, such as the administration's
industrial or van pooling programs, are based on infor-
mation transfer and public education programs which
provide the energy consumer with the detailed informa-
tion necessary to implement cost-effective energy con-
servation measures. Public information camgaigns are
also undertaken to overcome institutional or communi-
cation barriers wh ich often prevent the marketplace
from responding rapidly to conservation opportunities.
The voluntary approach also includes the establishment
of selected incentives for energy conservation, such as
tax credits for installing home insulation.
Secondly, regulatory strategies for energy conserva-
tion impose direct requirements on activities within an
economic sector to assure that energy efficient products
or capital investments are provided in the marketplace.
32
-------
As with environmental requirements, the standards may
be technology-forcing or may be based on other criteria
(i.e., cost-effectiveness, currently available technology,
etc.). An example of a regulatory energy conservation
strategy is the administration's proposed energy conser-
vation standards for residential and commercial build-
ings.
B. Energy-Pricing Strategies
Energy-pricing strategies have wide-ranging implica-
tions and generally effect every sector of the economy.
Energy consumption reductions are achieved by inde-
pendent (voluntary) decisions on energy use by con-
sumers and firms at all levels in the economy. At the
same time, energy-pricing strategies also trigger resource
development responses for the specific fuel or fuels
affected by the pricing strategies. The $1, $2, and $3
import fees under the Mandatory Oil Import Program,
and the Phased Decontrol of Old Oil are examples of
energy-pricing strategies.
C. Combined Strategies
Combined strategies include the mix of sector-
specific and energy-pricing strategies as well as energy
resource development strategies. These are generally
oriented toward restructuring basic energy consumption
and production patterns, so as to achieve certain energy
goals. For example, legislative proposals in the Energy
Independence Act of 1975 and related tax proposals
were designed to: (1) reduce the U.S. dependence on
foreign sources of energy; (2) increase the consumption
efficiency for certain fuels; and (3) expand the develop-
ment of domestic resources.
ANALYTICAL APPROACH AND TECHNIQUES
The Federal Energy Administration has devoted
much effort to the development and interpretation of
energy, economic, and environmental implications of
energy policies. Development of the analytical tools now
used by FEA was initiated by several Federal Agencies,
in conjunction with the private sector, prior to the
October 1973 oil embargo. I will briefly describe these
analytical tools and the procedures involved in their use.
A. Overview of Analytical Approach
A flow diagram is provided in figure 1 of the anal-
ytical approach utilized for conducting comprehensive
assessments of the consequences of alternative energy
policies. The diagram is most appropriate for a combined
strategy in which there may be environmental, resource
development, energy conservation, economic policy, and
foreign policy elements. However, the same approach
would be applicable to an isolated energy conservation
policy. A comprehensive national energy conservation
strategy is likely to have associated environmental,
resource development, and e<;onomic policy elem~nts to
assure a balanced and orderly transition to a more
energy-efficiency economy.
B. Energy Policy Elements
As shown in figure 1, a combined strategy may con-
sist of five categories of energy policy elements:
Environmental Policy
Energy Resource Development Policy
Energy Conservation Policy
Economic Policy
Foreign Policy
The environmental policy elements may include modifi-
cations to existing environmental legislation, such as the
Clean Air Act, to mesh energy production and consump-
tion patterns and environmental goals. Modifications
that strengthen, weaken or delay environmental require-
ments may be required to adjust to an energy strategy
that relies on an accelerated rate of domestic energy
resource development while continuing to progress
toward environmental goals. Furthermore, modifications
to compliance schedules and degree of technological
dependence may be desirable to achieve a more energy-
efficient approach to environmental protection.
For example, policy elements pertaining to energy
facility siting, sulfur dioxide control for coal-burning
power plants, automobile emission standards, and signifi-
cant deterioration were included in the Energy I nde-.
pendence Act of 1975.
Energy resource development policies provide for
expanded domestic energy resources. This category may
include:
regulatory policies which result in the development
of energy resources in areas that were not previously
under development;
energy-pricing policies to provide additional
incentives for domestic resource development;
financial policies that aid the development of
needed energy production facilities; and
measures providing for strategic and emergency
preparedness in the event of future energy shortfalls
or curtailments.
Energy conservation policies may, as we have dis-
cussed above, consist of voluntary or regulatory pro-
grams or energy tax and pricing policies to encourage the
adoption of cost-effective energy conservation measures.
These may include:
tax credits for insulation;
33
-------
Energy
Consumption
Environmental
I ~n.~ironmental L I R ate and Degree Anillysis
I ru.,,;;y I of POllution
Control . Utilities
. Residential
& Commercial
Energy Resource . Industrial
Development . Transportation
Policy Aggregate
Environmental National and
Impact Regional
Energy Production Analysis Pollution
Energy Impact Analysis Analysis
Conservation (1977 & 1985) Energy
Policy Production Air
Environmental Water
Energy Consumption Anal sis land
Impact
Economic Analysis .Coal
Policy (1977 & 1985) .Oil & Gas
w . Synthetics
~
. Refining
Consumption
Foreign Patterns
Policy
Fuel-Use
Shifts
Figure 1. Energy/environmental/economic impact analysis of energy policies.
-------
public information campaigns;
energy conservation standards for the construction
of new buildings;
efficiency requirements for vehicles, appliances, or
industrial processes;
gradual or immediate decontrol of crude oil and
natural gas prices;
import restrictions, gasoline rationing or allocation;
and
excise taxes on imported oil, all petroleum pro-
ducts, or gasoline only.
The economic policy elements provide means for
minimizing undesirable impacts on consumers, the pri-
vate sector, and the public sector. Economic policy
elements may include:
tax rebates to the consumer;
incentiyes for capital investments;
grant for assistance programs for State and local
governments; and
special provisions to assist low-income persons, such
as federally funded programs to improve the
thermal efficiency of low-income homes.
Foreign policy elements of a combined strategy
could include:
international agreements for the exchange of
information concerning energy conservation and
development;
international agreements pertaining to coordinated
curtailment of world energy demand in the event of
future emergencies; and
international agreements pertaining to sharing of
available energy supplies in event of future emergen-
cies.
C. Energy Consumption Analysis
Energy consumption impact analyses assess the
projected energy savings resulting from specific policy or
program initiatives intended to further energy conserva-
tion. Estimates of reductions in energy consumption are
based on:
estimates of anticipated energy demand growth in
the sector or subsector affected by the policy or
program initiative (taking into account the likely
impact of current energy prices);
projections of changes in consumption habits due to
increased energy prices (as estimated by projected
demand elasticities) or tax incentives to encourage
energy conserving investments;
projections of the replacement of existing capital
stock by more energy-efficient facilities, vehicles,
equipment, or appliances;
assumptions regarding economic growth and the
price of energy if no action is taken;
assessments of the overall potential for cost-effec-
tive energy savings and the likely savings to be
realized as a result of specific actions (such as instal-
ling insulation in an average home); and
assessments of the production capacity and tech-
nological capability of affected industries (such as
automobile or appliance manufacturers).
The level of sophistication of the analyses perform-
ed for each of the many program or policy initiatives
proposed by the administration or Congress has varied
widely. It has ranged from detailed studies of energy
demand elasticities, which are in some respects more
advanced than any other computer modeling effort, to
"quesstimates" of consumer response to measures such
as energy efficiency labeling of new appliances.
The computer model used by FEA to integrate and
project overall energy demand and supply levels to 1985
has relied heavily upon the results of independent anal-
yses for inputs regarding the savings likely to result from
sector-specific conservation measures. However, for over-
all pricing strategies and selected excise taxes, such as a
gasoline tax, the FEA forecasting model is now capable
of estimating energy impact internally. This analytical
system enables a wide variety of factors to be taken into
account simultaneously and also permits projections of
numerous variations of similar pricing or tax policies.
However, because of insufficient data on the complex
interactions between energy demand and factors other
than energy pricing, it has been impossible to use the
model to project the impacts of some nonprice initi-
atives, such as the insulation tax credit and appl iance
efficiency improvement proposals. I n these cases, esti-
mates of energy demand savings are developed independ-
ently and then fed into the computer to become a part
of the overall equilibrium solution that the model deter-
mines. At the same time, FEA is continually attemptin!J
to improve and expand the model through the incorpor.
ation of more sector-specific data.
An example of a sector-specific conservation initia-
tive which FEA has analyzed is the Building Energy
Conservation Standards Act, which is now receiving the
consideration of the Congress. The estimates of energy
savings resulting from the imposition of such standards
reflect a variety of considerations. First, estimates of the
energy demand by new buildings was projected to 1985
on the basis of data on construction rates in the residen-
tial and commercial sectors. These rates are affected by
35
-------
many factors, including economic growth, technological
change, and projections of the average energy-use char-
acteristics of new buildings. Studies were then con-
ducted to de termine the cost-effective portion of the
overall potent ial for energy conservation in new build-
ings, based or actual measurement and simulation. Fin-
ally, an examination of current efforts to improve build-
ing codes provided a basis upon which realistic estimates
could be mace of the savings achievable through the
institution of more advanced energy conservation stand-
ards. Compari:ion of the projected energy demand under
current imposition of standards yields the incremental
energy demand impact of imposing energy conservation
standards for buildings.
D. Energy Production Impact Analysis
Energy prJduction forecasts, by fuel type, are based
on energy derland forecasts and the availability of al-
ternative enen IV resources at the projected production
cost. Policy al ternatives in the energy development sec-
tor are subjectl!d to an analytical procedure analogous to
that described above for the evaluation of energy conser-
vation options.
E. Integratior' of Supply and Demand Analyses
The Pro ect Independence Evaluation System
(PIES) has beel used extensively to project energy con-
sumption and production levels for the 1985 period (ref.
1 I. Recently, 1'1 ES has been further developed for esti-
mates for 1980 and 1990.
The purpcse of the PIES model is to provide an
analytical tool which will generate planning estimates,
depicting poss ble equilibrium states of the national
energy system, recognizing the effect on demand of rela-
tive prices for the various fuels, the potential for fuel
substitution, and the technological constraints which
inhibit energy clevelopment. The PIES model is actually
a framework for the integration of the outputs from a
series of judgmental and quantitative submodels. For
example, separate submodels exist for energy demand
and for the supply of each of the fuel types. Outputs
from each of H ese submodels are then combined by the
integrating moclel to produce a supply/demand balance
consistent with the policy constraints which are being
analyzed.
The model assesses the energy requirements by fuel
type for each;nnsus region and then sums these esti.
mates for a narional total. The PI ES model was devel-
oped to analyze potential energy- related policies and
actions of national significance, and therefore, the aggre-
gate energy est;mates for the Nation are considered far
more reliable than the disaggregated regional estimates.
This condition presents clear limitations for the develop-
ment of reliable environmental assessments. Meaningful
environmental impact assessments are regional or site-
specific in dimension. Thus, the outputs of the PIES
model provide a limited basis from which to project
regional environmental impacts.
F. Economic Impact Analysis
FEA's economic impact analysis provides for the
estimates of:
direct energy price increases by fuel type,
direct household costs by fuel type, geographical
region, and household income categories.
Furthermore, direct and indirect ene~gy costs have been
computed by the use of a stage-of-processing model
developed by Data Resources, Inc. (DR!) (ref. 21. This
model is used for determining indirect energy price
impacts on the consumer price index (CPl).
G. Environmental Impact Analysis
The environmental impact analysis to date has been
focused on the determination of environmental residuals
for alternative energy strategies. I n addition, the number
of energy facilities needed to meet alternative demand
requirements are projected as an indicator of future envi-
ronmental impacts.
The residual and analysis is based on air, water, and
solid waste coefficients which were originally developed
by Hittman Associates (ref. 3). The data has since been
computerized by Brookhaven National Laboratory,
under the direction of the Council on Environmental
Quality, and organized into a system known as MERES
(Matrix of Environmental Residuals for Energy Systems)
(ref. 4). Through MERES, it is possible to estimate the
residual outputs associated with the alternative supply/
demand forecasts generated by PI ES.
Residuals estimates have been developed for all
major energy systems, and include solid wastes generated
and acres of land disturbed, as well as estimated quan-
tities of air and water pollutants emitted. The residuals
estimates for the 1985 period provide an analytical basis
for comparing the net changes associated with alterna-
tive mixes of energy conservation and energy resource
development policies. A major limitation of the method-
ology is that ambient air or water quality changes cannot
be predicted with desirable accuracy. That is, due in part
to the likelihood of unforeseeable changes in residual-
generating variables unrelated to energy matters, the
analysis cannot be carried beyond estimates of energy-
related residuals.
Nevertheless, the residuals approach to energy/envi-
ronmental analysis is a useful too, because the quantita-
tive output reflects the net effect of the opposing strate-
gies emphasizing energy conservation and/or domestic
36
-------
resource development policies. Qualitative statements
regarding future impacts rarely achieve the same level of
decision-relevant information as the residual estimates.
A second index of environmental impact may be
. derived by calculating the number of energy facilities (or
appropriate categories) needed to meet the PIES
supply/demand forecasts, and assuming some given level
of impact to be associated with facilities of a specified
type and size.
IMPACT INDICATORS
, FOR THREE STRATEGIES
Sector-specific, energy pricing, and combined strate-
gies have been selected for the purpose of conveying the
energy policy impact indicators that have been examined
previously pursuant to the requirements of the National
Environmental Policy Act and the administration's
inflationary impact statement requirement. The three
strategies provided herein for illustrative purposes are:
the Bu ildi ng Energy Standards Act of 1975 (Title X
of the Energy Independence Act) (ref. 5);
the Mandatory Oil I mport Program (MOIP) (ref. 6);
and
The Energy Independence Act of 1975 (EIA) (ref.
5).
A summary of the impact indicators provided in the
three impact assessments is shown in tables 1, 2, and 3
for energy, economic, and environmental impact indica-
tors, respectively.
AN EXAMPLE POLICY ASSESSMENT FOR A
NATIONAL ENERGY STRATEGY
An example of the results of an impact assessment is
provided here to illustrate the decisionmaking inputs
that have been developed by the aforementioned
methodologies.
An environmental impact statement (EIS) was
prepared on the President's proposal to establish a fee on
imported crude oil and petroleum production. The scope
of coverage of impacts was both national and regional.
Regional classifications were developed for energy
production areas, energy consumption regions and
summary regions (i.e., areas to be examined for aggre-
gate environmental impacts). to reflect the combined
environmental effects of the oil import fee on energy
conservation, fuel switching toward coal, and new
energy production.
Table 4 is a summary of the differential energy
impacts of the proposed $3 import fee, with alternatives
consisting of a lower fee and a tax on gasoline use. The
gasoline tax, for example, only reduces energy use, while
the fees also stimulate domestic energy production.
Other energy impacts that were identified in the EIS
included:
consumption effects by fuel and consumption
sector for .nine regions; and
regional energy production effects for each energy
source, e.g., strip-mined coal production.
Table 5 presents a nationwide summary of the changes
in pollutant loadings that would occur with an oil
import fee, for major air and water pollutants and solid
wastes. These estimates are based on more detailed
projections of pollutant loadings for production and
consumption effects by region. I n order to place these
estimates in a more meaningful context, the results were
presented in six summary regions identified in figure 2.
Table 6 contains a summary description for each of
these regions including:
which pollutants will rise or fall due to the import
fee; and
the sector causing the change in emissions, e.g., fuel
switching to coal in Region IV (Midwest) causes
increases in sulfur dioxide.
These pollutant-loading effects are also discussed in
comparison with the environmental quality problems in
each of the six regions. For example, the Midwest al-
ready has significant particulate control and fugitive dust
problems. Consequently, the increase in particulate emis-
sions in that region should be of greater environmental
concern than increases in many other areas of the
Nation.
D. Economic Impacts
The costs of the oil import fee were presented in the
context of the household. Table 7 shows the rise in
household costs of energy by region which would be
caused by the imposition of the $3 oil import fee. It
shows, for example, that the Northeast would be more
severely impacted than the national average, because its
energy use for heating is so heavily dependent on petro-
leum products. Table 8 reflects another aspect of eco.
energy costs for households with varying income levels.
This comparison indicates that the import fee could have
a greater impact on family incomes for the lower income
group.
METHODOLOGICAL LIMITATIONS
The analytical tools for energy and economic anal-
ysis of energy conservation strategies are relatively well
advanced. The PIES model has contributed significantly
to the understanding of the impact of supply and de-
mand components of future energy plans. The Project
Independence Report (PIR) and other energy policy
37
-------
Table 1. Energy impacts of energy strategies
Impact indicators
EIA
Title Xa
ENERGY STRATEGIES
MOIpb
EIAc
A. Short,.term (1975, 176, '77,
or 1713)
- Natural gas demand
- Pet 1~0 1 e um deman dua 11
sectors
- Coal demand--uti1ities
xd
e
x
x
x
x
B, Midte"m (1985)
1. Corsumpti on
- E y fue 1 (oil, gas,
coa 1, e lectri city,
nuclear, other)
- Py sector (residentia1/
commercial, industrial,
transportation,
utilities)
- By cause
. Auto Re 1 ated
. Residential/Commer-
cial Building Related
- B,y regi on
2, Producti on
- All fuels by production
rl~ gi on
3. Imports (crude products,
natural gas)
x
x
x
x x X
x x
X
x x x
x
x
x
x
a Build':ng Energy Standards Act of 1975 (Title X of the Energy Independence
Act of 19/5).
b Mandatory Oil Import Program ($3 Import Fee).
C The Pr'esident's Energy Program (Energy Independence Act of 1975).
d x denctes coverage in referenced sources,
e - denotes no coverage in referenced sources.
38
-------
Table 2. Economic impacts of energy strategies
Impact indicators
EIA
Title Xa
ENERGY STRATEGIES
MOlpb
EIAc
1. 'Pri ce impacts on petroleum
products
2. Direct energy costs to
household
- by fuel typef
- by Geographic Regiong
- by Household Income
Class!!!
3. Aggregate value of energy
savings
4. Employment
- Gas utilities
- Petroleum refineries
- Electric utilities
5. Capital requirements
'e
d
x
x
x x
x x
x x
x
x
x
x
x
a Building Energy Standards Act of 1975 (Title X of the Energy Independence
Act of 1975).
b Mandatory Oil Import Program ($3 import fee).
c The President's Energy Program (Energy Independence Act of 1975).
d x denotes coverage in referenced sources.
e - denotes no coverage in referenced sources.
f Gasoline and motor oil, heating oil, natural
energy.
g Northeast, Middle Atlantic, East North Central, West North Central, South
Atlantic, East South Central, West South Central, Mountain, and Pacific.
h Poor, lower-middle, upper-middle, well-off.
gas, electricity, and all
39
-------
Table 3. Environmental impacts of energy strategies
Impact indicators
EIA
Title Xa
ENERGY STRATEGIES
MOl pb
EIAc
A. National impacts
1. Energy consumption
activities
a. Utility conversion
to coal (1977)
b. Utility sector (1985)
c. Industrial sector
( 1 985 )
d. Household/commer-
cial sector (1985)
e. Transportation
sector (1985)
2. Energy production
activities
a. Coal production
(1977 and 1985)
e
xd x
x x
X X
X X
X X
x
- surface mining
- underground mining
- beneficiation I
- aggregate coal
impacts
b. Oil production
c. Petroleum refineries
d. Natural gas production
3. Agqregate consumpti on/
products
a. Air poll~tantsf
- na ti onwi de
- re 9 i on a 1
b. Water po11utantsg
- nationwide
~ regi ona 1
x x
x x
x x
x x
x x
x x
x x
e xd
x
x x
h
x x
x x
40
-------
Table 3. (can.)
ENERGY STRATEGIES
. Impact indicators EIA a tIO I pb EIAc
Title X
3. ( con.)
c. Solid wastei
- nationwide x x
- regional x x
B. Regional impacts (1985)
1. Energy consumption
act i viti es
a. Utility sector x x
b. Industrial sector x x
c. Household/commercial
sector x x
2. Energy production
acti vities
a. Coal production
(1977 and 1985)
- surface min ing x x
- underground mining x x
- beneficiation x x
- aggregate coal
impacts x x
b. Oi 1 production x x
c. Petroleum refineries x x
d. Natural gas production x x
e. Oil shale x x
f. Coal gasification x x
3. Aggregate consumption/
production
a. Air pollution e xd
x
b. Water pollution x x
c. Solid waste x x
41
-------
Table 3. (con.)
Impact ~ndicators
ENERGY STRATEGIES
EIA
TitleXa
MOlpb
EIAc
4. RE lationship of regional
emission changes to cur-
rEnt environmental quality
p "ob 1 ems
x
a Building Energy Standards Act of 1975 (Title
Act of 1975).
b Mandatory Oil Import Program ($3 import fee).
c The Presi dent's Energy Program (Energy Independence Act of 1975).
X of the Energy Independence
d x denotes coverage in referenced sources.
e - de10tes no coverage in referenced sources.
f Total suspended particulates, sulfur dioxide, nitrogen oxides,
hydrocar~)ons, and carbon monoxide.
g Acid~;, suspended solids, dissolved solids, organics, thermal emissions.
h MOIP impacts included alkalinity, acidity, chemical oxygen demand, and
biological oxygen demand, in addition to those references in g.
i Includes fixed land and incremental land impacts.
assessments ha\e provided the Congress and the public
with 10-year energy forecasts that contribute to a better
understanding ,)f the difficult and important choices
that must be me de.
However, procedures for environmental impact anal-
ysis of national conservation strategies have provided
information of marginal relevance for public pOlicy
determination. Whereas individual environmental impact
"hot spots" wi1 hin the country can be projected with
MI;RES, the ab lity to provide accurate ambient impact
estimates for 10 years hence is, simply, not now in hand.
At the same time, this is not to say that the environ-
mental impact methodology presented herein does not
contribute to and improve understanding of the conse-
quences of alternative energy policies. The methodology
provides a means for observing the interrelationship
between policy choices and the environmental factors
that may be directly or indirectly affected.
Furthermore, the overall methodology for conduct-
ing energy, economic, and environmental assessments
plays an important role. Too often, energy policies are
evaluated from only one point of view, such as consumer
42
-------
Table 4. Reduced oil imports under alternative
scenarios, 1958 (expressed in terms of differences
from base case)
Millions BID
Factors $3 Fee $2 Fee Gasoline tax
\
Reduced energy consumpti on 1.0 . 7 2.8
Domesti c oi 1 producti on 1.5 1.0 -. 1
Sh i ft to coa 1 2.3 .7
Sh i ft to natural gas .2 .2
Total import reduction
5.0
2.6
2.7
Table 5. Environmental impacts of oil import fee
Percent
change from base
1977
1985
Air
Total suspended particulates (TSP) +1.0 +0.4
Sulfur oxides (SO ) +3.0 -3.0
x
Nitrogen oxides (NO) -1.0 -4.0
x
Hydrocarbons (HC) -2.0 -5.0
Carbon monoxide (CO) -1.0 -2.0
Wa te r
Total dissolved solids +0.1 -1.0
Suspended so 1 i ds 0.0 -20.0
Nondegradable organics 0.0 -24.0
Alkalinity 0.0 +12.0
Acidity 0.0 -16.0
B;olog;cal oxygen demand (BOD) 0.0 -6.0
Chemical oxygen demand (COD) 0.0 -48.0
So 1; d was te
Waste 0.0 +1.0
43
-------
-,
~\.:-
" .!"'r'J
...
~s.
, . G
r 0 ,
.r:.
.r:.
c ,
! :; -
, ,
. . .
, .
"."":::" 4.
VI
-~..
----
--- -
T~'-
. 0
-'';:~~~~
- '.
....
o
- .
. -
f\ . :; '"
v
...."'",
~~~..!'
. . ;;. A .:. C
"':: .
"" 0 Q ,. ..
o ... ( 0 T ..
SOL T ...
D A Ie';" ..
,,--lV'
e ;:; A S t ...
, s
c \)
IliA
"
~r;~
- r
c, -,--- --:j'~-c'
/.' ~'" 0
/ 3~-' ~h ----'
..--~ ~
'.-: C':: _:~~Zr-" -
'oJ' .', 7~ -' '';'::=
.,} :;",--
III I"'''''' E. SOT ...
.
~
"...., ..-
/
-~~ ~~
',' . '" J 0
~l"'"
~'-~'. I .--..-"-
1/
.'/
..J -I
- )
. - -
1/ '
//
-;.
,J '" ~ \ '" l ....
(
I~~~J
...J
;
- ... s'"
r
-{
t:=. ,"
- --:;...
-. c. -:..:.:..: . ...... ~.
, ~~~-:~,.,-
--4-..-
PEM"S~\.I/""\.
"
".y /S-L~~:,', :~~1}" -
,( .
"
o "
~ I
-: ;-::;-.~~;~]~~.'
,-..,~.,- 'I o~:~:i
~Q:.v ~.,
" , ~ G \ No \.....:.,~:. i
- ....,
, ,-;;,
, . . c c ' ~ /~ ~:)
, ;.:.;
11'"
C ... A
50\,/1\04 ,,!
-.C"'RO\..\"'.~
... :. . .. .... 5 .. S
- - ---
~
~
",'
~'
y.'
j
-{
- r
5
o
c.
..,
~
~
'. .'- F
""..-= -~I...c:>~ 0
..
\,
, "
t'- ,,\
~ '
'~.
"; l L r
:;
H
....
'V.
./
../*' -
Figure 2. Summary regions.
'11"11.1.'0'''. .\t ,J'
'" Ttn.
1::-;IT~:i) s;r.\T~;S
-----'-~-~-
-------
Table 6. Regional environmental impacts: oil import fee
Region
TSP
Air
NO
x
HC
CO
Water
S02
I
Decrease
Oil refining reduction
Industrial demand reduction
Transportation demand reduction
So 1 i d was te
Decrease
Oil refining re-
duction
Industrial demand
reducti on
Decrease
Oil refining reduction
Industrial demand re-
duction
I I
Increase
Utility coal
substi tuti on
Decrease
Transportati on
reducti on
Decrease
Oil refining re-
duction
(alkalinity increases
due to coal mining)
Increase
Coal mining increase
Utility coal sub-
stitution
III
~
U1
Decrease
Demand reduction
Oil refining reduction
Transportation demand reduction
Decrease
Oi 1 refining re-
duction
Demand reduction
Decrease
Oil refining reduction
Demand reducti on
IV
Increase
Utility coal substitution
Oil refining increase
Gas production increase
Increase
Utility coal sub-
s ti tuti on
Oi 1 refining increase
Increase
Util i ty coal
t uti on
Oi 1 re fi n i n g
increase
s ub s t i -
V
Decrease
Demand reducti on
Oil refining reduction
Transportation demand reduction
.
Decrease
Oil refining re-
duction
Demand reduction
Increase
Coal mining increase
VI
Decrease
Demand reduc-
tion
Increase
Gas production
increase
o Oi 1 refining
increase
'0
Increase
Oil refining increase
Gas production
increase
Increase
Oi 1 refin ing increase
.
-------
Table 7. Regional distribution of the increased direct
consumer energy expenditures (per household) under
oil import fees ($!year)
Heating
Regi on Gasoline. oil E1ectri ci ty To ta 1
Nort'1east 41 42 14 84
Mi d-.~t1anti c 35 20 7 55
Eas t North Centra 1 47 9 1 57
West North Cent ra 1 55 6 0 61
South Atlantic 51 4 8 55
East South Cen tra 1 52 1 54
West South Central 52 0 2 54
Moun ta i n 62 1 1 65
Pacific 45 1 5 51
AVE'rage u. S. 47 8 4 59
cost increases. By providing rigorous analytical assess-
ments of the pnlicies, the focus of attention shifts to the
full societal co~ ts and benefits of energy policy choices.
Clearly, a bala!1ced approach is needed to adjust our
energy production and consumption patterns to be
consistent with our domestic energy resources, societal
goals, and techrological capabilities.
CONCLUSIONS
Studies co"ducted to date of the Mandatory Oil
I mport Program and the Energy I ndependence Act have
provided some !leneral conclusions regarding the implica-
tions of energy ;;onservation strategies:
Sector-spncific policies show mainly positive
environmertal effects.
Economic impacts of sector-specific policies are
generally positive because the conservation ap-
proach is f -equently the cheapest source of energy
presently a\ ailable.
A mix of both beneficial and adverse environmental
effects is expected under energy-pricing conserva-
tion strategies due to reduced consumption and
increased resource development. Furthermore, con-
versions of industrial and electric facilities from
natural gas to oil and coal or oil to coal also gen-
erate adverse impacts, due to the use of dirtier fuels.
The economic impacts of energy-pricing strategies
are frequently adverse unless combined with com-
pensating economic policies that minimize the
resultant consumer cost impacts.
With the range of positive economic and environ-
mental consequences of alternative energy conservation
strategies that have been observed, the Federal Energy
Administration continues to identify and implement
voluntary conservation programs that apply technologies
which are currently in hand and which produce cost-
effective savings. Policy analysis, communication with
the public, program demonstration, and implementa-
tion: all of these are essential elements in the emerging
national energy conservation strategy.
46
-------
Table 8. Current household energy costs by income class
Lower- Uppe r-
Poor, mi ddl e, middle, We l1-off,
average average average average
$3,000 $9,600 $16 , 800 $29,400
Gasoline $173 $432 $ 778 $ 914
Heating oil 66 66 66 83
Natural gas 91 108 117 140
Electricity 160 203 259 319
Coal 16 16 16 16
Tota 1 $506 $825 $1,236 $1,472
Percent of
average income 16.9 8.6 7.4 5.0
Source: WCMS Survey for 1972-1973, adjusted for pr; ce ; ncreases
to September 1974.
REFERENCES
1. Project Independence Report, Federal Energy
Administration, November 1974.
2. Data Resources, Inc., An Energy Price Impact
Model, October 1974.
3. Hittman Associates, Inc., Environmental Impacts,
Efficiency, and Cost of Energy Supply and End Use,
2 vols., 1974.
4. Cf. MERES and the Evaluation of Energy Alterna-
tives, Council on Environmental Quality, May 1975.
The Council of Environmental Quality has also
organized the preparation of a reference work based
on MERES for use by preparers of energy-related
environmental impact statements. See Energy
Alternatives: A Comparative Analysis, Council on
Environmental Quality, et aI., May 1975 (prepared
by the Science and Public Policy Program, Univer-
sity of Oklahoma, Norman, Oklahoma).
5. See Draft Environmental Impact Statement, Energy
Independence Act of 1975 and Related Tax Pro-
posals, Federal Energy Administration, March 1975.
Chapter 12 gives particular attention to the Buildinn
Energy Standards Act of 1975.
6. Cf. Draft Environmental Impact Statement, Manda-
tory Oil Import Program, Federal Energy Admin-
istration, June 1975.
47
-------
LEGISLATIVE PROGRAMS IN ENERGY CONSERVATION
Maxine l. Savitz, Ph.D. *
Abstract
As part OJ' the President's overall energy program,
four legislative initiatives were submitted to Congress in
1975. The Energy Independence Act of 1975 contained
three of these actions: Title X, Building Energy Conser-
vation Standards Act of 1975, which would establish
national themal efficiency building standards for all
new buildings to be implemented by States and local
governments; ritle XI, Winterization Assistance Act of
1975, which v.fOuld provide States with funds to help
purchase energl conserving materials to be installed in
homes owned ('r occupied by low-income families; Title
XII, National Appliance and Motor Vehicle Energy
Labeling Act of 1975. which would require labeling of
energy consuming major appliances. In addition, the
President requl'sted Congress to enact legislation that
would amend t7e Internal Revenue Code to provide in-
dividuals with G tax credit for 15 percent of the cost of
modifying thei,. homes to conserve energy. In addition,
other legislativE options for achieving energy conserva-
tion in building1: and industry have been proposed by the
members of Co.1gress. The status and implication of the
various legislath'e activities will be discussed.
Any paper presented on proposed legislative actions
during an ongo I1g session of Congress runs the risk of
being out of da.:e the day after it is written. This paper
will be a status report as of October 1 on several pieces
of legislation re ated to energy conservation in the end-
use sector.
Almost 1,200 bills related to energy have been intro-
duced into the ~14th Congress. Of these bills. about 40 to
60 remain activt:, with about 20 including a decrease in
fuel consumpticn as one of their objectives. This paper
will deal with several specific pieces of legislation related
to energy consel vation-more efficient use of energy-in
buildings and in industrial, and transportation sectors. It
will not deal with bills which might achieve reduction in
energy use through tax and regulatory actions to in-
crease the price of petroleum and natural gas. Most of
the legislation c iscussed in this paper has been passed
either by the H.)use or Senate. No energy conservation
*Director, Dil'ision of Buildings and Industry Conservation,
Energy Research and Development Administration, Washington.
D.C.
legislation has yet been passed in this session by both
Houses of Congress and sent to the President. Some of
the legislation discussed will have been proposed by the
Administration, and some by Congress.
BUILDINGS SECTOR
As part of the President's overall energy program,
several legislative initiatives to accelerate energy conser-
vation in the buildings sector were submitted to Con-
gress in January 1975. The Energy Independence Act of
1975 (5. 594. H.R. 2650) contained three of these
actions: Title X. Building Energy Conservation Stan-
dards Act of 1975; Title XI, Winterization Assistance
Act of 1975; and Title XII, National Appliance and
Motor Vehicle Labeling Act of 1975. In addition, the
President requested Congress to enact legislation that
would amend the Internal Revenue Code to provide in-
dividuals with a tax credit for 15 percent of the cost of
modifying their homes to conserve energy. The legisla-
tive proposals in the buildings sector are an integrated
package aimed at reducing energy waste in buildings.
They are interdependent in that they deal with both the
long- and short-term and cover all segments of the build-
ings sector. Title X, Building Energy Conservation Stan-
dards Act of 1975. would establish national. thermal
efficiency building standards for all new buildings to be
implemented by State and local governments. This pro-
posal is particularly important, since by 1985 approxi-
mately 30 percent of all residential units and 40 percent
of all commercial space will have been constructed after
1974.
Adoption of energy conserving design practices
could reduce the heating and cooling energy consum.p-
tion of single-family dwellings by 35 percent, of high-rise
multifamily structures by 24 percent and of commercial
buildings by 32 percent. This legislation would result in
energy savings of over 500,000 BPD by 1985, and the
savings would continue to grow in the future. The legis-
lation as proposed by the Administration would have
standards developed and promulgated in three phases:
. Phase 1
Secretary of Housing and Urban Development
would, within 6 months after enactment of the legisla-
tion. publish for public comment proposed component
performance standards for new residential buildings. The
recently adopted ASHRAE 90 is an example of a com-
ponent performance standard. Final standards would be
48
-------
developed and promulgated 6 months after publication,
effective 1 year after promulgation.
. Phase 2
New minimum performance standards (e.g., energy
budget) for energy in commercial buildings would be
developed no later than 18 months after legislation.
. Phase 3
Minimum performance standards for residential
buildings would be developed not later than 3 years after
legislation was enacted. These would supersede standards
promulgated in Phase 1.
Research and demonstration activities to assist in
the development of standards for Phases 2 and 3 would
be performed by HUD, Energy Research and Develop-
ment Administration, Federal Energy Administration,
and the National Bureau of Standards. The standards
would be developed also in consultation with Federal,
State, and local government agencies, industry, labor and
consumer groups. The National Institute of Building
Sciences (NIBS) would be utilized.
The legislation would require States to certify to
HUD that the new standards have been inco~porated
into local building codes and that they will be enforced.
If, within a specific period of time, a State has not certi-
fied that the Federal energy conservation standard or an
equal or better standard is being adopted and imple-
mented by the appropriate local or State code, Federal
financial assistance for the construction of new buildings
would not be approved in that area.
Title X authorized HUD to make grants to the
States to assist them in meeting the costs of developing
procedures to implement energy conserving building
standards.
In the spring, the Senate passed Title X essentially
as proposed by the Administration as part of an emer-
gency housing bill. It was dropped in conference to give
the House an opportunity to consider the bill more
thoroughly. In September, the House passed its version
of energy conservation performance standards, H.R.
8650. The version passed by the House dropped the
mandatory aspects of the standards bill, but directs HUD
to monitor progress of the States and local governments
and report to Congress within 12 months of enactment
and then semiannually report the progress of States in
adopting energy conservation standards. The House-
passed version has only two phases-development of per-
formance standards for new residential and commercial
buildings. It omits the component performance phase.
The Senate is still reconsidering Title X and other
options, but as of this date has not produced a standards
bill since the one passed earlier was rejected in confer-
ence.
Two proposals have been made by the Administra-
tion w,ith regard to existing buildings-tax credits and
weatherization. A tax credit, not a deduction for the
cost of purchasing and installing such items as ceiling
insulation, weatherstripping, and caulking was proposed.
As a part of the House-passed H.R. 6860 ("The Ullman
Bill"), a tax credit of 30 percent per homeowner on an
inve~tment of $500 was passed. The tax credit would
expire in 3 years. This bill would also allow changes in
investment credit related to energy conservation options
such as insulation for commercial buildings. The Senate
had passed earlier a tax credit for insulation as part of
the Tax Reduction Act of 1975, but it was eliminated
during conference because it was felt it should be part of
energy legislation. The Senate has not yet taken new
action on this tax credit.
It was recognized that there is an adverse impact of
higher energy prices on low-income persons, and that
many of these people would not be able to take advan-
tage of the tax credit; thus the President proposed a
winterization assistance program for low income per-
sons. The proposed legislation would call for a $55
million program in fiscal years 1976, 1977, and 1978 to
fund State winterization programs. The funding would
allow for purchase of energy conserving materials for
low-income use. State and local officials are to be en-
couraged to utilize the Labor Department's public ser-
vice employment program, as well as voluntary labor
programs, The goal is to winterize approximately 1.5
million homes by the end of fiscal year 1978. The
House-passed H.R. 8650 contained a title for weatheriza-
tion assistance for low-income persons, similar to the
Administration's bill. The Senate has not considered this
legislation.
Another aspect of the buildings sector concerns
appliances. The President proposed mandatory energy
efficiency labeling of room air conditions, water heaters,
refrigerators, freezers, ranges, washers, dryers, and telp.- .
visions (Title XII). In the spring, the Senate passed by a
large majority a Truth in Energy Act - S. 349. This bid
provides for mandatory labeling, and directs the Federal
Trade Commission to define average use cycles for
household appliances, and to devise procedures by which
average use cycles may be simulated and by which
energy utilized during cycle may be measured or calcu-
lated. I n September, the House passed H.R. 7014 ("The
Dingell Bill"). One section of the bill requires labeling of
all major appliances and has mandatory efficiency
targets for appliances.
INDUSTRIAL SECTOR
The Administration itself has not submitted any
legislation to Congress regarding industrial conservation.
49
-------
Current projects in industrial conservation are thought
to be sufficiently comprehensive at this time. These in-
clude a volun1ary program underway with the Federal
Energy Administration, the Department of Commerce,
and industry; ":he research and development program at
ERDA; and industry's efforts on its own. There are two
bills, S. 1908 clnd H.R. 8495 "Industrial Energy Conser-
vation Act of 1975," which are currently being
considered in Congress. Both bills have two identical
sections. One sl!ction directs the Administrator of ERDA
to establish an:l maintain a comprehensive research, de-
velopment, and demonstration program related to
energy conserv ng industrial technologies. It would allow
for contracts .md grants to private organizations and
business to conduct demonstrations. The other section
would allow 1 he Administration to make loans and
guarantee the payment on loans for the purchase, con-
struction, operntion, or maintenance of energy efficient
equipment in facilities. The Senate bill S. 1908 would
establish a mandatory efficiency reporting system and
energy efficienl:y targets. This section is similar to one
which was in H.R. 7014 (Dingell Bill, Section IV B). In
voting on this lIill, the House voted 220 to 187 to delete
the provision aimed at reducing industrial consumption
of energy. It would have set targets for the Nation's
2,000 largest industries and, although voluntary, would
have required those industries to make public reports of
what they were doing to save energy.
TRANSPORTATION SECTOR
The labeling bill proposed by the Administration
(Title XII' mandates labeling of motor vehicles in
addition to appliances. The Senate-passed version also
considered automobiles. H.R. 7014 contains provisions
for automobile ,~fficiency standards; H.R. 6860 provides
for an automobile efficiency tax. The fate of these two
measures in the Senate is uncertain. The Administration
has addressed the issue of reducing automobile fuel con.
sumption by obtaining voluntary commitments from
automobile manufacturers to improve the production
weighted average of their new cars by 40 percent in
1980. Measured against the 1974 model average of 14.0
mpg, the 40 percent goal translated into an average new
car fuel economy of 19.6 mpg in 1980.
Congress is considering an "Electric Vehicle Re-
search Development and Demonstration Act of 1975"
(H.R. 8800, S. 1632). The bill passed the House in
September and is currently being reviewed by the Senate
Commerce Committee. The House passed version man-
dates ERDA to procure within 15 months 2,500 urban
passenger and commercial vehicles which will meet
initial standards and criteria developed within a year of
enactment of the legislation. Within 42 months, 5,000
vehicles would be obtained. The emphasis of the bills is
on near-term demonstration rather than developm~nt of
improved electric vehicles. Due to the time constraints,
lead-acid batteries would be needed in the initial phase.
It is felt by ERDA that a minimum of a 3- to 5-year
program is necessary to develop advanced batteries to
provide vehicles with lower overall operating costs, and
that a demonstration program should be delayed until
battery development programs can provide the appro-
priate improved energy storage system.
As can be seen from the above discussion, en-
actment of proposed legislation-even what appears to
be straightforward legislation such as tax credits for re-
trofit of residences-requires much time, discussion, and
compromise. It is hoped that some of the above-
mentioned legislation-such as building standards, tax
credit, and weatherization-will be passed and signed
into law in order for some of the uncertainties in energy
policy to be removed and for the projected savings to
start to accumulate by using energy more efficiently.
50
-------
HIGHLIGHTS OF ENERGY CONSERVATION PROGRAMS
Howard Hagler* and Harvey M. Bernsteint
A bstrac t
Highlights of energy conservation opportunities in
the four consuming sectors (residential, commercial, in-
dustrial, and transportation) are presented along with a
brief description of the energy demand patterns through
1985 for all four sectors. In addition, a detailed discus-
sion of some energy conservation programs conducted
by Hittman Associates in the residential and commercial
sectors is presented to highlight how energy is being con-
sumed, how energy consumption can be reduced, and
the impacts associated with energy conservation mea-
sures and their implementation. Programs discussed in-
clude a technology assessment of energy conservation
innovations for residential structures, a study of residen-
tial energy conservation in the Baltimore/VVashington
area as well as in 10 other cities in the United States, and
a study to analyze factors which are related to energy
'use in the commercial sec tor.
INTRODUCTION
I n recent years a number of Federal programs have
been focused on energy conservation in the four con-
suming sectors: residential, commercial, industrial, and
transportation. The performance of these programs has
been timely in helping to identify energy consumptive
trends as well as ways to minimize consumption, and in
a number of cases has resulted in the development of
national energy policy. The energy conservation pro-
grams have covered a wide range of topics, embracing on
the concepts of energy use indexes, demonstration proj-
ects, life cycle analyses, improved processes, newequip-
ment, retrofitting old equipment, etc. This paper high-
lights some of the major energy conservation opportuni-
ties in all four consuming sectors and then presents a
more detailed description of some of the current energy
conservation programs in the residential and commercial
sectors. Prior to highlighting the programs, a brief de-
scription of the energy demand patterns for all four con-
suming sectors is presented.
.President, Hittman Associates. Inc., Columbia, Maryland.
tManager, Energy Conservation Department, Hittman
Associates, Inc., Columbia, Maryland.
ENERGY DEMAND PATTERNS IN THE
FOUR CONSUMING SECTORS
In 1972, the U.S. aggregate energy demand was 7.32
quadrillion Btu's (ref. 1). Of this total, the four major
end-use sectors consumed energy in the following pro-
portions:
Residential
(10.5 quadrillion Btu's) .. . . . . . . 14.3 percent
Commercial
(6.2 quadrillion Btu's) . . . . . . . . . .8.5 percent
Industrial
(21.9 quadrillion Btu's) . . . . . . . . 29.9 percent
Transportation
(17.8 quadrillion Btu's) . . . . . . . . 24.3 percent
The remaining 23 percent of the 1972 demand was com-
prised of two auxiliary end uses: nonenergy uses of
energy resources (petrochemical feedstock, etc.). and
losses in the generation and transmission of electrical
power.
In its recent report on the potential for energy con-
servation in the United States, the National Petroleum
Council (NPC) disaggregated the 1972 energy demand
by end uses presented above, and made energy demand
projections through 1985 (ref. 1). These projections
reflect a continuation of past and current trends, and
embody a number of economic assumptions (too numer-
ous to mention here), which resulted in projecting that
the 1972 U.S. energy demand will almost double by
1985. The projected increase is essentially attributed to
increasing nonenergy uses of energy resources (e.g.,
petrochemical feedstock) and to energy consumed due
to electricity production and distribution losses. These
represent 31.5 percent of the 1985 U.S. projected total
demand and an increase of 37 percent over the portioll
they consumed in 1972.
I n view of these trends, it is clear that in order to
maintain our Nation's economic health and standard of
living without jeopardizing the independence of our
foreign policy activities, we must exploit every possible
opportunity to use energy more wisely and efficiently.
Several ways to achieve this end are described in the
following section.
HIGHLIGHTS OF ENERGY CONSERVATION
OPPORTUNITIES BY END-USE SECTOR
In this section, highlights of some of the opportuni-
ties for conserving energy are listed for each of the four
51
-------
l:ntial and commercial sectors, based on work
performed b\ Hittman Associates, follows this section.
Residential S('ctor
Existing Homes
Add storm windows and storm doors,
Use forced attic ventilation when attic tem-
pera1 u res exceed 100° F,
Add ~eiling insulation,
Apply caulking and weatherstripping around
door'; and windows,
Insul3te exposed basement wall,
Set thermostats back.
New Homes
Reduce window area,
Spec ify installation of storm doors and
wind )WS,
I ncrease wall insulation,
i nsta I white roofing shingles instead of black.
Commercial S'!ctor
Existing Buildings
Establish 65° F as maximum indoor setpoint
temp'~rature,
Estatlish 10° F se~back during unoccupied
hour!,
Insulilte ceiling,
Redu-ce lighting levels,
Shut down cooling system 1 hour before
closirg.
New Builclings
Avoic' use of terminal reheat cooling systems,
Determine building orientation to optimize use
of sol ar energy,
Speci"y double-glazing, .
Use ventilation air and lighting heat recovery,
Use s1eam condensate heat recovery.
I ndustrial Sect)J"
Existing P.ant and Equipment
I nstitll1e energy management program,
Insulate bare steam lines,
Return steam condensate to boiler plant,
F I a.h high-pressure condensate to produce
low-pressure steam,
Eliminate leaks in gas lines,
Lower pressure in compressed air lines to
minirrum,
Uso 1!I\!lilW IJxhalist heat 101 51eam !IOIII)I al il.ln,
Check for proper fuel temperature lor luel oil
atomization.
New Plant and Equipment
Substitute coal derivatives for oil and gas
derivatives as feedstock for plastics industry
production whenever Dossible,
Implement raw materials with lower primary
energy requirements wherever possible,
I mprove grinding processes in cement pro-
duction,
Build cement kilns with heat recovery tech-
nology,
Invest in improved technology to improve the
efficiency of electrolytic reduction cells used in
smelting alumina to aluminum,
I ncrease use of sawmill residue and recycled
stock in paper production,
Increase use of recycled rubber hydrocarbon.
Transportation Sector
Existing Automobiles
Auto maintenance,
Carpooling,
Driver behavior improvements.
New Automobiles
Stratified charge engine,
Smaller, lighter autos,
Diesel engines,
Battery-powered auto,
Reduced rolling resistance,
Reduced aerodynamic drag,
Transmission modification,
Axle ratio modification.
ENERGY CONSERVATION PROGRAMS IN
THE RESIDENTIAL SECTOR
This portion of the paper will address two of the
more recent energy conservation programs performed by
Hittman Associates in the residential sector for the
Department of Housing and Urban Development, and
briefly mention a third ongoing effort.
Residential Energy Conservation in the Baltimore/
Washington Area
The objective of this study (ref. 2) was to examine
in depth the most significant ways in which energy is
consumed in our homes, and to evaluate various wch-
niques that can be employed to reduce that energy con-
sumption.
The analysis of residential energy conservation
measures followed the basic methodology outlined
below;
52
-------
1. The Baltimore/Washington area was selected as the
geographic and climatic region to be studied in
depth.
2. "Characteristic" dwellings were defined, using
demographic data, for the most prevalent single-
family. townhouse, low-rise, and high-rise struc-
tures. These dwellings were used as a basis for
evaluating the merit of energy-conserving modifica-
tions.
3. The program was directed toward energy savings
possible in new construction only.
4. All structures were assumed to use good quality
construction.
5. One weather year (1954) was selected as representa-
tive of the Baltimore/Washington weather.
6. A constant daily occupancy and appliance/lighting
schedule was used.
7. A computerized simu lation of the characteristic resi-
rlences was used, subject to computerized simula-
tion of the weather, internal loads, and occupancy
loads to calculate the precise energy requirements
for the characteristic residences.
8. The simulated residences were modified to include
certain energy conserving modifications, and the
residence energy requirements were recalculated for
the "modified" structure.
Energy conserving modifications to the "characteris-
tic" dwellings were limited to those which:
1. would save sufficient energy to pay for themselves'
within a 10-year time span;
2. would be within current construction capability and
be aesthetically acceptable from an architectural
standpoint;
3. would not significantly alter the lifestyle of the resi-
dents; and
4. would be applicable primarily to new structures.
The computer program used in the energy calcula-
tions was originally developed for the U.S. Postal Service
for the analysis of energy balance in post office build-
ings. However, this program was not usable in its original
form for residential buildings. Therefore, it was modified
and its capabilities were amplified.
Following the above methodology, the results of the
study identified quantitatively methods whereby energy
consumption can be reduced within single-family and
multifamily residences. Though the analyses were per-
formed for residences characteristic of current construc-
tion in the Baltimore/Washington area, the results can be
extended generally to both old and new construction,
and to residences in different geographic areas (with
limitations) .
Three areas of energy consumption within resi-
dences were identified for the study: the design and
structure, the internal loads (lights and appliances), and
the comfort control systems. A description of the ob-
served results of the study (ref. 2) with respect to these
three areas follows:
Design and Structure.- The potential for energy savings
in the area of design and structural modification is
dependent on the climatic conditions, and on
whether the residence has already been constructed
or is still in the design stage. For new construction,
those modifications applicable to colder or warmer
climates can be readily incorporated. However, for
existing structures, modifications such as increased
insulation in walls and floors may prove to be im-
practical. Emphasis should therefore be placed on
modifications that can be more easily accom-
modated such as addition of storm windows and
storm doors, cau I ki ng and weatherstripping, and
placement of window shading. Particular attention
should be given to reducing the amount of air in-
filtration through cracks in siding, windows, and
doors. All existing residences should be checked for
areas of air infiltration.
Internal Loads-The modifications identified for re-
ducing the energy consumed by lighting and appli-
ances are those controlled primarily by consumer
choice and would be applicable to both new and
existing residences regardless of their geographic
location. However, the ability to implement these
modifications may be restricted by the unavailabili-
ty of gas in some areas and the cost of replacing
existing equipment while it is still operable.
Comfort Control Systems-The applicability of modifi-
cations identified for reducing energy consumed by
the comfort control systems would be influenced by
the climatic conditions associated with the geo-
graphic area of the residence, and by the type and
age of the residence. For example, modifications
associated with cooling may be not applicable fer
northern climates and the conversion of one form 0'
heating to another in an existing building may be
impractical. However, all residences in all geographic
areas cou Id conserve energy by lowering the thermo-
stat setpoint for heating and raising it for cooling,
and by specifying high-performance heating and
cooling equipment.
Technology Assessment of Residential Energy Conserva-
tion Innovations
Our national experience has taught us that the side
effects or secondary impacts of a new technology's im-,
plementation frequently are unexpected and adverse.
The objective of this study (ref. 3) was to examine the
desirability to the Nation and to the consumer of
53
-------
selected tech;lical innovations intended to reduce energy
consumption in the residential sector.
The inncvations studied were chosen at the outset
of the study For single-family housing, they include
storm windows and storm doors, furnace energy re-
covery devices, and open-air-cycle air conditioning. For
multifamily ~ ousing, innovations studied were ventila-
tion energr recovery systems and double-glazed
windows. Th ~se particular innovations were selected
from those whose technical performance has been
studied as part of a continuing program on energy con-
servation reSf arch sponsored by the. Department of
Housing and Jrban Development, in conjunction with
the National ~cience Foundation and the Environmental
Protection Agl!llcy.
This techr!ology assessment had three objectives:
1. To quanti':y the energy, economic, materials, and air
emissions impacts of the selected single-family and
mu Itifamily modifications intended to reduce
energy cOlIsumtpion;
2. To deternine the cost effectiveness of the con-
sumer's in'Jestment in the selected innovations;
3. To examine and evaluate government policy options
capable of promoting mass implementation of those
innovations found desirable from a national view-
point.
An underlving objective was to develop the capabili-
ty to perform "technology assessments in the area of resi-
dential energy ~onservation. Other very important modi-
fications, both technical and operational, remain to be
addressed in this type of study, including ceiling and
wall insulation, reduced window area, heat pumps, and
various solar sy:;tems.
The performance of this study followed seven con-
ceputal steps (rd. 3):
1. To establish the energy conserving modifications to
be studied and the impact areas of principal con-
cern. The mpact areas include material resources,
energy reS(IUrCeS, environmental impact, and eco-
nomic impsct.
2. To develop a baseline against which results in the
four impac::ed areas may be measured. This involves
a detailed characterization of each of the modifica-
tions studied, including determination of the
materials and energy consumed in their manu-
facture, eVilluation of their current use ("satura-
tion") in various regions of the country, estimation
of the ene'gy to be saved per dwelling unit by
region, deturmination of the cost and manpower
involved in their manufacture, determination of the
annual dolla savings to the consumer resulting from
the energy saved. and examination of existing policy
and institutional structure relevant to the study.
3. To evaluate Federal, State. and local policy options
for promoting the installation of energy-conserving
modifications. This involves a study of their ease of
adoption (including precedents in previous legisla-
tion) and their effectiveness in promoting the modi-
fications, Le., how they influence the rate of intro-
duction of energy-saving modifications. It also calls
for testing the effect of policy options on the cost
effectiveness of typical innovation purchases.
4. To adapt an existing computer program (developed
by Hittman Associates, Inc., under a previous tech-
nology assessment program for the National Science
Foundation) to perform the detailed calculations
and bookeeping operations in each of the four im-
pacted areas.
5. To develop plausible case studies based on rates of
introduction consistent with feasible limits of in-
novation applicability, industrial production capaci-
ty. and the minimum elapsed time before change
could occur as a result of this study.
6. To determine the magnitude of potential impact in
each of the four areas and evaluate the conse-
quences of those impacts.
7. To recommend action by the Federal Government
to promote implementation of those innovations
found desirable from a national viewpoint.
The life cycle impacts of the innovations included
case study evaluations of energy, economic, air
emissions, and materials impacts. Evaluations are sup-
ported by quantitative examples drawn from 33 com-
puterized case studies.
The cost effectiveness of typical innovation pur-
chases is determined by cash flow analysis. It includes
initial cost, sales tax, finance charges. installments, fuel
bill savings or increased rent, maintenance and operating
costs, property taxes. and income tax advantages in
determining the cash payback period and total net cash
return of innovations as investments in various regions of
the country. Selected government policy options were
tested in the study for their effect on these cost-
effectiveness parameters.
The major conclusions reached during this study are
as follows:
1. I n creased implementation of storm doors and
windows on new and existing homes can be
achieved without serious negative impacts.
2. The furnace energy recovery device is potentially as
conservation effective and cost effective as storm
doors and windows.
3. Open-air-cycle air conditioning can save very little
energy. and does not appear cost effective.
4. Ventilation energy recovery can save very little
ene~gy and is limited in application.
54
-------
I ncreased new multifamily implementation of
double glazing can achieve modest energy savings
with no adverse life cycle impact and is cost effec-
tive in selected areas.
Residential Energy Conservation-Detailed Geographic
Analysis
The overall objective of this study is to quantify the
impact of selected single-family and multifamily resi-
dence modifications on residential consumption for dif-
ferent geographical areas. This study, for HUD, has just
recently started and will identify the energy consump-
tion associated with characteristic residences for 10
geographic areas of the United States and also identify
the energy savings associated with structural modifica-
tions. The 10 selected areas of the United States that
will be used for this study are:
Minneapolis, Minnesota
Chicago, Illinois
Boston, Massachusetts
Denver, Colorado
St. Louis, Missouri
San Francisco, California
Atlanta, Georgia
Los Angeles, California
Houston, Te.xas
Miami, Florida.
The overall study is very similar to the study described
earlier for the Baltimore/Washington area, with one
exception; the primary objective of this study is to iden-
tify the differences in energy consumption attributed to
location and climatic differences. The study will result in
reports being prepared for each of the areas being
studied as well as an overall summary report.
5.
ENERGY CONSERVATION PROGRAMS
IN THE COMMERCIAL SECTOR
This portion of the paper will address an energy
conservation program recently completed by Hittman
Associates in the commercial sector for the Federal
Energy Administration.
Factors Related to Energy Use in the Commercial Sector
The objective of this study (ref. 4) can be divided
into the following four areas of investigation:
1. To quantify the commercial space inventory in the
Central Business Districts (CBD's) of two major U.S.
cities (Baltimore and Denver) and three major sub-
urban shopping malls. This quantification was to in-
clude an assessment of the aggregate amount of
commercial space in each location, as well as an
'assessment of the physical characteristics of the
commercial space in each location (such as height,
age, equipment type, etc.).
2. To investigate and identify those institutions and
institutional practices which led to the current rates.
of energy consumption, as well as those which stood
as constraints or impediments to energy conserva-
tion through retrofit in each location.
3. To gather energy consumption data for a subsample
of the Baltimore CBD buildings, and to use these
data to:
Define average empirical energy use rates for
sets of buildings with different commercial
uses, and establish the range of energy use rates
shown for each commercial use type.
Investigate whether or not empirical relation-
ships could be shown between building physical
characteristics and energy consumption rates.
4. To develop and analyze a set of possible government
policy options which are capable of inducing
retrofit for energy conservation in the commercial
sector.
The overall study identified a number of observed
trends in the Baltimore and Denver CBD's, some of
which may be generalized to other CBD's and are identi-
fied here for each of the four major parts of the study
(ref. 4).
Physical Characteristics
Twenty-six hundred buildings were surveyed in
Baltimore and Denver and resulted in the following ob-
servations:
1. Office activity is the overwhelmingly predominant
use of commercial building space in both CBD areas.
For the two cities studied, the office space was
above five times as large as the next use (see figure
1) .
2. Small stores and department stores are the most pre-
dominant uses of retail space. In each of the citil's
studied, these two uses together comprised approxi-
mately 70 percent of the CBD retail space.
3. The trend for the future of CBD areas seems to be
toward greater util ization of space for nonretail
activities. From observations in both Denver and
Baltimore, only the nonretail uses have experienced
new CBD construction activity after 1961. Retail
space construction has been in the outlying subur-
ban areas (such as the building of enclosed shopping'
malls like the Villa Italia Mall in Denver).
4. Growth in CBD retail space virtually stopped about
the year 1950.
5. Larger, newer commercial establishments tend to
use distributed hot air for heating. Smaller, older
55
-------
~ BALTIMORE
15
14.IM DENVER
14
LEGEND
13
(/) 12
2
0
..J II
..J
,~
2 10
~
L&J 8
w
LI..
L&J
Q: 7
-------
6.
establishments tend to use radiant steam or radiant
hot water.
Fuel used for space heating is dependent on local
energy market conditions.
Offices and hotels, which are predominant in
Denver (see figure 1), are normally free-standing
structures, a.nd have the highest exterior glass ratios
of all the commercial and retail establishments when
expressed as a fraction of total floor space.
Baltimore CBD data showed that the glass fractions
(with respect to floor space) for offices built after
1960 were approximately 10 percent higher than
for those bu ilt before 1960.
7.
8.
Institutional Factors
Building codes, building design and construction
practices, and building regulatory practices have led to
the current energy consumptive trends in the com-
mercial sector. Some of these practices can be identified
as follows:
Building Codes
. Promulgated to protect citizens from fire
hazards and structural failures from poor work-
manship;
Designed with emphasis placed on reliability
resulting in large safety factors and general
overdesign in sizing building components;
Written such that lighting and ventilation levels
were specified to protect public health and
safety-these must be reassessed to determine
impacts on building energy use.
Building Trade Practices
Extensive use of exterior glass contributes to
heating and cooling system loads;
Buildings are designed for minimum first cost;
Life-cycle operating and maintenance cost for
equipment and systems are not specified;
Pe rfor ma n ce and re I iability servicing of
equipment after installation is not the engi-
neers'responsibility.
Practices of Regulatory Bodies
Utility rate structures are regressive and have
helped to encourage excessive energy use.
Master metering has been installed in many
build ings that were previously individually
metered.
Energy Consumption
Energy consumption data were collected for 383
buildings in the Baltimore CBD during the 12-month
period of May 1974 through April 1975.
The mean annual energy use rates (expressed in
Btu/ft2/yr) found for the 12 commercial use types are
given in table 1. along with the number of establish-
ments and number of square feet of commercial space
upon which each energy use rate is based.
In addition to the energy use rates, the following
conclusions were drawn from the analysis of Baltimore
CBD energy use data:
1. From an energy use standpoint, commercial build.
ings cannot be considered as a monolithic whole.
There exists substantial variation along numerous
lines in buildings' energy use rate.
2. Steam cooling results in significantly more in-struc-
ture cooling energy use than electrical cooling.
Other factors, such as utility system load factors
and building owners' economic concerns mitigate
this result.
3. Baltimore CBD energy consumption rates shower!
that energy use rate was dependent upon building
age, with older buildings having uniformly lower
energy use rate values than the average for their use
category. It was assumed that age represented a sur-
rogate for such variables as materials used, design
methods employed, and construction practices.
Policy Options
Eight policy options and their potential impacts
with regard to energy savings, costs to the building
owners, and costs to the government were discussed in
the Hittman study for FEA (ref. 4). These eight policy
options are listed below to identify the direction the
results of this study took:
1. Tax Credit Policy Option-Enact legislation granting
a 15 percent tax credit to building owners who in-
stall qualifying retrofit equipment.
2. A ccelerated Depreciation Policy Option-Enact
legislation permitting a 3-year depreciation schedule
for qualified retrofit equipment.
3. Excise Tax Policy Option-Enact legislation creating
an energy excise tax with an increasing tax rate for
consumption above a decreasing standard.
4. Small Business Loan Policy Option-Expand th.~
loan program operated by the small business admin.
istration to include special loan funds for retrofit of
small business establishments.
5. Disclosure Policy Option-Enact legislation re-
quiring a building owner to disclose the previous
5-year energy-use history of his building to any
potential buyers before the sale.
6. Public Information Policy Option-Undertake a
public information program to educate building
owners and operators regarding energy concerns.
7. Utility Rate Restructuring Policy Option-Enact
legislation instructing the Federal Power Commis-
sion to remove the regressiveness from their rate
structures, and encourage private utility companies
to do likewise.
57
-------
Table 1. Average EUla of Baltimore CBD commercial
building groups (ref. 4).
Number of Number of
Use Average establishments square feet
type EUI sampled sampled
Restaurants 300,000 25 70,991
Night clubs 253,192 23 42,479
Drug stores 232,672 6 15,303
Food stores 206,986 5 5,704
Department stores 164,412 27 1,142,175
Hotels/motels 146,597 6 950,400
Banks 144,634 15 68,743
Offices 124,647 87 6,477,049
Personal services 117,318 26 46,299
Sma11:;tores 95,378 132 383,443
Theaters 75,844 2 51,608
Wareho'Jses 61,973 29 439,470
aEUL
Energy Use Index, expressed in Btu/gross square foot/year.
8. Direct Subsidy Policy Option-Enact legislation en.
abling building owners to receive a subsidy for 20
percent of the cost of qualified retrofit installations.
For a more detailed discussion of these policies and their
potential impilcts, the reader is directed to reference 4.
CONCLUSIONS
Energy c(lp.sumption in the consuming sectors still
needs to be curbed and reduced to levels deemed ap-
propriate by Federal officials and many of the experts in
the field of ellergy utilization. Programs such as those
briefly discussl:d in th is paper help to meet and identify
many of the causes for our continued inefficient use of
energy. It should be noted that there are a number of
other studies, 100 numerous to mention here, that have
made significant contributions to conserving energy in
the consuming sectors. One study in particular, entitled
"Guidelines for Saving Energy in Existing Buildings"
(ref. 5). was recently published by the FEA and will help
building owners, manager, operators, and occupants to
conserve and manage energy usage without having to
invest significant amounts of money to do so. Informa-
tion obtained from studies such as those discussed in this
paper should be used in providing input to energy policy
makers and planners. .
REFERENCES
1. Potential for Energy Conservation in the United
States: 1974-1978, A Report of the National Petro-
leum Council, September 10, 1974.
2. Residential Energy Conservation (A Summary
Report), prepared by H ittman Associates under con-
tract H-1654 for the Department of Housing and
Urban Development, HUD-HAI-8, July 1974.
3. Teelmology Assessment of Residential Energy Con-
servation Innovations, prepared by Hittman Associ-
ates under contract H-2132R for the Department of
Housing and Urban Development, May 1975.
4. Physical Characteristics, Energy Consumption, and
Related Institutional Factors in the Commercial
58
-------
Sector, prepared by H ittman Associates under con-
tract CO-04-51888-00 for the Federal Energy Ad-
ministration, October 1975.
5. Guidelines for Saving Energy in Existing Buildings,
Conservation Paper Number 20, Federal Energy Ad-
ministration, June 16, 1976.
59
-------
THE EFFECT OF ENERGY COMPONENTS
ON SYSTEM COSTS AND EFFICIENCY
T. J. Thomas, Ph.D. *
Abstract
It is the oremise of this paper that the cost and
efficiency of proposed energy delivery system compo-
nents must be evaluated within the context of the sys-
tem with whicl-j they are to be used, as the system ineffi-
ciencies can scale up the costs of the proposed compo-
nents. Specific'jl/y, there exists a number of technologies
by which the wlfur oxide emissions from the burning of
coal can be minimized. These technologies are attractive
on paper, but their LIse may greatly affect total system
cost and efficiency.
A method'Jlogy is presented by which total system
cost and efficimcy may be easily calculated. Individual
component contributions to total system cost and ineffi-
ciency are also developed by this methodology.
This methodology is applied to six example energy
delivery systen1s based upon the delivery of 1 MWe of
electricity (froln coal) to a midwest area. Each system
has some form Jf sulfur removal, and each represents the
most optimal s Istem of a set of systems using nearly the
same components in various combinations.
The resu/~ show clearly the effect of major system
components, including sulfur removal techniques, on
total system cost and efficiency. Potential areas for
improving systEm efficiency at little or no cost are delin-
eated.
INTRODUCTION
Environmental gains in the past decades have been
achieved in part through the substitution of clean fuels
for coal. Howel'er, the recent OPEC price hikes, arriving
with a depletic n of national resources of clean fuels,
have made deal fuels expensive, and even unavailable.
The need for a fuel which is relatively abundant. avail-
able, and inexpl!nsive is resulting once again in a growing
utilization of coal.
Many proc1!sses have been developed to minimize
the sulfur oxides emissions from the burning of coal. A
proper evaluation of the effects of these processes upon
the delivered ccst of energy and the delivery efficiency
requires that these processes be evaluated within the
context of their proposed usage. In short, what needs to
*Re~l';)rch Environmentalist, Battelle Columbus Labora-
tories, Columbus, Jhio.
be examined is the conservation consequences of pro-
posed environmental benefit systems.
Future users of coal-based fuels will have several
technologies available by which their emissions of sulfur
oxides can be decreased to acceptable levels. These
include:
1. cleanup of the stack gases;
2. utilization of low sulfur coal;
3. liquefaction of the coal, with sulfur removal;
and
4. gasification of the coal, with sulfur removal.
Each of these technologies introduces a cost and ineffi-
ciency to the energy delivery system to which it is
applied. What is presented in this paper is a consistent
methodology by which an energy delivery system can be
evaluated, and the results of the application of the
methodology to six energy delivery systems that utilize
the sulfur oxides control technologies.
THE COST MODEL t
The Cost Model is based on "technology modules"
and "energy delivery systems," A technology module is
a distinct step in mining, transportation, or processing of
coal to produce energy. An example of a technology
model, shown in figure 1, is the Hy-Gas, Steam-Iron
Process, which produces pipeline-quality gas of approxi-
mately 1,000 Btu per cubic foot, hi all cases, except coal
mining, a technology model is described by technical
and economic p(jrameters which reflect all activities and
costs associated with that technology module, In the
case of mining, the cost of the technology module is the
cost of the coal. In all other technology modules the
cost of the fuel is excluded from the estimated costs of
the technology module, as developed in the individual
technical studies. In all cases, the cost of operating a
technology module includes the cost incurred for the
disposal of all by products of the process, whether those
by products are normally considered useful or are nor-
mally considered as wastes. No credit was allowed for
potentially useful byproducts, but all costs for the dis-
posal of waste byproducts are included in the operating
costs.
tThis is based upon work performed with internal funds
through the Battelle Energv Program.
60
-------
Interior
Hi-Sulfur
Deep Mine
-I
Hygas Steam-
Iron Process
Pipeline
.. Gas
INPUT DATA
. Initial Capital Cost
. Annual Operating Cost
. Load Factor
. Efficiency
$171,290,000
$ 17,885,000
90%
57%
Figure 1. An example of a technology module.
A variety of technology modules were developed
. which fall into the following five groupings:
1. Underground mining-modules varying by seam
thickness.
2. Surface mining-modules varying by size of
mine and method of mining (area mining and
contour mining) as well as by region of the
country (Eastern, Interior Province, Western)
and type of coal mined (bituminous, subbitumi-
nous, lignite).
3. Chemical refining of coal-modules varying by
process (aqueous leaching, hydrothermal proc-
essing, solid refining) and by type of coal
processed.
4. Coal liquefaction-modules varying by process
(H-Coal, CONSOL, etc.) and by type of coal
processed.
5. Power plants-modules varying by general type
(power plants producing electricity as primary
output, power plants producing process heat
and/or steam as primary output); varying by
application (base load, midrange, or peaking);
and by type of coal or fuel form (coal as a
solid, coal as a liquid, residual, crude, distillate,
or turbine fuel, etc.).
With an inventory of technology modules, it is
possible to arrange series of compatible modules to
represent the means by which energy, based upon coal,
can be delivered to a "common point," and the overall
costs of such energy delivery systems can then be com-
puted. The total cost of delivered energy is thus a func-
tion of the costs and efficiencies associated with the
component technology modules. Figure 2 shows an
energy delivery system based upon gasification of coal.
With the exception of the costs of transportation of
energy*, all costs were developed from data produced by
the technical studies of the Battelle Energy Program.
It should be noted th;:It the term "cost" as used in
this context is intended to reflect all the normal costs
required to operate the particular energy delivery system
plus a reasonable profit sufficient to maintain an ongo-
ing industry. This cost is to be differentiated from the
actual "price" that may be paid by the purchaser. That
price can differ widely from cost because of the compet-
itive situation that may exist at a particular time and
place.
Description of the Cost Model
With the definitions and concepts outlined above, a
computerized cost model was formulated. This model
provides a consistent framework for the evaluation of
energy system costs and efficiency. The primary output
of the model is the cost of delivering, from a specified
energy delivery system, 1 million Btu of energy in a
specified form.
Mode/Inputs
As noted above, the primary inputs to the cost
model consist of the cost data and operating efficiencies
of each technology module based on 1973 costs. Be-
cause of a difference in the cost calculations, the model
distinguishes between two types of modules: (1) those
that require new investment for implementation, and (2)
*Unit train transportation of coal, pipeline transportation of
high-, intermediate-, or low-Btu gas, and transmission of electric
power.
61
-------
c:=]~
Clean
~G-
Transport
Convert to
Electricity
106
Btu
Figure 2. An example of an energy delivery system.
those that require no new investment for implementa-
tion. An example of a module of the first type is a
high-Btu coal !Iasification plant, while a module of the
second type might be an existing pipeline for shipping
the gas.
For modules requiring no new investment, the cost
consists of an operating charge and a capital charge. The
capital charge includes depreciation, interest, corporate
income taxes, ilnd profit sufficient to maintain an indus-
try in a viable condition. All cost figures are expressed in
dollars per mi lion Btu of energy output for a given
module. In me dules that require new investments, the
capital charge is computed from the initial investment,
the expected I fe of a new facility, and a specified cost
of capital. In the capital charge computation, an annual
cost stream is calculated which, when discounted over
the life of the hcility, equals the initial investment.
The operating efficiency of each module is the
second primar~' input. This information must be pro-
vided for all m Jdules. Efficiency for the purpose of this
model is defimd as the ratio of the Btu output from the
module to the Btu input to the module. This definition
of efficiency i:; applicable to all phases of the energy
del ivery system, I n the case of transportation, the energy
required to perform the transportation is not included.
Transportation Modules. In the evaluation of energy
delivery syster,1S, three transportation modules were
utilized:
1. rail transportation of coal,
2. transrrission of electrical energy by 500- kV
power lines, and
3. pipeline transmission of natural and synthetic
gas.
Rail transportation costs are based upon the equa-
tion:
Te: = 0.5564 + 9.380 x 10-3 0
- 6.266 x 10-602
+ 1.542 x 10-903
where
TC = transportation cost in $/ton
o = distar ce in miles.
The equation is based upon freight tariff data com-
piled by the Peabody Coal Corporation. Based upon cost
data for 1971, 73 percent of total cost was assigned to
operating costs and 27 percent to capital charges.
Costs from gas pipeline transmission are taken from
H. C. Hottel and J. B. Howard (ref. 1). Specifically, a
transportation cost of 1.75 cents/million Btu per 100
miles was used. Of this total cost, 22 percent was as-
signed to operating cost and 78 percent to capital
charges. The cost of transmitting electrical energy was
obtained from Battelle data. The total cost of 12.3
cents/million Btu per 100 miles is slightly higher than
the corresponding cost given in Hottell and Howard. Of
the total cost, 44.5 percent was assigned to operating
cost and the remainder to capital costs. An overall effi-
ciency of 96 percent was used for electrical transmission
lines and 100 percent for pipeline and rail transporta-
tion.
Computational Scheme
From the above information, a total system cost is
computed. This is accomplished as follows. Let there be
n modules in a system and let Ci and Ei denote the cost
and the efficiency of the ith module. The cost incurred
in the (n-1)st module is Cn-1/En, since 1/En million
Btu of output at stage n-1 is required to produce 1
million Btu of final output. The total system cost is thus
obtained by summing the cost contributions of each
module, so that
n
C=2:
i=1
C.
I
n
IT
E.
J
j=i+1
where
C = total system cost expressed in dollars per million
Btu
Ci= cost of the jth module expressed in dollars per
million Btu
Ei= efficiency of the ith module
n= number of modules in the energy delivery system.
62
-------
Cost-Influence Coefficients. A cost-influence coeffi-
cient can be defined for each module in an energy deliv-
ery system. The cost-influence coefficient for the ith
module represents the change in total cost produced by
unit change in the ith module cost. For example if the
cost-influence coefficient for the unit train transporta-
tion module is 3.0, an increase of 1 cent in the module
cost will produce a 3-cent increase in the system cost.
The cost-influence coefficient may also be inter-
preted as the amount of energy output (expressed in
million Btu) required by a given module to produce 1
million Btu of delivered energy. Thus in the unit train
example, 3 million Btu have to be transported to pro-
duce 1 million Btu of delivered energy. It should be
noted that the reciprocal of the cost-influence coeffi-
cient of the first module yields the overall system effi-
ciency.
Efficiency-Influence Coefficient. The efficiency-
influence coefficient quantifies the sensitivity of the
total system cost to changes in the module efficiencies.
Thus the influence coefficient for the ith module yields
the effect of a 1 percentage point change in module
efficiency on the total system cost. The influence coeffi-
cient is always negative since an increase in the effici-
ency produces a decrease in the system cost.
In using the influence coefficients described above,
it should be remembered that only one parameter is
varied at one time. In the case of the cost-influence
coefficient, all module efficiencies are assumed to re-
main at the levels used for the coefficient calculations.
For the efficiency-influence coefficients, all cost and
efficiencies other than the module efficiency whose
effect is to be tested are assumed to remain constant.
APPLICATION OF COST MODEL
Through internal studies at Battelle, a number of
energy delivery systems were defined which utilized a
1,000-MWe power plant and some means of sulfur oxide
control. The most optimal systems investigated are
described below in table 1.
Basically, the systems compare the utilization of
central high sulfur coal with stack gas cleanup to low-
and high-Btu coal gasification, coal liquefaction, and
Western coal. Specifically, the prominent features of the
systems include:
1. Minemouth-the coal is' burned at the mine,
and electricity is transmitted 250 miles to the
demand center-
2. Western-strip mined low-sulfur coal is shipped
1,140 miles by unit train to the demand center,
where it is converted to electricity.
3. Low-Btu-a moving bed dry ash gasifier con-
verts the coal to a clean gas which is piped 250
miles to the demand center. The power plant
. must be converted to burn the low-Btu gas.
I - - -- u~ :0.- -' :...-:.....:
Table 1. Optimal energy del ivery systems
Identifier
- -..-. -- .-- .0.'-- ---"':' - -~ .;.:.~=..:...~.o.....-=----,-....;...:~.:_;:,==_""::,~:_-;,::,,,~-,;,~=-=,~:--:"'-~==-~'::..;:,=~"=~~'~--:_..._n_- - .;-.~=-
Processes
- .-. --- -- .-.-----.--- -. _._.~ - - - ..--- .-.-- -- '. ----.-.- ..------...--. -.----.---.. -.---.---------------.---.-.-------------
Minemouth Central Coal preparation Combustion Stackgas Electrical
high-sulfur 1,000 "1We cleanulJ transmission
coal 1,140 miles
Western Wes teru Coal preparation Unit train Comuustion
low-sulfur 1,140 miles 1 ,000 M\~e
coal
Low-IHu Central lIigh- Coal preparation Moving-ued Pipcline Cumhlls t i on
sulfur coal Ory-ash trall~mission (retrof i I.)
(;asifier 1,1110 mi les
Iligh-lItu Cen tra 1 high- Coal preparation lIydrane Pipclinc Comuusl. ion
sulfur coal Gasifier transmission comuined cycle
1,140 miles
H-C combined Central high- Coal preparation Unit train II-coal process Combustion
cycle su lfur coa 1 250 miles future C-C
II-C Centra 1 high- Coal IJreparation Unit train II-coal Combustion
sulfur coal 250 miles pr'ocess 1,000 MWe
-..--.- ---- -.-..- '..+. --+-_. .--.--.----..- --_...-- +.- ----- -+---- --.----------------.-------.-------'.-----
63
-------
Minemouth
.
Low Btu
.
High Btu
.
4. High.Btu-a Hydrane gasifier converts coal to a
fuel dcceptable for use by current combustion
equi~'ment. A pipeline ships the gas to a
comtined-cycle power plant at the demand
cente r.
5. H-C-the coal is shipped by unit train to an
H-Co.JI thermal coal processor. The resultant
clean fuel is burned at the demand center.
6. H-C Combined Cycle-the coal is shipped by
unit train liquefaction process. The resultant
clean fuel is burned at the demand center in an
adval1ce-combined-cycle power plant.
The elements described above were combined
together with ':he previously described costing methodol-
ogy. The resultant cost and efficiency of each delivery
system are portrayed in figure 3.
In figure 3, it is seen that the Western coal system is
the most effic,ent and least expensive. This is due to the
relative simplil:ity of the entire system, and the fact that
Western coal is strip mined, and thus is less expensive
than the deep'mined central coal. The principal reason
0.4
Weste rn
.
>-
u
c
Q.)
u
0.3
-
-
W
0.2
I
3
I
4
...-./Iv
that the combined cycle H-C system is cheaper t~an the
other H.C system is that the efficiency of the H.C
combined-cycle system is higher, thus decreasing fuel
costs.
This can be seen more clearly in figure 4, which
displays the contribution of each module to total system
efficiency. It is clearly evident from this figure that com-
bustion processes are the largest contributor to system
inefficiencies, while coal liquefaction or gasification also
have large contributions.
The coal processing elements within these systems
have each been touted as a means of converting coal into
an acceptable fuel at a low cost. However, the actual
cost of using these elements is inflated by system ineffi-
ciencies. Figure 5 demonstrates the actual, in-place
module costs for the various systems. It is seen that the
coal processing elements contribute significant costs to
the total .system costs.
The effect of component cost and efficiency upon
total system cost has been clearly delineated in figures 4
and 5. It is seen that the coal processing modules have a
H - C combined cycle
.
H-C
.
I
5
Dollars/Million Btu
I
6
I
7
Figure 3. System efficiency versus cost for six systems.
64
-------
0.2
Key-
0.3 Power plant
Coal preparation
>- Coal gasification
u 0.4
c Coal liquefaction
Q)
u Stack gas cleanup
'+-
-
w 0.5 Electrical transmission
0.6
0.7
0.8
0.9
1.0
M inemouth
Weste rn
Figure 4. Contributions of system components to system inefficiency.
large effect on both the system cost and efficiency. If
these processing techniques are to be a competitive alter-
native to Western coal, then their efficiencies must be
increased and/or their costs decreased.
To examine this further, the cost and efficiency
influence coefficients defined earlier can be combined to
yield the slope of a component efficiency versus cost
curve for which total system cost would remain un-
changed. In fact, an analytical curve can be developed
which defines the whole curve.
Two such curves are presented in figure 6. This
figure illustrates lines of constant system cost for the
coal preparation and combustion modules of the energy
del ivery system for Western coal.
Curves representing the actual cost of combustion
and coal processing modules can be imagined super-
imposed on figure 6. These curves would be concave
downward. Their intersections with the appropriate con-
stant cost curves yield maximal efficiencies for identical
system costs. As a result, a constant-cost curve which is
relatively level has a higher maximal efficiency than a
steeper constant-cost curve. This feature is a result of the
relative importance of fuel cost, and hence the unimpor-
tance of other costs, in inefficient systems.
65
-------
7
Key
6 ~ Power plont
. Coo I preparation
:J 5
- Coal gasification. lique-
£D faction, or stack gas
c . clean up
o 4 = Electrical transmission
~ Mining
'- 3
1/1 Unit train or pipeline
L...
0
0 2
0
o
High Btu
Low B tu
Western
Mine mouth
H-C
H - C combined cycle
Figure 5. Contributions of system components to system cost.
Figure 7 demonstrates the constant-cost curves for
all of the combustion modules examined. These curves
are roughly ar"anged by order of system efficiency - The
most striking feature of this figure is the relative cost
which can be incurred by the high-Btu system in order
to improve thn efficiency of the total system. Converse-
Iy, it would appear that increasing the efficiency of
Minemouth alld Western coal-burning modules could
only result in an increased system cost (a fact which is
borne out by 1 he lowered efficiencies of modern power
plants).
A similar set of curves is presented in figure 8 for
the coal proc ~ssing modules representing liquefaction
and gasification. Here, both the gasification modules dis-
playa relative cost insensitivity. However, here the gain
in efficiency to be achieved is minimal, as the gasifica-
tion processes are relatively efficient.
CONCLUSIONS
A systematic methodology for evaluating the cost
and inefficiency contributions of energy delivery system
components was developed and applied to six promising
systems for the conversion of coal to electricity designed
to minimize sulfur oxide emissions. The results of the
analysis demonstrated that:
1. The cost and inefficiency contributions of coal-
processing modules are multiplied by other
system inefficiencies, and thus have a signifi-
cant effect on total system cost and efficiency.
66
-------
>-
u
c
Q)
.~ 0.5-
.....
.....
W
WESTERN COAL SYSTEM
1.0
:>.
u
c
Q)
u
0.5
-
-
W
Not l': Thps(' ('II rv('s do
110 t imp Iy 'tlw t (' r r i l' i ('11(' it'S
of IOO/:'. ;tn' tt'('llIli(';llly
ft';\sihlL'.
0.00
100
Percent of Nominal Cost
200
Figure 6. Constant cost curves for coal processing and
power plant modules.
1.0
Western..
N'III': 'I'll"::,, t'III'V"~i jlll'St'ut
t.'I',HIOIIIII' "1 I (,' I vile it':' wh It'll
:In' unt (t','IIIII,';1I Iy h':I~;(bll"
(.w
.. (.'1
. {'au
(1\ \) \
C (.0
"',
.-,:--::::
--
--
--
0.00
300
100
200
Percent of Nominal Cost
Figure 7. Constant cost curves for combustion modules.
67
-------
1.0
H-C and H-C combined
cycle
>-
u
c
.~
u
-
-
IJ.J
0.5
Note: Efficiencies approaching
100% ~re not technically achiev-
ahle.
0°0
100
Percent of Nominal Cost
200
Figure 8. Constant cost curves for coal treatment modules.
2. In opposition to current designs, the combus-
tion facilities for liquefied coal could be de-
signed for higher efficiencies with no increase
in total delivered cost. The resulting increase in
efficie'ncy would greatly affect total system
efficie'ncy.
Furthermore, it should be pointed out that the
superiority of 1 he Western coal system, both in cost and
efficiency, is indicative of an increasing pressure for
development of strip mining of Western coal.
REFERENCE
1. New Energy Technology - Some Facts and Assess-
ments, The MIT Press. Cambridge, Massachusetts.
1971.
68
-------
ERDA's CONSERVATION OBJECTIVES
IN THE TRANSPORTATION SECTOR
John J. Brogan*
Abstract
My presentation is intended to provide an overview
of the kinds of activities that we in ERDA have planned
for the Transportation Sector and to explain specifically
what we are doing now in order to meet two major
energy conservation objectives.
ENERGY CONSUMPTION IN THE SECTOR
I would like to begin by bringing out some facts
related to this energy-consuming sector, thereby placing
its problems in proper perspective. The Transportation
Sector consumes about one-fourth of the total energy
consumed by all sectors in the United States. The
energy-consuming elements of the sector can be divided
conveniently into two categories: the nonhighway trans-
port modes, such as air, water, pipeline, etc., which
consume about 25 percent of the total energy consumed
by the sector, and the surface land transport or the high-
way mode, covering autos, buses, and trucks. This high-
way mode consumes about three-fourths of the total
energy of the sector, with the auto alone consuming half
of the total of the sector.
This sector epitomizes the United States as a whole
when it comes to its reliance on selected energy sources.
Consider the fact that transportation is almost totally
dependent upon an energy source that is one of the least
abundant domestically produced sources in this country.
Transportation is currently 98 percent dependent on
petroleum as its energy source. This sector and the
manufacturing directly related to transportation were
the most visible casualties of the Arab oil embargo. The
impacts of the embargo have lingered with the auto
industry especially. Further, when we look closely at
this sector, we find that it is almost totally dependent on
liqu id fuels. Powdered coal is not practical for current
auto powerplants. So transportation is a relatively big
consumer of energy from perhaps our most critical
energy source. It is compatible primarily with liquid
fuels, and we do not see early replacement for currently
used gasoline. Putting these facts together could lead us
to the conclusion that the sector is in potential trouble
and that serious consequences are likely if the traditional
business-as-usual attitude continues to prevail in this
. Acting Director, Division of Transportation, Energy Con-
servation, Energy Research and Development Administration,
Washington, D.C.
country. I believe this to be a correct view of the current
situation.
THE ERDA ROLE
Now, just what does ERDA have in mind in at-
, tempting to solve these problems?
I believe it is wise to start off by explaining what we
see as our role. Our transportation function is part of the
Office of Conservation. As such, we see our role as that
of providing technical leadership for energy-motivated
research in all modes of transportation. Our work will
span the full range of transport modes, from highway
systems to air, water, rail, pipeline, and other cargo and
people movers.
The types of activities included will range from
studies and technical development to hardware develop-
ment. We will conduct social, economic, environmental,
and energy impact assessments. We will conduct studies
that focus on the factors, which influence acceptance
and commercialization of different operational-pro-
cedure changes and of new technologies. We will con-
duct critical assessments of institutional barriers that
render achieving energy consumption reductions in the
sector. We plan to evaluate and, where warranted,
develop new transportation concepts. We intend to
develop technology from the research phase through to
advanced development phase (using DOD definitions
here). and we plan to develop not only technologies but
to build the prototype hardware necessary to demon-
strate adequately the technologies to the public.
I recognize that there are some individuals who
would prefer that we in ERDA keep a low profile when
it comes to carrying out hardware development, that WI!
should concentrate on carrying out only long-term basic
research in materials, combustion, etc., and that it
should be left to industry to pick and choose any tech-
nologies to be developed further. We do not see this as a
viable approach ~nder today's conditions. Rather, we
have embarked on an aggressive program in developing
big technologies in partnership with industry, and we
fully intend to pursue with industry-hardware develop-
ment and system demonstration, if the particular work is
cost-effective and if it offers the potential of early
commercialization.
Finally, at the present time we see the conservation
role in transportation as end-use oriented, because we
69
-------
believe that the biggest payoff is there. Also, other
organizational elements within the Office of Conserva-
tion are hard at work on reducing consumption in
industrial processes associated with manufacturing in
industry. I n addition, at present other organizational
elements within ERDA are hard at work reducing energy
waste in prod Jcing synthetic liquid fuels from fossil
energy sources which someday may appear as the pri-
mary energy so Jrce for the sector.
We recogn.ze that this perception of conservation in
transportation could change in the future. Our concern
here is that proposed changes made in the transportation
system eventucilly have to be traced back to their net
energy savings, and under certain circumstances the end-
use consumption could be reduced. But to accomplish
this the net change could be an increase. This comes
right back to thl: pervasiveness of conservation, and it
will take time to better understand the nature of con-
servation.
PROGRAM OBJECTIVE
The Transportation Program focuses on both near-
and midterm, with midterm covering 1985 to 2000.
There are two major objectives in the program, and each
can be traced back to the alarming facts cited earlier
regarding ener!ly consumption by the sector. These
objectives are: (1) to reduce energy consumption in the
earliest possibl ~ time consistent with environmental
constraints, and (2) to develop, demonstrate, and
promote commercialization of operational procedures
and technologit,s that will eliminate the dependence of
the Transportat on Sector on petroleum energy.
While we vlOuld like to make the major impact of
the program felt before 1985, we foresee our biggest
potential oppo'tunities coming after that date. This
means that we Must be astute enough to work at the
same time on achieving both objectives.
AC~IIEVING THE OBJECTIVES
Achievement of the first objective suggests that
ERDA consider known, but either unused or' partially
used, operatioral improvements and technologies, as
well as developi1g new ones: I n the area of known tech-
nologies, etc., we recognize that we may be plowing
ground that ha~ been plowed before by other agencies.
In this event, II\e propose to join forces. In promoting
commercialization of either known or new technologies,
we also proposu to join forces with other government
agencies. The point that I make here is that the ERDA
conservation function in transportation intends to
become involved, whether it be with known-technology
or new-technology development.
When it comes to near-term impact, we are con-
cerned about the 110 million autos on the road today. A
relatively small 1 mpg increase in average fuel energy
could result in 300,000 barrels-per-day reduction in
imported oil. This enormous leverage potential lies
primarily in the auto and secondarily in the long-haul
truck segment of the sector. We want to understand why
past attempts failed to convince the public that sched-
uled tune-ups save the public money. We want to know
whether i~ would make ~ood sense for the government
to pay a partial cost for a tune-up. We may be surprised
when we look at the areas. We want to know what
previously known and proven traffic-light control tech-
niques to improve traffic flow are practiced in few cities
in this country. We want to know why previously known
and proven aerodynamic drag-reduction devices on
trucks are not being used by truck owners. We want to
respond to many of the old idea files that have been
discarded for years because perhaps the economics of
their implementation simply were not right when con-
ceived. We feel that we can serve a useful role as a cata-
lyst in this area, and we are letting the public, universi-
ties, and industry, large and small, know that we are
interested in their suggestions for ways to reduce energy
consumption in this sector.
The two program objectives are compatible. As we
see it, the reductions in consumption will appear in
gradual changes at first in fuel-consumption data, unless
another embargo hits us or unless government regulatory
forces appear. The introduction of new and energy-effi-
cient powerplants, such as the gas turbine and Sterling
cycle systems, if successfully developed by 1990, could
have a pronounced impact in consumption in the sector.
Work is underway on these developments. The multifuel
capability and very low exhaust emissions make these
systems especially attractive. I ntroduction of the bat-
tery-powered electric vehicle is another event that we are
working toward at this time, with increased emphasis on
system and component developments-especially the
battery. The use of nonpetroleum-base fuels in the
sector appears high in our program priorities. Particular
emphasis is being placed on methanol from coal and on
synthetic gasoline and distillate from coal and shale.
Hydrogen as a candidate fuel in the sector is also being
pursued.
While I have discussed mainly highway vehicles
because they consume three-fourths of the total sector
energy, we are proceeding to firm up our plans for meet-
ing both program objectives for the other transport
modes. We recognize that in some cases other agencies
70
-------
have been established in research roles. We are currently
pulling together data on all of the ongoing work in air,
rail, etc., and are carefully reviewing directions and
objectives of this work to determine whether further
opportunities exist for reductions in energy consump-
tions. At a minimum, ERDA intends to keep abreast of
all ongoing energy-related work in each transport mode,
even where E RDA research funds are not being used.
I believe that covers an overview of ERDA's Trans-
portation program. I point out that conservation is new
to us here in the United States. I am not aware of any
conservation energy experts in the transportation field,
mainly because virtually none of us in the past have had
the chance to practice being an expert in this area.
71
-------
72
-------
3 November 1975
Session. II:
NONTECHNOLOGICAL METHODS
TO CONSERVE ENERGY
David R. Berg*
Session Chairman
*Acting Director, Environmental Technology Assessment Branch, Office of Energy, Minerals, and Ind!Jstry,
Washington, D.C.
73
-------
74
-------
ENVIRONMENTAL IMPLICATIONS OF NONTECHNOLOGICAL METHODS
TO CONSERVE ENERGY
John H. Gibbons*
Abstract
U.S. policy to improve health, safety, and environ-
ment has resulted in higher energy prices, reflecting
more adequately its true cost. The OPEC cartel price
shift raised prices further. In response to higher prices,
energy users are shifting to less energy-intensive con-
sumption patterns and higher efficiency of use. The envi-
ronmental implications of these altered consumption
patterns are generally salubrious, because the actions
taken to save energy are much less polluting than using
the energy. Most other policies to induce additional
conservation will not only save consumer's money but
will further reduce environmental insults. There are
some proposed actions to save energy by relaxing envi-
ronmental controls. While a small amount of energy
could be saved in this way, it seems far more productive
to save energy through increased efficiency of use rather
than a retreat from environmental goals.
It is very difficult for a technologist, such as my-
self, to constrain his remarks to a discussion of "non-
technological methods." Since that is my assignment,
however, I will do my best to stay on the subject with
only two caveats: (1) I assume that it is legitimate to
discuss actions which may be technological but do not
require the development of new technology; (2) I
remind you that most policy decisions about resources
(including energy) engender technological responses
because technology is one of our most powerful means
to respond productively to socioeconomic perturbations.
Sometimes the best way to understand the present
condition and to anticipate the future is to examine the
past. On the other hand, the famed Satch Page cautioned
us in discussing his philosophy of his life to ". . . never
look back: something might be gaining on you...."
U.S. energy policy prior to the emergence of the
environmental movement in the 1960's was relatively
simple and straightforward. It consisted of a tech-
nological policy to develop nuclear energy and a non-
technological policy to (1) maintain minimum price for
energy resources; and (2) provide maximum availability
of energy resources. The doctrine of minimum price was
achieved at the expense of several very important things:
--
"Director, Environment Center, University o' Tennessee,
Knoxville, Tennessee.
Environment and Health. Throughout the
domestic scene, major environmental pollution and
human health costs were allowed to be unfairly borne by
air, water, land, and people. These "environmental exter-
nalities" amounted to non market costs of energy pro-
duction which were not associated with the same places
or people where the benefit or low-priced energy were
enjoyed. For example, we now payout of general tax
revenue about 1 billion dollars per year in black lung
benefits and will continue to pay similar bills for many
years to come. These very real costs of underground coal
mining were not included in the price of the coal so
obtained in past years. Thus, the benefits of "cheap"
coal were enjoyed by one generation but the costs were
levied on the next generation.
Domestic energy resources. The policy of
maximum use of energy resources through a variety of
schemes to subsidize its price caused energy resource
extraction to be accelerated to a rate that simply cannot
be maintained for more than a few decades. At the pres-
ent time, fossil fuel consumption of oil and gas amounts
annually to the equivalent of at least a hundred thou-
sand years of production. Domestic natural gas supplies.
despite the major finds in Alaska. have already peaked
out; oil production is peaking at the present time and
the costs of incremental additions to oil and gas reserves
are exponentiating rapidly.
Domestic energy resource independence. The
very high growth rate of demand began to overwhelm
domestic production capability of oil in the early
1960's. Rather than alter our energy consumption
policies or patterns, we turned to imports as a means to
continue the "energy feast." This amounted to a willful
relinquishment of national independence; the real co~t
did not occur to us until 1973.
The collective impact of these nontechnological
energy policies was a de facto subsidy of energy prices.
which caused demand growth rate to rise from a little
over 2 percent in the 1950's to nearly 5 percent in the
early 1970's, despite the fact that the thermal efficiency
of power plants and the energy efficiency of railroad
locomotives increased sharply from 1950 to 1970.
Energy efficiency in the consumption infrastructure
(e.g.. automobile mileage) fell steadily.
The advent of the environmental movement in the
latter half of the 1960's can now be seen as the first
major step in altering our energy policy, even though it
was not viewed that way at the time. The environmental
75
-------
movement arose in part from the (!xplosiv(! growth of
energy demanc and accompanying pollution problems
that resu Ited not only from cheap prices but also from
spreading population and the growth of affluence known
as the "throw-away society." The environmental move-
ment also was precipitated by regional environmental
overload caused by the implosion of human activities
(e.g., increasin!l unit size in power plants, increasing
urbanization) .
Environmental legislation such as the clean air and
water acts, the National Environmental Policy Act, the
Coal Mine Hedth and Safety Act, the Occupational
Safety and Hea th Act, and others reflected a realization
of major external costs in our economy, unfairly born in
society, and a national decision to internalize these
costs. In effect, we moved from a "pay tomorrow"
philosophy to a "pay iJS you go" philosophy. The ration-
ale is well illU':trated by a tale from "Peanuts." Lucy
exclaims to LinL!s, 'This generation has been given the
works! All the world's problems are being shoved at us."
Linus responds, "Well, what do you think we should do
about it?" Luc,.: ". . . we'll stick the next generation'"
It is not hard tn see that we have been sticking the next
generation for tnany generations. It is heartening to see
that we are be9inning to change that pattern. The envi-
ronmental legislation impacted energy production and
use especially liard because that is where most of the
environmental ~ roblems were located.
The impacts of the environmental/energy policy
decisions were many. In addition to direct impacts on air
and water quality, there were at least two important
consequences in the area of energy:
1. Energy price increased significantly. We should
recall that the a\:tual total cost of energy did not change,
but rather the external environmental costs of energy.
production andJse were reallocated to the consumers of
energy. The consequence of this impact was the initia-
tion of a move nn the part of energy ,consumers toward
greater efficienc'l of energy use.
2. The mi,.,imal environmental impacts of natural
gas production and use were such that users of other
forms of energy began to turn to natural gas. As a con-
sequence the demand for natural gas, already high be-
cause its price 1ad been held artificially low through
'price controls, cccelerated to even greater heights. Our
national store 0 f natural gas nose-dived in the wake of
explosive demand growth, and the ratio of gas reserves
to annual demar.d fell steadily. The consequence of this
response was that we enjoyed a rapid move toward im-
proved environmental quality, but only by borrowing
from the future.
The second major event in our changing national
energy policy did not arise indigenously like the environ-
mental movement, but rather was externally imposed
upon us. This was the oil embargo placed on shipments
to the United States by the Arab states in the Arab-
Israeli conflict of 1973-74. In retrospect that traumatic
event was important and productive shock therapy for
the Nation. Adlai Stevenson once observed that "man
does not seem to see the handwriting on the wall until
his back is up against it." The handwriting on the wall
about the hidden costs of becoming dependent on
foreign oil was made very clear through the embargo. It
drove home a realization of the finiteness of oil and gas
res'ources and also the extent of our growing dependence
on nondomestic supplies. Perhaps more importantly for
the short-run, it gave a real specter of the uncertainty of
availability of energy and triggered a rapid consumer
shift in purchase decisions. particularly in the area of
automobiles. The embargo crystallized a national con-
census about the need to develop domestic energy
resources and the imperative to improve the efficiency
with which we use energy in the United States.
The myth that greater energy consumption, per se,
necessarily implies greater productivity and economic
progress was beginning to tarnish; so too was the myth
that it is not practical to save a significant amount of
energy.
The third major event in our new national energy
policy, again imposed by external forces, was the OPEC
cartel price shift and subsequent ripple of higher prices
for other energy resources. We are in the process of
responding to this era of higher energy price, and so it is
difficult at this time to anticipate exactly what the total
response will be. Nevertheless, several things are quite
clear: Energy prices have become largely uncoupled
from current production cost. Rather than reflecting
average cost, prices now more closely reflect the incre-
mental cost of bringing in new energy supplies (e.g., oil
from under deep water off the eastern coast of the
United States). The higher price level (which is currently
our most basic energy conservation strat()gy) was a yift
from the OPEC countries and should be viewed as a very
important assist, albeit traumatic, to the United States in
adopting a more enlightened energy policy. The implica-
tions of the new era of higher price are many and
diverse:
1. Our energy supply system, primarily oil, gas
and coal, now has the opportunity to become more
diverse through the addition of other energy sources-
heretofore uneconomic (e.g., geothermal). As the rela-
tive price of energy increases, the type and number of
"new" energy sources that become economically attrac-
tive increases sharply. The fastest way to discourage
investment in a more diversified energy supply system
would be to roll back present prices to previous levels or
76
-------
even to hint that this might be done as a matter of
national policy!
2. The new energy price immediately makes it
economically attractive to alter the energy consumption
infrastructure to one of much higher efficiency. This can
be effected through direct substitution (e.g., better insul-
ation instead of bigger furnaces and air conditionersl.
increased efficiency of conversion (e.g., more ~fficient
air condit'oners and automobiles), and minor modifica-
tions in lifestyle (e.g., set back thermostats at night,
returnable beverage containers, more careful automobile
trip planning). In almost every instance, these kinds of
moves toward increased efficiency imply an increase in
domestic economic activity and improved environmental
quality, contrary to the rumors and myths which pro-
claim such moves would surely deflate the economy.
3. The new price level means that it is now
important to reevaluate the benefit/cost of various
public policies that were created during an era of cheap-
ening energy price. In the public sector this includes
such policies as freight rules, depletion allowance, power
plant cooling tower requirements, etc.,) I do not want to
imply that I feel that all of these policies should be
changed; energy price may have very little to do with
their basic validity. However, if public policy like envi-
ronmental control is based on benefit/cost calculus, then
the associated policies should be reexamined whenever
significant changes occur in either the benefit or cost
parameters.
4. Several years ago the price of coal was such that
the environmental externalities were a significant frac-
tion of its market price. (e.g., coal price at 7 dollars per
ton, environmental externalities estimated to be 3 dol-
lars per ton). Now that the price of coal has risen in
proportion to the new oil price, the cost of reclamation
and of other externalities becomes a very much smaller
fraction of total coal price and should therefore enable
decisions to internalize the external costs to be much
more palatable.
From an environmental point of view, the points we
make in the preceedil1!1 para!Jraph bring both good news
and bad news. First the good news: the substitution of
materials, time, labor, and human ingenuity for energy--
simply because its cheaper that way--is very salubrious
in terms of the environment and is also an economic
imperative. Likewise, the shift away from dependence
on fossil energy resources toward a time when we learn
how to meet our energy needs with sunlight and possibly
fusion is undergoing rapid acceleration, and none too
soon because it may take us a half-century or more! At
the same time there is bad news beyond the near-term
economic dislocations caused by such a sudden change
of events. Perhaps the greatest area of concern is the
implied massive shift toward increased coal production
and combustion in response to the change in price and
lessened assurance of availability of oil and gas. If we use
present technologies for mining, conversion, and com-
bustion of coal, the environmental implications seem
little short of disasterous.
Realistically, however, there seems to be little incen-
tive for users to shift from oil to coal because the rela-
tive price difference between these two energy com-
modities is not all that great. Furthermore, the chance
for the international oil cartel price to be undercut dur-
ing the next several years is small but finite. When one
couples that fact with the uncertainty about the long-
range U.S. public policy on energy, one can only surmise
that private investors will be very reluctant to commit
the very large amount of capital investment required to
significantly expand U.S. coal production in the near
term. It is possible that for a number of years we may
have much more oil available to us than coal.
Finally, we should remember that, while the "mar-
ket place" has a great deal of appeal in terms of the
allocation of goods and services, it cannot be automati-
cally depended upon to allocate costs and benefits in.
equitable fashion. One needs only a cursory glance at the
coal fields of Appalachia to understand the dramatic
reality of existing socially uneven impacts of costs and
benefits of domestic energy production.
Now that we have reviewed the environmental
implications of energy policy developments that have
occurred over the past several years, it seems appropriate
to consider several future policy options and their impli-
cations on environmental quality:
1. Rollback prices. Some members of the Con-
gress opt strongly for this approach, especially for oil
price. To be sure a cut in oil price would be a welcome
palative for the short run, but I maintain that it would
be disasterous in the more important long run. First, i'
would deter-if not reverse-the present move toward
more efficient use of energy in the United States. Sec-
ondly, it would exacerbate the problem of decreasing
production of domestic gas and oil. Third, it would deter
se verely the investment and interest 01 American
industry to diversify our energy resource base. In gen-
eral, therefore, a rollback of energy price or even a hint
that this might be seriously considered would probahly
be the most effective single thing we could do as a na-
tion "to stick the next generation."
2. Assure a "platform" price for energy; deregu-
late oil and gas prices. At least a portion of this policy
option is currently being seriously considered by the
Congress and the Administration. If such a policy were
developed with appropriate time phasing it could induce
very productive responses. First, the assurance of future
77
-------
prices that are level or increasing will induce investments
to make our energy consumption more efficient and
. would also induce lifestyle accommodations that are less
energy-intensive. Second, the assurancE;! of future price
would undoubtedly precipitate increased investment to
expand and diversify domestic energy supplies. To be
sure, there wo Jld be some hardships, but that situation
is not new. In the days when energy was made cheaper
by allowing e>ternal costs, it was the economically dis-
advantaged Ar1cricans who lived downwind from the
smokestack and in the coal fields that bore an unfair
portion of eneqy costs. Today, higher energy prices take
their heaviest toll among the middle class and disadvan.
taged, but we can ill afford to subsidize energy price for
all energy customers in the name of "helping the poor."
3. Public policies to induce shifts to increase ef.
ficiency of USI~. There are a variety of policies now
under consideriltion which could accelerate the rate and
extent to whict we become more energy efficient. These
range from extl emely minor pertubations of the market
system to here ic measures such as rationing. We sum.
marize these as follows:
(a) Labeling, education, exortation. Using these
techniques through Federal, State and local actions, we
could provide consumers with better market signals
which could mable them to make more energy-
conscious decisi Jm. We could rev-ain architects and engi-
neers to think G bout energy in the design of new build-
ings or in their renovation of existing buildings. We
could influence industry and the financial community to
produce consumer hardgoods and buildings that are
designed for minimum total life costs, which generally
implies much hi!lher energy efficiency than is the current
style. The environmental impact of such actions would
be generally very good. People would tend to dress for
the season and in general to live less energy.intensive
lives.
(b) Minimum performance criteria through regula-
tion and incentit'es. As in consumer safety it seems sen-
sible to insist .)11 minimum performance for energy.
consuming products (consistent with minimal total
cost). Therefore, it seems sensible to establish uniform
building codes :Ind standards both for new structures
and for the renovation of old structures. Similarly, it
might make good sense for minimum efficiencies to be
established for a I' conditioners since there are many on
the market which sacrifice energy efficiency for a slight-
ly lower purcha.e price that results in a much higher
lifetime operatin;l cost for the consumer. The economic
and environment31 impacts of such actions are very good
in that employnwnt would tend to increase, and less
energy would be required to sustain a given standard of
living. However, wch rules would limit consumer choices
and "freedom" of the market place. On the other hand,
it can be argued that such minimal performance rules or
tax incentives and penalties would provide an inherently
more equitable consumer market.
(c) Allocation and rationing. Some people feel
that equity can only be achieved by establishing policies
that bypass the price system. Therefore, the concept of
price-induced actions to decrease per-capita energy con-
sumption is seen to be regressive compared, for example,
to rationing. Such policies, if put into place, would cer.
tainly make the reality of the energy problem much
more convincing, but would create a bureaucratic night-
mare of cost and complexity that hardly seems justifi-
able.
(d) Relax and alter environmental and other regula-
tions. Some argue that, in the face of higher energy
price and the imperative to become less dependent on
foreign energy sources, we should make major changes in
existing regulations. Power plant emission controls are
accompanied by an energy efficiency penalty of several
percent. Automobile emission control carries an energy
penalty of up to 15 percent. Energy transportation rules
force energy-inefficient modes in the transport of people
and goods and subsidize the use of energy-intensive vir-
gin materials over recycle materials. The environmental
impacts of these various policy changes are mixed. In
some instances there would be strongly negative impacts
but just the reverse in other cases. Even in those cases
where environmental impacts seem bad there are some
productive possibilities under certain conditions. Con-
sider electric power plants: the increase of conversion
efficiency that would result from allowing more direct
cooling and less use of cooling towers implies a decrease
in the amount of required capital, as well as a decrease in
operating expenses. These "saved" dollars could be
plowed back into the economy in a more productive
way, for example in building new medical schools. Alter.
natively (and, unfortunately, more realistically) the
"saved" dollars could be used to reduce electricity prices
and consequently accelerate energy demand growth and
foster further decreases in efficiency of use!
CONCLUSIONS
1. The impacts of energy conservation on the envi-
ronment (including not only pollution and health but
also employment and amenity level) are generally very
positive but depend very much upon the specific strate.
gies employed. This means that in developing energy
conservation policy, like everything else, we should LISP.
our brains.
2. Even with full-cost pricing additional incentives
for energy conservation are justified because there are
78
-------
non'market, national needs to become more energy effi-
cient beyond the signals that we get from the market
place. These additional imperatives for energy conserva-
tion include the need to reduce oil imports and the
long-term need to conserve fossil energy resources for
nonfuel uses. This implies that we need uniform Federal
legislation: a surrogate for productive "mutual correc-
tion" through the application of collective national wis-
dom.
There are strong impediments to instituting energy
conservation programs. First come the myths about
energy conservation, which require time to dispel. Many
of these were purposely created by industries trying to
sell more energy or energy-consuming items. The Arab
embargo helped to reinforce the idea that conservation
was equivalent to "tightening the belt" or "freezing in
the dark." To the contrary, conservation means "wise
use"; it is simply the practice of sensible economic deci-
sionmaking.
Any change in our economic system, particularly
that arising from technological innovation, implies some
job displacement and sometimes decreased employment.
For example, the tractor displaced a lot of farmworkers
and horses. It also enabled us to eat better than before.
Flourescent lights closed a lot of incandescent lamp
production lines, but enabled Americans to enjoy light
less expensively. As we examine the specific strategies
for energy conservation through increased efficiency we
see, contrary to much popular opinion, that there is
some job displacement, but generally greater employ-
ment rather than less.
There is also the myth that energy conservation
somehow denies the poor their chance for the "good
life." It is difficult to see how such an idea makes sense
when conservation is designed to maximize he extent of
goods and services that can be provided by a given
amount of resources. I personally believe that it is only
through repeated encounter with the specific strategies
for energy conservation that we will dispel the counter-
productive myths about its undesirabil ity.
Finally, in responding to the energy problem
through alteration of demand patterns we encounter per-
ceptual difficulties that arise from the ways we've
thought about problem solving ever since the beginning
of the industrial revolution. It was through the industrial
revolution and the move to the western hemisphere that
we achieved so many material gains for mankind. We
lived in an infinite world in which riches were to be
found by "moving west" and by expanding production.
Ours came to be known as the "cowboy economy." In
the energy problem we encountered more sharply than
ever before the reality of the anachronism of these time-
honored ideas. Now the exponential, which until very
recently was equated with progress, becomes in many
instances a mortal enemy. We take every opportunity to
try to avoid thinking seriously about how we might
transform ourselves into a society that is in dynamic
equilibrium. In the decade of our 200th national birth-
day, we can no longer escape confrontation with these
thoughts. Hopefully we have the capacity and will to
make a productive response, and an excellent subject on
which to begin is energy!
79
-------
FEDERAL INITIATIVES TO SAVE ENERGY IN LIGHTING AND APPLIANCES
Kurt W. Riegel, Ph.D.*
Abstract
Energy sa,lings to the Nation and to individual con-
sumers can accrue from the following approaches:
1. Research and development on new technologies that
will lead to more efficient appliances, lighting
systems, and control systems that govern their use:
2. Improving the energy efficiency of new appliances
and lighting equipment by better incorporating
presently available technology.
3. Motivating purchasers of appliances and lighting
equipment to make selections based on energyeffi-
ciency and life-cycle cost thus creating market pres-
sures for more efficient manufactured products.
The Fee/mal Government is engaged in programs
that follow all of the above approaches. The Energy
Research and Cevelopment Administration has programs
for research, dwelopment, and demonstration of energy
efficient technology in appliances and lighting. The
Federal Energy Administration and the Department of
Commerce conduct a Voluntary Appliance Efficiency
Improvement Program, the purpose of which is to en-
courage implementation of presently available technol-
ogy in new an j more energy efficient appliances. The
Federal Energ}l Administration encourges wise use of
energy in lighting to conserve energy through two
programs--the Federal Energy Management Program and
the Lighting al1d Thermal Operations Program. Both
Federal operations and the private sector are the targets
of these efforts. Finally, consumer awareness and moti-
vational issues lelating to efficient energy utilization are
being. addressed in such programs as Commerce's Volun-
tary Appliance Labeling Program, FEA and ERDA
research projecl's, and by legislation calling for stronger
programs in all of the mentioned subject areas.
INTRODUCTION
I am happ'/ to speak today about current Federal
programs for el1ergy conservation in lighting and appli-
ances. This is so particularly because of the close rela-
tionship betwem energy use by lighting and appliances
and environmental quality. Thus, the subject seems to be
'Chief, Tech10logy and Appliances, Division of Buildings
and Industry, Energy Research and Development Administra-
tion, Washington, D.C.
in perfect harmony with the principal subject of this
symposium. With exceptions that are rather minor on a
national scale, appliances and lighting draw energy for
their operation principally in the form of electricity or
gas. Both forms of energy playa very important role in
the current debates over energy and environmental
policy, for quite different reasons. Natural gas is one of
the cleanest-burning fossil fuels available for onsite use
to operate appliances and major climite conditioning
systems. Gas is currently affected by supply and demand
factors that are troublesome to large segments of the
economy, Some believe that electricity has associated
with it some of the most worrisome and difficult envi-
ronmental problems facing the Nation. Thus we expect
that energy conservation programs that reduce or defer
the growth of natural gas consumption and electricity
usage will make a strongly positive contribution to the
preservation of environmental quality and to the devel-
opment of an orderly national energy policy.
To give some numbers, lighting is virtually entirely
electric in this country, except for some decorative
types, and presently accounts for about 5 to 6 percent
of total national energy consumption, or about 20 to 25
percent of total national electricity production. These
figures are made up mostly from direct lighting energy
consumption, but also contain a small excess contribu-
tion due to indirect demands for energy, particularly in
large commercial buildings, for climate-conditioning
requirements that are driven by internal building-lighting
loads.
Appliances will be defined here to include water
heaters, refrigerators and freezers, ranges, telev ision sets,
room air conditioners, clothes dryers and washers, and
dishwashers. Excluded from this discussion are major
climate-conditioning equipment such as central heating
and cooling equipment. Small home appliances account
for less than 10 percent of all appliances energy con-
sumption. Altogether, residential-type appliances ac.
count for about 8 percent of total national energy con-
sumption, a number which is large enough to consider as
a part of the total national energy conservation program.
Of this, over 70 percent is in the form of electricity, and
most of the rest is natural gas. Thus, previous n!marks
about the importance of electricity apply strongly to
appliances.
At this point it will be useful to break the discussion
into two separate parts, appliances and lighting. The
various Federal programs that operate in these two areas
will be described.
80
-------
APPLIANCES
The three Federal agencies with energy conservation
activities relating to appliances are the Energy Research
and Development Administration (ERDAI. the Federal
Energy Administration (FEA), and the Department of
Commerce through their National Bureau of Standards
(NBSI. It is apparent that large energy savings can be
achieved in three different ways:
1. Research and development on new technologies that
will lead to more efficient appliances, lighting sys-
tems, and control systems that govern their use;
2. Improving the energy efficiency of new appliances
and lighting equipment by better incorporating pres-
ently available technology;
3. Motivating purchasers of appliances and lighting
equipment to make selections based on energy effi-
ciency and life-cycle cost, thus creating market pres-
sures for more efficient manufactured products.
ERDA, which started its conservation program in
July of 1975, is currently working with a FY 1976
budget of approximately $12 million for all of its build-
ings' energy conservation activities this fiscal year. Of
this total, approximately $0.6 million will be obligated
for the support of research, development, and demon-
stration activities in appliances. I ncluded in the RD&D
agenda for appliances are projects to develop more fully
the data base on actual appliance operation and energy
consumption according to appliance technical attributes
and according to climate and other factors that are
important. Shortly after the first of the year, we will
host a conference on "Technical Opportunities for
Improved Appliance Efficiency:' which will convene
experts to begin the process of developing information
on appliances useful to ERDA in setting priorities for
support of RD&D in this area, and useful to the appli-
ance industry in stimulating change and innovation with
respect to conventional appliances. Furthermore, ERDA
intends to explore the potential for developing inte-
grated appliances which combine several conventionally
separate appliances to take advantage of waste heat and,
thus, reduce energy consumption. There are several cases
of appliances that have complementary functions, for
which such integration is expected to provide large
energy savings? there are promising applications for such
integrated appliances. For example, integrated refriger-
ator-water heater appliances might be introduced in
factory-built mobile home units, a portion of the
industry which is projected to continue to account for a
large fraction of new housing. Finally, as a result of
ERDA activities, technical opportunities for changes in
both residential and commercial appliance will be iden-
tified, and RD&D efforts will be supported so as to
encourage and stimulate the introduction of energy-
saving appliances. It is estimated that by 1980, rather
minor changes, well w.ithin the bounds of existing tech-
nology and practice, could result in cost-effectively
improved appliances with efficiences 28 percent greater
than the 1972 base case, averaged over all appliances,
energy and production weighted. (See appendix A,
"Technical Opportunities for Improving the Energy Effi-
ciency of the Nation's Appliances.")
Technical measures for improving the energy effi-
ciency of appliances include, as a sample, insulating
water heaters, ranges, and refrigerator/freezers more
effectively; rerlacement of the standing gas pilot lights
with intermittent ignition devices in water heaters,
ranges, clothes dryers, and other appliances; improve-
ments in heat-exchanger efficiency in refrigerator/
freezers, room air conditioners, heat recuperation; and
improvements in controls in refrigerator/freezers, room
air conditioners, and other appliances. For example, the
present day automatic defrost refrigerator goes through
i~s defrost cycle mindlessly on a regular repetitive basis
regardless of need; controlling the defrost cycle so that it
is actuated only when actually needed would give energy
savings.
ERDA, FEA, and NBS are all very interested in
motivating purchasers of appliances to consider energy
efficiency as an important factor in final selection. NBS
has a voluntary appliance-labeling program, which pro-
vides some energy information in the form of a label
affixed to appliances displayed on the showroom floor.
For the air conditioner, the information has been EER,
energy efficiency ratio, the ratio of heat pumped to elec-
tric power input. For other appliances, such as refriger-
ators and water heaters, the information displayed will
be in the form of operating cost. Present information
about consumer motivation indicates that annual cost of
operation is probably the simplest and most effective
way to communicate energy efficiency information to
the consumer. From a public policy standpoint, the
necessity to communicate effectively with the consumer
should be an overriding consideration, since mere tech-
nical accuracy unaccompanied by consumer understand-
ing will not effectively influence decisions toward effi-
cient, life-cycle cost-justified appliances. It is taken for
granted that the long-term effect of large numbers of
individual consumer decisions in favor of efficient
appliances is to stimulate the manufacture of appliances
with larger industrywide average energy efficiency than
without a labeling program.
There is some very interesting and promising legisla-
tion pending in the Congress, which is intended to ex-
tend the present NBS voluntary program and to result in
greater effectiveness with the consumer. The House and
81
-------
Senate have rI'cently reported out of conference a bill
that would require mandatory annual cost-of-operation
labeling for th ~ major energy-consuming appliances, and
would furthermore set an energy-efficiency improve-
ment target of 20 percent for newly manufactured appli-
ances. If the J\dministrator of FEA determines, after a
specified period of time has elapsed, that progress
toward attainment of the 20 percent goal is not satisfac-
tory, then a mandatory standards provision of the legis-
lation can be activated. The standards provision is a very
useful backup to the labeling and efficiency target provi-
sions, since there is very little information on what
progress can retlsonably be expected from them alone.
Finally, I should mention that the Department of
Commerce and FEA have a program for Voluntary
Appliance Efficiency I mprovement. Targets have been
established, and these two agencies have established a
timetable for tJbtaining a 20 percent improvement in
efficiency by : 980. The program was in response to a
call for the President in the 1975 State of the Union
address for such a voluntary program.
LIGHTING
There are pomising technical approaches that might
be taken towal d the development and implementation
of new lighting t::!chnology, both with respect to compo-
nents (light sources and fixtures) and controls. ERDA
will support th~ RD&D activities leading to more effi-
cient light sources through such projects as one to pro-
duce a fluorescnnt light bulb suitable for replacement of
the standard i.1candescant light bulb. Efficiency im-
provement of Dver a factor of three seem technically
feasible and would provide the first real innovation that
is capable of j:;enetrating the large residential incande-
s~ent market. Controls that are inexpensive enough to
achieve significant national usage, with respect to time
scheduling or fvailable natural light, would also offer
energy savings .lOd will be explored by E RDA. Finally,
the re lation bet Neen human visual performance and the
visual environmmt will be investigated in order to pro-
vide it more ra tional and accurate basis for specifying
illumination levels and visual design of spaces in build-
ings; this phase of the program activity is essential to the
development of energy performance standards for build-
ings, a major ERDA activity.
In the days shortly after the oil embargo, the Fed-
eral Government moved to take emergency action on its
own energy consumption, through a program that is now
called the Federal Energy Management Prqgram, at FEA.
As a part of the Federal program, the General Services
Administration conducted extensive delamping of
Federal buildings to meet illumination guidelines refer-
red to as the 50/30/10 standard (50 footcandles on the
desk, 30 footcandles in areas surrounding work stations,
and 10 footcandles or less in hallways and other non-
critical areas). FEA built on this in the Lighting and
Thermal Operations program, which continues today.
The objective of this program is to take some of the
energy conservation measures that have worked best in
Federal buildings and encourage their application in the
private sector on a completely voluntary basis. A publi.
cation entitled "Lighting and Thermal Operations Guide-
lines" has been made widely available and,forms the core
of a marketing effort to encourage the adoption of
energy conserving lighting and climate conditioning prac-
tice. In Federal buildings, energy consumption has been
reduced about 18 percent below preembargo levels,
mostly as a result of the measures included in this pro-
gram. FEA also sponsored research on lighting energy
conservation in office buildings, in which the standard
Illuminating Engineering Society recommended illumina-
tion levels for typical office tasks were questioned on
the basis of new experimental data. This report, with
actual operating experience in a large number of Federal
buildings after implementation of the 50/30/10 guide-
lines, indicates that somewhat lower illumination levels
can give adequate light for office tasks. Finally, addi-
tional information produced by the National Institute
for Occupational Safety and Health, and by FEA's
recent Symposium on the Basis for Effective Manage-
ment of Lighting Energy, indicate that energy savings in
lighting can be accomplished without jeopardy to safety,
health, or performance.
82
-------
94th congress}
1st Session
COMMI'l'TEE PRINT
POTEXTL\.L FOR IMPROYE~IEXT OF
. .
EXERGT EFFICIEKCY OF ..:\."CTOMOBILES
AXD ~L\..JOR APPLL\.XCES
l'IU:PARF.D Ie'OR
SrBCO)DHTTEE OK EXERG Y ~U\D POWER
OF THE
CD
W
I:\TERSTATE AXD FOREIGX COMMERCE
CO)1BlITTEE
eXITED ST_\TES HorSE OF REPRESEXTATIYES
ALGLST :''9. Wiij
,:)1;-743
U.S. GOVER!oOME~T PRINTING OFFICE
WASHINGTON: 1975
APPENDIX A
TECHNICAL OPPORTUNITIES FOR IMPROVING THE
ENERGY EFFICIENCY OF THE NATION'S APPLIANCES
(By Kurt W. Riegel, Energy Research and Development Administr.a-
tion Washington, D.C.) .
I. IXTRODUCTION
Appliances are defined here as including water heaters, refrigera~ors,
freezers, ranges, television sets, room air conditioners, clothes dryers
and washers, dishwashers and any of a number of small appliunces
normallv found in residential use. Excluded from this discussion are
central 'climate conditioning systems such as furnaces, central air
conditioners, heat pumps. ilnd large commercial appliances.
Appliances presently account for about 8% of total U.S. energy
consumption. Electricity and gas represent the dominant forms of
energy required to operate appliances, with energy delivered in the
form of electricity accounting for about 70% of all appliance energy
consumption. Only in the case of the Xew England and the )'Iiddle
Atlantic states does fuel oil and liquified petroleum gas account for
~ome portion of water heater fuel consumption, and for this applil1.nce
It amounts to no more than I:3~ of total energy for hot water hel1tmg.
Residential-type appliances .ire found primarily in the residential
sector and account for I1bout 82% of appliances' energy consumption.
However, they are used to some extent in the commercial sector also.
Energy consumption b~' the Xation's residential appliances has been
determined by the Federal Energy Administration during the develop-
ment of the Project Independence Report. Data on energy con;;:ump-
tion, for the Illajor energy-consuming I1pp].iances, are given in Figure 1
below. This information comes from the Final Task Force Report:
"Residential and Commercilll Energv Use Patterns 1970-1990",
November 1974 Yolumc 1. The data' have been corrected to give
primary energy consumption, reilecting total energy consumpti~n
mcluding electric power plant los;:.es of 69%. ~o natural gas distribu-
tion losses are included. They ha,e also been interpolated IIDd ex-
trapolated to on the basis of trends th'1t appear in the report cit~
above, to 1972 and 1974. .
TABLE I.-PRIMARY ENERGY CONSUMPTION
[ThouS2nds of barrtls per day equivalent)
Appliance
'1.12
1914
. ,
Water heaters:
Electric.w.............----...........--....."w--.....--...u,,,,--,,- ..., 412 .~
Gas...... --...... u.',""""" uu.." --.. u. --.'." -- --.. --uu-. uu u.' 574 620
Refrigerator-freezers. --- .-.----------_... .._-_.*._. --______0" -- --p---------_:'_- . 589 636
~~"r:':;~" -. .uu.' .... --... --................ --................~... w. u ...u 139 I!fJ
a~;t:~~[[[ . m m
~~~i~~~~~itioners:::::::: ::::::: ::: :::::: :::::::::::::::: ::::::::::::::::::: . ~~ ~193
Clothes dryersu...uw.........--..u--....--.................u.w--....u_.' 135 . ~
g~w":s:~~.~.'_::::::::::::: ::::::::: :::::::::::::::::::::: ::::::::::: ::::: ::: '('~N)",~ (2m)~ ~
Other appliances.-. w....." u.... --...." w.....'," u u u..'''''''' -- --' on' . 293 JI5
Total..--.. no.""''''' u.. ...... .'''--'''''' .u... ..." u ...... --... '.' '2.916 3. I'll
." .. 'It
-------
G
CD
~
II. BENEFITS OF REDUCING APPUANCES' ENERGY COYSU}IPTION
One policy option for r: dueing Lhe energy consumption of. the
Nation's appli~l1et:s it thrt of setting minimum efficiencv fltu.ndards
foj' the major e~ergy .consuming ones. A~o.ut 70% 0/ appliances
@~Ig-.t \:.ui.i.bUiUV~.uu 1:::' JU lI~Je funu ui ~lccw""Jl.a~.y, a. iraCLion willeh is
J:!rojected to grow to 80% by 1990. This fraction is based on projec-
tions of current p~tterns and does not allow for the introduction of
new major appliances; new appliances are expected to he primarily
electric8.l, and the figure could be higher than SO%' .
, Electricity is a special energy irom whic:h has a number of asso-
cIated pl'oble~s t~at are natio~al in sco~e and which requiTe planning
over extraordmariJy long penods of tIme. Conservation measures
which result in reduced appliance energy demand will affect electricity
requirements appreciably, and wiJI thus haye the foJlowing desirable
results:
1. R.educed capital investment, requiremE'nts for facilities
E'XpansIOn.
2. Ameliorative environmental bE'nefits Ilsso<:iated with di-
minished requirE'ments for new fuel production, generation and
transmission fncilities. -,
~. Reductio!} in the pace at which fundamental national
policy decisions must be made about the most desirable mix of
nyclear/solar/geothe~al!fossil ele~tric power generating facili-
tIes, about the best miX of deep mmes versus strip mines for coal
development, and about other energy dE"velopr.lent, questions
of primary natio!lal significance. Energ}" savings that result from
con<>ervtUion program" reduce the pres<;ure for rapid expansion
of energy production facilities. Especially in vi!'w of the fact
that selection of major energy production option~ t{,J1d to foreclose
o,ther options.because ,)f the large.resourc~s required and the long
tIme s~ales mvob;ed, conservatIOn measures are particularly
attrac.tlve.
ill. TECfCiICAL ENERGY CONSERVATION :lfE,\.SURES I:\' ,\PPLIANCES
T~e mo:'\t app~opriate energy c;onservation m~aSllrE'S vary from
3pJ:>hanec to I1ppll!uce and nccordlIlg to the fuel, electricity or gas,
whIch powers th,) appliilnce. HerE" tlre discussed distinguishing aUri-
butes of each of th" major appli.ances leading to differc!lces in enerO'y
conserva.tion measures dUlt will be E'ffeftive in each case. ".
.d,. U'ata ~tf3rs
Hot water energy consumption might be rf'duced in two ways: by :
~creasingJ'I1ergy I.osses which ?ccur wi~h the water ht'ating appliance
itself, and by ma~ changes m the wa:ys hot water produced bv the;
lieater. is uSed. Of these t".o, b~ far the most promising energy' con- !
~atlon m.ethods are tho1*' whIch rl'duce hot water recluiremt'nts I1t i
@intofeIJl!.use.ForexamplE'.typic-aloverailefficiencie,;o.£bo, th elec-
fiHc and !1:~, water hellters are low, 2:3% and 44% respectIvely, when
~~~.due :to:~1ectricit,y I!'ener&ti::m and transmission, tank insulation, '
~olli~ht "Ilri;d ~ue, are taken int~ account. Redesign of the watel'
~qter .ltsel,t,:,m life-l'rcle cost E'ffectlve ways, offprs a target for n18xi-
Il11Jm I?cre.8;seq"effiC1~ncy to: stout 28% for rlectrir- (reduc~;; fuel
cbMurnpUbrl by 12%) and to'about 62% fo'r ga" t:reduces fliel con-
7
6U!i1ptio~ by 29~~). (':Re~ic!.e::lt;1l1 Water lIeati'lg: Fuel C(':L,cn'ctioll,
EeoJlonll<::~, ~nd P"',~\I!I: Poh,.y," Ra~d CJrpOI'lti(}11 Report pr",pured
for the ~RtJOnal :,,\('I€Il('e Foundr.tlOn. R-1498-XSF, :.Illv 197.1),
There ar~ .db~ '-E'ry I.a]'~e el'?ctl;c Ilnd ga" wolter heatp;' .'GJ",;"'rY'lti";n
~.~~~!.!t!~!t!?~ ~~~ ~~:.:~!~f~:~~ ~~ ~ ~~~'J c! ~i';;;6~. 'l'~kil~g ;"'~i.u;': '-:\.U~;. t ;vu
>ihowers instead oi Iuillub h~th", the use of devices to Cllt (,tf "ink hot
wat<:r when, the tl~er steps :1~a~-, set tin~ watt'r tt'mpemt!~;'l' tJ.~ Iv\\' as
possible, ('Olel Wal('r laundenng. dc., wIll al! cut consun,p!ion d hot
wat.e)'.
About 25% of residential hot water is ,'stimatl'd to 0'0 foi' dotlws
washing in homes that han dl\thE's washl'rs (a;JOut 71<;'; i~ 1372\.
(.r G, Mulle!", ~ederal Energy Admini,;tra tio:!). .-\Ilotl\l~r 5~c. g.)jth gas water heators. RE'pluccment of the stanllil1!! ~.t" pi]ot
lIght m ?;as water heatPrs, with a]tema tiv~ ig'lli!ion devices :.hii,- .[.) not
consume ene:'gy continuous],y may offer ad'Clirion31 energ\- ...:.1\in:::.-.
B. Refn'gerator.s (uul fne:zers U .
Thi" appliance class, thou~h :second in energy c<>nsunH1tion to
wa~er heater~, repr~si!nts the one that, offers the fargest pct;mj;1 for
natlOna] energy sav:ngs as a result of lmpro'.-ed d...sigll and effi('ipnfT
within the appliame itself. A target for elJergv cons~)'vati()n of abo~it
50% of energy lIsagp is possible and mi"ht be reaIizl'CI b,- chawrf''' in
!llanllfac.turir.g practice and appliance design that are ~f)"t..dr;c:tin'
m the life c~'eJe "ellse. ("Tht' Productivity (,f Sprvicin~ COli-tImer
Durable Products," .:\lIT report for the National Scil'I1ce Foundation,
CPA-74-4,1974).
Resista~ce heat('J's OIuJIion heaters) are 11,;ed to f)rp-'-t"1~ the
acc~unuJahon o~ condpnsed water 011 the outside sm-f a"e (Ii "'.JD1E"
relngerators m cllse of I ugh. humi~it.r. They use more dum lOC(. of
the ener!:,)' consumed by this appliance, and could be rep!:tC'E'd br a
system that u?es the waste hE'at rejected by the c.ondcn:'er coi!-. [t'- i"
already done III some modek Somt' refri£:erators 1IIreaa\' !tan" .:-esi"t-
anc~ heater disable smtches, called power silver switche;: th.."", mig:ht
be mcluded on all units. ~\Ianual defrost refrigerators allt! freezers
c,an be .mo~e conservativE' of energy than automatif' dl'fr05,t units.
Such urut,>, I! defrosted n:gu!arly as r~quired, are superior 10 ;;.ui('HJ:Jtic
defrost applIancp.s In botl1 tir>t and lif" cycle cost. ,'onn'1Ji"n('E' i" the
factor th.at h.as neatly driHn tJlallua! defrost ufiit~ out 'Jf tbe Ill::rket.
Techruca~ Imj)fOY~11lents that rpsult in redu<:cd enl'rO'\' cou-amntiun
deGre~ed Jif~ c.n]e (;ost, but )Jos"ibJ~' ilicreasrrl fi~t ro,,~ i,:-;cl;id~
better H~sulat!OJ1, better quality control in manufacture. and --e:e<'lioJ1
of supcnm" hardware including motors, cumjWt's:'ors, comi'!Q,.;<:..,. . hl'at
exchanger" and c0l1tl'01 :'ysterns.
-------
8
00
at
O. Ranges
There is some room for impl'oyement in the efficiency of conventional
cooking ranges. The main design opportunities include 5imple impro~.e-
ment of the insulation of ovens and the replacement of the standmg
gas pilot light with electric igniters. Energy savin~ can be achie,:ed.
by insulation improvements-the extent to which cost-effec~lve
modifications can be pushed depends on suc~ factors as oyen vent!lla-
tion requirements and incremental costs of Improved design. Savmgs
of 10% for electric and gas ranges are possible.
Pilot liahts use about aO% of the gas consumed by ranges. The
permanently-T usm~ m~cl'owave
o.ens should be examined. ConventIOnal oven msulatIOn tmprove-
ments that are cost effectiye in the life cycle sense should be promoted.
D. Tekviswn StJts
This appliance is a surprio.ingly large consumer of energy. ':acuum
tube sets consume over twice the power of a comparable solId-state
set., and are chosen preferentially primarily because of first cost ad-
vantuge that thry enjoy. 1115tant on vacuum tube sets waste It. g;r~at
deal of ener!r\'. Life n-ele costs that include both the cost of electl1.cIt.y
consumed ~d servicmg costs are lower for most solid-state receivers, .
on both count.s. The normal barriers associated with consumer reluc-
tance to give full weight to life cycle co~t. considerations ?perat.e here.
It appears that the eyaluation of teleVIsion technology IS rapId, a~d
substantial conversion away from vacuum tube sets toward sohd
st.ate sets, with a 50% energ}- savings opportunity, has already taken
place.
E. EOQTn' air conditioners
This applil!nce has been one of the strongest ~om~onents of. in-
creased appliance energy usage in recent years, and 15 qu!te sus~epuble !
to prO!!Tams in increase efficienc\-. For example, room aIr condItIOners
comm~nly found on the market" operate with energy efficiency ratings
t.hat range o.er about a factor of 2.5 (J. C. Moyers, ORNL-NSF-EP-
59, October 1973). ~Iany room air condition~rs have ~o switching de-
,ice to cut off the fan when the compressor IS off; domg so would cut
energy consumption a little. Presently available room air conditioners,
toward the top end of the efficiency ran~e, already use abo~t 40% of
the energy for the same delivered. coolmg power than Ulllts a~ the
bottom end of the ranae. ~trateQ"les to change the market mix of
presenth' available unit~ offer m~rked potential for increasing the
overall efficiencies of the Xation's room air conditioners. Further tech-
9
nical improvements in electrical motor efficiel1cy, in compressor.
effif'iencv, and in hea.t exchanger elliClency are all possib)£'. Currently,
small condensor area" clakE' the tl\sk of efficient heat ~xcha.nge very
difficul t. .
An effic;"ncy improvement tl1J"get of 25% over available 1972 units
i~ life-c,'cle cost-justified on a nationwide basis, and larger increments
in efficiency are life-cvcle cost-justified in southern regions where
annual operating hourS lI!e high. Finall):. .the technical measure of.
increasing fan motor efficiency, and proVIding for control of t!:l;e fap
motor together with t~e compressor motor ra~her t~an operating It.
continuously, can proVIde another 5--10% efficIency tmprovement.
F. Clothes dryer.~
Gas dryers already incorporate electric igniters in a larger fra.ction
of manufactured units than for any other appliance. Some efficlepcy
improvement is possible by manufacturer selection of more efficient
motors to drive both electric and gas dryers, and by improvement of
control systems so that drying energy is consumed only when the
clothes in the dryer still contain moisture. Finally, there are oppor-
tunities for recuperation.of the heat content of exhaust air that is
presently exhausted. Heat excha~ger rec?very devices mi.ght ~e used
more widely to pre-heat new dry atr ent~nng t~e dryt;r. It IS estlmat~d
that there is room for at least 10--15% efficIency Improvement, In
life-cycle cost-justified w~s, for this appliance. (W. David Lee,
Arthur D. Little, Inc.)
G. Clothes 1cashers
Technical measures for increasing efficiency include .~od}fications
that are J?0ssible to the geometry of the washer to mllllIlllze water
consumptIOn. hot and cold, manufacturer selection .of mo~e effi~ie':lt
electric motors. ~Iost of the energy consumed by this appliance IS m
the hot water that it uses. Changes in the opera.ting cycles to reduce
the use of hot water would thus weigh very importantly in reducing
total clothes washer energy consumption. Wider use of cold water
detergents would also help enormously.
H. Dishu:asheTs .
Changes in dishwasher geometry and control of the washing cycle
can reduce energy consumption markedly. Elimination of the auto-
matic dry cycle, with at least a provision for manual override, would.
save the "dry;ng energy. A very large indirect energy demand is placed
on the water heating systems by many dishwaters, since the manu-
facturers recommend that water be heated to 150°F at the water
heater, in order to provide water that is hot enough for the dishwasher.
No other appliance in the home requires wat~r this hot, and thus a
very large portion of standby water heater losses are. accounted for. by
the dishwasher. Although direct energy consumptIOn by the dish-
washer would go up in some cases by providing for more booster heat,
overall water heating energy requirement~ might be s1!bstan~ially
reduced. It is estimated that at least a 10% Improvement m .efficlency
is justified by changes that can be made on the average dishwasher
in the manufacturing process, apart from booster changes discussed
above (communication from U.S. Derartment of Commerce, National
Bureau of Standards). Including al possibilities, savings of at least
20% are attainable.
-------
10
IV. :;U~n'~hY CF t i'FiCIEXCY DfPHOVE\iEY!' TARGET:;
TaLle ~ ~e]ow cOI,tain;; hest current, f'nd cons/'r.'ative with respect
to remark" in the paragraph" lJreceding, e"timutes of efficiency im-
jiJ"" "-~::atl6lii~ L~.r.ni ure l~ciI11~l:ttji,~ ieusiuie ior r ilC flU tionai mIX ot the
product listed, in terIJE of lhp base year 1972. For almost all "ppli311f'P;;,
t!lere b'as b~n no appreciable ('hange in the PPl'rgy ~ffi,;ienc.'" chllr-
~cteristics of Uluts produced in 1974, with the except ion cf t..'levi~ion
sets. This apfJliaO"e has experienced a stron~' tren,l away froll1 \'II':lIum
tube sets toW',ud tho~ incorpor..ting solid-state circuitry
TABLE 2
-------
fto ppliance
inel"'.Y efficiency impl wverr.~! t
paten~ial (pefc~rl~)
1972 ene'gy
sa.."i:1i5 I
-_.
1972
1974
---~----_.
Wah.r heat7f'!:
flectric- --.--- -----_. -- -- ----. - .---- -----...--.. _.-.---.-_.
Gas_------ ------------------- n.-.-.-_.__....-.-.-._-_._---
Re~rjgerator-fr~rs- _n___u n_----.- _'.h_- _,hUU" '.h'_____-
freezers_- -..-. -'-'--.-.-- .-- - -. ,_. -. '--' --. '____'0 .- ._-. --- --.-
Ranges:
UectIic.u. '........,... ------ ........--uu...u"..,.... 10
Gas.____-- -.-.--.---..- ...---- -- -- -. -....-.-... .-.---- ----.- 35
Televisjon se1! h.un_, ..--.. --_.O'''''''',,,uuuUn._-, uu 45
Room air-condrt,>aer5. --'-n'-- un_.-.---.__... -".'nUU'.__'n 25
Clutl>esdrym.:... ""..._m......." ""'.... ouu"... u to
ClothesW2tSher~-nh_n.-n-__u u---- '''nn__.__.-....-----.- 10
OishN8Shf:5--_n... .-.__h._u_-nhh... .-.-.. -_u-- 'u_---u.- 20
Other a~piia1"res_n. --'-.--u. n__- ...... u_------.. nn' u__. _--_n.. u -.-.
12
27
45
40
12
27
45
4U
49
iS5
265
56
16
f.4
104
4;
14
i9
3
~~
30
25
i~
10
~
- _. -. -- -~. -... - _. - - _. -.. . - - -
CD
C)
Total.... --.Uh'--...."h... ..... '''''''''.h'",,, .UU
28
26
789
-----
I Thousand ~rrels per c:ay eqLli\'a~t.
The> energy-weightt,..I il'\"erage efficiency improvenh:nt potl'ntial
for all the named appliancps in this ]ist is 28 percent with re~pect
to !972, 'Imd 26 percpnt with re-spect to 1974. Tlw energy ~1Jxing"
indicat~d .-..re th~ th,~t wlJuld have resulted if the ef!icif,nc\' impro'.-e-
menb had been iIl1plp~ne:1ted in 1972. Saving" in later year.; !tf'l hugeI'.
1n Jt: C8£es. thl' li"ted en.~rg~' efficiency improvemPBt potential i~
consistent. with t1H\ C'omtraiJ't of life cycle cost effectiYeue:..~.
V. DtPRPV1XG Jl.PPLIJl.XCF EFFICIEXCY BY RESEARcn AX:> IJE\'ELOP\IEXT
OF XEW TECHXOLOG,
There is I\illple oppornmit,. for progres~ towllrd improl-iuf tllP
energy dE-:-iency cf the X :ltion's nppliancf's far beyond thl' enb
di;:cussed for the nea~ ; 'I'm ;,1 ('\)I1\-entiono) ,tpplia1w('s. The rt:,,-earcl.,
dp.vel, OplJient anJ dpIJ:onstr.nion program;; of the Energy He~earch
:.:nd Dev.elop:<1ent Adminis'rat;on are aimed "t deliwIlrl6 j..!, sueil
co'ntriblltioTIs. This a;:.tiritr rill};! con,lucted at ERDA \I ithin the
Pivi"ioD of Build:ng5 all; r;'dlbIry.
For exam"ie, int('~ratt\d or tn"bri{! ap::Jlinllces, whi('h eop;bine
..,;evera111J'i>lanc€5 tJllt are: ,'esentfy thought of il~ sep:lr:!.t" on.;", may
~rovid" )ar~e ene~g\ ,;t1Yings, "?articlilarly r A' mobile lu"ne .,pp~icu-
tlot"" hybnd refn~t'r3:or-Tater heat?!'s would nlJow for tl.e "~e of
..-aste p~ltt Iruni th~ n.frigcm tor l'oIr.ponent to heat clon'e;;tic hut
wdt~r.
Ii
Dn"doprnent ()I ,;en,ol' technolo~~- and other control sys1em
comvonel:t;;, incorpo:-3t:Il~ advnnced solid-state .11(: int~g-at{ , c: "cuit
techuolo!!\-, n1:1 t)!'ovi\l~ .:heMP and efl, c.t; ,e {'ontrol ov"r nooliAJH'p,;
eIleTI!Y con"umptio:l. For e~9rnple, the "ase of t11(' nrespnt -clay al.lto.
IIll\t~ ~c.fr()st refri~<'rtHO., which goe>, thrnu!:-ll ih defrost c.fde min~l-
les~ly Oil 11 re!;ular rel'('lit: ve basi;; regardles;;: of need ;s !'usC{'ptible: to
contrd so th:lt tr.e .Iefl't')st ('yde is Htuswd on" when a~tuullv
needed. :\Ian." othpr applia.nce"s can beuefit in l'eelIeed ~nergy (.c,ri-
5111nption fr.)m improvl'<[ controls.
Fim\lj \' RD&D work on basic components of appliances, in order to
exploit ;~d bring to the m8rke~plac" the most effeetiv,:, devir:es II.nd
idea.5, ~hOl"d result III innovative ttpplOll' hes 10 'l.-lplIlI.nce,; desIgn
tha t mll~' r~sult in marked cnerg)' reduc:,io:..
o
-------
THE DESIGN OF ELECTRICITY TARIFFS
Charles J. Cicchetti, Ph.D. *
(Text adapted from Public Utilities Fortnightly, Vol. 96, No.5 (August 28, 1975), pp. 25-33.
@1975 by Public Utilities Reports, Inc., Publishers. Used by permission.)
Abstract
This paper discusses time-of-day pricing as a method
of pricing electricity which could be more widelyadopt-
ed in this country. Ten basic reasons are enumerated as
. to why time-of-day pricing produces beneficial effects.
The design of electricity tariffs, as well as their level,
has been a source of consumer outrage. Many thoughtful
critics wonder out loud, "How can volume discounts for
electricity be offered at a time when energy conservation
is widely advocated and each year electricity prices are
increased because of growth in use?"
Consumers are upset over their increasing electricity
bills and angered by the conflicting explanations given
by utility spokesmen. In 1973 they were told that in-
. creasing prices were due to the higher costs of increased
consumption, but in 1974 the same spokesmen blamed
higher prices on conservation efforts. There is little
comfort from the fact that within the industry there are
simple explanations for the apparent inconsistency in
these two statements. But many, and their ranks are
increasing, feel that a different tariff structure would, in
addition to making better economic sense, eliminate
much of this confusion over tariff schedules by com-
municating varying costs directly to the consumer.
Behind the conflicting explanations for rising prices
is the concept referred to in the industry as load factor.
On the one hand, if increased usage at times in which
system peak is likely to occur makes it impossible to
avoid the need for new capacity, the cost of building
new generating capacity will result in increased unit
prices. According to utility spokesmen, this was the case
in 1973. On the other hand, the greater the use of exist-
ing facilities, the greater benefits there will be in holding
down costs and prices for all customers. This latter
phenomenon explains the 1974 statements, when con-
servation apparently led to a reduced use of existing
facilities without diminishing the need for capacity and
thus resulting in higher average costs. The effort to
design tariffs which incorporate both these considera-
tions simultaneously has been the main thrust of my
involvement in the electricity tariff controversy in the
United States.
. Director, Wisconsin Office of Emergency Energy Assist-
ance, Madison, Wisconsin.
Eliminating volume discount pricing and the tariff
philosophy of accountants, who do not seem to have
been informed about changing use patterns, and sub-
stituting time-of-day discounts is the particular reform I
have stressed. The reasons why I have come to this posi-
tion will be discussed, how it may be implemented, and
then some of the other tariff reform alternatives offered,
especially with low income consumers in mind.
Time-ot-Day Pricing ot Electricity
Ten reasons time-of-day electricity price reform
makes the most sense are listed below and discussed each
in turn. They are as follows:
1. Cost minimization for the utility;
2. Equity and fairness in tariff structure;
3. Social welfare maximization and economic
efficiency;
Load factor management;
Reducing environmental externalities;
Energy conservation;
Earning stability;
Tariff stability;
Consumer benefits; and
I ndustrial protection.
4.
5.
6.
7.
8.
9.
10.
Cost Minimization
Efficiently managed electric utilities attempt to
minimize costs in different ways. First, the systems
planner and engineer attempt to minimize the cost of
meeting what they consider to be the future demands
and load pattern for the utility and its connecting sy~-
tems. They do this by using various sophisticated engi.
neering and computer techniques. Their success in these
matters is often imitated by many countries around the
world. Additionally, electric utilities practice economy
dispatching for the operation of the electric system at
any given point in time. The unit cheapest to op~rate is
called on line to provide service to meet demands at 'a
particular time and place. Once again, elegant and
sophisticated, economic and engineering models are used
to practice economy dispatching. The simple fact is that
electric utilities cost minimize in both the operation and
the expansion of their systems. In the practice of this
cost minimization, it is understood that the cost of a
kilowatt-hour of electricity varies over time. The first
desirable feature of time-of-day electricity pricing is that
87
-------
it will attempt ':0 track the cost minimilation calculation
of the systems planners and the utility's economy dis-
patchers by reflecting this cost pattern in the tariff struc-
ture. The systems engineers' knowledge of cost would be
communicated to consumers and consumers' demands
and willingness to pay in the reverse direction.
Equity
Economist; are not the best ones to discuss the
question of fairness or equity when it comes to elec-
tricity prices. The reason for this is that economists
usually try to ,Ivoid taking any stand on whether or not
two alternative.. are fair or equitable; that is, economists
are more comfortable at making allocation or efficiency
decisions than in making social equity decisions. How-
ever, it is frequently asked at electric utility regulatory
proceedings w~ether or not time-of-day pricing will be
fair or equita;Jle contrasted with alternative pricing
systems. My response to that has usually been that by
pricing a kilowat~-hour of electricity for all customers on
the basis of th(~ actual cost of the utility would be the
fairest and the most direct way of pricing electricity
imaginable. It 1N0uid be fair in that it would price all
similarly prodcced kilowatt-hours from a cost stand-
point alike. Distinctions would not be made based
upon total levels of consumption over a billing period
and in this way if the fairness criterion is used, the cost
of a kilowatt-t our as it varies with time, it can be
claimed that equity and fairness have been increased as
contrasted with ~he present system of volume discount
pricing. The present arbitrary and often subjective prac-
tice of cost aliocation would be virtually eliminated
since kilowatt-hours produced at the same time would
be priced alike ether than for voltage differences.
Social Welfare Maximization and Efficiency
The main public pOlicy rule of economics is that to
achieve an effiGient allocation of society's scarce re-
sources, and at the same time to maximize the social
welfare of an ec onomy for a given income distribution,
the cost of each good and service consumed by that
society should te priced on the basis of the incremental
or marginal cos'~ involved in producing and distributing
that good or service. Setting second best considerations
aside to be adjusted after the basic tariff structure is
det~rmined. tim~-of-day pricing is a pragmatic attempt
. to bring this very elegant mathematical polky rule into
regulatory proceedings and into practice by the electric
utilities in the United States. Consumers would benefit
from more stabl.! electric prices. Time-of-day discounts,
that will make it possible for consumers to take advan-
tage of lower co;t consumption, will benefit themselves
and the utility. Time-ot-day penalties, which will indi-
cate to the consumer when it is that electricity is ex pen-
sive from a capacity expansion standpoint and opt:rating
standpoint, are the very essence of the pragmatic at-
tempt to translate the economist's notion of an efficient
allocation ot resources into the tariff schedules of elec-
tric utilities.
Load Factor Management
As indicated in above. electric utility management
attempts to minimize the cost of operating and expand-
ing its systems. The greatest benetits ot time-ot-day pric-
ing is that it will add an additional factor to the objec-
tive function of system planners when they consider the
various ways in which cost minimization might be
achiev!!d. When a systems planner considers the cheapest
mean~ of expanding an electric utility system to meet
growing loads, one ot those options that should be con-
sidered by the systems planner is the possibility ot offer-
ing discounts to existing customers in order to encourage
a change in load patterns and a change in peak consump-
tion habits on the part of consumers. If it is cheaper to
give discounts to existing customers, and thereby en-
courage a shifting in use, rather than to acquire a more
expensive new investment or undertake more expensive
operating costs, the tariff approach should be under-
taken and, it available, it would be willingly undertaken
by the nation's utilities. Time-ot-day pricing, along with
the generous use of interruptible tariffs for the larger
volume customers, will greatly hold down the need for
new generating and transmission capacity in the United
States. Note that, frequently, interruptible rates are
down played in the United States because it is pointed
out that no single tirm could afford a long interruption.
However, in other countries several industrial customers
are sometimes packaged into an interruptible group in
which no single firm has to be interrupted for the entire
peak period, but collectively. the capacity saving can be
great and so should the size of the discount. In these
ways the need for new capacity can be reduced without
necessarily diminishing the use of energy and the con-
comitant increase in economic activity and jobs for the
nation.
Environmental Externalities
Environmentalists entered the debate on electricity
pricing just prior to consumer and low income groups
who were concerned with rising prices. For the most
part. environmentalists have been opposed to the exter-
nal social cost associated with the construction and
operation ot electric generating plants and the transmis-
sion systems that connect such generating plants with
the consumers of electrical energy. Since time-of-day
pricing is expected to reduce the need for new generat-
ing and transmission capacity. environmentalists, who
are concerned about such matters, would find that under
88
-------
time-of-day pricing the amount of new electrical generat-
ing and transmission capacity needed for the United
States would be less than under any alternative form of
electricity pricing that might be considered. On the oper-
ation side, many electric power plants have peaking units
which are old generating plants which are often very
inefficient to operate. Such plants are typically fossil
fuel users and are the most polluting from a particulate
and sulfur dioxide standpoint. Because time-of-day pric-
ing would shift use away from such inefficient, and
therefore polluting units, time-of-day pricing is also
.expected to have a desirable pollution-avoiding effect.
Energy Conservation
Many people in the United States are concerned
with our uncontrolled and seemingly unending growth
spiral of energy consumption. Time-of-day pricing would
discourage the most expensive aspect of this energy con-
sumption by helping to avoid the use of new and more
expensive generating and transmission facilities. Time-
of-day pricing would also result in a shift to more energy
efficient base load and intermediate load units and away
from inefficient peaker units and old fossil fuel plants
which are inefficient from an operating standpoint. It is
quite possible, however, that if energy is measured in
terms of total kilowatt-hours consumed, as opposed to
total amount of energy that is required to meet a given
load, that time-of-day pricing might actually encourage a
greater use of electrical energy. To some this greater use
may seem inconsistent with energy conservation, but
time-of-day pricing will improve the energy efficiency of
the actual level of electricity consumed in the United
States, and that is a positive improvement in energy con-
servation.
Earning Stability
One of the serious problems today in electric tariff
controversies is that many have selected a tariff design
objective which is inappropriate. To many the goal of
electricity tariffs is to achieve gross revenue stability.
This is a short-run and foolish objective for those who
adopt it. Electric utility gross revenue stability can, of
course, be achieved by charging as much money "up
front" as is possible and which the electric regulatory
commissions will permit. In this manner the goal of gross
revenue stabilization can be achieved. But that leaves
very little revenue that can affect the discretionary use
patterns of the customers of an electric utility. It also
means retaining volume discount or promotional pricing.
It, therefore, means that nobody will be getting signals
as to when it is cheap to consume electricity from the
standpoint of generating capacity cost or system-operat-
ing cost. A pricing system, based on time-of-day costs,
that attempts to tie revenues for the utility and costs for
the utility together is the essence of time-of-day pricing.
But the objective that the pricing system is meant to
achieve is earning or net revenue stability, and not the
incorrect goal of gross revenue stability. The controversy
surrounding electricity tariffs often becomes a discipli-
nary conflict with accountants, who a're taught and
trained to think in terms of gross revenue stability, and
economists, who are taught and trained to think about
earnings or net revenue stabil ity. The two are in confl ict.
The economist's position, which closely follows the
engineering cost minimization position, is the most desir-
able pricing system for our nation's electric utilities.
Tariff Stability
There is no rule of nature that requires the unending
round of electric utility revenue increases to continue
unabated year in and year out. The customers, the in-
vestors, the regulators, the utilities themselves must and
would all find a common interest in ending this vicious
and unending spiral. By pricing electricity in a way that
varies with the costs the utility expends over time, this
chain can and will be broken by helping the utility and
the customers of the utility find tariff stability. This
objective will be furthered because changes in consumer
use patterns will result in changes in revenue and costs
that will move in the same direction and tend to offset
one another.
Consumer Benefits
As an economist, perhaps the best reason that elec-
tricity pricing reform makes sense for the Nation is the
expected consumer savings. Recently, the British and
French, who have been practicing time-of-day pricing for
almost 20 years, have been asked to estimate what they
think the annual savings of total electric utility cost are
by having time-of-day pricing installed for industrial
customers in both countries and on an optional basis fc,r
residential customers in both countries. Total installed
generating capacity in both countries is quite similar to
one another in total size. Both have aJ1proximately
30,000 megawatts of installed generating capacity. They
both have estimated a similar savings in annual cost for
that level of installed capacity of about a quarter of a
billion dollars per year. These estimates do not include
savings in operating costs that are undoubtedly associ-
ated with the change in the time pattern of use of elec.
tricity.
If we look at the United States, which has an in-
stalled capacity of about 10 times either France or Eng-
land, a similar annual savings would be approximately
$2-1/2 to $3 bi II ion per year for the capacity cost sav-
ings alone. Additionally, the French have made another
89
-------
calculation. Tiley have also estimated what the savings
would be if, ir, fact, they sold the same amount of elec-
tric energy toe ay, but using the load patterns that exist-
ed 20 years earlier when time-of-day pricing was first
introduced. ne French estimated that their annual sav-
ings would pre bably be five to six times the quarter of a
billion dollar !avings that already has been mentioned.
Additionally, there are significant operating cost savings
that are expected in the United States. If our industry
were as responsive as the French and the English, annual
savings of more than $10 billion per year could be
expected. By holding down costs all consumers will
benefit. Furthl!rmore, by offering time-of-consumption
discounts all who can and do switch will save additional
money.
Industrial Protl'ction l
Industrial uses of electricity in the United States
were priced in such a way that individual customer load
factors were improved. Today this incentive is in conflict
with economic, environmental, and conservation objec-
tives. Industri, I customers are typically good system
load factor customers, who would benefit from time-of-
day pricing. However, the volume discount pattern of
pricing does not provide the pattern of incentives that
will encourage them to save the utility money. Time-of-
day pricing will encourage the good system load factor
customers to tecome even more so. They will also en-
courage very pl)or individual load factor customers, who
use short durat on but intense power, to shift such eratic
use to off-peak discount periods.
In today's hostile political climate, volume dis-
counts are un':enable. Replacing them with time dis-
counts will not destroy industry and cost much needed
jobs; instead, it will benefit them.
With such jramatic savings to be gained, whether it
is characterized by using any of the 10 objectives that
have been men tioned previously, there is a strong and
undeniable cas!! for time-of-day pricing in the United
States. Now let LIS turn to the second part of this discus-
sion: How the particular problems of low income con-
sumers can be tuken into account.
The HOW ot Time-ot-Day Pricing
Time-of-daV pricing has many names. Sometimes it
is called peak bad pricing; sometimes it is called incre-
mental or marginal cost pricing. Each name attempts to
be descriptive, IJut because there are several names that
might be used, t can be inferred that none of the titles
by themselves is sufficiently all-encompassing to describe
the many characteristics that are implied by any of
them. If one th nks in terms of incremental or marginal
costs, it is important to understand that there are 8,760
hours in a year, several voltage levels, and various gener-
ating stations and customer .Ioad centers. Therefore, the
number of possible margins that might be used for set-
ting marginal cost prices is extremely large. Electricity
tariffs, therefore, have to be based upon a compromise
between the reality of a virtually unlimited number of
margins and the pragmatic requirement that electricity
tariffs must be understood by various customers of the
electric utility. The traditional separation of cost for
electric utilities into generating, transmission, and distri-
bution, along with distinguishing between energy and
capacity costs, is a desirable first cut that is useful to
retain when describing the time pattern of the structure
of the electric utility's cost. This pattern of cost, as it
varies over time, can then be used to develop tariffs
which are simple and easily understood by the customers
of the electric utility system.
Economists have used a device that goes under the
Latin phrase "ceteris paribus," meaning "other things
being equal." To the mathematician, when that phrase is
utilized, it means taking a partial, rather than a total,
derivative. In short, it means holding things constant,
which are otherwise likely to vary. In determining the
marginal cost of the various dimensions of electric utility
cost, use of the ceteris paribus or partial derivative con-
cept makes life much less complicated than it otherwise
would be.
To explain how this works, refer to a concept which
has received quite a bit of attention in electricity tariff
debates. The concept is called "long-run incremental
cost." It is used to determine the capacity cost of the
electric utility system looking into the future. It is deter-
mined by estimating what the future demand for the
utility will be in. the years immediately ahead, for exam-
ple, over the next 10-year period along with the ex-
pected pattern of that future load. Such projections are
a basic ingredient for designing the least cost plans of the
electric utility system by adding plant and equipment to
its currently available system. Once the development
plan to achieve the least cost expansion and operation
for the electric utility in the near future has been deter-
mined, the appropriate way for calculating long-run
incremental capacity cost is to look at the cost that
would be entailed if demand were to increase in such a
manner that the entire plan was moved forward by one
year. The difference in the present value of the two cost
structures converted to a per kilowatt cost basis is the
long-run incremental capacity cost. If reoptimization
would take place, and that information is available, it
should be used. In practice this might reduce to the cost
per kilowatt of moving a single plant planned for the
next 5 to 10 years forward by 1 year.
90
-------
An alternative, if information is more readily avail-
able, would be to ask what would be the reduction in
cost if the enti re plan were postponed by 1 year. For a
large electric utility system like those found in the
United States the difference between the two ap-
proaches is not expected to be great. The calculation
thus made uses the concept of holding other things con-
stant, ceteris paribus; it is quite simply the partial
derivative of cost holding all things constant other than
the level of supply, which is required to meet the expect-
ed level of demand.
When calculating long-run incremental cost there are
likely to be differences in fuel costs if new, more effi-
cient plants were moved forward or postponed 1 year. It
is appropriate to net out such cost savings or penalties
from the capacity cost in making such a calculation; this
step has sometimes been overlooked or omitted by some
who mistakenly limited the concept to capacity cost
differences. All those who participate in electricity pric-
ing discussions should make certain that the correct con-
cept is being utilized. The systems planners will be more
likely to understand the economists when the proper
definition is used.
If one looks at the operating side of an electric utili-
ty system, once again a partial derivative concept might
be utilized to measure marginal costs. Indeed, economy
dispatch makes use of a concept called System Lambda.
System Lambda is simply the derivative of the total
operating cost of providing electricity with respect to
the quantity of electricity supplied. Operating costs are
minimized by the system dispatcher for an electric utili-
ty by utilizing the available unit with the lowest system
lambda at any given time and place in order to meet the
load pattern of demand placed upon the electric utility.
System Lambda varies over time and across places and it
is, therefore, a very important concept that reflects quite
dramatically the time pattern of costs.
Transmission costs are an important ingredient of
the cost structure for the electric util ity. They vary
incrementally based upon the voltage level that the elec-
tricity is being transmitted at, as well as the distance and
the quantity of electric transmission that are taking
place. One can once again use a partial derivative, ceteris
paribus calculation in order to calculate the extra cost of
transmission at different times over different lengths and
at different voltage levels. Variations in transmission
costs calculated incrementally can and should be reflect-
ed by an electric utility in its time-of-use cost structure.
The basis for this calculation of incremental costs is
more long-run than operating costs, but closer to the
present than incremental generating costs. The basis for
the variation in costs is both extra line miles and trans-
former capacity per kilowatt of additional demand at
the various voltage levels.
Another major category is distribution cost. Distri-
, bution and transmission costs are distinguished by in-
cluding in distribution costs that component of transmis-
sion cost that can be easily identified with a particular
customer or group of customers further down the distri-
bution and transmission system. Transmission costs, on
the other hand, are those costs of moving electricity
from generating plant to consumer which are more com-
mon or collective to all customers who are supplied in a
similar geographic region. When it is possible to calculate
the additional costs of supplying (including metering) a
particular customer at a particular voltage level, these
can be used to establish the distribution or customer
cost component of supply.
The final step in analyzing costs is most important
because' it is the basis for the assertion that a kilowatt-
hour is not a homogeneous commodity from the cost
standpoint. There are several ways of incorporating that
simple fact into the final calculation that can and must
be made before tariffs can be designed. There are 8,760
hours in a year. However, each hour is not equally Ii'kely
to have demand exceed the available supply of the util-
ity. The available supply of capacity of the utility refers'
to the generating and transmission capacity which limit
the ability of the utility to meet the demands that might
be placed upon it. Additionally, the likelihood of
demand exceeding supply must also take into account
the fact that we are talking about demand relative to an
expected or anticipated supply, not the nameplate gener-
ating capacity of the system in question. Therefore,
scheduled plant maintenance and repair schedules, which
one has prior knowledge of and which will mean certain
units are unavailable during certain times of the year,
must be brought into the calculation of the probability
or likelihood of demand approaching or exceeding the
capacity of the electric utility.
Once the above has been determined, one can calcu-
late the probability of a loss of load; that is, the prob-
ability of demand approaching or exceeding the availablt,
capacity of the electric utility. It is important in recog-
nizing the time pattern of this probability of demand
exceeding supply to assign the cost of expanding the
electric utility systems, the long-run incremental cost of
generation and transmission capacity, to those hours for
which the likelihood of demand exceeding supply is
great relative to those hours of the year for which it is
small.
For some electric utility systems there might be
some intermediate or shoulder periods for which the
probability is not quite as great as those hours which we
might identify as peak, as opposed to those hours which
we might identify as off-peak. There is no simple gener-
alization that might be made in advance to fit every
individual electric utility system in the United States;
91
-------
nevertheless, it is possible to dl!filH! those hours (typi-
cally about 1,000 hours) in which we think the cost of
expanding the system when demand increases is greater
than those when demand growth does not imply capaci-
ty expansion.
If we havl! done all the above, we can calculate for
those hours w1ich are similar in terms of peak, off-peak,
'and intermediate probability of loss of load, the cost of
running the s {stem expected to operate during those
periods, and a.isign the incremental generating and trans-
mission costs to peak hours on either a kilowatt or
kilowatt-hour basis (depending upon whether we want
to perform an additional step of division). In addition,
we can determine transmission costs at different voltage
levels and customer costs further down the distribution
system. Once -Ale have done that, the structure of costs
for designing tilriffs is complete.
The tariff! that would logically follow from such
cost structure vlOuld not be so difficult for the industrial
customers of ar electric utility who would readily under-
stand them. Tt erefore, we do not need to worry about
major additional simplifications to develop tariffs for
that class. Additionally, most large industrial users also
have metering ,~quipment currently install'ed for which
we can measu:e use patterns sufficiently detailed to
encompass the cost structure described above after it is
translated into a tariff structure for this group. The
residential and ower volume customers present a greater
simplification problem. A discussion of this will follow.
For now, I strcngly urge that once this basic cost infor-
mation is knov,n, that time-of-day prices for industrial
customers be i'T1plemented along these lines with due
speed throughout the entire Unit.ed States.
To some the above may sound too theoretical.
Therefore, the actual experience of France and the
United Kingdorl in setting time-of-day varying rates is
described. I n a !Ieneral way there are some similarities in
the pricing systl!ms of these two countries. First, since
the 1950's both countries have adopted the economic
and engineering principle of marginal cost pricing and
have been moving forward in the direction of imple-
menting time-of day pricing. Today, almost all industrial
customers in both countries consume electricity under
time-ot-day and seasonal pricing structures. Smaller cus-
tomers in both countries, after paying basic customer
fees, are charged yor electricity according to either a flat
kilowatt-hour tariff or an optional time-of-day tariff. In
France about 20 percent of the domestic customers
select such a time-of-day pricing plan. The percentage of
kilowatt-hour sales in the United Kingdom is greater.
Overall, residentiC!1 customer consumption is substan-
tially less in both countries than in the United States.
and both countries have weather sensitive winter-peaking
louds. Thl~ customers who have adolJted till! tinll!-ol-day
pricing option are most likely to be the heating custom.
ers.
The French distinguish between their customers
very finely. For the industrial customer the price per
kilowatt-hour varies over five time periods: peak, winter
and summer full hours (intermediate), and winter and
summer slack hours (off-peak). They further vary their
rates by region and voltage levels. The customer also
pays a fixed charge to cover both the distribution costs
of the system that depend directly on the client's own
peak consumption and a portion of the costs that are
intermediate between collective costs and the individual
customer-related costs. This latter refinement is related
directly to the large hydroelectric proportion in the
French system.
A formula is applied in order to give a price incen-
tive to customers who adjust their capacity requirements
to system needs. If use at more costly times increases,
the customer pays a penalty. If alterations in use pat-
terns are made to benefit the system, prices are reduced.
Setting aside time-of-day variations in energy charges
and other use-category effects, this contractual formula
has been estimated to postpone the need to construct a
new generating unit of about 700 megawatts in size for a
full year. A further saving of about 1,300 megawatts is
attributable to the industrial tariff or tarif vert for a
combined estimate of 2,000 megawatts in a system peak
of roughly 30,000 megawatts.
Further adjustments are made for high load factors,
self-generation, emergency, and short-term customer
needs. Such refinements are primarily made in the fixed
charge. The basic charge for a contracted kilowatt is
about $55 per year. It is higher for self-generation and
lower for emergencies. Additionally, the energy charge
varies by a factor of about 6 to 1 in the base tariff and
nearly 10 to 1 in the short use emergency tariffs be-
tween peak and off-peak. The variation is less for the
standby service. The basic peak energy charge is about 5
cents per kilowatt-hour.
Low-voltage domestic customers are sold electricity
under the universal rate. Such customers have a contract-
ed fixed monthly charge and a proportional (flat) energy
charge. Circuit breakers are used to limit loads. Each
tariff usually has a day-night as well as a flat version.
Those selecting the day-night differential option pay
slightly more than one dollar per month for additional
metering costs. The day-night difference is about two to
one for an 8-hour night period.
In the United Kingdom, there are area distribution
boards who purchase bulk power from the Central Elec-
tricity Generating Board. Each board pays an annual
capacity charge with various adjustments, as well as an
92
-------
energy charge before fuel cost adjustment of about 1.6
cents per kilowatt-hour for daytime and early evening
peak months, about .65 cents per kilowatt-hour for
other daytime and evening purchases, and less than .5
cent for all sales between 11 p.m. and 7 a.m. throughout
the entire year. .
Area boards set their tariffs based upon this pur-
chase schedule. A typical industrial customer is likely to
pay a customer charge and a monthly varying capacity
charge (with as much as about a 10 to 1 difference
between peak and off-peak monthly charges). Day-night
differences of about 50 percent are in effect on the
energy charge for periods of night-time use. These are
typically of an 8-hour duration. Converting to a U.S.
price basis this would have been approximately 1.25
cents per kilowatt-hour charge in the daytime and 0.75
cents per kilowatt:hour charge at night before fuel costs
adjustment and peak monthly demand charges of about
$7 per kilowatt in peak months falling to less than $1
per kilowatt in the off-peak months.
Special terms for interruptible for flexible industrial
customers with a greater proportion of the revenue col-
lected in the kilowatt-hour charge are also utilized.
Prices during peak might be as much as 40 cents per
kilowatt-hour, but less than one-fourth of that during
the night-time off-peak hours. Capacity charges are col-
lected on a lower (about half the normal) price schedule.
But during potential peak periods (up to 50 hours per
year). the customer could face a capacity charge typical-
ly more than twice the normal peak monthly charge.
The hallmark of the pricing systems in France and
the United Kingdom is that they are pragmatic attempts
to have prices track costs. The principles articulated
above are utilized to design electricity tariffs. This prag-
matism is also reflected in the domestic charges in the
United Kingdom: since domestic customers are billed
quarterly in a staggered manner, seasonal billing is not
economic. The basic price is a flat charge of about 3
cents per kilowatt-hour. Previously, available off-peak
heating tariffs are maintained, but closed to new sub-
scribers. These offered customers 11 hours of off-peak
energy for heating only at about half the normal price.
Currently, an optional day-night tariff is offered with a
slightly higher daytime price (less than 5 percent), a
higher quarterly charge (about $2.40 per quarter) for
extra metering, but an 8-hour nighttime price of about
40 percent of the daytime price.
It is possible to learn from this European experi-
ence, but it cannot be directly transferred to our own
country and, therefore, we must consider how to imple-
ment time-of-day pricing in the United States. Some of
the problems are frequently mentioned by those con-
cerned with time-of-day pricing reform; the questions of
meter availability and cost, the impact 01 consumer
response and the so-called shifting peak problem, the
problem of allocating cost during a phase-in period
between those on the new tariffs and those not, and the
subsidiary question of how to deal with large volume
versus small volume users.
The question of metering availability is really a very
significant question for the smaller volume users. As
already noted, the cost of metering is trivial when com-
pared to the monthly bills of the large industrial users,
but for the smaller users we must proceed carefully and
make certain that we do not undertake tariff reform for
which the benefits to be derived are less than the costs.
But that is not our only question with meter availability.
There are also technical breakthroughs in metering on
the horizon that might make some of these currently
available meters seem quite primitive. A challenging
question for all to consider is whether or not we should
go forward with time-of-day pricing reform unless we
know when some of this new metering technology might
be available and what its cost might be.
The second problem area is predicting consumer
behavior. Implementing time-of-day pricing is not some-
thing you do and then forget. Load research and moni-
toring must be continued in order that, with any pro-
spective changes in use patterns, modifications in the
pricing practices on one schedule or another may be
utilized in order to help avoid creating new problems
and new costs. I n short, systems planners must be able
to use pricing incentives in order to find ways of helping
to hold down costs. In the final analysis, systems plan-
ners will pick the cheapest solution for all customers arid
for the utility as well when it comes to deciding what is
the best way to supply system needs. The introduction
of time-of-day pricing and the associated consumer
response to the calculations of the systems planner
would be a dramatic cost saving reform.
There is also a problem of allocating cost between
various customer categories, especially when some of the
users will be put on time-of-day pricing (typically the
larger volume ones). How will the revenue requirement
be allocated between these groups? What will happen if
one group generates too much revenue and another
group too little revenue? Those are really challenging
questions that the regulators must begin to tackle.
There are various implementation plans for the
large-volume users. Let us consider three options. First,
we might have a two-part tariff with customer charges
based upon the customer's maximum demand to recover
demand cost and a time-varying energy charge with
capacity costs of generation and transmission factored
into it on a peak kilowatt-hour charge. Second, we might
have a three-part tariff with a customer charge to recover
93
-------
the distributio/1 cost, a time-varying energy charge (with
the highest running cost plant during peak hours serving
as the peak I;ilowatt-hour charge), a capacity charge
based upon thl! hours of the day and perhaps months of
the year that consumption occurs. Third, an interrupti-
ble tariff with very large discounts is possible due to very
low capacity cost responsibility. Penalties for failure to
accept interruption, however, should also be high.
Time-of-dcY and seasonal discounts as a substitute
for industrial volume discounts make sense for customer
and utility alik~. Accordingly, it is important to demon-
strate this fact for industrial users by performing typical
bill analyses, ,s well as surveying these users through
statewide man,Jfacturing associations. Plans to do this
have' been tent atively made in several States. Consumer
understanding, even for large-volume customers, is an
important step.
The final c,uestion for large-volume users is to deter-
mine the speed of implementation. Several States are
moving to the position that cost-based tariff structures
(as the economist and engineer rather than the
accountant wOlild describe them) must be implemented
for industry, bllt there are still options present. First, all
large customers could be placed on cost-based tariffs and
their level could be constrained by current revenue allo-
cation formula j between the various customer classes.
Second, cost-ba5cd tariffs could be implemented and any
excess revenue mticipated could be used to reduce cus-
tomer cost ami off-peak prices in order to create the
maximum incentive for system cost reductions by keep-
ing all excess r ~venue in the class. Alternatively, other
classes could be subsidized if excess revenue were to be
generated. This latter step is probably illegal and it is no'!
good economic>, but it has political support in some
circles. Third, I)ptional time-of-day tariffs based upon
actual cost cOl..ld be offered, with future revenue in-
creases for the dass being placed disproportionately on
those who refU!e the option. Even if use patterns do not
change when metering costs are trivial, as they are for
the industrial class, time-of-day pricing is superior to the
existing accou ltant's approach as a cost-allocating
device. It is, thnefore, important to implement it to the
greatest extent possible for the industrial users. It will be
good for both t'1at user category and for the utility and,
therefore, for all customers alike.
For the smaller volume users, there are several
approaches, but in this category avoiding uneconomic
metering costs is a more significant constraint. The
various alternatives being considered for implementation
can be summarized as follows. First, time-of-day
schedules and r,eters and a sample of customers in the
category could be developed in order to find typical use
patterns by time-of-consumption, and then ,use these
data to develop a time-of-day rate for classwide bills. In
addition, an optional meter can be offered for those
customers who think they are atypical and can save
, money if they install meters at their own expense.
A second approach is to start with the larger small-
volume customers and implement time-of-day pricing for
them. If in retrospect the cost savings outweigh the
metering cost, the time-of-day pricing could be extended
to the next largest of the small-volume customers and
the analysis repeated. This process could be repeated as
long as the benefits exceed the cost. Third, tariff experi-
ments can be undertaken to measure the response of
customers to time-of-day varying tariffs and their will-
ingness to shift use away from peak. Such experiments
could complement the first two alternatives, but cus-
tomer compensation is absolutely essential in these
ex per iments.
However, for residential customers, plans to imple-
ment either of the first two approaches should not be
delayed by any pricing experiment. But only by pricing
experiments and load monitoring research, even after
new tariffs are introduced, can we determine how to
design tariffs and manage loads. Total class revenues can
be collected in many ways even with the time-of-day
adjustment. Only by such experiments can one find the
level of sensitivity that can best track actual costs,
manage loads, and have the greatest impact on reducing
the cost of generating electricity.
Finally, a tariff option can be offered in a wide
variety of formats even for the smaller volume users. The
fact that both the British and the French ultimately
went this route should not be overlooked. There are
several different forms of tariff at the low volume level
that can be considered. First, a two-part tariff with a
customer charge and a time-varying kilowatt-hour charge
could be implemented. A second alternative would be to
have a load-limiting three-part tariff with a customer
charge, time-varying energy component on a kilowatt-
hour basis, and a subscribed maximum demand with
charges varying for excesses at different times. Third,
less complicated load-limiting tariffs with seasonal
emphasis and with or without time-of-day va~ying
energy charges could also be considered. And finally,
various combinations based upon metering availability
and day-night and seasonal variability can be introduced
and considered.
It is important not to implement tariffs if the cost
of metering outweighs the benefits in utility cost savings.
This may be decided upon by the individual customer in
94
-------
the case of residential customers, but it is important to
go forward with time-of-day metering for industrial cus-
tomers, if only as a cost allocation improvement device
and because it will benefit both the utilities and indus-
try. Time of consumption discounts as a practical,
economically efficient and politically desirable alter-
native to volume discounts is a reform whose time has
come. When understood, its beneficial effects will be
widely embraced. The development and demonstration
of an applied methodology and implementation strategy
are now the tasks at hand.
Other Tariff Reforms
Time-of-day pricing reform, which in the eyes' of
utility executives responsible intervenors promote, is not
the only reform recommended by citizen activitists.
Some suggest inverted rate pricing or large-volume
penalties. Some recommend so-called lifeline rates and
some recommend public ownership. Reversing the cur-
rent practice of volume discounts and creating volume
penalties are sometimes offered as a panacea both be-
cause of the rising cost problems industry faces, and also
because it is fair. But reversing the "tilt" would be likely
to lead to large-volume user cutbacks or abandonment of
service that would result in higher prices for the remain-
ing customers. Additionally, any conservation that took
place under such inverted rates would erode revenue at
the most highly priced portion of the schedule. Unlike
time-of-day pricing, there is no guarantee that this re-
duced use would correspond to the highest costs of the
utility. Therefore, inverted rates are likely to lead to
serious earnings and revenue erosion problems. Thus,
rather than breaking the annual rate increase chain, they
are likely to extend it.
Another category of pricing reform is sometimes
proposed. It is the so-called lifeline rate. Its intended
purpose is to help the poor and the elderly. It would
make electricity available at a very low, or even zero,
price for a given level of use. Viewed as a tilt in the
direction of the poor, it would convert electric utilities
and public service commissions into welfare agencies. All
the problems of insuring that nonpoor do not benefit
would be present; e.g., adding additional meters in two-
family homes, luxury apartments, second homes, etc.
Economists frequently embrace the acronym
T ANST AAF L (there ain't no such thing as a free lunch).
which quite simply means that providing kilowatt-hours
below cost requires a decision to tax other customers or
stock holders to make up the subsidy. Such decisions are
not easily made by electric utilities or public service
commissions and, in fact, may be illegal.
I ncome redistribution is undoubtedly necessary but
it is the job of government, not private regulated indus-
tries. Legislatures and elected officials must do their
jobs, while always keeping in mind that what poor
people need is money. If a legislature passed a lifeline
tariff and provided the money for subsidy directly to the
utility or to the customer in the form of an energy
stamp, or explained who should pay into the fund to
finance the subsidy, electric utilities and public service
commissions would have less trouble with poor-people-
oriented or tilted electric rates.
Finally, in this regard, many of the current advo-
cates of electricity pricing reform, who have a pro low-
income stance, are missing a very fundamental point.
First, let me underscore the fact that declining block
volume discount pricing must be eliminated. Fifty years
ago electric utilities used to peak in the evening when
residential users of electricity all turned on their electric
lights. Most tariffs in the United States today are de-
signed with either high front-end separate charges or
very high prices for the initial kilowatt-hour of elec-
tricity consumed.
Proponents of volume discount or high front-end
pricing argue that this is the way to collect the fees from
consumers to pay for large generating and transmission
costs. But a new capacity is not built for a lighting load
in the United States. Instead, it is typically built for a
space-conditioning load. Further, a separate charge for
potential peak hours of electricity use to collect gener-
at ing and transmission capital outlays should be
adopted, especially for the large-volume industrial, com-
mercial, and residential users.
The low-income electricity user whose use is primar-
ily off-peak is paying a full share of generating and trans-
mission capacity at a time when his contribution to peak
is either zero or trivial. If we could meter time of use at
very low costs, this factor should, and with the reform
that has been advocated, would be recognized in a way
that would lower electric bills for the low income con-
sumer. Short of low-cost metering, participants in elec-
tricity rate proceedings should insist that all generatin~l
and transmission costs should be eliminated from the
customer charge and early block charges. In most States,
if only meter costs and specific customer-oriented dis-
tribution costs were permitted to be collected up front
and all other costs collected on a flat (or time-varying
basis), the reduction in low-volume consumer bills would
be staggering. Additionally, such a tariff is much closer
to the economists' notion of incremental cost pricing
than the tariffs currently in existence. Such a chan~le is
both more economically efficient and pro low-income,
and for these reasons the argument in favor of this step
is compelling.
When the future of energy prices is viewed, the po-
tential of a rapid reform of electricity pricing appears
relatively optimistic. We shall be well down the road
95
-------
towards redire"~ting the wasteful thinking of American
industry and consumers by providing ways to save
money and energy.
96
-------
OPPORTUNITIES IN ELECTRICAL LOAD MANAGEMENT
Douglas C. Bauer, Ph.D.*
Abstract
The importance of reliable and sufficient supplies of
electricity in our economy, coupled with the substantial
and increasing economic, energy, and environmental
costs associated with its production, make it essential to
insure that its development incorporates maximum prac-
ticable efficiencies. High among the underexploited
opportunities for efficiency is load management and its
complement, structural rate reform on the basis of long-
run marginal costs. The deterioration of load factors
(average load/peak load) has aggravated financial stresses
on the electric utility industry already burdened by
inflationary pressures that have caused marginal costs to
consumers to increase in constant dollars. Initial FEA
modeling work projects that load management could
have the effect of decreasing oil consumption by about
1.3 mi//ion barrels per day while maintaining a kilowatt-
hour sales growth rate of about 5 percent per year.
Capital savings in plant expansion would be about $48
billion. From a fuels management point of view, in-
creased reliance upon nuclear fuel and coal instead of oil
and gas could be accomplished.
Successive sections of the paper discuss the import-
ance of structural rate changes including peak usage
pricing, load management opportunities in space condi- .
tioning, heat storage, "cool storage," load deferrals, and
load control options. The FEA utility program goals and
features are described; these include a synopsis of
demonstration projects in 7 States (recently expanded to
another 12), the UCAN voluntary commitment program
with the largest utilities, and invited regulatory interven-
tions before local public service commissions.
BACKGROUND'
Electricity demand in the United States grew at
nearly 7 percent per year for most of the post-World War
" years. This increase is due primarily to increased per
capita electricity usage, rather than population, which
has grown at a rate of 2 percent per year and, more
recently, at less than 1 percent. In 1972, 25 percent of
the Nation's raw fuel supply was used to generate elec-
tricity. However, because of largely unavoidable ineffici-
encies in production (mostly in the power plant), elec-
. Associate Assistant Administrator, Utilities Program,
Federal Energy Administration, Washington, D.C.
tricity accounted for only 8 percent of the total energy
delivered to consumers (ref. 1).
For many decades, the U.S. was the leading ex-
porter of energy fuels. By 1950, however, the net
balance had shifted. In 1972, approximately 35 percent
of the U.S. oil supply was imported from foreign
sources. Twenty percent of that imported oil supple-
mented the 40 percent of U.S. coal supplies which were
used to generate electricity. The quadrupling of oil and
coal prices during the past year and a half has made the
generation of electricity much more expensive and has
accounted for most of the rise in electric bills.
The cost of future utility bills will depend on many
factors: the rate of inflation, the Nation's economic
health in general, the type of fuels used to generate elec.
tricity, and improvements in energy conservation and
utilization. The problems are complex and encompass
both energy supply and conservation, and domestic and
international sources of fuel.
Dwindling supplies of domestic oil and natural gas
for electricity production will require readily and abun-
dantly available substitutes. Power generated from coal,
nuclear energy, or hydroelectric installations is the most
usable available substitute for oil and gas. I n the long
term, the use of solar energy will be increased to satisfy
our energy needs, but adequate conventional electric
power will still be required.
The key to such a future, however, is a balanced and
~fficient growth for electric utilities. There remain sub-
stantial energy and economic inefficiencies in this sector,
not only in the generation of electricity, but also in its
transmission, local distribution, and end use. The elec-
trical utilities are also experiencing serious financial
stress resulting from rising costs of capacity and gener-
ator fuels and steadily deteriorating load factors (averag()
load/peakload). The load factor problem is particularly
important because capital requirements are driven by
peakloads, whereas revenues are derived from total load.
Deteriorating load factors also cause utilities to retain
'older, inefficient generators to meet peakloa,ds or to
acquire relatively inexpensive new peaking generators-
typically, simple-cycle combustion turbines inefficiently
burning scarce fossil fuels.
At the historic sales growth rate (nearly 7 percent)
the electric utility industry would consume 36.1 quad-
rillion Btu's (quads) of energy in 1985 and face probable
capacity shortages. A comprehensive energy-efficiency
program, reinforced by increasing energy prices and de-
clining growth of per capita income, could moderate this
97
-------
rate of sales g,owth to a manageable 5 percent (slightly
lower than the most recent Edison Electric Institute
forecast) with an associated 1985 energy consu mption
of 30.7 quad:;. Finally, an aggressive load-management
program could disengage and further reduce the growth
of peak kilowatt demand from this moderated growth of
kilowatt-hour sales.
Initial F E ~ econometric modeling work projects the
following 198:; consumption of generator fuels (quads):
Coal Oil Gas Nonfossil Total
--
High growth 12.3 8.0 3.7 12.1 36.1
Moderate growth 9.7 5.8 3.3 11.9 30.7
Load managen lent 10.9 3.0 3.4 13.4 30.7
As can bl! seen, load management could have the
effect of decrfasing oil consumption a full 2.8 quads, or
about 1,300,000 barrels per day, below moderate
growth while simultaneously maintaining sales growth
and increasin£ nuclear and hydro generation by 1.5
quads above moderate growth and 1.3 quads above the
high growth Cilse. It also has the potential to save about
$48 billion in capital for plant expansion capacity (see
figure 1 a-f) n lative to Moderate Growth and $140 bil-
lion relative to High Growth (historic). .
In addition, electricity's share of total delivered
energy could increase from 11 percent in Moderate
Growth to 13 percent in Load Management, largely by
end-use fuel switching from oil and natural gas to coal-
and nuclear-generated electricity. This end-use fuel
switching, which would conserve an additional 0.4 quad
of scarce fossil fuels, is caused by an 8 percent reduction
in the real dollar price of electricity, a reduction made
possible by increased capacity utilization.
ELECTRIC UTILITY RATES
I ncorporated into any load management scheme is a
basic restructuring of electric utility rates. Utilities liave
typically structured their rates on the basis of "fully
allocated" or fully distributed average costs. This basis
led to declining block rate structures; that is, the price of
a kilowatt-hour decreased as the number of kilowatt-
hours used increased. Such rate structures rarely, if ever
differentiate between on-peak and off-peak usage", even
."Peak usage" refers to the point in time of highest demand
on a utility system. It may be measured daily. weekly, monthly,
or yearly. In a daily increment. peak might refer to the period of
time between 4:00 p.m. and 7:00 p.m. when customers demand
the most energy.
HISTORICAL
BAUa
LOAD
MANAGE-
MENT
1975
2000
aSusiness as usual (Edison Electric Institute Forecast, June 1975).
Hgure 1a. Electrical generation and use: End-use conservation, high electricity
prices, appliance saturation, and demographics are expected to reduce
historic growth rates for capacity, and load management can reduce that
fu rther.
98
-------
S318
co
co
7%
Sales
Growth
$226 SAVINGS
5%% -0$48
Sales S178
Growth 5%%
Sales
Growth
HISlOR ICAL
"GROWTH
BAU
LOAD MGMT.
SOURCE: FEA
Figure 1 b. Capital requirements between now and 1985 are substantially reduced.
Historical vs. BAU vs. load management.
(billions of U.S. dollars)
-------
--
o
o
."..
,..
,/
,/
./
./
/'
--
-
-
,..
,/
TIME OF DAY (SUMMER)
SEASON
"-
./
/
/
,//
NOON
6 p.m.
JANUARY
AUGUST
DECEMBER
Figure 1 c. Load management reduces peaks. How? Peak rates, load control systems,
storage devices. We are now funding seven State demonstrations.
-------
......
LOAD MANAGEMENT CAN ALSO REDUCE
USE OF SCARCE FUELS
....
o
....
....
/ HYDRO ......
......
/ ......
// "", """
......
/ / ...... ......
..... -----
/ / ....... HYDRO
NUCLEAR ...... HYDRO
/ / .......~
/ / ------
/ /
/ /
/ / NUCLEAR NUCLEAR
/ 1 ",/', N.GAS --
/1 -
/ .........
/ "" ,/ 1', '
HYDRO / / ./ ......... '-
/ / ......... -----
/ OIL ' N.GAS N.GAS
NUCLEAR ./ ./ '-
./ - - - -- -
/ OIL
N. GAS ./"" OIL
,,'" . - -- -
-
" -- ---
OIL """" - --
.....
"
COAL
COAL COAL
COAL
1973
1985 .
HISTORICAL GROWTH
1985
BAU
1985
LOAD
MGMT."
.. A88uming BAU (~uain8S8-as-u8ual) price 01 electricity.
SOURCE: FEA
Figure 1 d. Load management can also reduce use of scarce fuels.
-------
915,000
CA_PAC!TY
(MEGAWATTS)
rADI\r'ITV! ITI. 1""71\T'Ir\fIol
-,--.. '-\_1 I I V I I L-1L..r"'\ 11\..11'4
(CAPACITY FACTOR)
--
o
'"
7%
Sales
Growth
783,000
5%%
Sales
Growth 673,000 57% 57%
5%%
Sales
Growth
49% 49%
461 ,000
1974 1985 1985
ACTUAL HISTORICAL BAU
GROWTH
1985
LOAD
MGMT.
1974
ACTUAL
1985
HISTORICAL
GROWTH
& BAU
1985
LOAD
MGMT.
1973
WEST
GERMANY
Figure 1e. A load management program reduces future capacity requirements
by improving plant utilization.
-------
ELECTRIC UTILITIES SUMMARY
.
LOAD MANAGEMENT IS THE KEY UTILITY
CONSERVATION EFFORT-MAYBE THE
SINGLE MOST IMPORTANT CONSERVATION
MEASURE
SAVES $48 BILLION OF INCREMENTAL CAPACITY.
ENCOURAGES COAL AND NUCLEAR BASE-LOAD
CAPACITY SA VI NG 1,300,000 BPD OF 01 L.
ENCOURAGES END-USE FUEL SWITCHING FROM
OIL AND NATURAL GAS.
Figure 1f. Electric utilities summary.
though it is clearly the former which is forcing the
additional cost of capacity expansion. These rate struc-
tures worked fairly well in the past when the incre-
mental cost of capacity was lower than the average cost
of installed capacity (ref. 2). Currently, incremental
costs exceed average costs. Rate structures do not reflect
this fact and, therefore, fail to communicate proper cost
signals to consumers.
I n the absence of more refined pricing patterns and
other load management tools, the industry attempted to ,
simultaneously expand capacity to meet rising demand
peaks while promoting the development of new loads
off-peak. Utilities with sum~er demand peaks often
promoted electric resistance space heating to fill in their
winter valley while continuing to promote air condition-
ing as well. Winter peaking systems, similarly, tended to
promote air conditioning.
MARGINAL COST (peakload Pricing)
Utilities should now begin to examine pricing
policies to insure that their rates adequately reflect the
true marginal or incremental costs of service-particu-
larly peak-demand costs in the case of electricity-and to
provide the proper economic signals to consumers.
Prominent economists agree that there is virtually no
argument in determining whether or not marginal cost
pricing is correct: as a matter of economic principle it is
correct by definition. In the case of electric utility
companies, rates must be based on marginal costs if
economic efficiency is to be achieved (ref. 3). Dr. Ralph
Turvey, former chief economist at the Electricity
Council of the United Kingdom, stated in reference to
cost-based rates that "it is better to be approximately
correct than absolutely wrong':'
Moreover, there is virtually no argument that mar-
ginal costs are of great significance in the case of electric
utility rates. It is generally accepted that the cost of
providing electricity varies significantly with the timing
of demand. Additional generating capacity must be built
in order for utilities to meet peak demands. In addition,
system-running costs are higher during peak than during
valley periods. The incremental fuel cost is twice as high
on-peak as it is off-peak (ref. 3). As Dr. Turvey wrote
(ref. 4):
. . . rate structure is very important. (It) influences
consumer's behavior as to when and how much they use
electricity. When consumers use electricity is equally as
important as how much they use during a billing period.
Like any other set of prices, the rate structure has in-
centive effects on the buyers: a low rate for ex tra pur-
chases encourages consumption and a high rate dis-
courages it (ref. 5).
The objective, therefore, is to provide consumers the
correct economic signals for consumption of electricity:
give them the incentive to economize when the cost to
serve is high and to consume when th.e cost to serve is
low. Berlin states that "marginal cost pricing alone per-
mits consumer sovereignty to achieve the allocation of
resources that is most consistent with the best interest of
the consumer, the company and the public interest (ref.
5).
Accordingly, electricity prices based upon marginal
costs would feature the following characteristics:
1. The cost of incremental (future) capacity
would be borne primarily by peak usage, i.e., there
would be an off-peak discount.
2. Declining block rates would be eliminated,
except to the extent strictly justified by fixed customer
service costs.
The resultant rate structure would have essentiall"
three parts: a usage (per kWh) charge, a demand (pea~
system kW) charge, and a service charge. Such a rate
would be based upon the best available approximation
of actual incremental, prospective costs and would pro-
vide the incentives to move loads to off-peak periods and
build new loads during such periods.
Potential benefits to be achieved through an incen-
tive system of this type are enormous. In 1973 only 49
percent of installed generating capacity was utilized. If
this capacity factor could be increased to 57 percent by
1985, assuming a 5-1/2 percent annual growth rate for
kWh usage, the need for installed capacity in 1985
would be reduced by 110 million kWh. This translates to
an estimated capital savings of approximately $56 billion
in 1975 dollars, which would substantially moderate the
103
-------
need for rate increases. Moreover, the costs of installing
the time-of-day meters ($6.5-$13 billion ") required for
true peak load pricing may now be justified by the bene-
fits, particularly if the technology also embodies remote
meter reading and selective load control capability.
Utilities and other systems with rapidly escalating
peaks now arE' expressing considerable interest in load
management tl!chniques, which are recognized as a tool
for improving system efficiency and limiting the need
for capacity e>:pansion. State and local regulatory auth-
orities, faced with consumer opposition to rate increases,
environmental.st attacks on utility expansion, and
Federal encoUI agement of energy conservation, are also
evidencing int,~rest in strategies to shave peaks and
improve system load.
LOAD MANAGEMENT POTENTIAL
As previolsly discussed, total U.S. kWh usage has
been growing c t an average annual rate of slightly less
than 7 percer, t over the last several decades. Peak
demand has been growing at slightly more than 7 per-
cent during the same time period. Although both growth
rates have been substantially lower since 1972, the pat-
tern of more I apid growth in peak than in usage has
become even more pronounced. This has led to a steady
deterioration i 1 load and capacity factors, thereby
. --
exacerbating ut lity financial problems and forcing larger
rate increases.
These diffil:ulties, in turn, have increasingly forced
utilities to purc lase low initial cost-peaking generators-
typically burnir g distillates or natural gas at inefficient
heat rates-to m~et peak demands.
A more b31anced growth pattern for electricity
usage and peak demand would be in the mutual interest
of such diverse ,]roups as regulatory authorities, environ-
mentalists, con:.umers, and utility executives. Such a
balanced growtl' pattern would cut the consumption of
scarce fossil fuels in electricity generation, moderate the
need for constrL ction of new capacity, reduce the pres-
sure for rate increases, and improve utility revenues. It
would also create a more stable base for future coal,
nuclear and hydrogenerated electrification of the econ-
omy, as an altel native to the direct combustion of gas,
oil and oil produ :ts.
In orde'r to reduce energy consumption within the
Federal Government, the Federal Energy Administration
developed energl/ use guidelines in January 1974, This
Federal Energy Management Program was responsible
for setting lightill9 and heating standards in government
buildings and speed limits on government vehicles, to
"Assuming $'JOO-$200 per customer for 65 million cus-
tomers.
name just a few. The Smithsonian Institution rigorously
followed the guidelines and reduced their energy COil-
sumption 22 percent in the first year. The Institution
recently installed a small process computer to provide
them with a load management capability. As a result,
energy consumption has been reduced an additional 16
percent; peak has been reduced by 17 percent. This one
example certainly presents a convincing argument for
load management.
Several avenues are open, which encourage balanced
growth patterns for electricity usage and peak demand.
Among those most frequently cited are: rate structures
that more closely reflect daily and seasonal variations in
the utility's cost to serve, and load management tech.
niques such as system control of deferrable loads and
heat storage,
Load Management: Space Conditioning
The increasing saturation of electric resistance heat-
ing in winter-peaking, systems and air conditioning in
summer-peaking systems suggests that the management
of these loads is becoming increasingly necessary. in
1966, only 20 percent of new single-family dwell ings
were constructed with electric heat. In 1974, the figure
was 50 percent (ref. 6). This type of electric heating is
typically direct resistance heating with no facilities avail-
able for heat storage. Electric air conditioning has exper-
ienced a similar growth rate: in 1956, 6 percent of new
homes were built with central air conditioning, while in
1974 the figure rose to 36 percent (ref. 7).
While it may be argued that peak loads are partially
a result of increased use of electric space conditioning,
the fact remains that electric space conditioning greatly
demands electric energy during system peak. Moreover,
this contributes to a waste of primary fuels due to the
bringing "on-line" of older, less-efficient peaking units.
As a result, the growth of electric heating in winter.
peaking systems and electric air conditioning in sum-
mer-peaking systems is causing deterioration of annual
system load factors.
Avoidance of electric space conditioning during
peak periods can be achieved by either interruption of
demand or by deliberately designing the system load to
draw electric power during off-peak hours. Interruption
of residential demands involves the cessation of electric
power for periods of time of up to 3 hours. One method
of achieving this is through the use of "ripple control:'
an audio frequency signal sent from the power station to
the residential unit. (I nterruptable loads will be dis.
cussed in more detail in the following section,) An alter-
native approach to peak avoidance involves the storage
of heat during off-peak hours for use during the re-
mainder of the day.
104
-------
Hea t Storage
Electric storage heaters consist essentially of a cen-
tral core of magnesite bricks, which are interspersed with
electric heating coils. The bricks are heated to as much
as 1,300° F during the night. The stored heat is then
drawn from the units as needed during the next day (ref.
8). The storage units range in size from a large, central
unit to an individual, room-sized one. The typical char-
gin times are 6, 8, or 10 hours, depending on the particu-
lar unit. Heat is dispersed in various ways, depending
upon the storage unit. In a "static" heat storage unit, the
stored heat leaves the device by radiation and convection
from the case surface. This results in high surface tem-
peratures, but poses no safety hazard. This system is
designed to maintain an average space temperature dur-
ing sixteen hours of discharge time, with no ability to
respond to changes in heat demand on a shorter time
interval. The unit also experiences a slow decrease in the
heat release rate as the storage core temperature de-
creases. There is no way of forcing the heater to emit
heat at a more rapid rate.
Dynamic heat storage units (fan assisted) are used
where greater flexibility is required for local space heat-
ing than can be provided by static storage heaters.
Dynamic storage units operate on a forced convection
principle and use fans to draw room air into the radiator
where it is heated in channels contacting the heater core.
Room temperature is maintained by a thermostat, which
controls the fan. This unit is much more flexible in con-
trolling room temperature than the static unit.
Space heating using hot water distribution of stored
heat is similar to a hot air system except that the air
heated within the storage core is contacted with water in
an air-water heat exchanger. The heated water is then
pumped throughout the residence where it gives up its
heat to room air in conventional-type hot water radia-
tors. A uniform heat release into the residence is main-
tained by keeping the water circulating at a constant
flow rate and temperature.
"Cool Storage"
Air conditioning load management is a relatively
uncharted field. The European countries have gained
expertise in heat storage in order to manage their
winter-peaking' systems. The United States, on the other
hand, is plagued by summer peaks and must therefore
investigate the' potential in storing "cool" as a load
management technique.
The two principle techniques for managing electric
heating (interruptible electric energy supply during peak
and heat storage) cannot be as simply applied for air
conditioning management. For example, interruption of
electricity supply during peak results in a more rapid
temperature change in summer than in winter. This is
partially due to the combination of internal heat with
solar heat which results in rapid temperature gain while
the air conditioning is shut off. (During the winter, these
conditions aid in reducing a drop in temperature as elec-
tricity is interrupted.) However, temperature drop alone
in summer is insufficient to create a comfortable living
environment; humidity must also be taken into account.
For these reasons, interruption of the air conditioning
load is not completely satisfactory.
Many methods for the storage of "cool" have been
proposed; however, the most promising alternative is the
storage of cold in the form of chilled water using con-
ventionaf air conditioning equipment (ref. 9). The con-
ditioning of the air space is accomplished by circulating
chilled water from storage as required. I n this way, suf-
ficient chilled water can be stored by operating the air
conditioning system off-peak and the cooling of a build-
ing space is done with stored chilled water using no
energy during peak.
Although this system is ideal for shifting peak day-
time loads to off-peak hours, a considerable economic
investment is required in both the storage system and in
the air conditioning system, which must be able to
accomplish an entire day-time cooling load in a 8-hour
off-peak time period.
Deferable Loads
One basic aid in load management is the existence
of a deferable load such as an electric portable water
heater. I n the U.S., household water heaters represent
the only load segment having a substantial energy stor-
age capacity. The volume of a typical water heater tank,
50 gallons, is sufficient to avoid any great inconvenience
to the customer under normal usage, if the heater is
disconnected from the supply for only a few hours.
Since the average demand of a water heater during peak
is approximately 1 kW, a utility with 200,000 electric
water heaters under control could reduce the load during
peak by 200 MW (ref. 10).
Control Options
Any discussion of load management techniques such
as interruptible loads, heat storage, and time-of-day
metering does not proceed very far without mentioning
the control aspects of such devices.
The clock option is the simplest approach; each
individual load unit is turned on and off automatically
by a clock-actuated switch. This method is currently
being used by a few U.S. utilities to control electric
water heater load. The customer acceptance of the con-
105
-------
trol can be se,:ured by reduced rates for off-peak water
heating.
In anothe' approach, water tanks are designed with
two heating ~Iements, each equipped with its own
thermostat. In the tank, cold water enters at the bottom
while hot waler is drawn from the top. One heating
element is nea' the top of the tank, set at 13cf F, while
one heating eh!ment at the bottom of the tank is set 10°
F higher. The upper heating element is permanently con-
nected to the )ower supply, while the lower element is
disconnected hy a clock-actuated switch during preset
time periods. During these times, power is used to heat
water only whIm the temperature in the (normally) hot-
test top layers of the water falls below the setting of the
top thermostat
The main ,Idvantage of systems using control clocks
is their simplicity, which is reflected in the relatively low
cost of the har:lware. The present cost to the utilities of
the basic residl!ntial single-phase watthour meter with a
built-in, clock-ictuated, on.off switch is nearly $50. The
register watthour meter with a built-in clock-controlled
switch used in conjunction with the two-element water
heater costs abetut $75.
The main problem facing systems with a large num.
ber of control clocks is lack of flexibility and a low
degree of reliability. For example, clocks often need
readjusting. To make a change in the on-off times of the
controlled loac, each clock must be manually reset,
which is a time-consuming, costly process. Moreover,
clock-controlled load management systems are not suit-
able if adjustml'nts in the control schedule are called for
by seasonal chilnges in the load curve or by changes in
switching from;tandard to daylight saving time.
Another indication of the inflexibility of such sys-
tems is that clocks normally operate on a 24-hour cycle.
The load is turned off even if this is unnecessary, such as
on weekends and holidays. Since the clocks run on line
power, power oJ1tages will delay the on-off times of the
load. An extended power outage in an area involving a
large controlled load may shift the on-period in such a
way as to pose a serious problem for the utility. This
situation is imploved by providing the electrically driven
clocks with a mechanical (spring) backup. The addition
of a 10.hour carryover adds $20 to the cost of the
meter.
Ripple control systems are in widespread use
throughout Western Europe, Australia, New Zealand,
and South Afric~. The system was first tested in Europe
over 50 years ag). The technology has steadily improved
since then and lias gained wide acceptance. fiy the end
of 1972, networks in Europe, Africa, Asia, and Australia
were equipped with audio frequency equipment, con-
trolling approximately 5 million receivers.
In this method of load control, a coded narrow-
band audio frequency signal is introduced into the
power grid at the intermediate or high-voltage level. This
signal conveys switching commands to receivers, which
are placed at desired locations in the network, usually at
the low-voltage (consumer) end.
The main appeal of ripple control load management
systems is flexibility and reliability. Depending upon the
type of sending-end equipment and the coding scheme
used, the number of different command messages ranges
to thousands, while the message transmission time may
be as short as a few seconds. The messages can be sent
manually whenever necessary, or automatically accord-
ing to any predetermined schedule. As a further step in
the degree of control, a ripple-controlled network can be
turned into a closed-loop system by using a computer to
monitor the network status via hard-line connections to
the appropriate network points, and to initiate auto-
matically the transmission of the required command.
In Europe, one of the most important uses of ripple
control is the control of chargins times on electric stor-
age heaters. However, ripple control signals are used not
only for load management purposes, but also for a
variety of other functions. These include control of
lighting (streets, monuments, public buildings, and dis-
play windows), remote switching of multiple rate
meters, etc. Since the injected control voltage is present
at every wall socket, ripple control equipment can in
principle also be used to transmit alert signals to volun-
teer firemen, policemen, members of the National
Guard, etc. In addition, ripple control signals could be
used by the utility in the execution of various functions
in its own distribution network, such as the shifting of
loads from one feeder to another, the closing and open-
ing of circuit breakers, the switching in and out of capa-
citor banks, the switching of voltage control devices, etc.
To cover the same number of load units, the ripple
control hardware is considerably more expensive than a
clock:controlled system. This is due to the need for
individual receivers which decode the control message
and provide the switching function. The ripple control
system requires a central control facility and a large
number of transmitters to inject the control signal into
the network. A hypothetical utility with 100,000 cus-
tomers and an average of 5 k VA of transformer capacity
pe r cu sto mer would have to pay approximately
$500,000 for the injection equipment to cover the net-
work. If this equipment were installed to control 20,000
water heaters, the ripple control receivers would cost an
additional $2,000,000. Thus, the total cost per customer
would average $125. The cost of any new metering
equipment, or the modification of the existing meters
would have to be added to these figures to arrive at the
106
-------
total cost of the ripple control system. For the sake of
comparison, the cost to replace 20,000 basic residential
meters with meters incorporating a clock-actuated on-off
switch with 10-hour carryover would be $1,000,000
(ref. 8).
Radio control, the principle of transmitting control
messages via radio waves, offers even more possibilities
than ripple control. This is because, in contrast to ripple
control, which is limited to the low frequencies not far
about the 50 or 60 Hz for which the network has been
designed, radio transmissions could be made at much
higher frequencies where wide bandwidths, and thus
. higher information rates, could be employed. As a prac-
tical matter, it must be recognized, however, that the
electromagnetic spectrum is already overloaded with
users, so that the designer of a radio control system
works under severe limitations when considering the
choice of the system's operating frequency and band-
width.
The principle of radio control is simple. Radio trans-
mitters are placed in suitable locations (or in vans, air-
craft, or helicopters) to cover the network by a suf-
ficiently strong signal. The transmitted waves are modu-
lated by some code containing an "address" for selecting
the desired receiver groups, as well as the control mes:
sage itself. At or near the location of the load under
control, a radio receiver detects and decodes the message
and performs the switching function as in the case of a
ripple control receiver.
The Motorola Company has developed a radio-
controlled load management system. The system is in
use by the Detroit Edison Company to control the load
of 200,000 water heaters. An additional system is now
being installed, also for control of water heaters in the
area served by Buckeye Power, Inc., of Ohio. According
to a recent article in Business Week (ref. 1). installed
cost of the system used by Buckeye Power will average
about $90 per customer. For control of water heaters in
areas having a relatively high customer density, the
Motorola system appears to be an attractive alternative.
On the other hand, it is probably true that for the execu-
tion of other command functions (street light, multi-
register meters, electric storage heaters). ripple control,
with its separate on and off commands and its greater
immunity to interference, would be preferable.
Other possibilities include phone service, which is
provided to a large percentage of households. It has been
suggested that phone lines be used as a one- or two-way
medium for electric load management signals. There is
no question that this would be technically feasible, but
the interfacing of the phone and load ma'nagement sys-
tems may turn out to be more expensive than some of
the alternative control methods. There is also the prob-
lem of incomplete coverage of the electrical load and the
questions of which utility should have priority over the
lines. These problems make it unlikely that phone lines
will be used for load management purposes in the fore-
seeable future. The same applies to the possible use of
the one-way cable systems, which are now used for the
transmission of TV and FM signals. Two-way cable tele-
vision (CATV) systems would offer a tremendous range
of possibilities, including automatic meter reading, but it
. is clear that the practical implementation of the so-called
"wired city" idea still is far in the future.
Finally, very little energy is consumed in the opera-
tion of load management systems. Control clocks use
about 3 watts per "clock continuously, i.e., about 69 kW
in a system using 20,000 clocks. Ripple control signals
are injected at about 3 percent of the line voltage, i.e., at
about 0.1 percent of the system power. The transmission
of ripple control signals thus requires about 1 kW for
each MW of load. Contrary to the clocks, this power is,
however, used only during the actual transmission of the
control signals. I n radio control, the power requirements
are even smaller than in the other control methods.
FEA'S UTILITIES CONSERVATION
PROGRAM
The Federal Energy Administration is attempting to
improve energy efficiency in the utilities sector by cul-
tivating the mutual interests of utilities, regulatory
authorities, consumers, and environmentalists in this
area. On the one hand, we must insure the capability of
utilities to supply adequate electricity to the Nation at
reasonable prices. On the other hand, we must conserve
energy by minimizing the inefficiencies and wastages,
which occur not only in the consumption of generator
fuels, but also in the ultimate consumption of electricity
itself. Although in some limited respects, these two'
objectives may conflict, we believe that this is not gen-
erally the case and that energy conservation programs
such as load management will not complicate, but actu-
ally enhance, the financial situation of electric utilities.
Cultivating this overlap of interests, FEA conserva-
tion programs for electric utilities will emphasize field
demonstrations, regulatory intervention, and encourage-
ment of regulators, utility leaders, and consumers to
adopt energy conservation/load management practices.
Assuming sustained mutual commitment to positive
action, such a voluntary program could achieve the fol-
lowing electric power objecti,!es by 1985:
1. I mprove capacity factor (average/capacity) and
load factor (average/peak) from the present 49 percent
107
-------
and 62 percent respectively, to 57 percent and 69 per-
cent;
2. Expand coal and nuclear generation from the
present 50 percent of kilowatt-hour output to 62 per-
cent;
3. Redu( e the use of oil in electric power genera-
tion by at least 1 million barrels per day; relative to
business-as-usual projections;
4. Incrca;e end-use efficiency of electricity con-
sumption by at least 10 percent;
5. Encourage end-use fuel switching to efficient
electricity applications.
In the case of gas utilities, national energy conserva.
tion goals for ':r.e period 1975-85 are not so complex,
although the need is more pressing:
1. Impro\'e end-use efficiency by at least 10 per.
cent;
2. RedUCI! nonessential consumption by elimi-
nating 60 percent of outdoor lighting and installing elec-
tronic igniters ill lieu of pilot lights on 45 percent of gas
furnaces, ranges, and water heaters;
3. I mplentent emergency curtailment plans as
necessary to co~ e with near.term supply shortfalls.
OFFICE OF UTILITIES PROGRAM
The Office of Utilities Programs within Energy Con-
servation and Ewironment has the primary responsibil-
ity for developing, managing, and coordinating all energy
conservation prcgrams for electric and gas utilities. The
office is divided into two suboffices, one of which is the
Office of Utilitil!s Policy and Demonstrations. Its mis-
sion is to gather facts and formulate recommendations
affecting the energy efficiency of the conversion, trans.
mission, and distribution processes as well as end-use
consumption pra';tices.
The office h 3S also begun to enter into cooperative
agreements with agencies of State and/or local govern-
ments to support "demand management" projects for
electric utilities. These demonstration projects highlight
certain options ilvailable to regulatory authorities and
individual utilitie'; to conserve energy in the generation,
transmission, local distribution and end-use of elec-
tricity. For example:
1. The Ariwna Fuel and Energy Office is develop-
ing and testing in.'lovative rate forms to give the residen-
tial consumer an incentive to defer electrical demand to
off-peak periods. These rates would be based on time-
of-day pricing and a summer/winter pricing. differential.
Load scheduling s'/stems and hardware will be utilized to
assist the custome - in managing his electrical demand.
2. The Ark,lnsas Public Service Commission will
monitor the consumption practices of residential, com.
mercial, and industrial customers in response to UIl-
announced changes in utility rate schedules, such as flat-
tening the (historical) declining block rates, time-of-day
rates, and an increased summer/winter differential.
3. Investigations by the Los Angeles Department
of Water and Power into alternative rate structures that
may be superior to those currently in use in the U.S. will
quantify the effects of peak.load pricing on residential,
commercial, and industrial customers; the administrative
and technical feasibility of alternative rate structures;
a~d adaptability of this system to utility systems in dif-
ferent geographic locations.
4. The Connecticut Public Utilities Commission
will examine residential consumer acceptance of peak-
load pricing. Such pricing will reflect the long-run
incremental costs of producing electricity during heavy
demand.
5. A study conducted by the New Jersey State
Energy Office will measure the economic reaction of
residential customers to a time-of-day rate structure. The
consumption behaviors of three groups of customers will
be investigated:
residential units with electric water heaters;
residential units without electric water heaters;
residential units with electric water and space
heaters.
The Office will also conduct a pilot demonstration
of an automatic remote meter-reading and control sys-
tem to shed electrical loads during peak demand.
6. The Public Utility Commission of Ohio will
randomly select 100 residential units and use multi-
channel recording devices to establish demand patterns.
Data will be collected under conventional electric rate
schedules, time-of-day metering, and under remote con-
trol of electric water heaters, space heaters, and air con-
ditioners. Computer models will be developed to esti-
mate the probable effects of various load management
options on the utility system's load.
7. A utility demonstration project in Vermont will
test rate designs incorporating cost incentives, disincen-
tives, and physical controls, which affect the tir,ne and
magnitude of electric energy (consumption) require-
ments. The Green Mountain Power Company will design
rates and install equipment to test the benefit/cost rela-
tionships. contribution to system peak, and feasibil ity of
off-peak and interruptible heating systems in comparison
to conventional electric residential heating systems.
The other suboffice, the Office of Utilities Policy
Implementation, is responsible for having appropriate
conservation policies and practices adopted by utilities,
including practices based on the outcome of the conser.
vation demonstration conducted by the Policy and
Demonstration Office. A .program sponsored by Policy
108
-------
~
Implementation that has recently gained notoriety is
"Utilities Conservation Action Now" (UCAN). UCAN is
a voluntary, yet structured, campaign to insure open,
two-way communication in the development and imple-
mentation of national energy policy, to elicit "affirma-
tive energy action" commitments from key institutions,
and to provide national coordination and other support
for these individual institutional efforts. I nstitutions to
be directly involved in this program include the 100-150
largest gas and electric utilities, State and local regula-
. tory commissions, national consumer and environmental
groups, and major utility trade organizations. The in-
volvement of these diverse groups should insure that
UCAN activities make sense in light of all pertinent
needs, constraints, and opportunities.
UCAN's challenge is not directed solely at the
American utilities, which have traditionally provided
safe, reliable service at reasonable rates. Rather, it is to
all interest groups whose activities bear upon utilities
matters and who have the potential of making positive
contributions to the solution of the Nation's energy
. pr.oblems, to cooperatively begi!:'. the work of building ~.,
better energy future. The challenge to the Federal
Energy Administration, as national UCAN coordinator,
is to stimulate mutual education through two-way com-
munication; to provide supporting policy guidance and
materials; to reconcile fundamental differences when
possible; and, perhaps most importantly, to insure that
UCAN action plans are responsive to regional and local
needs, while serving the national interest.
The UCAN campaign will be an ongoing FEA activ-
ity consisting of the following program elements:
1. Information materials and workshops designed
to familiarize participants with policies, practices, and
available technologies supportive of national energy
goals. These will be handled on a regional basis. .
2. Development and adoption, by all participating
utilities and other groups, of UCAN action plans featur-
ing organizational commitments to specific energy-
efficiency measures, quantified objectives, and imple-
mentation schedules.
3. Periodic and special reporting to FEA by all
participating organizations of progress in implementing
the Action Plans, together with requests for FEA sup-
port as necessary.
4. FEA intervention in State and local regulatory
hearings to clearly establish national energy policy with
respect to utilities and to advocate specific actions on
the part of utilities and/or commissions to achieve na-
tional energy policy goals. .
5. Ongoing national and regional publicity of
UCAN purpose and progress, including degree of support
received from utilities, regulatory commissions, and
periodic Federal awards for exceptional performance.
Related Activities
Utilities conservation policy in FEA is directed at
the development of guidelines, incentives, and regula-
tions to improve the utilization of generating equipment,
promote higher conversion and transmission efficiencies,
and moderate future growth in demand. In cooperation
with FEA, several other government agencies are also
actively pursuing utility-related research: The Energy
Research and Development Administration, the Federal
Power Commission, and the National Science Founda-
tion are just a few.
Paralleling government-sponsored research, several
State public utility commissions have taken the lead in
major load management studies. PUC's in New York,
California, and other States have begun holding generic
meetings on the design of rate structures with emphasis
on load management. Moreover, the Wisconsin PUC has
directed a major utility to implement full peilk-Ioad
pricing on specified existing industry and commercial
. accounts. Finally, the Edison Electric I nstitute, in joint
with EPRI, is examining rate design and- load- manage:"
ment techniques to lessen electric system peak demand
growth and, thus, contribute toward the conservation of
capital. .
While some aspects of the problems examined in the
above studies may be unique to the utility system
examined, the studies should indicate potential solutions
that, once modified, could be applied on a nationwide
basis to solve common problems experienced by utilities.
._.~--. -.------------
..._~-----
REFERENCES
1. S. David Freemanan, Energy: The New Era, August
1974, p. 28.
2. National Economic Research Association paper,
Energy Conservation Through Modified Practices
and Standards in Energy Consumption and Use, p.
31.
3. Edward Berlin Testimony, Michigan Public Service
Commission Brief pm Case No. U-4570, p. 6.
4. Private communication with Dr. Ralph Turvey, .
April 1975.
5. Edward Berlin Testimony, Michigan Public Service
Commission Brief pm Case No. U-4570, p. 7.
6. Committee on Challenges of Modern Society, Brus-
sels, Load Management Conference minutes and
commentaries, November 1974, p. 117.
7. Bureau of Census, Characteristics of New One-
109
-------
IIIII/llv IIUIlItJS, I'.S. I>UPWIIIU:III 01 <:OIllIllIllI:U,
05-/013, Hf/2.
8. V(~rmOI1I Electric Utility Load Management Pro-
(Jram, First Year Report, February 1975, NSF Grant
G141471.
U. Gonliill1 Asso(;iali~s, Elm:tri/: Utility Load, mlll1i1!J1)'
nmnt mpurt, April B, 1075, p. 92.
10. IEEE Spectrum, February 1975, p. 46.
11. Business Week, August 24,1974.
110
-------
POLICIES TO REDUCE TRANSPORTATION FUEL USE
Eric Hirst, Ph.D. *
Abstract
This paper reviews historical t~ends in passenger
traffic and energy use since 1950. Overall, transportation
fuel use grew from 8.9 OBtu in 1950 to 18.3 OBtu in
1974, with an average annual growth rate of 3.0 percent.
Energy use grew more rapidly than did traffic during this
period because of shifts from energy-efficient to energy-
intensive modes and increases in energy intensiveness for
most modes (due both to declining load factors and
reduced vehicle fuel economy).
A number of altematives exist for reducing trans-
portation fuel use. This paper discusses the energy sav-
ings possible due to expanded and improved urban mass
transit services, increases in new-car fuel economy, in-
creases in the price of gasoline, and increases in com-
mute-auto occupancy (carpooling). Expanded mass
transit is likely to save only small quantities of energy
during the next 5 to 10 years, primarily because of the
very low fraction of urban passenger travel now carried
by transit. Legislating increases in new car fuel economy
and/or higher gasoline prices can save substantial quanti-
ties of fuel in both the short and long run. Policies to
induce higher auto occupancy during peak periods are
unlikely to save much energy in the short term both
because of their political infeasibility and individual
reluctance to change habits.
Introduction
Between the end of World War II and 1972, trans-
portation fuel gas grew steadily and rapidly because of
increases in both passenger and freight traffic, shifts
towards the use of less energy-efficient modes, and de-
clines in energy efficiency for individual modes (ref. 1).
However, since 1972 a number of forces have emerged
that may significantly alter these trends in the future.
These forces include the Arab oil embargo and the
subsequent sharply higher prices for gasoline. After near-
ly two decades of falling "real" prices, the price of gaso-
line increased 26 percent between 1972 and 1974; since
then prices have risen even higher (ref. 2). Because of
these higher gasoline prices, the percentage of personal
*Research Engineer, Energy Division, Oak Ridge National
Laboratory, Oak Ridge, Tennessee 37830. The work described
here was performed while the author was Director, Office of
Transportation Research, Federal Energy Administration, Wash-
ington, D.C. 20461.
consumption expenditures devoted to gasoline increased
23 percent between 1972 and 1974 (to $36 billion in
1974). Prices of new automobiles are also rising
rapidly-up from an average of $3,700 in 1972 to
$4,400 in 1974 (ref. 3)-although they are rising no
faster than the overall Consumer Price Index.
In addition to these economic forces, a number of
institutional changes are underway or under serious con-
sideration. The National Mass Transportation Assistance
Act of 1974 authorizes the expenditure of nearly $12
billion during the 1975-1980 period. Unlike previous
Federal programs for mass transit, the 1974 legislation
authorizes operating, as well as capital, grants for transit
systems. The automobile industry under pressure from
the Federal Government agreed to improve new car fuel
economy 40 percent between 1974 and 1980 (from
about 14 mpg to 20 mpg); the Government is consider-
ing a number of measures to require these and greater
fuel economy increases. Modifications to the Federal
Highway Trust Fund allow funds to be used for mass
transit improvements and to encourage carpooling; sever-
al communities are beginning to institute significant car-
pool programs.
The extent to which these and other new forces will
operate on traditional patterns of personal travel and
land use to change the energy intensiveness and energy
use of our transportation system is the subject of this
paper. It examines the period 1950-1972 with respect to
personal travel and its energy use, reviews the relative
energy efficiencies of different urban and intercity pas-
senger systems, discusses several policies for reducing
transportation fuel use, and compares the energy savings
likely with each of these policies in 1980 and 1985.
Measures to reduce transportation fuel use can l'e
grouped into three generic classes, as shown in table i.
The policies discussed here include all three classes:
Policy. . . . . . . . . . . . . . . Improve mass transit
Class. . . . . . . . . . . Shift to more efficient modes
Policy. . . . . . . .
Class. . . . .
. . . . . . . .Increase carpooling
. . . . . . . Increase load factors
Policy
Class. .
. . . . . . . Raise gasoline prices
. . . . . . . . . . . All classes
Policy
Class.
. . . . . . . . . .Impose new-car mpg standards
. . . . . . . . . . . Improve technical efficiency
111
-------
Table 1. Transportation conservation measures'
Use existing transportation systems
Shift to more efficient modes
Increase load factors
Improve operational and usage
Improve technical efficiency of individual modes.
Reduce demand.
more efficiently:
characteri sti cs.
These four policies were selected for discussion because
they are among the most important and most widely
discussed, and because analyses exist with which to eval-
uate their effectiveness. However, several other impor-
tant options e;dst, such as stricter enforcement of the
55-mph speed imit, wider adoption of right-turn-on-red,
better urban ':raffic control systems, and a ,host of
changes related to air traffic and freight traffic.
This paper concludes that, in the short run at least,
significant pas!,enger transportation energy savings can
be achieved or IV by improving new-car fuel economy.
Behavioral cha:lges (greater use of mas transit and car-
pooling) are SL rprisingly insensitive to purely economic
forces unless they are so strong as to be politically in-
feasible. This suggests the need for more and better pro-
grams to encourage people to change their attitudes and
tastes towards f'nergy use and personal transportation.
Historical Trends in Passenger Travel and Energy Use
Total tram portation fuel use grew from 8.9 QBtu
(101 s Btu) in 1950 to 18.3 OBtu in 1974 (refs. 4,5)
with an aVera!le annual growth rate of 3.0 percent.
Between the m d-1960's and 1972, the growth rate was
much higher, a1: 4.7 percent a year. However, transporta-
tion fuel use increased only 3.3 percent between 1972
and 1973 and @ctually decl ined 3.2 percent between
1973 and 1974 (ref. 4). The 1974 decline was due to a
combination of sharply higher fuel prices, spot shortages
during the summer, the Arab oil embargo that winter,
and the 2 percent decline in GNP between 1973 and
1974.
Figure 1 shows actual transportation fuel use from
1965 through 1974 and projections to 1985 from three
different sourCES. The Department of the Interior (ref.
6) projection vras prepared in 1972-long before recent
oil price increa!:es. The other two sets of forecasts, by
Jack Faucett A!isociates (ref. 7) and the' Federal Energy
Administration (ref. 51. were prepared during the
summer of 1974 as part of the Project Independence
effort. These forecasts used crude oil prices (in 1973
dollars) of $7 and $11/bbl. The variation among fore-
casts is considerable. The Interior forecast is much
higher than the others, presumably because it assumes
the low oil prices of the 1960's; its growth is equal to
the long-run growth rate over the past two decades. The
other four forecasts show growth rates far below the
historical trend. If these latter forecasts prove correct,
considerable fuel savings will be achieved in the transpor-'
tation sector because of fuel price increases alone.
Intercity passenger traffic is carried primarily by
automobile and, to a lesser extent, by airplane, bus, and
train (ref. 11. The variation in energy intensiveness (EI)
among these modes (table 2) is considerable. Buses and
trains are the most efficient modes, followed by autos
and airplanes. In 1972, EI for airplanes was five times
higher than for buses. However, airplanes were the fast-
est mode and automobiles are the 'most convenient in
terms of schedules and routes.
Between 1950 and 1972, the fraction of intercity
passenger traffic carried by airplane climbed rapidly at
, the expense of trains and buses. Energy consumption
rose 192 percent as a result of a 156 percent growth in
traffic and a 14 percent increase in overall EI (refs.
1,8,9). This increase in EI was due to increases in EI for
individual modes and to the shift from buses and trains
to airplanes.
Urban passenger traffic is carried almost exclusively
by car, with only a small and declining fraction carried
by mass transit (buses and electric transit) (ref. 1 I. As
table 2 shows, mass transit is two to three times as
energy-efficient as autos are (refs. 9-11). Urban EI values
are double comparable intercity values because of poorer
vehicle performance (fewer miles/gallon) and poorer uti-
lization (fewer passengers/vehicle) in cities.
Between 1950 and 1972, the fraction of urban pas-
senger traffic carried by cars steadily increased. Energy
use grew 219 percent, caused by a 161 percent rise in
112
-------
30
10
ANNUAL GROWTH RATE DIFFERENCE FROM
1972-1985 FEA.S" IN 1985
I
(<70) (<70) (quads)
001 3.1 30 6.5
J FA
$ 7/ bb I 2.6 21 4.5
51 t/bb I 2.3 16 3.8
I
FEA
$7/bbl 1.9 12 2.5
5H /bbl 1.0 0 0
I
0
1965 1970 1975 1980 1985
/
001. 1972.,.,.
I
JFA. ~~~......
JFA. $11",>-<:;,..,.."
~~ .//
#~ <-
~ ~ /' FEA.$7
~~../ ./
~~,..,.. ~ ,/
~,....... -<"
~ ---- ------~FEA 511
::-- - - ----- . I
26
-
::J
m
o
~ 22
w
(/)
::>
...J
w
::>
LL
18
z
o
I-
«
I-
a:
~ 14
If)
z
«
a:
I-
Figure 1. Total transportation fuel use and forecasts.
traffic and a 22 percent increase in EI (refs. 1,9.11).
Increased EI was due to higher individual modal EI and
to the shift from mass transit to automobiles.
Figure 2 shows how EI for urban modes increased
between 1950 and 1973. Similar increases in energy
intensiveness occurred for the intercity modes, except
for railroads.
Figure 3 shows the distribution of transportation
fuel by mode, market, and purpose for 1972 (ref. 7).
Passenger travel dominates the fuel use budget, account:
ing for more than two-thirds of the total. The automo-
bile (defined in ref. 7 to include cars, motorcycles, and
personal use of trucks) uses 60 percent of the fuel.
Urban transit accounts for only 0.5 percent of the fuel
use. Intercity bus and rail passenger service also use less
than 1 percent of the fuel.
113
-------
Table 2. Energy intensiveness of passenger
modes, 1972
Mode
Mode
EI
Btu/PMa
E1
Btu/PMa
Intercity
Urban
Airline
Automobile
Rail
Bus
7, 700
3, 100
2, 700
1,500
Automobile
Rail trans i t
Bus trans it
6, 700
2,600
3,000
aEI = energy intensiveness; PM = passenger-miles.
Source: Updates using data sources and method-
ologies discussed in 'reference 1.
Trucks are the second most important energy-using
mode. However, trucks represent a much more heterog-
eneous mode than do automobiles. The 18 percent
figure shown fl>r trucks includes both local and intercity
truck freight tr 3ffic.
The third 1'Iost important energy-using mode is the
airplane, accounting for 8 percent of the fuel. Nontruck
freight modes use 8 percent and the military uses
another 5 perCE nt of the transportation fuel budget.
Improved Trandt
A I though transit is considerably more energy-
efficient than automobiles are, transit presently carries
such a small fruction of total urban passenger travel /2.5
.percent in 1973) that its short-term potential contribu-
tion to energy Gonservation is slight. The data shown in
table 3 from three recent transit demonstrations (refs.
12,13) suggest that the energy impacts of transit fare
reductions and service improvements (expanded area
coverage, reduc ~d headways) are almost negligible.
There are several reasons for the slight energy
impacts shown in table 3. First, transit accounts for a
tiny fraction 01 urban travel and an even smaller fraction
of the urban tnvel energy budget. Thus sizable increases
in transit traffic will have only slight impacts on total
urban traffic ilnd energy use. Second, while reduced
fares and improved service will increase ridership, the
experience cite.j above suggests that less than half the
increase comes from former automobile drivers.' The
remainder are auto passengers, walkers, users of other
transit systems, and people who formerly stayed home.
Only shifts from auto driver to transit reduce overall
energy use. Third, expanded route coverage and reduced
headways lower system load factors; this increases EI
and energy use. Fourth, automobiles are often used to
gain access to transit systems; this auto energy use must
be subtracted from the energy savings due to the shift
from auto to transit.
Figure 4 shows the sources of increased ridership
(refs. 13,14) due to the improvements in Atlanta's bus
system, summarized in table 3. In 1972, fares' were
reduced from $.40 to $.15 and a number of service
improvements were instituted: somE1 lines were extend-
ed, some were revised, new routes were established, and
headways were reduced overall. The net impact of these
changes was an increase in annual coverage from 19 to
22 million bus-miles.
Bus patronage increased 28 percent due to the fare
reduction and service improvements. Reducing fares
increased load factors by increasing ridership with no
increase in bus-miles. Service improvements, on the
other hand, lowered load factors because ridership-in
Atlanta, at least-increased more slowly than did bus-
miles. Overall, the combination of reduced fares and
increased service raised load factors slightly. Because
more than half of the new riders were not formerly auto
drivers, the fuel savings-9,300 gal/day-is slight.
114
-------
7000
.....8
"""'8
_8- _11-----
AUTOMOBILE
_8-
-
6000
~,
a..
'- 5000
~
m
en
en
~ 4000
w
>
en
Z
w
~ 3000
z
.
>-
<9
a:
~ 2000
w
-.-
1000
BU S ------ ...-1'".
.------- ' <;
RAI L '.L&-.
.
.
.
o
1950
1955
1960
1965
1970
1975
Figure 2. Urban travel modes energy intensiveness, 1950-1973.
The major conclusion from table 3 and figure 4 is
that transit improvements alone offer little hope of large
energy savings. Improving mass transit (time, costs, serv-
ice characteristics) can save energy only if the increased
transit ridership comes primarily from automobile
drivers. Increasing transit patronage by attracting people
from nonauto modes (other transit systems, walking,
bicycling, previously foregone travel) will probably
increase ur~an passenger energy use. Thus saving energy
via increased transit requires both the "carrot" and the
"stick." The carrot is to induce people to travel via
transit and the stick is to force people out of their cars.
Even if transit improvements and auto disincentives
are effective, transit is unli kely to provide substantial
energy savings during the next decade. The potential
energy savings are limited by the small size of the pres-
ent transit plant and the' small fraction of urban travel
moved by transit. Doubling the percentage of urban
travel carried by transit from 2.5 percent in 1973 to 5.0
percent in 1980 would require 100,000 new buses dur-
ing this 7-year period, compared with the 1973 fleet of
46,000 buses (ref. 10).
Assuming that funds can be found to finance ,h(~
purchase of these buses; that drivers, mechanics, and
managers can be trained during this period; that rider.
ship will increase; and that the new riders will come
from automobiles; the energy savings for 1980 are equiv-
alent to 52,000 bbl/day of crude oil (refs. 9-11). If the
115
-------
PASSENGER (69"!o)
URBAN PASSENGER (430/0)
URBAN AUTO
43 "!o
INTERCITY PASSENGER
260/0
A
//
INTERCITY AUTO //
17 '70 //
//
//
//
//
//
/
TRUCK
18"!o
AIRPLANE
8 '70
Figure 3. Distribution of transportation fuel use, 1972.
percentage of Lrban passenger travel carried by transit
increases to 7.E percent in 1985 (a tripling of its 1973
market share). the national energy savings would be
equivalent to 1 ~!2,OOO bbl/day of crude. As shown later,
the energy sav;.,gs possible by increasing transit travel
are small relativl! to the savings possible with other trans-
portation measures.
Although the short-term energy conservation poten-
tial of increased mass transit is slight, this does not mean
that transit improvement programs should be aban-
doned. Changes in urban travel patterns are likely to
require at least a decade because of long lags associated
with changes in land use patterns, auto ownership, and
individual attitu:Jes towards public transportation. Thus,
unless transit improvement projects are undertaken now,
the long-term pl)tential benefits of transit will never be
realized. Also, transit offers other benefits besides
reduced energy use: less congestion during peak periods,
fewer traffic fatalities, and increased mobility for those
with limited acc~ss to autos. Finally, combining transit
improvements with auto disincentives (such as the gaso-
line tax increa~e di$cussed later) provides a transporta-
tion alternative to those dislodged from their automo-
biles.
Carpooling
The average automobile load factor for urban work
trips is presently only 1.2 passenger-miles/vehicle-mile
(PMIVM) (ref. 11). There is thus an enormous energy
conservation potential in the empty automobile seats
traveling during the daily peak hours. Increasing this
116
-------
Table 3. Energy conservation impacts of
transit improvements
Strategy
-----------.-.--------..----.-----.---
----- ---- --- ----.-.-.- -..-------------.----.-
Estimated savingsa
Regional bus - Atlanta
40~ to 15~ fare
19 to 22 million bus-miles/year
28 percent increase in ridership
9,000 gal/day
(0.5 percent)
Corridor service, bus - D.C.
Shirley highway (1-95)
ll-illile busway in median
1,900 to 11,500 pass/day in 5 years
40 percent of riders were auto drivers
30 percent of riders get to bus by car
3,000 gal/day
(0.1 percent)
Corridor service, rail - Philadelphia
Lindenwo1d line
30,000 riders/day carried by line
28 percent of riders were auto drivers
90 percent of riders get to line by car
-450 gal/day
(0 percent)
aFigures in parentheses are percents of regional transportation
fuel use saved.
Source:
References 12 and 13.
load factor from 1.2 to 1.6 PM!VM would save 440,000
bbl/day of crude oil in 1980. However, serious questions
exist concerning the methods available to induce greater
auto occupancy and the effectiveness of these methods.
Cambridge Systematics, Inc. (CSt) is analyzing a
variety of carpool promotion opt ions (ref. 15) using a
modification of their disaggregate travel demand models
that estimate auto ownership and changes in work trip
and nonwork travel in response to these policies. Their
model is applied to Washington, D.C., using data col-
lected during the D.C. 1968 Home Interview Travel
Survey. CSI is examining a number of carpool alterna-
tives related to the cost and supply of parking, costs of
auto travel (tolls, gasoline taxes, carpool subsidies),
direct regulation of urban travel (auto-free zones, gaso-
line rationing), and employer incentives.
Their model's preliminary outputs for a number of
such alternatives are shown in table 4. These results are
short-run impacts; that is, they do not include long-run
changes in auto ownership that will occur because of
these policies.
Table 4 shows that both parking incentives for car-
pools and major increases in parking costs substantiallv
reduce the amount of solo driving and increase both
carpooling and mass transit use. However, these reduc-
tions in commuting travel are partially offset by in-
creases in nonwork auto travel. These increases arise
because more automobiles are available for nonwork
travel. In the long run, many of these "extra" cars will
probably not be replaced; thus the long-run energy sav-
ings are likely to be larger than the short-run savings.
As an example, 'consider the predicted impacts of an
increase in areawide parking costs of $3/day. The park-
ing surcharge reduces work-trip travel by 10 percent, a
3.9 percent reduction in overall urban travel. However,
the induced nonwork travel amounts to 1.4 percent of
117
-------
~
AUTOMOBI LE DRIVER
42 "10
NO TR I '
21 "10
Figure 4. Pmvious travel modes for new Atlanta
bus riders.
the urban total. The net impact of the $3/day surcharge,
therefore, is orlv a 2.5 percent reduction in urban auto
travel.
The percentage reduction in fuel use is less than the
percentage redlJ(;tion in travel because of fuel penalties
associated witt' increased trip circuity, extra weight for
carpooling, an( increased cold-starts for nonwork trips.
Thus, urban auto fuel use is cut by only 2.1 percent
compared with the 2.5 percent reduction in travel due to
the parking su 'charge. This is equivalent to a national
fuel savings of fi9,000 bbl/day of crude oil in 1980.
Let us assllme that, in the long run, only half the
induced short-run nonwork auto travel occurs. Then, if
long-run equilibrium is achieved in 1985, the 3.9 percent
reduction in u;ban travel due to increased carpooling
coupled with tile 0.7 percent increase in nonwork travel
reduces urban travel by 3.2 percent. This amounts to a
2.7 percent reduction in urban auto fuel use, equivalent
to 105,000 bbl/day of crude oil in 1985.
These results-fuel savings of 2.1 percent and 2.7
percent in 1981) and 1985 respectively-show a remark-
able insensitivity on the part of commuters to changes in
travel costs. Adding $3/day to the base cost of parking
increases the o'/erall cost of commuting ~y 100 to 200
percent. The re;ponse to this enormous change in cost is
only a 10 percent reduction in work-trip travel.
As discusse::llater, policies that affect all automobile
travel (both work and nonwork) are likely to be more
effective in saving energy for two reasons. First, work
trips account for only a third of all auto travel (although
they account for about 40 percent of all auto gasoline
use). Second, nonwork trips are likely to be more discre-
tionary than are work trips and therefore more sensitive
to changes in dollar and time cost.
The results of this ongoing CSI study should be
interpreted cautiously. The data used to construct the
model are from 1968-a time when fuel prices were low,
Washington's transit service was poor and deteriorating,
Americans loved their autos, and. incomes were rising
steadily. Thus the data are from an era in which our
attitudes and behavior strongly favored automobile own-
ership and use. Also, models such as that developed by
CSI capture the major demographic and economic varia-
bles affecting travel decisions but they cannot explicitly
model changes in attitudes. To the extent that attitudes
towards energy use, carpooling, and automobiles have
changed since 1968, the model's results are in error.
Finally, CSI's model evaluates only the short-run im-
pacts of policy changes. As households adjust their auto
ownership and home location in response to these poli-
cies, the impacts of these policies may increase.
Gasoline Taxes
One policy that is often discussed but rarely
embraced is an increase in the Federal tax on gasoline,
currently at $.04/gallon. Those favoring an increase
argue that it would be an effective way to reduce gaso.
line use, that it allows maximum consumer choice (in
terms of changes in both vehicle use and vehicle owner-
ship), and that it is simple and inexpensive to administer.
Opponents argue that its economic burden on low-
income families (especially those in which the breadwin-
ner must commute by car) would be intolerable, that
consumer demand for gasoline is insensitive to price
changes, and that further increases in gasoline prices
would adversely affect the economic recovery.
The Office of Energy Systems in the Federal Energy
Administration (FEA) developed a simple econometric
model for estimating changes in automobile travel, fuel
economy, and ownership in response to changes in gaso-
line prices (and also to exogenous changes in new car
fuel economy) (ref. 16). The model contains three
behavioral equations that estimate annual demands for
automobile travel, new car sales, and new car fuel econo-
my as functions of income, unemployment, gasoline
price, and the average age of the automobile stock.
Automobile gasoline use from 1965 to 1973 and
several projections to 1985 (using the FEA model) are
shown in figure 5. The gasoline tax in these simulations
is expressed as a constant percentage of the base price to
118
-------
Table 4. Energy impacts of carpooling policies in Washington, D.C.
---------
-------.--.--.-----
Policy
Percent change in work
trip mode shares
Drive
alone Carpool Transit
Percent change in VMT
(miles)
Work Nonwork Total
---.-------.---------..- ---.------------
Base values
excluding
weekend
travel)
--
Parking
incentivesa
Parking in-
centives &
parkigg
costs
Base parking
cost + $1
(areawide)
Base parking
cost + $3
(areawide)
Base parking
cost + $3
(CBO only)
Carpool sub-
s i dy, 5et/PM
52.9%
25. 4~~
14 . 5~b
10.4
(m;)
16.7
(m;)
27.1
(m;)
Percent
change
in fue 1
consumption
2.58
gal/day,
--
-0.55
-1.83
-0.68
-2.07
-0.78
-0.50
-----
- --
--
----
aparking incentives refer to the restriction that IIc10se-inll parking spaces
are reserved for carpoo1ers. Parking incentives and parking costs add a $2/day
parking charge for solo drivers. '
Source:
-10.69
22.05
0.40
-22.27
43.82
4.55
-5.07
4.50
10.61
-15.61
13.93
32.57
-6.52
2~97
17.83
-4.04
11.31
-5.06
Reference' 15.
-3.37
1. 02 -0.64
-9.82
2.50 -2.24
-3.27
0.71
-0.81
-10.20
2.29 -2.49
-4.04
1.04 -0.92
-2.47
0.43 -0.68
119
-------
14 -, "I 'T----T
/
HISTORICAL / BASELINE
~ 12 GROWTH "t/
I
m /
0
/ "
w ,.
/ ,.
Cf) "
::> / ,,-
wlO L-;;)
z ~-;7
:::i
0 " NEW CAR mpg
Cf) \Vr' STANDARDS
6
l/day in 1985. These fuel savings are
nearly an order of magnitude larger than those due to
programs that increase carpooling or mass transit use.
The short-run price elasticity of demand for gasoline
implied by the FEA model is -0.21; the long-run elastici-
ty is -0.72 (ref. 16). These elasticities are in good agree-
ment with those derived in other studies (ref. 17).
Figure 6 shows changes in new car fuel economy
predicted by the model. In the baseline, fuel economy
improves from 14 mpg in 1974 to 15 mpg in 1978 and
then drops back to about 14.3 mpg during the next 7
years. This decline in fuel economy is due to the as-
sumed rise in personal incomes: as incomes grow, the
price of gasoline becomes a less important determinant
of new-car purchase decisions. The 20 percent gasoline
tax simulated here causes a sharp increase in new-car fuel
economy to 17 mpg in 1978. Then fuel economy drops
to an average of 16.2 mpg during the next 7 years.
The curves of figures 5 and 6 show that the major
response mechanism to higher gasoline prices is an in-
crease in new-car fuel economy rather than a decline in
auto travel. During the first year, however, approximate.
Iy three-fourths of the change in gasoline use is due to
reduced driving and only one-fourth is due to improve-
ments in fuel economy (ref. 17). Auto travel is reduced
120
-------
22
~ 20
0-
a.
E
/
NEW CAR mpg /
STANDARDS~ ~
"7
/
/
/
/ ,,- 20 "/0 GASOLINE
/1'\ PRICE INCREASE
1 \
I \ ----
I \.....-
/I
>-
~
~18
a
u'
w
~
w
::>
Ii. 16
II::
-------
Table 5. Energy impacts of tra,nsportation
conservation measuresa
Policy
--
----
---------
Estimated energy savings
(thousand bbl/day)
1980 1985
Increase percentage of urban travel
cat'ried by mass transit from 2.5
pel'cent in 1973 to 5.0 percent in
1980 and 7.5 percent in 1985.
Increase carpooling sufficiently
to reduce work-trip auto travel by
10 percent in 1980 and 1985.
Increase gasoline prices by 20
pel'cent starting in 1975.
Increase new car fuel economy from
14 mpg in 1974 to 20 mpg in 1980
and 22 mpg in 1985.
52
122
69
105
484
700
568
1 ,327
C.These savings are calculated relative to a baseline in which
auto travel is 1.2 trillion vehicle-miles in 1980 and 1.4 trillion
vehi cl e-mil es in 1985. Urban travel accounts for 55 percent of
this total; average auto fuel economy is 14 mpg for both years
and 12 mpg in urban areas. Average automobil e occupancy is 2.2
PM/VM and urban occupancy is 1.6 PM/VM.
behavioral changes (with respect to personal travel) are
quite difficult 10 effect: people are extremely resistant
to purely ecor omic forces that seek to change their
travel modes and extent of travel.
These results also suggest (implicitly) the need for
much greater emphasis on activities that inform people
about energy I=roblems and the need for conservation
and that demol'strate the attractiveness and viability of
energy-efficient practices. These activities will encourage
consumers to c.lange their tastes and attitudes; changes
in consumer be1avior will then follow. In other words,
social norms nt!ed to be modified so that people will
want to carpool, will want to ride transit, and will want
to own small cars-quite apart from the economic incen-
tives to do so.
The analys1:s on which the above conclusions rest
use techniques md data that are far from satisfactory.
The models ger;erally capture the important economic
variables (e.g., prices and incomes). However, intangibles
, such as comfort, convenience, reliability, safety, what
the neighbors think, and whatever else goes into individ-
ual decisions on how, when, and where to travel are not
captured by these models. Therefore the models can pre-
dict behavioral chan~es in response to economic changes
only when these intangibles (those variables not included
in the models) do not change.
The models are generally based on data from the
1960's. To a surprising, perhaps frightening degree, the
United States today is quite different from what it was
10 years ago. This is true for fuel prices and also true
with respect to individual expectations for the future
and attitudes towards the environment, energy conserva-
tion, and automobiles. Thus decade-old data (even when
used with very good models) may yield inaccurate esti-
mates of behavioral changes. Also, the FEA model is
estimated using time-series State-level data. CSI, on the
other hand, uses detailed cross sectional data for a single
city for 1 year. Because of these differences in data, the
122
-------
results of these two models may not be consistent with
each other.
Finally, a comprehensive model with which to eval-
uate these policies in combination does not now exist.
Therefore, it is not possible to evaluate potential syner-
gisms among the four policies considered individually in
this paper. Combinations of auto disincentives (such as
an' increased gasoline tax) and improvements of alterna-
tives (carpooling, transit) are, intuitively at least, quite
appealing. Such combinations are likely to maximize
energy ~avings while minimizing adverse mobility im-
pacts of auto disincentives.
REFERENCES
1. E.' Hirst, "Energy Intensiveness of Passenger and
Freight Transport Modes: 1950-1970," Oak Ridge
National Laboratory report ORNL-NSF-EP-44,
April 1973.
2. Bureau of Labor Statistics, "Retail Prices and In-
dexes of Fuels and Utilities," monthly press re-
leases, U.S. Department of Labor.
3. National Automobile Dealers Association, Research
Department, 1975. '
4. Bureau of Mines, "U.S. Energy Use Down in 1974
After Two Decades of Increases," news release, U.S.
Department of the Interior, April 3, 1975; also earli-
er annual press releases.
5. Federal Energy Administration, "Project Indepen-
dence Report," November 1974.
6. W. G. Dupree, Jr., and J. A. West, "United States
Energy Through the year 2000," U.S. Department
of the Interior, December 1972.
7. Jack Faucett Associates, "Project Independence and
Ene rgy Conservation: Transportation Sectors,"
Federal Energy Administration Project Indepen-
dimce Blueprint final task force report, November
1974.
8. E. Hirst, "Total Energy Use for Comrnerciill Aviil-
t ion ," Oak Ridge National Laboratory report
ORNL-NSF-EP-68, April 1974.
9. Federal Highway Administration, "1973 Highway
Statistics," U.S. Department of Transportation,
1974; also earlier annual issues.
10. American Public Transit Association, '14-'15
Transit Fact Book, March 1975.
11. U.S. Department of Transportation, 1974 National
Transportation Report, December 1974.
12. D. Boyce, et aI., "Impact of Rapid Transit on Fuel
Consumption and Cost for the Journey to Work,"
University of Pennsylvania, draft final report for the
Federal Energy Administration, June 1975.
13. J. Curry, et aI., "Energy and Air Pollution Impacts
of Improved Transit Services," DeLeuw Cather and
Co., draft final report to the U.S. Environmental
Protection Agency, March 1975.
14. Metropolitan Atlanta Regional Transportation
Ad ministration, "Analysis of Transit Passenger
Data," October 1973.
15. W. A. Jessiman, J. H. Suhrbier, and T. J. Atherton,
"Speculative Statements on Shared Sedans: The Use
of Disaggregate Travel Demand Models to Analyze
Carpooling Policy Incentives," Cambridge Systemat-
ics, I nc., Cambridge, Massachusetts, August 1975.
16. J. Sweeney, "Passenger Car Use of Gasoline: An
Analysis of Policy Options," Office of Energy Sys-
tems, Federal Energy Administration, draft report,
December 1974.
17. S. Wildhorn, et aI., "How to Save Gasoline: Public
Policy Alternatives for the Automobile," Rand
Corp. report R-1560-NSF, October 1974. Also, B.
K. Burright and J. Enns, "Econometric Models of
the Demand for Motor Fuels," Rand Corp. report
R-1561-NSF,1975.
123
-------
124
-------
4 November 1975
Session III:
RETROFITTING'OUR PRESENT-DAY
ENERGY SYSTEMS
Roger S. Car Ism ith *
Session Chairman
*Head, Analysis and Evaluation Department, Holifield National Laboratory. Oak Ridge, Tennessee.
125
-------
126
-------
NEAR-TERM POTENTIAL FOR IMPROVED
AUTOMOBI LE FUEL ECONOMY
Karl H. Hellman, Ph.D., and John P. DeKany*
Abstract
There appear to be three broad vehicle-related
approaches toward improving automobile fuel economy
in the near term. These three approaches, which do not
involve changing or modifying the type of fuel used, are:
(1) retrofit programs; (2) inspection, maintenance, and
repair programs; and (3) improvement of new vehicle
fuel economy. All three approaches appear to be com-
patible with environment goals. Examples of each ap-
proach are given, and the pros and cons of each ap-
proach are discussed. Final/y, some of the types of ques-
tions that must be answered before decisions as to
which, if any, of the approaches should be undertaken
are listed.
Introduction
This paper discusses technological potentials for im-
provements in automobile fuel economy that are effect-
ed by changes to the automobile itself. Therefore, non-
technological changes directed toward reduced automo-
bile. fuel consumption such as carpooling, use of mass
transit, changes in driving habits, and the substitution of
bicycl ing or wa Ik ing for the most energy-intensive
(short) trips are not specifically discussed, although
some or all of the above approaches are candidates for
use in an overall energy conservation strategy. Also,
technological approaches such as highway design and
traffic control for reduced fuel consumption are not
considered, although approaches of this type are also
obvious approaches toward reduced fuel consumption.
The automobile or Light Duty Vehicle (LoV) is the
primary subject of this paper because of its visibility and
the great proportion of transportation energy consumed
by vehicles of this type. Much of the discussion, how-
ever, is applicable to Light Duty Trucks (LOT's) and
some of the approaches are also applicable to Heavy
Duty Vehicles (HoV's).
Since the automobile is the focus of this paper, the
near-term time frame can be conveniently quantified to
be the average lifetime of an automobile, approximately
10 years. Therefore most of the approaches considered
here should be able to be implemented within 10 years.
. Karl Hellman is Project Manager, Mobile Source Air Pollu-
tion Control Program, and John DeKany is Director, Emission
Control Technology Division, with Environmental Protection
Agency, Ann Arbor, Michigan.
This time frame implies that the automobiles under con-
sideration will generally be those powered by existing
engine types and fuels.
There are three basic approaches toward improving
the fuel economy of automobiles. Two approaches in-
volve programs that could affect the entire fleet of auto-
mobiles. These are retrofit programs and inspection,
maintenance, and repair programs. The third, improving
new automobile fuel economy, involves only new auto-
mobiles.
Retrofit
One approach toward improving automobile fuel
economy would be a retrofit program, that is, installing
fuel-saving devices or taking other measures on all or
some designated fraction of the automobile fleet.
A scheme that falls in the general retrofit area but
does not involve the use of add-on devices would be the
dismantling or removal of existing emission control
systems on automobiles. Such actions, which are called
tampering by EPA, are illegal if done by dealers. We have
studied this approach and have found (ref. 1) that when
done by gas stations and repair shops the results are
counterproductive, as shown in table 1. Since the fuel
economy went down and the emissions went up, the
$22.86 average cost for this tampering seems to be not
cost-effective.
Retrofits that involve the incorporation of devices
to improve fuel economy are more commonly con-
sidered under the heading of retrofits. These devices
span the range in complexity from simple air bleed~,
through carburetor replacement and gaseous fuel conver.
sions, to complete engine retrofits. EPA has tested many
such devices. Originally most were publicized as having
emissions-reduction capability; now many are being
touted as having fuel-consumption advantages. I n the
space of this paper it is not possible to discuss even a
portion of the devices, but some general comments are
possible.
Besides saving fuel, retrofit devices should ideally be
cost-effective for the customer; that is, the cost of the
device installed should be offset by the fuel savings ob-
tainable. The cost of the device can be computed as
parts plus labor and both out-of-pocket or discounted
accounting can be used. The fuel savings depend on the
type of automobile, the fuel economy improvement, the
number of miles driven, and the fuel cost. The results of
some calculations are shown on figun~ 1.
127
-------
Table 1. Fuel economy and emissions impact
due to tampering
---_._-
-------
- --,-------_..-
Fuel '~conomy
HC
3.5 percent
loss
-------------.--
--.. - ---------.----.------ _._. _.- -----
co
NOx
39 percent
increase
63 percent
increase
90 percent
increase
If a device cost lies above the line, it cost more than
the fuel savings; if below the line, less. For example, a
device that improves fuel economy by 5 percent over a
15-mp'g base, using 50,000 miles of operation and
$O.60/gallon fnr fuel cost, could cost about $92. Of
course, the dilcounted prices would be lower. What
figure 1 indicat~s is that relatively small improvements in
fuel economy may be justified if the device cost can be
made reasonable. It therefore appears that there may be
potential attral:tiveness to retrofit devices, if they do
work.
Another pntential attractive aspect of a retrofit pro-
gram involves the number of vehicles that can be affect-
ed. Theoretical!y, the whole vehicle population could be
retrofitted, thereby getting the maximum savings in fuel
economy as sac n as the retrofits are installed.
900
6= $0.60/ga1
0;; $0.75/ga1
Fuel
Savings,
Dollars
600
300
o
The outlook for retrofits, however, is not altogether
positive. There are several other aspects of such an
approach that must be considered:
1,. The emission characteristics of retrofitted
vehicles must be known, and be beneficial, before a sub-
stantial program begins, so that the gains made to date in
control of automobile emissions are not lost. This im-
plies a more comprehensive evaluation of the devices
than merely checking for their fuel consumption charac-
teristics.
2. Existing experience with retrofit programs can-
not be overlooked. California's retrofi't program cannot
be called an unqualified success, although its experience
does indicate some of the more practical problems that
must be faced.
3. The service industry will probably need some
l 10 mpg
J lOOK
miles
10 mpg
50K
15 rnpg
50K
20 mpg
50K
5 10 15
Fuel Economy Improvement, Percent
Figure 1. Fuel saving versus fuel economy improvement.
128
-------
increase in capability to be able to install the devices
correctly. I mproper installation was found to be a prob-
lem in California.
4. Any given retrofit device will not work on all
automobiles. An example is the air bleed. Air bleeds may
have some beneficial effects for older, less controlled or
uncontrolled automobiles, but their application to more
recent automobiles that have enleanment already as a
part of their emission control systems may be ineffective
or even counterproductive. Therefore, if the goal is to
retrofit a great percentage of the fleet, many different
retrofit devices will be needed.
5. To run such programs effectively, an expensive
administration, testing, and enforcement organization
will probably be needed. This cost should be added to
the other costs inherent in such a program to adequately
assess its effectiveness.
I n summary, retrofit programs for improving auto-
mobile fuel economy appear to require a great amount
of planning, testing, and cost-effectiveness analysis be-
fore they can be said to be (a) viable options for reduc-
ing fuel consumption, and (b) compatible with environ-
mental goals.
Inspection, Maintenance, and Repair(lMR)
Vehicles which are not kept tuned to the manufac-
turer's specifications can have poorer fuel economy and
emission performance than those that are tuned up. In
reference 1, a limited amount of data indicates that the
improvement between vehicles in the "as received" con-
dition and tuned up exceeded 10 percent, although the
more generally accepted value for an IMR program is
about 2 percent, since not all vehicles will receive main-
tenance. Therefore, it would appear to be desirable to
find some way to insure that vehicles receive the proper
ma intenance.
Most I MR programs contemplated to date have
actually been directed toward the control of emissions
from in-use vehicles. These programs usually are part of
the overall program of a State to insure that ambient air
quality requirements are met. Such programs involve the
periodic testing of automobiles, usually on a yearly ba-
sis, and the identification of those vehicles which are not
within specified requirements and which must be fixed
so that they will pass.
The advantages of IMR programs are:
1. All vehicles are affected. The benefits from
such programs can therefore be realized in a relatively
short time.
2. Compared to retrofit programs, there is no
expense involved for devices.
3. Since most programs are emission-control ori-
ented, the environmental effects appear to be beneficial.
As is the case with retrofits, there are also some
potential disadvantages to IMR programs.
1. 1M R programs have the potential to be regres-
sive. If the cost of repairs were equal for all segments of
the population, poorer people would have to pay a great-
er fraction of their income to get their automobile re-
paired. Actually, if (as it appears to be the case) poorer
people own older vehicles, which may require even more
repair than the average, the regressiveness is accentuated.
This impact on the older vehicle and its owner can be .
quite severe. If you own an automobile worth $50.00
that you need for transportation to work, and the 1M R
test indicates that you require a valve job which costs
more than your car is worth, what are you going to do?
The potential for getting around the system is great in
the above-mentioned case, since a fake or illegal sticker
for $10 would appear to be a better deal to the person
who could not afford a complying, presumably more
expensive, vehicle.
2. Since more IMR programs are emissions orien-
ted, the testing required can be complicated and expen-
sive, although work is underway at EPA to develop
short, cost-effective tests.
3. As is the case with retrofit programs, an expen-
sive administration, testing, and enforcement organiza-
tion will be required to insure that the goals of the pro-
gram are met.
4. The cost of the maintenance and/or repairs can
in some cases be high.
We conclude that IMR programs should not be con-
sidered as only fuel economy improvement options, but
that the improvements in fuel e<;onomy attributable to
such an emission control program should be considered
when analyzing its cost-effectiveness.
Improvement of New Vehicle Fuel Economy
There are several methods by which the fuel econo.
my of new vehicles can be improved. The near-term time
frame of this paper tends to rule out a switch to new
engine types, but there is potential for some new en-
gines, especially the diesel in some vehicles in just a few
years, if the work reported by Volkswagen and General
Motors is any indication.
The techniques for vehicles equipped with the con-
ventional engine involve both engine/emission control
system modifications and vehicle modifications. Engine/
emission control system modifications include more effi-
cient aftertreatment devices. These devices, such as more
efficient catalysts and/or exhaust port liners and air in-
jection systems, allow engine recalibration to improve
fuel economy. Changes of this type (catalysts) have al-
129
-------
ready been introduced in model years 1975 and 1976,
and have enabed the new car fleet average to improve
from 13.9 mpg in 1974 to 17.6 mpg in 1976 (ref. 2), a
gain of 27 perclmt.
I n the near future, advances in fuel metering-such
as fuel injection or sonic carburetion, electronic control
of engine and emission control parameters, and design
and development work to make the average vehicle as
good as the bl!st currently demonstrated vehicles-are
expected to ha\ e equally large benefits.
Other vehi :Ie-related improvements involve upgrad-
ing transmission efficiency. I n the near term, a massive
shift to manu a transmissions is probably not practical,
but improvemelts to today's automatic transmission via
the addition of extra gears and the introduction of
torque convertl!r lockup on one or more gears is now
being considerEd for near-1980 introduction. Continu-
ously variable t'ansmissions (CVI's), however, appear to
be beyond the r ear-term time frame of this paper.
The structL re of the vehicle can also be changed to
provide more p,lssenger room per pound of vehicle. New
body designs-such as unibody construction and the sub-
stitution of mOl e weight-efficient materials (plastics, alu-
minum, and hi~h-strength/low-alloy steels)-have poten-
tial. Improvements in vehicle aerodynamic drag and tire-
rolling resistanl:e also appear possible. Additionally,
more lightweigt t automobiles could be produced com-
pared to today'~ market share of such vehicles.
Performance reductions that allow the engine to
operate less thlottled more of the time can have fuel
economy benef.ts, although the marketability of such an
approach on a large scale must be investigated. All of the
above approach~s are now being considered for applica-
tion in the near term. The likely improvement in new car
fleet average f Jel economy has been estimated by
Malliaris (ref. :!). For changes of the kind described
above, which cculd be instituted around 1980, Malliaris
estimated a new car fleet average fuel economy of 25.9
mpg. This cormsponds to an 86 percent improvement
compared with' 974.
Like the other approaches discussed, improvement
of new-vehicle 1uel economy has advantages and disad-
vantages. Amon!1 the advantages:
1. More sophisticated technical approaches can be
used, compared to retrofit devices.
2. The new vehicles will meet the applicable emis-
sion standards.
3. The experience and expertise of the automobile
companies in thl! design and development of the vehicles
is utilized.
4. Possibl\ the greatest fuel savings can be realized
with this approach.
Among the disadvantages:
1. Only new vehicles are affected. This means that
the full benefits of such a program will not be felt until
all the older vehicles are replaced.
2. The fuel economy improvements will increase
the initial cost of automobiles.
Summary
There are no simple answers for improving automo-
bile fuel economy in the near term. All of the ap-
proaches have advantages and disadvantages. One thing
that all of the approaches have in common, however, is
that they will cost money. Whether or not the extra
costs will be justified in fuel economy savings must be
investigated thoroughly.
The actual route to be taken to achieve near-term
improvements depends on yet-to-be-made policy deci-
sions that are beyond the scope of this paper. Before
such decisions are made, however, we feel that detailed,
in-depth answers to the following questions must be
determined.
1. How much fuel needs to be saved?
2. How quickly to the fuel conservation benefits
need to be realized?
3. How much money, manpower, and time should
be expended?
4. How will the achievement of environmental
goals be affected?
REFERENCES
1.
"A Study of Fuel Economy Changes Resulting from
Tampering With Emission Controls," Report 74-21
DWP, Test and Evaluation Branch, Emission Control
Technology Division, Office of Mobile Source Air
Pollution Control, Environmental Protection Agen-
cy,January, 1974.
T. C. Austin, R. B. Michael, and G. R. Service, Envi-
ronmental Protection Agency, "Passenger Car Fuel
Economy Trends Through 1976," SAE paper
750957, presented at the Automobile Engineering
Meeting, Detroit, Michigan, October 13-17, 1975.
A. C. Malliaris, Transportation Systems Center, U.S.
Department of Transportation, "A Framework for
Evaluating Potential Improvements of Auto Fleet
Fuel Economy," presented to the Workshop on
Strategies for Reducing Gasoline Consumption,
Transportation Research Board, National Research
Council, Washington, D.C., October 7,1975.
2.
3.
130
-------
THERMODYNAMIC ANALYSIS OF INDUSTRIAL
ENERGY CONSERVATION OPPORTUNITIES
Elton H. Hall, Ph.D. *
Abstract
Thermodynamic analyses were conducted of the
processes employed in the steel, copper, aluminum,
glass, synthetic rubber, selected plastics, and paper in-
dustries. Results of the calculations for these seven in-
dustries include the minimum theoretical energy, the
efficiency of selected unit processes, and the effect of
certain process changes on the energy use. A computer
model was developed to perform the necessary calcula-
tions. The model performs the customary energy bal-
ances based on the first law of thermodynamics. More
importantly, availability losses are calculated according
to the second law of thermodynamics. The model is gen-
eral and can be applied to an entire industrial process or
to any of the unit operations which are employed within
the process. The results identify where large energy or
availability losses occur. The model is a powerful tool
for identifying those operations in which significant fuel
savings can be realized through straightforward modifica-
tions. On the other hand, the model will also show
where the potential for energy conservation is limited by
the irreversibilities inherent in the basic processes em-
ployed. Such a definition of the nature of the various,
availability losses is invaluable in directing the attention
of process engineers to those operations which offer the
greatest potential for overall fuel savings.
INTRODUCTION
Thermodynamic analyses of the use of energy in
each of seven, energy-intensive, basic industries were per-
formed to evaluate the potential fuel savings for various
energy conservation options (ref. 1). The industries in.
cluded were: steel, copper, aluminum, glass, rubber,
selected plastics, and paper.
Thermodynamic analyses were conducted to deter-
mine 'the theoretical minimum energy required to carry
out the processes associated with each industry. These
calculations were performed for each industry as a whole
and for each of the major unit operations within each
industry. Comparison of the results of these calculations
with data on actual energy use serves to identify those
unit operations for which large differentials exist be-
tween the actual energy use and the theoretical
. Associate Manager, Energy Systems and Environmental
Research Section, Battelle Columbus Laboratories, Columbus,
Ohio.
minimum energy requirement.
Conventional thermodynamic analysis of energy use
commonly consists of conducting energy (heat and
work) balances around the process, an approach based
on the first law of thermodynamics. Each industry was
analyzed by this method. However, for the purpose of
evaluating the potential for reducing the fuel require-
ments for a given process, it is more if1structive to calcu-
late the losses of availability. Availability is a thermo-
dynamic quantity which is equal to the maximum work
which can be obtained from a system by bringing it, by
reversible processes, to a state of equilibrium with its
environment. According to the first law, energy must be
conserved, but availability can be destroyed and its
destruction is associated with irreversibility. Calculation
of availability losses is done by means of analysis based
on the second law of thermodynamics and the results of
this type of calculation also are presented.
It is important to note that the theoretical
minimum energy calculated for a given process is that
energy which would be required if each step in the
process were conducted reversibly. I n addition, the
theoretical minimum energy requirement does not in-
clude any consideration of the time required for
processing. Since all real, processes are, in fact,
irreversible to a greater or lesser degree, and, since all
real processes require a finite time, it follows that the
theoretical minimum energy is an unachievable goal. The
value of the combined first and second law approach is
that those unit operations in which large availability
losses occur can be identified, and (by adjustment of
parameters) the amount of availability and consequentlv
the amount of fuel which might be saved (through modi.
fication of operating conditions and procedures) can be
calculated. In addition, the analysis also shows where the.
potential for fuel conservation is limited by the irreversi-
bilities inherent in the basic processes employed. In such
cases different and more reversible processes would have
to be employed in order to further reduce the fuel con-
sumption. If no viable alternative to the basic process is
available, then the minimum energy consumption achiev-
able in that industry may be significantly greater than
the theoretical minimum based on a hypothetical, com.
pletely reversible, instantaneous process. Such a de-
finition of the nature of the various availability losses
will be useful in directing the attention of process design
engineers to those operations which offer the greatest
potential for overall fuel savings. Furthermore, the
131
-------
ability to predict the potential fuel savings of various
design modifi ;ations by means of a combined first and
second law approach will enable the design engineer to
choose, within the constraints of the costs of the pro-
posed modifications, the design which requires the least
fuel.
A compu':er model was developed during this study
to facilitate the required computations. The code can be
operated in a batch/interactive mode, allowing answers
to be displayed in real time to questions such as, "what
is the potenti,,1 fuel saving if a certain heat exchanger is
operated at a higher temperature?"
In this pClper a brief summary of the thermodyna-
mic approach is given, and some selected examples of
analyses of thf potential fuel savings of proposed process
changes are pr,!sented.
FROCESS UNIT ANALYSES
Energy and Ef.r;ciency
The energy of a flow stream is determined by the
stream compo~;ition and mass flow and by the energy of
the components in the stream. Thus, the energy, Ej, of
stream i is give1 by the mass flow rate, Mi' times the sum
of the mole f-action. Xij' of component j in stream i,
times the enthalpy, hj, of component j:
E. = M'~x"h'
1- 1 IJ r
The efficil'ncy of a process unit was defined as:
Efficiency = Euse/Ein
(2)
where
Euse = total useful energy output,
Ein = total enmgy input including heat and work inputs.
Availability and Effectiveness
The availability of a stream is the maximum work
which can be obtained by bringing the stream to com-
plete equilibrium with the reference atmosphere. This
means that the stream is brought to the temperature of
the reference atmosphere (To), and to the total pressure
of the reference atmosphere (Po), and the partial pres-
sure of each component of the stream is brought to its
value in the refl!rence atmosphere.
Availabilitll per unit mass, A, is defined by the re-
lationship:
A == h - To s,
where h is the enthalpy, s is the entropy, and To is the
temperature of the reference atmosphere. The availabi.
lity, Ai, of stream i is related to the component availabil-
ities by:
I\i = MiLXij(hj - To Sij). (3)
The summatiOIl is done over all components in each
stream. The ellthalpy, hj. depends only on the com.
ponent and str.~am temperature, however, the entropy,
(1 )
Sij' generally depends upon the concentration of each
component in the stream as well as the temperature.
Effectiveness is the term used to measure the per-
formance of a process in terms of availability. Three
definitions of effectiveness have come into use and all
three have been calculated by the thermodynamic
model. However, for the purposes of this summary, only
one definition will be considered:
Effectiveness == Aout/Ain- (4)
Th is effectiveness definition is analogous to the
definition of efficiency and in many cases its value is
close to the value of the process efficiency.
In this definition the numerator is the total availa-
bility output, Aout, and not just the availability of the
product stream or the product stream and the streams
where waste energy is recovered. This approach is some.
what different than that taken in defining efficiency
where the numerator is the energy of the useful output
streams. The efficiency is a measure of how well some
specific product (the useful output) utilized the total
input energy. Effectiveness, on the other hand, considers
the maximum work (useful output) of all the flow
streams leaving the process, whether or not the availabil-
ity of these streams is, in fact, used. A process is charged
only with the destruction of availability in the process,
but is not charged with the destruction (by waste) of the
availability in process output streams. This represents a
fundamental difference between the calculation of
efficiency and effectiveness.
ANALYTICAL APPROACH
The analytical model for thermodynamic analysis
was set up to analyze the energy and availability fluxes
through any specified process unit. A process unit can be
any process, subprocess, or combination of processes up
to an entire plant or industry. The same analytical model
was used for studying processes in all seven industries,
there is no need to specifically tailor the method of
analysis to a specific process.
The system to be analyzed is defined in terms of the
input and output streams as shown schematically in
figure 1. The mass flows entering and leaving, the heat
inputs and losses, and the work inputs and outputs are
included. The specification of heat inputs and losses is
fairly straightforward. The work input and output
category includes electrical and mechanical work done
on (input) or done by (output) the process unit. The
electrical work inputs can be given in terms of direct
input or fuel equivalent, in the latter case a heat loss
should be specified to account for generation and trans-
mission losses. In the analysis performed in this study.
each process unit was first evaluated using the direct
132
-------
COOLANT
REACTANTS
ELECTRICITY
FUEL
AIR
CATAL YST &
NON REACTANTS
COOLANT
[PRIMARY
PRODUCTS &
BY-PRODUCTS
PROCESS
WASTE
FUEL PRODUCTS
WAST E
HEAT
INTERM EDIATE
PRODUCTS
Figure 1. Process definition for thermodynamic analysis.
input energy and availability values for utilities, steam,
and electricity. This procedure was used so that the
direct process losses could be quantified. Once this step
was completed, the process was reexamined, and the fuel
equivalents of the steam and electrical power used were
considered. The total fuel and fuel equivalent required
for a process or industry is the important quantity in the
larger view of national fuel use. The direct energy
consumption for a process, however, provides the basis
for studying that process and improvements in it.
The energy and availability of flow streams entering
and leaving a process unit provide a unique way of de-
. fining the process unit from a thermodynamic view-
point. The flow streams are specified by their composi-
tion, pressure, temperature, and mass flow rate. The
composition specification includes identification of the
chemical species, mole fraction or number of moles,
molecular mass, and partial pressure in the environment.
The enthalpy and entropy of the species may be input
directly, but usually are calculated by property subrou-
tines. The flow streams are designated as inflow or out-
flow by positive or negative val ues, respectively, for the
mass flow rate. Each outflow stream, and each outflow
of heat may be designated as a useful output, and thus
be used in the calculation of efficiency. All work out-
puts are considered useful.
The thermodynamic model performs a first law
(heat balance) analysis and a second law analysis on pro-
cess unit. The model does not check the mass balance
for the process, it is assumed that the investigator has
accounted for all the important flow streams and has an
adequate mass balance for the process unit. The model
first performs the first law analysis or heat balance. The
investigator may optionally specify one flow stream to
have a variable temperature or mass flow rate. If this
option is used, the temperature or mass flow (whichever
is designated) is adjusted so that the first law is more
closely satisfied. An iterative procedure is used to
successively calculate better solutions to the analysis.
The iterative process is ended when a maximum number
133
-------
of iterations i!
is satisfied tc
energy,
When the first law analysis has been satisfactorily
completed the model goes on to perform the second law
analysis. The 3vailability of each flow stream is calcu-
lated from thl! enthalpy, entropy, and pressure of the
stream. Heat losses are assumed to be availability des-
tructions and hence do not show up in the availability
summary. Work inputs and outputs are considered to be
fully equivalent to availability inputs and outputs. Since
the second law of thermodynamics is an inequal ity,
there is no way to balance the availability flows for a
process the wa'l energy flows must balance to satisfy the
first law. Thu:;, the mass flow rates and temperatures
which satisfied the first law are used in the model for the
second law ana 'ysis without change. Since the evaluation
of the thermodynamic properties is basic to the calcula-
tion 'of energy, availability, efficiency and effectiveness,
the property c.llculations are discussed in the following
section.
reached (usually 8) or when the first law
within 0.5 percent of the total input
Evaluation of 7hermodynamic Properties
The calculation of the energy and availability of a
flow stream ree uires the evaluation of the enthalpy and
entropy of the c:omponents of the flow stream. The ther-
modynamic medel was set up so that the enthalpy and
entropy of a s':ream component could be directly put
into the model jlong with the mole fraction and molecu-
lar mass of that component, or the values were calculat-
ed by the appwpriate property subroutine for the com-
ponent. The en':halpy was calculated at stream tempera-
ture. and the er tropy was calculated at the stream tem-
perature and the partial pressure of the component in
the stream. The temperature dependence of these pro-
perties was obtained by integrating the specific heat of
the component from the reference temperature (800 F)
to the stream temperature, and adding in all the
enthalpy and ernropy changes at transformations and
phase changes. 'fhe enthalpy and entropy of formation
also were added to the corresponding values as calcu-
lated above.
SUMMARY OF RESULTS
General
An attempt was made in each of the industries to
calculate the minimum amount of energy and availa-
bility required t.) convert the raw materials of the in-
dustry into the ,)roducts of that industry. In the steel,
container glass, ilnd aluminum industries, minimum
energy and availcbility requirements could be, and were
calculated. The minimum energy and availability so cal-
culated do not depend upon any process or process de-
tails. In the other four industries, such minimum values
were not calculated because the energy and availability
of the raw materials are greater than they are for the
products. Theoretically these four industries, copper,
rubber, plastics, and paper do not require energy to con-
vert their raw materials into their products. In fact, how-
ever, the processes currently in use for converting the
raw materials into products do require significant
amounts of purchased energy. In some of these indus-
tries, particularly rubber and some plastics, the heat of
polymerization (which is liberated during the making of
the rubber or plastic) is available only at temperatures
close to ambient temperature. Thus, it is only with great
difficulty and expense that the heat of polymerization
could be used.
In addition to determining the theoretical minimum
energy and availability required for each industry. the
inefficiencies of each industry and many of the indus-
trial processes were examined in detail to identify areas
for process improvement. Improvements are possible in
all industries, and all the processes studied have room for
some improvement. The efficiency and effectiveness of
energy use in industrial processes varies widely, some
processes waste relatively little of their input energy
while other processes waste nearly all their input energy.
A brief summary of the numerical results of the
analyses of seven major fuel-using industries is given in
table 1. The seven industries, or major product segments
in the cases of rubber and plastics, were analyzed on the
basis of both the first and second laws of thermo-
dynamics.
The data presented in table 1 are the overall values
of efficiency and effectiveness for the processes analyzed
in each industry. The rubber and plastics industries are
broken down in terms of the major products since these
products and their manufacturing processes are quite
different. The data are also presented in terms of direct-
equivalent and fuel-equivalent input. The direct equival-
ent is based on direct conversion of electrical input to
8tu's for each process, i.e., the electrical generation and
transmission losses are not included. The fuel equivalent
input is based upon the fuel required to generate all
electricity consumed (at 33 percent generation and
transmission efficiency). The direct input is representa-
tive of existing industry conditions since few industries
generate significant quantities of their own electricity.
The direct input does not, however, reflect the true de-
mand on resources which the fuel equivalent does. Thus.
under a total energy concept, i.e., each plant generating
its own power and recovering its own waste heat, the
fuel equivalent would be the only true basis for energy
and availability calculations. As shown in table 1, the
134
-------
Table 1. Efficiency and effectiveness summary for seven industries
Percent
Direct equivalent Fuel equivalent
Industry Efficiency' Effecti veness Efficiency Effectiveness
Steel 42.0 40.1 41. 0 39.0
Copper 3.5 9.2 2.7 7.10
o Aluminum 29.6 32.6 14.2 15.1
Container glass 23.9 22.1 21. 3 19.5
Rubber
High-temperature 92.4 96.2 92.4 96.2
" Selected plastics
Polyethylene
Low-density 83.5 93.1 79.0 '88.0
High-density 75.1 87.9 73.4 85.5
Polystyrene 82.7 97.9 82.2 97.2
Polyvinyl chloride 65.5 90.3 65.0 89.5
Paper 44.3 44.5 44.3 44.5
industries have efficiencies and effectivenesses which
vary greatly. The high efficiencies of rubber. plastics
and. to an extent, paper, reflect the fact that these in-
dustries are essentially processing fuels. The energy value
of the feed-stock and product in these processes leads to
a falsely high value of efficiency.
Another measure of process performance is given by
the availability destruction for the process and/or the
ratio of availability destruction to the availability of pur-
chased fuels and electricity. These two quantities are
listed for each industry in table 2. The fact that the
availability destruction is greater than the total pur-
chased availability for some industries indicates that
some of the feedstock availability is being destroyed.
Despite the high efficiency of several of the rubber,
paper, and plastics manufacturing processes, these pro-
cesses have large availability destructions relative to their
purchased availability. The'se cases illustrate that a high
first law efficiency does not always indicate good pro-
cess performance or even low availability destruction.
~ A further comment must be made on the analyses
rnducted. The input data used for each process were
based on industry average values wherever possible.
These analyses do not necessarily represent the true con-
ditions in any specific plant, process, or industry. In any
industry some plants will have better performance than
that shown. These plants will not be as easy to improve
as other less efficient plants. Specific conclusions for any
plant must be based on analysis of that plant in detail.
SOME SPECIFIC INDUSTRY RESULTS
The Steel Industry
The theoretical minimum energy required for steel
production is 7.52 x 106 Btu/ton. The steel industry
uses approximately 25 x 106 Btu/ton of steel produce~.
Although the theoretical minimum energy requirement
cannot be achieved in any real process, there appears to
be an opportunity for significant increases in efficiency
of fuel utilization.
An important aspect of this study was the determin-
ation of the potential effect of certain process modifi-
cations on fuel requirements. A rather simple example,
the blast stove, will serve to illustrate the approach.
135
-------
Table 2. Availability destruction, purchased
availability, and their ratio
-
Industry
Availability Purchased
destructiona avai1abi1ityb
(x106 Btu/ton of produc~)
A.D./P.A.
(percent)
High-temp. rubber
Low-dE!ns i ty
polyethe1yene
High-density
po 1yethe lyene 6.3 2.8
Polyvinyl chloride 2.7 1.6
Po 1yS1.yrene 1.9 1 .8
Paper 25.8 1.6
aBased on direct equivalence of electrical input.
blncludes direct equivalent of electrical input; does
feedstock availability.
Steel
Coppel"
Aluminum
Glass containers
12.0
62.2
61.8
8.9
2. 1
"
3.3
~
The blast stOl'es are fuel-fired air preheaters for the
blast furnace, providing air at 1500 to 2000 F. The
stoves are usually fired with blast furnace gas. A
summary of ener~ y and availability for a blast stove is
given in table 3, in units of 106 Btulton of steel. The
availability destruction in blast stoves, 0.66 x 106
Btulton, is approx'mately 54 percent of the input availa-
bility of the fuel, Combustion irreversibilities account
for about 40 percent of this total availability
destruction, and about 20 percent is due to heat losses
to the atmosphem. The remaining 40 percent of the
total availability destruction is due to transferring heat
across t~e temper ature difference between the com.
bustion products Clnd the air. This availability destruc-
tion would be reduced at higher operating temperature
with improved insl.lation to hold down heat losses.
The effect of producing higher temperature blast air
was analyzed toget'ler with other blast furnace modifica-
tions, including t~.e charging of hot or preheated raw
materials to the b:ast furnace, These process modifica-
tions do not chan.}e the total input energy to the blast
furnace, but rathu shift a portion of the energy re-
17.8
36.6
89.6
10.7
1.5
67
170
69
83
140
1.9
174
225
169
106
1,616
not include
quirement from the primary fuel to the raw material
inputs. The primary fuel rate can be reduced by the
energy required to heat the raw materials from ambient
temperature (current charging temperature) in the
modified process. The results of an analysis of higher
charging temperatures are summarized in table 4. In the
eight cases presented, only the material charging temp-
eraturesand the coke rate were varied, hence, the total
energy input to the blast furnace is the same for all eight
cases.
The rationale behind the particular temperatures
chosen for the analysis is: (1) not to quench hot
materials before charging (coke and sinter); (2) to pre-
heat large size solid material (pellets); (3) to increase
blast temperature; and (4) to preheat are.
Increased coke temperatures are attractive because
coke is produced at temperatures up to 2,000° F and is
presently quenched to ambient temperature. The sen-
sible heat of the coke is not utilized at all in present
practice. The reason for quenching is to prevent combus-
tion of the coke since its ignition temperature is near
1,000° F in air. The 1,000° F coke temperature is an
136
-------
Table 3. Energy and availability summary for
a blast stove
Energy Availability
(106 Btu/ton of steel)
Input
Blast furnace gas
Output
Air blast
Combustion products
Heat loss
Efficiency
1.19
1.22
0.91a
.18
.10
1.19
.76 percent
0.47
0.09
0.56
aConsidered useful output for efficiency calculation.
~pper limit on coke temperature without inert gas cover,
and 1,250° F might be possible with inert gas cover.
Increased sinter charging temperature is reasonable
since, like coke, sinter is usually produced onsite at high
temperatures and is currently quenched. No combustion
problems are likely as with coke, although some oxida-
tion could take place at high temperatures (1,000° F) in
open air. Pellets are produced hot, but offsite, and
would have to be preheated onsite as would iron ore.
Both could be preheated with inert gas from quenching
of coke, slag, iron, or steel. Preheating of pellets would
present fewer air pollution control problems than pre-
heating iron ore with a gas stream.
Higher blast air temperatures not only allow reduc-
ing the coke rate at the blast furnace but increase the
effectiveness of fuel utilization at the blast stove. In-
creasing the blast air temperature from 1,500° to 2,100°
F in the blast stove analyzed previously, results in a 15
percent increase in fuel consumption, a 2 percent in-
crease in efficiency, and 14 percent increase in effective-
ness for the blast stove without benefit of improved
insulation. Raising the blast temperature from 1,500° to
2,100° F, as in conditions 6 and 8 in table 4, results in a
savings of 0.4 x 106 Btu (per ton of steel) of coke with
an increase of 0.18 x 106 Btu (per ton of steel) of blast
~urnace gas. Since byproduct fuel gases are not fully
~sed in many steel plants, increased blast temperatures
offer a potential savings of 0.55 x 106 Btu (per ton of
steel) in coal requirements or a 2 percent overall fuel
saving.
The Aluminum Industry
The theoretical minimum energy required to refine
bauxite and reduce the refined alumina to aluminum is
28.3 x 106 Btu/ton of aluminum. The average estimated
energy consumption in the aluminum industry is 96 x
106 Btu/ton, if the electrical input is expressed in direct
equivalent.
An example of possible process modifications
analyzed in the aluminum industry involved alumina
calcining.
The calcining process removes water of hydration
from the alumina trihydrate produced by the Bayer pro-
cess and thus provides alumina for electrolytic reduc-
tion. The energy and availability summary for a repre-
sentative calcining process is given in table 5. The useful
output is the dry alumina, at 80 F, ready for transpor-
tation to the reduction cell. The off-gases have been used
for preheating the incoming alumina hydrate and some
energy from the output alumina is credited to air pre-
heating in this analysis. However, the calcining step is
not particularly effective in its present state, and a
fluidized-bed calciner has been suggested as an improve-
ment.
137
-------
Table 4. Fuel reduction allowed by increased raw materials charging
temperatures in a blast furnace
Charging temperatures (OF) Coke-rate
Sinter
Coke Ore & pellets Blast (relative to baseline)
1. Baseline 80 80 80 1,500 1.00
2. Coke, burden 1 ,000 80 500 1 ,500 0.9]
-'
w 3. 2 + blast 1 ,000 80 500 1,900 0.94
CIO
4. 2 + are 1 ,000 500 1,000 1,500 0.94
5. 2 + are a.nd blast 1,000 500 1 ,000 1 ,900 0.91
6. Maximum burden 1 ,250 1,000 1,000 1,500 0.93
7. 6 + blast 1,250 1,000 1 , 000 1,900 &0.90
8. 7 + blast 1 ,250 1 ,000 1 ,000 2, 1 00 0.89
-------
Table 5. Energy and availability summary for an
alumina calciner
Energy Availability
(106 Btu/ton of aluminum)
Input
Alumina (800 F)
Fuel
Tota 1
Output
Alumina (1,5000 F)
Off gas
Heat 1055
. Tota 1
0.00
9.98
9.98
0.00
9.07
9.07
2.07a
5.58
2.34
9.98
2.10
2.14
4.24
21 percent
Efficiency
aConsidered
calculation.
useful output for the efficiency
The energy and availability summary for a fluidized-
bed calciner is given in table 6. The primary operational
improvements over conventional rotary kiln calciners is
the improved effectiveness of preheating input materials
and the resultant decrease in exit temperature of
alumina and off-gases. The fuel rate requirement of the
fluid-bed calciner is lower than that of the kiln because
of the improved preheating of the alumina and combus-
tion air, and thus the fluidized-bed calciner can achieve a
30 percent reduction in fuel requirements. The off-gases
from this process could still provide some of the heat
required for the steam generation for the Bayer process.
The energy and availability flows for these two calcining
methods are shown in figure 2. The energy and availa-
bility of both the alumina and the combustion products
are lower for the fluidized-bed calciner, thus allowing
reduced fuel input.
The Glass Container Industry
The theoretical minimum energy required to pro-
duce glass for conta iners is 2.15 x 106 Btu/ton of glass
melted. The glass container industry currently uses 11.0
x 106 Btu of purchased fuel and electricity (direct
equivalent) .
An important process improvement currently being
studied by the glass industry for future application is the
preheating of the batch with exhaust gases from the
furnaces. The upper temperature limit for preheating th~
batch presently appears to be about 7600 C. The thermo-
dynamic analysis of the melt furnace and its regenerator
indicates that with no heat losses from the regenerator
there is enough energy and availability in the furnace
exhaust gases to preheat the incoming combustion air to
present levels and to preheat the batch to 7600 C. Al-
though heat losses cannot be completely eliminated, im-
proved insulation, within materials and process con-
straints, and the use of reheat and conditioning furnace
exhaust gases for initial preheating could be expected to
provide sufficient batch preheating. An energy and
availability summary for a combined melting furnace
and regenerator is given in table 7, and that for such a
combined melting furnace with air and batch preheaters .
139
-------
Table 6. Energy and availability summary for an
alumina calciner (fluidized bed)
Energy Availability
(106 Btu/ton of aluminum)
Input
Alumina
Fuel
Total
Output
Alumina (600° F)
Off gas (800 F)
Heat loss
Total
EffiCiency
Effectiveness
0.00
7.03
7.03
0.00
6.39
6.39
1. lOa
4.25
1.68
7.03
1.49
1.54
3.03
16 percent
47.4 percent
aConsidered useful output for the efficiency
calculation.
is given in tablE 8. The reduction in fuel consumption
for such an arrangement amounts .to 14.8 percent (fuel
input of 6.360 ,< 106 Btu in table 8 versus 7.465 x 101i
Btu in table 7). The effect of batch preheating on energy
requirements is shown graphically in figure 3. In this
case, the energy and availability of the product glass are
unchanged, but these values for the combustion pro-
ducts are lower with batch preheating.
The overall efficiency predicted. with batch pre-
heating, 38.6 percent based on the direct equivalent of
electrical inputs, approaches the maximum achievable by
melting furnaces of conventional design. The remaining
availability destructions result largely from combustion
irreversibilities and from heat transfer across a finite
temperature diffl!rence.
CONCLUSIONS
The results of the study show the value of the com-
bined first and sncond law approach in identifying those
operations in a process where significant fuel savings
might be achieved and in estimating the amount of fuel
which m!ght be saved through specific process modifica-
tions. Some examples have shown potential fuel savings
of 30 percent. Suggested process changes have been
analyzed from the energy and availability viewpoint
only. Final decisions regarding the desirability of
implementation of any process modification must be
based upon analysis of all of the costs required to
achieve the anticipated energy savings.
ACKNOWLEDGMENTS
This paper is based upon the results of a study con-
ducted at Battelle's Columbus Laboratories for the
Federal Energy Administration entitled, "Evaluation of
the Theoretical Potential for Energy Conservation in
Seven Basic Industries" (Contract No.
14-01-0001-1880). The author wishes to acknowledge
the assistance of the FEA Technical Project Monitors, W.
140
-------
~
:>
z
~
:>
..J
<{
u..
o
z
o
t
:>
f-
ee
'&
f-
:>
Cl.
z
~
:>
z
~
:>
..J
<{
u..
0
Z
0
f-
--
:>
f-
ee 4
'&
f-'
:>
Cl.
f-
:>
0
ROTARY KILN CALCINER
FUEL
10
'//
- ---- --
- ---- ---- --
'//,
'/.
;;;:;/
~/
--.- .-
~ ---- - ~
-- ...... ""'"'"
ALUMINA
1500 F
-- - - --
-- --
......
8
6
4
2
o
2
6
COMBUSTION
PRODUCTS
FLUIDIZED.BED CALCINER
...
> 3
FUEL..
~,
~
'/
---
ALUMINA
- 600 F
---
-
COMBUSTION
PRODUCTS
Figure 2. Comparison of rotary kiln and fluidized-bed calciner.
141
0%
D
ENERGY
~
AVAILABILITY
o
-------
Table 7. Energy and availability summary for a combined
melting furnace and regenerator
Energy Availability
(106 Btu/ton of glass)
Input
Batch
Fuel
Electricity (direct
equivalent)
T ota 1
Output
Glass
Combustion
Losses
products
Total
Effi ciency
Effecti veness
0.637 0.571
7.465 6.785
0.220a 0.220
8.322 7.567
2. 787b
1.464
4.071
8.322
1.327
0.621
1. 948
33.5 percent
25.7 percent
aElectricity expressed as fuel equivalent = 0.677.
If this value is used as the electrical input. the
efficiency and effectiveness rates would be:
Efficiency - 31.7 percent; Effectiveness - 24.2
percent.
bConsidered useful output for efficiency
calculations.
142
-------
Table 8. Energy and availability summary for a melting furnace
combined with batch and air preheaters
Energy Availability
(106 Btu/ton of glass)
Input
Batch
Fuel
Electricity (direct
equi va len t)
0.637
6.360
. O. 22a
0.571
5.786
0.22
Total
7.217
6.577
Output
Glass
Combustion products
2.787b
0.670
1. 327
0.2
(es ti mated)
Losses 3.76
T ota 1 7.217 1.527
Effi ciency 38.6 percent
Effectiveness 23.2 percent
aElectricity expressed as fuel equivalent = 0.677.
If this value is used as the electrical input, the
efficiency and effectiveness rates would be:
Efficiency - 36.3 percent; Effectiveness - 21.7
percent.
bConsidered useful output for the efficiency
calculations.
143
-------
en
en
<{
~ 6
u.
o
2
o
I-
--
::>
~ 4-
&
I-
::>
Q..
2
en
en
<{
-oJ
<.?
u.
o
2
o
I-
:3 2
I-
co
&
~.
::>
Q..
~ 4
::>
o
MELTING FURNACE AND
REGENERATOR
8
MELTING FURNACE AND
REGENERATOR WITH
BATCH AND AIR
PREHEA TING
FUEL
,....-
~
-----.-
'/.
- BATCH
.....- .,.
"'
-
COMBU STiON -
PRODUCTS
.......
-GLASS
FUEL
2-
-
~
~
~
/'
---- --
---
/.
-- BATCH
........ ,.
/ ~
~
.......
COMBUSTION
PRODUCTS
-
Figure 3. Effect of batch preheating on energy requirements.
. Mutryn and T. Gross, and the efforts of other Battelle-
Columbus staff who participated in that project: W. T.
Hanna, L. D. Reed, J. Varga, Jr., D. N. Williams, K. E.
Wilkes, B. E. JOlnson, W. J. Mueller, E. J. Bradbury, and
W. J. Fredericl" Battelle's Columbus Laboratories is
~Jrateful to the Federal Energy Administration for the
financial support of the study and for permission to
summarize the n!sults in this symposium.
0-
---GLASS
REFERENCE
} 148%
D
ENERGY
~
AVAilABILITY
o
1. Elton H. Hall, et al. "Evaluation of the Theoretical
Potential for Energy Conservation in Seven Basic
Industries," final report by Battelle's Columbus
Laboratories to the Federal Energy Administration,
Contract No. 14-01-0001-1880, July 1975. Avail-
able through NTIS as PB 244-772/AS.
144
-------
4 November 1975
LUNCHEON ADDRESS
ENERGY AND ENVIRONMENTAL
RESOURCES
John A. Green
Region VIII Administrator
Environmental Protection Agency
Denver, Colorado
145
-------
146
-------
ENERGY AND ENVIRONMENTAL RESOURCES
John A. Green*
Abstract
Over 50 percent of the Nation's coal resources,
almost 50 percent of the reserves, and more than 60
percent of the strippable reserves are located in the
States of Colorado, Utah, Wyoming, Montana, South
Dakota, and North Dakota. This area supplied about 7
percent of the Nation's 1973 coal production. Shale oil
located in the region's Green River formation represents
all of the Nation's reserves of this energy resource. Oil
and gas reserves, uranium ore reserves and output, and
nuclear energy in the area will supply an increasing
proportion of national requirements. The economic and
social impacts of environmental programs in this six.
State region are discussed below, and, after an outlining
of various technologies of energy resource development,
the environmental impacts to the land and people
involved are listed.
I am indeed pleased to address this joint EPA/
ERDA conference on the environment and energy
conservation. Hopefully, sessions like these will all.ow
members of industry, government, and the public to
exchange ideas and help minimize the adverse impacts
while supplying energy demanded by the Nation as a
whole as well as State and local entities.
Our two agencies, the States, local government, and
other appropriate Federal agencies must work closely
together because of the obvious interrelationships of the
environmental impacts from extraction and conversion
of energy resources.
True, there will have to be some tradeoffs between
energy resource production and protecting the environ-
ment. Developing our natural resources will be accom-
panied with some environmental degradation. That
degradation must be minimized. .
As I see it, we have two priorities. First, we must
make sure that the Nation's energy needs are satisfied.
Second, we must consider local priorities. In both of
these categories, however, we must be sensitive to the
economic and social impacts of environmental programs.
The environment must not be sacrificed, nor be the
scapegoat for the production or the delay. of energy
resource development initiatives.
'Region VIII Administrator, Environmental Protection
Agency, Denver, Colorado.
We are all aware of the problems and challenges
ahead. We as a Nation must become more energy
independent. To reach that goal the Administration has.
supported and initiated energy research, demonstration,
and commercialization programs. We are here to help
balance seemingly conflicting goals through cooperation,
technology development, and cooperation with the
States.
In this region, we are farming and ranching and
living on top of a vast coal field. We realize that some of
this coal must be burned to supplement Appalachian
coal and finite supplies of natural gas while other alter-
natives for energy resource development are being
researched. However, before we can burn any coal we
must, of course, mine it, and there is a whole host of
environmental, social, and cultural problems associated
with the mining of coal.
You here today know about the blighted lives,
lungs, landscapes, human harm, and environmental
havoc that accompanies the mining of coal-not only in
this country but everywhere in the world. We can no
longer accept these kinds of impacts as the inevitable, if
unfortunate, price of getting coal out of the ground and
into our power plants. The mining of both eastern and
western coal must be done according to careful and
comprehensive advance planning and accompanied every
step of the way by effective efforts to keep adverse
impacts to the lowest levels possible.
For its part, EPA will continue to do whatever it
can-consistent with its responsibilities under the law-to
assist and encourage the greater use of coal for electric
power. Under the clean fuels policy we identified 12
major coal burning States whose emission limits wen
stricter than required to meet national primary ambiem:
air standards. We estimated that, by the second half of
this year, these standards would have prevented the
burning of roughly 110 million tons of current coal
production, thus leaving some areas no choice but to
turn to oil in order to meet clean air standards. Thus far,
seven of these States have revised their implementation
plans. We have thus freed approximately half of the 110
million tons of coal for burning, and expect to free the
other half by the end of the year. We have also advised
the remaining Statf!s and territories of any regulations in .
their plans that are more stringent than necessary to
meet air quality standards to protect public health. We
hope that, wherever possible, these regulations will be
revised to permit even greater use of available domestic
coal.
147
-------
In addition, EPA is currently cooperating with the
Federal Energy Administration (FEA) to identify power
development fccilities that could convert to the burning
of coal, but WI' are insisting on consistent health stand-
ards at the samn time.
In the we~;tern States much of the so-called over-
burden that stands between the "stripper" and the coal
is prime grazin!1 land. The mine-mouth conversion plans
that many cornpanies hope to construct here would
consume vast amounts of already scarce western water
that is required for both crops and cattle. Shouldn't we
consider a poss;ble food shortage as well as the shortage
of enl~rgy resources?
A 70-year-old western rancher whose home and
range rest upon a vein of coal ripe for stripping
remarked, "We know coal is going to he mined and I
think we can 1 ace the fact that the development, the
population, is g.)ing to come into this country more. But
we don't want '0, be deluged with it all at once. If it can
just come here nore naturally, and have it on a sensible,
rational basis ,nstead of suddenly being inundated.
That's what we dread the most."
That ranch,!r has a lot of company out here in the
West.
Let's look at some energy development technologies
and their environmental consequences.
Coal conversion into electricity, gasification, and
liquefaction are processes that utilize coal. Coal gasifica-
tion has been successfully demonstrated since the 1950's
in a few fore'gn countries, but in these countries
environmental (onsiderations were not then considered
important. Sev ~ral gasification plants are currently
planned in this ,:ountry and it has been estimated, based
upon $13/bbl crude oil, that by 1985, coal gasification
can produce 360 billion cubic feet per year synthetic gas
at a normal pro!jram rate, and 1,300 billion cubic feet of
gas per year at an accelerated program rate.
Coal liquef u;tion is a developing technology and
lacks a history c f demonstration. Currently there are no
plans for conslruction of commercial plants in the
United States, t ut it is estimated that 500,000 bbl/day
of synthetic oil could be produced on an accelerated
program rate.
Now for th,! environmental impacts of these tech-
nologies. Sulfur .jjoxide emissions and solids disposal are
potentially signi1icant problems associated with low-Btu
gasification. Air emissions of particulates, hydrocarbons,
and carbon monoxides vary according to the process
used. Water eff uents could be a problem depending
upon the method of disposal. Air emissions of particu-
lates, nitrogen and sulfur oxides, hydrocarbons and
carbon monoxide, solids disposal, land use, water
consumption, and water effluents are all environmental
concerns of high Btu gasification. Environmental
impacts associated with coal liquefaction are anticipated
to be from air emissions of particulates, nitrogen and
sulfur oxides, water effluents containing total dissolved
and suspended solids, and organics. solids disposal, and
land requirements.
Oil shale, known to the Indians and early settlers as
the "rock that burns," has long been known as a poten-
tial energy resource and today economically recoverable
reserves of shale oil are estimated to be 600 billion
barrels. By 1985 it is estimated that shale oil production
can reach 250,000 bbl/day at a business-as-usual supply
rate, and up to 1,000,000 bbl/day at an accelerated
supply rate.
In addition to the impacts directly associated with
the mining of the oil shale, ther.e are environmental
concerns related to anticipated high water consump-
tion-reclamation, revegetation, large land requirements
for spent shale and other solid waste disposal, air emis-
sions of particulates, nitrogen and sulfur oxides, hydro-
carbons, carbon monoxides and aldehydes, and water
effluents containing total dissolved and suspended solids,
organics, BOD and COD.
Geothermal energy may make a substantial contri-
bution to United States energy needs by the year 2000.
This technology is of particular interest to EPA in this
region because most geothermal activity is located in the
western third of the Nation and geothermal power is
most effective when transformed into electricity at the
wellhead.
The major environmental problems associated with
geothermal energy production are: hydrogen sulfide
emissions, saline water disposal, land subsidence, and
land use guidelines. Of the three types of geothermal
power plants either now in existence or under develop-
ment, the closed system hot water type, in which iso-
butane is vaporized by contact with the hot water
through a heat exchanger, may be least detrimental to
the environment.
EPA has given increased attention and funding to
the development of technologies for the conversion of
solid wastes to fuel. A number of technologies for this
recycling are now in the pilot stage or early commercial
operation stages. Resource recovery of both organics and
inorganics is of prime importance. Although the poten-
tially available energy is small, perhaps 1 percent, utiliza-
tion of these wastes will decrease problems associated
with landfills and result in a low sulfur fuel.
Nuclear energy, the nonconventional and controver-
sial energy source, bears close study. Radiation release
into the environment cannot be tolerated. As of last
148
-------
year, 6.1 percent of tlw Nation's d.~cll ical ~W1wralin!1
capacity was hy conv(mtional lission reactors, and fully
16 percent will be produced by 1980.
The solar energy resource system is a final alternate
technology. Such a system includes direct solar radia-
tion, wind power, organic farming, and ocean thermal
gradient energy. Land area requirements in this case are
considerable.
Our six-State region-including Colorado, Montana,
North Dakota, South Dakota, Utah, and Wyoming, is
being cast as the supplier of vast supplies of energy for
the Nation's needs. Over 50 percent of the Nation's coal
resources, almost 50 percent of the reserves, and more
than 60 percent of the strippable reserves are located in
these States. We supplied about 7 percent of the
Nation's 1973 coal production.
Essential)y all of the Nation's shale oil reserves are
located in the region's Green River formation. ERDA's
synthetic fuels program proposes two oil shale plants in
this area, each producing 50,000 barrels of crude oil per
day. Resource State governors have requested a final
veto power to opt out of this resource development
program.
Production of oil and gas in this region is expected
to increase slightly. We are sitting on top of 5 percent of
the Nation's oil reserves and 4 percent of the gas
reserves. Tight gas resources in our Region VIII States
are about 2% times the known conventional gas reserves.
This region contains over 40 percent of uranium ore
reserves and produced almost 60 percent of the national
1973 uranium ore output. ERDA has predicted that the
nuclear industry will satisfy 25 percent of U.S. electric-
ity demands by the year 2000. If that prediction mate-
rializes, this region will be called upon to supply a major
portion of the requirements.
Our limited water resources are necessary for hydro-
electric power. Our States rank second only to the west
coast in geothermal development. High rates of solar
insolation make this region an attractive solar energy
resource. Wind energy, which is in the research and
development stage, will benefit from above average wind
speeds in the Great Plains region.
People who live here are conservation-minded.
Conservation is extremely important to us because it
means using less gasoline, which will help us solve air
pollution problems we have in Denver, Salt Lake City,
and other western cities. Conservation means more effec-
tive development of energy resources, which means less
environmental degradation. Conservation means less
disturbance of the land, less water consumption, and less
impact on air quality.
We westerners are also concerned about the socio-
economic impacts of energy development. We have seen
how hooll1 (1lOwth allects drinkin~, waWr sllppli(~s and
also wastewater tr!!atment factilities. We see the n(!(!d lor
more schools, roads, and housing for miners, engineers,
construction people, and their families. Over 100 small
communities in Region VIII may be impacted with
growth rates up to 400-500 percent.
Some time ago, EPA recognized the need for a
coordinated effort to assess and minimize the environ-
mental impacts of a potential energy resource develop-
ment program in the western States. A coordinated
. Federal/State program to evaluate potential environ-
mental impacts from development of coal in the north-
ern Great Plains was set into motion in 1972. This pro-
gram came into being as a result of my letter to
Administrator Ruckelshaus and his subsequent contact
and recommendations to secretaries Morton and Butz.
EPA Region VIII instituted a total energy resource study
effort a little more than a year ago when I established
the Office of Energy Activities in order to carry out the
coordinated investigatory and planning functions which
we felt necessary. This office is in constant communica-
tion with State and local agencies and the public regard-
ing energy issues, and coordinates the baseline air and
water monitoring activities, technical investigations, and
integrated planning and management functions required
to determine and predict the before and after effects of
energy resource development in Region V III.
EPA Region VIII's efforts to assess and minimize
the environmental impacts from energy development fall
into six major categories:
1. 208 water planning activities,
2. air quality maintenance area (AQMA) planning,
3. community assistance programs,
4. 201 facilities design and construction,
5. State program assistance, and
6. Technical investigations.
Section 208 of the Federal Water Pollution Control
Act Amendments of 1972 provided funding to alloll\'
States to develop the technical framework for meeting
the 1983 water quality goal of "providing a water
quality which provides for the protection and propoga-
tion of fish, shellfish, and wildlife and provides for the
recreation in and on the water. . . ."
Areawide waste treatment management planning is
to be conducted by local planning agencies in coordina-
tion with the State environmental and planning agencies
. in areas designated by the governor of a State. Section
208 provides 100 percent funding for a 2-year pro!Jram
and will provide local !Iuvernment with resources in
order to plan and assess the environmental impacts from
energy development. Twenty-year projections of popula-
tion and industrial growth and the attendant water
quality impacts are to be analyzed. Local councils of
149
-------
government an: providiny this planning function
throU!lhout Rqion VIII. Over $12 million has been
obligated to thl~se planning agencies for the formulation
and implement
-------
4 November 1975
Se~sion III (con.):
RETROFITTING OUR PRESENT-DAY
ENERGY SYSTEMS
Roger S. Carlsmith
Session Chairman
151
-------
152
-------
ENERGY CONSERVATION TECHNIQUES
APPLICABLE TO THE IRON FOUNDRY'S CUPOLA
Dennis J. Martin*
Abstract
The cupola is the device most often used-primarily
due to economic factors-for the melting of gray iron.
Two events in recent years, however, have caused signifi-
cant difficulties in the use of the cupola. The first event
was the Clean Air Act, which promulgated an avalanche
of air pollution regulations. The second event was the
substantial increase in fuel prices usually associated with
the OPEC embargo. Added to these is the energy waste-
fulness of the cupola. All three problem areas are now
part of the difficulty involved in running a successful
and economical cupola operation. There do exist tech-
nologies to partially alleviate all three problem areas; this
paper concerns the results of an investigation into the
feasibility of these technologies.
ENERGY WASTE
The cupola is simply a device for melting iron to a
specific chemical composition and to a specific tempera-
ture. The physical characteristics of conventional type
cupolas are shown in figure 1.ln preparation for melting,
the bottom doors of the cupola are propped shut and a
layer of sand up to 10 inches in depth placed above
them up to the iron tapholes. Next, a layer of coke is
placed on top of the sand. This layer is then ignited,
usually with natural gas, and more coke is added to a
specified height above the cupola's tuyeres. Alternate
charges of coke (mixed with a fluxing agent) and iron
are then fed into the cupola through the charging door.
Combustion air is forced into the cupola through
the tuyeres. The heat generated by combustion melts the
iron charge and the molten metal flows through the coke'
layers to the sand bottom, where it can be tapped. For
an in-depth explanation of cupola operating practices,
the reader is referred to the manual entitled 'The Cu-
pola and Its Operation" prepared by the American
Foundrymen's Society.
The cupola is an energy-intensive device with over
50 percent of the heat generated being exhausted to the
atmosphere. For any good foundry operation, an energy
balance should be constructed. The procedure for doing
such, while somewhat involved, lays the basis for the
correct understanding of the cupola. An example of the
.Project Manager, Engineering and Applied Research Divi-
sion, York Research Corporation, Stamford, Connecticut.
results of a heat balance can be found in table 1 for a
cupola with such operating conditions as those I isted in
table 2.
Two other energy inputs to a cupola system that are
not usually considered are fan horsepower and after-
burner requirements. These would usually add 6 to 10
million Btu per hour to the energy input.
ENERGY WASTE - TECHNOLOGICAL SOLUTIONS
Recuperative Heat Exchangers
F rom the heat balance it is obvious that a majority
of the energy input to the cupola is wasted in the stack
exhaust. The most economical and useful method of
recovering some of this energy is by the use of a recuper-
ative heat exchanger. The heat recovered is then return-
ed to the cupola in the form of a hot blast. A schematic
Slock
Chor\lin9.
deck
Wind box-
Prop
Figure 1. Illustration of conventional I ined
cupola (ref. 1).
153
-------
Table 1. Heat balance of cupola
Heat input
l.
2.
3.
4.
5.
6.
Potential heat from coke
Carbon loss from dissociation
Carbon loss from iron pickup
Oxidation of iron
Oxidation of manganese
Oxidation of silicon
87,595,017 Btu/hr
5,751,896
3,873,136
111,480
405,216
735,840
Total heat input
Heat output
l.
2.
3.
4.
5.
6.
Iron
Slag
Dissociation
Stack gas
Stack gas
Radiation
sensible
latent
from cupola
Total heat output
/I
/I
/I
/I
/I
79,222,521
/I
31,326,180 Btu/hr
2,824,012
1,733,044
10,686,368
30,303,354
2,349,553
/I
/I
/I
/I
/I
79,222,521
/I
of a typical recuperative hot blast system can be seen in.
figure 2. Air to the cupola can be preheated to 1,200° F,
which can aCC(tunt for a maximum of 30 percent of the
energy in the slack gas of a medium-sized cupola.
There are sHveral possible variations of the normal
recuperative system. Generally the takeoff point to the
afterburner is located adjacent to or above the charge
door and is cornmonly referred to as "above charge door
takeoff." This oositioning allows enormous quantities of
tramp air to irlfiltrate the system. "Below charge door
takeoff" norm(llIy has the flue gas being drawn off from
4 to 6 feet below the top of the charge material. Infiltra-
tion is usually cut in half. A relatively new variation is to
increase the amount of charge material to the point
where the flue gas takeoff is below approximately 25
feet of charge This reduces infiltration to zero and
allows cleaning of the flue gas prior to afterburning. Care
must be taken "that air does not.enter into such a system
to avoid a CO or H2 explosion.
While recuperative heat exchangers are not a new
idea, it has only been recently that design problems
caused by high-particulate grain loading and temperature
have been overcome. As of yet, however, there are only
20 recuperative systems installed on iron foundries in
this country. Our investigation has shown that the equip-
ment on the market exhibits a high degree of reliability.,
The system that utilizes particulate removal prior to
afterburning has not yet become operational but is
expected to in the near future.
Other Waste Heat Uses
While there are numerous other possibil ities for
waste heat recovery, the dominating factor is the plant
layout. One foundry was able to utilize the stack gas
waste heat for plant heating but only because the neces-
sary duct work was already in place near the recupera-
tive heat exchanger. Other possible ideas, such as sand or .
coke drying, depend upon the proximity of the drying
154
-------
Table 2. Cupola data
Charging rate lb/hr
Coke rate 1 b/hr
Limestone rate lb/hr
Blast volume ft3/hr
Blast temperature, of
Tapping temperature, of
C02 content of stack gas, percent
CO content of stack gas, percent
Dry bulb temperature of blast air
Wet bulb temperature of blast air
Barometric pressure
Stack temperature, of
Tapped iron
-Carbon content, percent
-Silicon content, percent
-Manganese content, percent
Charged iron
-Carbon content, percent
-Silicon content, percent
-Manganese content, percent
Slag
CaO content, percent
FeO content, percent
CaC03 content of limestone
Fixed carbon content of coke
56,000
6,588.2
1 ,120
678,000
70
2,762
12.00
14.9
70
68
29.92
900
3.0
0.60
0.48
2.52
0.70
0.72
22.00
2.5
98
92
facility to the cupola. One idea that has been brought
up, bl:Jt not put into practice, is drawing off a portion of
the CO-laden air before the afterburners and piping this
to the drying facilities. The obvious difficulty in doing
this is preventing an explosion from occurring due to the
CO and H2 present. This is possible to circumvent, how-
ever, if infiltration through the charge door is com.
pletely eliminated and there is no oxygen in the system.
An ambitious s,cheme to produ~e electricity from
the waste heat has also been proposed. This does not
appear justifiable from an economic viewpoint. The
steam produced from a waterwalled combustion section,
however, can be used to drive the 10 fan and the com-
bustion air fan economically.
155
-------
AIII.IENT A.II
FR 0111 tlATURAL GAS H10
CHARGE 00011
COPULA WATER
OFF-GAS AFTERBURNE QUENCH
-
HIATED
A.II
TO
STACK
DIIIIOBAL
Figure 2. Schematic of heat recovery system.
AI ~ POLLUTION CONTROL
Air pollutic n regulations are necessary to preserve
the health and well being of our population. The regula-
tions, though,lave caused major difficulties for the
foundry industPI due to the fact that the emissions from
a cupola have inherent properties which make the emis-
sions difficult to control. The particulate grain loading in
a cupola's exhaust is very high. To control this within
acceptable limit;, a pollution control device must be
highly efficient (usually an efficiency of 90 percent
removal is the minimum required, while a removal effi-
ciency of 99 pE rcent is common). This efficiency re-
quirement limit! the choice of control equipment to
three devices; nnmely, electrostatic precipitators, high-
energy scrubbers, and baghouses.
The particle size of the dust emitted from a cupola
is also a factor in control device selection and design
since an appreciable percentage is usually below 10
microns in diam~ter (see table 3). This results in the
necessity of a pollution control device not only with a
high overall efficioncy but also with a high collection
efficiency for smf II particles.
Finally, the exhaust gases from a cupola have an
appreciable carbon monoxide content, usually averaging
around 13 percent. To comply with air pollution codes
that severely limit the amount of carbon monoxide
emitted to the atmosphere and to avoid the possibility
of explosions in the ductwork, the exhaust gases are
afterburned to convert the CO to CO2. The temperature
of the exhaust gas is therefore usually in the range of
2,000° F by virtue of the CO combustion. This tempera-
ture is, of course, far above what any pollution control
device can handle and consequently must be reduced.
The two most common methods of accomplishing this
are dilution with ambient air and water quenching. Both
of these methods, however, present further problems.
While diluting with ambient air is an effective method of
reducing the exhaust temperature, it also significantly
increases the gas volume through the system. Sincl: the
cost of pollution control equipment is directly related to
the air volume to be treated, this method would result in
a large cost increase- WaWr quenching systems, while
effective to a degree, arc not useful in reducing tho m
-------
Table 3. Particle size distribution - cupola emissions
(ref. 2)
Cumulative percent by weight.
Plant diameter in microns
-1 -2 -5 -10 -20 -50 -100 -200
1 30% 50% 65% 82% 90% 99%
2 64 82 98 99
3 2 12 34 92 99 99%
4 13 28 45 55 60
5 54 86 98 99 99 99
6 14 15 15 21 99
7 19 25 99
8 99 99
9 0.6 2 3 8 99 99
10 4 5.5 7 13.7 75 80
11 11 13 32 53 75 94
12 8 12 17 28 69 89
13 18 . 25 38 62
14 17 26 36 53
15 24 28 23 42
16 26 30 32 44
17 0 7 25 32 34 41 56 61
18 0 7 24 41 47 32 69 81
of the water flashes into steam. This excess water causes
damage to the refractory and ductwork. If a baghouse is
used after a water quenching system, cool spots in the
baghouse could allow condensation of the water vapor
on the bags. This increases' the weight on the bag, caus-
ing caking and bag failure. Due to the aforementioned
reasons, neither system is used exclusively but rather a
combination of the two techniques is employed.
With the necessity of employing a highly efficient
pollution control device that incorporates a temperature
reduction mode, the cost of such a system becomes an
important consideration to the foundry management.
Table 4 shows the economics of air pollution control for
the three most prevalent devices. As noted, the cost of
any pollution control device is dependent on flow. Using
ambient air as a temperature controller can increase the
flow by a factor of 4 and above. The cost of controlling
the pollution from a large cupola runs into the millions
of dollars. Costs for smaller cupolas are often in the
neighborhood of $1 million.
157
-------
Table 4. Cost equations (ref. 2)
-------------.------------ _._-- _.-------
--- --.-----------.---. --_. ------ .---
Investment Limits of Stand-
Equipment cost observa- Correlation F ard Data
type equation. tion coefficient ra t i 0 e rro r points
High energy 1=49,519+2.84 6,000 gas .82 25 16,000
wet scrubber gas vol vol 20,000 25
1=43,519+8.97 20,000 gas
gas vol vol 92,000 .99 139 29,000
Low energ,t 1=38,744+2.05 4,500 gas
wet gas vol vol 67,000 .84 55 22,000 34
Fabric fi"ter 1=-55,000+8.95 10,800 gas
gas vol vol 100,000 .98 321 48,000 19
AI R POLLUTION CONTROL-
TECHNOLOGICAL SOLUTIONS
Recuperative H'Jat Exchmgers
As noted previously, one of the more innovative
systems propCls!!d for recuperative heat exchanging
inherently reduces the cost of air pollution control. In
this section th ~ cupola is top loaded and the depth of
the charge material is approximately 25 feet above the
flue gas takeof".: point. Infiltration is effectively reduced
to zero. The flue gas is immediately quenched to about
1600 F and then scrubbed for particulate removal. The
gas is then aftmburned to recover the latent heat. The
primary advantage to this system is that the volume of
gas to be cleanml is at least one.fourth of the volume of
conventional systems; and the cost of pollution control
is appreciably rt)duced. .
Ordinary "balow charge door takeoff" is also effec'
tive in reducinlf gas volumes to be treated by approxi.
mately 50 percent, thereby reducing air pollution cost.
This alone is al1 appreciable improvement over existing
systems.
Steam-Hydro A" Cleaning System
This system solves two problems simultaneously. It
is a low-cost pcllution control device of high efficiency
and is run by waste heat from the cupola. The flue gas
from the afterbJrner goes to a waste heat boiler. Steam
produced creates the draft that draws the dust-laden air
into the system. The steam is injected into a mixing
tube, and encap>ulation, nucleation, droplet growth, and
massive turbulellt scrubbing takes place. The particulate
can then be removed by low-pressure drop cyclones.
Efficiencies for the device are above 99 percent and
since the draft is induced by the steam there are no
horsepower requirements. The system can also be
adapted to use the steam for other purposes. Startup
problems can be eliminated by using a steam.driven com.
pressor to store compressed air, which the system can
use instead of steam for pollution control.
FUEL COST
The cost of coke has risen from $44.00 per ton in
1971 to $112.00 per ton in 1975. Deregulation of oil
prices will probably increase this somewhat more. The
increased emphasis on the use of coal as a fuel and the
environmental regulation governing stripmining opera.
tions will further tend to increase coke price~ in the
years ahead. Such fuel price increases, if continued,
would make cupola operation uneconomical.
FUEL COST - TECHNOLOGICAL SOLUTIONS
Recuperative Hot Blast
The major benefit of hot blast is a reduction in coke
consumption. The relationship between the degree of
preheat and coke reduction is shown in figure 3 (ref. 3).
Converting from a cold to hot blast usually conserves
about 25 percent of the coke. For a 30 ton/hour cupola,
this could result in $500,000 saved per year. Also, since
the emissions from a cupola are dependent on the
amount of coke used, the efficiencies required for poilu'
tion control devices are reduced.
158
-------
300
275
.,;
..
...
z"
o
0:
...
o
z
o
I-
0:
...
.. 25
o
...
0:
:;
o
...
II:
...
"
o
u
225
200
10
AIR TEMPERATURE.,..IOO
Figure 3. Effect of hot blast on coke.
21 % O2
79% NO!
Reduce wind
by 11%
2 10 C FM
790 CFM
Divided Blast
The concept of introducing blast air into the fur-
nace through a double row of tuyeres has been around
for some time. It was not until recently, however, that
the control and monitoring system necessary for holding
the blast rate balanced have become avai lab Ie . The use of
a balanced blast on cold blast cupolas has been credited
with fuel savings of 20 to 25 percent of the coke. Opera-
tions in this country have been less successful since the
practice has been to have only one control system and
windbox. In Canada, where the divided blast has been
used successfully, the practice has been to use separate
windboxes and controls for both rows.
12
Oxygen Enrichment
By increasing the flow of air to the cupola, the rate
of combustion of the coke is increased. The use of pure
oxygen added to the combustion air allows for an in-
crease in melting rate without appreciably increasing the
air volume. This effect can be seen in figure 4. Using a 2
percent enrichment, the air volume necessary for the
same equivalent air is reduced by 8.7 percent. Higher
flame temperature and a reduction in the oxidation zone
results along with a decrease in the amount of nitrogen
supplied to the furnace. Using a 3 percent oxygen en-
richment, coke reductions of 10 to 20 percent are possi-
ble, depending on the foundry. The cost of the oxygen,
however, reduces the cost benefits of reduced coke con-
sumption.
23 CFM
~
187 CFM
210 CFM
23% O2
---
703 CFM
703 CFM
77% Nz
.._--------~-----_.. -.-------.------------.--.------
1000 CFM Air 890 CFM Air 913 CFM
Enriched Air
210 CFM OXYGEN
913 CFM Enriched Air
=
23% or 2% Enrichment
Figure 4. Equivalent enrichment.
C()"rt~sv 01 Air Producls and Chemicals
159
-------
SUMMARY
Our investigation has shown that the technologies
exist for foundries to economically combat pollution
wh ile conservi ng energy. 0 ifferent strategies, however,
must be adopted for different situations. A new plant
should have a layout that would maximize the use of
recovered heat. Existing plants must decide which tech-
nology would give the most economic benefit. For
instance, a plant with a cold blast and an adequate pollu-.
tion control system would benefit the most by simply
converting to divided blast, which is a relatively low-cost
technique but one which yields an appreciable economic
benefit.
Perhaps the most important aspect of these tech-
nologies is th ~t the small foundry operator now has
some cost-effective techniques that may be employed to
compensate for the spiralling raw material costs and
increasingly stringent pollution control standards that
have been threatening to force him out of business.
REFERENCES
1. Metals Handbook, 8th edition, Vol. 5, Forging and
Casting, American Society for Metals, 1970.
2. "Systems Analysis of Emissions And Emissions
Control in the Iron Foundry Industry," A. T.
Kearney and Company, Inc., February 1971, pre-
pared for the Environmental Protection Agency,
Contract No. CPA22-69-106.
3. J. E. Rehder, "Hot Blast in the Cupola," Modern
Casting, July 1968.
160
-------
ENERGY USE, EFFICIENCY, AND CONSERVATION IN INDUSTRY
Robert H. Essenhigh, Ph.D.*
Abstract
Energy conservation can reduce pollutant emissions,
reduce pressure on energy sources, and possibly reduce
manufacOJring costs. The basis for energy savings cam-
paigns must be: first, knowing current consumption and
efficiencies of use; and second, having targets of con-
sump tion and of use efficiencies for comparison. Current
use is determined by energy audits. Target use is deter-
mined on the basis of several factors, including thermo-
dynamic considerations. Furnace Analysis also provides
a basis for identifying the principal factors controlling
efficiency of furnaces; factors include output, excess air,
processing temperaOJre, etc. In evaluating and comparing
systems and tasks, four efficiences and three efficiency
ratios are involved: the operating efficiency of the sys-
tem (T/), its intrinsic efficiency (ex), its maximum effi-
ciency (T/':nax)' and the maximum task efficiency (T/':nax)
of the optimum system for the task; the efficiency ratio
of the system (Es = T//T/max)' the efficiency ratio for the
task (eT = T/1r1':nax), and the figure of the merit for the
efficiency of the system at its optimum efficiency (EST=
T/max/T/':nax)' The task efficiency ratio (ET) is shown to
be identical or nearly identical to the so-called "Second
Law" efficiency. Evaluation of the material deployed in
terms of efficiencies and the Furnace Analysis perform-
ance equations provides a basis for developing a sys-
tematic energy savings procedure. However, in spite of
low overall efficiencies (25 to 35 percent) in the indus-
trial sector of the energy market, where potential for
savings exists, it is nevertheless concluded that the chief
barrier to more rapid progress in energy conservation is
out-of-date or invalid cost accounting procedures.
INTRODUCTION
It is a truism that, if pollution emissions at any
moment in time are roughly proportional to energy use,
cQnservation of energy will greatly aid pollution control.
This is well recognized, but it has waited for the develop-
ing energy famine to provide the incentive for action
that will reduce pollutant emissions by this means.
Energy conservation indeed serves several joint interests.
. .
It can reduce pollutant emissions, reduce pressure on
*Professor of Fuel Science, Department 01 Material Sci-
ences, Pennsylvania State University, University Park, Pennsyl-
vania.
energy sources, and-sometimes-reduce manufacturing
costs. These triple objectives deserve more emphasis.
This paper will therefore focus on conservation, particu-
larly in the industrial sector of the energy market. This
particular sector has been chosen as the one that uses the
most energy, while at the same time it contains the most
diverse array of energy applications and yet has had the
least attention.
THE SCOPE FOR CONSERVATION
The starting point is, clearly: What is the scope for
conservation, and where does the energy go?
As a basis for the initial answers to these questions,
figure 1 illustrates the energy flow in the United States
in 1970 in the form of a Sankey or energy flow diagram
(ref. 1). This particular diagram has been reproduced
many times in the last few years, but it is reproduced
here once again to emphasize two points. The first point
is the small magnitude of energy supplied by electricity
to end use; and the second point is the magnitude of
waste, and the implied inefficiencies of use, in this modi-
fied diagram.
Taking first the matter of energy supply by elec-
tricity, the fact that this intermediate conversion market
of electricity generation draws on 25 percent of all gross
energy supplies and operates at an overall efficiency of
approximately one-third is well Known and has been
very widely quoted. The consequence of waste, however,
has not been so fully emphasized with respect to the
relatively small magnitude of energy flow to end use. As
the diagram shows, in the year 1970 this amounted to
about 8 percent of gross energy flow in the system. The
percentage of energy delivered by electricity to end use
is, therefore, 8 parts in approximately 72, which is about
only 10 percent of total end use demand. The point is
emphasized because of the feeling that still seems to
exist among many people that with the development of
nuclear fission or fusion power systems, all proble.ms will
ultimately be solved, and there will be no further energy
worries. It should be noted, however, that even if the
expected increase of nuclear production of electricity is
attained, and by the end of this century electricity wi II
draw on a gross of 50 percent total energy supplies, it
will still only deliver about 25 percent of end use
requirements, of which approximately 10 to 15 percent
will be derived from nuclear sources. This means that we
shall still have to rely on fossil energy sources even by
161
-------
-
0>
I'.)
NUCLEAR POWER 0.30
- HY ROPOWER 4.1 :t,
GROSS
CONSUMPTION
I - AIR CONDITIONING
2 - COOKING
3 - LIGHTING
4 - WATER HEATING
5 - MACHINES; APPLIANCES
6 - RAILROADS
7 - MARINE
8- AVIATION
SUPPLY
IN THE UNITED STATES IN 1970.
Figure 1. ENERGY FLOW f 1, Figuresarepercentage~~fg~oss
(Adapted from Earl Cook (re,) rt ) adopting use efficiencies as
. (' e less expo s . d
domestic consumption I.., . 1 ercent. (2) domestic an .
follows' (1) electricity generation, 3 ~ 20' Percent' (4) industrial,
, '(3) transportation, , -
ommercial, 50 percent, . '64 6x101S Btu per year =
c t G ross domestic consu mptlon. .
33 percen ,
2.2x109 kW.
-------
the end of this century for 80 to 90 percent of energy
supply to the end use markets. This reemphasizes the
critical dependence that we shall still have on fossil
energy by the year 2000, and the strong need to imple-
ment all conservation practices as soon as possible.
Coupled with the assumption of a still rising GNP over
the next 25 years, the absolute fossil fuel requirements
will likewise be rising, and the need to conserve fuel just
to control pollutant emissions becomes imperative.
The second point regarding efficiencies of use is,
therefore, even more significant. Commonly the energy
flow diagrams published elsewhere have shown approxi-
mately 50 percent of total energy going to waste. In
figure 1, this is increased to a little over 70 percent. The
two main changes have been: reduction of space heating
from about 75 percent overall efficiency to 50 percent;
and, even more drastically, a reduction on the industrial
sector of about two.thirds overall efficiency to about
one-third (which may still be high). This really targets
the objective of this paper. It is to explore the sources of
inefficiency of energy use in industry, and to develop a
set of priorities for action in conservation efforts.
Industrial energy use, overall, accounts for about 40
percent of total energy use, of which roughly 12 percent
is used for generating the electrical requirements which
delivers nearly 4 percent (as electricity) for use. The
remaining 28 percent is energy nominally delivered as
raw fuel to end use. This energy supply is delivered to
four subsectors. These are process steam, raised in boil-
ers; process heat, used in industr ial furnaces; electric
delivery, which is mainly electric drive; and feedstock,
which is quite minor. As can be seen from table 1, show-
ing the breakdown in percentage of industrial use, proc-
ess heat accounts for about one-third of all use, and
process steam nearly half. The problem of efficiency
exists only to a minor degree in the steam raising process
itself, because the boilers are overall respectably efficient
if properly designed, specified, maintained, and oper-
ated. The inefficiencies of industrial use come about
firstly, in the use of the steam in the varied processes;
and secondly, in the direct.heat use in the process fur.
naces. This article will focus mainly on the process fur-
nace subsector, which parenthetically will include the
problem of steam use in appropriate process operations.
The nature of the problem is that this process furnace
subsector is seen as a multitude of dissimilar functions
because of the sign ificant differences of equ ipment,
operation, control, end products, and so forth. This sub-
sector therefore lacks the common unity found in steam
raising, or in transportation, or electric generation, or
domestic and commercial space heating. It is therefore
claimed that argument from one industry or company to
another is invalid. One of the points to be made in this
article is that, contrary to general belief, the very wide
range of process furnaces can nevertheless be treated
largely as a single subset entity.
EFFICIENCY OF THERMAL SYSTEMS
The basis of efficiency analysis is the heat or energy
balance, which in principle is simple enough. In practice,
however, there are a number of subtleties that have
caused considerable confusion in the past, and would
appear to be continuing to do so. The sources of part of
Table 1. Energy use in industry
Percent of gross
national
use
Percent of net
industrial
use
Process steam 16.5 48
(boilers)
Process heat 11.5 33
(f u rnaces)
Electricity (gross) 9
(net delivered) 3 9
Feedstock 3.5 10
----.. ------
Total 40.5 percent 100 percent
----
163
-------
the confusion appear to be the methods of using the
laws of therm)dynamics and the sometimes conflicting
requirements ,)f dynamic behavior. This section will
present a specific view of efficiency that may be at vari-
ance in some I espects with some other presentations of
this subject.
Objectives of E. fficiency Analysis
The objectives of efficiency analysis are essentially
practical: they are to determine the energy flows in a
system, and from that to determine the scope for energy
savings. There are two ways in which this can be ac-
complished. The first is to examine a system in use and
to determine the scope for upgrading its efficiency; and
the second is 10 investigate the scope for substituting a
different energ'l supply system which may even involve a
different way altogether of achieving the same end
objective.
I n the fir,1 approach it is necessary, as strongly
emphasized by Lyle (ref. 2) nearly three decades ago, to
determine the ,Ictual use, or efficiency, and to compare
this with some target use or efficiency. The development
of targets is therefore central to any effective energy
analysis intendn! for practical action; this article there-
fore focuses al! a on the method of developing such tar-
gets.
Upgrading the system can include rearrangement of
existing pieces of equipment: for example, the rear-
rangement of a production line which, as actually carried
out in one instance, eliminated two reheat furnaces.
The seconc' approach is to investigate the scope for
using some m£ thod of upgrading the heat and fund-
amentally chan ~ing the heating system. Again reference
can ba made to Lyle (ref. 2a) for an early discussion on
the scope for de ing this by use of heat transformers.
The role of thermodynamics in these procedures is,
first, to aid the development of the heat balance, util-
izing the First Law; and second, to utilize the Second
Law for investigating the scope for upgrading the heating
systems themseves. These points will be amplified be-
low.
Definition of Ef.'iciency
Efficiency is something of an intuitive concept. The
most general de:in ition is: the ratio of useful return on
input. In the context of thermal processing, the useful
return can be measured by the energy in the product,
and input is measured by total Btu's supplied by fuel or
any other energy sources. Most commonly it is expressed
by the following equation:
Useful output, in energy units
Efficiency =
Total input, in energy units
Values of efficiency mostly lie below 100 percent.
Where efficiency exceeds 100 percent it is common to
divide the efficiency by 100 and rename it Coefficient of
Performance (a procedure described by Lyle (ref. 2) as
cowardice!}. An alternative specification commonly
found in industry is essentially the reciprocal of effi-
ciency: Btu per unit of product. As this is reduced, the
process is more efficient. This is also known as the
"Specific Energy Consumption,"
It must be emphasized that efficiency is the ratio of
a part to the whole. Since selection of the part involves a
value determination of what is useful, this means that
efficiency is essentially a value judgment, and, as such, it
is neither a scientific concept nor is it a thermodynamic
function in spite of the central place that efficiency
plays in part of classical thermodynamic theory. Effi-
ciency must be defined on a clearly specified device, in a
precisely specified function. Since efficiency is a value
judgment, its numerical magnitude can be changed by
changing one's mind. For example, a heat engine work-
ing as a heat pump has one efficiency, but the same
device working as a refrigerator has a different efficien-
cy. What has happened here is no change in the device,
but a change in the specified function. In discussions of
efficiency, it is sometimes found that devices are com-
pared while changing the ground rules, and invalid con-
clusions are obtained. Nature, however, has no concern
for efficiency: all energy is utilized by nature in its own
way, and the concept of value has no meaning. More-
over, two different living systems can very well place
different values on different components of energy sup-
ply. Consequently, two such different living systems
could produce significantly different conclusions about
the magnitude of efficiency. Those strictures, however,
do not necessarily apply when calculating the maximum
possible energy obtainable from a given device or re-
quired for a given task. Such cases ate governed by
natural laws, which properly introduces thermodynamics
and kinetics.
Thermodynamic Aspects of Efficiency
Thermodynamics enters efficiency analysis in three
principal ways. First, the First Law provides the basis for
the energy balances; second, the First and Second Laws
can be used to provide bases for constructing target effi-
ciencies; and third, the Second Law can be used to iden-
tify thermodynamic sources of inefficiency, in contrast
to operational sources of inefficiency.
The First Law energy balance provides the essential
basis for checking the accuracy of numbers to be used in
the efficiency calculation, and in that sense it is also the
basis for the efficiency calculation. On that account, the
164
-------
efficiency as defined by equation 1 is often called the
First Law efficiency, but this writer considers such a
name to be misleading as it could imply that selection of
the numbers to be used in the efficiency calculation is
governed by the First Law, which of course is not so.
The First and Second Laws are used to provide a
basis for constructing target efficiencies for different
systems. This requires the division of all thermal process-
ing devices into three subsets, as follows:
1. First Law limited system: process furnace,
involving heat-to-heat conversions.
2. Second Law limited systems: power devices,
involving heat-to-work conversions.
3. Combined heating and power where the maxi-
mum efficiency is still First Law limited, unless
a heat transformer is involved (ref. 2a).
The distinction between the first two subsets is
whether the device is a heat-to-heat conversion system or
a heat-to-work conversion system.
The target efficiency of power systems is well
known, with a long history, being nominally determined
by the Carnot efficiency. Less well known is the avail-
able heat* limitation for the process furnaces. Targets
may be specified as follows:
1. For First Law limited devices, the maximum or
target efficiency is equal to the available heat, using the
processing or stock temperature as the limiting tempera-
ture. The available heat is the heat in the thermal supply
(or flame), at the specified temperature of the thermal
supply, in excess of the processing or stock temperature.
This can be written approximately:
EfficiencY,Tlmax = (Tad - Ts)/(Tad - To) = (}'o
where Tad is the adiabatic flame temperature, or the
temperature of the thermal supply; Ts is the stock or
processing temperature; and To is the ambient tempera-
ture. Equation 2 is an approximation, but it is adequate
for our purposes here; it represents a linear decline in
efficiency with increasing Ts' In practice, the line is
slightly curved, as explained (with eXilmple) below.
2. The Second Law efficiency is given by the classi-
cal Carnot efficiency equation.
Efficiency, Tlmax = 1 - TofT
where T is the source temperature for the engine. In
passing it may be noted that the conventional proof of
. Available heat is a combustion engineering term, not to be
confused with Available Energy or Availability.
(2)
Carnot's theorem, stating that all reversible engines have
the same efficiency and that no engine can have effi-
ciency greater than the Carnot engine, is essentially spur-
ious (ref. 3). The conventional proof utilizes tWo engines
driving each other alternately forward and backward,
but exchanging heat only at the source and sink reser-
voirs. By exclusion of heat exchanges on the transiso-
thermals, the transisothermals must be adiabatics. The
conventional proof therefore establishes that the effi-
ciency of theCarnot cycle is the efficiency of the Carnot
cycle, and no more. Reference 3 provides the modifica-
tion to the conventional proof required to recover Car-
not's theorem. At the same time, however, it is also
shown that there are restrictions to application of the
theorem that are not normally recognized, and evidently
not adhered to in practice. In particular, distinction
must be drawn betWeen the reciprocating engine, and
the continuous engine in which there is also flow work.
Analysis of the continuous Carnot engine shows that it
has zero or negative efficiency. Use of Carnot's theorem
as a basis for target efficiencies, particularly of contin-
uous engines, is therefore cast in doubt. For the pur-
poses of this article, however, the conventional view will
be adhered to since it is appropriate in the context of
this paper.
3. The use of the Second Law to identify sources of
thermodynamic inefficiency, in contrast to sources of
operational inefficiency provides a basis for comparing
the inherent efficiency of tWo different systems. Opera-
tional inefficiencies, on the other hand, are concerned
with the behavior of a single system under different
modes of design or operation. In using the Second Law
for comparisons of efficiency, what we are looking at is
a comparison of the target efficiencies or efficiencies in
achieving a given Task objective by tWo different sys-
tems or devices. In addition, the use of the Second Law
for the comparison may be pointless or inapplicable
unless one or both of the devices involve engines lIr
energy transformers. Engines convert heat 10 work, or
vice versa. Energy transformers are devices to upgradl'
work or heat. Figure 2 illustrates tWo examples. Figure
2a is a simple work transformer, in which a system at
pressure Po is able to expand and, by suitable adjust-
ment of a fulcrum, is able to increase the pressure from
Po to P. Figure 2b is a particular example of a heat
transformer involving tWo engines, one working as an
engine and the other working as a heat pump, operating
betWeen two different temperature reservoirs. This com-
bination is not the only method of upgrading heat. Lyle
(ref. 2a) list three different methods:
a. Thermocompression-steam injector, mech-
anical compressor;
b. Boiling point elevation;
(3)
165
-------
[
[
p -
'f
o
L
Po ,A
T
Q
w
L
o
T'
To
Figure 2. Energy transformers.
166
Figure 2a. The work transformer:
p = po(Lo/L)
Figure 2b. The heat transformer:
T.T
0'= ~ 0
T -To
For T = 1,300 K,
T' = 400 K,
and To = 300 K,
0'/0 = 10
-------
c. Heat engine and heat pump combination as in
figure 2b.
Of these three methods, all discussed cursorily or
extensively by Lyle, the first has limited appJ.ications.
The second is still evidently an untested concept, even
though proposed 30 years ago. The third, however, is
currently of considerable interest as a topic for analysis.
The heat engine and heat pump combination is, of
course, well known as a device for generating efficiencies
in excess of 100 percent. As such, it is essentially a
traditional subject requiring no further comment here.
Somewhat related to the heat engine and the heat
pump combination is the use of a backpressure or pass-
out turbine in a combined heating and power system.
The common factor in the combined heating and power
arrangement compared with the combined heat engine
and heat pump arrangement is the thermodynamic
objective to avoid useless degradation of heat. Combined
heating and power is another traditional subject, also
well covered by Lyle (ref. 2b); of relevance here, how-
ever, is some additional history, and additional terminol-
,ogy in relation to this point.
Efficiency Ratio and Second Law Efficiency
The examination of the degradation of heat, and its
avoidance, has generated the possibly unfortunate term
of "Second Law" efficiency. I t can be argued that th is
possibly represents a misunderstanding of the meaning
and objectives of the Second Law, and its applications in
this context. The older terms Efficiency Ratio or Effec-
tiveness may be more appropriate. The use of the
Second Law is to identify the system arrangement that
will draw on the least amount of energy to make it
work. Having done that, the efficiency can be calculated
by equation 1, as before. Clearly, the practical test of
any improved system is that the fuel supply will not go
down so quickly for the same amount of production.
Consequently, this will be accurately measured by effi-
ciency as defined by equation 1. The effective or prac-
tical use of the so-called Second Law efficiency for
choosing operational devices must provide the same
resu It: i.e., a set of operational devices must be ranked
in exactly the same order by both methods, for other-
wise if they were ranked in a different order, we should
be able to choose a system on the basis of a better
Second Law efficiency that did in practice use more
fuel. The theoretical source of concern derives from the
awareness that entropy is increasing, or that energy in its
flow from one system or reservoir to another is being
degraded. This has, of course, been fully recognized in
the case of power systems for over a century. I n the case
of process systems, however, where the conversions are
heat-to-heat, the history is somewhat shorter.
The earliest paper on heat degradation in processing
that this writer is aware of is that by Thring in 1943, in
which he introduced the concept of the Virtue of energy
(ref. 4). Examination of his analysis shows that the
Virtue of energy is indeed equivalent to the net work
that could be obtained from a gross amount of heat if
this was utilized in a perfect cycle. As a fixed quantity
of energy flows as heat through a system, but being
transferred from one body to another at lower and lower
temperatures, the Virtue of that same amount of energy
declines progressively with temperature. The objection
that can be leveled at this concept is that Virtue is then a
quality attaching itself to something that only exists
when it flows, namely the heat flow. There is therefore
an implication that for this purpose, energy flow, and
particularly heat flow, can be treated as something tangi-
ble, such as a caloric. If this identification is acceptable
for purposes of argument, then the Virtye concept can
be regarded as valid.
A later study on somewhat similar lines was pro-
vided by Denbigh (ref. 5), who used a Second Law
analysis, again to develop a target for comparison of
efficiency. This was recently amplified further by
Riekert (ref. 6) (noting that he defines the normal effi-
ciency as the First Law efficiency, and again introducing
a Second Law efficiency). Similar views are taken in a
study supported by the American Physical Society (ref.
7), in which a Second Law efficiency is defined, but as
an operating efficiency rather than as a target efficiency.
What this does, in effect, is to normalize the true opera-
tional efficiency by dividing by an optimum efficiency
for the task, thus providing an effectiveness or efficiency
ratio (see appendix 1).
The comparisons of appendix 1 would seem to indi-
cate that the Second Law Efficiency must be used with
care. The test of an efficiency analysis is ultimately that
less fuel is used for a given quantity of product, which
means that a dip-stick in an oil tank, for example, will
show that the level has not gone down so much. In tha',
case, as remarked earlier, there cannot be any difference
in the outcome between efficiency defined by equation
1, wh ich is strictly a practical, operational efficiency,
and in the other definitions, unless the other definitions
have misleading results or are used for unintended pur-
poses. Clearly, if the alternative definitions advise us to
select systems that use more fuel, they have failed in
their purpose. If all definitions lead 10 the sam(~ OIH!ra-
tional result, then (!quatioll 1, !J(!ill!jtlw sirnplest, would
seem to be the olle to use in all operational CaSf!S.
Kinetic Aspects of Efficiency
The kinetic aspects of efficiency refer to the influ-
ence of time on the efficiency. In classical thermo-
dynamics, it is always assumed that heat exchanges haile
infinite time to be accomplished. Under such limiting
167
-------
theoretical cin:umstances, a source or sink can exchange
heat at identic ally the same temperature as a body being
heated or cooled. With a constraint on time, however, a
temperature difference must be allowed, and it is then
found in som!! circumstances that what is desirable for
high efficienc) from the thermodynamic point of view is
quite the 0ppl)site from what is required from the kin-
etic point of ',iew. Indeed, a key conclusion, discussed
below, is that efficiency (defined by eq. 1) increases
with temperature difference between the flame gases and
the stock or bad. It may be noted that operation in a
time frame does not require any alteration to the effi-
ciency definiti,)11 as given by equation 1, even though it
is no longer h£ at or work that is being used in a calcula-
tion but, instec d, heat fluxes and work fluxes, or rates of
heating and ntes of working (or power). This again
emphasizes that the efficiency definition used here is not
a thermodynamic efficiency. It is also noted that the
distinction betNeen heat and work on the one hand, and
rates of heat e (change and rates of working (or power).
on the other hand, are not normally made. The distinc-
tion is being rr ade here because, as already indicated, it
has been found to be important. For that reason it seems
desirable to di! tingu ish between heat and rate of heating
or process, just as a distinction is made between work
and rate of working (defined as power). In this article,
therefore, proc~ssing will be used as a parallel to power,
to mean a rat£ of heat exchange. It may also be noted
that, in what 101l0ws, virtually all the calculations and
examples will b~ concerned with systems in time.
Effidency Ranf'es and Limits
This section may be summarized by considering
briefly the I imi ts to efficiency as set by equations 2 and
3, and a brief Gomparison of the differences between
process and po\V~r. Referring first to equation 3, this is
well known, and indicates that the maximum efficiency
is limited by the temperature of the source reservoir,
with efficiency for a given reservoir temperature maxi-
mized by equation 3 if the exhaust to atmosphere is at
ambient temperature. If the exhaust is not at ambient
temperature, it is in principle possible to improve the
efficiency by including a heat exchanger in some way.
By contrast, alt'lough equation 2 provides something of
a parallel to thf Carnot efficiency limit, but for process
systems, nevertheless there is still a difference in that the
efficiency can be lifted, in principle, to 100 percent by
suitable heat recovery equipment. Equation 2 does re-
present a limit 10 efficiency, so long as the gases leave at
the processing t~mperature, Ts. Equation 2 calculates an
intrinsic efficiency, representing the condition at which
the maximum a:nount of heat goes into the stock, with
minimum stack or exhaust gas loss, and no wall loss. A
graph illustrating equation 2 is provided in figure 3. This
will be discussed in more detail later, but it shows essen-
tially the almost linear variation of efficiency with proc-
essing temperature for a variety of processes, with addi-
tionally the effects of excess air included, and the effect
of variable specific heat (that was ignored in formulating
eq. 2) also included yielding the curvature on the lines.
Clearly, as processing temperature goes to the ambient,
the efficiency rises towards 100 percent. (In practice,
there is a break at 90 to 95 percent, on account of the
loss of heat of condensation of the water from the fuel).
If, however, the process temperature itself cannot be
reduced to ambient, it is still possible, nonetheless, to
increase the efficiency of the overall process by heat
recovery. This is not indicated in figure 3; discussion of
this effect is provided below.
In essence, therefore, First Law limited systems can
in principle attain 100 percent efficiency on the basis of
the conventional method of heating furnaces, which is
direct or muffle firing; but cyclic power systems can
never reach 100 percent efficiency because of the Carnot
limitation (unless this is indeed avoided in special cir-
cumstances as discussed in ref, 3). When we are con-
cerned with heat-to-heat conversions or exchanges, how-
ever, in the First Law limited systems, then we can, in
principle, get around even the First Law limitation, if we
can develop an appropriate energy transformer. The heat
transformer of figure 2b wi II do this in principle, as it is
also doing it in practice in the case of some very low
temperature heating systems. What is needed for some
furnace operations is a heat pump that can deliver heat
at appropriate efficiency or COP, at significantly higher
temperatures than have yet been achieved with heat
pumps. Clearly, as T' approaches T (see figure 2bl. then
the value of the heat transformer progressively diminish-
es, partly on account of inefficiencies of the heat pump
itself, with COP's significantly lower than their theoreti-
cal maxima, and also on account of the need for a signi-
ficant temperature difference to enable heat to be trans-
ferred in a reasonable time. The time scales of interest
can be said to range mostly between 0.1 sec and 10 sec.
This will bracket the residence time of most gases in
most furnaces. On account of this short residence time
the temperature of, for example, a coal flame in a colel
wall boiler will generally lie between 2,800° F and
3,200° F while the theoretical adiabatic flame tempera-
ture is about 3,600° F, and the water walls are in the
range 200° to 500° F (sometimes up to red heat, 4000
C), with radiant superheaters at closer to 1,000° F.
These temperature differences do, of course, represent
substantial Second Law inefficiency, or virtue degrada-
tion, but they are at the same time necessary to accom-
plish the heat transfer within the time available. If a
168
-------
4000
IL.
0
CD
...
::J 2500
-
c
...
CD
~
E
{!!
0 2000
c
..
..
CD
U
.... 0
...
en a.. 1500
co
E
::J
E
.E
:e
1000
500
PROCESS
.
Alumino Melting Point
..
Mullite Melting Point
Blast Furnace
Silica Melting Point
0%
..
..
..
Open Hearth
Copper Smelter
-Silica and Basic Bricks: Rotary Kiln Firing
'Cement Roosting
Steel Welding
Gloss Melting and Forming
Stoneware and Porcelain
Steel Forging and Roiling: Lead Smelting
Lime and Dolomite Roosting
Copper Melting
Building Bricks
-High Temperature Coal Carbonization: Lead Glazed
""Steel Carburizing Earthware (Slipware)
-Vitreous Enammeling
"'Brass Melting
-Steel Annealing
Raku
Aluminum Melting
1
Excess Air (%)
-Metal Recuperator Limit
Tar Stills: Low Temperature Coal Carbonization
-J Soft Metal An.~ling
Domestic Oil Heaters (Kurylka)
Japanning Oven-
-Domestic Gas Heaters (Kurylko)
Drying Ovens: Boilers
1 BoB."
300% ......
-
......
-
-
-
--.....
~
00
10
Maximum Intrinsic Thermal
Figure 3. Variation of maximum intrinsic thermal efficiency with minimum processing
temperature for a range of thermal processes without heat recovery.
-------
1,0000 temperature difference is reduced to 10 to reduce
the energy de! radation, the area of the heat transfer
surface must bl: increased by 1,000 to transfer the same
amount of heat in the same time. It is therefore neces-
sary, when excmining efficiency, to take into account
the kinetic or time constraints as well as the thermo-
dynamic requirl!ments and constraints.
In summary:
1. The F"irst Law provides the numbers for both
ideal {thermodynamic) and for real time (kin-
etic) systems.
2. Value judgments are then used to select the
approJriate numbers from the First Law bal-
ances for the calculation of efficiency using
equation 1.
3. The First Law also provides the basis for eval-
uating target efficiencies of purely process
operations. Likewise, the Second Law provides
a basi; for determining target efficiencies for
power, combined power and heating, and heat
transformer systems.
4. The u timate basis for evaluation and compar-
ison of different systems remains the standard
efficie ley calculation based on eQuatjon 1 for
both deal thermodynamic and real time or
kinetic systems.
FURfllACE ANALYSIS: SUMMARY
Furnace aralysis is the formal study of the develop-
ment of equations for the efficiency of furnaces, where
furnace includes all such thermal processing devices as
boilers, stills, tc nks, autoclaves, and the like. In this sec-
tion the point is made first that these represent a re-
markable array of diverse systems, but at the same time
it is possible, a~. shown in the summary here, to develop
a single analysi'; for all these diverse systems. The ele-
ments of the furnace analysis have been described in
previous publicdtions (ref. 8) which also provide details
of the derivatio 15 of the equations and extensive review
of other work.
Systems of Com'ern: Classification of Furnaces
The most g meral furnace that can be described is a
rectangular, or substantially rectangular box, of high-
temperature ref 'zctory brick, comprised of a roof, four
walls, and a floor. This description represents quite a
wide array of fL maces, particularly of the smaller proc-
essing furnaces and many heat treatment furnaces.
Variations on this basic structure are then to be
found for genenlly good logical reasons. The essential
logic of the basic construction is self-evident; variations
1
in shape, size, and so forth are then the result of require-
ments in the different industries. For example, if the
material being processed is a liquid, the furnace bottom
must provide a bath: hence the Open Hearth furnace for
steel melting and refining, the reverbatory furnace for
copper, and the varied glass tanks, for example, If the
material tends to be lumpy, such as in smelting, cupola
melting, or lime burning, then the furnace is erected
vertically, to make use of gravity for feeding the material
through the furnace. At the same time, the shape
changes, for obvious structural reasons, from square or
rectangular cross section to circular. Additionally, if
there is fear of hangup inside the vertical shaft thus
formed, the walls are then flared out, as in the blast
furnace for smelting (and particularly iron).
As another aspect of logical development, the roof
of the typical furnace will collapse unless it is built as a
sprung arch, or unless (a relatively modern development)
the roof is flat but supported by an elaborate framework
above the furnace roof. With the sprung arch there must
be abutments to carry the thrust, and buckstays behind
the abutments to prevent the abutments being pushed
out, with tie bars holding the buckstays in position. If
the furnace is very large, the walls themselves must be
relatively thick, both to reduce wall losses and also to
provide the necessary structural strength required to
withstand both the weight of the roof and thermal
stresses, The crushing strength, particularly of the
bottom brick, must be appreciable. This generally re-
quires high-density brick, which correspondingly tends
to have high thermal conductivity. To offset the high
conductivity, it is then customary to provide some
degree of insulation, frequently by use of insulation
brick on the outside of the furnace. Since the bricks are
very expensive, it is also customary to optimize design
against cost by building large furnaces with two or even
three different types of brick, with the outer brick being
generally the more highly insulating and with the inside
brick able to stand the highest temperatures. The middle
bricks are therefore often of lower quality and carefully
selected so that their softening temperature will not be
exceeded in the furnace as designed. Other factors that
have to be considered in selection of the brick include
(as well as density and thermal conductivity) refractori-
ness, the crushing strength, refractoriness under load,
spalling, slag resistance (where this is important), and the
softening temperature or melting point. Therefore,
where refractories are carefully selected to meet the
maximum demand on the furnace, the unthinking addi-
tion of further insulation can lead to disaster, since the
internal temperatures will rise, and if they rise above the
design point the furnace can collapse. If further insula-
tion is to be added, it would be wise to recover the
170
-------
original design data, if these are still available, and recal-
culate the effect on the furnace to prevent the possi-
bility of collapse.
It may also be noted that the wall losses through
most furnace walls are 10 percent or less of the heat
supplied by the fuel. This rises as the operating tempera-
ture rises, and in special cases, for example, the glass
tank, it can be as high as 30 to 35 percent, where these
very high values are required to maintain the refractory
at a temperature below which they will not melt away or
collapse. For the higher temperature furnaces the ques-
tion of reducing wall losses is generally' a problem of
materials, not of failure to understand fundamental
principles.
There are other aspects of furnaces that must also
be considered; for example, furnaces are used in differ-
ent ways depending on whether the material to be heat-
ed is in units, or can be treated in a continuous flow. For
example, very large ingots being heated for stress relief
or soaking may be held in a furnace for periods up to 3
or 4 weeks. Such furnaces are generally referred to as
batch or "in and out." More commonly the batch fur-
naces are the smaller ones where material is being pre-
pared in relatively small quantities or small batches.
However, the ideal requirement, naturally, is to be able
to process the material continuously. G lass tanks are a
good example; however, many uniform solid units can
also be processed continuously, such as steel billets bei ng
pushed through billet heating furnaces, or parts for stress
relief, that can be glass, metal, nonmetal, etc., being car-
ried through the furnace on conveyers.
An intermediate type of furnace is the one that is
generally run over a period of months or years, such as
the Open Hearth, the blast furnace, and so forth, in
which the material is added in batches, processed over a
period of hours, removed, and a new batch added. In the
case of the blast furnace there is a steady burden moving
through continuously, with material added with some
degree of frequ~ncy, but slag and iron only removed by
tapping at very definite intervals of several hours. On a
daily basis, such furnaces appear to be batch or periodic,
yet on a monthly basis, or over the period of the
campaign lifetime of the furnace, the output can be
treated as continuous. This is important since such fur-
naces can likewise be analYl.ed by the furnace analysis
procedure, which originated with analysis of continuous
furnaces.
A further aspect of furnace variation is whether the
furnace is direct fired or indirect fired. If indirect fired,
it is known as a muffle, and in such furnaces the com-
bustion gases are separated from the stock being heated
to prevent contamination. A typical but not unique
design is the radiant tube furnace.
There are also, of course, additional furnaces whose
designs depart to a minor or a major extent from the
above general descriptions; for example; multiple hearth
furnaces, generally used for ore roasting; or tar stills,
which generally have bare metal tubes, although some-
times with refractory sections; or coal stills or coking
ovens. Nevertheless, the above general description pro-
vides a reasonably common context for consideration of
most furnaces. Table 2 provides a classification of many
furnaces based on the primary division into batch,
. periodic, and continuous, with the periodic further
, divided into 5 subsets. It will be appreciated from this
list how nominally complex the problem of furnace
analysis is. Nevertheless, we can still represent any fur-
nace as a box or enclosure formed by a thermodynamic
surface and across which surface there are energy flows
in and out. The First Law of thermodynamics then
enables us to set up an energy balance between the
flows, more particularly if the system is in steady state
and there is no net accumulation or loss of energy inside
the enclosure. This is the basis for the analysis.
Heat Recovery and Classification
Furnace analysis can be applied to a single furnace,
or to a furnace with heat recovery equipment, or to a set
of furnaces with or without heat recovery equipment.
Furnace analysis develops equations to predict behavior
of the single furnace unit, without heat recovery equip-
ment. It does not follow a priori that the same equations
will apply to the furnace with heat recovery equipment,
or to a compound set of furnaces. This has to be estab-
lished independently. To do this, furnaces are classified
according to their methods of heat recovery. The classifi-
cation used here follows:
Class 1 No heat recovery.
Class 2 Heat recovery on the gas side, for
example a boiler.
Class 3 Heat recovery on the materials side, for
example billet heating.
Class 4 Heat recovery on both the gas and
materials side, for example tunnel kilns.
In the unit furnace (the class 1 qesign with no heat
recovery) it is assumed that the material being processed
is essentially at the sam!! temperature. This is not also a
requirement for the gases. For such a system the maxi-
mum efficiency according to equation 2 is that obta ined
when the gases leave the furnace at the (uniform) proc-
essing temperature and there is no wall loss. Normally
speaking the gases do leave the furnace at temperatures
above the processing temperature, and this can be small
to substantial, meaning anything from 100° to 1,000° or
more. In addition, there can be anything from a few
percent wall loss, as in a boiler, up to 30 percent or more.
171
-------
Table 2. A classification of furnaces
A. Classification by type
1. Group I - Batch
Group II - Periodic
Group III - Continuous
2. a.
b.
[)irect fund
I ndirect fired (muffle)
B. Classifica-.:ion by heat recovery
Class 1 - r Jo heat recovery
Class 2 - lIeat recovery on gas side (e.g., boiled
Class 3 - Hoat recovery on materials side (I).g., billet furnace)
Closs 4 - ~eat recovery on both gas and materials side (c.g.,
'unnel kiln)
C. Type ar~~ification ,
1. Batch (in and our)
a. p,)! and crucible furnaces, e.g., melting - metals,
some glasses (day tanks), salt bath heat treatment,
antimony ore preparation, etc.
b. C~ramic kilns (j) for drying (stoves!, and Iii) for
fhng (e.g.. vertical downdraft I
c. Heat treatment - metallurgical and others (e.g.,
gl ass) reheat, soak, stress relieve. forge, etc. (many
are car bottom without or with recycle)
a.
2. Perioc ic (I ntermittent or semicontinuous)
Vertical shaft 1. Blast furnace for smelting - iron,
tin, antimony, lead, phosphorous
2. Cupola. for melting - cast iron
production; copper
3. Lime burning
4. Mercury ore
b.
Horizontal hearth
1. Open hearth - steel refining [also
AJAX, Brymbo]
2. Reverbatory - coppor "matto"
production; copper refining; Ni,
tin, bismuth smelting
c.
Stills
1. Horizontal coke oven retort .
coke oven and town gas
2. Metals refining. mercury, zinc,
cadium/zinc/lead
d. Converters
1. Bottom blown - iron (Bessemer).
copper
2. Top blown - LO; LOac (BOF);
KALOO
e. ' Others
1. Brick kiln
2. Fluid-bed devices
3. Gasifiers (water
!J
-------
transformer hecause of the high temperature the brick
must be raised to, to get the brick-forming reactions to
occur in a reasonable period of time.
Operational Performance of Furnrxes
For the purposes of analyzing efficiency, the be-
havior of interest is the variation of firing rate, or heat
supply rate with output, representing the heat removal
rate as useful heat. I n spite of the diversity of systems,
all furnaces are found to have effectively the same per-
formance characteristics so far as the variation of ther-
mal .input and thermal output an~ concerned. The varia-
tion of heat supply rate with heat removal rate is known
as the firing curve, and it is illumated in figure 4. Qual-
itatively this curve has a characteristic shape for all fur-
naces so far examined, and also, incidentally, for power
producing dev ices (ref. 81.
The characteristics of the firing curve to take note
of are as follows. First, when the furnace is at idle but
ready to produce, or the power machine is also at idle
ready to drive, output is zero, but an idle heat input or
zero load firing rate is always required. This serves to
balance the wall losses or the int'~rnal frictional behavior
of the device. When the furnace is producing, or the
machine is driving, energy is now being removed from
the system, and the firing rate must increase correspond-
ingly. The increase, however, is nonlinear with output
because, as the energy is withdrawn, there is an associ-
ated loss of heat to exhaust. A I inear increase of firing
rate with output would mean that each new increment
of fuel shared its heat with the output and the exhaust
loss in a constant proportion. This does not, in fact,
happen, because as firing rate increases the residence
time of the gases in the combustion chamber or furnace
falls. With less time to exchange- heat with the stock, or
to produce power, an increasing fraction of the heat
released goes into the exhaust. As the exhaust tempera-
ture rises (and a smaller fraction of the heat goes into
useful outputl. the firing rate must again be increased to
offset the diminishing transfer to useful output. The
consequence is the concave-upwards shape of the firing
curve, with a final limit as the firing rate theoretically
goes to infinity at a finite output.
Clearly this infinite limit is a theoretical condition
only, and not realizable in practice, because the fuel
supply system would obviously choke long before that
point is reached and the combustion efficiency would
also start to fall off drastically. It is still a limiting condi-
tion with the theoretical significance in that, at infinite
firing rate, the transit time through the combustion
chamber is zero, and if reaction of the fuel is infinitely
fast, the gases in the chamber will be at the adiabatic
flame temperature. At that temperature the maximum
o
60
30
8
50 I
8
40 /
8
Rate of Fuel I
Consumption in
30
Thermal Units 8
Btu per Hour I
8
20 /
8
/
8
10 /
8
/
8
. Idle Heat
o
o 10 20 x 106
Heat in Output: Btu per hr.
Figure 4. Variation of fuel consumption rate
with output for a continuous billet heating
furnace. In this figure the heat supply rate
and heat output rate have been plotted to
the same scale to emphasize the (normally)
great disparity between the two.
output of heat or work is then obtained. This provides a
basis for estimating the maximum theoretical output of
a furnace or engine.
Even more striking than the common qualitative
behavior of the firing curve is the establishment of a
common equation describing that curve. Writing the
thermal input from the fuel as Hf' Btu per hour, and the
equivalent useful thermal output as Hs. Btu per hour,
the common equation has the following form:
Hf = H? + Hsiao (1 - Hs/H~1
(51
173
-------
where H7 is t'1o idle heat requirement, H~ is the maxi.
mum output, and aO is the intrinsic efficiency. Using the
definition of oquation 1 we obtain from equation 5 the
expression for the thermal efficiency:
1/= Hs = a°[1-(HslH~)(Hs/H~)]
Hf {ae (H;/H~)+[1-ao(H;/H~)](HslH~)}
This equOition describes a curve having an asym-
metric inverted U-shape as illustrated in figure 5. It will
be seen from t:,is that efficiency reaches a maximum at a
particular or optimum output, There is a further intrin.
sic efficiency that can be defined, where we will describe
the efficiency of equatipns 1 and 6 as the operational
efficiency, an(! we will define the intrinsic efficiency as
;P
I
>.
u
c
CD
u
....
....
w
c
E
~
...
..c
~
U
ell
c:
c:
v
c:
o
CI
c:
o
o
...
...
Q,
o
70
(6)
being the limit of Thring's (ref. 4) Heat Utilization Fac.
tor, a. Thring argued that the idle heat was used only to
balance wall losses and the associated stack gas loss, and
that the wall losses were relatively insensitive to output.
Therefore, the energy available for transfer to the load
was only (Hf.Hf). He therefore defined the Heat Utiliza-
tion Factor as a factor that distributed the energy actual.
Iy available between the useful output and the stack
losses. This factor (a) is formally defined as:
a = H/(Hf-Hf) = a°(1-Hs/H~) .
(7)
Equation 7 is the equation of a straight line of nega.
. tive slope, with a maximum value (which is the intrinsic
.: 1\ Operat lanai Thermal Efficiency
Intrinsic 0: a Intrinsic Thermal Efficiency
aO Efficiency
Foctor
0"
o
"0
"0
,
o
. "
/::-H:~.~:~"
t... ~
. \............. . ~~
......, ...........~
.... Effect of ""'",-
" Wall Loss '. ,
.... Increasing ...".~
~....tI'al Ma,imom 00"0" H~' .,
~
4
60
50
40
30
20
10
.
:
.
.
0. =
HS
a
Hf - Hf
00
10
20
Heat in Output: Btu
Figure 5. Variation of operational and intrinsic thermal effi-
ciency with output for the billet heating furnace (see
figure 4); also illustrating the effect of increasing wall loss.
174
-------
efficiency) at zero output, and declining to zero at maxi-
mum output. This is also illustrated in figure 5, where it
is seen that the operational efficiency curve is tangent to
the Heat Utilization Factor curve at the maximum out-
put. Figure 5 also illustrates the method of obtaining the
limiting extremities, (x0, and H~', by extrapolation of
experimental data each way. This has been done for
numerous data sources obtained from the literature, and
the results have been tabulated in the previously refer-
enced articles (ref. 8). We may note that construction of
the Heat Utilization Factor line or curve is possible if we
have data for a firing curve, as in figure 4, with the
ability to extrapolate with some reliability to zero out-
put to obtain the idle heat. We
-------
furnace by itself can be increased again by preheat
because of th,~ rise in flame temperature and the increase
in the temperature difference. This results in a higher gas
exit temperature, but figure 3 only continues to apply
with correctic1ns to take the preheat into account. If the
preheat is by a separate preheater, there is a further loss
in efficiency represented by the increased fuel consump-
tion due to the air preheater, but there is the offsetting
increase in efficiency of the processing furnace due to
the higher in':ernal temperatures. On balance, the effi-
ciency of the combination can be increased. If the pre-
heat is provid,)d by a heat exchanger to recover the heat
in the exhaust gas, the efficiency of the combination can
be increased very considerably. It is also noted that the
improved effic iency by increased 6T is a kinetic require-
ment that is dimctly opposed to thermodynamic require-
ments for efficiency.
5. Load ,:missivity is also a factor that can have
very considerable influence on the efficiency. The effi-
ciency will dr,)p drastically as the load emissivity is re-
duced. In the limit of zero load emissivity (or more
accurately load absorptivity) the efficiency would be
zero if there VIas no convective heat transfer. It follows,
therefore, that the traditional conclusion that radiation
is the prime heat transfer factor is valid only for loads of
fairly high absmptivity. This is generally the case, parti-
cularly where the furnace is for high-temperature proc-
essing since absorptivities generally rise significantly with
temperature. Exceptions are certain types of close-
packed loads (in brick kilns) where very.high-velocity
burners (up tc 200 mph jets) have been found to in-
crease efficiency and throughput very considerably. At
lower processir g temperatures, the greater temperature
difference can make up for reduced absorptivity. Never-
theless, comparisons of different furnaces should take
the absorptivitv of the load into account as it is by no
means an unim,Jortant factor in determining efficiency.
Load absorptivity is, of course, something that is most
usually uncontlollable. Nevertheless. opportunities for
affecting it sho Jld not be disregarded. Variable absorp-
tivity is most :ikely to occur in the glass tank as the
result of highly rdlecting batch being allowed to spread
out over the surface of the melt. Some of the marked
changes in fuel requirements of operating glass tanks
could be due to this effect. In other systems. more con.
sideration mighl be given to packing or spacing of the
materials being heated, when this is possible. so that the
overall absorptivity of the assemblies is increased.
6. Wall Loss is also a very important factor affect-
ing the operational efficiency, but not the intrinsic effi-
ciency. at low outputs. The graphs are not reproduced,
but the effect oi changing the wall loss is to change Hf
in equation (51. and this is found to affect significantly
only the fuel requirements at the lower output condi-
tions. On the efficiency graph of figure 5, the curve, and
the position of the peak, is significantly influenced by
wall loss between zero load and the optimum load.
Beyond that, changes in wall loss have only second order
or less effect. Nevertheless, the results do indicate the
desirability of adding insulation in most instances, so
long as it is not likely to cause problems from overheat-
ing and melting or collapse of the inner wall sections.
7. Flame Emissivity has least effect of all the fac-
tors so far discussed. This is an interesting conclusion in
view of the attention given to flame emissivity in the
past. The reason seems to be that there is a close trade-
off between the increase in heat transferred from the
flame when the flame emissivity is increased, and the
reduced heat transferred from the walls due to the great-
er fraction being intercepted by the flame. An additional
compensating factor is the generally lower flame temper-
ature of the more luminous flames, partly because of the
greater thermal load due to the higher emissivity, and
partly it would seem, due to the longer burning time of
those fuels that are likely to generate more luminous
flames. At all events, emissivity would appear to have
negligible effect on efficiency to a first approximation;
any influence would seem to be second order. A further
consequence is that furnaces are apparently relatively
insensitive to change of fuels from the overall thermal
efficiency point of view.
Of the seven items listed above. all affect the opera-
tional efficiency, but only five affect the intrinsic effi.
ciency. In determining the intrinsic efficiency, the ther-
mal input rate data should first be corrected for the
excess air and then the holding heat, Hf, should be sub-
tracted out. The Heat Utilization Factor. a. can then be
calculated, and by plots. such as figure 5, the in itial
value; aO, can be determined. The values of aO can be
used as numerical estimates of the intrinsic efficiency for
comparisons of different furnace designs and operations.
The results of such determinations show that the intrin-
sic efficiency as defined here is evidently quite insensi-
tive to the design details. dimensions, burner placement,
fuel type, etc. These factors influence rather the heat
flux distribution inside the furnace while the gross or
overall behavior remains relatively independent of such
factors. It should therefore be remembered that the con-
clusions of this article relate to overall performance;
their application to local heat flux distribution problems
should be handled with great care. At the same time,
however, the relative independence of gross behavior
from the factors mentioned would seem to imply some-
what simpler internal behavior. with self-compensatory
mechanisms in operation, than is usually considered to
be the case.
176
-------
In summary, considering the seven factors signifi-
cantly affecting the thermal efficiency, we find: proc-
essing temperatures are preset, flame emissivity is rela-
tively unimportant, load emissivity is difficult or impos-
sible to control, and flame temperature is only controll-
able by oxygen enrichment or by preheat. Wall loss is
nominally controllable by added insulation, but even
this has economic limits where it Gan be applied, and in
very-high-temperature fu~naces, it can cause such over-
heating that it does more harm than good. Indeed, in
one particular case (of a carbon gl'aphitising furnace) it
was proposed that increasing the wall loss by water cool-
ing might permit higher outputs at a higher thermal effi-
ciency in spite of the increased wall loss. Consequently,
the scope for improved efficil3ncy lies principally
with: (a) better matching of the load to the furnace to
insure that operation is as close to the maximum effi-
ciency point, on average, as possible; (b) control of
excess air by proper fuel proportioning and damper con-
trol; and (c) by heat recovery.
4.5 Other Factors
The influence of other varied factors has also been
examined, either briefly, or in some depth. Summaries
of the results obtained follow,
1. The effect of heat recovery on the gas side was
examined in one of the previow; publications (ref, 8).
but it is now known that the conclusion then reached
was incorrect. The original analysis was based on the
widely used assumption that th~ fraction of heat re-
covered from the stack gas a:; air preheat can be
balanced, on a Btu basis, by a reduced firing rate. Closer
examination of the more complex firing equation thus
obtained showed that it leads to incompatible physical
behavior in the furnace. The a:;sumption is therefore
wrong and a better assumption was looked for. The
source of the error in the initial assumption turned out
to be that exact replacement of potential heat in the fuel
by sensible preheat means not only a reduction in firing
rate but also an increase in the residence time of the
gases, if the fuel/air ratio is maintained at its correct
value. Consequently, there is more time for the gases to
exchange heat and it is possible to reduce the firing rate
still further. It also follows that without this further
reduction the gases leaving the furnace after the air pre-
heat and the initial fuel reduction would be higher than
with the full firing rate without preheat. A better as-.
sumption is that the gas exit temperature should be the
same whether the air is preheated or not. With this
assumption it is found that as indicated, there is more
fuel saving than a direct tradeoff on a Btu basis. More-
over, it is found that the firing equation (eq. 5) remains
unchanged in form, with only modified values of the
maximum output, the intrinsic efficiency, and the effec-
tive idle requirement. The effect, as one might intuitive-
ly expect, is to increase the intrinsic efficiency and the
maximum output.
2. As mentioned previously, periodic furnaces run-
ning over a long campaign, with output averaged over
the day or the week, can be treated essentially as contin-
uous furnaces so far as the furnace analysis equations are
concerned. A practical problem of applying the analysis
does occur, however, when such furnaces are operated at
an approximately constant output point. This means
. that for all practical purposes there is only one effective
point on the firing curve, and extrapolation to the idle
heat becomes impossible. Nevertheless, if such furnaces
can be operated without output for experimental pur-
poses, the measurement of the gas exit temperature dur-
ing such operation will provide the basis for calculating
the value of the intrinsic efficiency, Assumptions about
the joint emissivity of heat exchange between the flame
and stock at the theoretical maximum can also be used
to estimate the maximum output, and it is then possible
to construct the curves for the efficiency and heat utiliz-
ation factor to check the relation between the actual
measured efficiency at the prescribed output. An exam-
ple is provided in figure 6 for a metal melting furnace in
which a number of different assumptions were made to
construct several different efficiency and heat utilization
factor curves. It will be seen in this instance that the
output was close to the optimum, but with an unexpect-
edly wide scatter in the range of efficiency from 20
percent to nearly 40 percent. Further analysis of this
particular situation established that this variation in effi-
ciency was most probably due to the different forms in
which the metal scrap was introduced into the furnace,
some being ingots, some being fine baled wire and sheet,
and some being plain sheet. This analysis indicated,
therefore, that to improve efficiency one significant
point could be the method of baling the scrap and in-
troducing it into the furnace.
3. Little has been done yet on extending the anal-
ysis to batch furnaces. Essentially, a periodic furnace is a
batch furnace with fresh material inserted as soon as the
previous sample has been removed. This means that
there will be a minimum reinsertion rate above which
the system behaves effectively as a continuous furnace,
but below which the heated material is up to tempera-
ture and just remaining there at a nominally idle condi-
tion. For such a furnace it can only be said that the
efficiency will then start to drop off quite drastically,
depending quite strictly on the number of units being
processed in unit time. If the number of units being
processed is very small in relation to the minimum re-
quired to approximate continuous operation, it implies
177
-------
aO
o
,
,
"-
,
...
"
,
,
~...\, "",
~ "
,.~. '"
'.,:-... "
," ""
' .
"
....~ ',". "
. ""....' ......... '"
.....' .
...' "
...' . ...
... , ....... ...
" . ""
..., .......
" .
" " ,
". ,
,'........ ,
". "'
"......... ,
~. ,
''.,., " . ..........
" ......... m "'
"~.... . " H s . E = 0.4 ...,
~ " \
m - '" "
Hs . ~ - 0.3- , . ~
'" ..........
" ,
'" .
"'-.
6 7 8 9 10 " 12 13 !4
Output BTU/hr ft2x 104
~ 60
o
I
00
o
N
.- 40
::>
-
o
Q)
I 30
"t)
c
o
-
-
w
10
00
_...!.-- I
I 2
I
5
I
.3
I
4
. = a Heat Utilization Factor
. = 7J Efficiency
15
Figure 6. Variation of operational efficiency and heat utilization factor (intrinsic effi-
ciency) with output for an aluminum reverbatory melting furnace (with curves
drawn for different flame and load emissivity assumptions; see text).
that the furnace is significantly oversized for the job,
and the cure (fin,ances permitting) would be to replace it
by a furnace beuer matched for the duty.
The elements of the analysis have, however, been
applied with sc me success to another type of batch fur-
nace, a vertical downdraft kiln. This was fired steadily
on a preset program to give a rising temperature over a
period of about 2 weeks. Treating the brick inside the
kiln with its ri~ ing temperature as equivalent to a steady
removal of he a: from the firing gases, it was found that
the equations cescribed above could be somewhat modi-
fied to fit this case. This is an area, however, in which
more research h; required.
4. Furnace analysis has been applied successfully to
the electric furnace. There is of course no stack gas loss,
but there is still an idle condition. The variation of stack
gas loss, howev~l, is replaced by variable wall loss, which
becomes proportionately much more important in such
furnaces. It was possible to establish from the analysis
that although the operational efficiency of the system
lay between 10 and 15 percent, nevertheless, the intrin.
sic efficiency was very good, being about 60 percent.
The system to which the analysis was applied was a
Sanders graphitizing oven. This is a furnace with a design
such that there is recuperative preheat between carbon
blocks entering and leaving the furnace. The high intrin-
sic efficiency is due to this recuperative element of de-
sign.
5. It is also remarkable that the influence of fuel
type, i.e. coal, oil, or gas, is relatively unimport,lI1t in
affectiny the efficiency of furnaces. In tho first place,
the effect of flame emissivity has alrciJdy been discussed
above, with the findin(J that then! is a significant trade-
off between increased heat from the flame as flame
emissivity increases and decreased heat from the roof
and walls as the flame becomes blacker due to its inter-
ception. A second factor contributing to the insensitivity
of fuel type is the remarkable uniformity of adiabatic
178
-------
4000
ADIABATIC
FLAME TEMP.
OF ........."'. ... ......... ..........
A.... ........
"".. ....................................... 8H
.............. I 2
.""'c:>;.. COG 3500
8CH4
Rich Gases
flame temperatures of fossil fue Is burning in air. This
derives from the purely statistical result that the amount
of heat in 1 cubic foot of a stoir:hiometric fuel and, air
mixture is approximately 100 Btu's for virtually all the
fuels. This rule tends to drop off only in the producer
gas and blast furnace gas region. What this approxima-
tion implies is that 100 Btu's are being released into 1
cubic foot of mixture. Since in most mixtures the air
outweighs the fuel by a factor of 10 to 1 or more, a
reasonable approximation is that in the majority of
cases, 100 Btu's are being released into 1 cubic foot of
cold air. The approximate calculation shows that the gas
temperature reached in such circumstances is approx i-
mately 3,600° F. A more accurate calculation that takes
into account the increasingly significant weight of fuel as
the heat of combustion or the fuel drops is illustrated in
figure 7 for fuel data obtained by Rosin and F reidlander
(ref. 9). As can be seen, the range is reasonably approxi-
mated by the dotted line which is given by the following
equation:
1
Liquid Fuels
:-Colculoted
~ T = T + 3750-
: oc 0 1+750/h'f
Low Limit
2200°F
Tad = To + 3750/( 1+ 750/hf)
(9)
where hf is the heat of combustion of the fuel. The form
of the equation can be obtained from a more accurate
expression of the air requirements for burning fossil
fuels, as follows:
hf/Gs = A/(1+B/s . hf)
(10)
where A and B are empirical constants whose values are
given in table 3; Gs is the stoichiometric ratio of air to
fuel by weight; and s is the specific gravity.
EVALUATION
The purpose of furnace analysis is to develop a plan
for action that will lead to energy savings. The evalua-
tion procedure that will lead to any savings, however, is
complex. The fundamental basis is clear enough: it is a
comparison of actual use against targets for use. This is
3000
2500
2000
70000
o
4000
8000
12000 16000 20000 30000
Heot of Combustion - Blu I lb. (hf)
40000
50000
60000
Figure 7. Variation of adiabatic flame temperature with
heat of combustion. (Note: change of scale at 20,000
Btu/lb.)
179
-------
Table 3.
R.~lation between heat of combustion
(hf) and stoichiometric air require-
ments (Cs) and common fuels adapted
from Rosin and Friedlander (ref. 9)
(hf/Cs - A/(l + B/s.hf)
Heat of Specific
combustion gravity
Fuel hf (Btu!lb) A B s
Solin fuels
-- 10,000 1,375 890 (1)
10,000 1,375 1,000 (1)
Liquid fuels
18,000
1,637 4,233
(1 }
Gaseous fuels
Blast furnace
gas
Btu 1ft.
',.000.1.500 1.749
o
1.0
90-11 5
Producer
gas
1,250-4,000 1,596
o
0.9
90-300
Blast furnace
gas + coke
oven gas
1,500-8,000 1,302 .210 0.75:1:0.25 90-450
Medium Btu
gases
8,000-14,000 1,275 -320 0.45:1:0.05 300-500
co 4,368 1,780 0 0.97 341
H1 61.035 1,780 0 0.07 343
CH. 23.863 1.386 0 0.55 1.067
simple enough: the complexities emerge in the details.
The first step-l he comparison of use with targets-yields
estimates of lo.ises. It is then necessary to evaluate the
losses, and cla;sify them according to their different
possible source!. The nature of source of loss can affect
the way in which the actions to save energy are imple.
mentlJd. The initial evaluation must be against targets set
for the eQuipm(!nt and plant as it exists. Subsequent to
that, there wol.'ld then be a review of the processes to
determine what alternative procedures or processes
could be adopt.~d, and set new targets accordingly. We
therefore have an initial set of five steps, as follows:
1. Deterrr.ination of use - this is the energy audit,
2. Setting targets for the ex isting equipment.
3, Compa,'ison of use with targets to determine
sources and magnitudes of losses,
4. Review of overall procedures to determine pos-
sible alternative processes and set targets based
on alternatives, and
5. Developing and implementing plans for action
to save the fuel.
1, The energy audit is a procedure with a well
established history of four or five decades at least. Lyle
(ref. 2), for example, provides numerous examples both
of energy balances or audits on individual items of
equipment, and also on complete plants. What is impera-
tive, however, is the use of a systematic procedure in
developing the audit. The following is a description of a
typical procedure in a relatively simple case. The essence
of the procedure is to divide the whole plant up into a
number of smaller and smaller subdivisions and sub-
subdivisions defined by thermodynamic surfaces. Across
these surfaces there will be energy flows in and out: the
energy flows between all the different subdivisions must
of course balance overall. One objective of the division is
to reduce a plant to a set of energy conservation centers.
where each center will be the basis for an individual
evaluation. In this way, small flows that are small in
relation to the total energy use in the whole plant may
show up as significant on their local scale. For the simple
system we can then construct the following six set
sequence as follows:
a. Plant balance: all sources of energy supply into
the plant are identified and summed (not forgetting
potential chemical energy to be released in processes
other than from fuel). All energy flows leaving the plant
are identified and summed (remembering that. for exam-
ple, iron as metal represents an energy flow out as poten-
tial chemical energy with respect to its oxide if the mate-
rial arrives as oxide). The difference between energy in
and out is the total losses and errors. to be accounted for
in the subsequent enclosure balances. This is a first gross
target figure.
b, Conversion units balance: if electricity and/or
steam or some other energy stream (such as producer
gas) is generated on site in a central conversion unit, an
energy balance on such a unit is identified as a "second-
ary" energy supply. The balance on all such conversion
units will of course provide a figure for losses that in this
instance can be split into stack gas losses and "wall"
losses.
c, Intershop balance: an energy balance between
the total gross {)nergy supplied to plant minus that used
in the conversion units plus the "secondary" output
from the conversion units and minus the sum of the
energy supplies to each individual shop is now con-
structed, The difference will represent intershop trans-
mission losses and errors of measurement. Such a bal-
ance is not always possible since the individual shops are
180
-------
not always metered individually for all their input
energy flows. Where such individual metering is not
employed, clearly it should be considered as a future
task. In the absence of such metering, estimates of trans-
mission losses can be worthwhile and may already be
available in the files as original design calculations. The
results of this sequence of calculations is to generate a
value for the net energy supply to the process shops.
Subtracting the total energy out of the plant as in a.
above leaves a net target figure to be accounted for in
the remaining enclosure balances.
d. Shop balance: steps 1. and 3. are now repeated
on the individual shops inciudin~1 step 2. if appropriate,
and with the inclusion in some instances of a tertiary
energy supply such as sensible heat of ingots, etc. This
shop balance yields a difference that is the sum of the
intrashop losses plus measurement errors that have to be
accounted for in the final process units balance. Shop
heating requirements are included at this stage.
e. Thermal process units balance: each individual
process unit (furnace etc.) is now subjected to the same
procedure yielding a balance that should provide data
on: ener~y in product, energy in the stack gas, and
energy in wall losses. At this point, discrepancies in the
balance are solely errors of measurement, and these are
frequently found to be considerable.
f. Final balance: in principle, the data for a total
balance in detail for the whole plant are now available.
Assuming this to be so, with unaccountable losses re-
duced to acceptable, if not tolerable levels, the data are
now reduced to graphical flow form (Sankey diagram) in
which energy flows are represented by bands of width
proportional to the quantity of energy to be repre-
sented. At the same time, the numerical balance provides
a gross energy input to the pic nt, a net output in the
produce, a sum of all identified losses, and a balance of
unidentified losses with errors (of measurement. As men-
tioned previously, the identifiecllosses split in two: the
losses during process; and inte"process losses. The dis-
tinction between the two is important since the method
of dealing with each category is fundamentally different.
This completes the process of "getting the num-
bers" that action must be based on, with the data now
presented visually in a Sankey (Energy Flow) diagram,
first for the whole plant; then for the onsite conversion
units, then for the intershop balance, if this is signifi-
cant; then for the individual shops; and finally for each
individual thermal process unit. The diagrams then also
provide ready summaries for plant managers, plant engi-
neers, shop managers, and so on.
2. Setting the targets has already described in out-
line above, and in greater detail in the previous publica-
tions (ref. 8). Further comment is redundant at this time
except to emphasize that this first target setting should
be for the equipment as it exists. Having set targets, the
Sankey diagrams for the individual energy conservation
centers and then for the total plant should be redrawn
with the targets as the basis for the revised Sankey dia-
grams.
3. With the energy audit and the target settings
complete, the basis then exists for their comparison and
for the determination of losses on the plant as it ex ists.
The losses can then be identified as due to a number of
different causes: we may identify poor operational
procedures, inefficiencies of furnace and/or plant design,
and losses due to the use of less efficient systems than
others available.
a. The poor operational procedures can be technical
procedures or even scheduling procedures. Reference to
figure 5 shows that efficiency peaks are at an optimum
output or throughput. If scheduling of material through
the plant causes wide swings in output, this can result in
unnecessary reduction in efficiency. First, of course, it is
necessary to carry out appropriate experiments and
measurements on the furnace to determine where that
optimum may be.
The poor operational procedures may include such
factors as uncontrolled excess air, as discussed previous-
Iy, or for maintenance that allows excessive air in leak-
age, and so forth.
b. Considering inefficiencies of plant design, one
can turn up several unexpected factors. For example,
one point already discussed above is the matter of de-
signing the furnace with an optimum wall thickness and
specified brick, costed on the basis of costs prevailing at
the time the furnace was designed. This can mean that
design procedures developed 20, 30, or 40 years ago may
lead to unnecessarily thin quantities ,of insulation, and
unnecessarily low qualities of brick, in the light of
day's fuel prices. If the furnaces were built at the time
the original calculations were made, then they represent
an optimum for that original period, and it may not b.!
possible to do too much about it now in terms of added
insulation, as also discussed above. What one should bear
in mind, however, is the use, particularly, of computer
programs containing value judgments of relative signifi-
cance of cost that are now out of date but are still being
used. Again, for a furnace already built, it may be too
late to do much about it; but, clearly, it is an obvious
move to check on the assumptions in optimizing costs in
new construction.
A closely related point is the question of wall thick-
ness and need for insulation on day furnaces. Many years
ago, calculations were carried out that showed that, in
dense brick, the heat stored in the wall was anything
between 10 and 100 times the hourly loss rate through
181
-------
the wall. Therefore. it was argued. one lost more energy
heating up tt'-e wall to operating temperature. day by
day, if one had a thick wall than was lost by conduction
and convection through the wall. Therefore there was
littlp. point in having thick walls. so long as they were
strOllg enough for structural integrity. I ncreased hourly
loss, integratee] through the day, was still likely to be less
than the amolnt that had to be stored in the wall before
the furnace was up to temperature. Consequently, a rule
of thumb eml!rged according to which "day furnaces
need little insJlation." This is in fact unsound on two
grounds, yet t appears that many day furnaces were
built according to this recipe. The two grounds
are: first, that 11 furnace does not cool down completely
overnight, so 1:hat it is not in fact necessary to have to
bring it back up to temperature from cold everyday--
indeed there i, a tradeoff: the more heat stored in the
walls, the largm the time constant for heat loss, and the
closer to operating temperature the furnace may be the
following day; second, this prescription ignored the
development elf light weight, therefore, low-density in-
sulation brick (particularly the hot face insulation brick
that can be used with adequate support from a steel
frame structun and in many cases without the use of
dense brick). It is doubtful that furnaces today are de-
signed on such principles, but since the life of a furnace
can be. if necessary, several decades, it is not impossible
that day furnal:es built according to that prescription are
still in use.
A third pJint is the problem of heat recovery. It
looks like a very simple procedure to recover waste heat
from the exhaust of a hot furnace. However, it turns out
that there are restrictions not commonly appreciated.
The chief proble;-n is the use that the heat recovered is to
be put to. ThE simplest use, of course, is probably to
raise steam in a waste heat boiler. However in the case of
a large plant w th a larger number (say 100 or more) of
high-temperatu:es furnaces, each with its individual
stack releasing into the shop, the amount of heat so lost
is very substan tial; but to collect it all in 100 or more
waste heat boilers is obviously absurd, and to collect it
all for diversion to a single waste heat boiler would re-
quire an almost impossible. spaghetti-like array of ducts
from the furnaces to the waste heat boiler. The more
obvious solutio.1 is to install an air heater. But this too
presents proble.ns. If the temperatures are significantly
above about 1,0000 F, then the degree of air preheat will
still be restrictE d to about that because of the need to
use a parallel flow recuperator since otherwise the allow-
able metal temperatures may be exceeded. One solution
to this problem is to use more expensive. i.e_. more
refractory metals for the recuperator, but this would
make the return on investment look even worse than
usual. If ceramic or refractory heat exchanges are used.
these are famous for cracking in due course, and leaking,
and the introduction of combustion products into the
preheated air stream can cause severe problems in some
circumstances. There is a further problem of how to use
the preheated air. Particularly where so many furnaces
are fired by gas inspirator burners. drawing atmospheric
air by virtue of the jet action of the inspirator, the pro-
vision of preheated air can mean removing all those
burners and replacing them entirely. Again this means
the use of metals capable of withstanding the necessary
preheat, (in addition to the capital investment of the
replaced burners). Frequently it turns out that the re-
turn on investment does not justify such procedures. at
least on the basis of conv{!ntional costing. It is easy to
say that heat is being wasted. Methods of recovering that
heat are also well known. Nevertheless. it turns out that
frequently the recovery methods are too expensive to be
justified.
4. I n the review of possible alternative procedures
and processes, this is where attention can be given to
such things as rearrangement and redesign of the plant
layout, investigation of the potentially possible applica-
tion of heat transformers, and so forth. This is where the
so-called Second Law evaluations of efficiency may be
appropriate.
5. The final step in evaluation is to develop and
implement an action plan. The all important factor here
is establishing probable magnitudes of expected savings
and using, jointly. the magn itude of the savings and the
expected cost of implementation as a basis for priorities.
A set of do's and don'ts without any evaluation of prior-
ities can be worse than useless. It can lead to tying up
expensive equipment and manpower on some jobs that
provide relatively trivial savings. while the big losses go
unremarked and unchecked. A common procedure has
been to pay marked attention to savings in electricity,
for example, particularly by turning off the lights. How-
ever, if turning off the lights saves 10 percent of the
electricity used for lighting, which is itself 20 to 40 per-
cent of electric use, the overall national saving is less
than 1 percent. hardly justifying the excessive attention
given to lighting. There can indeed be alternative reasons
for emphasizing saving on electricity: on occasions it
can provide a useful savings in cost and it also has the
advantage of immediacy and a form of visibility. The
psychological effect in a plant of showing that action is
being taken can be most important. At the same time,
however, this can backfire completely if plant operators
are aware of much more significant factors that manage-
ment are ignoring. The overwhelming advantage of the
systematic evaluation procedure, on. the other hand, is
that it compels attention to the factors of prime import-
182
-------
ance. As Lyle (ref. 2) frequently observes, the systematic
data gathering procedure can also lead to an immediate
awareness of how energy stream!; previously considered
waste can be immediately and rel;:ltively cheaply diverted
to use. Consequently, therefore, the systematic data
gathering procedure and evaluation tends to run several
of the steps together, with chang~s in procedures, plant
. layout, and ordering installation of new equipment
sometimes taking place almost at the same time. The
essentials, however, remain unchallenged: the energy
audit is required to determ ine th~ actual use of energy;
the development of energy use targets is required for
comparisons and identification 01 losses, and to provide
the basis for effective action for sclving energy.
CONCLUSIONS
We may draw a number of sequential conclusions
from the foregoing review, starting with the question of
the approximate overall efficiency of energy use in
industry. This as remarked earlier is usually quoted at 70
to 80 percent. First, we can see from table 1 that half
the energy used in industry is for raising steam. Effi-
ciencies on this in a well-run boiler plant can be 75 to 85
percent. There are, however, many boilers throughout
industry that are given far too little maintenance and
other attention, and efficiencies of some smaller boilers
could easily be as low as 50 percent even if we take 70
to 80 percent as bracketing thn average efficiency of
steam raising in the first place. Then we have to consider
the efficiency of the actual use of the steam. Good esti-
mates for the efficiency of use of the steam are hard to
find. We may, however, note that in the last year or two
there has been much comment on two sources of waste
of steam: first, excessive steam leaks to the point that a
steam cloud above major steam using plant was said to
be quite a common sight; and :iecond, there has been
further comment on the excessive use of steam where
steam is used in processing. In addition to that there are
the actual thermal losses through the walls of the steam
heated reactors and sensible he;:,t losses of effluents in
some cases. The overall efficiency of use might be rea-
sonably estimated as 50 percent. This gives us an overall
efficiency, including the steam raising itself, of 35 to 40
percent of half the energy use:n industry. To that we
can then add the efficiency of the heat used in thermal
processing as direct heat. This is the furnace sector we
have been mainly discussing; and here we may refer first
of all to figure 3 as a basis for estimating. In that figure
we see that maximum intrinsic thermal efficiencies of
the higher temperature reactors are under 50 percent to
start with. We then add to figure 3: the increased gas
exit temperature above the processing temperature, a
higher actual processing temperature than the minimum
listed in figure 3, plus excess air that can be anything
from 20 to 100 percent, plus wall losses that may run
anything from a few percent to 35 percent. The final
figure for efficiency for the higher temperature furnaces
is then found to lie on average somewhere between 10
and 20 percent. This in fact is the range given by Trinks
(ref. 10) for steel plant furnaces. The average would be
about 15 percent, and Trinks also makes the point that
many low-temperature furnaces are even less efficient
because the stock cannot take the high-temperature
gases which are therefore cooled by very high excess air
before admittance to the furnace. The third component
of industrial use is electric drive, which is very much
more efficient in its initial conversion of electric power
to shaft output; however, as this represents such varied
uses as driving fans, pumps, and mechanical devices in
furnaces such as conveyor belts, and so forth, again the
overall use will not be as efficient as all that. The final
efficiency of the industrial sector would therefore
appear to I ie somewhere between 25 and 35 percent, and
it could indeed be as low as 20 percent.
This would seem to indicate a potential for major
improvements in efficiency of energy use. Examination
of the sources, however, indicates that the biggest loss is
still a matter of high-temperature gases leaving the high-
temperature furnaces. Secondly, there is no doubt that
there are major losses due to bad operation, particularly
the age old problem of too much excess air. It is not
,fully appreciated by most operators, and even more so
by most company administrators just how much money
is lost due to excessive excess air. The difficulty of
appreciating this derives from the inability to see the air
traveling through the furnace, and the failure to meter it
by weight. It is not appreciated, for example, that very
often the greatest weight of material moving through an
industrial furnace is nitrogen from the air. To provide a
comparison figure: a 1,000-MW electric power statior'
will use about 400 tons of coal an hour. It will also
require about 5,000 tons of air an hour, of which 4,000
tons per hour is nitrogen. A single plant processing 4,000
tons per hours, continuously for about 11-1/2 months in
the year, day and night, week by week, would normally
be regarded as a major plant. Yet when it comes to use
of air, the point is missed entirely. What is very often
needed here is a relatively small investment in gas anal-
ysis equipment for each furnace, and proper (preferably
automatic) draft control. The stack gas losses that a
company has to live with, unless it installs expensive
heat recovery equipment, are already substantial: there
is no need to increase them unnecessarily by failure to
control the excess air.
183
-------
By comparison, the attitude towards steam leaks
appears to hav,~ undergone a transformation with the last
1 or 2 years. --he simple calculations to show the magni-
tude of losses, and then therefore the potential for dollar
and fuel savin~~;, have been carried out. Moreover, the
steam is very I'isible. The consequence recently has been
the provision by many companies of maintenance teams
dedicated to ':he one job of tracing and curing steam
leaks. At the same time there have been parallel reports
of inquiries in ~o the real steam needs of different plant
components, particularly in the chemical industry, with
the conclusion reached in many cases that steam use was
excessive or glossly excessive. Actions to cut down on
fuel by cutting down steam demand have been evidently
very active.
The actiors to be taken, in general, to reduce losses
and reduce em rgy demand therefore tend to fall mainly
into three principal categories:
1. Tightening up maintenance and operating proce-
dures, as for eX3mpie eliminating steam leaks or reducing
unnecessary excess air, or providing better maintenance
of furnace fabr cs to reduce or prevent air inleakage.
2. Conside ration is also being given to installation
of new add-oil equipment, notably heat exchangers.
However, to H is writer's knowledge, there has not yet
been too much activity in that direction. The equipment
is expensive, ar d experience has shown that it generally
requires very t ght maintenance and operation by more
highly skilled, md therefore more highly paid workers.
The question .)f costs is therefore crucial, and more
often than not it is still reported that the cost of such
installations car not be justified.
3. The third action can be altering the process. This
can either be a 'earrangement of the existing equipment,
or attention to such new factors as the heat transform-
ers. For example, Powell (ref. 11) describes an improve-
ment in procedures in a transportation equ ipment plant.
An axle forging process required three basic operations;
billet to roll-formed blank, blank to preformed forging,
and preformed 10 finished forging. Because of the layout
of the system e.lch of these operations started with cold
metal stock. By rearranging the location of equipment
into a single continuous line, the use of heat was reduced
by two-thirds and the labor and handling needed was
also greatly reduced.
Investigatiol1 of energy transformers, on the other
hand, has so fa.- been far less successful, at least with
respect to industrial operations. It is also generally
recognized that the high-temperature operations will not
benefit from sue h devices. Nevertheless, the time can be
seen where a variety of new approaches to heating at the
lower temperatures utilizing the smaller temperature
differences will come about. It should certainly not be
thought, as does seem to be implied by some recent
writers, that the potential of heat pumps and the like has
gone unremarked. Neither has the significance of energy
degradation been missed. These points, as indicated pre-
viously, have in fact been under fairly active and detailed
discussion for a period of 30 to 40 years. The principles
have been appreciated; what has been missing up to now
has been the necessary equipment.
It is indeed a striking fact that what is needed to
improve efficiency of industrial furnaces, and industrial
use of energy, are actions that have been known for
decades. Technical difficulties are not really a problem.
The principal reason for now embarking on much more
extensive energy-saving operations is costing, which in
turn depends on the price of fuel. What does not seem to
be generally appreciated is the relatively small cost of
energy as a percentage of the cost of manufacturing. For
heavy manufacturing, such as steel, glass, cement, heavy
chemicals, and so forth, the cost of energy is between 10
and 25 percent of the manufactured costs. But for the
whole of industry, as an average, the cost is about 5
percent, having evidently risen to that from 4 or 4.5
percent in the last 1 or 2 years. Since there is a major use
on the high side of that average, there must also be
major use on the low side of the average to obtain the
average. On the high side of the average, companies have
generally known of the importance of saving energy, and
within what was always permitted by standard cost
benefit analyses, or the return on investment, substantial
investment was indeed made in energy saving operations.
With the further rise in the cost of energy, more energy
saving operations can still be undertaken and show a
return. On the low side of the average, however, the cost
of energy as a percentage of manufacturing costs was,
and still is, so low that companies paid little attention to
it in the past, and still pay relatively little attention to it
today even with the drastic rise in the cost of fuel. The
central issue is always the return on investment, and
even with the cost of fuel as high as it is today, the
return on investment is still too unattractive to pay
much attention to much energy saving beyond immedi-
ate concerns such as turning out the lights. This is essen-
tially a microeconomic point of view.
From the macroeconomic point of view, on the
other hand, what we then see is that all fuel not saved
and not domestically available has to be brought from
overseas, thus representing flow of dollars overseas that
are not used inside the country for further capital invest-
ment. Even if the money does return from overseas as
capital investment from outside, the return on that still
leaves the country, and again from the macroeconomic
point of view this represents a significant loss that would
justify very much more intensive efforts toward~ energy
184
-------
saving. This would seem to indic ate a direct contradic-
tion between the microeconomic point of view (or the
point of view of the individual company) on the one
hand, and the macroeconomic point of view represented
by the country as a whole. There now seems to be some-
thing of a growing appreciation that this contradiction
exists, and it would seem to be indicated that the resolu-
'tion of the contradiction may po!;sibly lie in the recogni-
tion that fuel has a value that exceeds its cost. This
excess value over cost exists because, within the country,
fuel is in limited and diminishing supply. Clearly, if the
value exceeds its cost, and if value is then used in place
of cost in return on investment calculations, it must fol-
low that some or many decisions about whether to make
the investment or not would tun,. in favor of the invest-
ment.
The question therefore aris~s: on what basis can
value be estimated as a multiple of costs? One simple
way of viewing this is to ask the additional ques-
tion: what does a company stand to lose if it does not
have the fuel? This is the question that a number of
companies found the answer to by involuntary experi-
ment in the last 2 years when they had no fuel and
temporarily had to close down" They generally found
that there were at a minimum close-down costs, startup
costs, union benefits, maintenance expenses, fixed
charges, and so forth. As the very simplest method of
obtaining an estimate of the ratio of value to cost, one
can simply look at the value o'f product per dollar of
fuel, which is just the reciprocal of the cost of fuel as a
fraction of manufacturing costs. If energy costs are only
5 percent, the reciprocal view states that there are $20
worth of product for every dollar of fuel. In the case of,
for example, a company marlufacturing high-quality
glass products, which sell for many thousands of dollars,
and the gas used is a very sma II quantity, the value of
product per dollar of fuel can run between $100 and
$1,000, or even $10,000, in special cases. This starts to
emphasize very clearly the valu€' of that fuel if it is not
available. Indeed, the reciprocal 'factor of dollar worth of
product per dollar of fuel is a rough estimate of fuel
value, and has been used, for t!xample, by Powell (ref.
11) as an estimator for a multip.lier to convert fuel costs
into fuel value. It is also interesting to note that this
approach emphasizes to the small user just how import-
ant it is to him also to take steps to conserve energy,
where the large user with fuel as a large percentage of his
manufacturing costs knows this already.
This is a problem that might perhaps be called the
cost accounting barrier. It also !;eems to this writer to be
the most important problem to overcome if energy
saving is to become realistic on a large scale. This writer
is able to say at present from h is personal contacts that
u
some of the larger companies are engaged in energy sav-
ings campaigns in which justification for improvements'
to be carried out is made on the basis of such things as
improving a product mix or updating an old component
or plant, rather than the precise objective of saving
energy merely to justify the expenditures and surmount
or tunnel through the cost accounting barrier. The use of
the cost multiplier to convert cost into value, with value
used in the return on investment calculations, would
seem to be another possibly effective way of achieving
the same end. This it would seem would be worth ex-
ploring further.
In summary, therefore, we conclude that the effi-
ciency of energy use in industry is probably between 25
and 35 percent, and could be lower. The principal source
of inefficiency would appear to be the sensible heat
losses from many high-temperature furnaces from which
heat is discharged at very substantial temperatures to the
atmosphere without any heat recovery. There are also
large losses that combustion engineers have complained
about for decades in the excessive use of excess air,
where the use of relatively simple gas analysis equipment
could produce very large savings with a payback in some
instances of months, if not weeks. There is also scope for
substantial tightening-up of operating procedures, im-
proved maintenance, and so forth. There is also need for
closer examination of new plant layout, revisions to the
processes, and possible benefits from examination of
heat transformers. The biggest barrier of all to improved
efficiency of energy use, however, is not technical, it is
financial: it can reasonably be called the Cost Account-
ing Barrier. The cost of energy as a percentage of manu-
facturing costs is still on the average so low that pur-
chase of necessary energy-saving equipment such as heat
exchangers, and so forth, is still difficult, if not impos-
sible to justify on current accounting rules. This perhaps,
is the area where, by reassessment of accounting rules
and recognition that the value of fuel exceeds its cost,
there could in fact be the greatest benefit in improvin!l
efficiency of energy use in the shortest possible time.
APPENDIX 1
SUMMARY OF THERMAL EFFICIENCY
AND EFFICIENCY RATIO EQUATIONS
The operational efficiency of any system is only a
partial gu ide for evaluating the system performance. The
operiJtional efficiency is certainly the central factor
regarding actual fuel consumption, but it also needs to
be compared with the best efficiencies that the system is
capable of, and als.o compared with the best efficiencies
that some other optimum system is capable of. The
maximum efficiencies can be used for normalizing the
185
-------
operational e':ficiencies, thus generating the efficiency
ratios.
Efficiencies
Four effi:iencies can be defined, with some iden-
tical in some i:lstances (see table A 1):
1. T/, the operational efficiency, defined by equa-
tion 1;
the intrinsic efficiency, defined by equation
5
3. 71max' the maximum possible operational effi-
. ciency of the system of concern; and
4. 71~ax' t~e maximum possible efficiency of the
optimum system that will achieve the given
task (best task efficiency).
2. 0:°,
With further definitions below we have:
71
Q
= Eu/=g
= Eu/:Eg-Eg)
= Eu/:!},m
= Eu II:T ,m
(Al)
(A2)
(A3)
(A4)
71max
71~a)(
where, with un ts of either heat or work:
Eu usef J~ energy used or delivered by system,
Eg = gros; energy actually supplied to system,
E3 = "idh:" point energy,
(Eg - Eg) = net I/tilizable energy supplied to system,
Eg,m = minimum gross energy required by system,
and
Ei,m = minimum gross energy required by the opti-
mUlTI system that will achieve the given task.
It is noted that the minimum required input energies
(Eg.m and ET,nt) will be equal to the available energies
(6B) where:
81 = U + p(V . T oS for batch systems
B" = U + pV - ToS = H-ToS for flow systems
[ = H - ToS+ (V2/2g) + Z for general flow systems]
Efficiency Ratio,
These are df'fined as:
€s
€T
EST
= 71lnm 1)( = Eg,m/Eg
= 71ln~,I)( = ET ,m lEg
= 71maxhl~ax = ET,m/Eg,m'
(A5)
(A6)
(A7)
Equation (,l15) comPilres the system, as operating,
with itself at optimum operation. Equation (A6) com.
pares the syster1 as operating with the optimum task
system. Equatio.'1 (A7) compares the optimum of the
system in use with the optimum task system.
When Eg is supplied as work, equation (A6) is clear-
ly the "Second Law" efficiency as defined in reference
7. This is also found to be true for energy supplied as
heat from a reservoir, and approximately true for energy
supplied from fuels. Reference 7 compares the "F irst"
and "Second Law" efficiencies for nine single source and
single output systems based on the following:
Sources (1) work; (2) energy from fuel; (3) heat
from a reservoir .
Outputs (1) work; (2) heat supplied to a device; (3)
heat extracted from a device.
The matrix of each source satisfying each output
generates nine systems.
Table A 1 compares the same nine systems for three
of the four efficiencies (with Q appearing only in systems
five and six), and for the three efficiency ratios. The
efficiency commonly referred to as the "First Law" effi-
ciency is 71; that commonly referred to as the "Second
Law" efficiency is €T' However, the rationale for choos-
ing these two as the "First and Second Law" efficiencies
out of the six or seven options is not clear.
REFERENCES
1. E. Cook, "The Flow of Energy in an Industrial
Society," Scientific American, Vol. 224, 1971, pp.
134-144.
2. O. Lyle, The Efficient Use of Steam, Ch. 20, "The
Heat Balance," H.M.S.O., London, 1947 (1st ed,l,
1958 (2nd ed.).
2a. Ibid, "Stepping up Heat," Ch, 25.
2b. Ibid, "Combined Power and Heating," Ch. 3.
3. R. H. Essenhigh, "Enquiry Into the Validity of the
Thermodynamic Assumptions Regarding Efficiency
of Power Cycles," Proceedings of Frontiers of
Power Technology Conference, Oklahoma State
University, September 1971, pp. 7.1-7.29.
4. M. W. Thring, "The Virtue of Energy, Its Meaning
and Practical Significance," J. Inst. Fuel, Vol. 17,
1943-44, pp. 116-123; also, "The Science of Flames
and Furnaces," Ch, 2, The Thermodynamics of
Furnace Heating, Wiley, N.Y., 1952, 1962,
5. K. G. Denbigh, 'The Second Law Efficiency of
Chemical Processes:' Chem. Eng. Sci., Vol. 6, 1956,
pp.1-9.
6. L. R iekert, "The Efficiency of Energy Utilization in
Chemical Process," Chem. Eng. Sci, , Vol. 29, 1974,
pp.1613-1620.
7. W. Carnahan, K. W. Ford, A. Prosperetti, G. I.
Rochlin, A. H. Rosenfield, M. H. Ross, J. E. Roth-
berg, G. M. Seidel, and R. H. Socolow, "Efficient
Use of Energy: A Physics Perspective," report to
186
-------
Table A 1. Efficiencies (71 and a) and efficiency ratio (E) for single-source, single-output systems
(adapted from table 2.3 of ref. 7)
Note: Task efficiency ratio, €T' is equal to or approximately equal to "Second Law" efficiency of ref. 7.
--
CD
'-I
. Source T/ &f Energy supply from fuel Heat supply from hot reservoir
Demand Work supplied Es = Ws Es = toH :: toBs Es = Os extracted at T
T/ = WulWs = T/ (1) = Wu/toH = T/ (2) = Wu/Os = T/ (3)
Useful
wur~ T/max '" 1. TofT ad (engine)
del ivered = T/~ax = 1 = 1 . TofT
T/~ax T/max = 1 (fuel cell)
fS = T//T/~ax = T//(1 . T ofTad) = T//(1 . TofT)
Eu = Wu = T/
fT = T/ ::Wu/toBs
fST = 1 = (1.TofTad) = 1
(4) (5) (6)
Useful T/ = Q~lWs = T/ = Q~/toH = T/ = Q~/Qa = T/
heat = (Tad.T')/(Tad" 0) = a = (T.T')/(T.T 0) = a
supplied T/max = 1/(1 . TofT') (heat pump)
T/~ax = 1/(1 . TofT') = (1 . T ofT)/(1 . TofT')
Eu = ~ fS = T//a = T//a
Process at = T/(1 . TofT') = 11(1 . T ofT')/(1 - TofT)
fT = T/(1 - TofT')
T'>To = 1 = a(1 - TofT') = a(1 . T ofT')/(1 - TofT)
fST
= Q~ IWs = T/ (7) = Q~/toH = T/ (B) = Q~/Qa = 11 (9)
Useful 11
heat = (1. TofTad) /[(T "/To).1)
extracted l1max = 1/[(T"fTo)-1J = (1.T o/T)/[(T'/ fT 0).1)J
T/~ax = 1/[(T" /To) - 11
Eu = ~ fS =T/[(T"/To)-1J/(1.TofTad) = l1[(T"/To)-1J/(1 . TofT)
= 11 [(T"fT 0)-1J
Process at fT = 11 [(T" fT 0).1)
T" < To fST = 1 = 1-TofTad = 1
-------
the American Physical Society: National Technical
Information Service Doc. No. PB-242773, U.S.
Department of Commerce, January 1975.
8. R. H. Es::enhigh and T. Tsai, "Furnace Analysis
Applied to Glass Tasks," Glass Industry, Vol. 50,
pp. 333-3:18 and 378-387, 1969; Vol. 51, pp. 68-72
and 108.111,1970; R. H. Essenhigh, "The Potential
for Impro'/ing Furnace Performance," Proc. Fuels
Utilization Conference, Cleveland State University,
October HJ72, pp. 106--136; R. H. Essenhigh, A. C.
Thekdi. G. Malhouitre, and T. Tsai, "Furnace Anal-
ysis: A Ccmparative Study," Combustion Technol.
ogy: Soma Modern Developments, Chapter XIII,
Academic Press, 1974; R. H. Essenhigh, "Factors
Determining Efficiency of Furnaces," Conference
on Fuel Efficiency in Industry, The Pennsylvania
State University, April 1974; "Optimum Effi-
ciencies 01 Furnaces: Setting the Targets," Indus-
trial Heating XLI, 1974, pp. 21-24; "Evaluation of
Fuel Consumption Rates and Thermal Efficiency of
Automobil,~s by Application of Furnace Analysis,"
J. Transportation Res., Vol. 8, 1974, pp. 457-464.
9. P. O. Rosin and G. Friedlander, "The It Diagram,"
in Technical Data on Fuel H. M. Spiers, ed., Brit.
Nat. Comm. World Power Conf., 6th ed.. 1961, pp.
98-109.
10. W. Trinks and M. H. Mawhinney, "Industrial Fur-
naces," Vol. 1,5thed.,1961,p. 111.
11. A. S. Powell, "Energy Use Efficiency Costs and
Values," Conf. on Fuel Efficiency in Industry, The
Pennsylvania State University, April 1974,
See Also
E. A. Bruges, "Available Energy and the Second
Law Analysis," Academic Press, 1959.
ACKNOWLEDGMENTS
Preparation of this paper has been made possible by
contributions to the Cooperative Combustion Labora-
tory Fund by the ALCOA Foundation, Combustion
Engineering, Exxon Research and Development Corp.,
The General Electric Foundation. Mobil Oil Corpora-
tion, and PPG Industries Corp.
188
-------
4 November 1975
Session IV:
CENTRAL POWER STATIONS
Fred L. Robson, Ph.D. *
Session Chairman
*Chief, Utility Power Systems. United Aircraft Research Laboratory. East Hartford, Connecticut.
189
-------
190
-------
WASTE HEAT UTILIZATION/REDUCTION
A. G. Christianson and D. J. Cannon*
Abstrac t
Projections indicate that waste heat rejection from
central power stations in the year 2000 will nearly equal
the total U.S. energy consumption in 1970. Although it
is a waste product under our current methods of opera-
tion, a significant amount of this waste heat could be
utilized. Three of the more promising options for waste
heat utilization/reduction are (1.1 byproduct waste heat
utilization, (2) byproduct electdcal generation by in-
dustry, and (3) integrated energv production! use com-
plexes. The implementation of these options could yield
sizable short-term, midterm, and long-term environ-
mental, economic, and energy conservation gains; how-
ever, because of the long lead times involved for plann-
ing, design, and construction of these facilities, work
should be initiated on each of thf! options now.
In order to succ.essfully implement these options on
a commercial scale, many technical, economic, and in-
stitutional problems must be o~'ercome. The degree of
success achieved will be dependent on the ability of
energy producers, energy use,,>, and government to
mutUally solve these problems.
INTRODUCTION
This presentation is based on the premise that our
nation's environmental, energy, and economic consider-
ations are increasingly interdependent. A prime example
is the impact that the trends in electrical generation have
on the production and managemmt of waste heat.
Electrical energy requirements are projected to be
the predominant energy growth factor over the next 30
years. Whereas the total U.S. energy demand doubled in
22 years (between 1950 and 19i'2 . ref. 1 L the historical
trend of electrical production 01/er the past 50 years has
been a doubling every 10 years (about a 7 percent
annual increase). These trends, predicted to continue,
indicate the increasing role of el€'ctricity relative to other
energy forms in meeting our enel"gy demands.
I n terms of national energ'{ conversion efficiency,
growing demand for electricity, plus the anticipated con-
tinuing increases in transportation uses (as much as 3%
*A. G. Christianson is Director of the Energy Systems En-
vironmental Control Division, Industrial Environmental Research
Laboratory, U. S. EPA, Cincinnati,. Ohio. D. J. Cannon is an
engineer with the Criteria and AsseSI;ment Branch, Corvallis En-
vironmental Research Laboratory, U.S. EPA, Corvallis, Oregon.
percent per year) will result in an increase in overall
conversion losses from about 50 percent in 1970 to
almost 55 percent in 1990 (ref. 2). If such projections
hold true, the impact on resource requirements and
waste generation are rather startling. A national decrease
in energy conversion efficiency from 50 percent to 45
. percent results in an 11 percent increase in fuel input
requirements on a unit Btu basis, i.e. not considering the
growth in energy demand. The same efficiency decrease
results in a 22 percent increase in waste heat production,
again on a unit Btu basis.
Looking more closely at electric power generation,
in 1973, it accounted for a resource requirement of 19.8
quads (19.8 x 10' S Btu's) out of a total national energy
input demand of 75.5 quads-i.e., aDout 26 percent of
the total (ref. 4). The waste heat losses in conversion
amounted to 13.5 quads, which is about 18 percent of
our total national energy requirement. Waste heat is also
rejected from nonelectrical sources, but the amount
while exceedingly high is not easily quantified.
For a final comparison, it is interesting to consider
both growth rate and magnitude of waste heat produc-
tion. With electrical demand doubling every 10 years,
the predicted heat dissipation from central power
sources in the year 2000 is about 72 quads (ref. 3). This
quantity of waste heat is equal to the total U. S. energy
demand of 1970.
Rejected heat is considered a waste product in terms
of our current methods of operation. At an equivalent of
more than one-sixth of the Nation's total annual fuel
requirement, it must be considered for utilization or reo
duction through innovative planning and management
by both electric utilities and other industries. A new
emphasis must be given to the possibilities for improvin!1
conversion efficiencies and waste heat utilization-for
the sake of energy, environmental, and economic re-
sources.
In terms of energy impact, ERDA (ref. 3) indenti-
fies the potential savings in energy in the year 2000 for
these relevant categories as follows:
I ndustrial Energy Efficiency (Highest Priority,
Near-Term Efficiency [Conversation] Tech-
nology)-8.0 quads.
Waste Heat Utilization (Important, Midterm Tech-
nology)-4.9 quads.
These objectives are significant in terms of reduced
environmental impact because of reduced emission of
heat and other pollutants and also because of reduced
economic impact of pollution control.
191
-------
The Environmental Protection Agency (EPA) has
been involved in waste heat reduction and control acti-
vities primaril'l because of its mandates for environ-
mental impact mitigation. However, the economic
impact of pollution control technology has become an
increasingly prominent consideration, i.e., as the degree
of pollution control is increased, the incremental
benefit/cost ratio becomes more important and requires
more exact dnfinition. In other words, we are at the
state where "i ittle things mean a lot" with respect to
degree of control and related economic feasibility.
There are .1umerous options available for waste heat
utilization/redl'ction. We have focused on three options
which we feel t.ave good potential for future application.
These are (1) utilization of byproduct waste heat dis-
charged from conventional industrial and utility plants,
(2) generation of byproduct electricity in industrial
plants, and (3: development of integrated energy pro-
duction/use complexes which utilize energy more effici-
ently. The first option deals with the utilization of waste
heat after it has been discharged from a process or
facility, while the second and third options involve
optimizing. the design and management of the process
itself to reduce the waste heat discharge.
In order lJ capitalize on these waste heat utiliza-
tion/reduction :)Ptions, we see the need for emphasis in
research and dl!velopment in all three areas. Because of
the long lead ti'TIes and the long lifetime of the systems
proposed, it is 'mportant that planning be initiated now
in all three area!:.
INCENTIVES
111 the pas':, low fuel prices, plentiful supplies of
natur;J1 gas and oil, and the absence of pollutant control
standnrds gave limited incentives for the utilization of
waste heat, gereration of byproduct electricity by in-
. dustry, or const ruction of multipurpose central generat-
ing stations. Unjer these preenergy-crisis conditions, the
economic and ,)perational ben.efits of total energy or
waste heat utilization systems were marginal. However,
in the past three years the resource, economic, and en-
vironmental cOllstraints on the conversion and utiliza-
tion of energy have changed drastically. The costs of
natural gas, oil, and coal have risen significantly. Natural
gas is in short supply and many large utility and in-
dustrial users am being forced to convert to oil or coal.
The availability of oil is uncertain due to reduced pro-
duction in the United States and the tenuous political
situation in the Middle East. Environmental limitations
on SOx has necessitated the installation of scrubbers or
the burning of low-sulfur coal. Limits on thermal dis-
charges have nfcessitated the construction of closed
cycle cooling systems or other control measures. Finally,
the costs of nl!w construction as well as the cost of
capital have risen considerably, making the raising of
capital more and more difficult. Predictions point to a
continuation of these trends in the future, forcing the
cost of energy still higher (ref. 5).
The rates of electrical energy to total energy used in
this country have steadily increased and are expected to
continue this trend in the future. It is estimated that by
the year 2000, 41 percent of our energy needs will be
supplied by electricity (ref, 5). Efficiencies for direct
fuel conversion range from 70 percent to 85 percent for
most industrial facilities. However. the efficiency for
single-purpose electric generating stations ranges from 32
percent for light water nuclear plants to 40 percent for
HTGR and fossil fuel-fired stations, or about half that of
industrial energy conversion efficiency. This difference
in operating efficiency means that approximately double
the amount of fuel must be consumed by these central
electric stations (relative to industrial energy use) to de-
liver a similar amount of energy to the user. For this
reason, the pollutant emissions for central sta ions will
be double that of industrial combustion emissions,
assuming equivalent pollution control systems. The
differences in waste heat generated are equally dramatic.
In the future, the problems of fuel availability, pollutant
emissions, and waste heat discharges will be compound-
ed by the fact that the electrical energy to total energy
use ratios are increasing.
These current and projected constraints on energy
utilization have resulted in a significant increase in the
incentives for waste heat utilization/reduction concepts.
Many of the environmental, economic, and energy con-
servation incentives for waste heat reduction are inter-
related. For example, in reducing waste heat discharges
through more efficient production of electricity, fuel
will be conserved, pollutant emissions will be reduced,
and overall costs will likely be reduced.
Byproduct Waste Heat Utilization.
The economic incentives for utilizing heated cooling
water in agricultural, commercial, and industrial facilities
are currently improving, and are projected to continue
improving as the cost of fuel rises. In these applications,
heated water (or stack gas) can provide much of the
energy formerly supplied by natural gas or oil, thus con-
serving fuel, reducing the waste heat discharged in the
cooling water, and reducing overall costs. There are
several aquacultural and agricultural projects of demon-
stration and commercial scale now underway. The
success of these ventures will provide economic and
operational information needed to support further
development of these applications.
192
-------
Byproduct Electrical Generation bl' Industry.
Currently, good economic inGentives exist for pro-
duction of electricity by industrv. as a byproduct of
process steam generation. A recent study (ref. 6) implied
that in plant production of electricity by industry would
yield a 50 percent savings in fuel, a 50 percent reduction
in air emissions, and an 80 perCE!nt reduction in waste
heat generated (compared to eleGtrical generation in a
conventional power plant). Byproduct electrical produc-
tion will also yield reduced overall costs and an attrac-
tive return on investment for the industry. On a national
scale, byproduct electrical generation would reduce the
total capital needed by utilities for electrical generation.
Integrated Energy Production/Use Complexes.
These dual purpose or integrated facilities offer re-
ductions in fuel usage and in pollutant emissions due to
their high energy utilization factors. Studies have indi-
cated that a 10-15 percent savings in energy could be
achieved in an integrated energy facility. Savings in the
cost of delivered electrical power have been estimated at
six percent or more (ref. 6). Relatively few of these
integrated facilities have been built due to the problems
in matching production facilities to users, gathering the
necessary capital, and overcoming, institutional and other
constraints. However, the increa.ing cost of fuel is im-
proving the incentives for deVE!lopment of integrated
facilities.
OPTIONS FOR WASTE HEAT
UTI LlZA TION/REDUCTION
The discussion of options for waste heat utiliza-
tion/reduction is focused on three options which we feel
show good promise for development on a commercial
scale.
The first option, involving lIse of waste heat after it
leaves the conversion facil ity, has the greatest potential
for nearterm energy savings. Currently, most waste heat
applications are based on this approach. The waste heat
addressed here is the conventional reject heat from an
industrial process (e.g., electric power generation) which
is an end product or byproduct (if the process. Typically,
it is low-quality heat-usually between 10° F and 40° F
above incoming water temperature-such as that avail-
able in condenser cooling water from steam electric gen-
erating stations.
A number of direct-use waste heat applications have
been investigated to varying degrees-the more promising
methods receiving effort toward demonstrating their full
scale feasibility. At preenergy crisis fuel prices, the most
economically ,attractive waste heat applications have
been in agriculture and aquaculture. These efforts should
be continued and other uses evaluated more thoroughly
in light of current energy prices and potential savings.
Agricultural demonstrations of waste heat use have
shown promise in areas of irrigation, frost protection,
undersoil heating, and greenhouse heating and climate
control. Greenhouse applications appear most econo-
mical, based on production of high cash-value crops such
as flowers or vegetables. Refined systems are currently
being demonstrated which should result in commerical
involvement in 3 to 4 years.
EPA supported this application of waste heat in the
agricultural project it sponsored near Eugene, Oregon,
from 1968 to 1973. The project, jointly funded by the
Eugene Water and Electric Board, utilized cooling water
from a pulp and paper mill to provide spring frost pro-
tection, irrigation, crop cooling in the summer, and soil
heating. The project was most successful in providing
spring frost protection and in increasing crop yields in a
greenhouse which was heated by underground pipes.
A second EPA funded demonstration project is now
underway at the Sherburne County power plant in
Becker, Minnesota. The project will utilize heated water
from the condenser cooling loop of the two-unit,
1,400-MWe plant to provide soil heating, and air heating
for a half-acre greenhouse. Heating and cooling a half-
acre greenhouse in Minnesota for a year normally con-
sumes 25,000 gallons of oil or 3.5 million cubic feet of
natural gas. If the half-acre experiment is an economic
success, a commercially developed 100-acre greenhouse
complex could be in operation by 1985, with savings of
5 million gallons of oil or 700 million cubic feet of gas.
The success of this project will be gauged by crop pro-
duction results and. the economics of operation.
The Tennessee Valley Authority (TV A) has recently
received funds from EPA to use waste heat to stimulate
the growth of algae and amur fish in a project to recycle
nutrients from livestock operations. The project will use
livestock wastes to grow algae which will subsequently
be used to feed amur fish. The amur will then be harvest.
ed for livestock food.
TV A is also planning and conducting demonstra-
tional tests of open field soil heating, greenhouse heat-
ing, and catfish production using waste heat from con-
denser cooling water.
Numerous other systems have been proposed or are
under development for utilizing low-level waste heat.
Currently those projects that involve stimulation of
biological growth show the most promise for future
development. These include, aquaculture, mariculture,
algae production for animal food, and biological waste
treatment processes. Additional work, however, is need.
ed to identify species of plants 'and animals which re-
spond most favorably to waste heat stimulation, and to
193
-------
adequately control the chemical and biological wastes
from these act ivities. Since they have potential for high-
ly efficient protein production, these processes can also
be expected to supplement conventional, yet energy in-
tensive, agricultural practices.
The seco,d option, involving the generation of
byproduct elf ctricity in industrial plants, has good
potential for midterm reduction of waste heat. Nearly
20 percent of the nation's primary fuels are used to
produce indus':rial steam. Only 30 percent of this steam
is used to generate electrical power before using the
steam for process heating. Most industrial facilities now
generate their :>wn process steam with natural gas-or oil-
fired package :>oiler and purchase electricity from utili-
ties. Package boilers have been employed rather than
field-erected b:>jlers because of their low capital cost,
relative simpli( ity of operation, and the availability of
low-Gost natural gas and oil. Package boilers usually
operate at 10\0\' steam pressure (below 400 psia) with
operating efficie!1cies of about 75 percent. Field-erected
boilers, on the other hand, are relatively high in capital
. cost, and more complex to operate, but have an operat-
ing efficiency (If 88 percent and twice the useful life of
package boilers
Approximc tely 43 percent of the industrial steam
load is produce:! in amounts of 400,000 Ib/hr or more at
single locations. At facilities this size and larger, it would
be advantageoLs to install field-erected boilers and pro-
duce byproduct electricity inplant. For example, coal-
fired, a field-en!cted boiler producing steam at 900 psia
and 8250 F, a,d a turbine back pressure of 150 psia
would have an effective conversion efficiency of 4,500
to 5,000 Btu/kwh as compared with a rate of 9,000 to
10,000 Btu/kw.l for an efficient central power station
(ref. 6). The elEctrical power, therefore, could be gener-
ated with half the fuel required by a utility plant, not to
mention a lart e reduction in electrical transmission
losses.
The superior efficiency in byproduct electrical gen-
eration is due to the fact that the turbine exhaust (steam
at 150 psia) can be utilized for process heating, whE;!reas
conventional pO.Ner turbine exhaust heat is "wasted" to
the environment in the cooling water or the atmosphere.
In addition to the greater efficiency in electrical gener-
ation, the utiliution of a field-erected boiler rather than
a package boile' will result in about a 13 percent in-
crease in boiler efficiency. Add to this the longer
equipment life, and the field-erected boiler with by-
product electricc.! generation yields significant economic
benefits over ste)m production with package boilers and
purchase of electrical power.
Potentially, more than 33,000 MWe of additional
power could be generated by industrial byproduct power
units by 1985, resulting in an equivalent savings of
680,000 barrels per day of oil. It is estimated that this
could be achieved with a return on investment of 20
percent or more per year before taxes (ref. 6).
This application has significant potential for re-
ducing environmental impact of thermal and air pollut-
ants, conserving fuel, reducing the overall capital needed
for electrical generation, and reducing the cost of pro-
ducing power and process steam.
The disadvantages of this appl ication are that in-
dustry will be required to generate the new capital for
the new facilities and will need to upgrade the skills of
the boiler plant personnel. However, as the cost of
energy continues to rise, it is likely that industry will
find that the incentives for generation of byproduct
power will outweigh the disadvantages.
The third option, .often referred to as integrated
production/use facilities or dual purpose energy conver-
sion facilities, has potential for long-term waste heat
utilization/reduction. In this application both the energy
production facilities and the user facilities are designed
for operational compatability, in terms of capacity, load
characteristics, and equipment lifetime. These facilities
provide opportunities for increasing the efficiency of
energy utilization by 10 or 15 percent through optimiza-
tion of alternative energy blends. These facilities, how-
ever, must be large in scale (the equivalent of 1,000,000
Ib of steam/hr) to interest utilities in a joint venture.
Development of integrated energy facilities has been
limited to a few applications because constraints in-
volving compatibility of production and user systems,
the financial risk involved, lack of necessary capital, and
inappropriate long-term planning. Hopefully, the econo-
mic incentives for integrated facilities in the future will
encourage industry, utilities, and government to solve
some of the problems which now limit development.
An example of the benefits to be gained from an
integrated facility are summarized below. An integrated
coal-fired central station designed to supply 660 MWe
and 2,000,000 Ib/hr of steam at 150 psia, with a 35
percent load factor, could yield an equivalent savings in
fuel of 2,000 barrels of oil per day (ref. 6). This savings
in fuel would result in a proportional reduction in air
emissions and a large reduction in the waste heat dis-
charged.
EPA is initiating a study to (1) examine projected
cost of fuel, materials, labor, etc., (2) formulate four
promising integrated energy complexes, and (3) assess
the economic, environmental and energy conservation
aspects of these facilities at the projected cost levels. The
project should provide some indication of the environ-
mental benefits and economic viability of these applica-
tions.
194
-------
OTHER CONSIDEFIATIONS
The technical feasibility and the economic viability
of any application to utilize waste heat or reduce the
amount generated will depend on many factors such as:
1. compatability of the supplier and user systems,
2. economic, fuel conservation, or environmental
gains,
3. institutional constraints,
4. availability of capital, and
5. geographical constraints.
It is essential that these and other factors be addressed
and resolved before a joint wastl! heat utilization/reduc-
tion project is to be successfully initiated.
Each application will have different factors such as
fuel availability, fuel cost, site 'characteristics, distance
from production to load centers, availability and compa- .
tability of producing and usin9 facilities capital, etc.,
which must be evaluated on a case-by-case basis. It is
expected that each application will have to be designed
with specific objectives within fixed technical, econo-
mic, and institutional constraints.
CONCLUSION
Research is needed to indmtify and demonstrate
feasibility, energy-saving potential, and economics of
potential waste heat utilization/reduction applications.
Some of the approaches described in this paper are well
developed and near commercial application; others are in
the infancy stage and need rigorous efforts to evaluate
the!r ultimate potential. The impact of waste heat use
and reduction on our nation's energy supply is difficult
to quantify based on current methodology. As with
many projects, the ultimate impact depends on the de-
gree of successful technology development. However,
the potential for significant waste heat utilization cannot
be denied because the waste heat already is available in
tremendous quantities. As noted previously, unless we
take the initiative to (1) better utilize waste heat and (2)
improve the efficiency of energy production and utiliza-
tion, the amount of waste heat discharged to the en-
vironment by the year 2000 will equal the total U.S.
energy demand in 1970. Waste heat utilization, by-
product electrical generation, and integrated energy pro-
duction/use complexes will help to reduce the near-term,
midterm, and long-term discharge of waste heat, respec-
tively. However, because of the state of development,
the complexity of the applications, and the long load
times required, it is necessary for industry, utilities, and
government to take action on all three options at this
time. The results should yield significant environmental,
economic, and fuel conservation benefits.
REFERENCES
1. T. W. Bendixen and G. L. Huffman, "Impact of
Environmental Control Technologies on the Energy
Crisis," U.S. EPA-NERC-Cincinnati, January 1974.
2. U. S. Congress, Joint Committee on Atomic Energy,
"Understanding the National Energy Dilemma,"
published by the center for Stragegic and Inter-
national Studies, Georgetown University, Washing-
ton, D. C., September 1975.
3. U. S. Energy Research and Development Adminis-
tration, "A National Plan for Energy Research, De-
velopment, and Demonstration," ERDA-48, Vol. 1
of 2, U.S. GPO, June 1975.
4. Radian Corporation, "A Western Regional Energy
Development Study," for CEQ and FEA, RC NO.
100-64, August 1975.
5. Council on Environmental Quality, "Energy and the
Environment-Electric Power," U.S. GPO, August
1973.
6. Dow Chemical Company, "Energy Industrial Center
Study," prepared for the National Science Founda-
tion, draft copy, April 1975.
195
-------
STAGED COMBUSTION IN COMBINED-CYCLE
SUPPLEMENTARY FIRED BOILERS*
D. R. Bartz, S. C. Hunter, and J. K. Arandt
Abstract
A scaled laboratory combustion experiment was
conducted to determine the effectiveness of staged com-
bustion in red,Jcing nitric oxide present in the combus-
tion air supplied to large diffusion flames. Significant
chemical reduction of the inlet nitric oxide suggests the
possibility 0' application to supplementary fired com-
bined-cycle sy~ tems as an alternative to water injection
as a means of controlling NO x to anticipated EPA emis-
sion limits.
INTRODUCTION
One of thE' more acceptable approaches to adding
capacity to exi~ting municipal and regional power net-
works is that based on the installation of combined-cycle
units. They have been favored by many utilities recently
because of the ;omparatively short delivery and installa-
tion time, favorable unit cost, and low firing rate. A
combined cycle is a combination of open Brayton cycle
gas turbines and closed rankine cycle steam boilers and
turbines. The components are arranged so that the
exhaust of the Has turbines passes through the steam
boiler and prol'ides a means of recovering part of the
waste heat nor mally exhausted by the gas turbines.
When this heat recovery is accomplished with a convec-
tive waste heat boiler only, the combination is desig-
nated as an un1 ired combined cycle. Normally the gas
turbine exhaust contains oxygen concentrations in the
15 percent to 1'~ percent range, which can be utilized as
a kind of vitiatl!d air supply with which to burn addi-
tional fuel to t~ereby generate more power than would
the unfired boiler. This arrangement is designated as a
supplementary f r~d combi,ned cycle.
Aside from all of the favorable attributes of gas
turbines and co nbined cycles, the gas turbine as built
today for power applications experiences difficulty with
pollutant emissi )ns. While at present there are no na-
tional new-source standards for gas turbines-only local
standards in pric rity I regions-such standards are under
consideration bV the EPA and have been circulateq for
"Based on wo'k performed for the Electric Power Research
Institute, Palo Alto, California, under Contract EPR 1-224,
tD. R. Bartz is General Manager, S. C. Hunter is Manager,
Technical Assessmwt, and J. K. Arand is Principal Engineer,
with KVB, Inc., Tustin, California.
industry comment. The standards under consideration
for nitrogen oxide emissions (calculated as N02 at air
dilution characterized by 15 percent oxygen) are 55
ppm for gas fuel and 75 ppm for liquid fuel. These corre-
spond to 0.2 Ib/MBtu and 0.3 Ib/MBtu of N02 respec-
tively, the current new-source standards for boilers.
These standards are sufficiently low that they cannot be
met with current gas turbine combustion technology for
liquid fuel burning, and the use of natural gas in most
regions is only a fond memory. As a consequence, most
installations designed to meet regulations as stringent as
these are forced to utilize water injection as an NOx
control technique. While this does work, it is costly,
inconvenient, and in the long range may cause mainte.
nance or unit life problems. Thus there is a motivation
to reduce the NOx emissions to the levels of the poten-
tial new-source regulations by an alternative approach.
This was the objective of the analytical and experimental
work performed for the Electric Power Research Insti.
tute (refs. 1,2) and reported in this paper.
One of the most effective techniques for the control
of emissions of oxides of nitrogen in boilers has been the
use of staged combustion. The principle involved is to
conduct the combustion of most of the fuel in an envi-
ronment in which there is insufficient air for the fuel to
be fully combusted; then to transfer some heat out of
the flame prior to adding the remaining air necessary to
fully complete the combustion of all the fuel. This
achieves two purposes favorable to limiting the forma-
tion of oxides of nitrogen. First, it limits the maximum
flame temperature reached in the combustion zone, and,
secondly, it limits the availability of oxygen during the
combustion of most of the fuel. Since the kinetics of the
thermal NO reactions are extremely sensitive to the
temperature in the flame zone, even a reduction of 100°
F or so has a large effect. The presence of an abundance
of oxygen atoms is known to have an important effect
on the rate and extent to which nitrogen-bearing radicals
contained in the liquid fuels are oxidized to nitric oxide,
as compared with the competing reactions which convert
these radicals to nitrogen molecules. In fully premixed
(gaseous or vaporized liquid) flames, these effects are
very strong, almost dominant, in limiting the formation
of nitric oxide. In the large diffusion flames, which are
characteristic of large boilers, much of the combustion
goes on at near stoichiometric conditions in a fairly thin
flame front. However, there is some premixing that goes
on due to turbulent mixing processes in the flame zone.
196
-------
Thus whih~ sta!led comhustion hus not h()tHl found to he
as etfuctivl) in these lar!le diffusio.n flames as it would be
in premixed flames, nonetheless the use of staged com-
bustion with liquid fuels has limited NO emissions to 50
percent to 70 percent of what they would have been
without staging.
Another effective NO control technique is that of
flue gas recirculation in which 10 percent to 30 percent
of the flue gas is. recirculated into the combustion air.
The principle involved is that by loading up the combus-
tion air with a significant concentration of inert species
such as CO2, N2, H20 from the exhaust gas, the energy
released from the combustion of a given quantity of fuel
with its required oxygen will not yield as high a maxi-
mum-flame temperature becaus(! of the fraction of the
energy soaked up by these inerts in reaching that flame
temperature.
The connection between these two NO control
principles and combined cycles is that the supple-
mentary firing of the boiler gives an opportunity to take
advantage of both with little effort or penalty. The
equivalent of flue gas recirculation is achieved in the
supplementary fired boiler by virtue of the fact that the
exhaust gas from the turbine, when used as the combus-
tion air for the boiler, is diluted with inert species from
the turbine combustion process. Operation at conditions
in which the oxygen concentration is 17 percent is
equ ivalent to about 30 percent flue gas recirculation.
Now, if in addition the turbine exhaust is split into two
streams before entering the boill~r-one stream fed to the
burner windbox (providing 70 percent to 80 percent of
the oxygen needed for the fuel), and the other 20 per-
cent to 30 percent of the oxygen (and a portion of the
inerts) fed to some overfire air ports (i.e., above the
primary combustion zone)-the combustion can be
effectively staged. These two NO control techniques
working together should be eXI::>ected to limit NO forma-
tion from the burning of the supplementary fuel in the
boiler to 30 percent to 60 percent of what it would have
been otherwise. While this is attractive, the limitation of
the boiler-produced NO would 110t be sufficient to offset
the excessive turbine-produced NO, so as to render a
combined system emission below the potential regula-
tion level on a pounds-per-mill ion-Btu basis-not unless
the initial fuel-rich zone of the staged combustion could
actu a II y chemically reduce some of the turbine-
generated NO by stripping e,ff oxygen (i.e., the NO
acting as a substitute oxidizer in the absence of suffi-
cient air). The possibility for doing this was suggested by
some premixed flame experiments conducted by Wendt
et al. (ref. 3).
Thus the objective of the program conducted was to
determine the most favorable applicable combined-
cycle-system opcratiny conditions and then to conduct
scaled combustion experiments to determine whether
the chemical reduction of the NO could be achieved to a
useful extent under practical diffusion flame combustion
conditions.
SYSTEM ANALYSIS-GOVERNING PARAMETERS
System analysis was conducted in order to examine
all the pertinent combined-cycle-system design-point
variables and their implications with respect to control
of NOx by staged combustion in the supplementary
fired boiler. The effects of interest were those of gas
turbine exhaust temperature, oxygen and NO content,
fraction of gas bypassed around the supplementary
burner, air fuel ratio of the supplemental burner, and
power generation split between steam and gas turbines.
A schematic representation of a combined-cycle
power plant with staged supplementary-fired boiler is
shown in figure 1. The plant is comprised of the gas
turbine, supplemental burner, steam generator (boiler).
and the steam turbine. The system is different from con-
ventional combined-cycle systems only by providing the
bypass duct through which a portion of the gas turbine
exhaust can be routed around the supplementary burner.
By means of this bypass, the burner can be operated fuel
rich, and the bypassed oxygen and accompanying inerts
can be added downstream to complete the combustion
and insure the normal amount of excess (unburned)
oxygen in the boiler exhaust stack. Key to effecting a
reduction in the boiler-generated NOx is a loss of heat
between the initial fuel-rich combustion and the subse-
quent addition of the staged oxygen downstream.
Although the system analysis conducted and de-
scribed in detail in the final report to EPRI on this work
(ref. 1) used a 60-Mw gas turbine for illustrative pur-
poses, the results can be interpreted on a relative bas,s
and applied to any size and combination of gas turbines
and steam cycles. The steam cycle data was obtained
from manufacturer's data for a typical200-Mw system
operating at 1,800 psi, 1,000° F steam temperature. The
steam turbine heat rate (exclusive of boiler efficiency)
was taken to be 8,170 Btu/kWhr for full-load supple-
mentary-fired conditions. For the unfired conditions
where gas turbine exit temperatures in the range of 800°
F to 1,000° F precluded 1,000° F steam, a higher heat
rate of 10,000 Btu/kWhr was adopted. At intermediate
loads between these limits, the heat rate was scaled
linearly. A stack temperature of 300° F for the boiler
exhaust was assumed constant .for all conditions.
The key parameter governing the supplementary
combustion is the burner equivalence ratio; i.e., the ratio
197
-------
BYPASS
GAS
FLOW
FUEL -
FLOW
AIR -
FLOW
STEAM
--
GENERATOR H
T
S
C
-
BURNER
SYST EM ~
I
- GAS TU
EXHAU
--;.. GAS
TURBINE- - POWE
~ OUTPU
GENERATOR
.~
STACK
GAS
EXHAUST
EAT
o
TEAM
YCLE
STEAM
TURBINE-
GENERATOR
POWER
OUTPUT
FUEL
FLOW
RBINE
ST GAS FLOW
R
T
Figure 1. Combined cycle power plant schematic.
of actual fuel-tn.air compared with that for stoichio-
metric combust'on, or for a given fuel flow rate, the
ratio of stoichicmetric air flow rate to actual air flow
rate. A gas turbine at full load normally operates at
about 300 perc.mt excess air, which corresponds to a
0.25 equivalence ratio. Boilers typically operate with 10
percent to 20 pE'rcent excess air or an equivalence ratio
of from about O..~ to 0.9.
If an existirg boiler operating at a normal equiva-
lence ratio of 0.9 were to have its air supply replaced
with an exactly equal total mass flow turbine exhaust,
the boiler burner equivalence ratio for the same fuel
flow would incmase to over 1.0, i.e., become fuel rich
due to the lowe- oxygen concentration in the turbine
exhaust flow, as shown in figure 2. In order to adjust the
boiler equivalence ratio back below 1.0 due to safety
and efficiency considerations, it would be necessary to
reduce the burner fuel flow. For such a system as this,
the overall stack (exhaust) equivalence ratio is uniquely
related to the combination of burner equivalence ratio
and percent of the turbine flow bypassed around the
burner. Figure 3 shows that, in the limit of complete
bypass of the supplemental burner, the stack equivalence
ratio is of course exactly that of the gas turbine. At the
other limit of no bypass, the stack equivalence ratio is
exactly the burner equivalence ratio. The stack equiva-
lence ratio is seen to vary linearly between these limits.
To limit the stack equivalence ratio below 1.0, more
198
-------
1.4 4>qt=.4
GAS TUftBINE AIRFLOW = ORIGINAL BOILER FLOW .3l
NO BYP#'!~SSt NO MAKE - UP
LaJ NO RECIRCULATION
z BOILER FUEL FLOW SAME
in 1.2 AFTER .ADDING GAS TURBINE .2T
a: 4>0 .4> b (NORMAL FIRING)
::>
.... I NORMAL
(!) . GAS
z TURBINE
Q
Q 1.0 0 RANGE
ct
a:
LaJ
....
u..
ct
0 .8
~
LaJ
0
Z
LaJ .6
...J
ct
>
::>
a
LaJ
a:
LaJ .4
z
a: ~
::> NORMAL
m BOI LER
.a RANGE
-&- .2
o
o
. .2
.4
.6
.8
1.0
4>bot INITIAL BURNER EQUIVALENCE RATIO BEFORE
ADDING GAS TURBINE
Figure 2. Change in burner equivalence ratio when adding a gas turbine.
199
-------
t4
01 L FUEL
. 02 gt = 15 %
" gt . .3
f . BYPASS GAS FLOW
1.2 GAS TURBINE EXHAUST FLOW
o
~ LMAXIMUM POSSIBLE FIRING
.8
a ~5%
IIJ
~
0 ~7%
0
- ,4
0
-e-
15Ofo
.2 NO SUPPLEMENTAL
FIRING
o
o
20
40
60
80
100
f, TURBINE EXHAUST BYPASS, PERCENT
Figure 3. Effect on gas turbine exhaust bypass and makeup air on
overall equivalence ratio.
200
-------
than 50 percent of the flow must be bypassed around
the burner to achieve a burner fwd-rich equivalence ratio
as high as 1.5.
Finally the payoff variable, that which measures the
effectiveness of NOx reduction, must be define~1. In this
investigation the so-called NO reduction factor was
defined as the ratio of the sum (If the NO generated by
the gas turbine plus that generated in the supplementary
combustion less that leaving the stack all divided by the
turbine-generated NO (figure 4). Thus if the NO con-
centration in the exhaust stack was reduced to that
which was generated in the boiler supplemental combus-
tion, the reduction factor was unity, and the turbine NO
could be considered to be 100 percent destroyed or
chemically reduced. In the experiments to be described,
the turbine exhaust gas was a synthetic mixture of O2,
CO2, H20 and NO heated to clppropriate turbine ex-
haust temperature. The NO of course was independently
controlled over the range from absolute zero to 100
ppm. By measuring the boiler-burner combustion-gener-
ated NO concentration when the turbine NO concentra-
tion was maintained at zero, the contribution of the NO
generated in the boiler could easily be determined. As
will be shown with the data, up to 50 percent of the
turbine NO was reduced by the staged combustion.
EXPERIMENTAL INVIESTIGATION
The laboratory test setup utilized for this investiga-
tion is shown schematically in figure 5. The simulated
gas turbine exhaust was achieved by adding carbon
dioxide, nitrogen, water vapor, and nitric oxide to the
combustion air supply and then heating the gas mixture.
to turbine exhaust temperature levels with external
ceramic heating elements. This supply of vitiated air was
NO
GAS l1JRBI NE
NO
BOILER
GENER'!~TED
NO
STACK
I
R f =
NOGAS TURBINE + NO 1301 LER - NO STACK
NO GAS TURBINE
Figure 4. NO reduction factor definition.
delivered to the 200,000 Btu/hr combustion tunnel,
where a portion was fed directly to the single burner and
the remainder bypassed around the primary combustion
zone, to be added to the combustion flow about 4 diam-
eters downstream. Approximately 50 percent of the
burner heat input was lost through the tunnel walls prior
to the injection of the bypass air. This loss of heat is
essential to the success of staged combustion as an NOx-
control technique. By varying the bypass air flow with a
fixed burner fuel input rate and fixed total air flow, the
equivalence ratio of the primary combustion zone could
be controlled. By varying the absolute air flow with a
fixed burner fuel flow rate, various ratios of gas turbine
power to steam turbine power could be simulated.
Several burner configurations were used to investi-
gate the importance of mixing by the supplemental
burner to the anticipated NOx reductions. During the
investigation, the oxygen level and temperature level of
the vitiated air were varied to simulate changes in the
turbine load. The range of oxygen concentrations tested
was from 17 percent to 21 percent, and the range of
temperatures from 8000 F to 1,0000 F. I n addition, the
NO content of the combustion air was varied over the
range from zero, to establish the supplemental burner-
only generated NOx to 100 ppm, a typical level for
uncontrolled turbine emissions. The experiments, con-
ducted with both natural gas and light turbine oil fuels,
covered a range of fuel flows from 60,000 Btu/hr to
200,000 Btu/hr. The latter rate corresponded to a
volumetric heat release rate of 30,000 Btu/hr-ft3 in the
14-in. diameter, 75-in.-long tunnel.
In addition to normal fuel and air flow and tempera-
ture measurements made with standard techniques, the
gas composition measurements were the key to the suc-
cess of the program. Flue gas samples were withdrawn
from three equally spaced uncooled probes, while the
vitiated air supply was sampled from a single point. The
gas samples were passed through a filter and then refrit'-
erated to condense out water vapor before analysis. {,
TECO chemiluminescent instrument was used to meas-
ure NO, a Beckman electrolytic type sensor to measure
oxygen content, a Beckman NDIR instrument for CO.
Also, a Beckman hydrocarbon analyzer was utilized.
Samples were also drawn occasionally during the pro-
gram to permit mass spec analysis for trace species such
as HCN, NH3, etc.
TEST BURNERS
The burner that might be selected for an actual
combined-cycle supplementary boiler could range in
design and mixing quality all the way from a poorly
201
-------
GAS TURBINE PRODUCTS
STEAM
NO
C02
N2
ROTAMETERS
MIXING MANIFOLD
CERAMIC HEATING ELEMENTS
~
o
~
TOTAL AIR FLOW -
ROTAMETER
GAS SAMPLE
NO a 02
VALVE
~
GAS SAMPLE
NOt CO 02 a UHC
BYPASS AIR
MEASURING PROBE
~
~ I TEMP. a
PRESSURE
PROBES
VALVE ~
ROTAMETER
FOR METHANE
FUEL
SUPPLY
Figure 5. Combined-cycle test setup schematic.
TUNNEL
f
TUNN EL
TEMP. PROBES
-------
mixed simple flame holder type to a relatively well-
mixed high swirl burner. Since the burner design is
known to have a pronounced effl!ct on NO prqduction,
it was desirable to investigate a wide spectrum of burner
designs to determine the influenc~ of burner mixing on
NO reduction in fuel-rich combustion. One measure of
mixing used to characterize aerodynamic turbulence is a
mixedness parameter which is proportional to the square
root of the ratio of the pressure drop across the burner
to the dynamic head of the combustion gas flow in the
flame zone. Burners having a hi!Jh value of this ratio-
i.e., high pressure drop and low dynamic head-would
have high aerodynamic turbulence and hence good
mixing.
The can burner shown in figlJre 6 is similar to a gas
turbine combustor, which is normally characterized as
being nearly a totally stirred mactor. Intensive aero-
dynamic mixing is achieved with the impinging jets of
air. The flame stabilization is achieved by the intense gas
recirculation zones created by the jet flows. This type of
burner typically has a high pressure drop and a low
dynamic head, for a mixing parameter value of 351 in
arbitrary units.
The nozzle mix burner shown in figure 7 has swirl-
ing air flow similar to that of typical utility boiler reg-
ister burners. The flame stabilization is achieved by both
the swirling air flow and the high-temperature zone
created by the refractory throat. The mixing parameter
is only about one-fifth that of the can burner.
The grid burner, shown in figure 8, is the poorest
mixed of the three, with its very low pressure loss and
simple fuel injector. This type is 'Jsed in some combined
cycles where the supplementary firing occurs in the
boiler air supply duct rather than in the boiler itself. The
mixing parameter for the burner tested was about one-
tenth that of the can burner.
A problem encountered during the test program was
that of flame instability and blowout at many of the test
conditions at 17 percent O2 or below. The problem was
most severe away from stoichiom~tric air fuel ratios. It is
not surprising that flame stability decreased with greater
vitiation of the combustion air since this is well docu-
mented in the literature. Of the burners tested, the noz-
zle mix burner exhibited the highest degree of flame
stability at low oxygen concentrations. Because of this,
and because this is the most conl/entional of the burners
tested, results from only this burner will be presented.
TEST RESULTS
Typical of the results obtained during the program
are the data presented in figure n in which the NO emis-
sions rate expressed as Ib/hr of N02 is plotted versus
burner equivalence ratio. The lower curve with zero NO
in the inlet air characterizes the burner-generated NO as
the burner equivalence ratio is increased up to about 35
percent fuel rich. As is normally expected in burners
operating with such staged combustion, the NO emis-
sions dropped to about half the value at near stoichio-
metric or unstaged conditions. With 100 ppm of NO
added to the inlet air supply, the stack emission level
would be expected to be about 0.014 Ib/hr higher, i.e.,
the uppermost curve. The actual result, the middle
curve, was significantly lower, thus indicating that the
fuel-rich staged combustion was actually chemically
reducing a portion of the inlet NO to N2 by stripping off
oxygen. The more fuel rich the burner was operated
(higher equivalence ratio). the greater the fraction re-
duced, reaching about 57 percent at a burner equiva-
lence ratio of 1.33. Literally hundreds of such curves
were obtained (ref. 2) for a wide range of the many
variables of interest: fuel, burner type, inlet oxygen
level, inlet temperature, burner heat release rate, etc.
A number of such reduction factor data are corre-
lated against burner equivalence ratio in figure 10, and in
other such figures of the final report (ref. 2). Each of the
correlation curves had a similar shape. Maximum reduc-
tion factors somewhat higher than shown were achieved,
but generally occurred when there was unacceptably
high CO concentration. As can be seen, the results were
not sensitive to overall equivalence ratio or to inlet oxy-
gen concentration when oil was the fuel.Curiously, it
was found that even without staged combustion, some
10 percent of the turbine NO was reduced by passing
through the flame zone.
APPLICATION OF R ESUL TS
Results such as these were used to predict just how
effective would be the use of staged combustion as an
NOx control technique in supplementary-fired com.
bined-cycle boilers. This was determined as a function of
the power generation split between the gas turbine and
the steam turbine. To do this it was necessary to com.
pute the net system heat rate and stack equivalence ratio
as a function of this power split (as shown in figure 11).
Zero on the power split scale represents all gas turbine
power, arid 1.0 represents all steam turbine. As can be
seen, up to a 35-percent power split can be achieved
with 'an unfired boiler, and the addition of that unfired
boiler improves the heat rate markedly. I n fact, some
increase of system heat rate occurs as the power split is
taken further, requiring the supplemental firing. The
benefit from the supplemental firing from a system NO
emissions standpoint is shown in figure 12 which was
derived from the experimental data obtained in this
203
-------
IINLET
,AIR 0
PARTITION .. I
WALL .""
-0 ~p. 30" H2O @ 10000F
Sa MAX AIR FLOW
TSD = 351
VREF = 49.3 FPS
PLEN U M
TUNNEL
FLAME STABiliZATION
RECIRCULATION ZONES
~
II.)
o
~
FUEL OIL ORo
GAS EOUS FUEL
FUEL SUPPLY TU BE
CAN BURNER
BODY
tiN LET
AIR
Figure 6. Can burner.
-------
PRIMARY AIR FOR OIL FIRING
GASEOUS FUEL FOR GAS FIRING
TUNNEL
~
SECONDARY AIR
I
t S~i:LER
REFRACTORY
~n
I>J
o
U1
AIR SWIRLER
SWIRLING FLOW
ATOMIZING AIR
~P=37"H20 @IOOooF
(GAS FUEL)
TSD = 69 (GAS FUEL)
V REF = 62.6 FPS
a AIR FLOW
Figure 7. Nozzle mix burner.
-------
--- ...---.---~--------------....
AP=2" H20@1000oF a MAX FLOW
T SD = 4 I
VREF = 100 FPS
TUNN EL
FUEL
SPRAYBAR
DISTANCE FROM SPRAYBAR
rTO FLAME HOLDER,
3u GAS
15"01L
r
I -----------
FLAME HOLDE R
10 D
'"
o
0>
. COMB~
A~
FLAME ZONE
--
I ---------
----
RECIRCULATION ZONE
FUEL INJECTED PERPENDICULAR
10 AIR 4 LOCATIONS 2 ON EACH
SIDE OF SPRAYBAR
Figure 8. Grid burner
-------
~
o
......
a:
:I:
""-
CJ)
m
...J
N
o
Z
en
ct
o
Z
NOZZLE MIX BURNER
HEAT INPUT 100000 BTU/HR
FUEL ** 2 DIESEL OIL
INLET AIR 21 % OXYGEN
DATA POINTS 169 -171
OVERALL EQUIVALENCE RATIO "'.85
"'11"\ ern. 'f"TI ".... C'J\f"T"D AT ++ ....,,'" J:''' D I"" DDU I J:'
.030 ~~ ="~96'" ""'''' ,~"'. v.. ~, {+' / ;E~uCTio~' D'I'O'" ~OT"O~UR
R I = .0299- .0269 = 226 '.:"
f .0299- .0167 . ~,
~ '+.....
INLET NO 100 PPM /" -._+
.020
.010
INLET NO 0 PPM~
o
.4
.6
.8
1.0
1.2
BURNER EQUIVALENCE RATIO ~ B
NO REDUCTION FACTOR AT
~B = 1.33
R I = .0202 - .0126 = .572
f .0202 -.007
1.4
1.6
Figure 9. Reduction of inlet NO at high overall equivalence ratio for
21 percent O2 inlet air concentration and #2 oil fuel.
-------
1.0
NOZZLE MIX BURNER
~ 2 DIESEL OIL FUEL
100 PPM INLET NO
ALL DATA POINTS
.8 17 % OXYGEN 21 % OXYGEN
. cP 0 = .4 6 CPo = .4
-.... 0 CPo = .6 0 4>0 = .6
(Ie: + 4> 0 = .85 . 4>0= .85
QC: .6 CO < 100 PPM
()
~-
()
c:( .
L\.
-,.
~.
() .4
_..
~-
(,)
:)
C~
Ld 0
Q':
() .2
;2':
o
.6
.8 1.0 1.2
BURNER EQUIVALENCE RATIO CPa
1.4
Figure 10. NO reduction factor for #2 oil fuel at all test conditions
with stack CO less than 100 ppm.
program, some of which was presented on previous
figures. As is seen, the gas turbine emission rate corre-
sponding to 100 ppm is about 0.38 Ib/MBtu, somewhat
above the anticipated regulation level of 0.3. Some gas
turbines withou~: water injection emit considerably high-
er concentrations. As the power split is taken beyond 35
percent and supplementary firing is used, the system
emission level d 'ops rapidly both with the staged com-
bustion and ever with normal unstaged combustion. The
latter drop is a (onsequence of boiler-generate emissions
being lower for the case considered and the modest (10
percent or so) chemical reduction with normal firing.
With staged combustion, the drop in system emissions is
significantly greater, dropping to about one third of the
anticipated regulation level at a power split ratio of 0.75,
which corresponds to one-fourth of the system power
from the gas turbines and three fourths from the steam
turbine.
Such a power split as this is unusual for current
new-unit, combined-cycle design practice. The results
208
-------
taJ
~
~ ;- 11000
~ :I:
~ = 10000
2 0::
:I: 9000
&IJ I
~ ~
en ~
>- ...... 8000
en ::>
~ ~
&IJ m 7000
z
1.0
o
ti
0::
&IJ
c.>
Z
&IJ
~
::>
S
~
o
~
en
..
.0
-&-
GAS TURBINE HEAT RATE = 11000 (LHV)
STEAM TURB:-GEN. HEATRATE =8170 (HHV)
Pr >.7
STEAM TURB.-GEN. HEATRATE = 10000 (HHV)
UN FIRED
OIL
,8
17°/q
800°F
02 = 15°1c
gt °
T gt = 10000F
.6
.4
UNFI RED
FI RED, CONSTANT STACK
TEMPERATURE = 300° F
.2
o
o
.20 .40 .60 .80
Pr, RATED POWER SPLIT, P gt
Pst+ Pgt
1.00
Figure 11. Combined-cycle net heat rate and overall equivalence ratio
ciS a function of power split at design rated load. .
209
-------
OIL FUEL °2gt = 15 %
NO MAKE -UP AIR 19t =IOOooF
::> 5 NOZZLE MIX BURNER Tstk = 300°F
...... . I N
m 0
U) at
o Cgt .100 PPM@ 15 % 02 ~
~ @J
o .4 ~.
z 300
m B= .93 (NORMALLY Q.
-J ..
.. FIRED) en
fJ) z
.z 0
o .3 -
- en
en en
en ~
- 200
~ iLl
iLl X
X 0
o .2 z
z iLl
W -.J
o
-.J >-
o 0
>- 100 '
u C
I .1 iLl
c Z
w B I: 1.3 (STAGE D
z m
- COMBUSTION) ~
m
~ 0
o 0
u 0 0
0 .20 .40 .60 .80 1.00
POWER SPLIT, Pst
Pst + Pgt
Figure 12. Combined cycle, LB N02 /1 06 Btu, for oil fuel operation.
suggest howevE r a possible way of adding capacity to an
existing boiler by repowering with a gas turbine replac-
ing the FD fan and air preheater, thereby increasing total
system power significantly at what are likely to be
acceptably low NO emissions levels (ref. 4).
Other resul1s are presented in the EPRI report on
this work that shows how the combined-cycle heat rate
and t:missionslilry with percent of rated power output
for several opti,)nal throttling modes.
CONCLUSIONS
The use of staged combustion has been shown
through a series of laboratory scale tests to be a prom-
ising method for NOx emissions control in supple-
mentary-fired combined-cycle systems over a wide range
of system operational conditions. Specifically the results
suggest that the use of staged combustion in the supple-
mentary.fired boiler may provide a viable means for
210
-------
compensating for gas turbine NOx emissions in excess of
anticipated EPA standards without having to resort to
water injection.. It remains for full-scale tests to be con-
ducted to confirm these results, although there is no
reason to believe that it will not be as effective at larger
scales. The NOx reduction effectiveness was found to be
. the highest at the highest power splits (more steam tur-
bine power than gas turbine power) where the turbine
exhaust is fired to the minimum practical oxygen con-
centration. Reduction of up to ~;5 percent of the tur-
bine-generated NOx was found to be achievable under
practical operating conditions with diesel oil fuel. The
effectiveness of the staged combustion was found to be
dependent on the fuel type, burner, and firing rate much
more than on any of the gas turbine exhaust conditions
such as temperature or oxygen concentration.
REFERENCES
1. S. C. Hunter, "Reduction of NOx Through Staged
Combustion in Combined Cycle Supplemental
Boilers, Vol. I - System Optimization Analyses,"
KVB Report 6900-183, January 1975.
2. J. K. Arand, "Reduction of NOx Through Staged
Combustion in Combined Cycle Supplemental
Boilers, Vol. II - Experimental Program," KVB
Report 6900-183, January 1975.
3. J. 0. L. Wendt, C. V. Sternling, and M. A. Matovich,
. "Reduction of Sulfur Trioxide and Nitrogen Oxides
by Secondary Fuel Injection," Fourteenth Sympo-
sium (International) on Combustion, 1973, p. 897.
4. L. 0. Tomlinson, "Steam Plant Efficiency I mprove-
ment by Combined Cycle Repowering," presented
at American Public Power Association Engineering
and Operation Workshop, Schenectady, N. Y., 1974.
211
-------
ELECTRIC POWER TRANSMISSION
MODES, MEANS, AND STYLE
Robert W. Flugum*
Abstract
Transmission of electric power faces a challenge in
the future as the Nation shifts to a more electric econo-
my to conserve scarce fossil fuels. The modes of trans-
mission must "e expanded in power transfer capability,
and environmEntal aspects and efficiency must be con-
sidered as eSSt'ntial elements in design and installation
practices.
INTRODUCTION
Electric power has become, over the last seven de-
cades, the mos t important single technology maintaining
our way of life. It is destined to become even more
dominant as it! inherent flexibility permits for electrici-
ty production use of basic energy sources that do not
contribute to the undesirable attrition of scarce fossil
fuels. Critical a5 electric power is to our Nation's needs,
many practice!. and characteristics associated with its
generation, transmission, and distribution have come
under fire as not being responsive to today's concerns
about the environment and man's well-being. Further, as
the demand fer electric energy inevitably will exceed
industry's supp y, the need for conservation in its usage
must be addres~ed.
GENERAL
If the premise is accepted that our society, as we
know it today, could not be maintained without elec-
tricity, then it follows logically that we must generate,
transmit, and distribute that electricity. Without being
totally defensivl!, however, we find that our populace, at
'Ieast in part, r,lises strong objections to any intrusion
from the neces ;ary construction of electric power in-
stallations-in particular, generating plants and trans-
mission lines. There are undoubtedly historical, political,
eco'nomical, anc social rationales for this. negative atti-
tude, most of vlhich could be effectively debated. The
positive approach, however, is to address the questions
directly, using research and development programs
where necessar), and education where the debate is
philosophical rather than technical.
*Assistant Director, Transmission, Division of Electric
Energy Systems, Energy Research and Development Administra-
tion, Washington, D.C.
In particular, environmental concerns about high-
voltage (HV) transmission had been limited, prior to
1970, to aesthetics, landscape disturbances, and plant
life effects, with only radio and television interferences
as direct electrical considerations. More recently, it has
been suggested that possible biological consequences
from the presence of electric fields under high-voltage
transmission lines may appear. If true, this could pose a
hazard to the public and to the electrical HV trans-
mission worker. The public is now also aware of low-
frequency noise generated by HV transmission lines par-
ticularly during fog or drizzle. Such psychoaccoustic
irritations have been cited in intervenor actions against
new construction proposals for 765-kV transmission
lines.
Attention is also being focused on the impact of
electric power transmission on our national goals of
energy conservation. I n the power system loss structure,
only 10 to 15 percent is found in the transmission
system from generator terminals to ultimate user. How-
ever, even if only direct losses are considered, because of
the vast amounts of power to be transferred, a 25 per-
cent improvement in transmission efficiency would have
a significant effect in direct energy conservation. If, in
addition, conservation is expanded in definition to in-
clude the energy saved by reduced use of materials and
improved manufacturing techniques, the impact of
successful research in transmission equipment and
methods can yield further important savings.
ELECTRIC POWER TRANSMISSION MODES
In broad categories, electric power transmission can
be described as either overhead or underground-with
"on" or "over" ground considered part of the latter.
Which mode is used depends on a number of factors,
among them capacity requirements, system voltage;
distance, siting, terminal location, economics, and en-
vironmental constraints. There are many situations
where there is no choice. For overhead, these would be
applications requiring thousands of megawatts to be
transferred over hundreds of miles. On the other hand,
for overwater power transfer from 2 to 20 miles, under-
ground types of cable are usually the only practical
transmission scheme.
It is generally assumed that underground trans-
mission constitutes a large part of the total of electric
power transmission lines in the United States. Two of
212
-------
the above factors keep this from being true: economics
and power capacity requirements vis a vis overhead
transmission. Table 1 illustrates the disparity in installed
miles and capacity when comparing the two. In the case
of underground at 765 and 1,100 kV alternating current
(a.c.) and overhead at 1,100 kV a.c., the loadings shown
have not actually been attained; they are considered,
however, within the capability of present-day tech-
nology. .
While use of underground cables at distribution
voltages is at a parity with overhead construction in
many areas of the country, no such trend is evident in
transmission. Even the most recent line construction
figures in transmission voltages show a ratio of overhead
to underground of about 100: 1, comparing mileage of
each installed. This is particularly true at the higher
voltage levels, where high-power transfer capabilities are
absolutely necessary and where, as can be seen in table
1, there are no underground installations.
This almost total reliance on overhead transmission,
while undeniably affected by cost and power capacity,
also reflects a large element of design deficiency and,
probably more important, past application practices and
needs.
For example, if the underground cable circuits in
the United States were identified as to location and
application, it would be found that over 90 percent of
the installed circuit miles are in two or three east coast
utilities where population density and siting restrictions
have literally forced the use of underground, at whatever
costs. Further, the vast majority of these circuits are less
than 10 miles in length and carry no more than 100 MW
per circuit. In effect, then, for long, high-capacity cir-
cuits, overhead is the only mode of transmission utilized
by the electric operating companies.
The goal of the research effort in underground
cables is, in reality, to produce a credible alternative to
overhead transmission, so that where an application de-
mands underground, capacity and voltages can be com-
parable and costs can be more nearly equal. Such an
alternative does not exist today.
In that research, a considerable percentage is de-
voted to development of the cellulose and synthetic
cables for the higher transmission voltages-that is, 500
and 765 kV. There is no hope, however, that such de-
velopments, even if successful, would be used on circuits
for loads greater than 1,000 MW, over any distance
greater than 20 miles. This must be left to the new tech-
nologies in underground-gas cables, cryoresistive cables,
and superconducting systems.
Table 1. Per circuit capacity and installed miles
of transmission lines in the United Statesa
:=;l~~~~:~~==="~Z;~;'i=;;(;=)==-=Mi1~~~D=;&~
rating O.H.* U.G.* O.H. U.G.
138,000 400 400 159,500
230,000 900 700 54,500
345,000 1350 900 25,500
500,000 2000 1200 12,300
765,000 4000 2000 1,400
1,100,000 7000
1900
200
200
-
a
From a report by IEEE T&D Committee.
bAs of 1-1-75.
*Legend:
O.H.--overhead; U.G.--underground.
213
-------
Of these, the most promising from a standpoint of
economics and loss reduction are gas-insulated cable
systems and wperconducting transmission lines, both of
which could be installed under or on ground.
ELECTR IC POWER TRANSMISSION MEANS
There an ~everal electrical parameters that must be
considered in comparing transmission modes, and they
do not react or affect the critical limits of each mode in
the same way.
Surge Impedance
Overhelld line surge impedance varies from 500
ohms for 145-kV lines to 250 ohms for 1,200.kV lines.
For underground conventional cables it varies from 25
to 35 ohms au;1 is relatively independent of system volt-
age, while for gas cables it is about 60 to 70 ohms. The
surge impedar,c2 is a function of the line dimensions,
and construction numerically is equal to L/C. It affects
power transfer capability and line transient character-
istics.
Line Capacitanr:e
There is a1 inherent capacitance to ground for any
energized transmission line. The capacitance varies from
about .012 to .025 J.lfds/mile for overhead lines as a
direct functiOl1 of configuration and distance from
ground. For ur'cerground conventional cables it is about
0.5 J1fds/mile regardless of voltage. Gas cables have a
lower capacita,lce, between 0.1 and 0.2 J1fds/mile, be-
cause the gas dielectric constant is unity. The amount of
compensation required is directly related to line capaci-
tance.
Loading or Power Transfer Capability
For overhead lines, loading is thermally limited only
at 230 kV and below. At the higher voltages, system
stability, as de':ermined by surge impedance loading, is
the limiting fa :tor. Equally important is the fact that
conductor size and number of conductors per phase is a
function of corona start voltage limitations. At 345 kV
and above this J5ually results in multiple conductors per
phase; the cond uctors are thermally capable of carrying
sever
-------
Table 2. Comparison of costs for various transmission modes
- _._-
-------..----- -----. - ---
kvar/f.1W- $4 MW-MI
KV MAX. OH* UGI* GIS* SC*t OH UGI GIS SC
145 .25 8 500 2,000 600
242 .33 10 275 1,200 400
362 .55 12 1.25 220 1,000 350 250
550 .88 14 1.5 6 210 600 300
800 1.5 20 2.5 130 500 230
1,200 2.3 3.3 110 200
*Legend: OH--overhead; UGI--underground gas insulated;
GIS--gas insulated system; SC--superconducting.
tRough Estimate, source Brookhaven National Lab.
onerous and distasteful results of man's careless use of
nature.
The environmental effects of overhead transmission
can be separated into two classes, the known and the
presumed. I n the first category are such well-
documented and studied effects as radio noise, television
interferences, and audible noise. These are all the result
of corona that may appear on a line during foul weather
operation, or from breakage or damage to line insula-
tion, hardware, or conductors. Radio and television
interference problems, while responsible for the majority
of complaints, are still small in number relative to the
many miles of installed transmission line. They can
occur on lines of any voltage level, and are usually clear-
ed up by local maintenance or filtering.
Audible noise, on the other hand, although an even
rarer problem in the past, has the potential of being a
real barrier to the use of higher transmission voltages.
This is because the probability of annoyance increases
significantly with the higher voltages, and because
psychoacoustical responses to such steady noises, where
ambient levels are of similar magnitude, are an unknown
quantity. These direct effects of corona are being
studied, particularly the audible noise. Acceptable limits,
as determined by research programs, will be attained by
proper line construction.
From the aesthetic or visual aspect there is also
much to be gained from research. Although the tradi-
tional lattice structure, against which most outcrys have
been directed, is till the most efficient in terms of
material usage, it represents to the public the outmoded,
the archaic, and the undesirable in structural design.
While the utilities have made considerable progress in
tower design-evident from many unique structures,
primarily using steel poles and curved graceful cross-
arms-no attempts have been made to integrate the new-
est of insulating materials into designs of conductor
support systems., These new materials have strength-to-
weight ratios more than an order of magnitude greater
than the traditional materials such as porcelain and glass,
and thus invite use of totally different concepts in line
support structures. Transmission line support systems
using these materials, and designed with minimum en-
vironmental impact as a major criteria, may some day
inspire as much admiration as certain other lattice
structures supporting catinaries and cables crossing our
rivers and waterways. To attain this millenium, however,
it will probably be necessary to convince the public that
electric power is as important to them as the automobile
and getting home from work every night.
In the other category of environmental effects are
those which have biological implications. These basically
215
-------
concern the I!ffect of the electric fields under high-
voltage lines ,m life forms. either human, animal, or
plant. Howover, one corona-induced consequence.
ozone, has also been included as a possible hazard. The -
latter concern should have long ago been laid to rest,
since the mea!ured ozone production of transmission
lines, even whon operating at 1,500 kV, has been shown
to be an orde r of magnitude less than that normally
present in the so low that a harmless spark discharge
occurs every ti.ne another object is touched. That there
is psychoelectrical response is undeniable. This can occur
both under transmission lines with certain atmospheric
conditions and with a high enough field at ground level.
The electr.c utilities have, until a few years ago,
maintained .1 head-in-the-sand attitude about the
possible biological effects of HV transmission. This, of
course, should oot be tolerated. It is a responsibility of
the industry and the government to prove that no hazard
to the public e>cists under high-voltage lines, either those
in operation or those contemplated for the future. This
can be done w:th well-organized and controlled experi-
ments both in the laboratory and under operating lines.
We cannot afford to build 1,200-kV lines and fence in
the right of wav as we do interstate highways. The over-
head transmission lines of the future can be designed so
that field strennths on the ground are at levels that are
safe, presenting no danger, seen or unseen, to human,
animal, or plant life. Indeed, we may find that. as with
X-rays or alcohol, a little electric field is beneficial.
CONCLUSIONS
The transmission of electric power is as much a
mainstay of our form of life as our highways, railroads,
and waterways. Transmission lines are now and will even
more in the future be essential to success in attempts to
conserve our scarce fossil fuels, since only with extensive
transmission facilities can the needed shift to a more
electric economy take place.
In the transmission system itself it is possible to
conserve energy directly by reduction of losses through
use of higher transmission voltages, better materials, and
more efficient cable systems. The choice of whether to
use overhead Or underground circuits is not only one of
economics but also of technology and the basic. charac-
teristics of each system. The systems of the future will
be more composite, including both overhead for long-
distance power transfer through sparsely populated or
recreati.onal areas, and underground where urban areas
must be traversed, converting from one mode to the
other on a one-to-one basis both in voltage and power
capacity.
Underground circuits have a minimum impact on
the environment, in that there is only a temporary
disruption of the ground during installation. Overhead
transmission has a number of effects on the immediate
area around and under the lines. Proper and imaginative
design of these lines can produce installations presenting
minimum consequence to the environment no hazard to
the public.
REFERENCE
1. "Overhead And
Electric Power,"
Committee.
Underground Transmission of
Information Paper, IEEE - T&D
216
-------
GEOTHERMAL ENERGY DEVElOPMENT*
Paul Kruger t
Abstract
Projections of electrical energy demand over the
remainder of this century indicate the need for greater
reliance on energy resources other than oil and gas, even
with improved technologies for their increased ex(rac-
tion efficiency. Several alternate strategies are being
considered to reduce the apparent need for increased oil
and gas import. Among them are: conservation in the
form of improved utiliz(1tion efficiencies; synthesis of
fluid fuels from abundant coal and oil shale resources;
development of advanced electric energy generating
systems, such as breeder, thermonuclear, and solar con-
version reactors; and increased utl1ization of under-
developed resources, such as geothermal heat, solar
thermal energy, and combustible biological and waste
products. .
It is estimated that improved efficiencies in end-use
can reduce energy consumption in the year 2000 by
some 25 percent compared to the strategy of no new
initiatives. Replacement of this demand by synthesis of
fluid fuels from coal and oil shale could reduce substan-
tially the levels of oil importation. Intensive electrifi-
cation to meet the same demand would require
maximum near-term use of abundant resources of coal
and uranium and eventually .the advanced electrical
generating technologies. Geothermal energy might playa
small but significant role.
Each of these energy sources has its own spectrum
of technical, economic, environmental, and social prob-
lems which determine the magnitude and time scale of
its contribution to the future total energy supply, either
as a substitute for fluid fuels or as an alternate source of
electricity. Examples of underdeveloped energy
resources are oil shale for the production of oil and geo-
thermal heat for the production of electricity, for which
status evaluations of their technical, economic, environ-
mental and social problems are available.
*This presentation was based on reports prepared during
Leave of Absence 1974-75 with the National Science Foundation
and the Energy Research and Development Administration.
tProfessor of Nuclear Civil Engineering, Civil Engineering
Department, Stanford University, Stanford, California.
INTRODUCTION
The Nation has embarked on an aggressive program
to develop its indigenous resources of geothermal
.energy. For more than a decade, geothermal energy has
been heralded as one of the more promising forms of
energy alternate to oil and gas for electric power genera.
tion, but during the last 15 years, the total capacity in
the United States has reached 502 MWe, about half the
size of a single modern nuclear power plant. And yet,
the United States, especially its Western and Gu If Coast
States, is believed to possess a vast resource base of geo.
thermal heat at depths up to 3 to 10 km. Many estimates
of these potential resources suitable for the production
of electric power have been published, and they range
over a spectrum of more than a factor of 100. This
variation suggests that the potential is essentially un-
known.
Table 1 gives a range of published forecasts for the
year 1985 and the equivalent potential in number of
1,000 MWe power plants and in oil consumption in mil-
lions of barrels per day. In view of the estimated con-
struction of about 200 to 250 nuclear power reactors by
1985-90, the pessimistic forecasts clearly show that the
contribution of geothermal energy to the Nation's
energy supply may indeed be small. The optimistic fore-
casts represent more than 15 percent of the total electric
power requirements estimated for the year 1985. The
Task Force for Geothermal Energy, in the Federal
Energy Administration Project Independence Blueprint
report of November 1974, established a national goal for
i 985 of 20,000 to 30,000 MWe, the latter value repre-
senting an equivalent energy supply of 1 million barrels
of oil per day. This goal was clearly a compromise
between what is worth a national effort and what might
be realistically achieved. The potential for adding or
replacing the equivalent of some 25 nuclear power plants
or for conserving 1 million barrels of oil per day should
be an adequate incentive for the Nation to accelerate the
. development of a viable geothermal industry.
A puzzling enigma appears. If the potential resource
base of geothermal energy is so vast, why has significant
utilization not occurred? The entire U.S. production of
electric power from geothermal resources occurs at one
location, the Geysers in California, where, over a 15-year
period starting in 1960, generating capacity has grown
\
217
-------
Table 1. Forecasts and national goal for development of U.S. goethermal resources
~:-===_.-
Pessimistic National goal Optimistic
forecasts 1985 forecasts
Electric and thermal
energy capacity (MW) 2000-4000 20,000-30,000 182,000-400,000
Equivalent in
nuclear reactors
(No.1 , JOO-MWe units) 2-4 20-30 182 -400
Equiva1en't consumption
of oi 1 : 1 06 bb 1 / day) 0.07-0.14 0.7-1 6-14
from 12 MW Sl pplied by Unit No.1 to the total of 502
MW attained with the startup of the 106-MW Unit No.
11 in May 1975. The Geysers is the largest geothermal
electricity generating station in the world. The entire
worldwide capa:ity of electric power generation by geo-
thermal resources is slightly more than 1,000 MW, the
equivalent of t 1e capacity of a single modern nuclear
power plant.
Utilization of geothermal fluids for thermal energy
in the United States is almost negligible. And yet
throughout the country, fossil fuels are consumed in
large quantities to boil water for heating and electric
power generation, both at very low thermal efficiency.
Some countries already use geothermal fluids for ther-
mal energy, notilbly Iceland, where municipal heating is
an important u1 Hization. Several countries, responding
to increased public awareness that future supply of fossil
fuel may be very limited, are examining the potential use
of indigenous thermal waters for industrial and munici-
pal heating.
How is this ,migma to be solved; how is the United
States (and other countries). endowed with potentially
bountiful geothe-mal resources, going to develop these
natural resource~; as a significant contribution to its
energy supply? T1e attainment of a national goal to con-
tribute an equivalent of 1 million barrels of oil per day
from geothermal resources clearly requires accelerated
development of a geothermal industry capable of provid-
ing 20,000 to 30,000 MW of electric power and thermal
energy in the nex1 10 to 15 years. And this objective will
require a natioml effort to accelerate and coordinate
development in t 1ree parallel tasks: (1) the discovery,
provin!J. and extr,lction of geothermal resources to pro.
vide a significant supply of hydrothermal fluids for
direct utilization and to produce more than 5 x 1012
kWh of electricity over the amortization period of the
investment in resource development and power plant
construction, (2) the technology to convert the re-
sources as found in its various natural forms and qual-
ities into electricity, and (3) the removal of unnecessary
institutional constraints to' the rapid development of a
cost-effective and environmentally acceptable industry.
A major factor which helps create the enigma of
vast resource base and little utilization is the variability
of geothermal resources. The geothermal energy cycle,
although simple compared to other alternate energy
sources, is actually complex in that geothermal resources
occur in many types of geologic, thermodynamic, hydro-
dynamic, and chemical quality. As a result, the major
problems in the energy cycle vary by type of resource.
Table 2 lists the key aspects of the cycle from explora.
tion to utilization that must be evaluated for each type
of resource.
Several general reviews of the state-of-the-art of
geothermal energy resources and technology are listed as
references 1-6. One is the proceedings of the 1970
United Nations symposium on the development and
utilization of geothermal resources. Another is the 1973
compilation of an ad hoc working group convened by
UNESCO. A general introduction to geothermal energy -
is the proceedings of the American Nuclear Society con-
ference on geothermal resources, production, and stimu-
lation held in 1972. Among other compilations of papers
on geothermal energy are the proceedings of the second
and third All-Union conferences on geothermal rmp.r!JY
organized by the Scientific Council for GeolhNrnal
218
-------
Table 2. Problem areas in the development of the
geothermal energy cycle
Variability of geothermal resources
Location of subsurface reservoirs
Reservoir evaluation
Extraction technology
Conversion technology
Potential for multiple utilization
Environmental impact control
Legal and institutional constraints
Investigations of the U.S.S.R. Academy of Sciences.
Translations of these proceedings are not generally avail-
able, but much of the technical content is given in the
Soviet papers of volume 2 of the United Nations Sym-
posium and in the ARPA reviews of Soviet literature in
geothermal energy. The proceedings of the Second
United Nations Symposium held in San Francisco in
May 1975 adds another major contribution to the litera-
ture of geothermal energy.
GEOTHERMAL RESOURCES
The upper 10 km of the Earth's crust may contain
more than 3 x 1026 cal of heat, a resource base readily
classified as vast. However, much of this energy is too
diffuse to be exploitable as an energy source. Geother-
mal resources may be defined as localized deposits of
geothermal heat concentrated at attainable depths, in
adequate volumes, and at temperatures sufficient for
commercial exploitation.
The only geothermal resources presently used for
electric power generation are high-quality hydrothermal
convective systems which contain high-enthalpy geo-
fluids suitable for transferring the geothermal heat to the.
surface for direct use in low-efficiency steam turbines.
Unfortunately such resources have been discovered at
only a few places on earth. More than 75 percent Qf the
world's geothermal electric power capacity results from
vapor-dominated hydrothermal systems which produce -
dry or superheated steam for direct conversion. The
remaining capacity results from high-temperature, low-
salinity, water-dominated hydrothermal systems in
which the geofluids are flashed on production, and only
the separa,ted steam is used for electric power genera-
tion. The liquid fraction is either wasted or reinjected
into the ground. These systems are commercially less
desirable because only a small fraction of the water
flashes to steam, thermal efficiencies are low, and plant
operational problems are more severe.
Liquid-dominated hydrothermal systems are ex-
pected to be many times more abundant than vapor-
dom inated hydrothermal systems. Moderate-to-high
salinity hydrothermal resources may be more abundant
than low-salinity resources. And other types of geother-
mal resources, such as hypersaline brines, geopressured
fluids, volcani~ and magmatic deposits, and impermeable
hot-rock massives, which are not yet commercially
exploitable, may be even more abundant than the cur-
rently exploited hydrothermal resources. Thus the
answer to the utilization enigma may lie not so much
with the magnitude of the resource base, but more with
the ability to locate suitable concentrated deposits of
geothermal heat and with the technology to extract the
energy in quantities which are economically and environ-
mentally feasible.
Although estimates of the geothermal resource base
are available, the maglJitude of the potential reserves is
not yet well defined. The location of underground
deposits of geothermal heat, especially where thermal
manifestations are not visible at the surface, is a difficult
task. Over 1 million acres of "hot spots," areas of known
geothermal energy, were identified as early as 1967 by
the U.S. Department of the Interior in designated Fed-
eral lands in five Western States as having current poten-
tial value as geothermal resources. An additional 86 mil-
lion acres of land in 13 States were designated as pro-
spectively valuable for geothermal resources. Since then
several other inventories of known geothermal resource
areas (KGRA) have been compiled. A current assessment
of U.S. geothermal resources has been completed by the
U.S. Geological Survey and a summary of the resource
base, by resource type, is given in table 3.
Exploration for geothermal resources has been
undertaken by industry on private lands and, through
the Federal Leasing Act of 1970, on Federal lands by
competitive and noncompetitive leasing under super-
vision of the Bureau of Land Management. Although
total values are difficult to ascertain, it is estimated that
about 100,000 acres on Federal public lands and about
200,000 acres on Federal Indian lands were under lease
for geothermal exploration in mid-1975.
Resource exploration and assessment of potential
reservoirs of geothermal energy are made by the variety
of earth science methods listed in table 4. Details of
these methods are available in the general references
listed in the introduction. The final phase of geothermal
exploration is the drilling of exploratory wells. It is from
these wells that data for evaluating the suitability of the
219
-------
Table 3. Summary of geothermal resource base of the United States
Estimated heat content (1018 cal)
Identified Potential
Hydrothermal convection systems
Vapor'-domi nated (steam)
High temp. - hot water
(T > 1500 C)
Mod.tempo - hot water
( 9C 0 - 150 0 C)
TOrAL
Hot igneous systems
, Magmcl and hot rock
Geopre~;sured basin part of
regiunal conductive systems
Total resource base
26 50
370 1~600
345 1~400
740 3~000
25,,000 100,000
10,920 44,000
36~660 147~000
aFroni D. F. White and D. L. Williams~ eds.~ Assessment of Geothermal
Resources of the United States - 1975, U.S. Geological Survey Circular
726~ 1975.
resource as a production reservoir are obtained. Majo"--
factors in t,c economics of exploration and production
of geothernal fields are the success of techniques avail-
able for sUliace exploration of potential resources and
exploratory drilling of potential reservoirs. Improve-
ments and novel methods for reducing costs in these two
initial phaSf's of the geothermal energy cycle are thus of
great importance. In order to achieve the goal of provid-
ing an equivalent of 1 million barrels of oil per day by
geothermal resources, it is evident that exploration for
geothermal resources, especially hydrothermal, must
receive a very high priority by the U.S. energy resource
industry.
The m,;gnitude of hydrothermal resources required
can be esti Tlated from the following calculation for a
100 MWe gl!nerating plant operating with flashed steam
of 555 kc,il/kg (1,000 Btu/lb) heat content. The re-
quired geofuid production rate for a hot-water system
yielding 10 percent steam on flashing with a thermal
efficiency of 20 percent would be 7.75 x 106 kg/h (1.7
x 10' Ib/hr). The amortization of the 100 MWe plant
over a period of 30 years would require a total produc-
tion of 2.1 )( 1012 kg hot water, and a mean reservoir
porosity 01 10 percent would require a geothermal
reservoir volume of about 2 km3. At a 50-percent con-
densation efficiency, the plant would discharge a hot
water supply of about 100,000 m3/d (2.5 x 10' gpd).
For a national capacity of 20,000 MWe, these values
are multiplied by a factor of 200. Thus reservoirs sup-
porting 200 units of 100 MWe generating plants must be
located. These reservoirs will produce about 1.5 x 109
kg/h of hot water. For a mean well production flow rate
of 250,000 kg/h, a total of 6,000 production wells will
be needed, and for a mean spacing of 100,000 m2/w,JII
(25 acres/well), a total reservoir area of 6 x 1011 m2 (1 5
X 105 acres) of geotherm~1 resources must be found. It is
evident that if hydrothermal systems are to provide the
Nation with 20,000 MWe, very high priority for resource
exploration and assessment is indeed required.
UTILIZATION TECHNOLOGY
Utilization of geothermal energy varies with the--
quality of available resources. It has been noted that the
present geothermal industry has focused on high quality
hydrothermal resources. Extraction and conversion tech-
nologies for dry-steam reservoirs are sufficiently ad-
vanced to be commercially attractive. Conversion tech-
nologies for hot-water resources are more complex, and
220
-------
Table 4. Geothermal exploration methods
Exploration surveys
Airborne
Aeromagnetic survey
Thermal infrared survey
Geological
Tectonics and stratigraphy
Recent faulting
Distribution and age of volcanic rocks
Thermal manifestations
Hydrologic
Surface discharge of geofluids
Temperature of fluids
Chemical composition of fluids
Groundwater hydrology
Meteorology.
Geochemical
Chloride concentration
Si02 content
NA-K-Ca ratios
Isotopic composition of
hydrogen and oxygen
Geophysical
Geothermal gradient
Heat flow
Electrical conductivity
Seismic activity
Exploration hole drilling
Reservoir characteristics
Temperature-depth profile
Pressure-depth profile
Lithology and stratigraphy
Permeability log
Porosity log
Fluid composition
221
-------
for hot brines, geopressure'd basins, and hot dry rock
formations, they are even more complex; commercial
utilization :s still further away. Since these latter types
of resource hold great promise, technology to exploit
them mustJe developed.
Stimulation of geothermal energy production can be
achieved b'l research and development to (1) increase
the modes of resource utilization, (2) improve energy
conversiorl technology, and (3) provide advanced
methods 01 energy extraction. Increased efficiency in
each of thnse three aspects 9f the geothermal energy
cycle is attainable.
Develol)ment of a geothermal field generally in-
volves the ~eofluid characteristics, steam separation and
gathering hcilities, turbine and generator equipment,
cooling systems, and condensate disposal methods. Such
developmert presupposes that electric power generation
is the sole purpose of the field development. It may turn
out, howevnr, that for many geothermal reservoirs, non-
electric utilization of the resource may make the reser-
voir econor"ically feasible, with significant conservation
of fossil and nuclear fuels. Several modes of utilization
of geotherr'1al resources are listed in table 5. Hydro-
thermal fluids with temperature or enthalpy too low for
economic electric power production may be useful for
water or m: neral sources and for industrial, agricultural,
and municipal heating. 'However, since major interest in
geothermal energy is for the production of electric
pt)wer, combined or total utilization may help make
many gee thermal reservoirs submarginal in power
production alone become economically feasible. The
possibility cf building a community around a geothermal
resource, with municipal heating, an industrial park of
process firms requiring hot water and concomitant elec-
tric power production appears feasible. Thus, research
for method; stimulating geothermal resource utilization
in all forms is well warranted. '
Genera: methods for producing electricity from geo-
thermal fluids are summarized in table 6 and are de-
scribed adequately in the several cited references. The
choice of a conversion cycle is generally dependent on
the thermodynamic and chemical properties of the geo-
fluid. Presm'!t commercial plants utilize low-salinity
hydrothermal systems with steam or water at tempera-
tures above about 2000 C in the single-stage direct steam
turbine con wrsion system. To utilize lower temperature
fluids, investigations are underway to develop other con-
version sys':ems; among these are multiple-flash low-
pressure ste am turbines, single and multiple stage binary
cycle systens, and hybrid systems combining these two.
The binary system appears to be the most promising for
utilization of geofluids with temperatures between 1000
C and 2000 C. However, only one experimental facility"
c
Table 5. Utilization of geothermal energy
Electric power production
Direct use of dry steam
Flashing of hot water to steam
Surface flashing
Insitu flashing
Binary and hybrid cycles
Innovative single-well converters
Direct use of thermal waters
Agriculture
Aquiculture
Space hea ti ng
Industrial processing
Medical therapy
Byproducts
Mineral extraction
Water resources
the Pauzhetka station in the Kamchatka peninsula of the
U.S.S.R., has been constructed to date. The binary sys- ~
tem most likely to be successful in the United States will
require a downhole pump to prevent flashing. a heat
exchanger which can operate without excessive corro-
sion and deposition, and a circulation system which
allows for reinjection of the geofluids for environmental
control purposes.
Several types of downhole pump are under devel-
opment, 'involving design concepts which use: (1) in situ
heat to operate a closed steam-generator-turbine to drive
the pump, (2) a high-speed, high-temperature, hi'lh
length-to-diameter electric-motor driven pump, or (3) a
hydraulically driven unit with hydraulic power from the
surface. Heat exchanger concepts include fluidized sand'
beds to enhance heat transfer rate and maintain clean
surface, and liquid-liquid systems with direct contact of
immiscible fluids, tray-tower contactors, or subcritical or
supercritical power cycles.
F lash and binary systems are usefu I in large power
plants having capacity in excess of 50 MWe. They re-
quire complexes of multiple-well field development and
extensive networks of gathering lines. I nnovative conver-
sion systems are under development in which small
power plants, in sizes of 1 to 15 MWe, may be installed
at individual wells. These systems may involve a total
flow concept in which both the thermal and kinetic.
222
-------
Table 6. Types of geothermal power plants
Heat source
Generation mode
Dry steam
Hot water (T >
Hot water (T <
1800 C)
1500 C)
Hot water
(moderate salinity)
Hot brine (pressurized)
Hot brine (flashed)
Steam turbine
Steam turbi ne
Binary cycle
Hybrid cycle
Binary cycle
Impact turbine
Helical screw expander
B1ade1ess turbine
energy of the geofluid is used for production of elec.
tricity.
One of these is the impulse turbine, in which the
thermal energy is converted to kinetic energy through a
converging-diverging nozzle, and the high-velocity out-
put drives a hydraulic impulse turbine operated at low
pack pressure. Calculations indicate that a large unit
(e.g., 200 MWe) might be feasible for the Salton Sea
geothermal brines, which contain as much as 230,000
ppm total dissolved solids. The material handling prob-
lems of such brines are indeed enormous, but the dis-
solved solids may also represent a source of valuable
minerals, such as lead, manganese, and copper, if they
can be processed economically.
Another total-flow concept is the helical rotary
screw expander which expands the vapor from hot satu-
rated liquids by continuous pressure reduction in the
expanding screw, in essence creating an infinite series of
flashing stages. A small 62.5 kV prototype model was
tested successfully with moderate salinity geofluids with
indications that it can accept the total flow of untreated
brines. Still another concept is the bladeless turbine, in
which a series of closely spaced disks are rotated by vis-
cous drag exerted by geofluids introduced by a nozzle.
The device seems simple and self-cleaning, but the over-
all efficiency may be small.
Increased extraction efficiency represents a major
means to stimulate geothermal energy production, espe-
cially for nonhydrothermal reservoir systems. Calcula-
tions show that hot-water reservoirs contain a larger
amount of available energy than steam-filled reservoirs
under the same reservoir conditions because of the much
larger mass of water; but in either system, the heat con-
tained in the rock formation is much larger than the heat
in the fluids. Thus recovery of the formation heat would
be of major economic significance. Extraction of forma-
tion heat must be a non isothermal process, which can be
achieved either by flashing geothermal liquids to steam
within the formation or by recycling colder fluids back
into the formation. Laboratory investigations and
theoretical calculations of reservoir models are underway
to determine the extent of heat extraction from frac-
tured reservoir formations.
The natural extraction efficiency of energy from
impermeable hot dry rock formations is extremely small.
And yet hot dry rock in the upper 10 km of the Earth's
crust represents a major potential resource of geothermal
energy. The volumetric energy extractable from hot dry
rock, calculated for average expected properties and
possible technical extraction efficients, is of the order of
1.2 x 109 kWh/km3 of fractured rock, equivalent to a
volumetric power extraction of 1.4 MW/km3 for one
century. The technical challenge is the ability to fracture
such volumes of hot rock massives and achieve an extrac-
tion efficiency of the order of 10 percent.
Fracture stimulation methods are useful for many
types of geothermal reservoirs. In vapor-dominated
systems, stimulation may restore declining pressure or
connect dry holes in commercial steam fields to produc-
ing sections. I n liquid-dominated systems lacking suffi-
cient productivity for economic power generation, frac-
ture stimulation may provide larger wellbore diameter
for increased flow rate, greater surface area for heat
transfer, or restore porosity or permeability around wells
223
-------
having deposi':ed silica, calcite, or other precipitated
minerals. In :lry geothermal systems, stimulation is
needed to plOvide large fracture volumes for heat
transfer to an ,Irtificial convective extraction system.
Several fncturing methods are under study; these
include hydraulic fracturing, thermal stressing,. and
chemical and r,uclear explosive fracturing. Hydraulic and
explosive fracturing methods have already proven suc-
cessful in stimulation of natural gas reservoirs.
Experimellts to evaluate the potential for hydraulic
and thermal st'ess fracturing for recovery of geothermal
energy from hot dry rock formations are underway. In
this concept, large diameter vertical crack is created
hydraulically ,It the bottom of a borehole in the geo-
thermal formation. A second hole is drilled to intersect
the upper par1 of the fracture, and a pump is used to
initiate artifici.11 heat-extraction circulation. It is hoped
that pumping can be discontinued if a natural convective
circulation is a.:hieved. The major technical problems are
the attainmen1 of a vertical crack of about 2 km diam-
eter with suffi :ient fracture area, the creation of addi-
tional fracture areas by thermal stress of cold water
injection, and ':he ability to achieve a natural convective
circulation wit,out undue losses of water, especially in
arid regions. Calculations indidte that under favorable
conditions, thE system might provide an average power
of about 100 W'W (thermal) for 20 years.
IIIJSTITUTIONAL ASPECTS
Although much remains to be done in locating ade-
quate reserves and developing adequate technology to
meet the goal! for exploiting the Nation's geothermal
resources, there is great confidence that these will be
achieved. Thes~ problems involve advances in physical
research and Bchnology. I nstitutional problems, how-
ever, also exist. Such problems are complex; they involve
public acceptar ce, vested interests, historical precedents,
existing regulations from other resources, overlapping
jurisdictions, alld economic and financial factors. These
problems are often more difficult to resolve than are
engineering prcblems, and they may in the long run be
the major constraints to an accelerated, but orderly
development 0:' geothermal resources. The solutions to
many institutic,nal problems may require broad public
interaction, changes in regulations and legislation, and
perhaps change,; in traditional investment and marketing
procedures.
Economic factors affect all forms of energy supply;
they involve tctal capital costs per installed power unit
and operational costs per l!nit of energy production. For
geothermal eni!rgy, both of these cost factors are
strongly depen :lent on the specific characteristics of
individual reservoirs and the size of the installed power
plant units. Important capital costs include the invest-
ments for exploration, drilling and completion of wells,
gathering lines, and waste-handling systems for all util-
izations. For thermal energy applications, they also
include the distribution system, and for electric power
production, they include the power plants and the trans-
mission network. The production costs are influenced by
the cost of capital, operations and maintenance, and
. plant utilization factor. In the United States, additional
costs must also be added for environmental pollution
control.
Factual cost data for geothermal electric power
production in the United States are available only for
the Geysers field. The electric utility purchases steam
from only one supplier, but has negotiated to purchase,
steam for future plants from additional suppliers. I n the
development of future geothermal power stations, an
option exists for an integrated operation from explora-
tion to power production in contrast to the traditional
roles of an electric utility purchasing steam or hot water
from an independent supplier. The general effect would
be an increased investment cost per kilowatt hour of
energy.
Data for costs of recently constructed power plant
units at the Geysers are sparse, but estimates for the
original plants range from about $100 to $150 per kW.
Production costs were estimated at about 7 mill/kW~ of
which about 3.5 mill/kWh was the price of the pur-
chased steam. These estimates included a cost of 0.5
mill/kWh for injection of water condensate as a disposal
method. With escalation of drilling, construction, and
environmental reporting costs over the past few years,
these cost values are not useful for estimating costs of
new facilities, especially for reservoirs which do not
produce dry steam. Recent estimates indicate installa-
tion costs may range from $500 to $700 per kW and'
operating costs for binary conversion systems of the
order of 20 to 40 mill/kWh. These costs, of course, are
hypothetical, and more precise costs will be generated as
other major reservoirs and plants are developed and
operated. A large uncertainty in the total cost is the
fixed exploration cost for the resource, which is inde-
pendent of plant capacity, and the average drilling costs
of the production, dry, and injection wells. Computer
models to evaluate the relative importance of these
resource and utilization costs are under development.
Because of large uncertainties in the technical costs
of exploration and drilling, conversion efficients, and
stimulation techniques, and because of the rapid escala-
tion rate of these costs, it is difficult not only to esti-
mate costs on an absolute basis, but even to compare
costs of other forms of electric power generation.
224
-------
Besides the costs affected by these technical factors,
other factors more social in nature must be considered.
Among these are public acceptance and government
stimulus for accelerating the development of geothermal
energy in relation to other energy sources, the inter-
pretation of compliance with the National Environ-
mental Protection Act of 1969, and the availability of
investment capital for development of geothermal
resources and electric and thermal power plants. These
socioeconomic factors may require much public and gov-
ernment deliberation before general philosophies are
widely accepted in practice.
Although geothermal energy is considered to be one
of the least polluting of the many forms of energy avail-
able, it should be assumed that the public will insist that
the environmental impact of producing geothermal
energy in all of its natural and stimulated forms be
thoroughly investigated in accordance with NEPA and
any additional requirements under State and local legis-
lation. Furthermore, in addition to environmental
impact, it is also evident that assessment will be required
of the operational aspects of the various types of reo
sources which affect personnel safety and plant mainte-
nance.
In the evaluation of a benefit-risk analysis, geo-
thermal energy is expected to compare favorably with
respect to other energy resources, especially when
viewed over the entire fuel cycle. Since geothermal
energy must be utilized or converted in the vicinity of
the resource, the entire "fuel cycle" from reservoir to
transmission is located at one site. This is in contrast
with material fuels in which the cycle involves mining,
storage, refining, transportation, reprocessing, and waste
disposal, many or all of these at different locations.
Furthermore, increased utilization of geothermal energy
may result in a correspondingly reduced demand for
material fuels in short supply, such as natural gas, oil,
coal, and uranium. And still further, geothermal fluids
may provide byproduct sources of water with reduced
demand for cooling water.
Geothermal energy, nevertheless, has its array of
potentially deleterious environmental impacts. A list of
potential environmental impacts is given in table 7. A
review of the more important ones has recently been
completed in a workshop sponsored by the National
Science Foundation (ref. 7) as the basis for a program to
support research for baseline data and technology for
monitoring potential impacts and controlling actual
Table 7. Environmental impacts of geothermal
power production
Land
Air
Water
Land util i zati on
'V2 km2/100 MW
Drilling operations
'V20 wells/km2
Power plant construction
1 pl ant/l 00 MW
Steam gathering lines
'VO. 2 km/l 00 MW
Condensate reinjection
Lines and pumps
Potential for
Land subs i dence
Sei smi cacti vi ty
Disposal of
drilling fluids
Need for supplementary
cooling water
Bui 1 dup of
salt concentrations
- --- -- -
Release of steam
and other gases
during drilling
and testing
Noise pollution
Re lease of H2S an d
other noncondensab le
gases
Potential changes in
micrometeorology
Disposal of
condensor cooling water
steam condensate
'V3.5 Mgpd/100 MW
Potential mineral and
thermal pollution of
fresh surface waters
Potential for well
blowout
225
-------
hazards. The major impacts include gaseous emissions,
liquid waste disposal, and geophysical effects such as
seismicity and ;ubsidence. Other concerns involve ther-
mal releases, su:face water contamination, land use plan-
ning, cooling Vlater consumption, and visual and noise
pollution.
An array Jf legal problems associated with geo-
thermal resource development also exists. These have
been reviewed in another workshop sponsored by the
National Science Foundation (ref. 8).
The legal )roblems of geothermal resources begin
with resource definition, which varies from State to
State. For example, in California geothermal resources
are defined a! "the natural heat of the earth, the
energy-which 'nay be extracted from naturally heated
fluids-but exe luding oil, hydrocarbon gas or other
hydrocarbon substances." This definition leaves open
the question whether geothermal resources are legally
defined as water, mineral, or gas resources, and results in
large uncertainty with respect to Federal, State, and
local jurisdictions. On the other hand, the State of
Hawaii considHrs geothermal resources as minerals,
whereas the StHte of Wyoming has declared them water
resources. As VI ater resources, they would be subject to
the very complicated set of State laws concerning water
rights and regulation. As minerals, they would be subject
to mining laws and such problems as ownership, deple-
tion allowances, and write-off of intangible drilling costs.
Geothermal re;ources have already been classified in
court decision! in different ways. In one case a U.S.
District Court in San Francisco treated the Geysers geo-
thermal resource as "nothing more than superheated
water" and therefore not a mineral, but in another case,
the resource was held to be a gas within the meaning of
the Internal Ftevenue Code provisions for depletion
allowance and intangible drilling costs.
Ownership. rights are also a serious institutional
problem. The Federal Government has given some 35
million acres cf land to the homesteaders, States, and
railroads, but generally reserved the mineral rights to the
Federal Government. However some State grants in-
cluded mineral rights and thus many problems exist in
the ownership aspects of Federal and State lands under
the leasing of t'lese lands for geothermal energy develop-
ment. Land utilization for geothermal resources also
comes under the jurisdiction of local. governments,
except for resources on State or Federal lands.
Other institutional questions at the State level in-
clude the acre,lge level for commercial development, the
need for long-range financial and land use planning, and
the overlappinq of State regulatory agencies with each
other and wit1 jurisdictions of local governments for
. permits, licenslis, taxation, and especially environmental
control. The latter may be affected at the Federal, State,
regional, county, or city government levels. For exam-
ple, in some areas, authority may be divided betWeen
such agencies as a Regional Land Development Commis-
sion and a County Air Pollution Control Board.
The institutional aspects of licensing and regulation
of power plants is very complicated; they cover the
spectrum from Federal to local jurisdictions. Regulations
already exist with respect to the exploration, drilling,
and operation of water and mineral wells in all States.
The extension to geothermal wells should be relatively
simple. Yet the need to satisfy the provisions of NEPA
and any specific State environmental requirements may
make geothermal resource development a slow process.
For example, in California, the State Lands Commission,
before it can lease any lands under its jurisdictions, must
make a finding at a public meeting that the lease will not
have significant detrimenta~ environmental effect and
must prepare an environmental impact report available.
to the legislature and the public. The corresponding
problems of environmental impact from geothermal
resources in private lands are not yet fully resolved.
Once the field is developed to the point where a
utility contracts to purchase the resource and construct
a power plant, other regulatory agencies come into the
picture, such as the Federal Power Commission and cor-
responding State and local agencies. Site selection and
environmental analysis criteria are becoming of major
importance in power plant licensing for all types of
energy resources and their effect on geothermal energy
development will probably be determined by solution of
these problems On a generic basis, rather than specifi-
cally for geothermal energy alone.
. Institutional problems thus involve many social,
legal, environmental, and economic questions. The prob-
lems become more complex for land use planning when
geothermal resources span Federal, State, and private
lands. They involve capital investment problems for geo-
thermal development which may be considered to be
high-risk and involve long-delay times until they become
income producing. They involve interindustry arrange-
ments when mu Itiutilization or total utilization is need-
ed to support economic development of electric power
generation, thermal power heating, desalination and
mineral recovery. And they involve multigovernment
arrangements in the realms of regulation, licensing, and
environmental control.
NATIONAL GEOTHERMAL PROGRAM
Although significant growth of the one natural
steam field in the United States has occurred since 1960,
it has become apparent that a major national effort of
226
-------
industrial development supported by Federal stimulation
is needed to develop the potential of geothermal re-
sources in its several forms as an alternate energy source.
Early efforts to achieve a coordinated Federal program
for the support of research and development were under-
taken by an informal Interagency Panel for Geothermal
Energy Research. From these efforts evolved a 5-year
program whose objective was the rapid development of a
viable geothermal industry for the utilization of geo-
thermal resources for electric power production and
other products. The goals and plans for this program
were prepared by the Interagency Task Force on Geo-
thermal Energy under direction of the National Science
Foundation in the Federal Energy Administration "Proj-
ect Independence Blueprint" (ref. 9). The task force
evaluated two alternate strategies. The first was "busi-
ness-as-usual," which assumed continuation of current
policies affecting levels of geothermal production. The
second was "accelerated demand:' which assumed
. specific changes that would result in a more rapid expan-
sion of potential production.
The task force estimated that under the "business-
as-usual" assumptions, electric power capacity could
reach 4,000 MWe by 1985 and perhaps 59,000 MWe by
1990. The corresponding numbers for the "accelerated
demand" assumptions were 20,000 to 30,000 MWe by
1985 and 100,000 MWe by 1990. These latter values
were adopted as the primary goal Of a proposed National
Geothermal Energy Research Program, which was dir-
ected toward (1) providing the necessary technological
advances to improve the economics of geothermal power
production, (2) expanding the knowledge of recoverable
resources of geothermal energy, and (3) providing care.
fully researched policy options to assist in resolving envi-
ronmental, legal, and institutional problems.
The major research funding agencies which con-
tributed to the task force program were the Atomic
Energy Commission, the Department of the Interior, and
the National Science foundation which served as lead
Federal Agency. The status of the research carried out
under support from these agencies is described in the
proceedings of a conference on research for the develop-
ment of geothermal energy resources (ref. 10).
During 1974, tWo acts of Congress resulted in a
marked change in direction for the national development
of geothermal energy. The first was PL 93-410, the Geo-
thermal Energy Research, Development, and Demonstra- .
tion Act of 1974, which established a Geothermal
Energy Coordination and Management Project. The
Project was given responsibility for the management and
coordination of a national geothermal development
program which included efforts to: (1) determine and
evaluate the geothermal resources of the United States;
(2) support the necessary research and development for
exploration, extraction, and utilization technologies; (3)
provide demonstration of appropriate technologies; and
(4) organize and implement the loan guarantee program
authorized in Title II of the Act.
The second law was PL 93-438, the Energy Re-
organization Act of 1974, which established the Energy
Research and Development Administration, ERDA, with
responsibility as lead Federal agency for activities related
to R&D of all energy sources. The Act abolished the
AEC and transferred the geothermal development func-
tion of the AEC and NSF to ERDA. On January 19,
1975, ERDA assumed responsibility for the national
program of geothermal energy development. It has also
assumed direction of the Geothermal Energy Coordina-
tion and Management Project, which has completed the
Final Report required by PL 93-410 (ref. 11). In addi-
tion, ERDA, in response to Congressional requirements
and internal needs, prepared a comprehensive R,D&D
plan (ref. 12) for developing energy technology options.
The geothermal section of the plan built upon the prede-
cessor plans of the Task Force for Geothermal Energy
and the Geothermal Project and has based the goal for
the national program on the rational given in volume 2
of the Plan.
The objectives being considered in the ERDA prog-
ram for geothermal energy include methods to stimulate
the industtial development of indigenous hydrothermal
resources to provide the Nation with 10,000 to 15,000
MW of electric power and thermal energy during the
1985 to 1990 period and to develop new and improved
technologies for cost-effective and environmentally-
acceptable utilization of all types of geothermal re-
sources as a long-term alternate source of energy.
The strategy of the program which might accom-
plish such objectives would be to accelerate industrial
development of the Nation's geothermal resources by (1)
coordinating efforts for exploration and assessment of
geothermal resources necessary to establish reserves by
1978-1980 which can support production of 20,000 to
30,000 MW of power, (2) demonstrating near-term a.ld
advanced technologies needed to utilize many types ,)f
geothermal resources in a cost-effective and environ-
mentally acceptable manner, and (3) fostering rapid
development of a viable geothermal industry by appro-
priate incentives, timely reduction of institutional
impediments, and direct participation of the private sec-
tor in development and demonstration of geothermal
energy technology.
Although ERDA assumes overall responsibility for
effective management and coordination of Federal geo-
thermal activities, the scope of the Federal program
includes the efforts of many Federal agencies. The Geo-
thermal Steam Act of 1970 authorized the Department
of the Interior to lease Federal lands for geothermal
227
-------
. .
resource explora~ion, development, and production of
energy and use1ul by products (such as methane, desali-
nated water, and valuable minerals). The leasing program
is conducted by the Bureau of Land Management, which
is responsible for selecting lands for lease and holding
lease sales, and the U.S. Geological Survey, which classi-
fies the lands by appraised value.
The U.S. (Ieological Survey's geothermal research
program is focu ;ed on the characterization and descrip-
tion of the na':ure and extent of the geothermal re-
sources of the United States'. The output of the U.S.
Geological SUrYIiY'S program is the determination of the
magnitude of the geothermal resource base on a national
and regional basis. The Survey's program includes devel-
opment of exp loration technology, methodology for
estimating energv potential of geothermal systems, envi-
ronmental effec1s of geofluid withdrawal, and geochemi-
cal aspects of reservoir permeability.
Some of thl' problems that have retarded the delin-
eation of the Nation's geothermal resources through the
leasing of public lands include the lack of reliable infor-
mation regardin{ suitable resources, even on lands classi-
fied as KG RA';, insufficient requirements for early
exploration and development of leased lands, and legal
problems involvilg ownership and control of use of geo-
thermal resourcl!S. Under a national program, coordi-
nated effort by I:IRDA and the Department of the Inter-
ior would help to accelerate the establishment of geo-
thermal reserves by the resource industries. Potential
actions include m accelerated estimation by the U.S.
Geological SurveV of the available resources by geologic
type, (2) improved technology for resource exploration
and assessment a:,d for reservoir evaluation, (3) easing of
leasing impedimE nts by better methods for designating
KG RA's and establishing minimum acceptable bids, (4)
incentives for early development of leased lands, and (5)
recommendatiom for legislation to resolve legal uncer-
tainties pertaininu to geothermal resources.
The second part of the strategy for the Federal
program would center on ERDA efforts for demonstra-
tion of near-terrn and advanced systems for resource
utilization, deve opment of supporting research and
technology, and I)xecution of the Federal loan guarantee
program. Demcnstrations of utilization technology
could occur as (1) commercial-scale demonstration
plants to provide the public sector with operational
experience with "ull-scale electric power plants capable
of generating enl)rgy at design production cost under.
pertinent environmental and institutional conditions, (2)
pilot-plant facilit:es to prove technical feasibility, pro-
vide preliminary economic data, and provide capability
for testing new and improved extraction and conversion
systems for electric power production, and (3) field test'
facilities to improve reservoir assessment technology,
evaluate reservoir characteristics and performance, test
and evaluate energy extraction and conversion compo-
nents and processes, evaluate material compatibility with
geothermal fluids, and test environmental control tech-
nologies. A supporting research and development pro-
gram could provide for development of hardware sys-
tems, components, processes, and control techniques for
installation in the demonstration facilities and field test-
ing of reservoir evaluation technologies for the range of
resource types. Supporting research and development
programs could also provide advanced research'and tech-
nology to the geothermal industry and its supplier and
support industries for improved productivity and utiliza-
tion.
Implementation of the loan Guarantee program
should be coordinated with the Bureau of land Manage-
ment's geothermal leasing program and ERDA's
research, development, and demonstration program. The
program could involve venture capital companies, reser-
voir developers, and lease holders to maximize the
impact of the loan program in stimulating early develop-
ment of commercial electric and thermal power facili-
ties. The loan Guarantee program might be used pri-
marily for income-producing projects, such as field
development and power-plant construction. Smaller
industrial firms could benefit from guaranteed loans by
gaining access to necessary private capital. Regulations
and procedures governing the implementation of the
loan guarantee program are currently being drafted in
coordination with other Federal agencies. such as the
Small Business Administration and the Economic
Development Administration. Approved regulations and
operating procedures setting forth specific information
requirements to be met by the applicant and criteria
governing the approval process should be widely publi-
cized as early as possible.
The third part of the strategy of the Federal pro-
gram would involve several Federal agencies, notably the
National Science Foundation, involved in assessing envi-
ronmental, legal, and institutional problems of advanced
energy technology under its RANN program, the Envi-
ronmental Protection Agency, involved in environmental
emission standards, monitoring, and control technolo-
gies, and the Federal Energy Administration, involved in
institutional aspects of the national energy situation.
With the mutual efforts of the Federal program and.
228
-------
the geothermal industry, the attainment of the National
electric and thermal energy goals for geothermal re-
sources could add a significant alternate energy source to
the national economy before the end of the present
century.
REFERENCES
1. Proceedings, United Nations Symposium on the
Development and Utilization of Geothermal Re-
sources, Geothermics, Special Issue No.2, 2 vol-
umes, 1970.
2. H. C. H. Armstead, ed., "Geothermal Energy: Re-
view of Research and Development," Earth Science
Series No. 12, UNESCO, Paris, 1973.
3. P. Kruger and C. Otte, eds., "Geothermal Ener-
gy: Resources, Production, Stimulation," Stanford
Un iversity Press, Stanford, 1973.
4. Scientific Council for Geothermal Energy, "Geo-
thermal Investigation and Utilization of the Heat of
the Earth," U.S.S.R. Academy of Sciences, Nauka,
Moscow, 1966.
5. Scientific Council for Geothermal Energy, "Study
and Utilization of the Deep Heat of the Earth,"
U.S.S.R. Academy of Sciences, Nauka, Moscow,
1973.
6. Informatics, Inc., "Recent Soviet Investigations in
Geothermy," ARPA-1622-3, May 1972, and "Soviet
Geothermal Electric Power Engineering," ARPA-
1622-3, December 1972.
7. Proceedings, Workshop on Environmental Aspects
of Geothermal Resources Development, National
Science Foundation Report No. AER 75-06872,
1974.
8. Proceedings, Conference on Geothermal Energy and
the Law, National Science Foundation Report No.
NSF-RA-S-75-003, 1975. ..
9. Federal Energy Administration, Project Independ-
ence Blueprint, Final Task Force Report, Geo-
thermal Energy, November 1974.
10. Proceedings, Conference on Research for the Devel-
opment of Geothermal Energy Resources, National
Science Foundation Report No. NSR-RA-N-74-159,
1974.
11. Energy Research and Development Administration,
Definition Report: Geothermal Energy Research,
Development and Demonstration Program, ERDA-
86, October 1975.
12. Energy Research and Development Administration,
A National Plan for Energy Research, Development
and Demonstration: Creating Energy Choices for
the Future, ERDA-48, 2 volumes, June 1975.
229
-------
CONSERVATION AND ENVIRONMENTAL IMPLICATIONS
OF OPEN-CYCLE MHD*
Abstract
Finn A. Haist
I NTROD UCTION
It is widell recognized that the development of
MHO power gt'neration would enable a much more
effective utilizarion of our fuel resources, in particular,
our vast coal res1rves with drastic reductions in pol/ution
of air and water, It would serve the following important
national goals:
1. More effective coal utilization,
2. Conservation of aI/ types of fuels,
3. Reduced pollution of air and water,
4. Signific ~nt economic savings, and
5. ImprovI?d reliability of electric utility power
systems.
The important potential advantages of MHO relative to
existing technology are possible because of the inherent
characteristics 0:' the MHO process. This paper discusses
how the importmt goals can be reached and served by
MHO, the status of MHO technology today, and its
future prospects, The discussion centers on coal-burning,
open-<:ycle MHO central station power generation. The
conclusion reach(]d is that the development of MHO can
contribute significantly to solving the serious conserva-
tion and environmental problems associated with elec-
trical power genuation.
.This work n ported is presently supported by ERDA and
EPRI. The MHD development program at AERL, Inc., has pre-
viously been suppcrted by the Interior Department through the
Office of Coal Resnarch, by the A VCO Corp., and by the follow-
ing electric utility I:ompanies and organizations: American Elec-
tric Power Co., I~ow York, N.Y.; Appalachian Power Co.,
Roanoke, Va., (Subsid iary of American Electric Power Co.!;
Baltimore Gas & :Iectric Co., Baltimore, Md.; Boston Edison
Co., Boston, Mass. Central Illinois Light Co., Peoria, III.; Con-
solidated Edison CII. of N.Y., Inc., New York, N.Y.; The Dayton
Power & Light Co., Dayton, Ohio; Edison Electric Institute, New
York, N.Y.; Electri: Power Research Institute, Palo Alto, Calif.;
Illinois Power Co., )ecatur, III.; Indiana & Michigan Electric Co.,
Fort Wayne, Ind. (Subsidiary of American Electric Power Co.!;
Indianapolic Power & Light Co., Indianapolis, Ind.; Kansas City
Power & Light Co., Kansas City, Mo.; Louisville Gas & Electric
Co., Louisville, Ky.; NEGEA Service Corp., Cape & Vineyard
Electric Co., Hyal1nis, Mass.; New England Electric System,
Westborough, Mas!.; Northeast Utilities Service Co., Hartford,
Conn.; Ohio Power Co., Canton, Ohio (Subsidiary of American
Electric Power Co,); Union Electric Co., 5t. Louis, Mo.; The
United IIluminatinE Co., New Haven, Conn.
tPrincipal Re ;earch Engineer, A VCO Everen Research
Laboratory, Inc., E 'erett, Massachusetts.
MHD power generation is now widely recognized as
a promising new method for large-scale electrical power
generation. It promises a substantial improvement in the
thermal efficiency of power plants. This will enable a
much more effective utilization of the fuel burned.
Furthermore, the MHD power generator can use all
types of coal (as well as other types of fossil fuels) with
their inherent ash and sulfur impurities and still drasti-
cally reduce pollution of air and water. It will allow the
burning of coal directly without preprocessing and
assure strict emission' control of all pollutants. This will
make power generation much more compatible with the
environment. MHD also promises significant economic
gains and improved reliability of electric utility power
systems and networks. I n summary, the development of
MHD power generation can greatly contribute to solving
the serious conservation and environmental problems
associated with electrical power generation.
This paper will discuss how the above important
goals can be reached and served by MHD, the status of
MHD technology today, and its future prospects. The
discussion here will be limited to open-cycle MHD power
generation which by far has met with the greatest at-
tention and largest development efforts both in this
country 'and abroad. It will naturally center on coal-
burning MHD power generation since coal represents our
. largest and most important fossil fuel resource.
Since the advantages of the MHD energy conversion
process follows from its inherent characteristics, it is
important to understand the principles of the MHD
process. Therefore, a brief review of this is first present-
ed before the potential application of MHD to practical
electric utility utilization is described along with its con-
servation and environmental implications. This is fol-
lowed by a brief review of the status of MHO technology
and its further necessary development.
MHD PR INCIPLE
The principle of MHD is old and based on Faraday's
well-known laws of electromagnetic induction. The
magnetohydrodynamic (MHD) generator is a heat engine'
which transforms the heat energy of a !J
-------
Contrary to a turbogenerator, the MHO generator
extracts electric energy directly from a hot gas without
the intervention of moving parts. The basic conversion
process is a volume process where the gas itself is made
to conduct electricity and passes at high velocity
. through a magnetic field generating electricity. Of great-
est importance is that this inherent simple direct process
makes it possible to handle high-temperature combus-
tion gases from direct burning of coal and other fossil
fuels containing ash and sulfur impurities.
The basic d.c. combustion generator is schematically
illustrated on figure 1. Hot combustion gases are pro-
duced by burning of the fuel under high pressure and
temperature. A small amount of an alkali salt, called
seed, is added to the gas to enhance the electrical con-
ductivity of the gas. The hot gases expand at high veloci-
ties through the MHO generator channel or duct inserted
in a strong magnetic field produced by a superconduct-
ing magnet surrounding the duct. This induces an EMF
in the gas which drives a current through the gas and
external load, which is connected to electrodes inserted
in opposite walls of the duct. This type is the basic linear
Faraday-type generator. Other generator types with
different electrical load connections and of different
geometry are also possible.
COOLANT ~
OUT :.:
MAGNETIC
FIELD
COOLANT PASSAGE
CHANNEL
REINFORCED
PLASTIC
TV PICAL INSULATOR WALL
CON FIGURATION
82816
From a practical engineering point of view, it is
important to recognize that the walls of the MHO gen-
erator channel assembly as well as the upstream high-
temperature burner are cooled by a coolant such as
water. The reason for this cooling is the same as that for
cooling a combustion engine which operates at the same
high peak gas temperatures as the MHO generator. In
this way, the operating temperatures of the wall mate-
rials are controlled so that ordinary construction mate-
rials can be used. Thus, the high gas temperature does
not represent a materials temperature limitation for the
MHO generator. In fact, any temperature limitation is
rather imposed by the theoretical flame temperature
which can be reached in combustion of the fuel.
The characteristics of the MHO power generator
which make it so well suited for applications to electric
utility power generation can be summarized as:
1. the ability to handle high temperatures,
2. the ability to handle very high power levels,
3. no moving parts or close tolerance, and
4. the ability to start and reach full power prac-
tically instantaneously.
Because of these important characteristics, the MHO
generator has great potential for electric utility power
generation applications.
/
COOl ANT
IN
SUPERSON IC
NOZZLE
FLAME
HOLDER
COMBUSTION CHAMBER
SEED
Figure 1. Basic d.c. combustion MHD generator.
231
-------
APPLICATION TO ELECTRIC UTILITY
F'QWER GENERATION
General
As a heat (mgine, the MHO generator utilizes a
thermodynamic Gycle similar to that of a turbine, The
intrinsic feature af the high-temperature MHO process
for attaining ver', high cycle efficiencies provided from
the early stages in MHO development provided the im-
petus for development of coal-burning central station
baseload power plants, This application has always been
considered the major objective of the development of
MHO. Therefore, the discussion here will center on this
type of application. However, other types of applica-
tions such as intermediate load, peaking, or reserve
power also repre!:ent important potential appl ications of
MHO power generation for electric utility service.
Two basic MHO power plant designs are shown on
figures 2 and 3. I~ binary power cycle is utilized in both
designs where COJI is burned directly with preheated air
and d'.c. power is extracted from the hot seeded combus-
tion gases in the MHO generator. The heat in the MHO
generator exhaust gas is utilized for regeneration and for
production of sUJplemental power in either a bottoming
steam or a gas turbine plant. The important feature of
the latter combir ation is that there is no need for cool-
ing water for cor densation of steam. The significance of
this plant design ft:ature will be reviewed in more detail
later together w th the important gas cleaning and air
pollution control aspects of MHO power plant designs.
However, it sholld be noted at this point that the gas
turbine in a combined MHO power cycle operates with,
air instead of combustion gases as the working fluid,
which avoids tUl'bine erosion and corrosion problems
from use of comlJustion gases as working fluid directly.
Another alternate MHO power plant design which
recently has rece:ved attention is to combine MHO
power generatior, with coal gasification. The MHO gen-
erator offers a lnique opportunity in this respect be-
cause the heat clmtent of the MHO generator exhaust
gas can be utiliZI!d in the gasification process avoiding
fuel efficiency 10 ises otherwise generally associated with
coal ga~ification i., connection with power generation.
Plant Efficiency, Fuel Conservation, and Heat Rejection
It is widely recognized that substantially higher
power plant efficiencies are attainable with MHO as
compared to conventional power technology. Reported
results from severCJI power plant efficiency studies and
investigations are summarized in figure 4 (refs. 1, 16,
17). Important riarameters for determining the effici-
ency of the power cycle are the preheat temperature
which can be obtained of the combustion air and the
magnetic field strength which can be employed, since
both of these two parameters, particularly the preheat
temperature, determine the amount of enthalpy or frac-
tion of fuel energy which can be extracted from the
MHO generator.
The conclusion has been reached that early MHO
power plants based upon present status and knowledge
would attain roughly 50 percent efficiency and that at
least 60 percent efficiency could be expected from
further development. This means that developed MHO
power plants would generate 50 percent more electricity
from the same amount of fuel burned as compared to
today's most efficient power plants. Thus, MHO would
enable power plants to utilize and to conserve our coal
reserves and other fossil fuel resources more effectively.
All power plants must reject heat. but the amount
of heat rejected decreases rapidly with increasing plant
efficiency. This is so because the fraction of the input
fuel heat energy which is converted into electricity is
just the efficiency, 17, whereas the fraction of heat
energy rejected to the environment equals (1 - 1/). Thus,
the total amount of heat rejected per unit of electricity
produced (0) is 0 = 1 - 17117. This relationship is shown
by the curve plotted in figure 5.
Comparative data on heat rejection from different
thermal power plants are listed in table 1.
The heat rejection is divided into two parts. The
first part is specified as the amount of heat rejected to
cooling water for condensation of steam and the second
as the amount of heat rejected directly to the atmos.
phere.
In the case where the MHO generator is combined
with a gas turbine bottoming plant, there is no need for
cooling water for condensation of steam and, thermal
pOllution of water is eliminated. In addition, it should be
recognized that a large fraction of the waste heat in this
case is rejected at relatively high temperatures (400° to
8000 F) in the form of hot combustion gases and hot air
exhausting from the gas turbines. Thus, the high-temper-
ature MHO process which extracts energy from the gas
at temperatures between 3,000° and 5,000° F permits
heat rejection at higher temperatures than other thermal
power plants while still maintaining very high plant effi-
ciency. This offers better prospects for utilization of
waste heat, such as for generation of low-pressure proc-
ess steam, for heating purposes. for desalination, etc.
Air Pollution Control
Another pressing pollution problem facing coal
burning power plants today is emission control of air
pollutants formed during combustion of the fuel. This
important item has received particular attention in the
232
-------
SEED
MAKEUP
RECYCLED
SEED
I'J
W
W
C8793-1
COMB". AIR
STEAM GENERATOR
3S00 ps; 110000F /1000oF
L.T. AIR
HEA TER
A.C.
EL. GEN.
B.F. WATER SYSTEM
r ....-m. u",. ]
COND
PUMP
S' INDICATES SEED & FLY ASH COLLECTED FROM BOILER AND AIR HEATER TO SEED
PROCESSING AND RECYCLING.
Figure 2. MHO - steam power cycle.
HP
ECO
S'
02
J
LP
ECO
S'
02
PLANT
AIR
! .!
STACK
GAS
GAS
CLEANING
RECOVERED
SEED
SULFUR
REGENERATED SEED
FOR RECYCLING
AIR
-------
MHD SUPPL.
POWER POWER AIR AIR
riVE RTER EL.
GEN.
SE~D
BY-PASS AIR
FOR TEMPERING)
COAL
H.T. COMB. AIR
LOW TEMP.
STACK
GAS
RECUPERATOR
Figure 3. MHD - air turbine power cyCle.
development of MHD combustion systems and will be
reviewed here in more detail.
The MHD ,:ombustion systems have two distinct
features which are different from conventional systems.
First, the comb Jstion occurs at much higher tempera-
tures, and secondly, a small amount of an alkali seed is
added to the combustion gases to enhance the electrical
conductivity of the gas. This latter feature necessitates
recovery of the major portion of the seed added for
economic operH~on. Therefore, highly efficient gas
cleaning equipm1mt for recovery of seed is a natural and
integral part of the power plant system. Experimental
work has demonstrated that seed together with any
particulate matt£ r in the gas can be removed at efficien-
cies as high as !J9.9+ percent by conventional electro-
static precipitators, bag filters, or scrubbers. Further-
more, the alkali seed has a high chemical affinity to
sulfur and it hab experimentally been verified that the
alkali seed will combine with sulfur from the fuel so that
sulfur can be eft~ctively removed from the gas together
with the seed (rufs. 2,3,4). This sulfur removal process
requires processing of the recovered seed together with
recycling of prol:essed seed and it produces elemental
sulfur as a valu,!ble byproduct (ref. 2). The method
developed for processing of the recovered seed with sul-
fur removal is sch~matically shown in figure 6. In the
particular process shown on this figure,K2 504 together
with any remaini 19 flyash removed from the gas is treat-
ed with reducing agents such as carbon monoxide and
hydrogen. in ordE r to reduce K2504 to K25. K2504 is
formed in the gas under oxidizing MHD generator ex-
haust gas conditions which have been assumed in this
case. However, the possibility exists for operating with
slightly fuel-rich conditions at the higher exhaust gas
temperatures which will promote the formation of K25
in the gas instead of K2 504, With this mode of opera-
tion, K25 can be directly removed from the gas which
eliminates the need for reducing K2804 to K28 as
shown in figure 6. Complete oxidation of the combus-
tion gas is in this alternate mode of nperation performed
at lower gas temperatures. In any event, the reducing
agents necessary for any reduction of K2504 can readily
be produced directly from coal by partial oxidation.
Coal itself may also be used directly as a reducing agent.
K25 produced in the first reduction step or alternative Iy
removed directly from the gas is then reacted with water
vapor and carbon dioxide to form potassium carbonate
(K2 C03) and hydrogen sulfide (H25). The hydrogen sul-
fide produced is fed to a conventional Claus plant where
elemental sulfur is produced as a byproduct. The K2 C03
from the process is recycled to the MHD burner.
5ince MHD is a high-temperature combustion proc.
ess, particular attention has been directed towards the
control of nitrogen oxides formed at the high combus-
tion temperatures.
There are principally two' routes available for con-
trol of nitrogen oxides emission (ref. 5). One route is to
minimize the nitrogen oxides in tbe gas so that it is
acceptable for direct emission to the atmosphere. The
other is to maximize the oxides of nitrogen in the gas so
234
-------
o
()' 64
I
>-
U 62
Z
W
U 60
lJ..
lJ.. 5 8
W
;i 56
~
ffi 54
J:
.... 5 2
....
Z 50
<0
1500
.
o
S. WAY ET AL. (WESTINGHOUSE)
SHEINDLIN ET AL. (USSR ACADEMY OF SCIENCES, MOSCOW)
6
'f;J
BIENSTOCK ET AL. (US BUREAU OF MINES)
, I
MASSE ET AL. (ELECTRICITE DE FRANCE)
~WSNER (BROWN BOVERI & CO., BADEN)
o
+
ANDRZEJEWSKI ET AL. (IN ST. OF HEAT ENGINEERING, WARSAW)
GEBEL (SIEMENS SCHUCKERTWERKE, ERLANGEN)
*
.
ST ANFORD UNIVERSITY
AVeO
x
(8 TESLA)
.
(5)
o
( 7)
X
X (4)
(6) 0
X
(6)
6
(6)
.
(5)
*
(5) (4)
. "6
0(4)
'f;J
(6)
(5)
o
0(4)
(6)
X
X
-------
8 2.5 r
~
Q:
W
Z
w
C> 20
>-
!:::
u
Q:
.....
~ 15
..J
w
!:::
z
::>
Q: 1.0
w
Q..
o
LtJ
t-
~ 0.5
"">
LtJ
Q:
~
w
J:
TOTAL HEAT REJECTED PER UNIT ELECTRICITY GENERATED:
QTOTAL = U
.,.,
't'J" OVERALL THERMAL EFFICIENCY
o ~___.L_~___I....._--_. L......-...-.., .....L_' '-'-"" ..-.---
. 30 40 ~O 60 70 OVERALL THERMAL EFF, 'n '70,
8)722 .,
Figure 5. Heat rejection.....:. efficiency relationship for
thermal power plants. .
Table 1. Comparison of heat rejection from different
thermal power plants
(heat rejected per kWh generated) ,
----.-.
ThEmna 1 eff. Condenser' Stack }as Total
Plant type (percent) (kWh) (kWh kWh
Fossil-steam 40 1.25 0.25 1.50
Nuclear-steam 32-42 , 2. 12-1 .14 0 2.12-1.14
(Present-advanced)
MHD- s tea'tl 50-60 0.64-0.50 0.18-0.17 0.82-0.67
MHD-air turbine 50-60 0 0.82-0.67 ' 0.82-0.67
236
-------
- -
I
GAS ABSORBER
,---+ H2 S. C02
COOLER
C02. H20. N2 REMOVAL TO CLAUS PLANT
CO. H2 FOR SULFUR
PRODUCTION
. H 2S. C02
H20. CO.
H2' N2
,
REACTOR K 2S ~ REACTOR K2 CO) RECYCLED
~ KT
REDUCED TO -.- CONVERTED TO TO MHD COMBUSTOR
K2S K 2CO)
RECYCLE GAS
K2S04 FRO~
MHD POWER PLANT
REDUCING GAS (CO + H2) PRODUCED
DIRECTLY FROM COAL ON SITE OR
AL TERNA TIVEL Y DELIVERED TO PLANT
AS SYNTHESIS GAS.
"* WASTE GAS
TO BOILER FURNACE
E7716
* WITHOUT N2 IN SYNTHESIS GAS ALL GAS EXITING FROM ABSORBER IS RECYCLED
Figure 6. Seed recovery process with sulfur removal.
that recovery of fixed nitrogen becomes economically
attractive. The main attention has so far in the develop-
ment work been directed towards the first route with
the use of direct reduction as an emission control tech-
nique for nitrogen oxides in the gas.
I n order to attain a very low emission level of nitro-
gen oxides by this first route, a two-stage combustion
technique with afterburning in the MHO generator
exhaust gas is employed. This control technique has
been proposed by several investigators in the MHO field
and its practical application has been experimentally
demonstrated and verified in analytical studies (refs.
6.7 ,8,9,10,11). These studies indicate that it is possible
with this technique solely through direct homogeneous
gas reactions occurring in the gas to reduce; the level of
nitrogen oxides in the stack gas to 100-150 ppm for
coal-burning plants or to about 25 percent of the recent-
ly adopted Environmental Protection Agency emission
rate limitations for such plants. The kinetic history of
nitric oxide throughout the power plant system, when
tWo-stage combustion technique is employed, is ill us-
trated on figure 7. Additional reduction through cataly-
tic effects from refractory materials employed in high-
te mperature air heatE1rs have been repbrted (refs.
6,11,12,13). This will make it possible to !ower NOx
concentrations in the effluent gas further, and to less
than 50 ppm which is less than one-tenth of present EPA
standards. Emission control of nitrogen oxides through
two-stage combustion and direct reduction is presently
considered the basic emission control technique for
nitrogen oxides from MHO power plants.
Another alternate control technique based upon
fixation of the maximum amount of nitric oxide formed
in the high-temperature MHO combustion process for
recovery of fixed nitrogen is also being pursued because
of its important economic potential (ref. 5). This poten-
tial will probably become more and more important ill
the future as new and alternate sources for fertilizer
nitrogen must be sought to replace and to conserve the
limited available resources of natural gas. Natural gas
presently represents the dominant raw material for
production of fertilizer nitrogen through the ammonia
synthesis process.
The mode of operation for this alternate control
techniquewith nitrogen fixation is different from that of
two-stage combustion with initial fuel-rich mixture for
direct reduction of nitric ox ide in the gas. The combus-
tion in the high-temperature MHO burner must instead
occur with a slight amount of excess air followed by a
rapid expansion and cooling of the gas in the MHO gen-
erator in order. to fix the maximum amount of nitric
oxide in the gas. The fixed nitrogen is contemplated
237
-------
--- ---
- - ---- -- ----
----------
.----.-- ._- -
MHD-STEAM POWER PLANT WITH TWO STAGE COMBUSTION
FOR DIRECT REDUCTION AND CONTROL OF NITROGEN OXIDES
BASIS: COMBUSTION OF COAL WITH AIR PREHEATED TO 30000F
2
.Q.
Q.
Z
- 10,000
C/)
ct
(!)
Z
C/)
Z
o
!:i
a:
~
z
LAJ
o
Z
o
o
,
o
z
IDOO
\
\
\
\
\
\
/
\
\
\
\
\
\
\
\
\
FROM EQUILIBRIUM
100
10
DI371
L.T.AIR
STEAM H.T. AIR STEAM HEATER
GEN. HEATER GEN. AND
ECONOMIZER NOX ~ 150PPM
STACK GAS
. KINETIC CALCULATED
NO- CONCENTRATIONS
. THEORETICAL
NO- EQU ILiBRIUM
CONCENTRATIONS
Figure 7. Kinetic history of nitric oxide in coal burning central
. station MHO power plant. .
recovered by set ubbing of the effluent gas before the gas'
is emitted to the Oitmosphere (ref. 5). Preliminary investi- '
gations indicate that such a. recovery process for fixed
nitrogen can be very effective for removal of nitrogen
oxides and also of any sulfur oxides from the effluent
gas. The proces~. has great economic potential because
fixed nitrogen is a valuable fertilizer material. The
---.--------
economic attractiveness of the recovery process is
unique to the high-temperature MHO process, since con-
ventional power plants with their .Iower combustion
temperatures canl10t produce as much fixed nitrogen.
Expected emission factors for MHO power plants of
the three major pollutants produced in combustion of
coal are shown. in table 2. These emission factors are
- - --~_._--- -~ ,---. ,.--...
--- ------ -
238
-------
Comparative emission factors for coal-burning power plants
(Basis for comparison, coal containing 3 percent
sulfur burned with air.)
Table 2.
Existing EPA
Pollutant s teama MHD standard
S02 lbs/106 Btu 4.587 0.045-0.018b 1.2
NOx 1 bS/NO/1'06 Btu' 0.81 0.19-0.065c 0.70
Particulates i bs!l 06 Btu 1.054 d 0.10
0.10-0.01
aEmission factor from Public Health Service Publication No. 999-AP-42.
bBased on experimental demonstration at U.S. Bureau of Mines and Avco
Everett Research Laboratory, Inc.
CBa~ed on experimental demonstration
Everett Research Laboratory, Inc., and
dBased on experimental demonstration
Inc., in U.S.S.R. and England.
compared to current EPA emission standards and to
reported emission rates from conventional steam power
plants. .
Economics
Oesign studies with economic estimates of the
capital costs of central station baseload power plants and
costs of generating electricity in such plants have been
conducted in order to assess the economic potential of
MHO (ref. 14). It is recognized that such cost estimates
can only be considered reasonable projections and that
actual economics can only be established after construc.
tion and operating experience with large commercial
installations have been obtained.
Estimated costs for generating electricity in coal
burning MHO power plants are presented in table 3.
The cost estimates of MHO power plants are com-
pared with those of generating electricity in conven-
tional coal-fired steam power plants. Estimated costs
have been added separately for cooling towers and for
S01 removal from the gas for pollution control. For
MHO power plants, costs related to gas cleaning are in-
.at U.S. Bureau of Mines, Aveo
in Japan.
at Aveo Everett Research Laboratory,
eluded in the basic plant estimated costs. Since early
MHO plants here have been considered combined with
steam power, comparative costs for cooling towers for
this portion of the plant have been included. For fully
developed and advanced power systems the combined
use of a gas turbine is considered with no need for cool-
ing towers. The economic attractiveness of MHO i~ clear-
ly indicated. Actual savings can of course only be cor'-
jectured at this time. Fuel savings alone of $11 billion
before year 2,000 have been estimated in a report pub.
lished by the Office of Science and Technology in 1969
based upon 25 cents per thousand Btu (ref. 15). Present
coal costs would triple this figure and further rising fuel
costs and possibly more extensive use. of MHO would
increase savings further.
STATUS OF MHO TECHNOLOGY
The development of MHO technology has spanned a
period of 15 years and was pioneered in the United
States. The major effort is now occurring in the
U.S.S.R., but the U.S. effort has expanded in recent
239
-------
Table 3. Comparative estimated costs of coal-burning
1,OOO-MW power plants
(Estimate based on 1973 U.S. dollars.)
Conventional MHD
Steam Early Developed
Efficiency, percent 40 50 60
Basic plant cost, $/kW 250 235 170
Basic generating cost - Mill s/kWh
Capital charges (16 percent, 80 per-
cent C.F,) 5.7.1 5.37 3.89
Fuel (3S~:/MBtu) 2.99 2.39 1. 99
Operation and maintenance 0.30 0.30 0.30
Seed 0.08 0.05
Basic generating cost 9.00 8..14 6.23
Pollutio~ control cost - Mills/kWh
Dry cooling towers 1.25 0.70a Db
Sulfur removal 2.00d O;50c O.SOc
Total add. generating cost 3.25 1.20 0.50
aMHD - steam cycle, cost prop. to steam plant.
bMHD - gas turbine cycle, no steam condensation.
cSulfur removal - seed regeneration cost.
dS02 - removal cost, Federal Interagency SOCTAP Report 1973.
years with the trowing concern and recognition of the
importance of improving fossil fuel technology.
The development work has centered on the MHO
generator itself but major efforts have also involved
important plant components and subsystems such as
. .
superconductin!l magnets, combustion systems, air
heaters, seeding and seed recovery systems, and methods
for pollution control of sulfur and nitrogen oxides. For a
detailed description of this technical development work,
reference is made to numerous papers on the different
subjects presented and reported on at National and
International MHO Symposia (refs. 16,17).
The study of the MHO generator itself has proceed-
ed along two complementary paths as illustrated by
240
-------
~figure 8. The first has included construction and opera-
~ion of relatively large experimental generators designed
for short term operation. Parallel to this, studies have
been conducted with experimental generators of lower
power levels to establish the integrity and performance
of electrodes, insulators, and channel mechanical design
features for continuous long-term operation.
I n the United States two large short-duration experi-
mental MH D generators were built by A VCO and suc-
cessfully operated in the middle sixties. One of these
generators called the Mark V produced a gross power
output of 32 MW. The other generator called the
100,000
LORHO generator and located at the Arnold Engineer-
ing Development Center, Tullahoma, Tennessee, deliver-
ed an electrical output of 18 MW.
In addition to these larger short-duration experi-
mental generators considerable work has been conducted
with several fossil-fueled generators at much lower
power levels. This work has mainly been carried out at
AVCO Everett Research Laboratory, Inc., University of
Tennessee, and Stanford University. It has provided valu-
able data and information for the engineering design and
practical long-term operation of MHD generators under
conditions appropriate to central station power plant
BASE LOAD
PEAKING
EXPERIMENTAL
PILOT PLANT
c9~25
'--MK :'iZ: \
o
10,000
(J)
t-
~
~ 1000
o
...J
:.:
100
o
UTSI
10
cf-MK~b
c:=J U-02
JUNE 1967
1
SEC
10
SEC
10
MIN
I I
HR DAY
OPERATING TIME
I 3
MO MO
1
YEAR
30
YEARS
I
MIN
E610~-1
LEGEND
o SINGLE RUN PERFORMANCE
o CUMULATIVE PERFORMANCE
Figure 8. Performance in MHO generator development.
241
-------
applications. Operation with experimental MHO gen-
erators at AERL and UTSI has demonstrated that MHO
generators perform successfully with coal combustion
products containing ash and can utilize coal slag as an
effective and -protective lining of the channel walls.
Figures 9 and 10 show the slag coating of the channel
or.~
~
,-
t
1:1
~
.-..-...
and combustor walls obtained in the experimental opera-
tion at AE R L. Work is now underway at AE R L aimed at
extending long-term channel operating experience with
coal ash up to power levels of roughly 500 kW.
Different experimental MHO generators which have
been designed, built, and operated during the MHO
«it:
-1IiII"'I111...,j{iII""
2
IilliiJilMII ..
~ w
1!1!
~ .ot'1l
;..;...
\~
"I
.~
tI
a
. ::~,
u
'::-..8-
Figure 9. Slag deposit on MK VI diagonal channel wall.
242
-------
i'.J
.j>.
w
.~ "8--
, <-
-= lJ-~',
.
I
I
..
fl
'>
\
\
,~
"
6h'
ii' ",
'it: >ii
Figure 10. Slag deposit on MK VI combustor wall.
-------
develop me ,t program at A VCO Everett Research Labor-
atory, Inc.. are shown in figures 11 and 12. Significant
milestones achieved are indicated for the different exper-
imental gerlerators.
The N HD development work in the United States
has been f Jnded both by private industry and govern-
ment. The Nork is being carried out principally at AVCO
Everett Fresearch Laboratory, Inc., MIT, Stanford
University, the University of Tennessee, Westinghouse,
GE, STD ~orporation, the Bureau of Mines, and the
Arnold En!lilleering Development Center.
By far the most impressive MHD effort today exists
in the Sovint Union, with the strong support of the State
Committee on Science and Technology and the joint
sponsorship of the Academy of Sciences and the Minis-
try of Electrification. The most significant accomplish-
ment to da~e has been the design, construction, and
operation of a natural gas-fired pilot plant designed for
an ultimate MHD generator output of 25 MW. It con-
stitutes the first MHD pilot plant in the world feeding
ganerated "ower directly into the Moscow electrical net-
work SYStE m. Basic development for this plant was
carried out in the smaller U-02 installation with 5-MW
(thermal) capacity located in the center of Moscow. It is
an integratld system experimental facility where devel-
opment of the different elements of an MHD system
continues. Officials of the U.S.S.R. have recently
announced that design work now is underway on a large
commercial demonstration MHD steam power plant
incorporatirlg an MHO generator of about 700-MW elec-
trical outpJt scheduled, for operation in the early
eighties. A I:ooperative MHD program has been initiated
between th~ United States and the U.S.S.R. which is
being coordinated by ERDA.
A serio JS effort in MHD development has also been
underway in Japan for many years which incorporates a
smaller scalI: integrated MHD power system test facility
and a largl!r experimental MHD generator of 1,000
kW-capacity. This larger experimental MHD generator
has been operated with a superconducting magnet which
produces a field strength of about 5 T.
The CUI rent MHO development work in the United
States is bliing expedited and specifically directed to
support the national objective of achieving a commercial
MHD powe' demonstration in the late 1980's. A pro-
gram plan te> this aim is outlined in reference 18. The
most critical goals of this program are identified 'as:
1. MHD generator demonstration,
2. MHD system demonstration, and
3. Co llmercial MHD combined cycle electric
power gener.i\i,on. .
SUMMARY AND CONCLUSIONS
It is widely recognized that the development of
MHD power generation would enable a much more
effective utilization of our fuel resources, in particular,
our vast coal reserves, with drastic reductions in pollu-
tion of air and water. It would serve the following
important national goals:
1. More effective coal util ization,
2. Conservation of all types of fuels,
3. Reduced pollution of air and water,
4. Significant economic savings, and
5. Improved reliability of electric utility power
systems.
The important potential advantages of MHO relative
to existing technology are possible because of the inher-
ent characteristics of the MHD process. It is a volume
process which converts heat energy directly into elec-
tricity. The mechanical simplicity of the MHD generator
with no moving parts makes it practical to handle high-
temperature gases produced from combustion of coal or
other fuels with their inherent ash and sulfur impurities.
Because of its high temperature capability, a much more
efficient use of the fuel is possible. Prospects are that
developed MHD power plants can attain 60 percent
efficiency or even higher. Thus, 50 percent more elec-
tricity can be produced from the same amount of fuel as
compared to the most efficient conventional steam
power plants in use today.
The much better fuel utilization results directly in
less rejection of waste heat so that thermal pollution can
be drastically reduced. Detailed analysis of MHO power
cycles has resulted in the development of MHD power
plant designs which can eliminate thermal pollution of
water and reduce air pollution to far below current EPA
emission standards. Because MHD power plants can be
designed to operate without condenser cooling water
they can also be located in the western part of 1'he
country where coal is most abundant.
Experimental and analytical work by several MHO
investigators in this country and abroad has shown that
the emission of nitrogen oxides under realistic MHO
power systems operating conditions can be reduced to
very low levels and to only a smaller fraction of current
EPA standards.
Experimental investigations have also verified that
seed, together with any flyash impurities, can be re-
moved from stack gas at very high efficiencies with con-
ventional gas cleaning equipment resulting in strict con-
trol of particulate emission. Furthermore, experimental
work at the Bureau of Mines has shown that sulfur
244
-------
.
.
g ,
f
Figure 11.
Two larger short-duration experimental MHD generators:
MK V--1965, 32,000 kW, 1 minute, (AERL); LORHO--
1966,18,000 kW, 1 minute, (built and operated by AERL
for USAF at AEDC).
245
-------
!
1
r
,
tt
Figure 12.
Long-duration experimental MHD generator at AERL MK VI - 1975,
120,000 kWh at an average power level of 300 kW.
246
-------
oxides can be effectively removed from MHO generator
exhaust gas together with recovery of seed and that
elemental sulfur can be produced as a byproduct. This
permits the .combustion of coals having high sulfur con-
tents in MHO power plants.
The conservation and environmental benefits offer-
ed by MHO with better fuel utilization, elimination of
thermal pollution of water, and with reduction of air
pollutants to far below EPA standards are perhaps the
most attractive feature of MHO.
Its other major attraction is the projected favorable
economics which has been clearly shown in many plant
design and cost studies. This follows from a better util-
ization of the fuel and also from increased utilization of
invested capital. Fuel savings alone could amount to
several billions of dollars before the end of this century.
During the past 15 years, impressive research and
development efforts have been conducted towards the
development of coal-burning MHO power plants. Several
experimental MHO power generators have been oper-
ated. Power outputs as high as 32,000 kW have been
obtained from experimental MHO generators designed
for short-term operation. Testing of experimental
generators appropriate to operating conditions of coal-
burning central station power applications have been
conducted at power levels up to roughly 500 kW. The
Mark VI experimental generator at AERL has to date
produced close to 120,000 kWh in long duration testing
at power levels between 200 and 500 kW for a total
accumulated operating time of roughly 400 hours.
At the same time, other components required for
the MHO power system have been developed and
de monstrated experimentally. These include super.
conducting magnets, coal combustion systems, high-
temperature air heaters, and other subsystems. Through-
out the development program, MHO power system
design studies and analyses have been conducted in order
to assess the technical and economic feasibility of MHO,
its environmental effects, and guide the direction of the
developmental work.
Further developmental work is necessary for com-
mercial realization of MHO power generation. Such
work should take full advantage of past accomplish.
ments. A future MHO program has been proposed dir-
ected towards specific component and systems develop-
ment problems. This incorporates a new Engineering
Test Facility (ETF) to perform full MHO system tests.
This is then followed by an MHO combined cycle com.
mercial demonstration power plant.
To conclude, the development of MHO offers
unique opportunities in helping to satisfy and balance
society's need and demand for electric energy, conserva-
tion, and environmental quality.
REFERENCES
1. F. A. Hals and W. D. Jackson, "Systems Analysis of
Central Station MHO Power Plants," presented at
the Fifth International MHO Symposium, Munich,
April 1971. See also references listed in this paper,
and proceedings from all five International Aspects
of MHO in the United States (refs. 16,17).
2. D. Bienstock et aI., "Air Pollution Aspects of MHO
Power Generation," 13th Symposium on Engineer-
ing Aspects of MHO, Stanford University, March
1973.
3. H. F. Feldman et aI., "Kinetics of Recovering Sulfur
from Spent Seed in a Magnetohydrodynamic (MHO)
Power Plant," Env. Sc. & Techn., Vol. 4, No.6
( 1970).
4. F. A. Hals, Symposium on MHO Electric Power
Generation, Rapporteur, Auxiliary Equipment Sec-
tion, Warsaw, Poland, July 1968.
5. F. A. Hals and W. D. Jackson, "MHO Power Genera-
tion-Economic and Environmental Implications,"
Proceedings Tenth Symposium on Engineering
Aspects of MHO, M.I.T., March 26-29, 1969.
6. F. A. Hals et aI., "Development and Design of High
. Temperature Air Preheaters and Techniques for
Emission Control of Nitrogen Oxides for Open
Cycle MHO Power Systems," Proceedings of Sixth
International Conference on MHO Electrical Power
Generation, Wash., D.C., June 9-13, 1975.
7. F. A. Hats and F. P. Lewis, "Control Techniques for
Nitrogen Oxides in MHO Power Plants," ASME
Paper 72-WA/ENER-5.
8. F. A. Hals et aI., "Progress in Development of Aux il-
iary MHO Power Plant Components at A VCO Ever-
ett Research Lab., Inc.," ASME Paper 74-WAI
ENER-6.
9. D. Bienstock, K. J. Demski, and J. J. Demetu,
"Environmental Aspects of MHO Power Gener:J-
tion," Proceedings of the 1971 Intersociety Energl'
Conversion Engineering Conference, U.S. Bureau of
Mines Conference.
10. J. W. Pepper, R. H. Eustis, and C. H. Kruger, "NO
Concentrations in MHO Steam Power Plant Sys-
tems," Proceedings of Twelfth Symposium on
Engineering Aspects of MHO, Argonne National
Laboratory, Argone, III., March 1972.
11. Y. Mori, K. Ohtake, and T. Taira, "Kinetic Study of
Oxides of Nitrogen in MHO Power Plants," Proceed-
ings of Twelfth Symposium on Engineering Aspects
of MHO, Argonne National Laboratory, Argonne,
III., March 1972.
12. Y. Mori, K. Ohtake, and S. Irobe, "Decomposition
247
-------
of N-O on J\lumina Surface Under Reducing Condi-
tions for Emission Control of N-O in MHD Power
Plants," Proceedings of Sixth International Confer-
ence on MHO Electrical Power Generation, Wash.,
D.C., Vol. 1, June 1975, p. 61.
13. Y. Mori, K. Ohtake, and K. Nakamura, "Prediction
of N-O Concentrations in MHD Power Plants by
Axisymmetr ic Two Dimensional Analyses," Pro-
ceedings of Sixth International Conference on MHO
Electrical Power Generation, Wash., D.C., Vol. 1,
June 1975, p. 77.
14. F. A. Hals ,md R. E. Gannon, "Development and
Design Cha'acteristics of Auxiliary MHD Power
Plant Components," ASME Paper 73-WA/ENER-10,
presented a,: the ASME Winter Annual Meeting,
Detroit, Mich., November 1973.
15. MHD for Ce.'Jtral Station Power Generation: A Plan
for Action, prepared for the Office of Science and
Technology :JV Panel on MHD, June 1969.
16. National Symposia on Engineering Aspects of Mag-
netohydrodl'namics (MHD); 2d, Philadelphia, Pa.,
March 9-1(1, 1961, N.Y., Columbia University,
1962; 3d, University of Rochester, Rochester, N.Y.,
March 28-3'J, 1962; 4th, University of California,
Berkeley, JI.pril 1D-11, 1963; 5th, M.I.T., Cam-
bridge, Mas!;., April 1-2, 1964; 6th, University of
Pittsburgh, Pa., April 21-22, 1965; 7th, Princeton
University, Princeton, N.J., March 3D-April 1, 1966;
8th, Stanford University, Stanford, Calif., March
28-30, 1967; 9th, University of Tennessee Space
Institute, Tullahoma, Tenn., April 3-5, 1968; 10th,
M.I.T., Cambridge, Mass., March 26-28, 1969; 11th,
California Institute of Technology, Pasadena, Calif.,
March 24-26; 1970; 12th, Argonne National Labora-
tory, Argonne, ilL, March 27-29, 1972; 13th, Stan-
ford University,. March 26-28, 1973; and 14th,
University of Tennessee Space Institute, April 8-10,
1974.
17. International Symposia on MHO Electrical Power
Generation; 1st, King's College, University of Dur-
ham, Newcastle upon Tyne, England, September
6-8, 1962; 2d, Paris, France, July 6-11, 1964; 3d,
Salzburg, Austria, July 4-8, 1966; 4th, Warsaw,
Poland, July 24-30, 1968; 5th, Munich, Germany,
April 19-23, 1971; and 6th, Washington, D.C., June
9-13, 1975.
18. W. D. Jackson and P. S. Zygielbaum, "Open Cycle
MH D Power Generation. Status and Engineering
Development Approach," presented at the Ameri-
can Power Conference, Chicago, ilL, April 21-23,
1975.
248
-------
ENERGY CONSERVATION THROUGH THE USE OF
COMBINED-CYCLE POWER SYSTEMS
Albert J. Giramonti and William A. Blecher*
Abstract
Analytical studies are being conducted to identify
commercially feasible advanced technology combined-
cycle power systems which would reduce consumption
of fossil fuels, reduce utility-caused atmospheric and
thermal water pollution, arid produce low-cost electric-
ity. Technological advances are described which could
lead to gas turbines operating above 3,OOrf F and result
in combined-cycle efficiencies approaching 55 percent
(HHV) on clean gas or petroleum-based fuels. Because of
the high cost and questionable availability of these fuels,
the potential of combined-cycle systems using coal-
derived fuels was established. One of the most attractive
methods of utilizing coal would be to gasify it, remove
the various pollutants, and burn the clean gas in a
combined-cycle plant Careful integration of the gasifier,
cleanup, and power systems could result in coal burning
power stations using presently available technology
which would produce power at costs competitive with
conventional steam plants using stack gas cleanup. These
systems would also have comparable efficiency, reduced
thermal pollution, and lower emissions of S02, NOx,
and particulates. Systems utilizing advanced technology
gas turbines would have even more attractive power
costs, higher efficiency, and more favorable pollution
characteristics.
INTRODUCTION
. Uncertainties in the availability and cost of future
fossil fuel supplies have created an urgent need for more
efficient energy conversion machinery to best utilize
available fossil fuels. It is expected that more economical
use of fossil fuels than is now possible in electric utility
steam stations will be provided by future utilization of
combined-cycle power plants incorporating open-cycle
gas turbines and waste heat recovery steam systems. The
primary reason for this prospect is that gas turbines will
operate at substantially higher peak cycle temperatures
than steam stations.
Fundamental technological and economic obstacles
prevent the utilization of steam conditions beyond the
2,400 to 3,500 psig, 1,000° F /1,000° F level, thus limit-
ing steam station efficiencies to the 33-39 percent range
. Albert J. Giramonti and William A. Blecher are Senior
Systems Engineers with United Technologies Center, East Hart-
ford, Connecticut.
depending on the stack gas cleanup and heat rejection
procedures used. Conversely, industrial gas turbines with
firing temperatures of 3,000° F are now envisioned as
possible with sophisticated cooling schemes to insure
component durability. With such firing temperatures,
combined-cycle efficiencies approaching 55 percent to-
gether with high specific work output, are possible. The
heat recovery-type steam system functions to convert
gas turbine exhaust heat into shaft power, thus side-
stepping what would otherwise be the need for exces-
sively high gas turbine compression-expansion ratio to
fully utilize future high firing temperatures.
The high performance potential of combined cycles
is partly negated by the requirement for a clean gaseous
or distillate type fuel. Unfortunately, domestic supplies
of gas and oil are limited, and a serious gap between
domestic supplies and demand for these fuels could per-
sist for the next 20 years.
The energy crunch with its accompanying conserva-
tion measures has reduced somewhat the projected
growth of the electric utility industry. Nonetheless, even
conservative estimates (ref. 1) of the need for electrical
energy in 1980 indicate that about 631 GW will be re-
quired, of which about 400 GW will be fossil fueled. By
1995, it is estimated that 1237 GW will be installed,
approximately 680 GW of which will be fossil fueled.
With all the projected additional installed capacity of
. fossil-fueled power systems, the already difficult prob-
lem of utility-caused pollution could become intolerable.
Two different approaches are open for utility con-
sideration: (1) treatment of the power plant exhaust to
remove pollutants, and (2) treatment of dirty fuels
before combustion to remove harmful pollutants. S011e
stack gas cleanup processes show promise as a means of
meeting the near-term EPA standard for S02 and partIc-
ulate emissions. Currently, however, no commercial-scale
stack gas cleanup system has been demonstrated that
also removes significant amounts of NOx' Fuel pretreat-
ment would permit meeting all environmental regula-
tions, but would also result in a significant increase in
the cost of fuels delivered to the power generating
system and introduce significant energy losses. I n order
to offset some of the increased fuel cost and the coal
conversion inefficiencies, the thermal efficiency of the
electric power generation should be increased as much as
possible by using advanced-cycle systems.
Several feasibility evaluation studies of advanced
combined-cycle power systems incorporating fuel gasi-
fication and cleanup have been conducted by the United'
249
-------
Technologies Research Center. Widespread attention to
the concept of the integrated gasifier/cleanup/power
system was brought to focus by the publication in
December 197D of the United Technologies Research
Center/BurilS & Roe/FMC report entitled
"Technological and Economic Feasibility of Advanced
Power Cycles and Methods of Producing Nonpolluting
Fuels for Utility Power Stations" (ref. 2). A current
study for the E ~ A is putting into perspective the techni-
cal and econo 'nic characteristics of these advanced
power systems based upon the current assessment of
technological dovelopment in order to identify the po-
tential environncntal intrusion of the system (ref. 3)
and to evaluate the merits of high-temperature fuel gas
cleanup. The purpose of this paper is to present prelimi-
nary findings 01 the current EPA program and to identi-
fy the economic, environmental, and performance bene-
fits that might be attained through the use of advanced
combined-cycle power systems operating on distillate oil
and coal-derived low Btu gas.
ADVANCED COMBINED-CYCLE POWER SYSTEMS
One of the most promising advanced-cycle power
systems consist! of an open-cycle gas turbine used in
combination with a heat-recovery steam system. Such a
COmbined Gas And Steam (COGAS) power system is
BURNER
depicted in its simplest form in figure 1. A variety of
combined-cycle configurations have been investigated,
including supercharged and exhaust-fired, but the un.
fired waste heat recovery configuration appears to offer
higher performance at lower cost than others. Unlike
some present-day combined cycles in which the gas tur-
bilJles are essentially air pre heaters for the steam boiler,
advanced cycles would utilize large industrial gas tur-
bines operating at high turbine inlet temperatures. The
technology basis for these gas turbines represents spinoff
from the aircraft gas turbine industry. These gas turbines
would produce approximately 60 percent of the net sta-
tion electric output, and their exhaust gases would be
directed into waste-heat boilers which would generate
steam for a steam turbine system producing the remain-
ing output.
Advanced Technology Requirements
Future industrial gas turbines used in combined
cycles could be developed incorporating advanced aero-
space technology in substantially improved large-
capacity engines with appreciably higher thermal effi-
ciency. While meaningful improvements in aerodynamic
performance are projected for future gas turbines, the
key to high efficiency is raising the turbine inlet temper-
ature. The newer gas turbines currently being used in
utility applications operate at temperatures around
1,900° F (see figure 2). Turbine inlet temperatures are
POWER
TURBINE
ELECTRIC
GENERATOR
COMPRESSOR
. TURBINE
STEAM
BOILER
------
TO
STACK
ELECTRIC
GENERATOR
PUMP f
'-~-{CONDENSER .
Figure 1. Combined-cycle power systems.
250
-------
u..
. 3600
I
W
a:
:) 3200
t-
e(
a: 2800
w
Q.
:!
w 2400
t-
t-
W 2000
..J
Z
-
w 1600
z
-
m
a: 1200
:)
t- 1950
R & D ENGINES
MILIT ARY
AIRCRAFT
COMMERCIAL
TRANSPORTS
INDUSTRIAL 11-
APPLICA TIONSI
1958
1966 1974
YEAR
1982
1990
Figure 2. Estimated turbine inlet temperature progression.
expected to exceed 2,400° F by 1980 and 3,000° F by
1990. These increased temperatures will be a direct re-
sult of improvements in materials and blade cooling
techniques. Future blade materials include advanced
nickel, cobalt, and eutectic alloys. Recently, consider-
able attention has been given to ceramic materials such
as silicon nitride and silicon carbide.
The most significant gains in turbine inlet tempera-
ture have been made possible by the application of vari-
ous cooling techniques to the hot parts. Figure 3 summa-
rizes some of the advanced air and water cooling config-
urations. Turbine cooling can be accomplished with
coolants such as air, water, or liquid metals; but because
of the complex cooling system designs and mechanical
problems a'ssociated with liquid systems, air has been
used exclusively in commercial engines. Turbine cooling
schemes have progressed from simple convective cooling
configurations incorporating cast, round, radial passages
for vanes and single cavities for blades to advanced de-
signs utilizing impingement, shower head, film, and
transpiration concepts in several combinations. Ad-
vanced impingement-convection designs provide im-
proved cooling of the inside surface of the leading edge
and could be considered for operation up to 2,200° F.
Complex shower head-film designs have been developed
in which a layer of coolant is injected through slots and
holes to form an insulating air blanket which protects
the outside surface of the blade. These designs should be
useful up to 2,600° F. By using a multipass thermo-
syphon blade with supercritical water as the coolant,
temperatures in the 3,000° F range should be feasible.
Performance Potential
Performance estimates for future high-temperature
gas turbines incorporating conventional impingement-
convection cooling are presented in figure 4. Because of
the high performance penalty associated with the tur-
bine cooling air requirements, gas turbine efficiency does
not increase appreciably with increased turbine inlet
temperature. However, the specific power (net power
per unit air flow) does increase significantly. Specific
power is a measure of the amount of work which can be
done by a given size machine and is, indirectly, a guide
to cost.
A secondary advantage of higher turbine inlet tem-
perature is higher exhaust temperature resulting in
higher performance for the steam portion of the com-
bined cycle. For waste-heat recovery systems, the most
251
-------
ImpingHment - convection
Shower head-film
Water thermosyphon
2200°F
2600"F
3000"F
Figure 3. Advanced turbine cooling configurations.
attractive steam system would have a relatively simple
cycle configurction. Regenerative feedwater heating
would be limite::l to a single deaerating feedwater heater
so that essentially all the steam would pass through the
turbogenerator ilnd be condensed.
The potent ally high performance of the combined
cycle is one of i::s most attractive attributes. Referring to
a system using conventional distillate-type fuels and con.
ventional air cooling techniques, performance levels as
shown in figure 5 should be possible. At 2.2000 F tur-
bine inlet templ!rature, a level which should be attain-
able in baseload turbines in the late 1970's, an efficiency
of 43.5 percent could be realized. As turbine inlet tem.
peratures increase, efficiencies also increase, and it
should be possit Ie to approach 48 percent at inlet tem-
peratures of approximately 2,9000 F. The performance
values presented in figures 4 and 5 are based on u~'.! of
conventional sh)wer head.film cooling using air bled
from the compmssor and sent directly to the part, either
static (vane) or mtating (blade), to be cooled. Also, max-
imum metal temperatures are limited to 1,5000 F.
With use of other cooling techniques, system effi.
ciency could be increased to about 55 percent. This is
depicted in figure 6, where system efficiency changes
with various cooling schemes are presented.
By the simple expedient of cooling the compressor
bleed air to !bout 3Q~J: (fr~m_perhaps 8000 F at com-
pressor discharge) before using it for cooling, a lesser
amount of air would be required and an efficiency
increase would result. If the static parts were cooled by
water or made from some temperature-resistant material
such as ceramics, the amount of air bleed would be
further reduced to that required for rotating parts alone
and the efficiency would increase'. If both static and
rotating parts were water cooled, only air required for
buffering and disc cooling would be used and further
efficiency gains could be expected, the exact value being
a function of type of water cooling selected. Finally, if
the static and rotating parts were ceramic, no heat would
be removed from the gas path by cooling and the largest
efficiency gain, to almost 55 percent, could be realized.
While ceramics offer much incentive, their develop-
ment in large sizes poses problems and it appears that
improved cooling technology offers a more effective
near.term solution. Advanced cooling methods are not
, without technical difficulties; e.g., water cooling could
252
-------
40
-
>
:E:
:E:
-
~ 35
>
(.)
Z
w
-
CJ
-
u.. 30
u..
w
w
Z
-
£n
a:
:> 25
t-
fn
~
0
,
--
-
"
COMPRESSOR
PRESSURE RATIO
32
--
16
2200 2400 2600
. .
TURBINE INLET TEMPERATURE, of
20
80 100 120 140
NET POWER PER UNIT AIRFLOW
160 180
- KW/LB/SEC
Figure 4. Gas turbine performance (Distillate fuel oil,
conventional air cooling).
require treatment of water beyond the level typically
associated with power stations.
INTEGRATED COMBINED-CYCLE/
GASIFICATION SYSTEMS
Combined-cycle power systems, as with other ad-
vanced technology gas-turbine-based power systems, re-
quire fuels of utmost cleanliness. The level of this clean-
liness is usually well beyond that required by env'iron-
mental regulations. When considering the use of coal as
fuel for such power systems, it is apparent that consider-
able processing and cleanup must precede its introduc-
tion into the engine. Low-Btu fuel gas, produced by
gasification and desulfurization of high-sulfur coal,
appears to be very attractive for use in these power
systems. The following paragraphs contain brief descrip-
tions of selected advanced power and fuel conversion
processes which could be considered for use in inte-
grated combined-cycle/gasification power systems, para-
metric performance estimates for future systems, a
253
-------
"""
>
.s::
.s::
'-oJ
~
I
CI)
o
c
as
8
....
~
Q)
Co
C
o
.-
ofJ
as
ofJ
rn
rn
as
C)
o
()
50
Distillate. fuel oil
Conventional air cooling
5')
~-
Compressor pressure ratio
--
/./ --
~/.,// .//
/..'~ ./
'2200-2400
4U
44
4CI
1100
150
200
24
8
\2900
2600
Turbine inlet
temperature,OF
250
300
350
Net power per unit airflow - kw / Ib /sec
Figure 5. COGAS station performance.
summary of environmental pollutant emissions, and
preliminary ecor.omic comparisons.
Integrated System DeSl:riptions
During recent investigations of integrated
combined-cycle/)asification systems (ref. 3), four com-
binations of coa gasifiers and fuel gas cleanup processes
were examined. They are: (1) Bureau of Mines stirred-
bed gasifier followed by Selexol low-temperature clean-
up, (2) Bureau of Mines gasifier followed by Sintered
Iron Oxide hot gas cleanup, (3) BCR two-stage entrained
flow gasifier followed by Selexol cleanup, and (4) BCR
gasifier followeci by Conoco hot cleanup. Selection of
these processes was made to permit a meaningful com-
parison of hot ar d cold cleanup processes in conjunction
with realistic ga iification systems and is not meant to
imply that these gasification and cleanup processes are
superior to other prospective processes.
Those systems using the Bureau of Mines gasifier are
considered first qeneration because they have the poten-
tial for commercial operation during the early 1980's.
Those systems using the BCR gasifier are considered
second generation because they have the potential for
commercial operation by the late 1980's. In a similar
manner, the gas turbine technology judged to be repre-
sentative of engines used in first generation systems is
conventional shower head-film cooling at a turbine inlet
temperature of 2,200° F and a pressure ratio of 16: 1.
For second-generation systems, the use of ceramic vanes
with conventional blade cooling at a turbine inlet tem-
perature of 2,600° F and a pressure ratio of 24: 1 was
selected.
The Bureau of Mines/Selexol integrated system is
shown schematically in figure 7. The gasifier is supplied
with air and steam from the combined-cycle system.
Compressor bieed air is used to generate process steam
prior to entering a boost compressor and saturated steam
is extracted from the high-pressure section of the waste-
heat boiler. Because of the tar content, the gasifier exit
stream is quenched and the tars separated and returned
to the gasifier. The gas leaving the water quench is satu.
rated with water vapor limiting the amount of regenera-
tion of the fuel gas which is possible. However, the
254
-------
56
-
co
e""
mi 52
s;.U
""45
Q)Q.
-
u.
>->- 48
Uu
"Cc
Q)Q)
C.-
.- u
~i 44-
o
40
2000
3200
2200
2400 2600 2800 3000
Turbine inlet temperature .oF
Figure 6. Combined gas turbine and steam station
performance (A/C = air cooled; W IC =
water cooled; C = ceramics).
latent heat of the water vapor is quite significant and is
used in the reboiler of the sour water stripper. After
regeneration with the product gas, the remaining heat is
rejected to ambient to lower the temperature prior to a
water scrub and Selexol cleanup tower. Condensate from
the 5elexol stripper and the gas cooling process along
with water from the scrubber must be processed in the
sour water stripper. In the ammonia recovery process,
H25 absorbed in the water is recovered and sent to the
C1aus plant. After regeneration, the clean fuel gas is fed
to the combined cycle which functions normally except
that steam is extracted from the high-pressure section of
the boiler for use in the gasifier and ammonia removal
unit. Also, steam raised in the gasifier jacket is used to
supplement the waste-heat boiler. Because the fuel gas
steam provides heat directly to the sour water stripper,
the quantity of low-pressure steam required is relatively
small and the steam raised in the gasifier bleed air stream
is sufficient to satisfy this need.
The Bureau of Mines/5intered Iron Oxide integrated
system is shown in figure 8. Because the tar in the prod-
uct gas does not affect operation of the hot cleanup
system, the water quench can be eliminated. The flow
schematic diagram is quite simple due to the lack of
regeneration and elimination of the water scrub and tar
recycle. Gasifier performance is enhanced by not rec,,-
cling the tar since it permits operation at lower air- and
steam-to-fuel ratios. The net cleanup system reactions
are exothermic both during absorption and desorption,
adding sensible energy to the fuel and enabling the re-
covery of heat in the form of process steam from the
regenerator off-gas. However, the sulfur comes off the
iron oxide as 502 and in the sulfur recovery system
must be concentrated and a portion reduced to H25 so
that it can be used to form elemental sulfur in a Claus
unit. This process uses a considerable amount of fuel and
process steam. As a result, even with the steam raised in
the process, a significant amount of low-pressure steam
255
-------
BUMINES I SELEXOL SYSTEM
INLET
A'R
d
~~
~
GAS
TURBINE
E XHA,UST
.~
en
0)
STEAM CYCLE
CONDENSER
COOLING
TOWER
PROCESS
COOLING
WATER.
eOA...
HANDLING
- ~. ~ ~
BOOST
Cf)MPRESSQR
SfP,
ACCUM
Figure 7.
CLEAN GAS TO BURNt:.R
CYCL~
DUSTY.
COAL
H p. 5TM
F"
..R
STEAM,
CONDENSER
SOUR
WATER
STRIPPER
1i2SWS
REBOtLER
Bumines/Selexol system.
P~OOUCT GAS
RE"1EAT ...x
tooL.fR
C""OE"-SER
~ATER
QUE NC...
GAS
LtQUI:;)
SEP
TAR
WATEP
SEP
OIL'
WATER
SE'
COOLE R
TO WATE!=!:
SCRUBBER
WASTE
WATER
A.~""ONIA
RECOVERY
N"J
'=LEAN GAS TO SULFUR RECOVER'Y'
'lvA TER
SCRUBBER
SEP,
ACCUM
CLEAN
PRODuCT
GAS
SE'..E-OL
ABSQRBER
1= ROM
SWS
CONDENSER
SELExOL
STRIPPER
LP
STM
SULFUR
RECOVERY
- ~. ~ ~..
s ~.:.: 0(
'~.:.s
AIR IN
-------
"',-E.
...
BOOST
COMPRESSOR
:'AS ......c:ae,""'E
E~';;'.J5"
I'.)
U1
.....
STEAM CVCl.E
co...ce'.sED
COO:.. "\;G
.OWER
Figure 8.
BUMINESISINTERED IRON OXIDE SYSTEM
ClEAN GAS TO BURflr,jER
COAL
HANDl.ING
SY'STEM
DUS1'
H.P. STM
f.WI'.
STEAM
L.P S1'eAY
PROCESS COOLING WA TER
PART':U:,,"'''E
pe"'OVA-
Bumines/Sintered Iron Oxide system.
D.JS"
:'I'CL.~E
DUST
5VL~UP
GAS
SORPT,Orv
---------
ST ACIC GAS
S1.JI..~:..!Q REC.:)'wEpy
SULfUR
PURCOE
AIR IN
l..p SHAY
IRON
O,.ODE
REGEI'Ij.
COOLER
-------
must be taken ':rom the waste heat boiler to satisfy these
requirements. The lower gasifier steam requirement
means that nea.-Iy all the heat needed to raise the gasifier
steam can be taken from the gasifier cooling jacket. The
system suffers from the inability to remove ammonia
from the gas and the uncertainties associated with partic-
ulate removal. Because ammonia would not present a
problem if it were reduced to its equilibrium concentra-
tion, the possitility of catalytic decomposition was con-
sidered. Iron oHide appears to be potentially capable of
catalyzing the :1ecomposition while removing hydrogen
sulfide. This would remove the primary source of nitro-
gen; however, the tars would contain some nitrogen and
sulfur compoul1ds that would not be removed by the
cleanup system.
With the BGR/Selexol system, shown in figure 9, it
is possible to both regenerate and extract useful heat
from the gasifier exit stream. Ammonia and H2S remov-
al are quite similar to the Bureau of Mines system, but
the Selexol sys':em is complicated by a high percentage
of cas in the gas to be processed. Because cas is about
one-third as soluble as is the H2 S, its presence in the gas
stream requires that the process be tailored specifically
for removal of cas. The utilities for this system reflect
this situation. A significant amount of heat is recovered
from the gasifiel' jacket and the high-temperature gasifier
exit stream. Thh; heat is used to raise high-pressure steam
which supplem mts the high-pressure section of the
waste heat boil er by adding about 50 percent to the
steam raised there. After the steam for the gasifier is
extracted, the l1et increase in steam available for the
turbine is about 25 percent. All of the low-pressure proc-
ess steam requir ements are met by steam raised in the
boilers located upstream of the boost compressor and
downstream of the fuel gas regenerator. '
In the BCR'Conoco system,shown in figure 10, the
gas stream arrangement is very simple and the omission
of a water scrub makes process steam requirements quite
low. As a result, it is possible to raise high-pressure steam
in boilers located upstream of the boost compressor and
in the transport gas stream prior to the CO2 removal
unit. The low-pressure process steam needed to supple-
ment that raised within the cleanup and sulfur recovery
system can be Sl pplied by a boiler located between the
high-pressure boiler and boost compressor. The high-
pressure steam is used to supplement the waste heat
boiler and balarces the amount of superheated steam
extracted for inj€ ction into the gasifier.
Integrated Sysrer" Performance
A summary of the calculated performance charac-
teristics for each of the four systems described is given in
table 1. The perf ormance of the second generation gasi-
fication and cleanup systems mated with a first gen-
eration power system is also given to permit a gross com-
parison of the gasification and cleanup systems. In
addition, the integrated station performance variations
as a function of gas turbine pressure ratio are shown in
figures 11 and 12 for first and second generation power
systems, respectively. To provide consistency and a basis
of comparison, both figures include the performance
estimates for distillate-fired combined cycles.
Looking at overall net plant efficiency, it is clear
that there are two significant factors in performance im-
provement: turbine inlet temperature and high-
temperature cleanup. Increasing turbine inlet temper-
ature from 2,200° F to 2,600° F produces a 13 percent
improvement in cycle efficiency. For the second genera-
tion gasification systems, the difference in performance
between high- and low-temperature cleanup is about 18
percent. Detailed analysis of the high-temperature clean-
up case reveals that only 7 percent of the gain is due to
the high-temperature operation, approximately 7
percent is due to a higher cleanup process efficiency
(lower requirements for process steam and auxiliary
power). and the remaining 4 percent is due to the lack of
ammonia removal provisions. Even the 7 percent gain for
high temperature could be reduced by development of
fuel gas regenerators capable of operating above 1,000°
F and by rehumidification of the fuel gas stream to
replace the water vapor condensed during the cool down
process.
Integrated Station Cost Comparison
Historically, the U.S. electric utility industry suc-
cessfully reduced the cost of' generating power by
utilizing the latest available technology and taking
advantage of economies associated with large-scale gener-
ation facilities. This era of decreasing cost of electricity
has ended, and we are now in a new era with ri,sing costs.
This unfortunate situation is a direct result of rapidly
rising construction and fuel costs combined with public,
demands for effective control of atmospheric and ther-
mal water pollution. Rising costs plague all methods of
power generation, both fossil and nuclear. Because of
this situation it is extremely difficult to accurately
estimate the capital and operating costs of future power
systems.
Nevertheless, preliminary estimates of capital and
operating costs were made (see table 2) for each of the
four selected systems. In addition, the costs of power
generation in a conventional plant with stack gas cleanup
and in a gasified coal-fired steam plant have been esti-
mated to provide a basis for comparison. A summary of
total power generation costs for the six systems is pre-
sented in figure 13 in terms of base coal cost. These
258
-------
1~l.ET
GAS
TURBII'.oE
EXo-IAUST
1'oJ.
en
CO
STEAM CYCLE
COi\oOE.I\,jSE~
COOLING TOWER
,
o
PROCESS
COOLING
WATER
BCRISELEXOL SYSTEM
CLEAN GAS TO BURNER
COAl.
PROCESSING
SYSTEM
COAl. TRANsPORT
G"
M.P.
ST"
BOOST
COMPRESSOR
COAL
6 GA.S
CYCLONE
se"
GASIFIER
CHAR
WASTE
WATER
OUEHCo-I
WATER
SLAG
COOl.Ef:!
SOU"
WATER
STR IPPE R
AMMOf\IIA RECOVERY
Figure 9.
BCR/Selexol system.
COOl.ER
"'ATER
SCRUBBER
TO WATER
SCRUBBER
WASTE
WATER
N"J
SELEJ(O..
S:RU8BEA
GAS
WA TER
SE-
SHEXOl
STRIPPER
SOLveNT
COOLER
SULFUR RECOVERy
.
CLEAN
FUEL
AIR IN
SULFUR
STACK GAS
-------
I\J
0)
o
CLEAN GA.S TO BURNER
,'uET
,..
rtT~ riT~
_C\ C\
\.J ~
.
BOOST
CO,",PI=I:E SSO~
C('MPGESSOR
G4S TVABIf\lE
i
SUPE~":.---'-L ~----==l
- ---4 i IA'.
I
SCR
GAsr~IEA
- - - -
CHAR
OUENCH
- WATEA
SLAG
CO:.... &
""
S~E:'..' C>CI.E
c..S
CONVE:ATER
WATER
CONDE NSfR
:;'OL''\iG
':"0~"EC=
COAL
PROCE SSING
SYSTEM
PROCESS COOU~G
WATEA
Figure 10.
BCR/Conoco system.
DOL ()aojI ITE
.ABSORBER:
DOL~ITE
MAKE UP
",S
DOLOM ITE
SLUDGE
C02 &
STEAM
TR AN$PQR T
GAS
co,
SEPARATION
~PST'"
TRANSPORT GAS
TO c~ ABSORBER
Ll
OOLO,,",'TE
REGEl'., .
SPENT
DOLOMITE
'liA TEA TO SW5
L.P STEAM
SUl FUR
RECOVERY
STACK
GAS
SVlF'UR
I - COOLING WATER
L-A1A
CONDENSATE
TO 5.W S
-------
Table 1. Integrated systems performance summary
Second
generation gasification
First
First generation Second generation generation power system
BOM/ BOM/ BCR/ BCR/ BCR/ BCR/
Selexol Iron oxide Selexol Conoco Selexol Conoco
Gas turbine
Turbine inlet temperature - OF 2,200 2,200 2,600 2,600 2,200 2,200
Compressor pressure ratio 16 16 24 24 16 16
Exhaust temperature - of 916 913 1 ,107 1,115 913 920
Output power - MW 595.4 626.2 726.6 857.6 642.3 757.6
Steam cycle
Steam temperature - OF 816 813 1, 000 1 ,000 . 813 820
Steam pressure - psia 1,250 1 ,250 1,250 1,250 1,250 1,250
Condenser pressure in. Hg. Abs. 4.0 4.0 4.0 4.0 4.0 4.0
~ Net steam cycle output - MW 223.8 208. 1 293.3 296.6 273.5 271 .4
C»
-- Net steam cycle efficiency .280 .292 .307 .307 .282 .279
Gasifier and cleanup system
Coal feed rate - lb/hr 700,000 700,000 700,000 700,000 700,000 700,000
Air/coal ratio . 3,013 2,688 3,088 3,088 3,088 3,088
Steam/coal ratio .405 .349 .567 .567 .567 .567
Air preheat temperature - of 800 800 800 800 800 800
Steam temperature - of 584 584 1,000 1,000 913 920
Steam pressure - psia 1,250 1 ,250 1,250 1,250 1,250 1,250
Gasifier exit temperature - of 1 ,000 1 ,000 1,800 1 ,800 1 , 800 1,800
Cleanup system exit temperature - OF 265 1 ,070 1,000 1,700 1,000 1,700
Fuel gas higher heating value Btu/SCF 160.3 165.9 159.3 1 35. 8 159.3 135.8
Integrated station
Gross power - MW 819.2 834.3 1,019.9 1, 1 54 . 2 915.8 1,029
Boost compressor power - MW 43.4 36.1 40.1 40.2 40.1 40.2
Gasifier & cleanup aux. power - MW 28.2 36.5 58.7 27.6 58.7 27.6
Plant auxiliaries - MW 10.6 10.2 13.6 14.5 12.5 13. 1
Net plant output - MW 737.0 751.5 907.5 1,071. 9 804.5 948. 1
Net plant efficiency (KHV-Coal) .314 .320 .360 .425 .319 .376
-------
TURBINE INLET TEMPERATURE = 2200F
CONVENTIONAL COOLING
. 28 ~12,
32 -7 )..
SIMPLE CYCLE/
DISTILLATE
28 24
/
LI,S
>
J:
J:
40
'*
>-
U
Z
w
U
LL
LL
W
PR = 40
35
20
16
8
COGAS/BCR/CONOCO
40
COGAS/.
BCR/SE LEXOL
2~20 16
12
24 8
28
3'J
50
100
150
SPECIFIC POWEA - KW/LB/SEC
250
200
Figure 11. Performance estimates for first generation
turbine systems.
estimates are based on mid-1974 costs with escalation
through the construction period. It is quite clear that the
most attractive gasification/cleanup system for use with
a combined cyde powerplant is the BCA/Conoco com-
bination. This second generation combination appears
capable of gene-rating electric power at much lower cost
than any other alternative, including steam with gasi-
fication and steam with stack gas cleanup. The other
combinations are only marginally competitive with the
steam alternatives.
It is intereiting to note that the relative cost com-
parisons do no': change greatly over the coal cost range
of interest. Also, for a given gasifier and power system,
the low-temper ature cleanup is more c<;lstly than the
high. This is mostly due to the lack of ammonia removal
and associated equipment in the high-temperature clean-
up systems. In the case of the BCA/Conoco system,
however, that effect is amplified by the improved effi-
ciency of the gasifier/cleanup combination, resulting in a
greater plant output over which to amortize the capital
cost. Changing from first to second generation systems
also tends to decrease initial cost due to a higher gas
turbine specific output, which results in a lower cost per
kilowatt of the power system.
Environmental Impact of Integrated Stations
T he four integrated power stations previously
described were used as a basis to identify the various
pollutants which would be associated with this type of
powerplant. A summary of the estimated environmental
impact of the four selected systems is given in table 3.
262
-------
1 UIHlINE INLET TEMPER:' TlJRE = 7600F
CERAMIC VANES. CONVEIIITIONAL BLADES
50
45
COGAS/DISTILLATE
\
12X
40 36 32 28
COGAS/BCR/CONOCO (T/FUEL-85%)
EFFICIENCY -
%(HHVI 40
SIMPLE CYCLE/DISTILLATE
24
35
20
16
12
30
100
150
12
PR
16 COGAS/BCR/SELEXOL 11)FUEL-74%)
12
8
200
250 .
300
350
SPECIFIC POWER - KW/LB/SEC
Figure 12. Performance estimates for second generation
turbine systems.
Waste water, thermal, SOZ, NOx particulate, solids,
sulfur, and ammonia discharges from the gasification!
cleanup system and from the combined cycle power
systems are shown.
It is apparent from the table that SOz emissions
from the sulfur recovery plant could be significant (60
percent to 85 percent of total sulfur emission). However,
commercially available tailgas treating processes could be
used to further reduce sulfur levels in these cases. All
sulfur remaining in the fuel (as Hz S or in tar vapors from
the Bureau of Mines gasifier) were assumed to be con-
verted to SOz during the combustion process. Resulting
total sulfur emissions should all be less than the EPA
limit (1.2 Ib SOz!MMBtu) for coal-fired powerplants.
The formation of nitrogen oxides (NOx) is a mUl:h
more complex problem than that of sulfur oxides. Two
mechanisms, one involving nitrogen in the air and the
other involving fuel nitrogen, contribute to the forma-
tion of nitrogen oxides during combustion processes.
The thermal mechanism dealing with atmospheric nitro-
gen is reasonably well understood and many analytical
models have been developed to predict NOx formation
using this mechanism. Additional nitrogen oxides would
be produced by nitrogen-bearing compounds (NH3,
HCN, pyridines, etc.) in the fuel. These compounds are
produced from the nitrogen-bearing compounds in coal
during the initial gasification process. They would be
removed by low-temperature cleanup systems, but
263
-------
Table 2. Cost comparison for coal-fired power plants
Combined- Combined- Combined- Combined-
cycle cycle cycle cycle Steam Steam
BuMines BuMines BCR BCR stack gas BCR
Selexol iron oxide Se1exol Conoco cleanup Conoco
Capital Cost - $/kW
Power system 232 230 208 190 345 280
Gasification system 111 107 117 99 111
Fuel gas cleanup 88 48 90 35 44
N Stack gas cleanup 81
~
Total capital cost 431 385 414 324 426 435
Owning-Plus-Operating Costs - mi 11 sl kWh
Owning costa 11.9 10.7 11. 5 9.0 11.8 12.0
Operation & maintenance
Power system 1.3 1.3 1.2 1.1 1. 1 . .8
Gasification & cleanup 2.8 2.1 . 2.8 1.8 2.5
Stack gas cleanup 1.1
Fuel @ 60~/MMBtu 6.5 6.4 5.7 4.8 5.8 6.1
Total power cost 22.5 20.5 21.2 16.7 19.8 21.4
a of capital, 70 percent load fac tor.
Seventeen percent
-------
40
-----
GASIFIED COAL - CDGAS SYSTEM
GASIFIED COAL - STEAM SYSTEM
J:
3:
~
.......
CI)
...J
...J
::E
I
Ii;
o
u
z
o
i=
<
ex:
w
z
w
C)
ex:
w
3:
o
Q.
30
BCRISELEXOL
(SECOND GENERATION)
BCR/CONOCOISTEAM
20
100
0.50
1.00
COAL COST, $/MM Btu
1.50
2.00
Figure 13. Cost summary.
265
-------
Table 3. Estimated environmental impact 01 integrated gasification/combined-cycle power systems.
(All values in Ib/106 Btu coal unless otherwise noted)
Waste a b Ammoniab
water Thermal S02 NO Particulate Solids Sulfur
x
BOM-Se1exo1 219c 0.005 0.320f 0.356g O~Olh 14.6 2.96 1.38
2,200° F 27d e 0.088 0.174
To ta 1 246 0.005 0.408 0.530 0.01 14.6 2.96 1.38
BOM-iron oxide ---d 0.018e 0.800f 0.31g ---- 14.6 2.51
2,200° F Com.-~yc. 27 0.337 3.29 0.06i
Total V 0.018 1 . 137 3.70 0.06 14.6 2.51
BCR-Se1exo1 520c 0.016e 0.487f O. 300g ----h 7.1 2.74 0.96
2,600° F com.-cyc. 24d 0.080 0.012 0.01
I\J Total 544 0.016 0.567 0.312 0.01 7:T 2.74 0.96
C)
C)
BCR-Conoco 7Sc 0.002e 0.033 . 7.1 2.72
2,600° F comb.-cyc. 22d 0.520 2.65g 0.041 ----
Total 97 0.002 0.553 2.65 0.04 7:T 2.72
aMi11ions of Btu rejected per million Btu of coal.
bRecovered with no credit for subsequent sale.
cContains trace amounts of contaminants.
dIncludes boiler makeup and cooling tower losses.
eMechan i ca 1 draft cooling towers are used.
fInc1udes emissions from Claus plant. A 90 percent + reduction could be obtained with tail-gas cleanup.
gSum of therma 1 and fuel-related NO .
h x
Assumed to be equal to methane-fired systems.
iSum of carryover and combustion products.
-------
would pass directly through the high-temperature clean-
up systems. NOx emissions from high-temperature
systems would be higher than from low-temperature
systems because of the fuel nitrogen and because hot
fuel would produce higher combustion temperature lead-
ing to higher thermal NOx. As indicated in table 3,
emissions of NOx from power systems using hot cleanup
processes could be unacceptable.
Both first and second generation systems will dis-
charge substantial quantities of solid material for
disposal to the environment. This solid material will be
primarily ash and slag, with substantial amounts of
chemical wastes (such as sodium sulfate or spent dolo-
mite). Byproduct sulfur and ammonia can be stored in
solid or liquid form until sold and should present no
significant environmental problems.
The aqueous waste streams from the low-
temperature cleanup systems are potentially hazardous.
These strea.ms will contain nonstrippable contaminants
originally present in the raw fuel gas. Data on the nature
and quantity of trace contaminants are not available.
Design of a water treatment system to render the aque-
ous wastes acceptable for discharge would require'
detailed analysis for the specific gasification process
involved. .
CONCLUDING REMARKS
The purpose of this paper has been to identify the
economic, performance, and environmental character-
istics of advanced combined-cycle power systems opera-
ting on distillate oil and coal-derived low-Btu gas. First
generation coal gasifiers used in conjunction with
commercially available low-temperature cleanup systems
and combined cycles have the potential of producing
electrical power at costs competitive with conventional
steam plants using stack gas cleanup. Those integrated
gasification/combined-cycle systems would also have
lower emissions of S02, NOx, particulates, and reduced
thermal pollution. High-temperature cleanup processes
used in conjunction with advanced combined cycles have
the potential of providing significant advances in effi-
ciency and lower capital costs, resulting in power costs
20 percent lower than steam plants with stack gas clean-
up. Emissions of NOx from systems using hot cleanup
could be unacceptable.
Further work is underway to answer some of the
questions that have been raised. For each of the system
combinations, those factors that adversely affect either
performance, cost, or pollutants have been identified
and alternate approa.ches are being evaluated. In addi-
tion, other advanced gasification processes that would
minimize the use of steam in the process and thus en-
hance the overall cycle are being considered. These
efforts pl-omise. to show improved performance and cost
characterIstics while maintaining low levels of emissions.
REFERENCES
1. 26th Annual Electrical Industry Forecast,
September 15, 1975.
2. F. L. Robson, A. J. Giramonti, G. P. Lewis, and G.
Gruber, Technological and Economic Feasibility of
Advanced Power Cycles and Methods of Producing
Nonpolluting Fuels for Utility Power Station,
UARL Report No. J-970855-13, December 1970
(NTIS No. PB 198392).
3. F. L. Robson, A. J. Giramonti. W. A. Blecher, and
G. Mazzella, Fuel Gas Environmental Impact,
EPA-600/2-75-078. November 1975.
267
-------
GAS TURBINE HTGR FOR ECONOMICAL DRY COOLING OR HIGH-EFFICIENCY
COMBINED CYCLE
John M. Krase*
Abstract
This paper describes the current reference design
and the design e'.lolution for a large, helium-gas-turbine
nuclear power plant under development by General
Aromic Company. This type of nuclear electric gener-
ating station offers major improvements in plant simpli-
fication, efficienr:v, resource utilization, and economics,
as well as substantial reductions in the environmental
effects of pOIM'r generation. The dry-cooled reference
design of the Gas Turbine HTGR (GT-HTGR) as of
October 1975 is described; it includes a 3,OOO-MW High-
Temperature G~s-Cooled Reactor (HTGR) and three
power conversion loops, each containing a 370-MWe,
horizontally oritmted helium turbine-generator. Further
studies in 197ti have established an improved design
baSlJd upon the ~c1me concepts, which for a site with no
cooling water w')uld hrlVe a dry-cooled station efficiency
of 40 percent. For sites with some cooling water, the
reject heat from the helium turbine power conversion
loops can be USlJd in a wet-cooled secondary power cycle
to produce adc'ir:ional power, thus raising the overall
station efficienc}l to about 48 percent.
INTRODUCTION
The Gas Turbine High Temperature Gas-Cooled
Reactor (GT-HTGR) makes use of a helium-cooled
graphite core to provide hot helium to a closed-cycle gas
turbine power conversion system which operates on the
reactor coolant helium. The helium gas turbines drive
conventional elf ctric generators to produce the electric
output of the plant. The High-Temperature Gas-Cooled
Reactor (HTGII) core, which is also used in the
GT-HTGR, has been developed for commercial appli-
cation in large central stations with a steam turbine-
generator.
BACKGROUND
Gas Turbines
The closed-I:ycle gas turbine (ref. 1), including the
helium turbine (ref. 2), has been developed primarily in
Europe for use Nith fossil fuels (coal and coal gas). A
* Project ManEger, Gas Turbine HTGR Project, with the
Advanced Concep~s Division of General Atomic Company, San
Diego, California.
number of small- to medium-size, fossil-fired, closed-
cycle gas turbine plants for power generation, plus
district heating are in service in Europe and Japan and
have demonstrated very long life and high availability.
The performance of this type of plant is particularly
impressive since overall utilization of input heat in the
form of electric power plus district heat can attain values
of 60 to 80 percent. However, one of the critical compo-
nents of the fossil-fired closed-cycle gas turbine has been
the input heater which is limited in temperature level by
materials and by combustion side corrosion and chemi-
cal effects. The availability of the HTGR core to provide
the high-temperature gas (helium) for the closed-cycle
gas turbine effectively. solves the input heat exchanger
problem which has limited the fossil-fired closed-cycle
gas turbine to relatively small units where district heat-
ing is needed.
The open-cycle gas turbine development and utiliza-
tion, in which the United States has been the leader, has
has been strikingly successful in a wide variety of appli-
cations ranging from military and commercial aircraft, to
vehicles, and to stationary application such as gas pipe-
line pumping, utility peaking generation, and interme-
diate load generation. The latter has made use of spe-
cially developed heavy-duty gas turbines with steam-
turbine bottoming cycles to achieve good availability
and efficiency. Almost all of the open-cycle gas turbines
so far have required relatively good quality petroleum-
based fuels. Problems in long-term availability of such
fuels are now seriously inhibiting further utility use of
open-cycle gas turbine power plants. However, the tech-
nology and manufacturing capability needed for the heli-
um gas turbines of the GT-HTGR already exist. In fact,
the gas turbine temperature levels and other technical
requirements for the GT-HTGR application are relatively
easier than those of the fossil-fired gas turbines.
HTGR
The first HTGR to produce electricity was the
40-MWe prototype built by General Atomic at Peach
Bottom, Pennsylvania, which began commercial opera-
tion in June 1967. The Peach Bott~m plant was shut
down in October 1974 after a total of 1349 equivalent
full-power days and producing more than 1.38 billion
kWh of electric energy for the Philadelphia Electric
Company. This reactor served as an invaluable test bed
for the fuel for the large HTGR and for reactor physics
studies. The average gross plant efficiency over its
7%-year life was 37.2 percent and the nuclear system
268
-------
availability was 88 percent, both of which are nuclear-
industry performance records.
The 330-MWe Fort St. Vrain HTGR has been built
by General Atomic for the Public Service Company of
Colorado under the Power Reactor Demonstration Pro-
gram of the U.S. Energy Research and Development Ad-
ministration (ERDA).* Construction of the plant is com-
plete, and power production is expected to start in
1976. The technology developed for the Fort St. Vrain
plant and its main components is directly applicable to
the large HTGR's scheduled for operation in the early
1980's. Development programs for these large reactors
were initiated in early 1968; the remaining portions of
these programs consist of component proof tests, con-
tinuing long-term metallurgical tests, and miscellaneous
confirmatory tests of design details.
In addition to the HTGR programs in the United
States (at General Atomic and Oak Ridge National Labo-
ratory), extensive development programs are in progress
in England (the DRAGON project), in Germany (the
AVR and THTR plants and other advanced projects), in
France (the commissariat a l'Energie Atomique), and in
Japan.
HTGR Features
The general arrangements of the large HTGR nu-
clear steam supply system are shown in figure 1. A sche-
matic flow diagram is shown in figure 2.
The HTGR's use helium gas as the coolant, graphite
as the neutron moderator and fuel-element structural
mate rial, with pyrolytic-carbon-silicon-carbide-coated
US2 fossile fuel particles, and pyrolytic-carbon-coated
Th02 fertile particles dispersed in a graphite matrix as
the fuel.
The choice of graphite as the moderator and core
structural material is based on its unique chemical,
physical, and mechanical properties at elevated temper-
atures and on its very low neutron cross section, good
radiation stability, ease of fabrication, and low cost.
The ThU2 3 3 standard fuel cycle (with U2 3 S as the
initial fissionable fuel) is used because of its potential for
achieving a higher fuel utilization and lower power cost
than any other thermal spectrum reactor system. The
neutronic characteristics of U2 3 3 are superior to those
of either plutonium or U235 in thermal systems. A sub-
stantial portion of the energy comes from fission of the
U2 3 3 converted from the fertile Th2 3 2. Efficient use of
the thorium fuel cycle is of major significance in extend-
ing the U.S. nuclear fuel resources as indicated in figure
3. Features of the HTGR's uranium and thotium fuel
cycle are summarized in figure 4.
* Formerly the U.S. Atomic Energy Commission.
Other features of the HTGR which apply also to the
GT-HTGR include inherent safety resulting from the
high-temperature capability and thermal capacity of the
core which avoid the need for a fast-acting emergency
cooling system, and the integration of the reactor and
entire helium system into the PCRV. A very high integ-
rity pressure vessel results from the use of multiple and
redundant structural elements in the PCRV. The radio-
active releases from the HTGR plant are negligible and
very much lower than from LWR's.
, GT-HTGR
A program now in progress is aimed ,at the design
and development of a commercial GT-HTGR power
plant.
Because of its high reject-heat temperature the
GT-HTGR is exceptionally well suited to economical
rejection of waste heat directly to the air through the
use of dry-cooling towers. For sites with cooling water a
secondary power cycle can be added to generate substan-
tially more power from otherwise reject heat, thereby
increasing overall plant efficiency from an already high
value of nearly 37 percent for a dry-cooled plant at ISOt
standard day conditions, to about 46 percent for the
water-cooled binary plant. The GT-HTGR plant designs
described in this paper are, based upon currently avail-
able technology and require a relatively low-cost design
and development program.
Preliminary evaluations indicate that an 1,1 OO-MWe
dry-cooled GT-HTGR plant can be constructed for
about the same cost as a wet-cooled conventional nu-
clear steam electric plant; and compared to a dry-cooled
nuclear steam electric plant, the GT-HTG R plant exhib-
its a capital-cost advantage. With the wet-cooled binary-
cycle GT-HTGR plant, an advantage in installed cost per
kilowatt is expected relative to the dry-cooled
GT-HTGR plant. In addition, there are fundamental
advantages of further reduced reject heat from the plant
and higher net plant output for a given thermal cere
power limit.
The preliminary design and program definition
phase currently underway is a part of an overall program
aimed at the design and development of the commercial
GT-HTGR power plant. Both the dry-cooled and the
wet-cooled binary-plant options are included. The con-
ceptual and advanced preliminary design phases have
involved substantial participation by the U.S. Energy
Research and Development Administration, 16 utility
companies, 3 leading turbomachinery manufacturers,
and the General Atomic Company. A prominent feature
of the proposed development program for the GT-HTGR
t I nternational Standards Organization (ISO) sea-level
standard atmospheric temperature of 59° F.
269
-------
HEll UM
PURl FICATION
WELLS
AUXILIARY
CIRCULATOR
I\J
-..I.
o
CIRCUMFERENTIAL
PRESTRESSING
SYSTEM
COR E AUXI LIAR Y
HEAT EXCHANGER
PRESTRESSED
CONCRETE
REACTOR
VESSEL
CONTROL ROD
DR IVE AND
RE FUELING
PENETRATIONS
CIRCULATOR
LINEAR
PRESTRESSING
SYSTEM
STEAM
GENERATOR
PCRV SUPPORT
STRUCTURE
Figure 1. HTGR nuclear steam system.
-------
HEL I UI1
CIRCULATOR
CJ
""
-
PCRV
STEAK
GENERATOR
TURBINE
GENERATOR
2~OO PSIG
950°F
CONDENSER
55~ PSIG
IOOO°F'
BOILER
FEED PUMP
CONDENSATE
PUMP
Figure 2. Overall HTGR plant flow diagram of typical
secondary cooling system.
( < $100/LB)
(Q = BTU X 1018)
"I .. :- ,i
.. ...., " 48' O~ Q" ': .i".,:,- '.
{. . .r ~:/. . .J:?
FROM U30a
F.ROM THORIUM
Figure 3. Probable U.S. uranium and thorium resources.
271
-------
URANIUM/THORIUM FUEL
. HIGH CONVERSION RATIO
. HIGH BURNUP
. LOWER FUEL INVENTORIES
. REDUCED SENSITIVITY TO CHANGING
U30S COSTS
Figure 4. Uranium/thorium fuel.
is the full-seal ~ demonstration test of a 370-MWe gas
turbine-generator and other primary system components
in a fossil-fired test facility. A preliminary engineering
study of the f,lcility by NASA is currently in progress
under ERDA sponsorship.
Also, a program is in progress in the Federal
Republic of Gurmany for the development of a direct-
cycle nuclear g,IS turbine p,lant designated as HHT. As a
result of a joint investigation in 1974 between the Gen-
eral Atomic Ccmpany and Hochtemperatur-Reaktorbau
GmbH (HRB) of Germany, the reactor designer and
systems integrator for the German direct-cycle project, a
substantial degree of commonality in plant configuration
was developed Similarity in plant configuration and
coordinatiop 01 development efforts is expected to bene-
fit both programs.
GAS TURBINE HTGR
Cycle Condit;Of's
The reference-plant cycle, figure 5, is dry cooled.
The binary-plar.t cycle conditions are shown in figure 6.
The cycle concitions are given for 590 F cooling-water
temperature.
The resulting net plant performances with a
3,OOO-MWt reactor core are 37 percent efficiency-about
1,100 MWe-fo' the dry-cooled plant, and 46 percent
efficiency-abollt 1,380 MWe-for the binary plant.
Major design p:>int and rated parameters for the dry-
cooled plant ar,~ given in table 1. The more important
parameters remain much the same since plant concep-
tion, consistent with the goal of maximizing the use of
ex isting technol,)gy.
1.
1,OOO,psi maximum design-point primary-
system pressure results in high gas density and
compal:t turbomachinery with only a modest
extensbn of PCRV structural technology. Pre-
liminary studies indicate this is the economical
choice as well.
2. 1,5000 F turbine inlet temperature is achieved
with the same peak fuel temperature used by
current steam-cycle HTGR's. Also, this is below
the temperature level where turbine-blade cool-
ing is required.
3. The design-point compressor compression ratio
of 2.35 is slightly less than optimum for effi-
ciency, but it was chosen to reduce the number
of turbine and compressor stages needed.
Dry-Cooled Plant Description
The GT-HTGR reference-plant design has an inte-
grated primary loop with all primary system components
including the turbomachinery contained within the
PCRV. A closed-loop circulating-water system carries
reject heat from the primary power conversion loops to
two natural-draft dry-cooling towers where it is dissi-
pated to the atmosphere. The capability for economic
dry cooling is considered a major advantage of this type
of plant.
The compact, integrated-circuit gas turbine power
plant provides inherently high reliability against
pressure-boundary failures that could result in rapid
system depressurization. This important safety feature
has been basic to the circuit design since early concep-
tual studies.
Plant Arrangement
Major components of the primary system are illus-
trated in figure 7. The 3,OOO-MWt reactor system incor-
porates three power conversion loops. Each loop
includes a single-shaft gas turbine and generator of
370-MWe rating, a recuperative heat-exchanger unit, and
a precooler heat exchanger for cycle-heat rejection. The
heat exchangers are contained in vertical cavities
arranged within the PCRV wall, two for each loop. The
three turbomachines and generators are horizontally
272
-------
q/////////////////////////////////////////////////////////////////////////////////////////////////.
~ ~
~ . . s
~ ~
~ ~~ ~
~ ~
~ 410 PSIA ~
~ 3340F ~
~
~ 962 PSIA
~
~
~
~
~
~
~
~
.~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~ ~
~ ~
~ ~
~ ~
~ ~
~ TURBINE COMPRESSOR ~
s ~
~ S
~///////////////////////////////////////////////////////////////////////////////////////////////h~
922°F
REACTOR
CORE
3000 MW(t)
RECUPERATOR
994°F
419 PSIA
15000F
HELIUM
PCRV
PRECOOLER
18°F
403 PSIA
PLANT EFFICIENCY = 37%
~.
~
~
~
~ 590F
~
~ AIR
~
~
~
~
~
~
~
~
~
~
~ 69°F
94°F
DRY
COOLING
TOWER
~
~
~
~
~
~
~
GENERATOR
1091 MW(e)
Figure 5. GT-HTGR with;.dry cooling..
oriented and spaced azimuthally 120 degrees apart. The
generators are external to the PCRV, and the turbo-
machines are within the radial PCRV cavities. The con-
necting output shaft penetrates through the PCRV clo-
sure plug. Purified-helium buffer seals on the shaft assure
no leakage of primary helium at the shaft penetration.
Horizontal orientation allows use of a conventional gen-
erator and employs a more conventional turbomachine.
Penetrations in the bottom head of the PCRV make in
situ access to the turbomachinery bearings possible. This
is a particular advantage of the horizontal arrangement.
The secondary containment vessel is a prestressed
cylindrical concrete structure of 159-ft inside diameter
with an oblate spheroidal dome. The generators are
housed in grade-level enclosures that extend radially
about 40 ft from the containment cylinder wall.
Heat rejection is accomplished by the recirculating-
water system and dry-cooling towers. Also included in
balance-of-plant are the reactor service building (includ- .
ing fuel storage and shipping). service systems, auxilia-
ries, and maintenance facilities.
Reactor
The fuel components, reactor core and reflectors,
core support, control-rod mechanisms, and fuel-handling
equipment for the GT-HTGR plant are essentially the
same as those used for the commercially offered stezm-
turbine HTG R. Modest core loading adjustments Ne
anticipated to account for the approximately 40 perce,t
higher helium flow rate and correspondingly lower coel-
ant temperature rise in the GT-HTGR application. With
1,500° F reactor outlet temperature, the peak fuel and
graphite temperatures in the core are essentially the
same as those in the steam plant. The lower reactor out-
let temperature in the 3,OOO-MWt steam-turbine HTGR
of 1,366° F is offset by the lower helium temperature
rise in the hot channel of the gas turbine reactor core
because of the higher helium flow rate, thereby retaining
the same maximum fuel-temperature conditions.
Components
A significant advantage of the closed-cycle gas tur-
bine is that the compact size of the turbomachinery
273
-------
5<'/'/'/""/"".'//'//////////////'/////"'/////////////"'/"/"/'//""""/"""""""''''/''''''''''''1-
~ . ~
~ ~
~ 5490F ~
~ 433 PSIA ~
~ 414°F ~
~ 991 PSIA
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~ ~
~ ~
~ ~
~ ~
~ TURBINE COMPRESSOR ~
~ ~
~ ~
~/""",,/,,/"',/ '/"//,//////////,/"////"",,/,,,///,/,,///,///////////////////////,//,///////////,s
817°F
<>
<,
RECUPERATOR
")
<>
<
REACTOR
CORE
3000 MWlt)
1000 PSIA
996°F
438 PSIA
15000F
HELIUM
PRE COOLER
VAPOR
TURBINE
89°F
GENERATOR
456 MWle)
590f
WATER
CONDENSER
~
~
.~
~
~
~
~
GENERATOR
922 MWle)
PCRV
PLANT EFFICIENCY = 46%
PLANT ELECTRICAL OUTPUT = 1378 MW
Figure 6. Binary GT-HTG R with wet cooling.
allows design of a completely integrated circuit with its
attendant safet\' and economic benefits. Figure 8 illus-
trates the confi!luration of a 370-MWe axial-flow turbo-
machine. Dimensions of the unit are similar to those of
heavy-duty ope ,-cycle gas turbine units of the 50- to
100-MWe range I:urrently in utility generating service.
To achieve high cycle efficiency. the heat
exchangers for tha closed-cycle GT-HTG R require signif.
icantly greater total heat-transfer surface area than, for
comparison, th{ steam generators in a steam-turbine
plant of the sarre output rating. However, tube sizes are
smaller and the requirements are substantially less
demanding in k~y structural areas for the gas turbine
plant heat excha.l!Jers.
The recupentor experiences maximum shell-side gas
temperature of a little over 1,0000 F with a maximum
tube-wall pressure difference of about 550 psi under nor-
mal conditions. Minor leaks between the two helium
streams do not constitute a serious problem but result
only in some pl~rformance degradation, depending on
the quantity of leakage relative to the total loop flow.
I n the precooler the temperatures are even lower,
with a normal maximum gas inlet temperature in the
range of 4500 F for the dry-cooled plant. Pressure differ-
ence is only about 150 psi, with the highest pressure on
the helium shell side. Thus, a leak would result in helium
flowing into the cooling-water loop rather than water
leakage into the reactor. This can be readily detected
and promptly isolated.
Heat Rejection System
(\ key feature of the GT-HTGR plant is its adapta-
bility to economic dry cooling. The elimination of water
consumption requirements for primary cooling will
greatly increase potential plant siting options in many
areas, without otherwise attendant economic penalties.
Two natural-draft dry towers are used for the pre-
liminary dry-cooled l,100-MWe plant design. Design
studies suggest tower dimensions of about 300-ft base
diameter decreasing to a 200-ft top'section diameter
274
-------
Table 1. Gas turbine HTG R performance
at design point and rating point
Ambient temperature, of
Reactor inlet temperature, of
Reactor outlet temperature, of
Reactor flow rate, lb/hr
Reactor power, MWe
Compressor outlet temperature, of
Compressor inlet temperature, of
Compressor flow rate, lb/hr
Compressor pressure ratio
Compressor outlet pressure, psia
Turbine inlet temperature, of
Turbine outlet temperature, of
Turbine flow rate, lb/hr
Turbine pressure ratio
Net electrical output, MW
Power plant efficiency, percent
_.
De sign
point
ISO rating
point
75
59
934
922
1500
1500
14,505,000
.3000
14,199,000
3000
349
334
95
78
4,909,000
5,015,000
2.35
2.41
971
1000
1500
1500
1018
994
4,915,000
4,811 ,000
2.129
2.178
1069
1097
35.62
36.57
above the heat-exchange modules, and an overall height
of 250 ft. With selection of mechanical-draft cooling, the
dimensions would, of course, be greatly reduced. These
towers are substantially smaller and of less cost than
would be required for a steam-turbine plant. Smaller
towers can be used because the higher water inlet tem-
perature results in a large air-temperature rise which
reduces the air-flow requirements. The hotter air induces
greater buoyancy and thereby higher air-flow velocities
for a given height. Also, further economies can be
achieved in the heat-exchanger modules by using multi-
ple passes with the large water-temperature drop avail-
able.
Although there is little experience in the United
States with large natural-draft dry-cooling towers, there
is extensive process-industry experience with
mechanical-draft dry-cooling systems and, in addition,
experience at a 20-MWe power station at Gillette,
Wyoming. Also at Gillette, the 33Q.MWe WYODAK
coal-fired station is under construction for 1977 oper-
ation. It will also incorporate a direct mechanical-draft
air-cooled condenser system. An ERDA-sponsored
dry-cooling-module test program is planned for the
WYODAK station.
275
-------
PRECOOLER
PRESTRESSED
CONCRETE
AUCTOR
\I&SSH
ACCESS PENETRATION
:1
~ AUXILIARY COOLER
O PENETRATIONS
O VERTICAL.
PRESTRESS
~ TENDONS
VERTICAL COMPRESSOR
DISCHARGE CAVITY
PRESTRESS CHANNEL
RECUPERATDR
CORE
SECONDARY
CONTAINMENT
HORIZONTAL
PRESTRESS
TENODNS
Figure 7 . Three-loop 3,OOO.MW GT-HTG R power plant,
I n Europe, a number of large natural-draft dry
towers for elec'~rical generating plants up to 220 MW are
in service, datiilg to 1962 in the case of the 120-MWe
generating station in Rugeley, England. A natural-draft
dry tower is currently under construction for the
300-MWe THTR reactor at Schmehausen, a West
German high-tf mperature reactor that closely parallels
the General Atomic Company's 330-MWe Fort St. Vrain
HTG A.
Ell NARY PLANT DESIGN
System Design
The ammorlia balance-of-plant systems are similar to
a steam power plant in their functional design
requirements. Vapor systems include a main vapor cir-
cuit, a vapor bypass circuit, and a feedpump turbine
circuit. The condensate system includes a main conden-
sate feed circuit, a condensate bypass line, a condensate
storage and makeup system, auxiliary services, and a pre-
cooler isolationmd dump system.
The main vapor circuit transports supercritical am-
monia from the precooler to the secondary turbine and
condensers. Piping from the three precoolers is joined
into a single inlet line to the turbine. Temperatures are
controlled by regulating inlet flow to the precoolers.
Turbine stop' and bypass valves are provided as in most
steam plant to control power. Vapor flows from the
double flow turbine to twin condensers beneath the tur-
bine. The slightly superheated vapor must be desuper-
heated in the turbine exhaust.
CONCLUSIONS
The dry-cooled GT-HTGR plant offers the means
whereby power plant siting flexibility can be substan-
tially increased without efficiency penalty or cost disad-
vantage. With other advantages of the basic HTGR, the
result is a nearly environmentally neutral plant. Option-
ally, the binary-cycle GT-HTGR maximizes electrical
output from available cooling water and nuclear fuel
resources because of the very high efficiency achievable
by combining the closed-cycle gas turbine with a vapor
secondary power cycle. For either option, the resulting
power plant offers the highest available nuclear effi-
ciencies with lower generating costs and with minimum
276
-------
TO RECUPERATOR
(lP SIDE)
"
II.)
--.J
--.J
I'
t
Q
fa
FROM
REACTOR
J
J
..,
BEARING SPAN
27.5 FT (8.4M)
TO RECUPERATOR
(HP SIDE)
)
~
1;;.(;;~;.;;;r\,~1{1:1i~w~jl;~;;:"\ ,f~~:fI.{!t'~~;~~.Ih
COMPRESSOR (20 STAGES)
V OUTER 1-
O DIAMETER CAVITY
11.5 FT DIAMETER
13.5 M) 13.5 FT
Q (4.1M)
~
OVERAll MACHINE
lENGTH 39 FT (11.9M)
Figure 8. Three hundred and seventy-MWe GT-HTGR turbomachine.
-------
Table 2. Performance and growth of GT-HTG R
._~ . - - ---
GT-HTGR Mark I GT-HTGR Mark II
(1,500° F) ( 1 ,8000 F)
Dry-cooled Binary cycle Dry-cooled Binary cycle
Station effi ciency, % 36.6 45.3 42 50
Output from three- 1,100 1 , 360 1 , 400 1 , 730
loop sys tern, MWe
Output from" four- 1 , 470 1,810 1 ,860 2,300
loop syste.m, MWe
Reject hec..t relative 0.71 0.58 0.57 0.48
to LWR
environmental impact. Moreover, with the relatively
near-term poterctial for increased temperature, consider-
able increases in efficiency are realized as shown on table.
2, both for 1 he dry-cooled and binary-cycle plant
versions.
REFERENCES
of Germany," paper to be presented at the ASME
Gas Turbine Conference, New Orleans, March 21,
1976.
2. K. Sammert. "The Oberhausen Heat and Power
Station with Helium Turbine." address on the occa-
sion of the inauguration of the Helium Turbine
Power Plant of EVO on December 19, 1974, at
o berhausen-Sterkrade.
1. K. Sammett, and G. G roschup, "Status Report on
Closed-Cycle Power Plants in the Federal Republic"
278
-------
ENVIRONMENTAL AND SAFETY CONSIDERATIONS FOR A FOSSIL FUEL-FIRED
POTASSIUM-STEAM BINARY VAPOR CYCLE*
A. P. Fraas t and Robert S. Holcomb~
Abstrac t
Design studies indicate that superimposing a potas-
sium vapor cycle on a conventional steam cycle should
increase the thermal efficiency of a fossil fuel-fired plant
to about 50 percent, thus reducing the fuel consumption
by about 25 percent, and the rejection of waste heat to
the environment by about 33 percent. Systems have
been designed for burning gas or oil in furnaces that
would employ burners' that recirculate flue gases to
reduce NO lC release. High-sulfur coal would be burned in
a fluidized bed furnace fed with limestone or dolomite
to react with the S02 in the combustion zone and re-
duce the emission of S02 by about 90 percent. Electro-
static precipitators operating at a flue gas temperature of
about BOd' F (421' C) could be uStJd to control the
emission of particulates to a very low level.
The reactive character of potassium dictates that
careful attention be given to the design and construction
of the system to insure that the safety of the plant is
maintained. A novel design for the potassium con-
denStJr-steam generator has been evolved to minimize the
energy releaStJ of a steam leak in to the potassium. By
employing this design along with the high standards of
fabrication and the sensitive leak detection methods that
have been developed for liquid metal systems in the
nuclear power program, the probability of StJrious con-
StJquences from a leak should be kept very low. The
worst accident from the standpoint of the public would
be the release of most of the potassium inventory to the
stack. Such an accident is unlikely since the combustion
gas pressure in the furnace would be higher or nearly
equal to the potassium pressure in the boiler. To guard
against this contingency, a wet scrubber would be in-
cluded at the base of the stack so that it could be acti-
vated in an emergency.
*Research sponsored by the U.S. Energy Research and
Devalopment Administration under contract with Union Carbide
Corporation.
tPotassium Vapor Program Manager, Oak Ridge National
Laboratory, Oak Ridge, Tennessee.
:j:Project Design Manager, Oak Ridge National Laboratory,
Oak Ridge, Tennessee.
INTRODUCTION
For over half a century, up to 1950, the electric
power industry continually improved its new steam
plants to give a higher thermal efficiency. I n nearly a
quarter of a century since 1950 there has been essen-
tially no further improvement. In searching for a new
approach that would give a much higher thermal effi-
ciency for power plants of the future, A. P. Fraas pro-
posed in 1958 that a potassium vapor topping cycle be
superimposed on a conventional steam cycle (ref. 1).
This approach would yield a substantially higher overall
thermal efficiency for a given peak cycle temperature
than can be obtained with a Brayton cycle because of
the inherently higher efficiency of a Rankine cycle.
Further, by using potassium in the high-temperature
topping cycle, it is possible to avoid the high pressures
and corrosivity of steam in systems designed for opera-
tion in the temperature range above 1,000° F (538° C).
Studies indicate that this yields an unusually attractive
power plant for use with fossil, fission, and fusion heat
sources, and presents relatively tractable development
problems when compared with other advanced cycle
systems. .
Potassium is a soft metal; at room temperature it is
about the texture of hard butter. It melts at 144° F (63°
C) and boils at 1,400° F (760° C) at atmospheric pres-
sure. At 1,540° F (838° C), the turbine inlet temper-
ature currently projected, the pressure would be only 15
psi (1.0545 kg/cm2) above atmospheric pressure, and
the vapor density and other properties affecting the
design of the turbine would be about the same as for
steam at 240° F (116° C). After expanding the p~tas-
sium vapor through a turbine, it would be condensed <,t
about the same pressure and density as currently used ill
steam turbines, but the temperature would be about
1,100° F (593° C). The heat rejected from the potas-
sium vapor cycle would then be used to raise steam at
around 1,000° F (538° C). By following this approach it
should be possible to obtain a thermal efficiency of
about 50 percent from the combined potassium-steam
binary vapor cycle. When compared with conventional
steam plants this would reduce the fuel consumption by
about 25 percent and would reduce the waste heat rejec-
tion to rivers, lakes, or cooling towers by about 33 per-
cent.
Critical studies of this cycle indicate that it is theo-
retically sound (refs. 1-5). The big question is whether a
279
-------
practical, depundable system can be obtained at a
reasonable capital cost. Major elements in any such
evaluation are the safety and environmental problems
that will be posed by an unorthodox system of this sort
employing an alkali metal as a working fluid. Fortunate-
Iy, an extensive background of experience has been ob-
tained with alkali metal systems under the atomic energy
program, and over 200,000 hr of boiling potassium
system operati,)n was obtained under the space power
plant program (ref. 6). This experience coupled with
extensive experience with mercury in the same labor-
atories indicatE-s that a potassium vapor system should
present substar,tially less difficult operational and safety
problems than a mercury vapor system-and several
mercury vapor topping cycle plants were successfully
operated in the United States in the period from 1920 to
1970 (refs. 7,81. (The mercury vapor topping cycle was
dropped by U.::;. utilities in part because its upper tem-
perature limit proved to be 900° F (482° C) as com-
1540° F
POTASSIl:JM
TURBINE
pared to over 1,000° F (538° C) for steam, thus giving
steam an edge in thermal efficiency, and in part because
the price of mercury rose to the point where the capital
investment became excessive.
This paper presents an evaluation of the environ-
mental pollution control and safety problems of gas-,
oil-, and coal.fired potassium vapor topping cycle sys-'
tems. The material iri this paper is based to a large extent
on a more thorough treatment of these problems given
in reference 9.
SYSTEM DESCRIPTION
Conceptual designs have been made for gas- or oil-
fired systems (refs. 2,10) and for fluidized-bed coal com-
bustion systems ( ref. 11).
Gas. or Oil-Fired System
The basic layout of the system envisioned is shown
in figure 1. Three basic systems are coupled to form the
STEAM
TURBINE
GENERATOR
a:
~5 a: w
0 ~~
:J(j) I- 80°F
Viz ~~ <1w
(/)w Wo
<:fa <:fw I-z 900F COOLING WATER
I- z wz (/)0
00 I-w U
Q.U (/)<.::>
Cm~BUSTION 10000 F
CHAMBER TO
FEED PUMP 5 TACK
100°F
660° F
9000 F 2500F
FEED
GENERATOR HEATERS
GAS TURBINE
Figure 1. Flow sheet for the proposed gas or oil-fired potassium
vapor topping cycle.
280
-------
power plant, i.e., a gas turbine that supercharges the
combustion chamber for the potassium boiler, the potas-
sium vapor topping cycle, and an essentially standard
steam cycle. For purposes of this paper, the most im-
portant components are the furnace-potassium boiler
unit and the potassium condenser-steam generator unit,
both of which employ tube bundle modules as their
basic unit of construction.
Potassium Boiler
As can be seen in figure 2, the potassium boiler tube
0RNl-{)WG "-t2i28
,«GOO
t
THERMOCOUPLES
/ SUPPORT STRlJCTIJRE
i
- -' - BOilER DRUM
PRESSURE VESSEL
INSULATION -
OOWNQOMERS -- .
TUBE BUNDLE
. --" FLAME
- ~ BURNER
MANIFOI !l
FLAME SENSOR
r.OM9JSTtON AIR ~'> --
{PRIMARY I
012345
...~E >'=-T =--,
FUE.L
Figure 2. Section .through the full scale potas-
sium boiler tube bundle and burner module
being developed at DRN L under the NSF/
RANN program. The unit is shown installed
in a special casing for test purposes.
bundle module consists of two annular rows of l-in.
(25.4-mm) diameter stainless steel tubes arranged around
a 22-in. (558.8-mm) diameter, long, vertical, cylindrical
combustion chamber. The tubes are bent at the top to
provide flow passages between them so that the hot gas
from the top of the combustion chamber can flow radi-
ally outward and then downward through an outer an-
nulus over the outer row of tubes. To reduce the amount
of heat transfer surface area required, the furnace and
combustion chambers would be operated at a pressure of
5 to 10 atm, i.e., substantially above the 2 atm which
would be the design operating pressure of the potassium
vapor system. Potassium would circulate by thermal con-
vection from a header tank at the top of the tube bundle
through downcomers to ring-shaped manifolds at the
bottom, and then vertically upward through the tubes
back to a header drum and vapor separator at the top of
the header tank. Around 200 tube bundles of this type
would be required in a 600-MWe power plant.
Furnace
The boiler tube bundles would be mounted in four
to eight furnace shells in the form shown in figure 3. In
this layout the burner assemblies are independently
mounted in the bottom head of the furnace shell with
individual flanges so that any burner can be removed
without disturbing the others. The boiler tube bundles
are also independently mounted to a grid at the top,
which in effect serves as the top head. The design is such
that any tube bundle can be replaced independently of
the others by disconnecting it and withdrawing it
throu'gh either the top or the bottom of the furnace.
Potassium Vapor Turbine
The potassium turbine of figure 4 would be essen-
tially similar to a low-pressure steam turbine except that
more elaborate shaft seals would be required. These
shaft seals would employ a three-stage labyrinth S( al
between two face seals as in figure 5. In the labyrimh .
seal, argon buffer gas would be admitted to the cente r
stage and bled from the two bleedoff regions between
the three stages, one bleed consisting of air-plus-argon
and one consisting of argon-plus-potassium vapor.
Potassium Condenser-Steam Generator
The potassium condenser-steam generator unit
would be mounted directly beneath the potassium vapor
turbine in much the same fashion as steam condensers
are mounted under steam turbines (fig. 4). The steam
generator tubes would be of the reentry type indicated
in figure 6 with feedwater admitted to the bottom of a
central tube about 1/4 in. (6.35 mm) in diarT]eter in
which it would boil as it rose to the top of the tube (ref.
12). The steam emerging from the top of the inner tube
281
-------
KIn
T_I Jnsulatfd ~fh
IX Vapor Out"
Pn!ssurQ VClSSQ I
60ffle
Coolin9 Air
- 19 Tubll IOvncll1l5 A!r Fur-noce
!
: ~
, I
I
o I Z. , .. 6'
.._~-~.:~ ,
"""Le PT.
CDmbu:otlon GO';
Air In
NoTE: 2 Ouch RQo.
0+ 180.
Individual Tube Bundle
and lOr Burner- ReploCllmenf .
Copabilii"y
M 1<. III.
Fig,Jre 3. Section through a furnace-potassium boiler unit showing provisions
for the burner installation at the bottom and the boiler tube bundle mounting
arrangement at the top.
282
-------
~
Q)
Co)
- TUR81NE
I
/
..~ 8UFFE..-ED
SEAL IlEAIIIII'-'
8G
ITI~
fj~
8a
G~
~ffi
~~
8j
. - "'DTASSlUII-TO-STEA18 IOIL(III
0-1 v&PC)fI DUCT
ItM" - It!N[AUlt 1
""-II -MM(AT!R II
ft\.:A" vIlE- Of TU8E BuNDlE \.OCATIOfiII
QA....N TO REDUCED SCALE
REENTRY BOILER TU8ES
I
I
I t
I .
; VAPOR
INLET
DUCT
I ,
: '
, '
, '
I :
, I
i !
L
WATER IN-
STEAM OUT":
\
\
,":,
" ,
" ,
" ,
" '
:'
"
"
~
I
I
/'
--
~
',I
, "
,I
--
h
"
"
"
J,
,-- ------- ----..,
,
,
--
rn----------n~
"
,- ,
/
..,
,
I "
I.::
, "
..,
( ~ '
I il I
II I
I II .
l !i I
n
"
I
"
"
;:
,:
'"
"
,.
, "
I I,
I ;'
I ,I
, ,I
"
I II
II .'
II I '"
I I I'
I, I'
I II I"
I I. 'i
I II 1;1
: i: : II
I II I II
\ ;~ : I I
.~;---('r--;..~;" r-,
"J . II: L--------_____.i
,
I
I I'
, I
,
, ,
, I
r'-'C;:~ , :
)-{ -, I I
.: t. I !
) ~..t- -- - - -----_S"\:
'~
,
,
,
'~;.o:
L
:"_1 ::\'~--~------_u_----,...I'
,r-.: ;,
..J -
SCALE-FT
'0
,
POTA551UM TURBINE AND MEAT £ XCHANCiFA 'j,,".r
FOR BINARY VAPOR CYCLE POWEA PlA""T
Figure 4.
Potassium vapor turbine, condenser, and steam generator for
the reference design.
-------
ORNL ~G. 74-7062
FACE SEAL
L L.ARGON +AIR
L ARGON
ARGON + POTASSIUM
[Figure 5. Proposed potassium turbine shaft seal and bearing arrangement.
would then flow vertically downward in the annulus
between the inner and outer tubes and would emerge at
the bottom sU;Jerheated. Potassium vapor would con-
dense on the outside of the outer tube, and a film of
liquid potassium would flow down over the outside of
the outer tube to the sump at the base of the condenser.
The reentrant tube design was devised to avoid large
thermal stresse!. that might be induced by the high heat
transf~r coeffic'ents inherent in the boiling of water and
condensing of rotassium if these processes took place on
opposite sides of a single tube wall if the temperature
difference betwaen the boiling water and the condensing
potassium became too large under startup or transient
conditions.
Fluidized Bed Coal Combustion System
The flow sheet for the potassium vapor topping
cycle coupled to a fluidized-bed coal combustion system
is shown in figure 7. The arrangement is similar to that
for the oil.fired furnace. The major difference is that
clean combustion air instead of flue gas air is used to
drive the gas turbine. The air is compressed to about 110
psia (7.733 kg/cm2 I, heated to 1,500° F (816° C) in the
economizer, and then sent through the gas turbine. It
284
-------
!
.-4'"
1118.
I:
I
{i'~
Outer Tube
Inner .Tube
f I
...
QI
..
:I
'2
QI
180
Figure 6. Section through a reentry tube
for the potassium condenser-steam boiler
of the reference design.
leaves the turbine at about 32 psia (2.25 kg/cm2) and
980° F (52"1 C) and from there goes to the air plenum
of the furnace to serve as fluidizing and combustion air
for the furnace. I n this system the gas turbine is used
only to provide power for raising the combustion air
pressure high enough to make up the pressure drop
across the air distributor and fluidized bed,
-------
1500. F
1540. F
GENERATOR
COMBLSTION
CHAMBER
fl
POTA~;SI UM
BOILER
80.F
COOLING WATER 90. F
CYCLONE
SEPARATOR
1650. F
1500.F
ECONOMIZER
980'F
32 PSI A
600. F
110 PSIA
800' F
ELECTROSTATIC
PRECIPITATOR
COMPRESSOR
TURBINE
80. AIR
COMPRESSOR
STEAM
TURBI NE
FEED
HEATE
POTASSIUM
TURBINE
STEAM
CONDENSER
POTASSIUM
CONDENSER
STEAM
GENERATOR
FEED
PUMP
TO
STACK
250.F
660.F
FEED HEATERS
Figure 7. Flow sheet for the proposed coal-fired potassium vapor
topping cycle.
Economizer
The economizer will transfer heat from the flue gas
to raise the tempsrature of the combustion air to 1,500°
F (8160 C) so that it may be used to power the turbine
that drives the compressor and thus contribute to the
cycle thermal efficiency. The heat load of the economiz-
er will be about 20 percent of the heat input to the
furnace. The fl Je gas temperature will be reduced to
about 8000 F (427° C), a temperature level that should
present no special problems in the design of the electro.
static precipitator.
The other cJmponents in the coal.fired system, such
as the potassiur.' turbine and the potassium condenser.
steam generator, will be the same as those for the 9as-
fired system.
ENVI RONMENTAL EFFECTS
The potassium vapor topping cycle has the advan.
tage that it will reduce fuel requ irements by about 25
percent. thus reducing the demand on our fuel resources
and the adverse effects of mining and drilling operations.
Further, if a fluidized-bed coal combustion system can
be employed, high.sulfur coal can be used. The higher
efficiency also reduces both the total gas flow to the
stack and heat rejection to the environment. The waste
heat rejected from the steam condenser will be reduced
by about 33 percent. On the other hand, the inventory
of alkali metal poses a potential fire hazard and environ.
mental problems. These and related problems are dis-
cussed in this section.
286
-------
(OM8U:-'TIOf\J GI\~~S
117000F
----- -~
AIR
.... 600°F
<;=
- I< VAPOR
4"540. F
!. .";:
"
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I'
I'
~ d
II
II
II
II
II
Ii
'I
II
I,
I ,
. If;:: '
11
II
'I
'I
'I
"
:1
II
II
:,
:: I'
:1
"
'I
'I '
I I I, I
L- ..JI,._.
o ,
11111
':11' .'.
II
II
1\
II
II
I'
II
II
II
,"
II
II
II
'I
"
I,
I
II
II
"
. '
~
,
..
5~Le ...:~.
.
,
'0
,
..
,
FLUIDIO!eD B,w COMBU~T'O'" C<;A"o~R
~ POTA5SlU/oI\ BOILER
Figure 8. Conceptual design for a fluidized bed combustion chamber.
and potassium boiler.
287
-------
Potassium Lea.'c to the Atmosphere
The pota~.sium inventory of the plant .might con-
ceivably escapu into the gas turbine exhaust stream pass-
ing up the stack, and in this manner might contaminate
the atmosphen in the area with potassium hydroxide.
The possibility of such an event is much reduced by the
fact that the potassium boiler would normally operate at
only 15 psi (1 ::>545 kg/cm2) above atmospheric pressure
(well below tho! furnace gas pressure) and the potassium
condenser WOL Ie:! operate with a potassium vapor pres-
sure of only aoout 1.5 psia (.10545 kg/cm2). Further,
cutting off the fuel flow to the burners would cause the
potassium boilt:r temperature to drop rapidly, and when
it drops to 1 AOO° F (7600 C) it would be down to
atmospheric pmssure.
Emissions 0' Scdium and Potassium in Stack Gases
Some pers.)cctive on the magnitude of this hazard
can be obtained from a comparison with conventional
coal-fired stearn plants. The amount of sodium and
potassium in tile coal burned in such plants commonly
runs about 2 percent of the ash content, and the ash
content commonly represents about 10 percent of the
weight of the I:oal burned. Better than average precipi-
tators in the stack remove about 98 percent of the ash
from the stacl: gases. Thus, even if precipitators are
employed, in c )nventional coal-fired plants the weights
of sodium and ,:>otassium commonly present in the gases
leaving the stac ( amounts to about 0.004 percent of the
weight of coal burned. The average coal plant burns
almost 1 Ib/kWe hr (.45359 kg/3.6J1. or about
12,000,000 Ib/day (5,443,080 kg/day) for a 500-MWe
(500-MJ/sec) plant. This gives a discharge of about 500
Ib/day (227 kg/day) of sodium and potassium up the
stack, or about 4 percent of the total potassium inven-
tory of 10,000 Ib estimated for 1 of the 4 furnaces
envisioned for a proposed 600-MWe (600-MJ/sec) potas-
sium-steam binilry vapor cycle plant. Note that if no
attention is givJn to minimizing the potassium inven-
tory, the boile; design may give '\-20 times that of
OR N L design (refs. 2,5). Over 20 years of experience
with liquid meBI systems at ORNL indicates that one
would never expect to have as much as 4 percent of the
liquid metal invuntory get into the stack gases even in a
bad accident, pa rticul atly in a system in wh ich the pres-
sure in the furrace would be above that in the liquid
metal system. rhus, the release of alkali metal com-
pounds to the a1 mosphere in a bad accident would prob.
ably be no more than that emitted daily from con-
ventional coal-fi:'ed plants with precipitators. However,
the potassium wl)uld probably be emitted as the hydrox-
ide rather than a sulphate, so that it would present a
much more serious problem if a potassium leak occur-
red. To avoid trouble, a wet scrubber could be installed
at the base of the stack and activated if needed.
Stack Emissions Under Normal Conditions
The emission of pollutants up the stack under nor-
mal operating conditions would be very low for all of
the fuels being considered; gas, oil, or coal.
The gas burner built for the potassium boiler tube
bundle makes use of recirculation of part of the combus-
tion gas to lower the temperature in the gas flame and
thus reduce the amount of NOx produced. Tests on a
one-tenth scale model of the burner with flue gas recir-
culation of 12 percent gave an NOx emission of 35 ppm
at atmospheric pressure (ref. 13). This 'is equivalent to
100 ppm at the full power operating pressure of 8 atm,
well below the EPA limit of 165 ppm.
The gas burner will operate with an incoming com-
bustion air flow of 10 percent excess over the theoretical
air. The CO emission level found from the tests was 20
ppm for 10 percent excess air supplied and with flue gas
recirculation of 12 percent. .
Oil-Fired Furnace
The burner for the potassium boiler has not been
tested with oil but the results of one investigation indi-
cates that the NOx emission level of oil.fired furnaces
can be reduced about a factor of five with flue gas recir-
culation (ref. 14). A typical value for NOx emission
from conventional burners with no flue gas recirculation
is about 250 ppm. It might be expected that this level
could be reduced to 50 ppm for advanced burners at
atmospheric pressure. For the potassium boiler furnace
operating at 8 atm pressure the corresponding NOx emis-
sion would be about 150 ppm, slightly less than the EPA
limit of 165 ppm.
The CO emission level for the oil-fired burner is
expected to be similar to that found from the tests of
the burner with gas firing, Le., 20 ppm at 10 percent
excess air.
The S02 emission from an oil-fired potassium boiler
furnace will have to be controlled by the use of low-
sulfur oil. The use of high-sulfur oil would also probably
lead to unacceptably high corrosion rates on the flue gas
side of the boiler tubes. Limited tests indicate that high-
sulfur oil could be burned in a fluidized-bed furnace.
Coal-Fired, Fluidized.Bed Furnace
The release of NOx from a coal-fired fluidized bed
will be reduced below that of a conventional furnace
because of the lower combustion temperature in the
fluidized bed. This avoids the formation of NOx from
288
-------
the nitrogen in the combustion air. Nitrogenous com-
pounds are present in the coal, however, and NOx will
be formed from thissource. The amount of NOx formed
will depend on the nitrogen content of the coal and the
furnace operating pressure. The NOx emission level in
coal-fired fluidized beds has been found to decrease as
the pressure is increased, which is opposite to the trend
in gas-fired furnaces where the NOx is formed from the
gases in the air. Experimental results of the British
National Coal Board give values of about 400 ppm of
NOx at 1 atm to about 200 ppm at 3.5 atm (ref. 15).
Argonne National Laboratory workers found the same
results with the NOx level varying from about 400 ppm
at 1 atm to about 200 ppm at 8 atm (ref. 16). Their
results indicate that a furnace operating at 2 atm pres-
sure, as would be the case for the potassium boiler fur-
nace, would have an NOx emission level of about 300
ppm, or 0.4 Ib/106 Btu (0.172 kg/1 06 J), as compared to
the EPA standard of 0.7 Ib/106 Btu (0.3 kg/106J).
The emission level of CO from coal-fired fluidized
beds has been found to decrease as the furnace pressure
is increased. The results found by Argonne National
Laboratory were that the CO level decreased from about
2,000 ppm at 1 atm to 200 ppm at 8 atm (ref. 16). They
found a level of about 1,000 ppm at 2 atm pressure.
The principal advantage of the fluidized-bed furnace
is that high-sulfur coal can be used as a fuel and the S02
can be removed from the flue gas by reacting with cal-
cium in limestone or dolomite in the combustion zone
of the furnace. The S02 emission level will depend pri-
marily on the Ca/S mole ratio of the feed material. At a
furnace pressure of 2 atm and a temperature of 1,6500 F
(8990 C) a Ca/S mole ratio of about 2.5 should give a
reduction of S02 emission of about 90 percent (ref. 15).
For 3 percent S coal, for example, this would give an
S02 level of about 200 ppm. This corresponds to a re-
lease rate of about 0.4 Ib S02/106 Btu (.172 kg S02/
106 J). much lower than the EP A standard of 1.2 Ib
S02/106 Btu (0.5 kg/106 J).
The particulate level in the flue gas leaving the bed
was found by Argonne to be about 40 Ib/1 06 Btu (17.2
kg/106 J) for a velocity of 5 ft/sec (1.57 m/sec) (ref. 16).
A cyclone separator with an efficiency of 90 percent.
would reduce this to 41b/106 Btu (1.72 kg x 106J). An
electrostatic precipitator would remove about 98 per-
cent of the particulate matter in the flue gas leaving the
cyclone separator. The particulate emission level to the
exhaust stack would thus be of the order of 0.08 Ib/1 06
Btu (0.035 kg/1 06 J) or slightly less than the EPA stand-
ard of 0.1 Ib/106 Btu (0.044 kg/106 J).
SAFETY PROBLEMS
The most serious set of questions with respect to
the safety of the potassium vapor cycle is associated
with the possibility of leaks into or out of the potassium
system, particularly those that would lead to a potas-
sium-water reaction. This section presents first a discus-
sion of the nature of the failures to be expected, then a
review of experience with liquid alkali metal leaks and
attendant reactions, and then an appraisal of the prob-
able course of events if failures occur.
Stresses, Strains, Cracks, and Ruptures
A review of experience with tube failures in both
conventional boilers and high-temperature liquid metal
and molten salt heat exchangers indicates that anyone
of several different mechanisms may induce difficulty. A
common source of difficulty is corrosion on the combus-
tion gas side when the fuel has contained sulfur and
vanadium and temperature conditions have favored con-
densation of volatile corrosive materials on tube walls.
The potassium boiler will employ fuels that either con-
tain little sulfur or capture the sulfur with calcium to
minimize this sort of difficulty. Local overheating of
steam boiler tubes as a consequence of irregularities in
gas temperature and velocity distribution has been a
cause of many tube failures in superheaters, but this
type of problem should not be present in either the
potassium boiler or the potassium condenser steam gen-
erator of the reference design.
The recirculating potassium boiler should operate
with all of the surfaces exposed to hot gases thoroughly
wet with liquid potassium so that the heat transfer coef-
ficient on the boiling potassium side of the tube will be
many times higher than that on the combustion gas side,
and hence the tube wall temperature will be held close
to the boiling potassium temperature.
I n the potassium condenser-steam generator, the
tube wall metal temperature will be very close to the
potassium condensing temperature and hence will be
very uniform. By far, the most important source of tube
failures in liquid metal or molten salt heat exchangers
and an important source of failures in conventional
steam generators has been thermal strain cycling (ref.
17). This may be coupled with pressure stress effects in
high-pressure steam generators, particularly in the water
walls of pressurized furnaces (ref. 18). The bulk of the
failures to be expected in the heat exchangers for a
potassium vapor topping cycle will probably stem from
combined thermal and pressure stresses.
289
-------
Thermal Strain Cycling
Differentia thermal expansion in a complex struc-
ture commonly leads to high local stresses that are above
the elastic limH, or at least above the creep limit, and
hence local pla:itic flow takes place. A few cycles of this
sort ordinarily do not lead to difficulty, but many cycles
may induce what is called a low-cycle fatigue failure
(refs. 19,20). 111 high-pressure, low-temperature systems
a fatigue crack is likely to lead to a burst type of rup-
ture, but in low-pressure, high-temperature. systems
where the ratio of the ultimate tensile strength to the
design stress is much higher, low-cycle fatigue cracks of
this sort progress very slowly and simply result in a
gradually increasing leakage rate; they do not lead to a
burst type of failure. This is an extremely important
point in apprai!,ing the types of failure to be expected in
the heat excharlgers for a potassium vapor topping cycle
and the course of events associated with such a failure. It
should be noted that experience with high-temperature,
high-pressure, fossil fuel-fired steam generators is consist-
ent with this a~praisal; failures ordinarily develop slowly
and often do nl)t become apparent until the amount of
makeup water -equired to offset the leak becomes sub-
stantial. In fact, steam generators are commonly
operated for hl>urs or days after a leak becomes large
enough to result in losses of thousands of pounds of
water per hour Even with such large leaks, burst type
failures are uncommon. Thus, the type of failure to be
expected is a sl,)wly developing, low-cycle fatigue crack
that will yield a small leak, and if this leak can be de-
tected readily ill a liquid metal system, its effects will be
negligible and limited to the leak itself. ORNL experi-
ence with liquid metal and molten salt heat exchangers is
consistent with this appraisal.
REACTIC NS BETWEEN ALKALI METALS
AND AIR OR WATER
It has been stated that the greatest handicap faced
by alkali metal systems is the impression made upon
most of us as junior high school science students when
our teachers introduced us to sodium by throwing a
small chunk of the metal into a beaker of water. The
vivid picture left by these demonstrations has made
spontaneous fir 35 and violent explosions seem almost
synonymous wi',:h alkali metals. In point of fact, if there
is not water around, a liquid alkali metal fire is much
like a charcoal fire.
An excellert summary of a wide variety of experi-
ments made by many different investigators is presented
in reference 21 for work carried out up to 1965. Some
90 references an cited. Much work has been conducted
since that date. The following presentation attempts to .
give perspective to the problems and cites representative
cases from this large body of work.
Reactions with Air
Unlike hydrocarbons, the liquid metals have a very
low vapor pressure, and as a consequence oxidation can
take place only at the liquid surface; the liquid metal
does not vaporize to yield a combustible or explosive
mixture of air and vapor. As a matter of fact, an oxide
blanket tends to form on the liquid surface, and this
interferes with the oxidation rate such that if the liquid
metal is in a stagnant puddle in an air atmosphere, the
reaction rate will be inhibited by the oxide. The oxide
density is greater than that of the liquid, hence it tends
to sink in the liquid, reducing its inhibiting effects on
the reaction.
Reactions with Water
If no air is present, the reaction between water and
sodium, potassium, or NaK is rapid and violent, and
both hydrogen and steam are evolved. If the hydrogen
mixes with air, an explosive mixture is formed and, once
a flame front starts in a hydrogen-air mixture, it is likely
to develop into a detonation wave so that a violent
explosion can ensue (ref. 22). This is the usual course of
the reaction in a high school science demonstration; the
sodium-water reaction commonly proceeds very vig-
orously for some time before ignition occurs and then,
when a substantial quantity of a hydrogen-air mixture
has accumulated, this ignites, detonates. and it is the
hydrogen that is responsible for the explosion.
Reactions with Steam
Because of its much lower density, steam reacts
with alkali metals at a much lower rate than water.
Steam is commonly used to clean out the residual liquid
metal from components of alkali metal systems when
they are removed for inspection.
Experiments with Small Scale Units
A series of small-scale experiments run at Mine
Safety Appliances Company, (refs. 23,24) has yielded
some valuable insights into the possible effects of leaks
in an alkali metal-heater generator. These tests entailed
injection of water into alkali metal or vice versa with
pressure differentials of the order of 200 psi at metal
temperatures around 6000 F (3160 C) using orifices
from 1/8 to 1/2 in. (3.175 to 12.7 mm). The results
indicated that, for the system tested, if a leak occurs in a
sodium- or Nak-to-water heat exchanger fitted with a
blowoff valve or a rupture diaphragm and a stack, the
peak pressures developed were only a little higher than
those for which the valve or diaphragm was designed. It
290
-------
was found that if the liquid metal temperatures were
above 6000 F (3160 C), there were no violent pressure
surges and there was smooth relief of the gas evolved via
the pressure relief valve (ref.. 22). Further, if the relief
valve can be located in the water system, the peak local
temperature in the reaction zone can be kept relatively
low because the heat released will go into vaporizing the
water (ref. 23). If the relief valve is placed in the water
system, the liquid metal system must be designed to take
the same peak pressure as the water system-a step that
will not be attractive.
It must be emphasized that the sequence of events
and the magnitude and frequency of pressure and tem-
perature fluctuations may vary widely from one system
to another as a consequence of differences in system size
and proportions, including surge volumes, size and
length of connecting piping, location of relief valves, etc.
......
..-i
\1 500
~
~ 400
~
,
,
~. ~ It ~'" ~
.~
!~ 'V f\ 17\ ~
~ ;
i \r "-~ / "\
I
'"', ~.' '\
I \
.
, \
i
~
! "'"
'..'.... .'''.
" ,
'.
'-
NaK Temp, = 6000F and 50 psi -
HaO Temp. = 200°F and 400 psi
Nozzle = 0.622" ID -
---- ./ -K .,.... HaO
-
I . . , I
900
800
700
600
300
200
100
o
4
o
2
, .
Differences Between Sodium, Potassium, and NaK
The tests at the Mine Safety Appliances Company
were carried out with sodium, NaK, and potassium. The
shape of the pressure-time curves for typical simulated
leaks between liquid metal and water systems differed
somewhat between the tests with sodium and those with
NaK, but there were no important differences between
the curves for NaK and those for potassium (ref. 23).
Typical sets of curves of this sort are shown in figures 9
and 1 O.
Experiments with Large Scale Units
A particularly fine series of experiments has been
run in the United Kingdom using both 1/6-scale and
full-scale tube banks to investigate the effects of leaks in
sodium-heated steam generators (refs. 25,26). The tests
were run with leaks designed to simulate the full range
ORNL DWG. 74-7068
6
TIME (see)
8
12
10
Figure 9. A typical curve showing the system pressures as a function of
time after initiation of injection of water into NaK in tests at Mine
Safety Appliance Company ,(reprinted from ref. 22).
291
-------
,/ ,- ~
- -
--v' -;-<" .,' ~ .......... ......... ,~" \
".". .'
(." '- 0... 0' -.;. ...... """-
-
V"
,.-- .J'
''''''''''
I
I
6000F .
K Temp. = and 200 psi
HaO Temp. = 200°F and 400 psi
Nozzle = 0.364" ID -
- K ........ HaO
-
I I I I I
900
800
700
600
'rl
U) 500
p.
~
U) 400
((J
~
11.
300
200
100
o
o
1
2
ORNL DWG. 74-7069
3
TIME (see)
4
5
6
Figure 10. A typical curve showing the system pressures as a function of
time after initiation of injection of water into potassium in tests at Mine
Safety Appliance Company (reprinted from ref. 22).
of effective orifice sizes that might be experienced in
service, Le., from pinhole leaks of 1 ml/min to a wide
open tube. The h:ak duration in the test ranged from as
much as an hour for the pinhole leaks to only a few
seconds for the large-scale leaks. The systems were
instrumented to give a detailed insight into the transient
local temperatu:es and pressures. The instrumentation
included strain~Hges on the heat exchanger shell. The
test work was carried out in close conjunction with
analytical work directed at predicting the effects of
changes in operating conditions.
Results of the tests were generally consistent with
the tests described above carried out by the Mine Safety
Appliance Company, but much more detailed insights
into the conditi Jns to be expected in a full-scale heat
exchanger were obtained. Detailed examinations of the
heat exchangers after the tests were made to determine
the character of the reaction products. This work indi.
cated that about half of the hydrogen from the water
that reacts with the sodium is blown off as gaseous
hydrogen, and the balance remains combined either as
NaOH or NaH. The hydrogen bubbles may go into solu-
tion in the sodium depending on the amount of hydro-
gen released. The NaOH and NaH tend to settle out and,
because of their higher melting point tend to freeze in
the lower temperature regions.
Test Conditions
The tests were carried out with water pressures up
to 2.300 psi (161.69 kg/cm2) with initial sodium tem-
peratures generally in the vicinity of 9000 F (5000 C).
The tests were run with plain carbon steel tubes, ferrittic
292
-------
chrome-moly tubing, and austenitic stainless steels. Parti-
cular attention was given to corrosion effects when the
heat exchanger tubes were examined subsequent to each
leak test. The carbon steel tubing showed substantial
corrosion, the chrome-moly tubing showed a mild
amount of corrosion, and there was very little corrosion
of the stainless steels. Tests run by other experimenters
have sometimes turned out less favorably. One major
factor is the length of time that elapses before the cor-
rosive products of the reaction are cleaned out of the
equipment.
Damage to Tubes Adjacent to a Leak
A major concern in attempting to envision the
course of events in the event of a leak in the steam
generator is damage to adjacent tubes and the possibility
of a rapid spread of the failure. This may occur as a
consequence of corrosion, erosion, or local overheating.
The temperatures that develop in the reaction zone may
reach around 2,4000 F (1,3000 C), and this may so
weaken adjacent tubes as to lead to failures. Whether or
not such failures occur appears to be very much a func-
tion of the character of the jet that emerges from the
leak. For example, tests involving 2,300 psi (161.69
kg/cm2) water jets in full-scale tube bundles led to fail-
ures when the jets impinged at close range on adjacent
tubes (ref. 25). These failures occurred in a matter of a
few seconds, am;! hence provided a mechanism for pro-
gressive failures of the sort that occurred in the NaK-
heated steam generator of the Fermi reactor plant (ref.
27). This phenomenon, commonly referred to as "tube
wastage," places a strong premium on detecting the
. development of such a leak very quickly so that the
. pressure on the water side can be relieved and the water
and sodium system pressures brought into pressure equil-
ibrium (ref. 28). When this has been done, the reaction
rate drops sharply and the leak may even plug with
sodium oxide when the pressures in the two systems
equalize (ref. 23).
If steam free of liquid leaks into the liquid metal
side of a steam generator, the reaction between the
steam and the liquid metal takes place at a greatly re-
duced rate. This happens partly because the density of
the steam is much lower than that of water, greatly
reducing the rate at which the two can react at the inter-
face, and in part because the hydrogen formed inhibits
the movement of steam to the liquid metal surface by
convective circulation (ref. 29). The reaction proceeds
quietly without the violent pressure fluctuations and
noise commonly associated with liquid water-sodium
reactions. This in turn suggests that steam generators for
liquid metal systems should employ a reentry tube steam
generator design (ref. 11) of the sort shown in figure 6
so that any leak that might develop would consist only
of steam with essentially no liquid water. When this is
the case, there should be no problem with tube wast'age
because the. latter is a phenomenon that apparently
involves both erosion and corrosion, and requires the
presence of liquid water in the jet emitted from the leak.
This thesis is borne out by experience with steam leaks
in the superheaters of conventional power plants; if the
region around the leak is reasonably free of ash deposits,
the steam jet from the leak will not "cut" the adjacent
tube.
Implications of Leak Test Experience
A number of important points emerge from a review
of the experience gained from investigations of leaks in
an alkali metal-heated steam generators. These observa-
tions may be summarized as follows:
1. Both analyses and test experience indicate that
a high-pressure jet of steam leaking into a tube matrix
filled with alkali metal commonly will not induce fail-
ures in adjacent tubes, whereas high-pressure jets of
water usually induce leaks in adjacent tubes.
2. If a condensing alkali metal vapor is employed
to heat a steam generator rather than a liquid alkali
metal, the probability of a high-pressure steam jet induc-
ing a failure in an adjacent tube should be very low, the
pressure fluctuation associated with the leak should be
minor, and the temperature excursions in the reaction
zone should be much less than if the jet were going into
liquid metal.
3. There is a strong incentive to detect the leak
before large quantities of reaction products can be form-
ed because these may plug the lines and make drainage
and cleanup of the system difficult.
4. Detection of a leak into an alkali metal con-
denser can be made much more rapidly than detection
of a leak into a liquid alkali metal system if the con-
denser is made to concentrate the noncondensibles in a
pocket at one end. This stems from the much highH
velocity of the liquid metal vapor and the much morn
effective collection of noncondensibles in a condenser
than of gas bubbles in a liquid system.
5. It is highly desirable to limit the inventory of
both liquid metal and liquid water in an alkali metal-
heated steam generator so that the quantities of material
that can react can be limited by closing valves or other
operator action.
As will be discussed further in the following sec-
tions, the referehce design for the potassium condenser-
steam generator considered here takes advantage of all of
the above observations to obtain a design that should be
singularly unlikely to develop a leak, and if one devel-
ops, it should be much less sensitive to the leak than a
liquid metal-heated steam generator.
293
-------
ANTlCII)ATED FAILURE MODES IN THE
POTASSIUM TOPPING CYCLE
Operation,1I and safety problems were major factors
in the design rhilosophy on which the system of figures
1 through 5 was based. Every effort was made to mini-
mize the pos:;ibility of a failure, and, if one should
occur, to minirnize its consequences. An effort was made
to take advantage of the fact that, in the system pro-
posed, the pressure in the potassium system will be be-
low that of ei':her the combustion gases in the furnace-
potassium boiier units or the steam in the potassium
condenser-steam generator units. As a consequence, if a
leak develops, leakage will be into the potassium system
rather than fro'n the potassium system out into the com-
bustion chamb,!r or the steam generator. .
Combustion GC/s Leak into the Potassium Boiler
A leak 01 combustion gases into the potassium
boiler will lead to the formation of potassium oxide in
the recirculatirg liquid potassium. Little or none of this
potassium oxice will be carried out of the boiler into the
balance of the system. The low pressure in the potassium
boiler implies ~cw-pressure stresses and hence, if a leak
does develop, it will probably stem from thermal strain
cycling and call be expected to progress slowly. This in
turn means that the damage can be limited if the leak
can be detected in the early stages. This is particularly
important bec.)use corrosion in the liquid potassium
system will prcceed fairly rapidly if the oxygen content
of the potassiL m builds up to more than 500 to 1,000
ppm. On the other hand, if the leak can be detected in
its early stages, the corrosion will not have an opportuni-
ty to become appreciable, the quantity of reaction
products ~ill be small enough to remain in solution so
that "crudding up" of the system will not be a problem,
and the repair operations can be limited to sealing the
leak itself and :'emoving the oxygen from the potassium.
(Sealing of the Icak in a heat exchanger tube is likely to
require remove. I and replacement of a complete tube
bundle module.)
Steam Leak into the Potassium Condenser
If a leak dove lops in the steam generator, the steam
jetting into the potassium condenser will react with the
potassium vapor to form potassium oxide and hydrogen.
Inasmuch as the potassium condenser will have a large
vapor volume .;pace available, there will be plenty of
space for the t.ydrogen and no explosion or even large
increase in pre!;sure will occur. (This situation is com-
pletely different from that in a liquid-metal-heated
boiler in which there is no free volume on the liquid
metal-side and into which the hydrogen from the reac-
tion can expand.) As the hydrogen builds up in the con-
denser, it will block the flow of potassium vapor into the
condenser and produce a back pressure on the potassium
turbine. Only a few pounds of steam leaking into the
condenser would produce a marked change in the tur-
bine back pressure, which would be obvious to an oper-
ator and could be used to trip a warning signal.
Although extremely unlikely, if a large steam leak
were to develop as a consequence of a burst type of
failure, the inherent nature of the double-walled tube
configuration and inlet orificing of the inner tube of the
reentry tube boiler is such that vapor rather than water
would be injected into the potassium region (fig. 5), and
the rate of injection would be relatively low-about 4
Ib/sec (1.816 kg/sec) per ruptured tube. This would lead
to an increase in the pressure in the condenser at a rate
of about 1 psi/see (.0703 kg/cm2 -see), or about 60
psi/min (4.218 kg/cm2-min). Thus, if the potassium
condenser were designed to take an internal pressure of
60 psi (5.218 kg/cm2), and if the flow of either steam or
potassium into the condenser could be stopped within 1
min of the first evidence of the rupture, the damage
would be limited to the broken tube. For the extreme
case, assuming the abrupt, complete rupture of a steam
generator tube, the potassium condenser pressure would
rise to about 4 psia, or about three times its normal
values, in about 3 sec. This would be easily and reliably
detectable and could be the basis for closing valves in the
feed water supply line. If this were done in an additional
10 sec, the inventory of superheated water in the boiler
design proposed would be exhausted in another 15 sec in
the course of the "coast-down," and the peak pressure in
the potassium condenser would be held to about 15 psig
(1.0545 kg/cm2). To protect against the contingency
that no action might be taken, a rupture disc should be
provided to blow off at perhaps 40 psi a (2.812 kg/cm2).
It should be emphasized that a complete rupture of
a tube appears highly unlikely; all experience to date
indicates that a leak would develop gradually as a result
of thermal strain cycling and, if the leakage were suffi-
cient to be detectable, the system could be shut down in
an orderly fashion before any great amount of leakage
had occurred. Thus, it is apparent that there is a strong
premium to detecting trace amounts of steam or gas
leakage into the potassium system.
LEAK DETECTION
SteafrJ Leaks Into Potassium
One good way of detecting just a trace amount of
steam leakage into the potassium condenser would be to
294
-------
design the potassium vapor flow passages to sweep any
noncondensibles to a small region at one end of the
potassium condenser in much the same fashion as is
done in conventional steam condensers. In the latter, the
noncondensibles, largely air, are swept to a small region
at one end of the steam condenser and the air is removed
with a steam ejector. Similarly, a potassium condenser
could be arranged so that noncondensibles would be
swept to a small region in which a hydrogen sensor could
be located. This might take the form of a nickel or pal-
ladium window separating the potassium vapor region
from an evacuated region. An ion gauge in the vacuum
region would quickly register the presence of hydrogen
that would diffuse through the palladium window if
there were hydrogen present in the potassium vapor. A
device of this sort is under development for the LMFBR
(ref. 30). .
Gas Leaks Into Potassium
It is difficult to measure quantitatively the amount
of oxygen in liquid potassium. However, experience with
NaK cold traps indicates that it would be possible to
make use of a sensitive plug indicator in the potassium
system if a few percent of sodium were added to the
potassium. This would reduce the vapor pressure over
the potassium by a small amount-less than 1 psi (.0703
kg/cm2) for around 3 percent sodium in potassium-but
it would provide a sensitive indication of the presence of
oxygen in the potassium.. Further, it would provide a
means for oxide removal by cold trapping, and this
would be advantageous from the standpoint of other
aspects of system operation. Alternatively, an electro-
chemical, on.line O2 concentration meter is being devel-
oped for sodium circuits under the LMFBR program,
and this may prove suitable for potassium systems.
In practice, a cold plug indicator probably ought to
be employed in connection with each tube bundle in the
furnace-potassium boiler units.. Thus, about 39 pipes
about 1/2 in. (12.7 mm) in diameter would have to be
carried out from the 39 tube bundles in the design pro-
posed, and would have to penetrate the furnace pressure
vessel wall. Note that these penetrations would not be
subject to the stringent leak.tightness requirements of
the potassium system because they simply contain com-
bustion gases.
STEAM EXPLOSIONS
The possibility of a steam explosion, although small,
is inherently present in any steam boiler. In the type of
boiler proposed, the inventory of superheated water in
the boiler is only a few percent of that present in con.
ventional coal- or oil-fired steam boilers, and hence
should reduce the hazard potential associated with this
type of accident to a much lower level than in conven.
tional coal.fired steam boilers.
FUEL FIRES
There is always a fire hazard associated with the
inventory of fuel maintained at a power plant. If the
amount of fuel in storage is determined by the number
of days' supply considered necessary, a plant of a given
power output but with a higher thermal efficiency will
require a smaller inventory of fuel. The proposed potas-
sium-steam vapor cycle will have a thermal efficiency
about double that of a conventional gas turbine plant,
and hence would require about half the fuel inventory.
If compared with a conventional oil.fired steam plant,
the fuel inventory would be three.fourths of that for the
conventional plant.
REFERENCES
1. W. R. Chambers, A. P. Fraas, and M. N. Ozisik, "A
Potassium-Steam Vapor Cycle for Nuclear Power
Plants," USAEC Report ORNL-3584, Oak Ridge
National Laboratory, May 1964.
2. A. P. Fraas, "Preliminary Assessment of a Potas-
sium-Steam.Gas Vapor Cycle for Better Fuel Econ-
omy and Reduced Thermal Pollution," ORNL-
NSF.EP-6, Oak Ridge National Laboratory, August
1971.
3. J. G. Collier, "Current Status of Methods of Recov-
ering Energy Directly fro~ Liquid Metal Circuits""
AERE.M-1724,1966.
\
4. B. Wood, "Alternative Fluids for Power Genera-
tion," paper presented at the IME Meeting in
London, April 8, 1970.
5. R. J. Rossbach, "Final Report: Study of Potassium
Topping Cycles for Stationary Power," GESP-741,
NASA Contract No. NAS3-17354, General Electric
Company, Energy Systems Programs, November 13,
1973.
6. H. C. Young and A. G. Grindell, "Summary of
Design and Test Experience with Cesium and Potas-
sium Components and Systems for Space Power
Plants," USAEC Report ORNL-TM-1833, Oak
Ridge National Laboratory, June 1967.
7. W. l. R. Emmet, "Mercury Vapor for Central Sta-
tion Power," Mech. Eng., 1941.
8. H. N. Hackett and D. Douglass, "Modern Mercury
Unit Power Plant Design," Trans. ASME, Vol. 72,
No. 89 (1950).
295
-------
9. A. P. Fraas, "Operational, Maintenance and Envi-
ronmental Problems Associated with a Fossil Fuel.
F ired Potassium-Steam Binary Vapor Cycle,"
ORNL.NSF.EP-30, Oak Ridge National Laboratory,
August 1974.
10. M. E. Lad.ey, "Second Iteration Analysis of a Fossil
Fuel-F ired Gas Turbine.Potassium.Steam Combined
Cycle," CiRNL-NSF-EP-39, Oak Ridge National
Laboratorlf, Ju Iy 1973.
11. A. P. Fra3s, "A Fluidized Bed Coal Combustion
System G:>upled to a Potassium Vapor Cycle,"
paper presmted at the 65th Annual Meeting of the
AIChE, New York, November 26-30. 1972.
12. A. P. Fraas, "A New Approach to the Design of
Steam Generators for Molten Salt Reactor Power
Plants," USAEC Report ORNL.TM.2953. Oak
Ridge Nati:>nal Laboratory, June 1971.
13. J. K. Arand, "Subscale Gas Burner Test Report,"
Report No. 41-99, KVB Engineering, Inc., January
1973.
14. Lubomyr lCurylko, "Control of Nitric Oxide Emis-
sions from Furnaces by External Recirculation of
Combustioi1 Products," paper presented at the
ASME Wir.ter Annual Meeting, Washington, D.C.,
November :Z8-December 2, 1971.
15. "Final Report on Reduction of Atmospheric Poilu.
tion," Fluidized Combustion Control Group, Na-
tional Coa. Board, London, England, September
1971, (pre~)ared for the Environmental Protection
Agency.)
16. G. J. Vogel et aI., "Reduction of Atmospheric Poilu.
tion by thE! Application of Fluidized-Bed Combus.
tion and Regeneration of Sulfur.Containing Addi-
tives," ANLYES-CEN-1007, Argonne National
Laboratory Annual Report, July 1973-June 1 Q74.
17. A. P. Fraas, "Estimating the Reliability of Sys.
tems," USA.EC Report ORNL-TM-2200, Oak Ridge
National Laboratory, May 1968.
18. T. H. Gladr.ey, and H. S. Fox, "TVA's Power Plant
MaintenancH Program; Philosophy and Experience,"
presented a:: the American Power Conference, April
18-20, 1972.
19. L. F. Coffin, Jr., "An Investigation of Thermal
Stress Fatigue as Related to High Temperature Pip-
ing Flexibility," Trans. ASME, Vol. 79, No. 1637
I (1957).
20. A. P. Fraas, and M. N. Ozisik, Heat £xchan.Qer De-
sign. John Wiley & Sons, Inc., 1965.
21. J. A. Ford, "Literature Review of Sodium-Water
Reactions," Atomic Power Development Associates,
Inc., APDA-167, March 15, 1965.
22. E. C. King and C. A. Wedge, Jr., "Reaction of NaK
. and H20," NP-1423, Mine Safety Appliances
Company, 1950.
23. E. C. King, "The Reaction of NaK and H20,"
NP-3646, Mine Safety Appliances Company, Sep.
tember 7,1951.
24. E. C. King, "The Reaction of NaK and H20,"
NP-334,Mine Safety Appliances Company. Sep.
tember 7,1951.
25. J. A. Bray, "A Review of Some SodiumNVater Reac-
tion Experiments," J. of the British Nuclear Energy
Society, Vol. 10, No.2 (April 1971), pp. 107-119.
26. A. Lacroix, J. Lions, and M. Robin, "Safety Investi.
gations on Sodium Water Steam Generators - Calcu.
lation Methods and Experimental Results," Proceed-
ings of the International Conference on the Safety
of Fast Reactors, Aix~n-Provence, September
19-22,1967.
27. John Graham, Fast Reactor Safety. Academic Press,
1971.
28. D. J. Hayes and G. Horn, "Leal< Detection in Sodi-
um Heated Boilers," Journal of the British Nuclear
Energy Society, Vol. '10, No.1 (January 1971), pp.
41-48.
29. C. C. Addison and J. A. Manning, "The Reaction of
Water Vapour with Liquid Sodium, Sodium Per-
oxide, Sodium Monoxide, and Sodium Hydride;
Vapour Pressures in the Sodium Hydroxide.Water
System," Journal of the Chemical Society, D ecem-
ber 1964.
30. "Reactor Development Program Progress Report for
April-May 1971," ANL-7825, June 23, 1971.
296
-------
A TOPPING CYCLE FOR COAL-FUELED ELECTRIC POWER
PLANTS USING THE CERAMIC HELICAL EXPANDER*
B. Myers, R. Landingham, P. Mohr, and R. Taylort
Abstract
We advocate ceramic helical expanders as the work
output element in a 2,50d' F direct coal-fired Brayton
topping cycle for central power station application.
When combined with a standard steam electric power
plant cycle, such a cycle could result in an overall ther-
mal conversion efficiency in excess of 50 percent. We
enumerate the performance, coal tolerance, and system-
development-time advantages of the ceramic helical
expander approach. We provide a perspective on the
choice of design and materials.
A preliminary consideration of physical properties,
economic questions, and service experience has led us to
a preference for the silicon nitride and silicon carbide
family of materials. A program to confirm the perform-
ance and coal tolerance aspects of a ceramic helical
expander system is planned.
INTRODUCTION
In the United States, about 25 percent of our fuel
consumption is in the area of electric power generation.
Most projections show that total fuel consumption will
double before the end of the century and that the per-
centage used for the generation of electricity will
increase also. A projection for the year 1990 is shown in
figure 1.
The present average thermal efficiency of the elec-
trical conversion process is about 34 percent, and for
modern fossil-fueled plants it is 40 percent. Technologies
for increasing the conversion efficiency in the power
plant to the range of 50 to 60 percent are near at hand,
as is well known. In the main, these new technologies
involve the addition of a topping cycle to new or exist-
ing conventional steam plants. A topping cycle increases
efficiency by extracting additional energy from a fixed
quantity of fuel. A perspective on the prospect for
progress of power plant thermal efficiency with time is
given in figure 2. An increase to 50 percent could enable
us to conserve as much fuel in the year 2000 as we now
consume for electric power generation.
-This work was performed under the auspices of the U.S.
Energy Rel4iarch & Development Administration, under Con-
tract No. W-7406-Eng-48.
tLawrence Livermore Laboratory, Livermore, California.
The principal contending concepts for the topping
cycle are:
the Brayton-topped steam cycle or gas-turbi ne
combined cycle with cycle temperatures above
2,300° F.
th eRa nk i ne -to pped s tea m cycle or potas-
s i u m -vapor-turbine combined cycle with cycle
temperatures of -1,500° F.
the open-MHD-topped steam cycle with cycle
temperatures of -4,800° F.
Each has its advantages and disadvantages. All, however,
share in common the challenge of accommodating the
available fuel resources, particularly the more abundant
coals.
We would like to describe an alternative in the
Brayton-cycle family that makes use of a novel ceramic
engine, a helical rotor expander, which is presently
unexploited. We see evidence that this system can be
built from available ceramic materials in a shorter time,
with greater fuel tolerance, and at lower cost than com-
peting systems. It would provide topping-cycle service at
temperatures approaching the limit for conventional
heat sources.
SOME PERSPECTIVES ON HELICAL EXPANDERS
We have for some time considered alternative con-
version machinery that might accomodate the strength
prop~rties and relatively low debris toierance of ceramic
structures (ref. 1). The most promising candidate identi-
fied is the helical rotary machine originated by Lysholm
(ref. 2). This positive displacement machine, operating in
the expansion mode, has a combination of character-
istics that are favorable for ceramic construction:
Characteristic centrifugal stresses, as compared witll
those in a bladed turbine, are down by as much as a
factor of 10 because of lower tip speed.
Rotor temperatures are significantly lower than the
inlet temperature because of the cyclic nature of the
device.
The geometry is rugged, providing high large-debris
impact resistance.
Small debris such as fly ash has a low velocity
relative to the rotor and can pass freely through the
machine without appreciable erosion potential;
there is virtually no way to design a ceramic blade
af'ld vane immune to rapid erosion or impulsive
damage from debris.
297
-------
I'.)
co
00
1M
~~~~~~~~~
~"0:'N'{cLEA~~~~~"" ' ELECTRICAL CONVERSION LOSSES
32.0"" '" " "" GENERATION 33.0
'" '" 49.6
HYDROElECTRIC1'11~: . /:~
GEOTHERMAL 0.8 /) 'If . ~ ~ .
. '., _. .' .~..o.: t/j~ ~/2.0 . ~ j.1 RESIDENTIAL
';~H~[:~.{.::,;t;;:7~.'? "',J.'\'...::.!.«',:.:.'){.}.?'.:'.:'~[\.:~.:~.~~.; -..':...' : '/'.,:, :'"i:/./ CO:~:.C'AL
POOT.{j;/'~':::" .~.: ,:'>""'.,. -. :". "',',,' ....,.X:.... . S..
::AL ~~:,,!~"""""":'" '-:::U::.",:,-
-~~~~~;:s~4'.~:~T.IA'
~~uA,~.~. J. /~:: -to
. . -.4 - ~ NONENERGY
~PETROL~UM-a:N.G.L'--.-=-= :- --=-==- - -=-. ---: - -. - .~=- ~ 10.4
c- - 29.0- - - - - - - - - - - ..";""~0.1
I- - - -'- - - -'-- --...:.. - -- - - 8'~1.4'\.'
~ - - -: - - - - - - - - - - - - TRANS
C =. ~~=-~;-- _~7- - .-~~~~~POR;:.~ION
- - -- - - - - - - - - - - - - - - - - - - -
- - - - .-
~- -
~-~ IMPORTS
~-- 24.1
~-
\
REJECTeD
ENERGY
75.9
/7.8
/ ~/9.1
~ ~ / / 24.8
~ ~)' .
g. )Q
I / ~ ~ ;.xXX
/ /'23.3 x
usefu~
ENERG'y
6'8.7~
Xy
x..)f
x,
y
1<
Lx
~ 8.
.,
~
X}Y
XXYYY
Figure 1. The U.S. energy flow in 1990, modeled after the National Petroleum
Council intermediate-demand case. All values are in 1015 Btu (2.12 x 1015
B!t: == i06 bbl/day of oil). Total energy consumption is 148.6 x 1015 Btu.
yX
I
)
-------
100
~
>,
u
c:
QJ
U 50
~
~
QJ
....
~
~
QJ
.s::.
I-
Waste heat (typical)
Advanced
Fossil-fuel
steam ""
o
Gas
work (typical)
1950
1800 '
1900
Range for
topped
steam cycles
turbine
2000
Year
Figure 2. Power plant efficiency trends, actual and projected.
An exploded pictorial view of the helical expander
is shown in figure 3. Its operation, like that of a recipro-
cating engine, involves admission, cutoff, expansion,
release, and exhaust phases as illustrated in figure 4.
Axial-flow gas turbi.nes develop power symmetri-
cally in the tangential direction. Blade and spool thrust
loads are present, but they can be carried to any arbi-
trary thrust.bearing location,' and they can be balanced
by compressor thrust or passive discs. In helical expand-
ers, by contrast, flow is essentially diagonal, as can be
deduced from figure 4. The gas enters in the top front
pocket, as shown in the view at the left, and it progresses
toward the sides and back end, releasing, as shown in the
view at the right, when expansion terminates as the
shaded pocket reaches the opposite end of the machine.
There are radial loads in the machine due to the
integral of the pressures over all rotor surfaces. There is
also thrust and torque due to the unbalanced pressure on
the net area of leading flanks as opposed to trailing
. flanks in each chevron. The differential area is consider.
ably smaller than the total so that turning forces are
small in comparison to the radial loads. The radial loads
are highest on the male rotor and toward the inlet end.
A bearing buried in the rotor or stiff rotor extensions
would probably be necessary to reduce dynamic effects
. and to keep the alignment necessary for close clearances
between rotors and case. The design of these high-speed'
bearings and the associated seals will constitute an
appreciable part of any development program.
The power per unit of rotor diameter is similar for
expanders and turbines of similar power output. The
screw expander runs at about a third of turbine tip
speed, but the gas enveloping regions occupy a much
larger fraction of the cross-sectional area. In addition,
the expander has two rotors instead of one.
The principal inefficiencies in the expander are
associated with the intake and exhaust processes, the
leakage between cutoff and release, and the bearing and
seal losses. The leakage is due to rotor and case clear-
ances and inlet end seals. Tighter clearances and seals
reduce the leakage, but they also increase mechanical
losses and design complexity.
Figure 5 shows test data on adiabatic efficiency
(T/ad) versus the pressure ratio for a 21-in. helical com.
pressor used in the first. U.S. Navy experimental gas
turbine (ref. 3). A helical expander of this size would
deliver about 5 MW under our topping cycle conditions.
We note in figure 5 that T/ad has a fairly broad peak in
299
-------
Figure 3. Exploded view of the helical expander.
High.pressure gas
is forced into the
pockets, hetween
the lobes of the
rotors and the
case
Turning cuts off the
intake port and
traps gas in the
pockets
Further turning
increases the volume
of the pockets
allowing expansion
to occur
Discharge occurs
when the exhaust
ports are uncovered
by the lobes
Figure 4. The operation of a helical expander.
300
-------
~.tric .ffici~ncy
.. 3300
----
8 8 1200 rpm
.~k
~
100
>,
u
c:
QI
U
'r-
'+-
'+-
QI
U
'r-
~
+J
~
::J 50
,...
o
>
"0
c:
"'
U
'r-
+J
"'
.Q
"'
'r-
"0
cc
0
1.0
rpm .
.
2500 rpm
.~8-
..
.
.-
..
-.~
8-
3300 rpm.
From C. R. Soderberg et al.. Trans. SNAME. 53: 249-289. 1945.
1.5
2.0
2.5
3.0
Pressure ratio
Figure 5. Test data on the adiabatic efficiency versus pressure ratio for 21-in.
helical compressor (from ref. 3).
the region of 85 percent for the higher speeds. At this
pressure ratio, as will be seen in figure 6, an expander
should have the same efficiency, which means that it
would be deficient by six to eight points compared to a
corresponding turbine. The falloff of T/ad for the com-
pressor at 1,200 rpm is probably due to greater leakage
for the longer time of expansion.
Figure 6 shows an estimate of the effects on T/ad of
pressure ratio, size, and expansion or compression oper-
ating mode for a constant-tip-speed family of machines
(ref. 4). The gain with size is due to surface-to-volume
effects, the larger sizes of geometrically similar machines
having less relative leakage area. Further advantage can
come from scaling clearances at less than linear. An opti-
mum pressure ratio r is shown because at low r, the
power increases more rapidly with increasing r than does
the leakage, and at high r, where some of the leakage is
at sonic velocity, it increases more slowly with increasing
r. The persistence of the efficiency of the expander
mode with increasing r is mostly due to the fact that
expanders are generally run at higher temperatures than
compressors so that the relative leakage is lower.
A perspective on the steady-state radial temperature
profile across the rotors in the expander is shown sche-
matically in figure 7. The temperature rar:'ge in the left
zone is set by the requirements of the topping cycle, and
the base temperature in the right-hand zone is set by the
lubricating-oil temperature limit for conventional bear-
ings. The intermediate profile depends on the relati'le
thermal conductances of the gas fil m, the rotor, and the
oil film. The ripple shown on the outside portion of th~
rotors is associated with the cyclic gas temperature, the
heat transfer conditions at the surface, and the thermal
diffusivity of the rotor material. It is seen that in con-
trast to that for axial flow machines, the peak steady-
state rotor temperature can be considerably lower than
the expander inlet temperature. Figure 7 also shows
combustion and ash temperatures, material strengths,
and thermal stress indices for reference.
The. .temperature drop across the hollow rotor as
shown in figure 7 gives rise to thermal stresses that are
compressive on the outer surface and tensile on the inner
surface. These stresses are superimposed on the rota-
tional stresses, but they are different in character be-
301
-------
100
~ 80 ~ Expander
>, 100 hp -=
V
I::
QJ 60
'r-
u
'po.
4--
'+-
Q)
u 40 -Compressor
'po.
+-J
rtI
.0
rtI
'po. 20 From H. R. Nilsson, private
"C
C; communication, 1965.
0
1 2 3 5 10 20
Pressure ratio
Figure 6. Qualitative size and mode effects for screw machines.
cause they will tend to relax somewhat with time as
creep adjusts the rotor geometry very slightly to accom-
modate thermal expansion. Stress analysis of the rotor is
further complicated by its geometry, stress concentra-
tions, residual Hresses, the natural variation of physical
properties with temperature, and any compositional
changes with rotor radius.
MATERIALS REQUIREMENTS
FOR HELICAL EXPANDERS
Each of thl! advanced coal-fired energy-conversion
technologies sp<)ken of in the beginning has unique
materials requirements. The helical expander requires
structural ceram cs with predictable strength in the range
of 2,500° F, adequate for rotating power-conversion
machines. The ::J1~bris tolerance advantage has already
been rnentionec. Also, such machines might run un-
cooled and theraby gain a performance advantage over
cooled metal alloy counterparts. -
One of the }overning considerations in this machine
performance tradeoff is illustrated in figure 8, which
gives the strength-to-density ratio (a/p) versus tempera-
ture. This is a governing criterion for rotating machines
where the inerti,,' stresses dominate. The dashed line at
80 x '03 in. rep:esents a typical requirement for a/p for
long life in axial-flow turbines. It is seen that the nickel-
base alloys are limited to something like ',600° F by
this criterion. All present air-breathing gas turbines and
the great part of those advanced models envisioned for
the future are designed for the ',6000 F or lower metal
temperature, with air or water cooling as required. This
cooling is a loss to the cycle, reducing the gain from
higher inlet gas temperature (T 3)' It finally prevails in
setting a particular limit- for T 3 in the 3,200° F region.
Silicon nitride or silicon carbide are advocated for
the hel ical expander construction material. They are
seen to have a temperature advantage of 8000 or 900° F
over the nickel-base alloys for the 80 x 103 -in. alp cri-
terion. Actually, as we have noted, the characteristic
centrifugal stresses in the expander are much lower than
for turbines. The resulting large margin on rotational
stress would be used to accommodate thermal stresses
and the brittle nature of these materials.
There are two other material regimes on the figure:
graph,ite, which is another 8000 F higher at the same a/p
level, and the refractory alloys (Nb, Mo, Ta, WI. which
are in the same temperature range as the ceramic mate-
rials, but with about half of the a/p, mostly because of
their increased density. We dismiss both of these other
material classes from the present expander material con-
siderations because of the principal difficulty of taking
302
-------
4500
4000
Service limits as shown
for alp = 8x104 in.
Graphite
Stoichio-
3500 metri c
combustion
3000
Silicon carbide
I.L.
o
Expander
inlet
I 2500
UJ
cr::
:::>
~
~
UJ
i: 2000
UJ
~
Ash melts
Sil icon nitride
Ash
softens
Expander
exhaust
1500
1000
Thermal stress
differential
(see insert
for index)
500
o
Combustion
products
(incl. ash)
Gas film Rotor ripple Rotor dc
- Rotor-
640 (Nickel base
320 alloy)
Graphite
160
80
40 Sil icon nitride
Silicon carbide
20
10 Alumina
Thermal stress index
W/in. .
oK (l-~)
Eex:
Oil 1 imit
Oil fil m
Lube oi 1
as coolant
Radius
ZONE
Figure 7. Schematic radial temperature profile for a screw expander under steady-
state operating conditions with various temperature constraints.
303
-------
c
'fo
\1'("
.().....
~\1'
~?
("
t1;..
d9>
\1'..,
':..'
'"
o
~..
200 -
a.
.......
\;)
o
.,-
~
t'O
~
~
100 -
~
:bC9
"',I,
6'
J't.
fI).
~)
C9J'
J'
J't.
C9C91
'r-
III
C
~
~
o
..J
50-
In
In
OJ
~
~
VI
20-
10
120;)
1600
2000
"'i'il'J!I~lliil!,~~I~III~
Tungsten
2400
2800
3200
-Minimum alp for axial gas turbine applications
OF
Figure 8. Comparative alp data (temperature (OF) for 1 percent creep in 10,000 h).
care of the oxjdation problem in coal-fired expanders.
Graphite migh t be used if the combustion products
could be kept nverlastingly on the fuel-rich side and the
ash content kept very low. Such a graphite system de.
serves separate ~ttention, because of the very large work-
ing temperatul e advantage. Either alternate material
might be used with inert gas as the working fluid, but
that just trans',:ers the oxidation problem to the heat
exchanger and presumably reduces the peak cycle
temperature.
We conclude, therefore, that the greatest advance is
to be had in p'-lrsuing the development of ceramic ma-
chines, in partcular the helical expander because, it is
best suited to the capabilities of ceramic materials now
available or SOOIl to be available.
There are iI number of refractory compounds that
can probably meet the primary mechanical and physical
properties critnia (strength, stabil ity, thermal strain
tolerance, and compatibility with coal combustion
products) for tt e helical expander. We will not dwell on
all of the altern ~tives here, but we have considered most
of the oxides, nitrides, carbides, and intermetallic com-
pounds, especi;llIy of silicon, aluminum, beryllium,
,
boron, titanium, and zirconium. We have made our ini-
tial selection of silicon nitride and silicon carbide, not
only on the basis of attractive properties, but more
especially on the basis of fabricability and economics.
To our knowledge, only these two among the candidates
, with superior properties have been demonstrated to be
practicable in the large sizes desired for the screw ex-
pander application, being at the same time economical
and in plentiful supply. In the most elementary form,
the fabrication process for both silicon nitride and sili-
con carbide parts involves powder consolidation to the
rough final shape by any of a number of methods, bond-
ing or partial conversion of the powders to give suffi.
cient strength for green machining at high removal rates,
and then final conversion. For both materials, the
shrinkage on final conversion can be held to around 0.1
percent. Both would be finished by diamond wheel
grinding. For each, there are maximum wall thickness
levels based on tolerable times and temperature differ-
ences during reaction bonding and powder conversion.
These levels are below the 1-ft or so half.thickness of
characteristic features of full-scale rotors so that they
would have to be significantly cored.
304
-------
EXPERIENCE WITH HIGH-TEMPERATURE
STRUCTURES AND MACHINES
We believe that two earlier experiences illustrate the
potential for near-term development of ceramic expand-
ers.
The first experience with structural ceramics design
came in connection with an advanced version of a
2,200° F gas-cooled nuclear reactor. A support was re-
quired for the 150-psi air-pressure-drop load across the
reactor over a free span 36-in. in diameter. The design
concept chosen W8S an arch configuration that elimi-
nated tensile stress in the ceramic material. A seven-piece
self-locking assembly shown in figure 9, was made of K T
silicon carbide. Prototype elements were also fabricated
of reaction-bonded silicon nitride in the approximate
"
-"7Mf
shape of a 24-in. hexagon with the same hole pattern as
shown in figure 9. The silicon carbide assembly was
proof tested in jet aircraft fuel combustion products at
2,700° F for over 30 h (ref. 5). The steady pressure-drop
load was 150 psi. Cyclic loads at 10 to 50 Hz with a
30-psi ampl itude were superi mposed on the steady load.
The structure was essentially unchanged by the test.
Later, with a challenge to design a machine for
extracting power from a high-temperature gas-cooled-
reactor heat source, we thought about the best way to
minimize the tensile stress problems inherently present
someplace in any rotating power-conversion equipment.
We wanted to make best use of ceramic and graphite
materials having a statistical distribution of strength and
low impact resistance. We chose the helical expander on
the grounds which have been cited, namely, low tensile
!"
~
,...
a
...
Figure 9. Ceramic support designed to withstand a 150-psi air-pressure-drop load
over a free-span 36 in. in diameter.
305
-------
stress and rugged geometry. A small helical expander and
Gompressol' demonstration machine was constructed of
graphite fer testing in argon at a temperature of 2,750°
F. The rotors were 4 in. in diameter. A view of the
rotors, a s,Jbassembly of the two helical machines, and
the furnace setup is given in figure 10. The expander
produced power from an electric generator on the out-
side of the furnace by means of a Brayton cycle. The
system accomplished its 300-h demonstration mission
successfullv (ref. 6). A discussion of the brittle-materials
criteria used in designing the machines is given in refer-
ence 7. A follow-on study considered the performance
. of a high-'.:emperature reactor power plant using the
graphite ex.'ander (ref. 8).
CERAIV IC HELICAL EXPANDER MATERIALS
IN COAL COMBUSTION PRODUCTS'
The combustion of coal produces CO2, steam, ash,
unburned C1ar, and many other substances. Coal com-
bustion pro:!ucts have proven to be corrosive for a wide
variety of cmwentional metal and ceramic materials. Sili-
con nitride and silicon carbide have not been tested in
this regard. They have demonstrated remarkable corro.
sion and thlirmal shock resistance in jet fuel combustion
products flowing at Mach 1. In one series of tests, they
were heated at 2,190° F for 1 h and then cooled for 5
min (with the cycle repeated 100 times) without failure
(ref. 9).
We havl: noted that ash velocities in the expander
are relativel,! low, i.e., in the vicinity of the rotor tip
speed. The kinetic energy will therefore be down by an
order of me gnitude or so compared to that of ash in
turbines. Th e effects of ash flow on ceramic materials
are expected to be substantially different than on
metals.
We have initiated a series of experiments to investi-
gate the rate of erosion and corrosion of silicon nitride
and silicon c.u'bide in flowing coal combustion products
and ash. The effects of the gas composition, ash temper-
ature, and th,~ temperature of the material will be tested.
Other variables to be investigated. are test material
composition, ash composition, ash particle size, and the
angle of impact of the ash. Erosion by particles in a
stream is the ;ubject of a number of recent investigations
(refs. 10,11).
Coal-fired boilers are subjected to fouling from con-
densation of various combustion products. Control of
boiler foul in!! is a well-developed art at conventional
plants. It has been learned that there can be a guide to
the control of fouling in a ceramic expander. Helical
expanders he ve been operated in geothermal brines
where fouling is a severe problem. Salt deposits did form
on the expander rotors, but they "wore-in" to very low
clearances without detrimental effects (ref. 12).
THE HELICAL EXPANDER COMBINED CYCLE
A block diagram of the helical expander combined
cycle is shown in figure 11. The cycle starts with a
state-of-the-art axial-flow bladed compressor with a com-
pression ratio of 10 and an adiabatic efficiency of 89
percent, supplying air to a coal burner.
We presume for the figure that the fuel preparation
and burner boxes together would include some emission
control features. Actually, we are investigating the two-
stage fluidized bed burner (refs. 13,14,15), which has
the potential for good ash scavenging in the bed. The
cyclone burner is also to be considered. It is in use today
in large coal-fired plants and removes 90 percent of the
ash as molten slag directly in the burner. Its use to date
has been limited to coals containing low-melting ash.
The relatively low combustion temperature should not
cause any NOx problems. We think of exhaust gas scrub-
bing for final SOx control and precipitators for final ash
control as required.
Hot gases with some ash carryover would be sup-
plied to the expander at 9 atm and 2,450° F. We project
at this time that a reasonable maximum size for a ceram-
ic rotor is about 5 or 6 ft, and two units of that size
should deliver about 90 MW net electrical output for the
conditions shown. Such expanders may have an adia-
batic efficiency in the high 80 percent range according
to extrapolations of figures 5 and 6. We entered an over-
all expander efficiency of 75 percent in the calculations
so as to allow a margin for the mechanical efficiency of
bearings, seals, and timing gears.
The expander exhaust goes to a convection-
dominated boiler to raise steam for a standard steam
cycle. Matching the expander exhaust energy to the
requirements for final feedwater heating, evaporation,
superheating, and reheating of the steam gives a steam
cycle output of about 100 MW. The pinch temperatu,'e
or the minimum difference between hot gas and steam
cycle states was presumed to be 100° F. An artist's
depiction of this system is shown in figure 12, with a
two-stage fluidized bed burner and the topping cycle
compressor operating on the steam turbine shaft for
better rotational speed matching.
Based on figure 11, the combined-cycle efficiency
from burner input to steam cycle rejection is.
Input - Reject
1/ =
C Input
- 565
- 1,200
= 0.53.
306
-------
~
External drive assembly
-- ~~_.
..
Graphite compressor-
expander case in furnace
Graphite rotors
Figure 10. Views of a graphite helical expander and compressor test assembly
ran successfully at 14,000 rpm for 300 h at 2,750° F in an argon atmosphere.
307
-------
.~ -- _. _.. -- ---.
, Coal Fuel prep.
I
1200 MBtu/h
-
Axial flow 9atm Heli ca 1
compressor Burner 24500 F . expander f--- 86
l1ad = 0.89, l1ad I1mech = 0.75
r = 10
1 atm
Air 15600 F
.
Standard 3500/
1050/1000 4 -
steam cycle
Temp. pi nch - 100° F
\
MW
100 MW
565 MBtu/h
Figure 11. Block diagram of the helical expander combined cycle
(efficiency = 1 - 1 ~g~ = 0.53).
This is well into the 50 percent range that we postulated
in the introduction.
CONCLUDING REMARKS
We are currently underway on the necessary mate.
rials development and expander design studies for this.
system. The near.term goal is construction of a 70. to
120-kW expander operating at an inlet gas temperature
of 2,500° F for performance and compati,bility studies.
Favorable findings will lead to construction of a 6. to
10-MW, 24-irl., intermediate version of the expander and
later of a full-size 50.MW machine.
We intend to emphasize high efficiency and the
ability to OpE rate directly on coal combustion products. ..
REFERENCES
1. P. B. Mohr, and F. Rienecker, Topping Cycles and
AdvanceQ Conversion Machinery for Central Power
Stations, Lawrence Livermore Laboratory, Report
UCRL.75,)75, Rev. 1, 1973.
2. A. J. R. Lysholm, "A New Rotary Compressor,"
Proc. Inst. of Mech. Engr., Vol. 150, (1943). Pp.
11-16.
3. C. R. Soderberg, R. B. Smith, and A. T. Scott, "A
Marine Gas Turbine Plant," Trans. Soc. Nav. Arch.
Mar. Engrs.. Vol. 53, (1945), pp. 249-289.
4. H. R. Nilsson, private communication, 1965.
5. Pluto Quarterly Report No. 13, Third Quarter,
1962, Lawrence Livermore Laboratory, Report
UCRL-7079,1962. .
6. W. M. Wells, D. W. Hanner, J. L. McElroy, and E.
Robinson, "High Temperature Testing of Graphite
Helical Screw Expanders and Compressors," Journal
of Spacecraft and Rockets, Vol. 4, No.6 (1967), p.
761.
7. E. Robinson, Some Problems in Estimation and
Application of Weibull Statistics, Lawrence Liver-
more Laboratory, Report UCRL-70555, 1967, and
E. Robinson, Some Theoretical and Experimental
Aspects of Design with Brittle Materials, Lawrence
Livermore Laboratory, Report UCRL-7729, 1964.
8. W. M. Wells, "On Central Station Application of I
308
-------
i
~
~ ..
-"- "'. .~
~ %.i
, ~
I~"~"'"",,'
..::.,. ~
....r\
fl.' .
. .it\
l1li\,
.-
.,
~,
Topping cycle
compressor
r ,
expander'
! ~I'
1 f ,
\
j
-.
=
---'1
,.
-~, _:~-I'
I
.-
~'IO-
~
--
'"
r.
t
'*
Ir.......
,
~:
I i-~
.\ " ~ \.
,.
'At'
-
Figure 12. Artist's conception of a power plant incorporating the helical expander
in a combined cycle.
309
-------
Graphite Helical Rotor Expanders," Proceedings of
the Sacond Energy Conversion Conference, Miami
Beach Florida, 1967.
9. W. A. :5anders, and H. B. Probst, "High Gas Velocity
Burner Tests on Silicon Carbide and Silicon Nitride
at 121)()o C," in Ceramics for High Performance
Applic.ations, Proceedings of the Second Army
Materials Technology Conference, Hyannis, Mass-
achusetts, 1973, J. J. Burke, A. E. Gorum, R. N.
Katz, l!d., 1974, pp. 493-531.
10. G. L. Sheldon and A. Hanhere, "An Investigation of
Impinnement Erosion Using Single Particles," Wear,
Vol. 21, (1972), p. 195.
.11. I. Finilie, "Erosion by Solid Particles in a Fluid
Strearr:' ASTM Special Technical Publication No.
307, ASTM, Philadelphia, 1962.
12. H. Weiss, R. Steidel, and A. Lundberg, Performance
Test of A Lysholm Engine, Lawrence Livermore
Laboratory, Report UCRL-51861, 1975.
13. A. A. Godel, i'Oix Ann~es d'Application de la Tech-
nique de Combustion du Charbon en Lit Fluidise;"
Revue General de Thermique, Vol. 5, No. 5~
(1966), pp. 349-359. .
14. A. M. Squires, "Clean Power from Coal," Science,
Vol. 169, No. 3948 (19701. p. 821.
15. J. Yerushalmi and Arthur M. Squires, "The Ignifluid
Boiler and the Godel Phenomenon," Proceedings of
the EPRI Clean Fuels Conference. Monterey, Cali-
fornia, 1974, page 46-51.
NOTICE
"Th is report was prepared as an accou nt of
work sponsored by the United States Govern-
ment. Neither the United States nor the United
Stetes Energy Research & Development'Admin-
istretion, nor any of their employees, nor any
of their contractors, subcontractors, or their
employees, makes any warrenty, express or
implied, or assumes eny legal liability or respon-
sibility for the accuracy, completeness or use-
fulness of any information, apparatus, product
or process disclosed, or represents that its use
would not infringe privately-owned ri!jhts."
310
-------
BALANCE-Of-PLANT POWER REQUIREMENTS fOR ADVANCED POWER SYSTEMS
Irving L. Chait*
Abstract
The growth in unit size of central station steam
power plants in the last 20 years has been accompanied
by a parallel and even more rapid growth in the auxiliary
power required to operate the unit. From a nominal 4 to
6 percent of rating, onsite auxiliary power has steadily
increased and, today, often takes 10 to 12 percent of the
plant's output. The increases have resulted from a com-
bination of environmental regulations and more severe
operating conditions.
Many proposed advanced power-conversion plants
include, in addition to the usual auxiliaries, process
equipment not found in today's plants. The. power
demand of all these auxiliaries may exceed the totals of
present-day plants and must be included in determining
the overall performance of advanced power systems-
particularly when comparisons are made among such
systems.
To achieve the efficiencies and environmental ac-
ceptance desired, power plants using advanced power
conversion systems-such as MHD generation, fuel cells,
advanced steam and gas turbines in various combina-
tions, fusion reactors, and solar collectors-are combined
with topping and bottoming plants, which look very
much like present-day plants. For equally cogent
reasons, they are integrated with coal gasification and
fuel-conditioning processes, cryogenic refrigeration,
magnets, oxygen plants, scrubbing systems, seed re-
covery, and similar energy-consuming auxiliaries. The
power load imposed by these additional processes may
significantly reduce the overall plant thermal efficiency.
From the viewpoint of energy and environment con-
servation, considerable benefit can be achieved by
optimizing the balance-of-plant systems. Efforts for such
improvements should proceed on parallel paths with,
and at the same pace as, the efforts' being made in devel-
oping the main advanced power-conversion systems.
INTRODUCTION
Modern central-station, steam power plants have
grown in unit size over the years, and so has the auxili-
ary equipment required to operate them. The major
-Manager, Advanced Power Systems, Burns and Roe, Inc.,
Hempstead, New York.
power demand in the balance-of.plant systems in coal-
fired plants has always come from the draft fans, circula-
ting water pumps for condensing exhaust steam, coal-
. conditioning and -handling equipment, soot-blowing
, , systems, and condensate pumps. Boiler feed pumps, the
, largest single service requirement in steam electric power
plants, are usually turbine driven because of the econom-
ic benefits, increased net output of kilowatts, and about
a 60 Btu/kWh improvement in heat rate.
Connected and operating electric power demands of
the auxiliaries ,of typical fossil and nuclear plants are
shown in tables 1 and 2. The operating loads are about
65 percent of the connected load in the fossil plant and
70 percent in the nuclear plant. Table 1 indicates that
the auxiliary power operating load at design capability
for coal-fired plant is 7.6 percent of plant gross rating.
This plant fires a low sulfur coal and has no scrubber. A
scrubber would add about 50 percent to the auxiliary
load. Table 2 indicates that the auxiliary power oper-
ating power demand for a nuclear plant is 4.7 percent of
the generator rating. Steam bottoming plants for ad-
vanced power conversion systems will have many auxili-
aries similar to those shown in tables 1 and 2.
The growth of the utility industry has been parallel-
ed by the proliferation of regulations requiring the addi-
tion of new auxiliaries to meet the environmental cri-
teria. Electrostatic precipitators, scrubbers, cooling
towers, waste treatment systems, water management and
other environmentally related facilities are becoming as
ubiquitous in power stations as fans and water pumps.
Simultaneously, operating conditions have become more
severe, and lower quality fuel has been increasingly USE-d.
Thus, from a level of about 4 percent of gross output for
the onsite usage of power, the auxiliary power of present
day plants now often amounts to 10 to 12 percent (ref.
1). For a 1,OOO-MW (net) plant with an heat rate of
10,000 Btu/kWh, as the auxiliary power increases from 4
to 10 percent, the thermal efficiency decreases from 34
to 30 percent. A loss of efficiency of this magnitude
wipes out decades of developments and improvements in
power plant technology.
Many advanced power conversion plants-particu-
larly the ones in which fossil fuels are required, such as
advanced steam and gas turbines, MHO, fuel cells, com-
bined cycles, etc.-will, of necessity, be integrated with
process technologies not common in present-day power
plants. These technologies include fluid bed furnaces,
311
-------
Table 1. Auxiliary power requirements (motors 25 hp or larger only)
for 42D-MW coal-fired steam turbine generator plant
Connected load Operatiny load
(hp) (hp
Forced d,'aft fans 16,000 10,600
Induced draft fans 12,000 9 , 800
H.P. service water pumps 2,000 675
Boiler startup pump 700
Gas reci rcul ating fans 6,000 5,520
Low-pressure service water pumps 900 405
Closed cooling water LP p~mps 800 360
Condensate pump 1,600 711
Boiler circulating pumps 5,200 5,024
Isolated phase bus blower 120 54
Stator cooling water pump 80 36
Auxiliary oil pump 50
Control air compressors 250 113
Condenser air exhauster 250 113
Service air compressor 200 180
Seal oil reheater 27 ~kW)
Carbon dioxide heater 36 kW)
Glycol pumps 250 113
Motor-generator set 120
Ash pit sump pump 400 180
Gas recirc. fan turning gear 150 68
Closed cooling water booster 50 23
HVAC supply fan 100 90
Ventilating supply fans 300 270
Roof exhausters 150 138
Prec. area fan 100 90
Circulating water pumps 2,000 1 ,732
Screen wash pumps 200 180
Fuel conditioners 1 , 800 1,200
Ignition oil pumps 100
Crusher 500 450
Con veye r 905 725
Cyclones 350 262
Precipitator 3, 155 1,276
Hydraulic fluid pump 300 180
Heater drain pump 125 113
Sea 1 ai r hlower 40 36
Total hp 57,292 40,717
kW =hpx o~746 44,989 31,974
0.95
Percent rating 420 MW 10.7 7.6
312
-------
Table 2. Auxiliary power requirements (motors 25 hp or larger only)
for 1,1 OO-MW nuclear power plant
Connected
load
(kVA)
Operating
load
( k VA)
Condensate pumps
Condensate booster pumps
Circulating water pumps
Pl an t servi ce wa ter pumps
High-pressure core spray pump
48O-V auxiliaries
Reactor recirculating pumps
Cooling tower fans
Low-pressure core spray pumps
Reactor HR pump
Standby service water pumps
Makeup water pumps
CRD pumps
Substation auxiliariesa
Total kVA
Percent rating (1,230,000 kVA)
3,315
7,725
13,211
"2 ,678
2,588
225
15,194
10,000
1 ,330
2 , 1 75
3,060
2,868
340
18,000
82, 709
6.7
2,210
5 , 150
13,211
1 ,339
75
15,194
7,318
1 ,434
276
11 ,680
57,887
4.7
coal gasification or liquefaction plants, cryogenic refrig-
eration for superconducting leads and magnets, scrub-
bers, and seed and sulfur recovery processes. The new
processes, superimposed on most of the "old" auxiliary
systems, will undoubtedly increase the fraction of power
absorbed by the auxil iary systems. I f more than 10 to 12
percent of all power generated is to be used inside the
power plant before it is transmitted to any customer, the
savings in energy expected by the application of some of
the higher efficiency, advanced conversion systems may
be disappointing. Increases in auxiliary power not only
result in a requirement for larger installed unit capability
but also reduce the net efficiency of a plant. To arrive at
realistic levels of efficiency attainable by the various
combinations of advanced power conversion systems and
their associated process technologies, a consistent and
complete determination of balance-of-plant require-
ments is necessary whenever a complete plant is con-
sidered.
The limits of the balance-of-plant systems for an
advanced power conversion cycle plant are necessarily
somewhat vague. Even in present-day plants, the bound-
aries may be defined differently, depending upon the
purpose for which they are established. All the energy
expended-in gathering, processing, treating, transport-
ing, generating, converting, and transmitting a natural
resource to the ultimate consumer of the heat, light, and
power-should be included if the true net power avail-
able from a plant is to be determined. If, in addition to
cost faCtors: energy cons"ervation and environmental
benefits are to be considered primary desiderata for
advanced power conversion system plans, they must be
evaluated on as equivalent and comparable bases as
practicable. The auxiliary systems for the advanced
power systems discussed hereinafter are indicative of the
magnitude only of their onsite pow~r requirements.
MAGNETOHYDRODYNAMIC POWER PLANTS
The major auxiliaries in magnetohydrodynamic
power-generating plants are the air compressor, cryo-
genic refrigeration plants for superconducting magnets,
oxygen production plant for enrichment (if used), gas
cleanup (if not in the power cycle), seed recovery plus
all the auxiliaries for coal handling and preparation fora
coal-fired plant and a bottoming steam or gas turbine
plant (whichever is utilized), Such components as air
heaters and combustors are not included in balance-of-
plant systems as used in this paper.
Projected thermal efficiencies of first generation
313
-------
central station MHO power plants are in the 48-50 per-
cent range. C ne of the cases studied for a 1,000-MW
(net) total MHO steam power plant (ref. 2) assumes that
air is preheate1 to 2,0000 F and that a 6-Tesla supercon- .
ducting magm t is used with the generator. Fifty percent
of the output is produced by the MHO generator and 50
percent by a bottoming steam plant. The air compressor
requires 157 rvw, steam plant auxiliary power is taken as
4 percent of 1 he gross steam plant output, and inverter
losses as 3 percent of MHO generator output, for a total
of about 16 ~ercent of gross output. Power or energy
requirements of the seed recovery, superconducting
magnet cooling system or gas cleanup system are not
listed separate y. The paper also discusses the potential
improvement of efficiency if oxygen replaces air in the
combustor. The oxygen compressors alone require up to
16 percent of the total plant output. Such an oxygen
plant requires 13 kW per 1,000 seth, or 0.15 kW /Ib/hr of
oxygen produced at essentially atmospheric pressure. If
a higher pressure, say 30 atmospheres, is required, the
oxygen plant \ViII require 16 kW per 1,000 scfh or 0.2
kW/lb/hr.
The refrhleration system load during steady-state
operation of a superconducting magnet is relatively low;
the initial cooldown sets the capability requirements of
the refrigeration plant. For a 1,000-MW MHO plant, the
magnet power supply would require about 5,000 kW,
the compresso' 1,150 kW, and cold box about 50 kW-
or a total of le:;s than 1 percent of power rating.
Seed reccvery technology for MHD generators is
one auxiliary system that requires greater definition.
Coal ash and seed will be mixed. Processes that can
separate and recover the seed with the required high-
process efficiency are not yet developed. One proffered
scheme for po':assium seed recovery is shown in figure 1.
It indicates thE' possible sequence of steps, including one
in which a fornate compound is formed as an intermedi-
ate compoun j, followed by filtering, crystallizing,
evaporation, arid calcining. For a 2,000 MW (total) MHO
steam plant, t 1e equivalent of 265 MW of heat is re-
quired in the evaporator and an additional 180 MW for
the crystallizer-based on a 7,000 Btu/kWhr plant heat
rate. It is appa,'ent that this energy must be recovered if
the process is to be considered viable. The recovery of
low-temperature steam is not always practical in a power
station. Anothl:r seed recovery system for an 80Q-MW
plant is described in reference 3. The method described
is similar and 1.1e power usage appears to be of the same
order of magni':ude as required for the formate method.
Electrostatic precipitators, scrubbers, and other gas
cleanup comp01ents in the MHO plant have large power
demands. Scrut bers are discussed hereinafter.
Overall. th.~ auxiliary systems associated with mag-
netohydrodynamic plants will require further develop-
ment'along with the power cycle, not only to perfect the
technology and r~duce capital costs, but to reduce the
power required by these systems and the attendant costs
. and energy losses.
FUEL CONVERSION AS AN
AUXI LlARY SYSTEM
The viability of many advanced power systems for
power station applications is largely dependent upon
the development of suitable fuel converted from coal. In
particular, fuel cells and advanced gas turbines, among
the advanced power technologies, are linked with the
world's petroleum fuel problems. Therefore, large utility
stations of the future will be integrated frequently with
a coal conversion process. The coal conversion system
will, for the most part, be independent of the generating
plant even though it may be partially interconnected
thermodynamically with the power cycle.
For example, the molten carbonate fuel cell is con-
sidered (refs. 4,8) to be one type, which has the best
potential for application to large power plants. Molten
carbonate cells operate at 5000-7500 C temperature and
at about 10 atmospheres for long time periods-condi-
tions which may be compatible with many coal conver-
sion processes. The molten carbonate electrolyte is com-
patible with carbon dioxide, thereby easing the fuel con-
ditioning problems-which were encountered with many
types of fuel cells. Also, the electrodes for these cells do
not require the. costly catalysts that are mandatory in
low-temperature fuel cells. It is expected that it will be
possible to tailor gas quality to the fuel cell require-
ments. The fuel cell exhaust can be recycled, as in figure
2, to maintain the coal steam shift reaction, or the
exhaust can be used in a steam or gas turbine bottoming
plant, as in figure 3. Nevertheless, the need to reduce the
levels of particulates, sulfur, and NOx will still be there
but shifted to the gasification plant.
The Kopper-Totzek process (ref. 5) is an example of
an intermediate-Btu (300 Btu/lb gas) process that can be
integrated with fossil-fueled advanced power systems.
In one study of a 1,OOO-MW plant, about 17 modules
will be required (ref. 6). The plant will convert about
14,450 tons of coal per day and requires an equal
amount of steam and 11,500 tons of oxygen per day.
Such an oxygen plant requires about 190,000 kW or 13
kW per ton of coal per day. The steam requirements of
1.2 X 106 Ib/hr if of central power plant size. Scrubbing
for desulfurization and particulate removal, sulfur
recovery, ash handling, and other conditions may be
necessary to obtain a suitably "clean" fuel. Use of air
314
-------
. .
"FORMATE" PROCESS
RESIDUE FROM STEAM GENERATOR ANO L TAH .- K2S04 + Ca(OH)2 + 2CO
RESIOUE FROM ELECTROSTATIC PRECIPITATOR -- 2KHC02 + Ca S04
ST EAM 2KHC02 + °2 - K2C03 + C02 + H20
H20 MIXING
TANK
H20 + C02 H20 Na2C03 + H20
2120f K2C03 !
Na2C03'
CALCINATOR MIXER FRACTIONAL K2COJ
to) ASH 'fILTER CRYSTALLIZER
-. (CRYSTAL LINE)
U1 f
°2
KOH CLARIFIER Fe(OH)2
Cr(OH)2
Mg(OH)2
CO
KHC02
K2S04 REACTOR NaHCD2 KHCD2
K2C03 MIXER (3920F) C!!S04 NaHC02
Na2S04 fiLTER CRYSTALLIZER
Na2C03 ~~ ~
Ca(QH)2 STEAM CaS04 H20
Figure 1. Process flow chart for seed conversion system.
-------
-
RBON -
.. - -
- .. -
M jTEAM
- -
-
REFORMING SULFUR SHIFT
STEP REMOVAL CONVERTER
'ACK FUEL
- -
13 -- CELL -
COAL JR
HVDROCA
AIR
STFA
51
CC
H2.
TRACE~; H2J CO
Figure 2. Fuel cell with gasification and shift conversion.
instead of oxyglm, which is considered feasible, probably
will not reduce the energy requirements.
A second study (refs. 7,8) of a 1,00D-MW (nominal)
integrated Lurgi low-Btu gasification and combined gas-
steam turbine gonerating plant estimates that the fuel
conversion plant will require 83 MW. On the basis of
6,670 tons of I:oal per day, the auxiliary requirements
are 12.4 kW per ton of coal, which is close to the
Kopper-Totzek plant requirements. The integrated
combined plant thermal efficiency is expected to drop
to 35 percent based on the electrical plant efficiency of
45 percent.
A 10,000 tons/day coal liquefaction plant study
(ref. 9) estimaHs that the electric power plant require-
ments will be ~bout 6 kW per ton of coal per day and
that an additional 164.5 X 106 Btu/hr of energy in the
form of fuel w-ll be burned in the process. A liquefac-
tion plant will not necessarily be integrated with the
electric'generating plant. .
The coal conversion plants will undoubtedly result
in a large increase in auxiliary power and, therefore,
adversely affect energy conservation. From the environ-
mental conservation viewpoint, many of the polluting
sources of coal-fired power plants, such as sulfur, NOx'
hydrocarbons, and particulates, will be transferred to the
coal conversion plant. The goals of less dependence on
oil for power generation and better control of environ-
mental effects may often be in conflict with the goals of
energy conservation. The energy losses incurred by con-
verting of one fuel into a more environmentally accept-
able one is one price paid for energy independence.
SCRUBBERS AND OTHER
CLEANUP DEVICES
Advanced power systems such as MHD, coal conver-
sion. plants integrated with fuel cells, gas turbines, and
steam turbines will require scrubbers, desulfurization
316
-------
COAL
(,,)
...
"
GASIFIER
SCRUBBER
MOLTEN CARBONATE
FUEL CELL
FUEL
GAS
(H2'CO)
(N2,C02)
FUEL
SIDE-
ANODE
AIR
SIDE-
CATHODE
RECYCLE
GAS
Figure 3. Molten carbonate fuel cell system with gasification
and bottoming plants.
BOILER
WATER
VENT
HEAT
EXCH.
4
AIR
-------
processes, or ether particulate removal processes. One
reason why scrl bbers have been particularly vexatious to
a few large power utilities is that scrubbing systems
often increase in-plant auxiliary power requirements by
30 to 50 percent. Reports of scrubber performance (refs.
10,11,12) indic ate that the scrubbing systems can con-
sume up to 6 percent of the power produced by a unit.
As shown in table 1, electrostatic precipitators are
among the usen of large amounts of power, but require
considerably le~s than is consumed by scrubbers.
SOLAR AND FUSION
The principal auxiliary power demand of solar-
generating plan':s is for pumping of the circulating fluid.
Pumping power requirements for a 200-MW plant using
concentrating p~rabolic trough collectors integrated with
a steam plant have been calculated to be about 0.4 per-
cent of the p(iWer generated in the integrated steam
plant (ref. 131. The main heat loss through the walls of
the circulating fluid system is given as 4.6 percent of
kilowatts available from the collectors.
Fusion real:tors have yet to demonstrate all condi-
tions for self-su$taining operation with a net payoff.
Many types of ,Jower-generating systems have been sug-
gested as bottoming plants. The parasitic and auxiliary
power requirements of fusion plants are expected to
result in a negative' power output for many years to
come. The firs', generation Princeton reference design
(ref. 14) for a 2,OOO-MW Tokamak fusion power plant
projects a power output of 2,405 MW(e) gross, with 15.6 .
percent (or 371> MW) required for station service and
2,030 MW net-.~nd an overall plant thermal efficiency
of 38 percent.
SUMMARY
The paper shows, by example, that in many ad-
vanced power s1lstems proposed for utility power appli.
cations, the balance-of-plant systems will require up-
wards 'of 10 pen:ent of the power generated. I n addition,
substantial amounts of energy in the form of steam,
additional fuel, or other sources will be required. These
onsite requiremonts are partly the result of the effort to
safeguard the ervironment, partly the result of the need
to move away from the use of petroleum fuels, and
partly due to the lower level of concern usually given to .
auxiliary system; than to main power cycle systems.
Since man\, of the auxiliary systems require no
fundamentally new technology, they are taken for grant-
ed and not given enough consideration in advanced
power conversicn system work. However, as they will
consume a considerable amount of the energy produced,
more attention should be focused on them. In some
cases, improvements in the auxiliary technologies will
result in as much benefit as may be gained in improved
efficiency of the main cycle of the advanced power
plant.
There appears to be a real possibility for a satisfac-
tory payoff to justify increased study and effort to
develop the balance-of.plant auxiliary systems so that
when the new advanced systems arrive, all the auxiliaries
will match them' in maturity and in the maximum pos-
si~le e~ergy and environment conservation. The goals of
energy and environmental conservation will be achieved
with these advanced systems only if sufficiently high
plant efficiencies are attained so that they will more
than just compensate for the increased auxiliary and
parasitic losses.
REFERENCES
1. "Selection of Plant Auxiliary Power Voltage Lev-
els," Electric Light and Power, August 1972.
2. F. A. Hals and W. D. Jackson, "Systems Analysis of
Central Station MHD Power Plants," Fifth .Interna.
tional Conference on Magnetohydrodynamic Elec-
tric Power Generation, 1977.
3. D. Bienstock et aI., "Magnetohydrodynamics-Low
Air Pollution Power Generation," ASME Paper
73-WA/Ener-3, Winter Annual Meeting, November
11-15,1973. .
4. S. Baron, "Fuel Cell Power Generation and Gasifica-
tion," American Public Power Association Annual
Conference, San Francisco, California, June 26-28,
1972.
5. J. F. Farnsworth, H. F. Leonard, and D. M. Mitsak,
The Production of Gas from Coal Through a Com-
mercially Proven. Process, Koppers Company,
August 1973.
6. National Academy of Engineering, Evaluation of
Coal.Gasification Technology, Part 1/: Low-and
Intermediate.Btu Fuel Gas, prepared for Depart-
ment of Interior, March 1974.
7. F. L. Robson et aI., Technological and Economic
Feasibility of Advanced Power Cycles and Methods
of Producing Non-polluting Fuels for Utility Power
Systems, United Aircraft Research Laboratories,
prepared for National Air Pollution Control Admin-
istration, U.S. Department of Health, Education and
Welfare, December 1970.
8. H. C. Hottel and J. 8. Howard, New Energy Tech-
nology-Some Facts and Assessments, The MIT
Press, March 1973.
9. Ralph M. Parsons Company Demonstration Plant
Clean Boiler Fuels From Coal-Preliminary Design/
318
-------
Capital Cost Estimate, Volume " prepared for
Department of Interior Office of Coal Research,
1973.
10. E. A. Sondreal and P. H. Tufte, "Scrubber Develop-
ments in the West," 1975 Lignite Symposium,
Grand Forks, N.D., May 14.15, 1975.
11. "S02 Removal From Stack Gases," Power, Special
Report, September 1974.
12. Pollution Equipment News, June 1975.
13. University of Minnesota and Honeywell, Research
Applied to Solar-Thermal Power Systems, Report
No.4, prepared for the National Science Founda-
tion, July 31, 1974.
14. R. G. Mills, ed., "A Fusion Power Plant," Princeton
Plasma Physics Laboratory Report MATT -1050,
1974.
319
-------
320
-------
5 November 1975
Session V:
RESIDENTIAL AND COMMERCIAL
ENERGY SYSTEMS
Robert Rosenberg, Ph.D.*
Session Chairman
.llIinois Institute of Gas Technology, liT Center, Chicago, Illinois.
321
-------
322
-------
THE POTENTIAL ROLE OF SOLAR ENERGY
IN CONSERVATION FOR BUILDING HEATING AND COOLING
Frank R. Biancardi and Maurice D. Meader*
Abstract
A number of solar collector configurations and air
conditioning subsystems are being investigated for use in
solar-powered building, heating, and cooling applica-
tions. The pertinent characteristics and operating
features of representIJtive fla t-plate and focusing-type
collectors as well as typical Rankine-cycle, absorption,
and dessicant air conditioning systems are briefly de-
scribed. Estimates of the energy savings possible for the
remainder of this century, with widespread utilization of
solar-powered systems, are also presented, based on
ERDA datIJ.
As part of the NSF/ERDA program to explore new
solar-powered systems, a Rankine-cycle turbocompressor
air conditioning and heating system has been designed,
fabricated, and tested. The design features of the system
are described and results obtained during the perfor-
mance testing of the system are presented.
The test results provide a convincing demonstration
of the feasibility of operating a Rankine cycle-turbocom-
pressor air conditioning and heating system at system
temperature levels consistent with present-day, flat-plate
solar collectors. During the testing of the demonstration
system, more than 4 tons of cooling and in excess of 60'
hours of safe, reliable operation were obtIJined. Most
significant was the operation of the system over a range
of turbine inlet temperatures from approximately
16ft F to more than 21ft F, during which the system,
performance goals were met or exceeded.
INTRODUCTION
Approximately one-quarter of the energy consumed
annually in the United States is used for building hot
water, heating, and cooling. Since systems to provide
these functions do not require high-temperature heat in-
put for successful operation, energy from the sun could
be converted into usable low-temperature energy to re-
place a significant fraction of our valuable fossil fuels
presently used for heating and cooling. ERDA believes
(table 1) that solar technology for building heating and
cooling offers the potential for supplying about 1.5 per-
cent (2.0 0) in 2000 and by 2020, as much as 10 per-
'Frank Biancardi is Chief, Advanced Energy Conversion
Systems, Energy Research, and Maurice Meader is Research
Engineer, with United Technologies Research Center, East Hart-
ford, Connecticut.
cent (15 0) of the Nation's future energy needs if costs
of collecting and utilizing solar energy can be reduced
substantially. The annual savings associated with the
conservation of fossil fuels brought about by solar
energy utilization thus could approach $35 billion dol-
lars even at today's energy prices.
Although most of the technology involved in direct
thermal applications is more advanced than for other
solar technologies, substantial research and development
is still required if the full potential for solar heating and
cooling for buildings is to be realized. A brief review of
the major subsystems required for building solar heating
and cooling applications is presented before describing
work performed at UTRC under NSF/ERDA sponsor-,
ship on a unique low-temperature, Rankine-cycle heat
pump system.
SOLAR BUI LDING SYSTEMS
A typical system for the solar heating and cooling of
buildings consists of a collector exposed to the sun's
radiation, a heat transfer fluid (liquid or air), which
carries the collected energy to the points of storage or
use, thermal storage devices, cooling or air conditionin'g
devices, and air handling systems for distributing con-
ditioned air within the building spaces, and an auxiliary
energy source (fig. 1). Provisions may be made for
domestic hot water; and appropriate pipes, ducts,
pumps, fans, heat exchangers, and controls are included
as required. However, the critical system subsystems are
the solar collector, thermal storage, and air conditioner.
Solar Collector
A solar collector consists conceptually of (1) a
radiation transmitting-shielding system designed to pass
short wavelength energy from the sun and atmosphere to
a collecting surface, while minimizing back radiation and
convection heat losses, (2) a highly absorbing surface to
accept the radiation, and (3) an insulated heat exchanger
system to transfer the energy to some working fluid for
transport to the energy storage or utilization system.
The flat-plate collector with one or more parallel glass
plates spaced approximately 1 inch apart, placed before
an insulated absorbing black surface has been widely
used. The black surface is generally a coated metal plate
which is integrated into the heat exchanger.
Variables which affect collector performance in-
clude collector tilt angle from the horizon, type and
323
-------
Table 1. Estimates of the energy and fuels to be supplied
in the United States by solar energy: direct thermal ap-
plications (in units of Q=Quads=101S Btu per year)
(ref. 1)
1985 2000 2020
Heating and cooling 0.15 2.0 15
Agricultural applications 0.03 0.6 3
Industrial applications 0.02 0.4 2
Total 0:2 j 20
Total projected U.S.
energy demand 100 150 180
number of bac!: vadiation shields, absorber surface mate~-
rial and coating, heat exchange configuration, insulation,
and seals. Unfortunately, present-day, flat-plate collec-
tors have typic~ I performance characteristics as shown in
figure 2 and thus are not capable of providing heated
fluids above 2C CO F while capturing a large fraction of
the incident solar energy. Medium concentration collec-
tors in which the incident energy is focused on collec-
tion surfaces have the capability of providing fluid tem-
peratures of 3000 to 5000 F with reasonable efficiencies.
These collectors, however, require movement of the
reflector or c )tIector surfaces but simple, reliable
methods of providing these features are being investi-
gated. Extensivl1 research and development efforts sup-
ported by NSF and E RDA are also underway on both
types of collectors to reduce costs and improve collec-
tion efficiency ("ef. 2).
-~. ------------..- ---
---------
ENERGY COllECTOR
Energy Storage
The design of the energy storage system involves a
selection of both the storage medium and system capaci-
ty. For sensible heat storage type materials, water is
attractive on an energy per unit volume basis. Also of
interest are phase change materials such as noncongruent
melting point salts which can store chemical energy
and/or operate at higher temperatures than water. Some
difficulties exist with their use in the form of supercool-
ing and segregation in the storage tank. The capacity of
the storage system is a function largely of economics.
Work to date has suggested that it should be proportion-
al to the solar collector area at a value of 10 to 15 Ib of
water (or its heat capacity equivalent for a different
media) per square foot of collector area.
To make the most effective use of both the solar
collector and the air conditioning unit, the storage sys-
'- .-- -, -_.~....-._-- ~.- -_. -. -
.- -'---'-'-----
--
AIR '0
OlSTA18U110N
OUCTING
Figure 1. Solar-powered heating/air conditioning system.
324
-------
WIND VElOC. ,y - 10 MPt-t
AMBIENT TEMPERATURE. 80F
70
80
II'
I
>-
U
Z
w
~
50
40
...
...
w
a:
o
...
u
W
...J
...J
o
U
30
:L
120 140 160 180 200 220 240
ABSORBER TEMPERATURE - F
280
300
260
Figure 2. Typical solar collector'
efficiency.
tem should be operated so that it enhances thermal
stratification. During winter operation, the coldest water
is removed from the storage tank bottom, heated in the
solar collector, and ideally returned to the tank at a level
where it can mix without turbulence with fluid of the
same temperature. Similarly, the cooling cycle will work
most efficiently with the hottest fluid available. The
piping, valving, and pumping arrangements necessary to
achieve these objectives can become complicated.
For combined heating/air conditioning operation,
there are some advantages to physically splitting the
thermal storage into two sections or separate tanks.
During the heating season, the two tanks could be oper-
ated at different temperature levels to improve collector
performance. During summer operation, one tank could
be operated hot and the other cold. Then, any excess
refrigeration capacity when the air conditioning load is
low could be used to chill water. (The use of a phase
change storage material would be particularly attractive
when operating in this mode.) The dual system would
also allow a maximum of flexibility during seasonal
changeovers. Programs are being supported by NSF/
ERDA to investigate improved energy storage systems.
Air Conditioner
The cooling process which is utilized in any thermal-
ly driven air conditioning system must be simple and
reliable and result in high system operating efficiency in
order to minimize overall systems cost and maximize the
utilization of solar energy. If these requirements are not
met, the collector, energy storage, and air conditioner
are likely to become large and excessively costly. and the
entire system will become uneconomical. Past studies
preformed at UTRC have indicated that the thermally
powered systems which utilize the endothermic process
of vaporization of a working fluid could achieve good
overall system performance levels. In addition, recent
efforts by other organizations have indicated that
adsorption air conditioning systems using various types
of dessicants could have attractive features. Other proc-
esses (all of which ,are capable of producing significant
cooling effects) such as the dissolution of solids in
liquids, the formation of azeotropic and nonazeotropic
liquid, mixtures, fusion and sublimation, expansion of
gases, and the thermoelectric and thermogalvanic effects
have 'been studied. All were rejected because of unattrac-
tive performance levels or the initial costs of the result-
ing systems were shown to be substantially higher than
thermally powered equipment already available or under
development.
Thermally powered air conditioning systems which
could utilize the vaporization of a working fluid as the
cooling process are:
1. ' Engine-driven vapor compression systems using:
Otto-cycle gas engine,
Brayton-cycle engine,
Stirling-cycle engine, and
Free-piston engine.
2. Absorption systems using:
Ammon ia-water,
Lithium bromide-water, and
Other absorbent-refrigerant combinations.
3. Double-loop systems with energy transfer via:
Ejectors,
Reciprocating expander-compressor, and
Turbomachineryexpander-compressor.
Engine-Driven Vapor Compression System
In a solar-powered engine-driven vapor compression
system, solar energy would be converted into shaft wo~k
which is then used to drive a compressor similar to tho,;e
used in electrical systems. The Otto- and Brayton-cycle,
engine-driven concepts suffer from the requirement for
relatively high maximum system temperatures for
efficient operation and would require more complex
collectors capable of providing temperatures of 1,000° F
or more. Even if the solar collectors could be provided at
reasonable cost, the engines themselves would be exces-
sively high in first costs and require burdensome mainte-
nance and overhaul procedures.
Work is presently in progress on a combination of
the Stirling heat engine and Stirling refrigeration cycle as
a means of using low-temperature heat input to provide
325
-------
air conditionin~. Tho system requires high- pressure
operation with '1elium or hydrogen as the working fluid
and extensive regenerative heat transfer. However,
small.scale cryogenic refrigerators have been built using
this system.
Engines operating on the Rankine-cycle principle,
which have received much attention in the last 10 years,
will be discussec with the double-loop systems.
Absorption Syslems
An absorption air conditioning cycle offers a simple
method of sub~tituting a heat.actuated device with few
or zero moving parts in place of a motor-driven vapor
compressor. Ahsorption air conditioning systems are
available in a w:de range of sizes for residential and com-
mercial applications and can operate with fossil-fuel or
low-temperaturE' heat inputs. As a result of many years
of operating experience, two basic absorbent-refrigerant
combinations are presently used, NH3-H20 and
LiBr-H20. These combinations have been selected based
on operating tE mperature, construction materials, and
performance cOllsiderations.
I n an ideali wd absorption system as shown in figure
3, the basic components are a generator, absorber,
evaporator, and condenser. Additional components are.
usually included to improve the thermal COP (Coeffi.
cient of PerforMance, defined as the ratio of heat
absorbed by the refrigerant in the evaporator to the heat
content of the energy required to operate the system),
and to reduce the quantity of absorbent vapor that
would be carried into the condenser with the refrigerant
vapor. These additions improve the system COP by
about 20 to 30 percent for representat.ive cycle oper-
ating conditions.
To further improve performance, still more compo.
nents and cycle modifications are introduced into the
absorption cycle. With sufficient heat transfer surface
area in the added components, a thermal COP of about
0.70 can be achieved. Although numerous absorption
systems using other absorbent-refrigerant combinations
have been suggested over the years, low predicted per-
formance levels, minimum advantage over the presently
used combinations, or the lack of detailed thermo.
dynamic data, have not encouraged further develop-
ment. Although LiBr-H20 systems can operate efficient-
ly at temperatures of 2000 F in the generator, the mate-
rials of construction must be carefully selected. Lithium
bromide-water systems have' been restricted ~y the:
operating temperatures in the absorber and condenser
and thus require water cooling to avoid crystallization of
the lithium bromide. However, modification to the
lithium bromide system is being investigated to permit
air cooling (refs. 3,4).
NHJ VAPOR
ilEA T RFJECTION
TO ATMOSPHHHi
. 0
PllMP
(aNlUA. 1 (lit
S I IH)NG
!iiOlUIION
, AbSORBER
WEA.l<..
$Ol.UTlON
o
lirA! IM"ur
a HEAT REJECTION TO
ATMOSPHERE
CONDENSER
NHJ LlQUIQ
EXP VALVE
EVAPORATOR
NHJ
V APOA
a
COOLING lOAD
Figure 3. Ammonia-water absorption system
schematic.
326
-------
Double-Loop Systems
An important class of thermally powered air condi-
tioning systems is based on the use of the so-called
double-loop cycles. Althou~h these cycles may differ in
their details, they all utilize a self-contained Rankine-
cycle power-conversion loop to provide the energy for a .
conventional vapor refrigeration loop. The energy from
the power cycle can be transmitted to the refrigeration
cycle by various means. One method is to use an ejector
system in which the energy of a high-pressure, high-
temperature fluid stream is transferred to the refrigerant
vapor through a momentum transfer process. Another
method would use a vapor expansion engine to drive a
reciprocating compressor. Another method is to use a
vapor-expansion turbine which drives a centrifugal com-
pressor in the refrigeration loop.
All of these approaches have been pursued at one
time or another with varying success. During the past
year and a half, UTRC has been under contract to
NSF/ERDA to demonstrate the operation of a Rankine-
cycle heat pump system using the turbocompressor
approach at system temperature levels consistent with
present-day flat-plate collectors.
TURBOCOMPRESSOR SYSTEM DESCRIPTION
The turbocompressor Rankine-cycle system concept
for solar-powered heating and cooling is shown schemati-
cally in figure 4. It utilizes a single working fluid and
common condenser for both the power and cooling
loops. Hot fluid from the solar collector/storage system
is used to vaporize a working fluid in the vapor generator
of the power loop. The vapor is expanded through a
._-~ ---.-.--
.Hot fluid
from
collector /
storage
system
Vapor
generator
Pump
turbine which produces mechanical power to direct-drive
a compressor. The compressor pressurizes working fluid
vapor leaving the evaporator to the level necessary for
heat rejection to the atmosphere ,in the condenser. The
exhaust of the turbine and compressor is mixed and
enters the common condenser which may be air. or
water-cooled.
A receiver is used at the exit of the condenser to
permit smooth distribution of the working fluid to each
loop. An expansion valve is used prior to the evaporator
to lower the cooling fluid pressure and temperature to
the desired level. A pump which may be driven off the
same turbocompressor shaft circulates the working fluid
in the power-conversion loop. A common nontoxic, in-
expensive refrigerant such as R 11 is preferable as the
system working fluid. When heating is desired, the turbo-
compressor can be bypassed and the evaporator coil used
as the heating element or the system can be operated in
a heat pump mode. Any deficiencies in solar-generated
heat would be made up by a supplemental burner.
. The turbocompressor system has many advantages.
For example, since the same working fluid is used in
both the cooling and power loop, a common condenser
can be utilized. Furthermore, elaborate seals are not re-
quired to separate the working fluid from each loop
when mixing in the turbocompressor. Lubrication can be
provided by the working fluid itself and thus the turbo-
compresor unit can be hermetically sealed. Single-stage
compressor and turbine designs based on technology
already demonstrated can result in a simple, yet efficient
unit. Excess shaft power can also be generated, if de.
sired, for auxiliary purposes and additional heat ex-
changers, Le., a regenerator or precooler can be used to
-----.-
-_.~ ... -- - _.-
Turbo-
compressor
Evaporator
Cooling
Condenser
Expansion
valve
Reservoir
Figure 4. Turbocompressor system.
327
-------
increase syste'Ti performance, if cost efficient. These
factors lead t) high overall system efficiency even at
relatively low ,)perating collector and temperature levels.
DErJlONSTRATION TEST SYSTEM
The technical approach used for the feasibility
demonstration of the Rankine-cycle turbocompressor
concept is bast!d upon the adaption of existing key com-
ponents while providing the capability of the system to
operate safely and efficiently at temperatures and condi-
tions consistent with present-day flat-plate solar collec-
tors. .
The demonstration turbocompressor system has
been assembled with two separate loops, one for power
generation and the other for cooling (fig. 5). Although
the same working fluid is used in both loops, separate
condensers are used. This configuration permits more
convenient OpE ration and measurement of temperature
and pressure conditions at the exhaust of the turbine
and compreSSCf elements, and therefore, provides a
direct means fc r correlating calculated and experimental
component performance. The two condensers are
normally reguilited to operate at the same temperature
-- ---
to simulate single condenser operation, although offset
condenser operations are possible for specific tests.
Heat transfer at all of the system interfaces is via
cold or hol wulor Houn:ItH. ThiH purmil!1 dOH'" IIIIIIP'"
ature control than would be feasible with air coolin[J.
Heat input from a solar cpllector is simulated by a steam
supply system which provides hot water to the power
loop vapor generator over a range of inlet temperatures
between 1700 and 2400 F. The condensers are capable
of operating over a temperature range from about 900 to
1300 F, thus simulating both water- and air-cooled
operations.
The turbocompressor unit consists of a two-stage
compressor mounted on one end of a shaft with a radial
inflow turbine at the other end. The compressor selected
for the turbocompressor demonstration unit is capable
of producing in excess of 8 tons of cooling with R 114 as
the working fluid and was originally designed for use in
an aircraft environmental control system. The rotating
elements are less than 4.0 in. in diameter and operate at
speeds of 20,000 to 32,000 rpm.
The shaft is supported by precision ball bearings
that can accept both thrust and radial loads. Cooling and
lubrication is provided by liquid refrigerant flow to the
BASELINE DESIGN CONDITIONS
FLUID-R 11
COOLING =3.7 TONS
COP=0.36
T= 200°F
COLLECTOR STORAGE
SIMULATOR
COMPRESSOR
T = 450F
N =28.560
RPM
T=1590F
H20
(COOLING LOADI
EVAPORATOR
FREON
PUMP
POWER lOOP
H20
T = 1 05 of
T=1050F
COOLING LOOP
Figure 5. Demonstration system flow chart.
328
-------
bearing cartridge. An expansion valve controls superheat
in the bearing exit coolant stream to about 70° F, thus
assuring a relatil/ely dry condition and minimum bearing
losses.
The other major components in the system are heat
exchangers. Two of these are evaporators and two are
condensers. All are commercial units of the shell and
tube type. The evaporators have water on the shell side
which is baffled to improve the film coefficient.
The two condensers use water on the four-pass tube
side. Both units are oversized for this application. As a
result, it is possible to use the condenser as receivers
since covering the bottom tubes with liquid refrigerant
does not prevent the unit from operating as required.
The demonstration system is shown in figure 6 as it
appears in the UTRC laboratory. The overall dimensions
of the system are approximately 8 ft long by 4 ft wide
and 6 ft high. Although a more compact arrangement
was possible, ample room was permitted for instrumen-
tation and any desired modifications.
TEST RESULTS
Over 60 performance test runs were obtained with
the turbocompressor demonstration system and more
than 60 hours of operation have been recorded. System
operation has been achieved at turbine inlet temper-
atures ranging from approximately 160° F to above
215° F, condenser temperatures from 85° to 125° F,
and evaporator temperatures from 30° to 70° F. About
half of the tests have been at conditions close to the
design point conditions of turbine inlet, condenser and
evaporator temperatures of 200° F, 105° F, and 45° F
respectively. The complete test results are summarized
and presented in the reference 5 report.
The analyses indicate very close correlation between
the test data and analytical predictions of system per-
formance at condenser temperatures of 100° to 105° F.
At other condenser temperatures, wider discrepancies
between the test data and analytical predictions are
noted with the test data generally higher than the analy-
tical predictions.
- ..-,-
Figure 6. Demonstration turbocompressor system.
329
-------
Specifc comparisons of the demonstration system
test data a ld predicted values of performance and oper-
ating conditions have been made at selected test runs
near the d!sign point conditions. The cooling produced
was generally somewhat higher than predicted (i.e., 4.05
10ns verSU5 3.7 tons in one easel. but the measured COP
was slightl,! lower than predicted. Some of the reasons
for the differences are described in the following para-
graphs. AI':hough the turbine inlet temperature during
the test run was higher than the design baseline condi-
tions of 2COo F. the higher temperature is in the form of
Rll superheat which does not significantly raise the
COP of the turbocompressor system.
While turbomachinery efficiencies are as predicted
or higher, !ystem Coefficient of Performance with these
conditions falls below that predicted mainly because the
evaporator is operating at substantially less than 45° F
and the flow to the compressor has an excessive amount
of superheat. At the lower pressure associated with the
reduced evaporator temperature, more compressor work
is required to pump the gases to the condenser pressure
and hence. ess flow can be circulated in the cooling loop
with the av,lilable turbine power.
Other test data obtained at approximately the
design point conditions also compare well with the
predicted s'lstem performance. For example, data taken
at a turbinE' inlet temperature of 2000 F produced 4.33
tons of cooling at an evaporator temperature of 35° F
with a COI' of 0.34. The test data also illustrate the
. sensitivity )f compressor performance to the desired
compressor head conditions. The highest performance
. levels are achieved with temperature differences between
the conderser and evaporator approaching 600 F or
more. principally because the compressor was originally
designed fol' relatively high-pressure ratio operation.
Althou!lh much of the demonstration system testing
was conducted at the nominal design conditions. opera-
tion at turbine inlet temperatures other than 200° F
were obtaiced with condenser temperatures of approxi-
mately 105° F. A comparison of the Coefficient of
Performanc!! and Cooling Capacity data obtained from
the demons':ration system tests with that predicted from
analyses is ;hown in figure 7 as a function of turbine
inlet temper ature. The predicted performance character-
istics are shl)wn in figure 7 by the solid line over a range
of turbine ;nlet temperatures for a condenser temper-
ature of 1050 F and a compressor inlet temperature of
50° F. Since the compressor inlet temperature includes
5° F of sup~rheat, the evaporator act'ually operates at a
temperature of 450 F. The experimental data are indi-
. cated in figure 7 by the open circles and have been ob-
tained over OJ range of condenser and evaporator tem-
peratures close to those used for the calculated perform-
- CAt.~LATlD"III'OAIIAM::I'OR 'OIl ~DI".R.IOP eVAl'OR"'T~
o ""RIM1NTAL DATA ATCQNOITIQNIAJIPM)XtlU,TlL,V THOll! 0' CALCULATIONS
0.7
iG.6
iu
~ 0.'
..
o
~ 0.3
~
~ 0.2
~ 0.'
tP~
'11) 200 210 "0 230 240
TURBINe INLET TEMPERATURE .. F
7.0
g 80
~ 1,0
~
O. '.0
..
~ 3.0
~
g2.0
'110 200 2.0 220 . 230 2'0
TURBINE INLET TEMPERATURE - F
Figure 7. Demonstration system
performance characteristics.
ance curves. The data shown indicate that the cooling
produced ranged from 2.05 tons to 4.33 tons with a
measured COP ranging from 0.20 to 0.34. Some differ-
ences between test and predicted performance are noted
at turbine inlet temperatures of 2000 F and above. A
portion of this difference can be attributed to the fact
that during the tests the system operating conditions did
not duplicate all the conditions used to analytically pre-
dict performance.
The results obtained when operating the demonstra-
tion system at condenser temperatures slightly lower
than the baseline condition of 1050 F are encouraging.
Data were obtained for turbine inlet temperatures from
190° to 1680 F. although evaporator temperatures were
generally higher than desired. During some of these tests.
a COP of 0.45 was obtained with a turbine inlet tem-
perature of 1800 F. The flow utilized for cooling the
turbocompressor bearings was reduced during some
selected runs and is believed responsible for the relative-
ly good performance.
Test results were al50 obtained when operating the
demonstration system at condenser temperatures o~
approximately 90° F. While only a limited number 0-,
330
-------
test runs were completed at this lower condenser tem.
perature, experimental performance met or exceeded
performance predictions and a COP approaching 0.40
was obtained. Substantial cooling was produced with
anticipated levels of Coefficient of Performance at tur-
bine inlet temperatures as low as 163.5° F together with
evaporator temperatures of 45° F or below.
A review of all the test data over the range of tur-
bine inlet and condenser temperature conditions investi-
gated indicate that only approximately a 70° to 80° F
temperature difference between the vapor generator and
condenser is required for the turbocompressor system to
operate. With selected improvements to the system even
lower temperature differences should be feasible for effi-
cient operation.
PERFORMANCE POTENTIAL
Although the turbocompressor demonstration
system has exhibited impressive performance, especially
at turbine inlet temperatures between 180° and 210° F,
substantial improvement is possible. One means of im-
proving performance is by incorporating specifically
designed and matched turbomachinery elements rather
than using modifications of existing equipment. While
the present turbomachinery has generally met or exceed-
ed the performance levels predicted analytically, much
higher performance levels are achievable, especially for
industrial turbocompressor systems designed to provide
from 25 to hundreds of tons of cooling and using larger
Welded housing,
Shrouded
compressor
wheel
Bearing
lubricant
Inlet
Pipe diffuser housing
compressor and turbine wheels than those used in the
present program. Even systems designed for the residen-
tial and small commercial markets; i.e., 3 to 20 tons,
can utilize turbomachinery which have component
efficiencies higher than those achieved in the present
system. An artist's concept of such an advanced turbo.
compressor is shown in figure. 8. Such a unit, approxi-
mately 6 in. in diameter and less than 9 in. long would
be suitable for use in a 3. to 5-ton domestic heat pump
system. With such machinery, the system COP can be
increased by 35 percent above the present levels and
system component sizes (including the solar collector
and storage systems) reduced accordingly. Turbocom-
pressor systems sized for the industrial market would be
more efficient because of the larger physical dimensions
of the components thus reducing some of the losses
inherent in small components. These larger components
should be capable of achieving efficiencies approaching
85 percent or more. As a result, the turbocompressor
product efficiency levels (i.e., compressor X turbine effi-
ciency) achieved with the present system (50 percent)
can be increased with further development to 64 percent
for residential systems and to 72 percent for industrial
systems. Coefficient of Performance levels of the turbo-
compressor system would correspondingly be increased
to levels approach ing 1.0 even when operating at 200° F
turbine inlet temperatures. The performance levels
achievable with the turbocompressor system as a func-
tion of turbomachinery product efficiency and turbine
inlet temperature are shown in figure 9.
Turbine wheel
Turbine nozzles
Turbine
inlet
Figure 8. Turbocompressor 'unit design.
331
-------
\ A"t)Q :;£fIIIfD.I't'(\A' \"', St,#{Q..U"
:QIIro;Of:twSl- + to ,
.'~""O'I"'~~ - ~ . - ~ . ~."IE"'1'
..
2.0
w
U
Z
..
~ 1.5
~
a:
. ~
...
o
t-
Z
w
U
~ 1.0
8
PRESENT Fl U PLAn
COllECT Of. nANGI
MeDIUM CONCIENTAATION
COlUCTOA RAHQI.
o
.110
200 260 300
TURBINE INLET TEMPERATURE - F
JIiO
Figure 9. Turbocompressor system
perfo"mance potential fluid R 11.
The result: shown in figure 9 for a condenser
temperature of !}Oo F also indicate the substantial poten-
tial for higher system performance when operating at
turbine inlet teMperatures above 2000 F. For example, a
COP greater than 1.1 can be achieved at a turbine inlet
temperature of :240° F with a turbocompressor product
efficiency of 64 percent. Thus in the future, marked
improvements i.) turbocompressor system performance
could be achievfid with the use of medium-concentrating
collectors and sr:ecifically developed turbomachinery.
CONCLUSIONS
1.
A turbocorr,pressor Rankine-cycle system has been
designed, built, and operated efficiently and reliably
at peak :iystem temperatures consistent with
present-day flat-plate solar collectors.
2. Substantial improvements in tUrbocol11pft!SSOI sys,
tem performance and therefore size and cost are
feasible with added component and overall system
development. The full-scale implementation of solar
heating and cooling with a turbocompressor system
could be brought to commercial development with-
in 3 to 5 years.
3. With continued intensive development, building
heating and cooling systems, using solar energy,
could make a contribution to significant energy con-
servation in the United States.
ACKNOWLEDGMENT
This paper is based largely on work performed
under Contract NSF-C903 for the National Science
Foundation and the Energy Research and Development
Administration. The suggestions and assistance provided
by Dr. John W. Leech, Program Manager at ERDA are
gratefully acknowledged. The authors also are particular-
ly indebted to the technical assistance provided by
Messrs. H. W. Simpson and J. L Warner of the Hamilton
Standard Division of United Technologies, and E. .R.
Schulman, W. Blecher, N. C. Rice. and W. G. Burwell of
UTRC during the program activities.
REFERENCES
1. Definition Report. National Solar Energy Research,
Development and Demonstration Program, ERDA-
49, June 1975.
2. Proceedings of the Workshop on Solar Collectors for
Heating and Cooling of Buildings, NSF-RA-N-
75-019, New York City, November 21-23,1974.
3. Solar Cooling for Buildings, workshop proceedings,
NSF-RA-N-74-063, Los Angeles, California, spon-
sored by NSF, February 6-8,1974.
4. Tenth Intersociety Energy Conversion Engineering
Conference, University of Delaware, Newark, Dela-
ware, August 18-22, 1975.
5. "Feasibility Demonstration of Solar-Powered Tur-
bocompressor Air Conditioning and Heating Sys-
tem," ERDA Report COO/SH-C903/75/1. UTRC
Report R75-951923-4, October 1975.
332
-------
EVALUATION OF ENGINE-DRIVEN HEAT
PUMP SYSTEMS OF SMALL CAPACITI ES*
Jaroslav Wurm and Gopal P. K. Panikkert
Abstrac t
One generally recognized and immediately available
method for conserving the energy used for space condi-
tioning is the heat pump. Studies have shown that
among heat pumps, those with certain component char-
acteristics could greatly increase the energy savings
usually resulting with heat pumps. One such preferred
system recognizes the special operating requirements of
a small air~to-air heat pump by utilizing an onsite gas-
operatedprimemov~asthedri~~
This paper discusses the results of performance
evaluations of various systems of this type and analyzes
the consequences of operating residential Rankine and
Brayton heat pumps driven by Otto, Stirling, Brayton,
and Rankine power subsystems. Since some of the can-
didates have already reached the level of hardware test-
ing, their characteristics are applied to machines with
two types of capacity modulation where the onsite
recovery of the engine's rejected heat is used in the heat-
ing mode.
The results show that when compared with an elec-
trically driven, residential heat pump applied to a typical
house heat load, the air-to-air heat pump driven by an
onsite gas-operated prime mover exhibits:
1. Superior performance (in terms of seasonal
COP),
2. A possibility of meeting comfort conditions
over a wider range of ambient temperatures, especially
for the heating mode.
INTRODUCTION
For several years the I nstitute of Gas Technology
(IGT) has been studying various concepts of residential
air-to-air heat pumps to assess their energy-saving poten-
tial and their practicability. Our approach has been quite
broad. The scope of the studies even included cycles,
components, and systems that had never before been
considered for conventionalt heat pump service.
*This work is part of a research project sponsored by the
American Gas Association. The authors wish to express their
gratitude for being allowed to present the material to the
Symposium.
tJaroslav Wurm is Supervisor, Space Conditioning Research.
and Gopal P. K. Panikker is Associate Mechanical Engineer at the
Institute of Gas Technology, Chicago. Illinois.
tsy "conventional," we mean a heat pump supplying heat
or cold to condition the living areas. .
Methodically, we have progressed from studying ideal-
ized systems (ref. 1) to evaluating practical systems at
design (ref. 2) and off-design operating conditions. Some
of our objectives called for identifying potential problem
areas and design and operating improvements for both
the existing and new components and systems. Finally,
we identified concepts with sufficient promise to war-
rant further development. This paper discusses the part
of our work specifically related to achieving these objec-
tives.
BACKGROUND
The new generation of conventional, residential,
air-to-air heat pumps is becoming an important factor in
our energy conservation efforts. The reliability of com-
mercially available units has improved significantly, and
the manufacturers have launched programs for dealers to
improve installation and servicing practices. The per-
formance and quality of heat pumps has been improving
based on continuing component and system design
research that had been initiated on a larger scale in the
1950's.
However, even the very best conventional, residen-
tial heat pump has its preferential range of operating
conditions, and so realizing the full energy-saving poten-
tial thus depends on the designer's ability to widen the
range of conditions over which a heat pump could be
advantageously used and/or on defining the application
limits. Generally, these limits (for conventional residen-
tial Rankine heat pumps) are inherently related to the
direct proportionality between the ambient (source)
temperature and the heat pump's capacity. In the hfat-
ing mode, this actually translates into two problems:
The heating capacity decreases as the temperature of the
ambient air decreases, and the heat delivered is at.low-!r
and lower temperatures. Less efficient supplemental
heating is commonly used to correct these inadequacies.
In the case where this is done by resistance heating. the
seasonal effectiveness of the system is penalized to such
an extent that in northern climates, the benefits of the
heat-pump effect vanish, and the seasonal effectiveness
approaches that of a conventiona I fu rnace (ref. 3).
The fact that, for marketing reasons, the present
residential heat pumps are being designed as both heaters
and air conditioners,g with the design point within the
!I For example, in the United States only recently have
"heating only" heat pumps been introduced on the market.
333
-------
cooling-mode range of conditions, certainly does not
moderate the detrimental effect of variable operating
conditions on the year-round effectiveness. Yet, system
designs that can achieve the maximum of the heat-
pump's energ',-saving potential have been conceived.
Lately, some t.ave reached the market place; others are
being evaluatej or have reached an experimental stage
(ref. 4).
The objective of the studies at IGT was to assess the
feasibility of CI special group of residential comfort sys-
tems: engine-driven heat pumps, a group that has high
potential for improved energy savings. We believe that
for certain residential heating and cooling applications,
engine-driven, air-to-air heat pump,s will provide total
year-round en ergy savings higher than that attainable
with conventional heat pumps. This results from:
1. The on-site recovery and utilization of the
engine's rejected heat, and
2. Operating the heat pump and its components at
conditions where high efficiencies are achievable.
The validity and feasibility of these two factors have
also been discus5ed in detail in reference 2. The systems
and. components selected for our investigati~n, the
method of eVi/luating the heat-pump design-point and
off-design performance, and the results of the evaluation
are presented below.
SELECTI::>N OF SYSTEMS, COMPONENTS,
JIND OPERATING MODES
An evaluation of a heat-pump system on a seasonal
and year-round basis requires consideration of many
individual step!. For every system, the sequence of these
steps (including an internal looping and iterative ap-
proach not indi :ated here) includes:
1. Selecting and sizing components,
2. Genemting and/or evaluating component char-
acteristics,
3. Comp,ment matching and generating heat-
. pump system characteristics,
4. Matching the system and load characteristics
(heat pumphow;e),
5. Selecti 19 the operating modes,
6. Applying the heat pumphouse system in selec-
ted climates by Jsing statistical weather data.
To better understand the evaluating procedures,
some of these ir.1portant steps are described below.
System Selectio.J
For the dl!sign-point and off-design performance
evaluation, we have selected systems on the basis of
evaluations of :dealized heat-pump candidates, the
suitability of thll components, and the availability of real
component characteristics.
Analytically, we have discarded the Stirling refrig-
erator as not showing potential for conventional heat-
pump applications (ref. 5). It shows superior perform-
ance only for greater differences between the sink and
source temperatures than that encountered in conven-
tional heat-pump service. In spite of its relatively poor
cooling performance, we have assumed that the Brayton
refrigerator is a reasonably good candidate because of its
stable and relatively high heating performance. The
Rankine refrigeration subsystem, and specifically its
mechanical configuration,* is the only one for which the
applicability has been verified, and systems using it have
been marketed. Thus, two candidates remained for con-
sideration: the Brayton and the Rankine refrigeration
subsystems.
Suitable prime-mover subsystems for driving the
refrigerators were selected purely on the basis of avai-
lability of desired component characteristics.t Although
diesel engines were not represented, the applied drivers
still include all of the important groups of prime-mover
subsystems. One hundred percent efficient, constant and
variable-speed electric motors were used, primarily for
reference cases.
For evaluation purposes, we have formally "design-
ed" several systems using su itable combinations of
prime-mover and refrigerator subsystems. The potential
and evaluated systems are depicted in table 1.
Component Selection
All of the components used in our evaluation were
represented either by measured or generated and experi-
mentally verified performance characteristics. None of
them was, however, designed for heat-pump service; we
assumed that the required design changes would be pos-
sible for the preferred configurations. The components
used and their characterization are as follows:
1. An automotive air-conditioning compressor-
open, single cyclinder, 98.3 cm3 (6 in.3) displacement,
500 to 4,000 rpm, driven by:
a. Hypothetical engine or motor with constant,
source energy efficiency of 30 percent,
b. Otto, 4-stroke, 1-cylinder, air-cooled, sta-
tionary engine with an output of 4 to 14 hp and 1,800
to 3,600 rpm,
c. Stirling, double-acting, 4-cylinder engine
with swashplate, 16 cm3 (1 in.3) displacement of
helium, pressure and rpm modulated,
d. Stirling, double-acting, 4-cylinder engine in
V-arrangement with pressurized crankcase.
* Absorption systems were stud ied separately (ref. 6L
t At the time of the investigation, not all of the necessary
characteristics of components desirable for evaluation were avail-
able to us.
334
-------
to)
to)
U1
CYCLE
~
UJ
~
CJ)
>-
CJ)
CD
=>
CJ)
C)
z
ti
a::
UJ
C)
0:
IL.
UJ
a::
I&J
~N
:.::-
Zl
eta::
a::
Component matrix of engine-driven, residential heat pumps
indicating potential and evaluated systems
PRIME MOVER SUBSYSTEM
Table 1.
HYPOTHETICAL
ENGINE
OR MOTOR
BRAYTON
Vacuum,
Regenerative
OTTO
DIESEL
MACHINERY
CONSTANT VARIABLE
SPEED SPEED
PISTON
ROTARY PISTON
TURBO
I DESIGN NO. I NO.2
P~TON :: -;'ai:")\\l')~'0t~
NO. I
NO.2
ROTARY
TURBO
zii ~A~
g I
~ ei'
:fi a a:: TURBO
~1f
f~.f~:~:~!!J.Ji;-r
D NOT EVALUATED
r::~;::.i DESIGN-POINT EVALUATION
~ OFF-DESIGN EVAWATION
* NO CHARACTERISTICS AVAILABLE
STIRLING
Helium
RANKINE
Steam
SINGLE FREE ROTAR-d-J
ACTING DOUBLE ACTING PISTON* PISTON ~Y"'"' TURBO
NO.1
NO.2
g\!!./x>/
:.:.~ ~ ;.?:)t:~:::=
.%mi\\)/!
.;.:.:.\::::::.:.:.:.\:::
8760I0OI5
-------
2. An automotive air-conditioning compressor
redesigned for a refrigeration application, open, single-
cylinder, 139 em3 (S.5 in.3) displacement, 500 to 2,000
rpm, driven by:
a. C tto, 4-stroke, 1-cyl inder, air-cooled, sta-
tionary engine with an output of 1.3 to 10 hp and 1,SOO
to 3,600 rpm,
b. Rankine semiuniflow, 4-cylinder, 262 cm3
(16 in.3) displacement, steam expander with an output
of 1 to 10.7 hp and 1,100 to 2,SOO rpm, 90 percent
efficient regenerative boiler [generating 1.72 kPa (250
psia) and 3250 C (615° F) maximum steam conditions]
with atmospheric or vacuum condensation.
3. Turbccompressor (Rankine, R-12) of an experi-
mental design, driven by an experimental, vacuum,
regenerative ga) turbine.
4. Brayt')f) experimental refrigerator consisting of
a radial turbocompressor (slightly backward vanes) with
a slightly curved diffuser, a nominal design specific speed
of 25,000 rpm, pressure ratio 1.9, and a radial inflow
expander with a nominal pressure ratio of 1.73. The
refrigerator was driven by an experimental gas turbine
consisting of a radial turbocompressor (straight), curved
diffuser, a rad al inflow expander, and an atmospheric
combustor with recuperator.
Operating Mod,~s
. One of the advantages of an engine-driven heat
pump is the flexibility in designing the controls that
establish the relationship between the instantaneous
capacity of thE machine and the heating/cooling load on
the house.. W~ call this relationship the "operating
. moda." This H!xibility gives the designer the possibility
of optimizing "the energy consumption of a heat pump
by modulating its capacity relative to the load. The
detailed explanation as to why the capacity modulation
could be beneficial to the heat-pump performance can
be found in re','erence 2. However, at this point we will
reemphasize two important aspects of capacity
modulaton:
1. Several ways of achieving desirable changes-in
the capacity oj a heat pump's refrigeration subsystem-
types of modlJlation-exist. For systems with piston
compressors, 1 hese include speed variation, suction
throttling, clea:'ance volume control, lifting of the suc-
tion valve, and discharge bypass. Some of these ways can
be used with systems having turbocompressors for which
an inlet quide-v me control should be added.
2. Severa~ ways of applying the preferred type of
modulation also exist. For example, a control mech-
anism can be designed in such a way that the required
load on the hOlJse is always matched by the capacity of
the heat pump.* Although this concept of applying the
modulation results in improved effectiveness (coefficient
of performance-COP), it does not yield the ultimate
benefits. Equally, applying modulation to follow solely
the operating points of highest system efficiency (with
respect to the variable ambient conditions) would not
yield the best economy. This is because of the relation-
ship between the combined component efficiency and
the changing temperature differences on heat exchang-
ers; each influences the COP's instantaneous values dif-
ferently and, at certain ambient conditions, even in
opposite ways. The resulting influence also varies from
system to system.
In our evaluation we used two types of capacity
modulation: (1) variation of speed for piston compres-
sors, and (2) speed variation and inlet guide-vanes for
turbocompressors. Also we have assumed only the
simplest form of applying modulation, that of matching
the capacity of the system with the load to the balance
points. For certain conditions [for turbosystems and
around an ambient temperature of 1So C (650 F)] an
on-off control was also used.
Although simplistic, our approach served the pur-
pose well. It identified the existence of improved energy
savings and quantified its approximate level, even for
systems with simplified versions of capacity modulation.
EVALUATION OF SELECTED SYSTEMS AT
DESIGN CONDITIONS
Early in the program we set out to prove that it
might be beneficial to the system COP to operate both
the prime mover and the refrigerator subsystems so that
the actual operating points would follow conditions cor-
responding to high component efficiencies, at least for
periods representing a high percentage of cumulative
energy inputs. Other COP benefits attributed to capacity
modulation and derived as a consequence of the de-
creased temperature difference on heat exchangers (due
to lower load at milder conditions) had already been
proven (refs. 7-9). Therefore, we decided to study the
former energy-saving potential separately and specifi-
cally for our selected systems. For this purpose, the
design-point evaluation is very well suited.
General Assumptions
The relative performance of various comfort sys-
tems may be evaluated by using values of the COP for
*The precticality of such an arrangement, the compatibility
with the house distribution system and auxiliary power con-
sumption, and the operating range of components are disre-
garded here.
336
-------
each system. For our purposes, we selected as the most
important characteristic, the COP versus the ambient-air,
dry-bulb temperature, assuming a constant indoor-
comfort level. We stress again that in this particular case
(and unlikE! the case of the off-design evaluation), all of
the COP values so calculated relate, for each point, to
variable "design conditions," meaning that the sizes of
the heat exchangers were specified only by constant
:
o
. 21
o
o
Q
. 18
en
I.tJ
i=
~ 15
Il.
~
U
en 12
2
I.tJ
I-
~ 9
en
36
temperature profiles.
To perform the graphical analysis of the perform-
ance behavior of our modulated systems, a load char-
acteristic had to be defined. The system and load char-
acteristic would then be superimposed and operating
points defined according to the desired control system.
We have done this for our design performance studies, as
indicated in figures 1 and 2.
120 .l--T~;GE---~~-- --
, MEASURED DATA
, 110 DISCHARGE P~ESS 1136 :;aSCOIS 'l
HEATING CAPACITY - - -
100 - ' ,', - ,
90 j----i - : -- - - -
33
30
27
24
--tu_--
'"
'; 80 --
a;
80"loY
I
/
1
/
. / '/'
/ 80"10/ ,/
, // ,/0"10
/ ' /
/' /
" /./ 0"10
I....,r\ . /;/ - /:
4 I //, /. ,/
I 1- ,/ '/ /800;'
. r- - ';- - //' 0 '
3 : // I ,/ . / .... / /O~o
2 ~/.::'b4' 1.-- / ~- 80"10;" ,
i --- - i ------- 1 -' -t -- --
3 -j-- - i ,-- -0"10
2 ----- --- r.. .
1 - ---
EERAGE HOUSE HEAT LOSS/.
o
g 70
------.
If>
'!! 60 -
I-
U
6
20 a:
o
If>
If>
!oJ
10 a:
Q.
::f
o
u
3
o
o
-20
-10
o
10
20
I
o
EVAPORATIVE
I
20
TEMPERATURE. of
I
50
10
30
40
COMPRESSORhMODEL 206l
REFRIGERANT R -12
RETURN GAS 18,3°C i
SUBCOOLING 9,4°C
ENGINE WITH CONSTANT
EFFICIENCY, ~.' 30 %
I I I :
--I
L-_-
~6'g'L~~~TI~~PA~~TM: ~~~C I
1 I ! ' I
1- ' , I j
: :~'" "("'~"O ;' ~.
o
o
c
-:.- -
-. ---
-
- --
o
w
..... . W
-...." 4 ~
- ....... 3 a:
o
If>
---2~
a:
I ~ I
--j-181
- -
-
-'-
'''''-
+--
I - -
-.-- -
-f--
i
- - - _. "--
- AVERAGE HOUSE HEAT GAIN
30
90
100 110 120 130 140 150 160
,'OtIDENSING TEMPERATURE. OF
I I I
80 90 100 110 120 130 140
40
60
70
TEMPERATURE. of
AMBlE N T
-30
-20
-10
CONDENSING TEMPERA'T\JRE,oC
eVAPORATIVE TEMPERATURE, °c
o
30
40
60
70
50
-20
-10
o
10
AMBIENT
20 30
TEMPERATURE, °c
40
50
60
"~IOO~I
Heating and cooling capacity of a Rankine piston compressor
heat pump driven by a constant-efficiency prime mover
(heating capacity lines correspond to 0 percent and 80 percent
recovery of the engine's rejected heat).
Figure 1.
337
-------
36
33
30
27
24
:? "'
i ~!
<
21 , .'
a
o c.
o "
a
Q "'
a
vi 18
"
1&.1 C
j: ~
~ 1:1 2
~
C
q ii
U ~
12 ~
en c
<.:
:I
1&.1
....
~ 9
en
6
3
0
24/
22 '
120
- --. COMPRESSOR-MODEL'206l
REFRIGERANT R -12
, RETURN GAS 18,3°C
SUBCOOLING 9,4 DC
ENGINE WITH CONSTANT
- . ]"""1'" ']":°-1"
EVAPORATIVE TEMP 4,4oc--tI---l
COOLING CAPACITY -'-'-, I
COOLING COP --6-- ~
""", I'" '1"""1' 1. J
RANGE OF MEASURED DATA
r-' -j
~ I RANGE OF
-- J - -~~~;~oDpTA .
DISCHARGE PRESS. 1136 kpoSCOIS
HEATING CAPACITY -.-..-' 50%
HEATING COP --0--
HEATING COP, 3500rpm"0..,
NO HEAT RECOVERY
110
20
100
18
90
c
I 6 - ~ 80.
m
- f-
10
/
/
Y
. / : .
/' '/'. I
.0' ,,"'9- / 'i""
A,"" / / ""
, "0" '/ / ' I
E f/ ,,' /' /
~ n /, /
w ,,// /
:;; 40 g /' /' /' l--
t; 2 h' // ,//
- 4 "", or./ / / /,
30 0" r' -y ,//,,/ ,
W / I .--:::::' /'"
~ 3'- /, ~ ---
20 :21 /_f1t-,.--. 'y/1/.>
~ 4. .-- I -+...-
10 ~ ~ -~~. :--=t-=:t--- '--:
~ I AVERAGE . Ho~i. HEAT LOS~
-'-",
o
I 4 - g 70
II>
1.2 - ~ 60
.~
u
...
Q
:.1) ~O
"..
), E I
~
0
, 0
! Q
08
-
+ --
o
"'. w
~'::::t 4 ~ j
---.1.3<>:
o
'"
'"
2 '"
<>:
a.
::f
I 0
u
0,6 -
-.-.
----
-,-
04
~._-_.
, . AVERAGE HOUSE HEAT G~IN
-;-
02
0__._-....._-
o
o
-20
160
-10
o
10
20
30
40 90 100 110
of CONDENSING
I
60 70 80 90
120
130
140
150
I
o
EVAPORATIVE
I I
10 20
TEMPERATURE,
I I
30 40 50
AMBIENT
TEMPERATURE, of
TEMPERATURE. of
I I I
100 110 120
I
130
140
-30 -20 -10
EVAPORATIVE TEMPERATURE,oC
o
30
40
:10 60 70
CONC£NSlNG TEMPERATIM'fE,"C
-20
-10
o
10
AMBIENT
20 30
TEMPERATURE,
40
:10
60
°c
..1'&OtOO~O
Figure 2.
Heating and cooling performance of a Rankine piston compressor
heat pump driven by a constant-efficiency prime mover (heating
COP lines correspond to 0 percent, 50 percent, and 80 percent
recovery of the engine's rejected heat).
Results
The important results of our design-point evaluation
are presented in ";igure 3. There, for reference purposes,
a band of COP's j,; accentuated. Its lower limit represents
typical values of COP's for a residential, electrically
driven heat pump, as given by a manufacturer. These
values include the average efficiency of electric-power
generation and transmission (29 percent); for cooling
they represent the total capacity. Supplemental heat,
auxiliary power, and motor efficiency are excluded. The
upper limit of our reference band is a COP curve calcu-
lated for a system comparable to an electrically driven
338
-------
w
u
Z
-
(/)
20
/
LLI.END
DRIVING A QS,3 em'. R 17 CUMPRESSOR
-.- )()"- EFF. ENGINE
~ STIR LING \ Nol MOle"rcllo. Br'l Eeo. IH'IIIY I j'
--0- OTTO 114 "pi V
DRIVINr; AN 13Q,3 em'. R 17 COMPRESSOR
- -/>-- RANKINE .
(PI\!OI! 5.11';1111 [11(1ll1r With 90"" HI',\!
Hf'(ovrrv, Nu PII"....,WP Dlop
Our IIItI POWf'1 C Vf If' t ..
-;}--OTIOIIOlipl
\
\
~\
BRA VI ON
.-.-'
10
..-
.; -
..-..-"-
o R,ILulff' 1'01111'.
'iJ-.- .
For RANKIN[ ',v.,tI'm'> 011111001 fod
',1: 6,6°( I ill 111',1111111. "," °c IOf ((Julu111
AfP'IIJ;IIV I'nf'r(jy r"f!LJllpd
Dly.11f oj)I',;,111I1I ,'''''..11111'11
1J1l1f'SS slatrll othrrw.\f', SOG.. hl';11 ff"(OVf'fY ;'SStHT'f'l~.
o
20
40
60
AMBIENT DRY BULB TEMPERATURE, of
-20
o
I
-30
o
10
20
BULB TEMPERATURE,OC
30
40
50
-20
-10
MBENT DRY
Figure 3.
\
80
100
120
'760100~Z
Coefficient of performance of various air-to-air heat pumps
(for engine-driven systems, the values do not include the
benefits of decreasing temperature differences on the heat
exchangers).
339
-------
heat pump. In this case, however, the compressor capa-
city always eq~Jals the house load. This is accomplished
by varying tht! compressor speed within the practical
range of 1,000 to 4,000 rpm. The driving hypothetical
prime mover has a constant efficiency of 30 percent,
and, for heating, 50 percent of its rejected heat is re-
covered. The curve, therefore, represents the poter:\tial
performance 0': a piston compressor with a selected type
and applicatior of capacity modulation. However, this is
still without the benefits resulting from decreasing
temperature di1ferences.
Superimposed, one can find the performance curves
of several practical systems, which were calculated by
using available characteristics of refrigerators and prime
movers.
Conclusions
Because of its limited value, the design point evalua-
tion was not curied out all the way to the assessment of
the year-round performance. However, it was sufficient
to comparativEly demonstrate the increased energy-
saving potential of various modulated systems with or
without recovel y of the engine's rejected heat. Also, the
results are sufficiently significant to suggest three basic
concfusions:
1. The positive effect of capacity modulation on
the COP of the ,malyzed systems, derived from operating
the refrigeratin!1 subsystems at high efficiencies, is signif-
icant, but it CHO be destroyed by improper matching
with prime mo~ers. Because of the other positive effect
of modulation-.namely, the effect of variable tempera-
ture gradient on heat exchangers-the systems with capa-
city modulation will always consume less energy as com-
pared with unm~dulated systems, when proper attention
is given to selecting and matching a driver. Therefore
capacity modulttion should be regarded as highly desir-
able for increasing the energy-saving potential of a heat
pump.
2. Substantial improvements in the heating COP
of a residential heat pump can be achieved by using an
engine to drive the refrigeration subsystem. I n practice,
at least 50 percent of the engine's rejected heat can be
used directly to boost the heating COP, heating capacity,
and temperature of the delivered air.
3. Capacity-modulated systems, specifically those
driven by an c;nsite prime mover with rejected heat
recovery, show
-------
Results
To show the significance of the off-design perform-
ance analysis and the significant differences in COP for
the design-point and off-design evaluations, we calcu-
lated the two characteristics for a Rankine piston, com-
pressor heat pump driven by a hypothetical prime mover
with constant efficiency of 30 percent. The comparative
performance results for this simple system are presented
in figure 4.
The significance of these results is related to our
definition of the design-point performance. As we have
defined it, the difference between the two approaches
was that the design-point evaluation specified the heat
exchangers only by a constant temperature profile (leav-
ing their size variable), while the off-design performance
corresponded to a totally designed system. Therefore, in
figure 4, the less steep COP curves respresent only the
benefits of operating the compressor at higher efficien-
cies when modulating by speed, and the steeper COP
lines reflect the additional benefits of decreased load on
the heat exchangers.
The rest of the results gives an example of the com-
parative performance evaluation of a speed-modulated,
Stirling engine-driven, Rankine piston compressor, heat
pump with engine-heat recovery. This system was one of
those that we selected as having high potential for be-
coming a practical, reliable, and efficient residential heat
pump. For this evaluation it was "designed" to serve a
typical single-family residence in the Chicago area, and
its off-design performance is given in figure 5. 11;1 com-
parison with other selected systems, its COP character-
istics as a function of the ambient air, dry-bulb tempera-
ture are given in figure 6, and its comparative energy-
saving potential is given in table 2. Also, in table 3 the
total annual emissions for the evaluated systems are
compared.
CONCLUSIONS
A study such as ours must be supplemented by an
evaluation of other factors related to operating and
manufacturing a system if it is to be used as a guideline
in selecting a system for preferential development. These
additional aspects would include:
1. Comparison between the performance of resi-
dential and district-type heat pumps,
2. Assessment of auxiliary power consumption,
3. Evaluation of other types of modulation and
other (optimal) ways of al;>plying modulation,
4. Assessment of control complexity,
5. Evaluation of interaction between the modu-
lated heat pumps and the distribution system.
The influence of some of these aspects on the sys-
tem performance might not be secondary. However, our
results are important because they confirmed our con-
clusions, made previously, of design analysis, and they
define and quantify incentives for additional system
design and experimental research.
. The components used in our evaluation were not
specifically designed for heat-pump service, and we have
necessarily assumed that the required design changes
were practical. We were able to assess the feasibility of
these changes only theoretically, but we are confident
that some of the systems (such as Stirling-Rankine
piston system and Brayton-Rankine turbocompressor
system) could be developed into marketable hardware.
Finally, as a result of our design and off-design
assessment of various residential heat-pump systems, we
have concluded that practical engine-driven heat-pump
systems can be developed that would indeed show per-
formance improvement relative to existing systems. One
such candidate identified as extremely promising is the
Stirling-Rankine piston heat pump.
REFERENCES
1. J. Wurm et aI., "An Assessment of Selected Heat-
Pump Systems," Annual Report, Project HC-4-20
for the American Gas Association, I nstitute of Gas
Technology, Chicago, Illinois, March 1974.
2. J. Wurm and W. F. Rush, "Evaluation of Engine-
Driven Heat-Pump Systems of Small Capacities,"
paper presented at the 14th International Congress
of Refrigeration, Moscow, September 20-30, 1975.
3. E. C. Hise, "Seasonal Fuel Utilization Efficiency of
Residential Heating Systems," ORNL-NSF-EP-82,
Oak Ridge National Laboratory, Oak Ridge, Tennes-
see, April 1975.
4. J. B. Comly, H. Jaster, and J. P. Quaile, "Heat
Pumps-Limitations and Potential," paper presented
at the Heat Pump Technology Conference, Okla-
homa State University, Stillwater, Oklahoma,
October 20-21,1975.
5. E. G. Granryd and J. Wurm, "Investigation of Sys-
tem Performance Improvements by Means of Varia-
tion in Component Characteristics of Engine-Driven
and Absorption Heat Pumps," Second Quarter 1975
Report, Project HC-4-20 for the American Gas
Association, Institute of Gas Technology, Chicago,
Illinois, Cctober 1975.
6. W. F. Rush and S. A. Weil, "Assessment of an
Ammonia-Water Type Absorption System as a Heat
Pump," paper presented at the 14th International
Congress of Refrigeration, Moscow, September
20-30, 1975.
341
-------
4.0
a:
o
u
-
~
N
w
u
z
~ 3.0
a::
f2
a::
w
a..
1.1..
o
~ 2.0
w
U
~
1.1..
W
o
u
(!)
z
:J 1.0
o
o
u
CI
z
«
(!)
z
~ 0
W
:L
""
"".
."",,'
."""
"""
".
.,,'
- OFF-DESIGN PERFORMANCE ANALYSIS
--- DESIGN-POINT PERFORMANCE ANALYSIS
--- DESIGN-POINT COP ASSUMING 50%
REJECTED HEAT RECOVERY
o BALANCE POINTS
SUPPLEMENTAL HEAT PUMP EFFECT PLUS HEAT PUMP
HEATING ONL~ I.. SUPPLEMENTAL HEATING -I EFFECT
"...-
",,""
""
"""
"
./
/
/
./
-if'"
"
"
"
"
"
.......
"
.......
"
-30 ...20 -to 0 10 20 30 40
AMBIENT TEMPERATURE (Dry Bulb), °C
I I I I I I I I I I I I I J
-20 -to 0 to 20 30 40 50 60 70 80 90 100 "0
AMBIENT TEMPERATURE (Dry Bulb), OF 816010014
I
-30
Figure 4. Comparison of system COP's as calculated for design point and off-
design operating conditions for a Rankine (R-12) piston heat pump
driven by a constant efficiency prime mover.
-------
'pm
4000
25 HEATING
STIRLING ENGINE - RI2 COMPRESSOR
8 . DESIGN POINT.
3500 TEMPERATURE OF EVAPORATOR AIR 500F
(t.T> EVAPORATOR 15°F
7 It. T) CONDENSER 30° F
20 3000
..E
'-
::>
iD 6 2500
~
~ Q HOUSE
.:£
~ COOLING COOLING
~ HOUSE LOAD
f- 15 ~ 5 HEATING~
U u LOAD 2000
it it
-------
a..
o
u
C)
Z
..J
o
o
u 2
a
z
-------
Table 2. Performance of selected residential heat pumps for
a single-fami~y residence in the Chicago areaa
Heating mode Cooling Mode Year-round
Energy Savings COP Energy Savings COP Energy Sa vi ngs COP
Input, compared Input, compared Input, compared'
kWh to System kWh to System kWh to System
A, percent B, percent B, percent
Heat supplied or removed
from conditioned space 36,883 4429 41,312
A. Thermal resista,nce
~ hea t i ng system 127 , 182 0.29
01
B. Electric heat-pump
system (constant
speed, 3,500 rpm) 43,856 65.5 0.841 5090 0.87 48,946 0.84
C. Electric heat-pump
system (speed
modulated) 35,248 72.3 1.046 2015 60.4 2.198 37,263 23.8 1 . 1 08
D. Stirling engine-driven
heat pump system
(speed modulated plus
heat recovery) 24,721 80.6 1 .492 2917 42.7 1.518 27,638 43.5 1.49
a heat in form of resistance heating; total cooling season, 1,287 hr; total heating season,
Supplemental
5,953 hr; same piston, R-12 compressor used for systems B, C, D; auxiliary energy input not included.
-------
Table 3. Total annual emissions for selected heat pumps serving a
single-family residence in the Chicago area
Heat pump system
Electric
{speed modulated
'1,000-4,000 rpm)a
1 b/l 06 Btu 1 b/yr
---~-
Po 11 utan~:s
Electric {constant
speed, 3,500 rpm)a
lb/106 Btu lb/yr
Stirling engine-driven
(speed modulated
plus heat recovery)
lb/106 Btu lb/yr
ParticulCltes 0.1 16.7 0.1 12.7
S02 1.2 200.4 1.2 152.6
NO 0.7 116.9 0.7 89.0 1.056 98.53
x
CO Negligible Negligible 1.56 153.3
a on coal-fired generating plant with input of 250 x 106 Btu/hr
Data based
or more.
7. E. R. Ambrose, "The Heat Pump: Performance
Factor. Possible Improvements," Heat Piping Air
Cond.. Vol 46. pp. 77.82. May 1974.
8. R. E. Cawley, and D. M. Pfarrer, "Part Load Effi.
ciency Advantages of Two-Speed Refrigerant Com-
pressors." paper presented at the Conference on
Improving Efficiency in HVAC Equipment and
Componen ts for Residential and Small Commercial
Buildings, Purdue University. West Lafayette.
Indiana, October 7-8, 1974.
9. E. G. Granryd. "Various Heat Pumping Processes,"
paper presented at the Seminar on Heat Pumps
Sponsored by the Educational Services of the Fin-
nish Society of Engineers, Helsinki, September
1974. (Swedish test).
10. U.S. Department of Commerce. Weather Bureau,
Climatography of the United States No. 82-11,
Decennial Census of United States Climate, Sum-
mary of Hourly Observations, Chicago, Illinois,
1951.1960, U.S. Government Printing Office, Wash-
ington, D.C., February 1974.
346
-------
DEVELOPMENT AND APPLICATION OF AN OPTIMUM DISTillATE Oil BURNER
HEAD FOR RESIDENTIAL FURNACES*
.
L. P. Combs, W. H. Nurick, and A. S. Okudat
Abstract
Experimental combustion research techniques have
been used to develop an optimum, low-emission, high-
efficiency head for conventional high-pressure atomizing
distillate oil burners such as are used in residential
space-heating furnaces and hydronic boilers. Application
of the optimum oil burner head would offer both signifi-
cantly reduced consump tion of fuel oil and reduced
emissions of air pollutants, particularly oxides of nitro-
gen. Current research is desCribed which is directed
toward applying the optimum head in two comple-
mentary ways. One approach is to commercialize the
optimum head immediately for new and replacement
OEM burners and for retrofitting existing burners at low
cost. The second approach seeks additional improvement
by further optimizing furnace components to match the
optimum burner head, and so would provide greater
benefits after a somewhat longer commercialization
period. ~
INTRODUCTION
The U.S. Environmental Protection Agency has
sponsored studies over the past few years to document
the emission of air pollutants from existing residential
and commercial oil-fired space-heating equipment (refs.
1,2). Concurrently, the EPA has also supported applied
research programs to determine the effects of design and
operating parameters on exhaust gas emission (refs.
3,4,5). These and related studies have shown that sub-
stantial reductions in total emissions from furnaces can
be realized by applying combustion control technology,
such as flue gas-recircula}ion, staged-combustion, and
advanced burner designs. .
In addition to providing low levels of pollutant
emissions, new technology for residential furnaces
should also provide some substantial increase in overall
thermal efficiency. This would provide the economic
incentive for homeowners to replace antiquated equip-
ment by insuring capital recovery within a reasonable
number of heating seasons. Equally important is the
.Work upon which this publication is based was performed
pursuant to Contract No. 68-01-0017, 68-02-1819, and
68-02"888 with the U.S. Environmental Protection Agency. Mr.
G. 8la\haS be.en EPA's Project Officer on all three programs.
~ tRockwell International!Rocketdyne Division, Canoga Park,
rlifornia. '
potential for conserving energy by reducing residential
consumption of clean fossil fuels. Thus, the dual benefits
of improving air quality and conserving the Nation's
dwindling energy resources are strong incentives for
timely development of advanced technology for residen-
tial heating systems,
This paper describes a series of related research
investigations, which began with the application of
experimental combustion research techniques to the
optimization of conventional, high-pressure atomizing
oil burners. Optimization was with respect to minimizing
pollutant emissions .and improving thermal efficiency,
and resulted in design criteria for the burner head. Con-
tinuing work in progress is in two directions: (1) opti-
mizing the application of the optimum burner head by
determining effective burner/firebox matching criteria
and applying them in a prototype furnace and (2) evalu-
ating the feasibility of commercial production of the
optimum head for installation in new furnaces and for
retrofitting old ones.
BACKGROUND
Some typical residential oil-furnace emission and
efficiency performance data are summarized in table 1.
Data on cycle-averaged emissions were' abstracted from
references 2 and 4. The existing furnace population con-
sists of a broad mixture of burner and firebox types and
ages so that its average emissions of incomplete combus-
tion products (CO, UHC, smoke, and particulates) are
slightly higher than the representative values for current
new furnaces.
An NOx emission goal of 0.5g NOx/kg fuel burned
was established by the EPA as one that is both' accept-
able from an air quality standpoint and achievable by
application of blue-flame burner'technology. However,
blue-flame burners characteristically have start transient
problems which have prevented them from being devel-
oped commercially, so it was hoped that the goal could
also be attained by conventional burners.
Wide ranges of thermal efficiency estimates for the
existing residential furnace population may be found in
the technical literature. Those listed in table 1 are repre-
sentative of the midrange and are generally consistent
with the flue gas CO2 concentr.ations (or excess air
levels) reported in references 2 and 4. Recognizing that
gross efficiencies, based on the higher heating value of
the fuel, are about 5 to. 7 percent lower than net. effi-
347
-------
Table 1. Typical residential oil furnace data
Average of existing
furnace population
Average of current
new units
~/cl e-averaged
emissions
Smoke (bacharach
scale)
CO
UHC
Particulates
NOx
Thermal
-~
efficiencies
Steady-state
Cycle-averaged
- No.2
- 0.6 g/kg fuel
- 0.08 g/kg fuel
- 0.3 g/kg fuel
- 2.5 g/kg fuel
65 to 75 percent
50 to 60 percent
- No.1
- 0.3 g/kg fuel
- 0.06 g/kg fuel
- 0.2 g/kg fuel
- 2.5 g/kg fuel
75 to 80 percent
65 to 70 percent
..
ciencies, it may be estimated that fuel utilization effi-
ciencies potentially might be increased by as much as 35
to 40 percent. Achieving such large gains, however.
would entail very significant departures from current
design concept~, and manufacturing, marketing, and
utilization practces. Considerably larger heat exchangers
would be needed to cool flue gases to near-room temper-
ature and cond~nse their combustion-generated water
vapor. A number of problems would arise immediately,
e.g., inadequate firebox draft, need for corrosion-
resistant furnaa! and flue construction materials, con-
densate disposal, and noncompetitive initial costs. For
these reasons, it was deemed appropriate for the current
research to add'ess efficiency gains achievable by less
drastic modificaj:ions of existing commercial technology,
Le., improvemer,ts that can be achieved relatively near-
term and remain cost-competitive in the marketplace. A
target value o.f 10 percent or greater efficiency increase
was thus selectee:.
.
Reducing Pollhtant Emissions ~
..
When clean fuels, such as nallnal gas and No.2 fuel
oil, are burned in residenti:!1 space-heating equipment,
t:wo classes of combustion-generated air pollutants are
emitted, viz., incomplete combustion products (CO,
UHC. and smoke) and oxides of nitrogen. The two
classes are similar in some respects. Their concentrations
in the flue gases are usually very low, so that their pro-
duction does not significantly affect combustor per.
fomiance or thermal efficiency, and their concentrations
usually are not particularly close to those rep.resentative
of thermal equilibrium, but are more likely to be deter-
mined by poorly controlled features of combustion, viz.,
regions of poor mixing, spray/combustor wall inter-
actions, size and strength of recirculation eddies, and
flame-zone cooling phenomena. These characteristics
tend to obviate quantitative analytical characterization
of the. pollutant emissions; as a result, such characteriza-
tions are usually studied experimentally. Conversely, the
two classes are dissimilar in some respects, principally
because their usual concentrations lie on opposite sides
of equilibrium. Thus, for example, NO production is
enhanced by higher temperatures and longer flamezone
residence times, both of which tend to drive the hydro-
carbon combustion processes to completion and reduce
carbonaceous emissions.
Improving Thermal Efficiency
The greatest source of thermal inefficiency in resi-
dential space-heating equipment is the convection of
heat up the flue. When the burner is being fired, ex.
hausted combustion product gases carry off substantial
348
-------
. sensible heat as well as the heat of vaporization of com-
,bustion-generated water vapor. These losses can be
reduced by decreasing the excess air level and by extract-
ing more heat, Le., lowering the flue gas temperatures.
Steady-state net thermal efficiencies might be raised by
5 to 10 percent before furnace operation is constrained
by pollutant formation, low draft, and condensation
(corrosion) problems.
Also, while the burner is being fired, most installa-
tions use heated, humidified, living-space air for both
burner combustion air and barometric draft control air.
Although the resultant heat losses are not charged
, against furnace thermal efficiency, fuel utilization may
be raised by 10 percent or more by using a sealed air
system to bring in outdoor ambient air for these uses
(ref. 6).
When the burner is not being fired, a natural draft
flow of air through the burner, firebox, etc., continues
to extract heat from the furnace and convect the heat up
the flue. This mechanism can reduce net thermal effi-
ciency by as much as 5 percent. These draft air heat
losses are a major aspect of the transient heat losses that
cause cycle-averaged efficiencies to be lower than steady
state. The lower the draft air ~oss, the less sensitive is
cycle-averaged efficiency to variations in cycle duration
and fractional burner on-time. Draft air heat losses can
be eliminated by providing a positive shutoff device in
the combustion air supply. They can also be reduced or
eliminated by firing the burner more nearly continu-
ously, e.g., ~y using modulated-flow or high/low/off
burner control. No thermal efficiency advantage can be
realized with such control schemes, however, because
the flue gas temperatures vary enough during modula-
tion to offset any reduction or elimination of draft air
losses.
Heat conducted to the exterior cabinet of a furnace
,is radiated and convected to the surroundings. Al-
though treated as a furnace loss, this heat, in some in-
stallations, may contribute directly to heating the
residence and may not be a true heat loss. These fur-
nace-setting losses average about 1.5 to 2 percent for
warm-air furnaces. They are higher (3 to 3.5 percent) for
hydronic boilers, because component temperatures are
more nearly constant during standby than are those in
warm-air units. Reduction of these losses nominally
depends upon a straightforward economic tradeoff of
fuel economy vs. more expensive insulation, although
manufacturers must also be concerned about first-cost
competition.
. ,
The electrical energy eXiJended in operating fur-
naces amounts to no more than 2 to 3 percent of the net
heat of combustion of the fuel. Nevertheless, minimiza-
tion of electric power consumption is a worthwhile goal
because electricity costs the homeowner more than fuel
oil and because there is a factor of 2.5 to 3 times as
much total energy savings when the inefficiency of the
electric generating plant is considered.
To maximize furnace thermal efficiency without
departing substantially from current manufacturing and
installation practices, the following measures should be
effected:
1. Design the burner/combustion chamber com-
bination to operate pollution-free at low excess-air
levels, e.g., 10 to 15 percent excess air [13.5 to 13.0
percent CO2 in the flue gases (dry basis)] .
2. Lower flue gas temperatures as much as pos-
, sible. The minimum practical temperature is probably
around 2000 C (3920 F).
3. Employ sealed air systems for combustion air
and barometric control. Filter the combustion air to pre-
vent long-term degradation of performance.
4. Close the combustion air supply during stand-
by.
5. Effect savings in electrical power consumption.
Areas to consider are proper matching of drive motors to
air fans and fuel pumps, interrupted-spark rather than
continuous-spark ignition systems, and solid-state igni-
tion and control circuits.
EXPERIMENTAL STUDIES
Oil Burner Optimization
The initial segment of the oil burner and furnace
investigations was concerned with reducing the emissions
of pollutants from conventional oil burners. The re-
search results have been fully reported by Dickerson and
Okuda (ref. 5) so are only briefly summarized in this
paper.
Advanced methods for flow visualization and for
mea~ring atomization, spray-gas mixing, gas-gas mixing,
and reaction product concentrations in flames were used
to study oil burners and combustion chamber design~
that are reasonably close to current conventional com-
mercial practice. Commercially available OEM oil burn-
ers representative of current technology were first tested
in several sizes to determine their atomization character-
istics and the spatial distributions of flow velocities, air/
fuel ratios, and reaction species concentrations that they
establish within a combustion chamber. The objective
, was to relate burner emission behavior to specific design
concepts and design parameters. Burner design features
important to pollutant emissions were identified and
"versatile" research burners, provided 'with continuous
adjustability of those features, were bu ilt in two sizes.
These research burners were fired in a range of refrac-
tory-lined and cooled-wall combustion chambers to
349
-------
w
(J1
o
(a) EXTERNAL VIEW
: ! ",{ r
--,'-
r---r,--,-
4/ r-:;r- I 1 '''''1
,.~~
(b) OPTIMUM HEAD
Figure 1. Optimum 1 mils (gph) oil burner.
-------
~termine, principally, optimum combinations of values
for their variable design parameters and, secondarily,
combustion chamber design attributes supportive of low
emissions and high efficiency. Optimum combinations
were found which, when fired in comparable combus-
tion chambers, reduced emissions of oxides of nitrogen
(NOx) by up to 50 percent with acceptably low emis-
sions of CO, smoke, and unburned hydrocarbons, and
without impairing achievable thermal.efficiency. It was
found that NOx emissions were about 50 percent lower
when the optimized research burners' axes were aligned
with the combustion chamber axis (tunnel-fired). rather
than made perpendicular to it (side-fired).
Two sizes of fixed-geometry "optimum" burners
were then built and tested to corroborate the versatile
research burner results. One of the burners was sized for
a nominal firing rate of 1 ml/s (gph) * (fig. 1(a)). which is
typical of burners for residential oil furnaces. The key
design parameters for optimizing the burner were con-
fined to the burner head, an internal view of which is
shown in figure 1(b). namely, the swirl vane size, or
effectivity, the swirl vane angle and the discharge choke
diameter. With sufficiently large swirl vanes, radially dis-
posed vanes canted 25 degrees from the burner axis were
optimum, irrespective of the burner firing rate and
~cess air level. Conversely, the optimum choke plate
!pening diameter was found to be proportional to the
oil firing rate raised to the 0.4 power (ref. 5).
Other burner components, viz., the housing, blast
tube, fuel pump, combustion air fan, drive motor, igni-
tion transformer, C~~ flame-sensing cell, and burner con-
trol circuitry, were. stock equipment supplied with a
commercially purchased burner. The stock ignition elec-
trodes were replaced by a smaller-diameter pair com-
patible with the more restricted space imposed by the
swirler vanes. Use of a 1.0-60° -A (1.0 gph, hollow cone)
Delavan spray nozzle provided good air/fuel spray mix-
ing while minimizing spray impingement on combustion
chamber walls.
The pollutant emissions from the fixediJeometry
optimum burners did reproduce those measured from
the versatile research burners. NO emissions are exempli-
fied in figure 2. It was also demonstrated, by testing this
burner in typical residential furnaces, that the optimum
burner research results are directly applicable to fur-
naces.
APPLICATION OF THE OPTIMUM HEAD
Two complementary approaches are being taken
.oward further exploration and application of ths opti-
, 01.00 mils = 0.951 gph. The 5 percent difference is neg-
lected in this paper when reference is mada to burner firing rates.
mum oil burner head technology. One approach is dir-
ected toward immediate applications in new and replace-
ment OEM burners and, possibly, in retrofit heads for
existing, installed burners. If found to be feasible, direct
commercialization of the optimum head alone would
offer prompt, albeit modest, gains in fuel economy and
reduced emissions. The second approach is directed
toward furnace manufacturer applications in new fur-
naces. This approach is somewhat longer term but
should offer greater gains in both efficiency and emis-
sion levels, because it involves further technology
improvements.
Feasibility of Commercialization
The feasibility of commercial application of the
optimum low-emission oil burner head is being investi-
gated by fabricating prototype heads to simulate those
which would be produced by appropriate commercial
fabrication techniques, and testing them to determine
their durability and whether their emission behavior
duplicates that of the preprototype research optimum
head.
Fabrication Techniques. After cursory consideration of
various types of fabrication techniques, three were
selected as being potential candidates for commercial
production of burner heads. These three methods are (1)
iron casting, (2) sheet metal stamping, and (3) injection
molding. A comparative evaluation of them was con.
ducted; considered were unit costs, manufacturability,
saleability, and design compromises likely to be encount-
ered. A comprehensive summary of the evaluation is
given in table 2.
The information in table 2 is essentially self-
explanatory. The table is concluded by assigning a
numerical ranking to each method as a result of compar-
ing all of the information. Reasons for the ranking are
also given in the table. The sheet metal stamping method
is clearly the best choice, and iron casting, the preferred
second choice. Aside from estimated unit production
costs, casting was downgraded with respect to sheet.
metal stamping because of uncertainty regarding the
effects on emissions of anticipated design compromises.
Prototype Optimum Heads. A preliminary design was
made for a 1 ml/s (gph) optimum head to be made by
sheet-metal stamping, and two prototype heads were
constructed to that design. (These heads were formed by
shearing, machining, and bending operations, rather than
by their being actually stamped.) The design layout is
shown in figure 3, and figure 4 is a photograph of a
prototype stamped-sheet-metal optimum head.
Experimental Testing. The prototype optimum head is
351
-------
2.0
.~""~~~~~~~~"""~.~'
~ OF~RCIALLYAVAILABLE -
~,~~~~~~~"~(Z,"',"""~ .
Q
...
z:
II:
::I
III
~ 1.5
w
::I
IL
IL
o
E
f 1.0
..
!
)(
o 0.5
~
II:
l-
i
- - - - - - - - - - - - - '- OIL NOZZLE
. 'F.. - - - - ---- '-.: 1.00600.A
- - FIXED GEOMETRY OPTIMUM BURNER'-I.o-900'A
1.II-INCH CHOKE DIAMETER
SIX 2110 SWIRLER VANES
o
1.0
1.10 1.20
nOICHIOMETRIC RATIO, ~AIR/FUE L1I14.49
1.30
Figure 2. NO emissions comparison (tunnel-
fired, refractory-lined combustor).
being tested extensively, first in research combustion
chambers and IHter in commercially available furnaces.
The research combustor testing will establish corre-
spondence with the preprototype research optimum
head, which has been tested extensively in these com-
bustors, as descl ibed in the next subsection. In the fur-
nace tests, therr"1al efficiency and ranges of operability
and applicabilitv will be established. This testing will
employ three n~w warm-air furnaces having different
firebox construc:tions and burner orientations. These
three were selected as being representative of over 50
percent of the residential warm-air furnaces found in the
field.
Currently, the prototype optimum head has been
tested only in II side-fired, refractory-lined, research
combu51:ion chamber. The measured flue gas pollutant
emissions corresponded very well with those from earlier
tests of the preprototype head. I n less than 50 hours of
testing, however, it was apparent that some design modi-
fications will be needed to improve the durability of the
head, as there WIIS significant scaling of the face of the
choke plate as \/IIell as some djstortion of its surface,
which was apparently caused by nonuniform heating. A
stiffening ring IT'ade of AISI ""1 ~tainless steel was
bolted to the fac e of the choke r-iate and testing was
continued. The results confirmed that the stamped sheet
metal optimum head should be made of 300 series stain-
less steel to elimi.late the scaling and should incorporate
substantial strengthening of the choke plate to avoid
warpage.
Future Effort. Following further evaluation of pollutant
emissions when a burner with the prototype optimum
head is tunnel-fired in a refractory-lined research com-
bustion chamber, a second series of prototype heads will
be fabricated and tested. An improved stamped-sheet- .
metal design will be used; it will involve either stamping
circular stiffening ribs in the choke plate portion of a
single-piece head or riveting two stampings together to
form a two-piece head. These design changes, together
with the material change, are expected to result in a
modest increase in the estimated cost of commercial
production of stamped heads. Furnace testing will pre-
dominate the experimental evaluation of the second
series of prototype heads.
Furnace System Optimization
This segment of the research is directed toward
reducing pollutant emissions, particularly NOx' even
further while effecting greater improvements in fuel
economy than can be attained by simply adopting the
optimum head in existing furnaces. The approach taken
was to determine optimum combinations of furnace
component designs, using both experimental arid analyti-
cal techniques, and to design, build, and test a prototype
furnace embodying the results.
352
-------
Table 2. Comparison of optimum burner head commercial fabrication methods
Ctat Yelpt "'"
hbrtt.tlOll ,.. [~,:t4 '(I" of "'11119
TlLhnIQ\18 "..hrlll r "0101,,.. loo,ooo/'r 1,000,000/" hrt h.t..", ~t'7 ~rG8I"1 rWM.tlon.1 (onlt4.rllton ,..OINt". Durin, lut
(a,t t",. .... UDt7 $1.10 ..ell 11.40 ..e. 0.'1 lot I.Loella ~.. 10 ,.., eo-l 1."1,.1 w.... to ,...ltt) I.M'" \88IrI1...twre .ull", I.e.. _ellIM off "'''''
",hl'lIt 4....., IMhMlI", 1",lud'''I 1.~::t:tI' IU ..rU, ""'" ,.C:::~~'::..:.roa'::';",. ,nh..nct -, be prcOl. ,.:,..~.~.(:t:I:;O::'
cII' t,. .Ilvtry ::f: ::1: te.., IIlth ~1'r 'rQ ,....
o.'mllft8d .., 'rG8 ,en.e, hol. 10 au :;fo~"'''' twit rnht..' brul addltt..1 ...,.
..I;h.tlDII _neall.bl., ,,,, Oft WI,," 0' 8101.:
,..,1"1_1. J.IIII.""" thlca",.... . "111 ,.. cool., UII.. '_I
pro\l'" "".. ,.", t.o _t.1.
4.i~::.~:':~', t. wi.
S_,...., o. \400 '1.""c" ".11..[11 .11". I."" 10 yet, 1011 '.bductlon 'ft VaM ,.",tIII LMoln ,roYII4 .dge ,11.. I. Vln,. nn be bent 10
It.t,,I... . ""I ,,...Iwdlnt l"dlllSl... IppeUIlltI ".rl', durl", to .1"I.h. UDUI)' 0' 118 ,Ir 1""'08. .ltf.,..nt '''91.. "
1,"1 111'1.." 121.'" ::1:: :r.knod.out t..t. .t.r1l1IllM. I.~ ..rpa" ., OCtwr ~utr".
looltng unter ...1. 0."...,.. tit I.C." b8 ..cSt to or'I,...1 4ur'1tI -UAI. '.Mlt.rt.1 u" be weldld
to IdJUlt ,,"'utlon ~:~:~:~OII .t - .11 to .'1111 or uftt.r hol,.
flow. Nlhtuu. )."''''1 or tlfthr "1.
).lIgI\t . "I' '.lnc:,.........I.to d. ell be trl"'.
to url'J
..E..,ta
I.,tal1.
:idt~- AIt.IM "... II.n ..cJI; 1.11 Hell .1'., ..Good 100"1'11 10 188r lOll 1..._,t- I. 'PI l..tIII "Put. wtll end If ttI.,.1 I.P.rt CPMt M ,erl-
ur.tc .....h tKI_IIII Inc1181811., _"I PI" ".rl" durllll to 111- .141.., :::t~I~ :rI:;':~~t M ... --.
"Itvery :,: ::;1: wtU. vood tnt. I."'tcbr IlCtlOM' '11.
Iwrflu ~=:~-, I.t rid" to .11081 1.11 probl. "tu a.tdatt-
,t,."" I. r~~n:i .,.10 to ..,1. It ope,.U", ~'It....
,."Ittl.. .tl1 rnhtanc..
.,... tf .,..,..
ped on hrd
Iwrllc..
-
r.brlutlcNl II.. 01 CMfllt... "1Irlcatl0ll ~1: g~I~~~:i::r In,t.ll-tlc.
T.elullque "'''rtal 101lrillCil "'.... <-to Recca.ndl t t on.
ellt troe Oftu INtt.,.. h.." "Productt. tole,.K.1 lequl,.1 .1110' toOling "'tell drtl1 hol" I.PMr toltrtllCl 0lIl " chatn blu",. 0' relSlIfI
:.::Tro:':::~I:;1 2.:~:: ...,1. t ..0 (OIt, "'S,ooo/dil chtnte In tub' I _d I 1.0. . 0.0. wlUII. Ibl. ,f"CMtutttCIII COtto ...
.tt.th w11" 4th. out co.tI, tool I.., (0".
CIA be "'d. J':£.~:U~ai:c~=:! Icr. _chlet..
,,,' It,p wltll lha,.,
cOtMr ridlul.
s.ll1I'_tal c... be _ell out ~:~:.:'::t~. a~.R'O Ma1 be abl. tAl u.. .ncKk. NI'e" drill ...1.. " cflot" MUu" of co,t.
0' lJ\16Kttl. ~::t~:i:t;:."~i~ 0:..,. I. t\ill I lMad a IIGhtwelotlt. nit ability
I"'et ..t.rI11. .n.c' wlOl ._t "wll~t n"atillty,
tlonal 'o:~'''' ..tallcr. 1",t81hIlOft..... FlIa-
~::1 ::'1:r ":;:u~t::..:~ ~:::'I:~ty t:'~~Tn~O:~.-
:~.::n'"~:II~:~~I~~:'~O
tlenal co.t -11ooo/c""", _n'..
:':i:~~on Toollnll good for ~:~::~I:n~~.t t::R1 Ctft '''""r: ...,. 111.1 I. Orll1"'I.I,,'. "Iota. ...)' to tn."" n c""u b'''1III1I or rh'
on" QM IIIt.rlll ~~;1~~ ~:'t~lt to .tdl -.01, t. 1.lot lu111 duelopH .f PIF' crn""'. ht.l",
... to .hrtAh" Mly .., durlny 118141. IICU,. "lttI hbrlc",tton proc.II 0' dew.lopllllon' YI"lttltt)',
of ptrt .ftel' I'IIIU n,. "rt ,tIMid I.,. I""t ..tat Ie.... lI),e ria. In .""'" lid 01 prodwttlOfl ftttHty
.1.t",. thl'CNCltl t"'. uhedu'. 1 .r. un. II"" cI,lhl Inw..-..t
==~r::C::. c.rtlt"')' t. procN(tlCIII Nqlll""
COlt.
J.Mo ..htt", productlClft
'.ctll"'. MIn.' "' WI
",hted tMlt14t"".. .tc..
to IIoust aqulc-tt.
353
-------
NOTE:
rRJU) IT FODT OF IMfE3
10.11£ a&-.z~ ~ ~AC£.
, ~...... Y"!!'! ~ fP'JY..L~
DIMENSIONS ARE INCHES
t.)
CTI
~
.10 T'(P
-
.0'"2 ...""
.,,( \ .
OS3 ,).
Y\~}!.OO
3.80
Figure 3. Layout drawing of the stamped-formed, sheet-metal
prototype optimum burner head (Mat'l: AISI 430 steel).
~.%o tIIA
-------
1m
----,;
p._~.t..
iIiII
'"
i!!
III
.
.,:~
~
.
~
,~,
Figure 4. Prototype stamped-sheet-metal
optimum burner head.
It was reasonably well established in the earlier oil
burner studies (ref. 5) that an optimized conventional
burner can operate reliably at low-excess air levels and
produce acceptably low CO, UHC, and smoke. Its NO
I emission levels, however, were a factor of 2 or more
higher than the target reduction below 0.5 g NO/kg fuel,
and whether or not the goal could be reached by opti-
mizing the furnace combustion chamber for this burner
was an uncertainty.
Therefore, the experimental investigation, which
used research combustion chambers rather than typical
furnaces, consisted of three separate, but related parts.
One part dealt exclusively with the 1 ml/s (gph) opti-
mum burner. This effort sought to optimize the match-
ing of the firebox (combustion chamber) to that fixed-
burner design. The other two parts branched out to
modified burners having, in one case, forced flue gas
recirculation (FGR) to the burner air intake and, in the
other case, forced combustion gas recirculation (CG R)
from the combustion chamber to the burner air intake.
These efforts were also focused upon optimizing the
burner/firebox combination for low emissions.
The first of these approaches-the optimizing of the
combustion chamber matched to the optimum low-
emission burner-was found to be the best approach.
The experimental investigation with the 1 ml/s combus-
tion gas recirculation (CGR) burner was frustrated by an
inability to simultaneously satisfy the NO, CO, and UHC
emission goals. Because of these problems and the inher-
ent potential problems of moderately high-temperature
I mixed gases being passed through the combustion air
fan, it was concluded that further work with this burner
concept was not warranted.
Tests of the 1 ml/s flue gas recirculation (FGR)
burner were more successful and showed good potential
for satisfying the emission goals under efficient operat-
ing conditions. However, actually achieving all of the
emission goals would probably be contingent upon the
use of a more complicated burner start sequence to
eliminate start-spike emissions of CO, smoke, and UHC,
which were experienced under conditions having accept-
ably low NO.
Thus, it was concluded that the optimized conven-
tional burner in optimized combustion chambers is a
preferred choice over both the CG R and the FG R burn-
ers. With this choice, it appears that the emission goals
can be met, low-excess air levels can be employed, and
fewer new or peripheral problems are likely to be
encountered in commercializing this burner than with
the more complicated recirculation types. Additionally,
burner simplicity will insure lower furnace costs with
this choice.
Experimental Apparatus. The 1 ml/s (gph) optimum
burner was cyclically fired in a set of three cylindrical
research combustors. I n conformity with the testing
reported in reference 2, the burner was fired for 10
minutes of 30-minute cycles. The three combustors con-
stituted a matched set, allowing considerable dimen-
sional, structural, and coolant variations. The basic ap-
proach for one of the chambers is illustrated in figure 5.
Each chamber was a 1.27-m (5.00 ft) )ong, flanged sec-
355
-------
REFRACTORY INSERT
1/2" PYROFLEX LINER " ,-'-, HEAT EXCHANGER
(OOUBLE THICKNESSI '" ", ~ SAMP, , ,WATER
-," ~'
// , '
, , ~
;" -. ,,~ -
;;9;-> e> -- ,,,?
" ' I~ ", . " / r ~"""'\ \
~ .- . , '~\)" 1.52 m - // / . - \ \ '! \
.. 'o,20m lci ' j ,I. , " ~/' - \{ (' ), ,/ '.'
o.26m, '/ ... .~~ .J
00 / ,- , t:
.... // ... .' BLANK
/./ /\\ ' FLANGE
( / - ~. SIDE-FIRED
\. ":. ,- ./, BURNER PORT
~"~" '-..-
"', ::\', ~MOVABLEHEAT
: "; ~ . .,; EXCHANGER (SPIRAL.
,f' \) \~ \ )'~ '/ // ",-. WOUND, FINNElHUBEI
. , ~,.I,r "~0.25 ....DIAMETER
)r, COMBUSTION CHAMBER
/\,
TUNNEL.FIRED BURNER PORT
END FLANGE
Figure 5. Experimental combustion chamber arrangement.
tion'of stee! pipe with a stubby, flanged, side-arm sec-
tion of the same size of pipe attached near one end. The
tunnel-fired orientation is illustrated in figure 5, in
which the oil burner fits into the annular flange depicted
at the left end of the chamber. In this orientation, the
side-arm was redundant and so was positioned at the
opposite end from the burner and simply blanked off.
The steel chamber could be either lined with a refractory
fiber insert to form an adiabatic combustion zone, or
left unlined and cooled by some means. A movable,
spiral-wound, finned-tube heat exchanger (shown in-
serted in the opposite end of the combustion chamber
from the burner) provided effective variation of the
combustion zone length. To achieve theside-fired con-
figuration, the combustion chamber was simply turned.
end for end, with the blank and burner port flanges
relocated as appropriate.
Three chamber diameters were selected (table 3),
such that addition of refractory linings to the larger two
would produce lined chambers having inside diameters
comparable with the smaller two unlined chambers. In
use, the major axes of the chamber were vertical, with
Table 3. Selected chamber diameters
Steel pipe 10,
m (in.)
Nominal combustion
chamber ID,
m (in.)
Nominal thickness
of Pyrof1ex,
m (in.)
O. 162 ' ( 6.36 )
0.222 (8.75)
0.162
0.222
0.175
0.279
0.22
0.279 (11.0)
(6.36)
(8.75)
(6.89)
(11.0)
(8.7)
0.024 (0.93)
O. 030 (1. 18)
366
-------
- burner firing vertically upward when tunnel-fired
IIIIPI"d horizontally when side.fired.
A water-cooled heat exchanger was used to ac-
complish rapid cooling of the combustion gases as they
flowed out of the primary "firebox" portion of any of
the combustion chambers. It was intended that the gas
temperature be quenched rather rapidly so that changes
in heat exchanger position (Le., firebox length) could be
readily correlated with variations in pollutant emissions.
The heat exchanger was a spiral-wound, finned-tube
assembly designed to fit inside all three combustors and
to be positioned anywhere along the length of a cham-
ber.
Experimental Results. Data were recorded during 160
tests of the apparatus previously described; in addition
to chamber diamete'r, construction, and burner orienta-
tion variations, the heat exchanger position and burner
excess air level, or stoichiometric ratio, were varied
systematically.
Typical measured emission data are illustrated for
one subset of these tests in figure 6. In this example, flue
gas pollutant concentrations varied fairly regularly with
both stoichiometric ratio and heat exchanger position,
although this was not always consistently so.
t The results of all the parametric variations tested
re synthesized into a tabular summary form amenable
o extraction of trends and design criteria. That sum-
mary is given in table 4, where the first three columns
describe the combustion chamber, and the next three
:Ii 30 --
~ 0.40
~ 0''"1 I
.;, 20 : I \ OJO m
~ I \ \
~ ~..J. \
i 10 '\ \ \
z \ \~ \
Ii! \'\
a:
-------
Table 4, Summary of results from 1 ml/s (gph) optimum
burner/combustion chamber matching experiments
REQUIREMENTS 'OR ACCEPTABLE JOR
CHAMBER DUIGN ATTRIBUTES MINIMUM) EMIIIIDNS NO. ,/k, FUEL
o L -0.& m II
CONFIGURATION COOLING METHOD IO,m CO II UHC SMOKE 1 NO. 11 I NO 1.2118 R
TUNNEL.FIAED AIR-COOLED 0.182 ROUGH-8URNING THROUGHOUT OPERATING RANGE
CCOAKIALI
ISTARTABLE OVER A 0.222 Lc:~O.6m SR ~ 1.06 L.. Qr 0." m 10.81
VERY NARROW 8R 0.279 Lc == 0.75 m SR ~1.1 OR Le = 0.4-5.0 m 10.4)
RANGEl lc>0.8m
WATEA-COOLED 0.222 La>O.6m SR > 1.15 SHORT Lc'S BEST 0.7
INSULATED 0.17S La }O.5", lc ~O.715 m SHORT lc'S BE8T 1.3
OR 8A >1.3
0.22 Lo ~O.5m Lc~O.l5m&
SA > 1,16 SHORT Lc'S BEST 0.8
SID£.FIRED AIR.COOLED 0.182 AOUa".aURNINO THROUGHOUT OPERATING RANGE
IPERPENDICULAR 0.222 La ~O.6 m Lc 2:0.6 m Lc - 0.76 m SLIQHT 0.8
PORTI SR 21.15 FAVOR
0.278 La ~0.4mll Lc:~o.~m.
SR ~ 1.18 8R .2: 1,08 SHORT lc'l BEST 0.8
IPORCED DRAPTI 0.280 Lo ~O.IS m La ~ 0.5 m, SA > 1.01 SHORT Lc'l BE8T 0.8
WATER.CooLID 0,222 La ~ 0.8 m La ~o.lm.
SR ~ 1.2 SHOAT Lo'S BElT 0.7
0.278 La l:O.5 m Lo ~ D.D II SR ~ 1.1 SHORT Le'S BEBT 0.4
INSULA TlD 0.178 La ~O.3m L. eo.smll
8A ~ 1,2& SHORT Lc'S BEST 2.0
0.22 La ~ 0.6 m La ~o.6m.
SR ~ 1.15 DATA MIXED 1.9
stronger and more complicated eddies of a side-fired
chamber.
Combustion Chamber Size. Chambers smaller than
about 0.20 m (0 in.) ID must be refractory-lined to
avoid operating with unacceptable combustion rough-
ness. Operation of all designs was acceptable in this
regard when thuir inside chamber diameter was 0.22 m
(8.3/4 in.) or Jar~ar.
, A tendency existed for larger-diameter chambers,
with or without insulation, to require longer chamber
lengths to achil!\le comparable levels of carbonaceous
pollutants and, concurrently, to produce substantially
lower levels of rlO. Both phenomena were undoubtedly
linked to the in!/f'stion of recirculating combustion gases
into the flame zone. For a given burner, larger and
stronger recircula'1ion eddies can be established in bigger
chambers, reducing the combustion intensity somewhat
and lowering tn3 rates of burnout of carbonaceous
species. Also, largsr diameter chambers have greater wall
areas for conveGtive and radiant transmission of heat
from the flame2one, which reduces peak flame temper- .
atures somewhat. Increased avc..;c~. gas residence times
should produce I)pposing trends for both NO production
and carbon burnout, which "ould help account for some
anomalies in thu data. Also, the steady-state recircula-
tion and radiation effects may be partially masked by
starting transient effects, particularly with respect to
smoke and UHC data.
Short combustion chambers generally were more
favorable for low NO but, if they are too short, the fuel
may not be completely burned before combustion
reactions are quenched in the heat exchanger. A com-
bustor length of 0.5 m (20 in.) is a suitable compromise
for most of the chamber designs tested. Slightly shorter
chambers might be appropriate if refractory linings were
to be used.
Chsmber Cooling Medium. Refractory-lined com-
bustion chambers generally exhibited less combustion
roughness and less sensitivity to starting conditions but
also produced higher NO concentrations than did
cooled-wall combustors. The air-cooled, side.fired
configuration had better starting characteristics (but
higher NO) than did the air-cooled, tunnel-fired cham-
ber. Water-cooling appeared to be preferable to air cool-
ing, particularly because of lower side-fired NO levels,
but also because of smoother starting and more con.
sistent CO and NO emission results. The water-cooled
chambers in shorter lengths, however, were prone to
produce excessive CO.
It was inferred that the key effect of both air- and
water-cooling is the reduction of flamezone temper-
atures that results from removing heat from the com-
368
-------
~ustor walls. The 0.279 m (11 in.) diameter of the
water-cooled combustor, which met all emission goals,
had approximately 20 percent of the heat of combustion
of the fuel extracted from the firebox. The single
"forced draft" air-cooled chamber was designed to dupli-
cate that heat removal rate, whereas considerably less-
effective natural convection prevailed in the other air-
cooled configurations. That forced-draft, air-cooled
chamber, however, had an intermediate 0.25 m (10 in.)
10. Its flue gas NO concentrations were approximately
midway between those from the 0.222 'm (8-in.) and
0.279 m (11-in.) ID water-cooled combustors, indicating
that forced-draft, air-cooling can be just as effective as
water-cooling in controlling NO emission levels.
Optimized Combustors for the 1 mIls (gph) Optimum
Burner. To reduce NO emission levels below 0.5 g/l
-------
Co)
en
o
...-.
All
11M(
_ao
All m.TD
ll@"~1
iwi -- [
.- h,...~. ,:!
1'1 I I ;1
I
L:
I--- 0.,",
_AIl
... .1- -
SECTiOll A - A
JIlSUUTtCII
CUMt. 1lUI8
OUIUI
_IISSY.
1.-
.... IISSY.
. ... All ILQID
.... AIl
Figure 7. Conceptual prototype: low-emission, warm-air furnace.
-------
-..-'-
\'
- .
&~ROf't\.ETR..\C.
CONTQ.ol..
*
E)(."TEI1..,"L FLUE.
/'.'!t .
. t-- ! --4<:1
r . /L-,! I
J/.. j
~\L~R.. ~\\
:.- 1
...=:;---.
E..x.'?NoJ Sol DN
T""',,-
co,,",w.
I
. A '
1._) "
. i'i
""""S T
,.--:
j /~"1'" .~.,
w
en
--
cxa..c........,."tIOtoJ
""..
R~",,--"'"
. ~I '
. 0.35 I~_- n
0.79 m .
SECTION A-A
SECTION B-B
Figure 8. Conceptual prototype: low-emission hydronic boiler.
-------
sheets. Betwenn the combustion chamber and the first
heat exchanger pass, combustion gases are reversed by
passage throu!Jh an insu lated lBO-degree return mani-
fold, whose volume is considered to be part of the com-
bustion chamber. Similarly, there are 180-degree return
manifolds between successive heat exchanger passes.
The con()~P1ual hydronic boiler design also has the
usual plumbing and control features needed for installa-
tion and operation in a residence, including a tank less
coil for heati.,g domestic hot water. Needed, but not
shown in figufe 8, is an exterior cabinet that should be
well insulated to reduce external heat losses.
Future Effort. Following comparative assessments of the
two preliminary prototype designs, including their
potential overall emissions and fuel economy impact, the
warm-air syst!.m has been selected as the more promis-
ing; therefore, a prototype warm-air unit will be built for
further study in the continuing work. The data and
experiences gained will be applied to the final design of a
cost-competitive, commercially producible, warm-air fur-
nace that embodies the derived low-emission, improved-
efficiency technology.
REFERENCES
1.
R. E. Hal-, J. H. Wasser, and E. E. Berkau, "NAPCA
Combusti.)n Research Programs to Control Pollu-
tant Emissions From Domestic and Commercial
Heating Systems," New and Improved Oil Burner
Equipment Workshop, NOFI Technical Publication
108 ED, National Oil Fuel Institute, Inc., New
York, New York, September 1970, pp. 83-93.
2. R. E. Hall, J. H. Wasser, and E. E. Berkau, "A Study
of Air Pollutant Emissions From Residential Heat-
ing Systems," EPA-650/2-74-003, Environmental
Protection Agency, Research Triangle Park, N.C.,
January 1974.
3. G. B. Martin, and E. E. Berkau, "Evaluation of
Various Combustion Modification Techniques for
Control of Thermal and Fuel-Related Nitrogen
Oxide Emissions," presented at the Fourteenth
Symposium (International) on Combustion, Penn-
sylvania State University, August 1972.
4. R. E. Barrett, S. E. Miller, and D, W. Locklin, "Field
Investigation of Emissions From Combustion Equip-
ment for Space Heating:' EPA-R2-73-084a (API
Publication 4180), Environmental Protection
Agency, Research Triangle Park, N.C., June 1973.
5. R. A. Dickerson, and A. S. Okuda. "Design of an
Optimum Distillate 0 il Burner for Control of Pollu-
tant Emissions:' EPA-650/2-74-047, Environmental
Protection Agency, Research Triangle Park, N.C.,
June 1974.
6. G. Peoples, "Sealed Oil Furnace Combustion Sys-
tem Reduces Fuel Consumption," Addendum to the
Proceedings, Conference on Improving Efficiency in
HVAC Equipment and Components for Residential
and Small Commercial Buildings, Purdue University,
Lafayette, Ind., October 1974.
7. L. P. Combs and A. S. Okuda, "Residential Oil Fur-
nace System Optimization - Phase I," EPA~
600/2-76-038, Environmental Protection Agency,
Research Triangle Park, N.C., February 1976.
362
-------
EVALUATION OF THE POTENTIAL IMPACT OF ALCOHOL FUELS
ON POLLUTANT EMISSIONS AND ENERGY REQUIREMENTS
OF AREA SOURCES
G. Blair Martin*
INTRODUCTION
Stationary combustion sources used in residential
and commercial applicatiom can be classified as area
sources. Although the pollutant emissions from an indi-
vidual unit are relatively small, the discharge into the
atmosphere is near ground level in populated areas and
the number of sources is large. Although residential and
commercial sources contribute only about 10 percent of
the mass emissions of NOx' the environmental impact
may be more significant. In addition, the.se sources
which generally require clean fuels (i.e., natural gas and
distillate oil) account for 20 to 30 percent of the station-
ary source energy usage. In 1972, transportation used
about 16 X 1015 Btu (1.68 X 1019 J) while residential.
and commercial heating used about 14 X 1015 Btu (1.48
X 1019 J). For the latter category, natural gas provided
about 60 percent of the energy, and fuel oil, predomi-
nantly distillate, provided the balance. Based on these
factors, it appears that both environmental and energy
flspects of these area sources require further attention.
Since these area sources require premium quality
fuels, the current energy shortage requires an examina-
tion of the options that may be available in the future.
These options include completely new approaches as
well as retrofit or modification of existing systems. The
first category includes solar, fluidized bed, and catalytic
combustion concepts; however, most of these are in very
early stages of development for area source applications
and must be considered as potential long-term solutions.
The second category includes increased use of electrical
heating, improved efficiency and emissions performance
on conventional clean fuels, and use of alternate clean
fuels. These approaches may be amenable to shorter
term applications and/or retrofit to existing types of
equipment.
The choice of new technologies to be developed
depends on a number of factors. From an energy stand-
point, the overall efficiency of the process needs to be
considered from resource extraction to end use. This is
particularly important since the overall efficiency is the
result of multiplication of the individual efficiencies of
the steps in the process. Cost considerations must in-
clude both energy cost delivered to the user and capital
-G. Blair Martin is with the Combustion Research Branch,
IERL, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina.
~
costs of the end use device. Evaluation of environmental
factors must also include all steps in the scheme from
resource extraction, through conversion processes and
transportation, to end use. Finally, policy decisions,
such as fuel allocation to various users, will playa major
role in the future course; however, the decisions should
take into account the best available information on the
energy, environmental, and economic aspects mentioned
above. It is clear that the required analysis of the avail-
able options is very complex and beyond the scope of
this paper. However, it is possible to identify the need
for additional information in technical areas related to
energy utilization in the residential and commercial sec-
tors. One of the near-term areas that is of particular
interest is the use of alternate fuels. From a practical
standpoint, alternate fuels can be used in conventional
equipment either directly or with a minimum of modifi-
cation. This allows consideration not only of the 5 to 8
percent of new installations each year, but also of the
vast number of existing units (e .g., about 55 X 106 resi-
dential units!".
For application to area sources, only high-Btu fuels,
which can be economically transported and/or stored,
should be given strong consideration. These sources are
fired predominantly with clean premium fuels, such as
natural gas and distillate oil. The alternate fuels must be
clean gaseous and liquid fuels that can be interfaced with
current combustion and fuel distribution systems with
fairly minor modifications. Synthetic Natural Gas (SNG)
should be virtually indistinguishable from the pipeline
natural gas and adaptable without change to existing
gas-fired equipment. For the distillate oil-fired units, the
alternative requiring the least change would be a clean
liquid fuel. One candidate is distillate oil derived from
coal or shale derived crudes; however, evidence indicates
that these oils may contain significant amounts of fuel
nitrogen which can produce fuel NO. The technology for
controlling NOx emissions for these high-nitrogen fuels
is just developing and may not be readily adaptable to
small residential sources. This is one area needing further
study. A second class of candidate liquid fuels is the
alcohol fuels, predominantly menthanol, which are sul-
fur and nitrogen free. For both types of liquid fuels
there are two major questions:
1. What are the combustion and emission characteris-
tics of the fuels, particularly for residential and
commercial equipment?
363
-------
2. What is the probability that the fuel will be available
in sufficil!nt quantity to have a significant impact on
area source energy needs?
The emphasis in this paper is on the use of alcohol
fuels in residfntial and commercial combustion systems.
A brief summary of information on methanol synthesis
is followed by a review of the data on alcohol fuels
combustion. Based on this information, emission and
efficiency aspects of alcohol fuels combustion in area
sources are dh,cussed.
BACKGROUND
The available information on alcohol fuels covers
both sources I)f supply and combustion properties. Each
of these subjet:ts is summarized briefly below.
Sourr:e of SupolV
The use of alcohol as fuel dates back into the 1800's
when methan)1 was derived from the destructive distil-
lation of woe d. The use of alcohols as motor fuels re-
ceived much attention in Europe during the 1940's.
Much of the historical information is summarized by
Pleeth (ref. 1), who also reports that 106 gallons of
ethanol can te derived frpm 2,000 acres of artichokes
(5.62 X 10-4 In3 per m2).
In current considerations of approaches for obtain-
ing large qualtities of alcohol fuels, three possibilities
are commonl,; mentioned. These are: (1) conversion of
remote natun I gas as a method of transporting energy;
(2) synthesis 1rom coal as an alternative to SNG; and (3)
generation by bioconversion processes. Most available in-
formation deals with the first two, which are discussed
briefly below.
Na tural Gas Conversion. Most chemical grade
methanol is IT ade by reforming natural gas followed by
methanol synthesis. While this is acceptable for the
limited amounts of chemical grade methanol produced,
it is not a vi~ble source of fuel methanol. In terms of
energy alone, it makes very little sense to take two
energy units cf a good fuel (CH4) and convert it to 1 to
1.2 units of another fuel.
Therefore. the only logical source of methanol fuel
from natural nas is in the situation where the methanol
provides a co;t and/or energy effe~tiveness advantage.
Duhl (ref. 2) has proposed that remote natural gas by
converted intc "Methyl Fuel" 'IS an alternative to liquid
natural gas (LI\lG), based on 10(lJ distance transportation
considerations. "Methyl Fuel" is i:I mixture of methanol
with about 10 percent higher alcohols and is produced
by a process that Duhl states is simpler than a chemical
grade methanl)l process. The price of "Methyl Fuel" is
estimated to range from $0.65 to $0.85 per 106 Btu
, ($0.615 to $0.84 per 109 J), which is one of the lowest
costs for alcohol fuels quoted in the literature. In con-
trast, Hoogendoorn (ref. 3) has also evaluated the con-
version of remote natural gas [$0.40 per 106 Btu ($0.38
per 109 J) ] to methanol and concluded that the produc-
tion cost is $2.55 per 106 Btu ($2.42 per 109 J). He
states that the thermal efficiency of methane to metha-
nol is 58 percent.
Coal Conversion. An alternate path of particular
interest in the United States for producing methanol fuel
is via a synthesis gas derived from coal. McGhee (ref. 4)
has proposed coproduction of SNG and methanol at 59
and 41 percent, respectively. In this case, the methanol
is required to sell as a chemical at $6 per 106 Btu ($5.70
per 109 J); however, this is the result of a decision that
the natural. gas must be subsidized to sell at $1.48 per
106 Btu ($1.40 per 109 J). Eastland (ref. 5) has studied
production of methanol alone from coal and shows a
range of cost from $2.50 to $4.00 per 106 Btu ($2.37 to
$3.79 per 109 J), dependent on coal cost and method of
financing assumed. Jaffe (ref. 6) has reviewed the full
range of methanol production technology and concluded
that a state-of-the-art plant can produce methanol at
$1.58 to $3.06 per 106 Btu ($1.50 to $2.90 per 109 J)
with an overall thermal efficiency of 45 percent. For
comparison, he listed a refinery price of premium grade
gasoline at $2.00 to $3.00 per 106 Btu ($1.90 to $2.85
per 109 J). Finally, Pasternak (ref. 7) estimates the cost
of methanol produced by in situ gasification of a low-
grade coal at $1.04 per 106 Btu ($0.98 per 109 J) and
contrasts this with $1.69 per 106 Btu ($1.60 per 109 J)
for methanol produced from a Lurgi synthesis gas. Al-
though there is a wide range of assumptions incorpo-
rated into these estimates, it would appear that metha-
nol can be produced at a cost that is reasonably close to
conventional fuels and SNG. In addition, the overall con-
version efficiency of coal to methanol in the range of 45
to 50 percent exceeds the current conversion efficiency
of coal to electricity. Viewed in this way, methanol may
have a future as a fuel for area sources if the production
facilities are established.
EnvironmentBl Aspects. The major environmental
concern in. methanol production is the coal conversion
process. Comprehensive analyses for a number of fuel
conversion processes have been carried out by Exxon
R&E under EPA contract and reported in a series with
numbers EPA-650/2-74-009 a to g. Another area of con-
cern is the toxic nature of the methanol itself, and fur-
ther study is needed to establish the actual health and
environmental problems related to handling this fuel.
364
-------
Combustion Properties
Although the data on combustion of alcohols in
stationary sources are limited, several recent papers have
provided some insight.
It can be seen that methanol produces 25 to 35 percent
of the level of NO from the two conventional fuels.
Figure 2, which presents the NOx emissions as a func-
tion of the higher alcohol content, shows that the 10
percent content of higher alcohols in "Methyl Fuel" will
result in a small increase in emissions. Finally, figure 3
shows the effect of flue gas recirculation (FGR) on the
NOx emissions. While external FGR is probably not a
Hot Wall Experimental System. Data comparing
alcohol fuels with two conventional clean fuels in a hot
wall experimental furnace are shown in figure 1 (ref. 8).
300
c
w
a:
::;) 200
CI)
c:t
w
~
CI)
c:t
>-
a:
c
~
Q..
Q..
W
C 0
x
0
t.J
a: 100
I-
Z
~
. .
.
ODISTlllATE Oil 6 ISOPROPANOL. METHANOL
OPROPANE 0 50% ISOPROPANOl- 50% METHANOL
o
o
2
4
6
8
SWIRL BLOCK POSITION
Figure 1. Comparison of baseline nitric oxide emissions for various fuels
as a function of swirl parameter.
365
-------
200
o METHANOL
o 50% METHANOL. 50% ISOPROPANOL
A ISOPROPANOL
Q
UJ
a:
;:)
U)
<
UJ
:!:
U)
<
>
a:
Q
-
150
:!:
a..
a..
100
.
w
C
><
C
u
a:
~
z
50
o
o
75
100
25
50
HIGHER ALCOHOL CONTENT, MASS %
Figure 2. Effect of higher alcohol content
on nitric oxide emissions.
feasible technique for small combustion sources, these
results indicatl! that a burner designed to induce internal
recirculation cf relatively cooleG combustion gases may
have the potential for a further reduction of emissions.
During these tests, the pmissions of other pollutant
species (e.g., GO, UHC) were small and comparable for
all fuel!:.
Cold Wall Experimental System. Results have also
been reported for a variety of fuels fired in a cold wall
experimental system simulating a package boiler (ref. 8).
The system was fitted with a modified commercial dual-
fuel burner and no problems were encountered using the
ultraviolet flame safety system. Figure 4 shows a com-
parison of the NO emissions for four fuels as a function
366
-------
300
I
o DISTillATE Oil
o PROPANE
6 ISOPROPANOL
o METHANOL
c
UJ
g; 200
en
,C:(
UJ
~
en
c:(
>-
a:
c
~
Q..
Q..
UJ
C
><
o
u
cc 1 00
I-
2:
o
o
0.05
0.10 0.15
FRACTION RECIRCULATED. f
0.20
0.25
Figure 3. Effect of flue gas recirculation on nitric oxide
emissions for various fuels.
of combustion air distribution between the burner
primary and secondary channels. It is particularly inter-
esting that the methanol results are not radically differ-
ent than those from the hot wall system, while the distil-
late oil emissions are significantly lower. The high level
for the residual oil is attributable to fuel NO from the
0.36 percent nitrogen oil.
In addition to these results, Duhl (ref. 9) has also
reported results for a cold wall system where the
methanol resulted in 25- to 50-ppm NO compared to
100- to 200-ppm NO for methane. These differences
from the results above may be attributable to the
presence of a hot refractory throat tile in the system
reported by Duhl.
367
-------
400
300
o
w
a:
:>
en
a:
o 200
-
~
Q.
0-
W
C
X
o
u
a:
....
z
100
o
20
. RESIDUAL Oil
. . DISTillATE Oil
. NA TURAl GAS
. METHAl\lOl
.
./
.~
/
.
.;/
./
/
.
./. .
/
.
L
30
40
PRIMARY AIR, % OF TOTAL
50
60
.
Figure 4. Comparison of baseline nitric oxide emissions for various
fuels as a function of burner primary air.
368
.
70
80
-------
. Practical Systems. Data on practical systems are
~Imost nonexistent; however, figure 5 shows data on a
50-MW utility boiler also reported by Duhl (ref. 9).
These results show relatively small differences between
180
160
:2 140
CL.
CL.
vi
w
c
X
c
z
w 120
c,:,
o
0:
~
Z
100
emissions for methanol and natural gas.
In addition, a field firing of methanol in package
watertube and firetube boilers is in preparation on EPA
contract 68-02-1498. These tests will provide direct evi. ".
80
o
60
20
o METHANOL @ 15%
EXCESS AIR
o NATURAL GAS @ 7-15%
EXCESS AIR
SOURCE: REF~RENCE 9
30
40
UNIT LOAD, MEGAWATTS
50
60
Figure 5. Nitrogen oxides emissions from
methanol and natural gas as a function
of utility boiler load.
369
-------
dence on the I:hanges required to fire methanol in exist-
ing commercial size package boilers and will also ex-
amine the imp,Ict on system efficiency.
Summary. The major aspects of the combustion
data can be summarized as follows:
1. For any given condition, alcohol fuels produce
lower emissions of NOx than distillate oil, methane,
or propare.
2. The NO~rnissions of alcohol fuels increase as the
percentage of higher alcohols increases.
3. The low I~O emissions for alcohol fuels appear to be
a functiol of the presence and level of oxygen in
the fuel r10lecule, which can be viewed as a diluent
carried ir. the fuel. The operative mechanism ap-
pears to :>I~ related to thermal effects of the fuel,
latent heat of vaporization, and/or decreased flame
temperature; however, chemical effects cannot be
totaily ru~ed out.
4. The magr,itude of the difference between methanol
and any conventional fuel is a characteristic of
system design; however, the difference appears to be
smaller for cold wall systems.
5. The emis!:ions of CO, UHC, and particulate for alco-
hol fuels were generally the same as, or less than
those for the conventional clean fuels tested (pro-
pane and distillate oil). Aldehydes do not appear to
be a significant problem based on limited data.
6. From a t~chnical standpoint, methanol appears to
be a satisfactory fuel for stationary combustion
systems.
DISCUSSION
Although the information for complete assessment
of the use 01 methanol in residential and commercial
systems is not available, it is possible to identify poten-
tial advantages and/or problems for the two types of
systems. Thesn are discussed briefly below.
CommercialB'Jilers
The type of equipment used in the commercial sec-
tor is almost exclusively package boilers generating hot
water or low pressure steam. I n recent years, the trend
has been towcrd burners capable of firing two fuels (i.e.,
gas and oil). The flame safeguard system is of the ultra.
violet (UV) t\,pe and, therefore, is compatible with the
requirements ;mposed by the nearly non luminous meth.
anal flames. The second cons;..ieration for methanol
firing is the ahility of the fuel noz:de to supply the rated
energy input of the boil"~. A total of 2.1 volumes of
methanol must be fired to achieve an energy input equiv-
alent to 1 void me of fuel oil; therefore, a larger capacity
fuel nozzle will be required in most cases. In addition,
the viscosity difference between methanol and fuel oil
may also influence the required nozzle size. Current evi-
dence indicates that conventional air or steam atomizing
provides acceptable methanol atomization.
The combustion performance. of alcohol fuels is
equal or superior to fuel oils. The levels of CO and UHC
are low, smoke formation and aldehydes are not signifi-
cant, and NOx emissions and the excess air required for
compelete combustion are significantly lower than for
oil. .
The main uncertainty is the impact of methanol
firing on the overall boiler efficiency. For the same
excess air level and stack temperature, the efficiency
based on gross heating value is 6 percent less ~or metha-
nol than for distillate oil. This is related to the partially
oxidized nature of the methanol and would require an
additional 6 percent fuel energy input to achieve the
same system output. There is also some evidence (ref. 8)
that the heat removal in the combustion chamber of a
scotch marine firetube boiler may be lower for methanol
than for distilla~; however, this cannot be directly re-
lated to overall system heat removal as the convective
pass heat transfer was not evaluated. Two factors may
provide a compensating effect: (1) the methanol can be
fired at a lower excess air level and, therefore, stack
losses may be reduced by several percent; and (2) the
clean combustion characteristics of methanol may re-
duce tube deposits, thereby maintaining efficient heat
transfer over the long term and reducing cleaning fre-
quency. System efficiencies for methanol, oil, and natur-
al gas are being compared under EPA contract
68-02-1498 for two package boilers; results should clari-
fy the question.
Residential Heating
The two primary fuels for residential heating are
distillate oil and natural gas with each fuel accounting
for roughly 40 and 60 percent, respectively, of the ener-
gy used in existing furnaces in 1972. In recent years,
natural gas furnaces have contributed the majority of
sales. It should be noted that these furnaces do not have
the dual-fuel capability that is found in many commer-
cial systems. There are also completely different furnace
configurations used for natural gas than for oil. All of
these factors combine to indicate that for residential fur-
naces, consideration must be given not only to new units
designed specifically for methanol, but also to retrofit of
existing units designed for distillate oil and natural gas.
To fire methanol in a furnace designed for natural gas
would require vaporization of the fuel prior to fuel
introduction to the burner or other major hardware
changes. For retrofit of a natural gas system, methanol is
370
-------
~robahly nul a tec:hnically or ecunomically viable alLer-
'ative; therefore, this paper will discuss only oil. For
liquid fuels, it should be possible to develop a furnace
capable of firing either methanol or distillate fuel oil
with only minor changes (e.g., nozzle).
Since there was no information available on metha-
nol firing in a residential furnace, an experiment was
performed in the EPA laboratory. A conventional resi-
dential warm-air furnace, designed to fire distillate fuel
oil at a nominal 1 gal per hr (1 ml per see), INas equipped
with a 2.25 gal per hr (2.25 ml per see) nozzle to main-
tain an equal energy input firing methanol. When the
furnace was started up, ignition was achieved with meth-
anol; however, the flame safety system shut the burner
down. Therefore, the burner was removed from the fur-
nace to diagnose the cause. Observation of the methanol
flame revealed that it was a translucent orange as
opposed to the highly luminous flame produced with
distillate oil. Apparently, the cadmium sulfide flame
detector, which operates on infrared radiation, does not
generate sufficient signal with the methanol flame. In
addition, the observations suggest that for methanol
firing, a redesign of the burner head may be required to
achieve the compact flame required by the residential
firebox. Finally, there is some concern that the metha-
nol does not possess the same lubricating properties as
~fuel oil and a different fuel pump may be required. The
'net result of these changes would be a replacement
burner for methanol firing, probably at a cost of over
$100. Based on the other combustion data from metha-
nol firing, it may be projected that emissions of CO and
unburned hydrocarbons will be comparable to those
from fossil fuels, while NOx emissions may be lower by
as much as 50 percent. Particulate and SOx emissions
should be nonexisting.
Another question is the impact of methanol firing
on system efficiency. Since the adiabatic flame tempera-
ture for methanoi is about 2200 F (1050 C) lower than
for fossi I fuels, the driving force for heat transfer is
lower; however, this may be offset by a longer residence
time of the combustion gases in the heat exchanger. The
other factor to be considered is the differences in heat
losses in the flue gas. These losses are related to the
temperature of the flue gas, the mass of flue gas, and to
the latent heat in the water vapor. For a given tempera-
ture and excess air, methanol has higher sensible and
latent heat losses than does distillate oil, due to the
greater mass of water generated for an equivalent energy
input. This can be observed from some calculations with
the WAFURN program (refs. 10,11), which can estimate
furnace performance as shown in table 1. The difference
is largely attributable to increased latent heat loss in the
flue gas from methanol. A potential compensating effect
should be noted. If the oil is burned at 85 percent,
excess air to avoid excess smoke formation and the
methanol can be fired at 15 percent excess without
smoke, the gross efficiencies are essentially equal. This
type of change is certainly a strong possibility for retro-
Table 1. Comparison of calculated furnace
efficiencies for methanol and distillate oil
in a fixed configuration residential furnace.
Excess air
%
Gross efficiencya
Distillate oil Methanol
% %
15
50
85
79.9
76.1
73.3
73.9
71.0
67.8
aGross Efficiency =
Heat Delivered To Home in Warm Air
Higher Heating Value of Fuel
x 100
371 .
-------
fit to existing oil-fired units; however, the potential is
somewhat less for new oil-fired units of modern design
that normally operate at 25 to 50 percent excess air.
CONCLUSIONS
The following conclusions can be drawn:
1. Use of alcohol fuels has the potential for reduced
emissions ')f NOx and other pollutants.
2. The efficiency of a system utilizing alcohol fuels
may be lower than for fossil fuels in the same sys-
tem; howEver, the most significant energy loss in the
overall Scll()me appears to be in synthesis of the
alcohol fu ~I from coal.
3. Alcohol fuels can be fired in commercial boilers
with a minimum amount of burner modification.
4. Additiona1 combustion technology development is
required for the utilization of alcohol fuels in resi-
dential fur naces.
5. Fuel supply appears to be the major obstacle in the
use of alc.}hol fuels in area sources. Fuel costs, over-
all energy efficiency, and environmental aspects
should all be factors considered in a decision on
alcohol fu ~Is versus other alternatives.
REFERENCES
1. S. J. W. Fleeth, "Alcohol a Fuel for Internal Com-
bustion Engines," Chapman and Hall Ltd, London.
1949, p. 32.
2. R. W. Duhl and T. O. Wentworth, "Methyl Fuel
from Remote Gas Sources," presented at 11 th
Annual M0eting of the SOIJthern California Section
AIChE, LeIs Angeles, Calif., April 1974.
3. J. C. Hoogendoorn, "New Applications of the
Fischer-Tropsch Process," presented at the IGT 2nd
Symposium on Clean Fuels From Coal, Chicago. 111..
June 1975.
4. R. M. McGhee, "Coproduction of Methanol and
SNG from Coal." presented at the Engineering
F ou ndation Conference, New Hampshire, July
1974.
5. D. H. Eastland. "Methanol from Coal with Winkler
Synthesis Gas," presented at the IGT 2nd Symposi-
um on Clean Fuels From Coal, Chicago, 111., June
1975.
6. H. Jaffe et al.. "The Production of Methanol From
Coal for Fuel Use," Report UCRL-76232. Lawrence
Livermore Labs, Calif., December 1974.
7. A. Pasternak. "Methyl Alcohol Production by In
Situ Coal Gasification." Report UCRL.51600.
Lawrence Livermore Labs, Calif., August 1974.
8. G. B. Martin and M. P. Heap. "Evaluation of NOx
Emission Characteristics of Alcohol Fuels for Use in
Stationary Combustion Systems." presented at the
Symposium on I mpact of Methanol on Urban Air
Pollution, 80th National AIChE Meeting, Boston.
Mass.. September 7-10. 1975.
9. R. W. Duhl and E. Allegrini, "Methyl Fuel - A
Boiler Alternate," presented at the Symposium on
I mpact of Methanol on Urban Air Pollution, 80th
National AIChE Meeting, Boston, Mass., September
7-10.1975.
10. L. P. Combs and A. S. Okuda, "Residential Oil Fur-
nace System Optimization - Phase I," Rockwell In-
ternational, EPA-600/2-76-038, NTIS PB
250 - 878/ AS, February 1976.
11. L. P. Combs, private communication.
372
-------
ASSESSMENT OF THE APPLICABILITY OF CATALYTIC OXIDATION
OF HYDROCARBON AND OTHER FUELS FOR CONTROL OF
NOX AND OTHER POLLUTANTS FROM AREA SOURCES*
J. P. Kesselring, Ph.D.,t G. B. Martin:!:
R. A. Brown,** and C. B. Moyer, Ph.D.,~
Abstract
A review of the state-of-the-art of catalytic com-
bustion concepts has been ca"ied out, and an assessment
of the applicability of catalytic combustion to gas-and
oil-fired home heaters and commercial and industrial
boilers has been made. Newly developed high-tempera-
ture support materials will greatly enhance the field of
high-tempemture catalytic combustion, but current cat-
alyst systems are limited by the catalyst coating to much
lovver temperatures than the supports. In order to keep
combustor temperatures below those that would cause
catalyst degradation, as vvell as to achieve high system
efficiency and also to prevent NOx formation, combus-
tion system concepts such as two-stage combustion, flue
gas recirculation, and bed heat removal appear necessary.
The application of these concepts to home furnaces
appears feasible, but the application to larger-size units
may be more attractive because of their larger initial
cost, generally more sophisticated controls, better super-
vision of equipment, and heat transfer characteristics:
INTRODUCTION
More than 150 years ago, Sir Humphrey Davy dis-
covered that platinum wires could promote combustion
reactions in flammable mixtures, and that the resulting
reactions appeared to take place on the surface of the
'This project was funded at least in part with Federal funds
from the Environmental Protection Agency under contract
number 68-02-1318. The content of this publication does not
necessarily reflect the views or policies of the U.S. Environ-
mental Protection Agency, nor does mention of trade names,
commercial products, or organizations imply endorsement by
the U.S. Government.
tLeader, Thermodynamics and Kinetics Section,
Aerotherm Division, Acurex Corporation, Mountain View,
California.
tCombustion Research Section, I ndustrial Environmental
Research Laboratory, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina.
"Staff Engineer, Aerotherm Division, Acurex Corporation,
Mountain View, California.
~Manager, Energy and Environmentel Sciences, Aerotherm
Division, Acurox Corporation, Mountein View, California.
wires, "without flame" and with the high radiative
fluxes associated with the solid surface emittance rather
than the low emittance of the combustion products (ref.
1). Subsequent study and investigation followed a
number of paths reviewed by Spalding in reference 2. Of
this work, a substantial fraction which might be termed
"fundamental studies of heterogeneous catalysis" has
attempted to identify and quantify specific heterogen-
eous reactions and define the mechanisms by which the
solid surface promotes or accelerates oxidation re-
actions. A second fraction, termed "computational", has
dealt largely with mathematical theories and analyses
attempting to describe the diffusion of reactants to the
promoting surface, the diffusion of products away from
it, and the combined heat and mass transfer. A third
fraction can be termed "reduction to practice", which
includes attempts to construct and employ practical de-
vices exploiting heterogeneous combustion. As is fre-
quently the case in combustion, work in this area has
proceeded with very little reference to achievements in
fundamentals and computation. Experiments through-
out the 1800's demonstrated that various materials pro-
moted heterogeneous combustion reactions to various
. degrees, and that with increasing surface temperatures
the number of materials with the ability to promote
reactions increased. The earliest successful practical de-
vices exploited heterogeneous combustion to provide
more efficient lamps. The best example is the mantle
lantern or Auer light, still sold in great numbers as a
"camping lantern", in which combustion takes place on
the ash residue of a fiber bag mantle surrounding the
fuel/air orifice. This mantle ash contains oxides, in-
cluding cerium oxide and thorium oxide, which, in
addition to constraining radiation emission to desirable
lines in the visible spectrum, contribute to the catalytic
activ ity of the mantle (ref. 3).
Concepts close to the original experiments of Davy
have been used as passive ignition devices and as com-
bustible detectors, particularly in the form of hydrogen
leak detectors. Related applications along this line in-
clude catalytic'fume abatement devices and automotive
catalytic converters, both depending on precious metals
to oxidize low concentrations of combustibles in gas
streams.
After extensive practical experimentation, Bone
publicized various high - temperature applications of
373
-------
surface comb.Jstion, using for the most part conven-
tional ceramic materials (refs. 4,5). He built and tested
surface combustion furnaces, boifers, and permeable
panels. This line of investigation has not, however,
proven very fruitful due to the complexity and expense
of the requind equipment and the limited industrial
need for high radiative fluxes, although some commer-
cial equipmen': of this type is available (ref. 6).
The range of commercial devices has recently led to
some specialb~tion of terminology in a field where
formerly several terms were used interchangeably. The
term "radiant burner" refers to burner types employing
a refractory surface situated within or near the flame so
as to provide for radiant heat transfer from the flame
area. Heteroge neous combustion occurs in these devices
to some degme, but generally does not dominate the
process and homogeneous combustion constitutes most
of the reac::ion mechanism. "Surface combustors"
employ refractory in a similar manner but use special
provisions of air/fuel distribution to maximize the
amount of heterogeneous combustion occurring.
"Catalytic combustors" employ special additives
(usually precious metals) to enhance the heterogeneous
activity of the surface.
There are several reasons for the recent upsurge of
interest in ca talytic combustion. Use of the diffusion
flame for corrbustion places a limit on the lower level of
control that Celn be achieved economically for clean fuels
(approximatel" 25 to 30 ppm of NOx for small sources),
and catalytic combustion appears to be able to reach
lower levels (If NOx control (perhaps < 10 ppm). In
addition, in order to minimize the combustion system
heat loss in t~.e exhaust, it is necessary to minimize the
exhaust mass, which implies running at stoichiometric
conditions. To minimize NOx production and also
maximize oVHall system efficiency, systems such as
two-stage corr bustion, flue gas recirculation, or cooling
of the catalyst bed must be examined.
The use ( f catalytic combustors in gas turbines be-
cause of their high excess air requirement is also of in-
terest. The Amospace Corporation, under an EPA con-
tract, reviewl!d the state-of-the-art of surface and
catalytic combustion concepts in 1973, and assessed the
applicability of these concepts to large utility boilers and
stationary gas turbines (ref. 17). It was concluded that
catalytic comhustors might be applicable to both exist-
ing and new s':ationary gas turbines, but it was unlikely
that catalytic combustors could be economically in-
stalled in utility boilers.
Finally, there is a continuir] interest in producing a
compact "high-performance", lov.-emission combustor,
and catalytic combustion IS a candidate for this system.
This stud'l, then was undertaken to assess the pre-
sent state-of-the-art in catalysis and catalytic com-
bustion, to assess the feasibility of catalytic combustion
for the control of NOx emissions from stationary
sources not covered by the Aerospace study, and to
define problems that might be encountered in the re-
design and/or retrofit of catalytic combustors for sta-
tionary sources.
AVAILABLE CATALYST MATERIALS
The materials needed for a catalytic combustion
system are those associated with the support, the
washcoat, and the catalyst itself. The importance of each
of these elements in the system is described below.
Support Materials
The catalytic support serves two important func-
tions in a catalyst system:
It increases surface area of the active metal or
metal' oxide by providing a matrix that sta-
bilizes the formation of very small particles.
It increases thermal stability of these very small
particles, thus preventing agglomeration and
sintering with consequent loss of active surface.
Most common support materials are AI203 and Si02.
The application of these supports in catalytic combus-
tion presents several significant problems that must be
considered:
The large space velocities used in a catalytic
combustion unit will result in severe pressure
drops if the catalyst beds are packed with con-
ventional pellets like those used in the petro-
leum industry.
The high reaction rates of complete combustion
usually present severe pore diffusion limitations
for the effective operation of the system.
The high temperatures and heat fluxes place
special constraints on the thermal properties of
the system.
Over the last decade, a number of new materials and new
preparative techniques have been developed to overcome
these problems. These materials have emphasized the use
of monolithic support structures and surface impreg-
nated pellets.
Monolithic supports are composed of small parallel
channels of a variety of shapes and diameters. These
structures may be in the form of honeycombed ceramics
extruded in on~ piece, oxidized aluminum alloys in rigid
cellular configurations, or multilayered ceramic corruga-
tions. The channels in honeycomb-I ike structures have
tubular diameters of 1 to 3 mm. The overall diameter of
the monolithic support may vary from 2 cm to 60 cm,
and is limited, in the case of extrusion processes, by
374
-------
availability and operation of the metal die. Materials of
fabrication are usually low surface area ceramics such as
mullite (3A1203 . 2Si02) or cordierite (2MgO . 5Si02 .
2A1203). The refractory monolith is produced with
macropores (11-1 - 101-1) and may be coated with thin
layers of catalytic materials, 5- to 20-vveight percent
coatings being common.
The two major advantages of monolith supports for
catalytic operations are the high superficial or geometri-
cal surface area and the low pressure drop during oper-
ation. Table 1 lists the currently available monolith
materials, and figure 1 shows typical extruded monoli~h
structures.
In addition to the monoliths, highly porous pelleted
materials can also be used as the catalyst support. Pellets
come in a wide variety of materials, shap'es, and sizes,
but have the disadvantage of a higher pressure drop
through the bed.
Wash Coat Materials
The low surface area of the monolith structure re-
quires the application of a thin coat of oxide such as
AI2 03, This wash coat, which strongly adheres to the
ceramic or refractory support, provides a uniform high
surface area while still insuring that the catalytic
material which is subsequently impregnated on the wash
coat is close to the main flow of reactants. The thickness
of the wash coat is usually between 10 x 10-6 and 20 x
10-ti m (10-20 microns).
Many different materials can be used as wash coats.
ZrO'2. for example, has been used in test runs for the
catalytic combustion of hydrocarbons in automotive
Table 1. Available monolith supports
Temperature
Manufacturer Product Description limit (OF)
American Lava Thermacomb Dense 96 percent 2,800
Corporation A£SiMag 614 Alpha-alumina
Thermacomb Porous 96 percent 2,192
A£SiMag 776 Alpha-alumina
Thermacomb Cordierite 2,192
A£SiMag 795
Corning Celcor Cordierite 2,200
Glass Works 9475(EX-20)
E. I. du Pont Alumina 2,732
de Nemours TORVEX
& Company Mullite 2,462
General Cordierite 2,550
Refractories Versagrid
Company Mu 11 ite 3,000
Norton Sil icon Carbide 3,000
Company RX387
SPECTRAMIC Sil icon Nitride 2,800
Honeycomb RX384
Sil icon 2,800
Oxynitride
RX385
375
-------
........------------
..........................
...........................
..........................
..........................
..........................
..........................
,.........................
..........................
..........................
..........................
,.........................
..........................
t.........................
..........................
..........................
..........................
..........................
..........................
t.........................~
I..........................
t..........................
...........................
I..........................
I..........................
t.a........................
...........................
~~~......................
-.';';'''''''' .... ................. .
,y..~~~.~~.~.................
.."..","'.Y... ..................
."~~..~.A.~~..............
-~~~ ~~ ~.............
.......,.......,,,....~"........... ..
"."''''.''''..''. "...........
,.."~. "".&~4"A.A '...........
""..A'f'A"..,......v."". Jj '~::::::::
~"''''''A'''..'''''''''''''.''''' a".... . . ..
~,,,.,,....y.,,,,,,,,,,,.,,,,,.. '.... . . ..
~"."'&"''''.A'''J''Y..",.,''f'",,~ ,...... '"
...............,......,...........~
".""'''.'''.'''.'''''....''.''''''''''''.''~
"'''..'Y....'''.'''...",.,.......,.:
'A"A'''...'''.'''''..'''.'''"...",,~
""''''.''''A..'''''.....",....""
''''''.A'''''''A......."".",,,~
iY4.A""''''''Y.......'''.,....~
"~.....w.........._._.__._.- _.
Figure 1. Typical extruded monolith structures
(courtesy of General Refractories Company).
exhaust. Variation in the type of wash coat can have
important consequences in the stability and life of the
catalyst system.
However, a vast amount of data exists which permits the
development of correlations useful for choosing promis-
ing candidates based on excellent catalytic activity and
catalyst stability. Over the last decade there has been a
rapid development of the art, science, and technology
necessary to synthesize, test, and manufacture ultrast-
able oxidation catalysts for the abatement of automotive
emissions. These catalysts operate under the most strin-
gent conditions, and sometimes at temperatures close to
2,000° F. The expertise developed in this area will be
Catalyst Coatings
Although the field of catalysis has progressed sub-
stantially over the past decade, its theoretical aspects are
not yet at the degree of sophistication which would en-
able one, a priori, to choose or design an active oxida-
tion catalyst with a given set of catalytic properties.
376
-------
useful in the development of a stable catalytic com-
bustion system.
Two broad classes of catalytic coating materials are
available: metals and oxides. The metals of catalytic in-
terest are listed in table 2. Of these metals, the only ones
Table 2. Metals of interest for
catalytic combustion
Group VIlla Group ISa
Fe Co Ni Cu
Ru Rh Pd Ag
Os Ir Pt Au
aEnclosed metals are considered
noble.
which have a possibility of remaining in the metallic
state in a high-temperature, oxidizing environment are
the noble metals; the others readily form oxides. Of the
noble metals, a large volume of data and correlations are
available for platinum and palladium because of their use
as automotive oxidation catalysts. They are among the
most active catalysts for the oxidation of a number of
fuels, including methane, methanol, and hydrogen. The
high activity of these metals is related to their ability to
activate H2' °2, and carbon-hydrogen and oxygen-
hydrogen bonds. Palladium and platinum are readily pre-
pared in a highly dispersed form on a number of support
materials. Because of the high activity of these metals
Per unit area of metal surface (specific activity) and the
ability to attain high dispersions, only small amounts are
necessary for catalytic combustion (0.1- to 0.5-weight
percent). Ruthenium metal is another possible candidate
for catalytic combustion; however, under oxidizing con-
ditions it forms a volatile oxide (Ru04), which is rapidly
removed from conventional catalyst supports. One
approach to solve this problem has been to anchor Ru to
a support by forming a relatively stable perovskite
structure with certain oxides such as La203. Osmium is
even more volatile and poisonous than ruthenium in an
a x idizing environment. Also, as with iridium and
rhodium, it is very costly, and available in limited
supply. The use of the latter two metals, if at all, would
be restricted to small quantities in multimetallic systems.
Silver, while quite active at low temperatures for the
activation of O2, melts at low temperatures and there-.
fore sinters to a significant degree at the high tempera-
tures of catalytic combustion. However, it could find use
as an additive in a multi metallic catalyst. Gold is very
inactive for oxidation, and would therefore only be con-
sidered as a structural or electronic modifier for
mu Itimetall ic catalysts.
Based on the above considerations, those metals
which show the greatest promise for use in a single-
active-element catalyst system are platinum, palladium,
and ruthenium (stabilized). Highly active and stable
catalyst systems can be produced through the use of
multimetallic systems (such as palladium/platinum), as
well.
The catalytic properties of metal oxides have been
investigated by a number of researchers, including re-
search in catalytic oxidation. As for metals, it has been
found that the activity for hydrocarbon oxidation
parallels the ability to catalyze the homo molecular ex-
change of oxygen. It has also been noted that the oxides
of transition metals containing ions with partially filled
d-orbitals have the greatest activity. Based on these con-
siderations, the most active simple oxides are C03 04,
Mn02' NiO, CuO, C0203, Fe203, and V20s. Multi-
component metal oxides of interest include stable spinel
structures such as C0304/CUO, Cr203/CuO, and
La20/C0304.
Temperature Capability of Catalyst/Support System
The choice of an optimum catalyst/support com-
bination requires a consideration of the temperature
capability of a given combination. The low temperature
limit is dictated by the so-called "light-off" temperature.
Solutions to this problem are not complex for most
catalytic combustion systems, since noble metal
catalysts exhibit good light-off characteristics for most
fuels of interest. The major thermal limitatio~ for
catalytic combus.tion is the high temperature of opera-
tion. The phenomena which lead to catalyst deactivation
at high temperatures are:
1. sintering of the catalytic species,
2. changes in catalyst stoichiometry,
3. sintering of the wash coat,
4. thermal degradation of the ceramic support
material, and
5. mechanical failures.
Sintering of the catalytic species will lead to a con-
comitant loss in surface area and therefore a loss in
catalyst activity per unit mass of catalyst. The tempera-
ture at wh ich sintering occurs is a function of the
catalytic material. I n an oversimplified manner this sin-
tering or crystallite growth is a function of the melting
point for metals. A number of techniques have been
377
-------
developed to minimize or prevent sintering. For ex-
ample, multinetallic systems can be prepared which are
more stable than the constituent metals. A number of
structural promoters are also known.
Stoichionl!try changes for oxides can lead to a
catalyst with dramatically different catalytic properties.
Loss of latti:e oxygen can cause an increase in the
energy required to activate O2 and thereby decrease
combustion cctivity. One method for overcoming this
problem is to dope the lattice with a material which
causes a decrease in the energy necessary to activate
oxygen and also stabilizes the stoichiometry of the
lattice.
A third mechanism of thermal deactivation is wash-
coat sinterin!l. At temperatures above 1,650° F high
surface area 17' or "}'-AI2 ° 3 undergoes a phase change to
Q-AI203 with concomitant sintering. The change may be
from a surfacl: area of 300 m2/g to about 5 m2/g. This
sintering results in pore cfosure and a "burying" of
active catalyt;c sites in the AI20J' To minimize this
problem, it is possible to work with catalysts for which
the catalytic phase has been deposited on a presintered
AI2 03, Alterr atively, other materials which are therm-
ally more res. stant may be investigated as washcoats.
These might irclude Zr02' Th02' and U02.
At very high temperatures, the ceramic support, be
it spheres or rronolithic honeycomb structures, begins to
degrade. Typil:al monolith fail temperatures have been
given in table I. Although catalytic combustion units are
not expected to operate continuously at the tempera-
tures listed, the possibility of local hot spots occurring
due to uneven air/fuel mixing and local catalyst bed
nonuniformitil!s must be considered.
Thermal expansion and contraction may also occur
at high tempHratures. 80th pellet and monolithic
supports have therma I expansion coefficients different
than their mt>unting hardware. Considerable catalyst
mechanical at1 rition can occur if the catalyst becomes
loose, thus callsing a decrease in catalytic performance.
E>:ISTING APPLICATIONS OF
CATAL YTIC OXIDATION CONCEPTS
The existi.1!1 application of catalytic oxidation con-
cepts in industry are largely in the low-temperature area
at this time. The primary applications of catalytic
combustion ar,~ in nitric acid plant tail gas cleanup, in-
dustrial odor control, small catalvtic heaters, and auto.
motive oxidatit>n catalyst systern5.
Nitric Acid Plant Tail Gas r'eanup
Nitric acid is produced commercially by reacting
ammonia with air to produce nitrogen oxides, which are
then absorbed in water to yield nitric acid. Large
amounts of unabsorbed nitrogen oxides in the tail gas
from the absorption tower pose a pollution problem,
which has been solved in some cases by the catalytic
processing of this tail gas over platinum-group metals to
yield useful energy in the form of steam and/or power.
The basic reactions for the catalytic treatment of nitric
acid tail gas over platinum, palladium, or rhodium for
methane reducing fuel are:
1. <;;H4 + 4N02 4 4NO + CO2 + 2H2 ° (decoloriza-
tion ),
2. CH4 + O2 ~ CO2 + 2H2 ° (combustion), and
3, CH4 + 4NO 4 CO2 + 2H20 + 2N2 (abatement).
The tail gas reactor outlet temperature is usually be-
tween 1,250° F and 1,3800 F. Support materials for the
catalyst have included nichrome wire, alumina pellets,
and alumina honeycomb materials.
Industrial Odor Control
The eiimination of organic fumes is desirable from
both air pollution and fire safety points of view. The
platinum metals have been used for several years as
catalysts for the oxidation of carbon monoxide and a
wide range of organic molecules in the presence of air or
oxygen. Most odors are caused by the emission of low
concentrations of organic molecules to the air, with
these organic molecules always containing carbon and
hydrogen and generally sulphur, nitrogen, and oxygen as
well. Since most industrial odor problems are caused by
organic compound concentrations well below the level
required for spontaneous combustion in air, it is necessa-
ry to raise the contaminated air stream temperature to a
level at which combustion can occur. The use of
catalysts lowers the temperature needed to achieve odor
removal, and also lowers the necessary residence time at
the combustion temperature.
Low- Temperature Catalytic Heaters
Low-temperature catalytic heaters are characterized
by the small portable radiant heaters that operate at
temperatures below 8000 F. Operating characteristics in-
clude heat release rates of approximately 75 8tu/in.2
and lifetimes of several thousand hours. Gaseous fuel is
distributed in a uniform fashion to all parts of a catalytic
pad, and combustion takes place on the surface of the
fibers of the pad.
The Control Systems Laboratory of the EPA has
conducted an extensive emissions testing program on a
number of catalytic heaters, as reported in reference 7.
No correlation between emissions and specific heat re-
lease rates was found, and CO and UHC levels were
generally quite high.
378
-------
Automotive Exhaust Catalysts
Since platinum metal catalysts were selected for
emission control on conventional automative internal
combustion engines for the 1975 model year, a great
deal of development work went into these emission
control systems. In general, NOx emissions are con-
trolled by a reduction mechanism, and CO and UHC
emissions are removed by catalytic oxidation with
secondary air added to the exhaust steam to insure com-
nlete oxidation.
A reactor for catalytic exhaust emission control has
three essential design features:
A ceramic support on which the catalyst is
deposited in a manner that prevents attrition,
Rapid warmup of the catalyst unit, and hence
rapid light.off, and
Nearly complete conversion efficiency.
The essential features of a satisfactory CO/UHC
oxidation catalyst for automotive emission control are
(ref. 8):
A light-off temperature in the region of 2500 C,
Conversion efficiencies in excess of 90 percent
for both CO and UHC at space velocities up to
150,000/hr,
High temperature stability of the support and
platinic crystallite system at temperatures to
9500 C, and
Poison. resistance to compounds containing
lead, phosphorous, and sulfur.
Platinum-palladium catalysts have been extensively test-
ed in full-scale vehicle tests (ref. 9). The differences be-.
tween platinum and palladium catalyst performance
were quite small.
Reference 10 discusses possible substitute catalysts
for platinum in automobile emission control. I n this
study, it was concluded that base metal catalysts gener-
ally are not as effective as platinum/palladium catalysts
for automobile emission control, especially for hydro-
carbon and carbon monoxide oxidation. For NOx re-
duction, certain base metal catalysts initially are more
active and selective than are noble metal catalysts; how-
ever, most suffer from rapid deactivation.
CURRENT RESEARCH PROGRAMS IN
CATALYTIC COMBUSTION
Since catalytic combustors have excellent potential
for low NOx emissions, a number of research programs
investigating their use in automotive, gas turbine, and
domestic appliance applications are currently going on.
Gas Turbine Catalytic Combustor Research
Becau.se the gas turbine operates under high
(approximately 200 percent) excess air conditions,
thereby holding combustion temperature down, ,it has
'excellent potential for catalytic combustion applica-
tions. Current or recent research programs in this area
include those conducted by the Air Force Aero Propul-
sion Laboratory (ref. 11), NASA-Lewis Research Center
(ref. 12), Detroit Diesel Allison Division, General Motors
Corporation (ref. 13), and E nglehard I ndustries. I n all of
the studies reported to date, no information has been
made public concerning the catalyst system, thereby
severely limiting the usefulness of the research. However,
the present operating conditions have been reported to
be:
1.
2.
3.
4.
5.
6.
Inlet temperature: 4000-1,0000 F,
Operating temperature: 2,0000 -2,6000 F,
Pressure: '-'0 atm,
Heat release rate: 100,000-20,000,000
Btu/hr-ftJ -atm,
Combustion efficiency: 99.99 percent, and
NOx emissions: 1 ppm.
Automotive Application Research
The Jet Propulsion Laboratory has been conducting
tests on a compact onboard hydrogen generator for use
with a hydrogen-enriched gasoline internal combustion
engine. This modified fuel system, called partial hydro-
gen injection, substantially lowers NOx emissions and
also increases engine efficiency considerably. It was
found that a nickel oxide catalyst speeds up the partial
oxidation reaction, and nickel oxide coated alumina
pellets have been used as the catalyst system. Typical
operating conditions for this system are:
1. inlet temperature: 4500 F,
2. catalyst temperature: 1,7740 F,
3. pressure: 1 atm.
Further information on the JPL system is given in
references 14 and 15.
Domestic Appliance Catalytic Combustion Research
The Institute of Gas Technology is currently con.
ducting catalytic combustion research on ventless home
appliances using reformed natural gas (nominally 80 per-
cent hydrogen and 20 percent carbon dioxide) as fuel,
and on range-top burners using hydrogen as fuel. The
reformed natural gas catalytic combustion burners used
a noble metal catalyst placed on a permeable support
plate, and are self. igniting at room temperature (ref. 16).
PRELIMINARY DESIGN CONCEPTS
Based upon the current catalyst materials, existing
industrial uses, and current research programs, pre-
liminary conceptual designs for the application of
379
-------
catalytic combustion to residential space heaters and
commercial a,1d industrial boilers have been made. The
conceptual drsigns presented are by no means optimum,
since very little data exist on which to design a catalytic
combustor, tut rather represent possible choices for
catalytic corr.bustors that appear promising but which
are not currert:y being done.
Home Heater Retrofit
Nearly all existing gas-fired burners consist of the
naturally aspimted "stick" or "log" type burner, for
either warm air or hot water application. These burners
ilre generally ::If the premixed type and rely on ample
quantities of ~econdary excess air 10 keep the "combus-
tion chamber" cool. 111 reality, there usually is no com-
bustion chamber ilS such; firing is directly to the heat
exchange surfice.
Figures 2 and 3 show two possible replacements for
the conventional log burners. In the first case (fig. 2)
heat is transferred from a monolithic bed by radiation
only. Preliminary calculations show that heat transfer
rates are sufficient to keep the bed below 2.200° F with-
out exceeding a maximum metal temperature of 9000 F;
a summary of the calculations for figure 1 is given in
table 3.
Figure 3 presents a catalytic bed combustor for
retrofit application to a gas-fired furnace utilizing small
U-tubes of cooling water. The heat absorbed by the cool-
ing water is then transferred to the circulating air just
prior to entering the furnace. This arrangement requires
an auxiliary pump and finned tube heat exchanger. The
water cooling circuit is arranged to come on after the
catalytic bed has reached a specified temperature. D,e-
pending on the len~lth of time required for the water
/,'--" "--
-",
//l
..-
".
/
".
..-
/'
,
~
'",
/
I
\
\
/
/'
Prellli XE d
air-fuEl
,
,
I
. )/
"--- - ~
~
\
\
./
;' 1(P
"-O~
~'
,/
--
Catalyst bed
Figure 2. Home heater retrofit design #1.
380
-------
CONVENTIONAL
'r"EAT
'1 A. EXCHANGER
. "{ V SURFACE
, \
'. '-.... '
'-.... .. .. -'~ \
ft'f-' \
--~._~---,.: -" /..;.> ;'8'"
-;..-=- "~. ',--- "- ..:~"
,'. '~.':'~.- ,'-~, ',",. -, -, -- ,--;>-~II"
. . . " II
" " ,,' I' I" ,
~.. : ]f~~:ii U
~ ~, i~\':'II;'
// .. I, 'I III
I.h..' - J1L
\ .
'~
COOLANT
OUT
COOLI NG IIIBE S
INSIII.ATINC; PAnS
"',~
(OOLING AIR
PREMIXED
FUEL AND
AIR
Figure 3. Home heater retrofit design #2.
system to heat, it may be necessary that the circulating
air fan come on sooner than current practice to insure
that heat is extracted from the water and boiling does
not occur. Further design data are !Jiven in table 4.
The two home heater retrofit designs just presented
could be applied to oil combustion by adding an air
preheater and oil vaporizer. In the case of an oil-fired
furnace, however, new design appears more feasible than
retrofit.
Home Heaters - New Designs
The freedom of a completely new design presents
fewer constraints on physical size and heat transfer
arrangement. There will still be an overall size limit
(similar to current practice) but this is easily achieved.
Two other basic criteria which will not be as easy to
meet are cost and noise, the latter because of the higher
pressure fans that may be required to achieve higher heat
transfer rates. Within these limits a great variety of gas-
and oil-fired warm air and hot water heating systems can
be considered. For oil-fired units the requirements of
preheated combustion air and vaporized oil still exist.
For either fuel it is necessary to take into consideration
either cyclic operation or modulation to match the vari-
ation in load. Coupled with these aspects is the transient
response of the furnace. Following are duscriptions of
three designs which warrant further study.
The first design, pictured in fi!lurr: 4, illvolv,:s riJ.
diant cooling of the catalyst bed, \uan catalyl ic first
stage, addition of fuel before the second stage, and
catalytic combustion of the second stage. The unit is
compact, involving a concentric cylinder arrangement of
the catalyst bed, fuel addition units, and outer shell. The
outer shell can either be water or air cooled, with addi-
tional fins placed in the outer shell if it were air cooled.
The catalyst bed could either be a packed bed, a fiber
mat, or a monolith bed. Fuels for this application would
most likely be the gaseous fuels, but a prevaporized oil
could also be used. Thermal or electrical startup similar
to the retrofit design would be required for oil to
381
-------
Table 3. Data on home heater retrofit design # 1
Length: 15 inches
Wi dth: 2 inches
Height: 0.5 inch
Volume: 15 in3
Surface area per unit volume: 50 ft2/ft3
Height required to be coated with catalyst: 0.25 inch
Heat release in bed: 25,000 Btu/hr
Heat transferred by radi ation: 11,400 Btu/hr
Heat transfer coefficient on air side: 8.9 Btu/hr-ft2_0F
Pressure drop through bed: < 0.25 in. H20
Bed temperature: 2,200° F
Wall temperature: 900° F
Bed type: Honeycomb monolith
Number of units: 4
Table 4. Data on home heater retrofit design #2
~~idth: 2 inches
Length: 15 inches
Hei ght: 1. 0 inch
Heat release per burner: 25,000 Btu/hr
Total number of burners: 4
Number of cooling U-tubes: 6
OD of cooling tubes: 1/16 inch
Heat transferred from bed by radiation: 4,612 Btu/hr
Heat absorbed by tubes: 6,788 Btu/hr
Coolant flow rate per burner: 21.5 ga1/hr
Exit gas temperature: 2,200° F
Coolant inlet temperature: 70° F
Coolant ~xit temperature: 1200 F
Pres:;ure drop through cooling tubes: 11.16 ft H20
382
-------
I I
--
12 x 8
Room air -
A.-I0880
33"
Fins
Catalytic surface
2nd stage
Room air
~
4"
Catalytic ,surface
1st stage
Fuel plus combustion
air
8"
Addit ional fuel
Figure 4. Home heater design #1.
383
-------
vaporize the fuel and preheat the combustion air. Oil
vaporizing tu:)es would be placed between the cold shell
and the catal'ltic surface. Table 5 summarizes the impor-
tant design parameters. Note that the overall dimensions
are well within current practices, although the room air
fan, combustion blower, and fuel controls must be pack-
aged into the unit as well.
It must 1llso be noted that sufficient heat must be
extracted from the first stage to achieve two goals:
1. To insurE' the maximum flame temperature in the
second stage does not exceed 2,200° F ~ 2,500° F,
to achievf) low NOx. and be within the limits of the
catalyst;
2. To lower the temperature sufficiently to prevent
thermal combustion prior to the second stage
catalyst t ed after introduction of the second stage
fuel.
The latter criterion can be circumvented if the
velocities are greater than the flame speed. This can be
achieved by a~propriately sizing the passages.
The second design, shown in figure 5, uses a conical
fibrous catalyst surface to radiate to an outer air-or-
water-cooled shell. A premixed mixture of air and fuel
would be routed behind the fibrous pad and ignited at
the outer surface. After exchanging heat radiatively with
the outer shell the flue gases are passed through in a
multipass crossflow arrangement outside the tubes. This
design is a single stage, operating with minimum excess
air but relying on radiation cooling to keep the catalyst
surface cool. It could be designed to operate on oil.
Figure 6 shows a concept utilizing heat pipes be-
tween sections of catalysts. The heat is picked up by the
room air stream flowing adjacent to either side of the
combustion chamber. This design could easily use oil or
gaseous fuel. The first combustion chamber would be
modified to utilize a startup oil spray nozzle firing ther-
mally and later switched to vaporized oil. The oil vapor-
ization tubes could also be located between catalyst
beds. The remaining heat is finally extracted through a
convective section at the right-hand side of the figure.
Table 5. Data summary, home heater design # 1, 100,000 Btu/hr
ls t Stage
2nd Stage
Heat re lease (Btu/h r)
Stoi ct iometry
Gas inlet temperature (OF)
Gas exit temperature (OF)
Catalytic surface requirement. (fi)
Surface area/unit volume (ft2/ft3)
00 (inches)
10 (inches)
L (inc~es)
Pressure drop (inches H20)
Heat transfer area downstream: 12.26 ft2
Averag'~ overall U = - 5 Btu/hr-ft2_0F
Exi t gdS temperature: 4000 F
Use 28 fins 1.3 inches long on either side
Co01in9 air flowr?te: 1,000 scfm
Pressu:roe drop ~ir side: 0.122 in. H20
68,750
1.60
70
2200
6. 12
200
2.86
2.00
4
- 0.04
31,250
1. 10
1401
2591
2.78
200
2.43
2.00
4
> 0. 1
384
-------
COOLANT OUT
FIBROUS CATALYTIC
MAT
FLUE. OUT
'\
l CONVECTIVE
J HEAT EXCHANGE
SECTION
RADIANT SECTION
l- PREMIXED FUEL AND AIR
Figure 5. Home heater design #2.
385
COOLI NG
FLUID
-------
Catalyst Beds
Premi xed Fue 1
and Air
.~---
./
/'
/~---~
,-- '-----
COJlvec t. i ve Ilea t.
Exchange Section
Fuel Out
i t t
C~JilJlt. Out
Figure 6. Home heater design #4.
Commercial and Industrial Boiler Designs
The princ'pal existing commercial boiler styles in
use are the Cest iron and firetube designs. Within the
firetube design, packaged scotch and firebox are preva-
lent. All of the.e designs have relatively large initial com-
bustion volumlJs, whether gas or oil fired. The firebox
designs, including some of the cast iron boilers, employ a
refractory or r,~fractory felt-lined combustion chamber.
The others (packaged scotch) utilize a large water back-
ed volume for the initial combustion chamber. The on-
off cycles are :,ighly dependent on the individual load.
The principal design difficulty is to extract suffi-
cient heat froln the bed to achieve low NOx' All the
problems associated with oil.fired home heater appli.
cations also apply, but the must ~ignificant difference is
that the capital cost of these units is an order of magni-
tude higher than the homn heating systems. This allows
considerably more latitude in potential designs.
Figure 7 shows a potential new design concept for a
firetube boiler. The firetubes are filled with catalytically
impregnated pellets. A combustion chamber is added of
any desired length for first-stage thermal combustion, if
required. I n addition, a second-stage can be added as
desired, followed by additional cooling. Flue gas recircu-
lation can be added to this design as well, and it also
applies to the small industrial boiler.
A catalytic water wall design is shown in figure 8.
Actually, this is very similar to the Home Heater 8esiyn
#1 (fig. 4) in that it employs a cylindrical catalytic com.
bustor radiating to the water wall. This radiant section is
followed by a convective section. This design is a single
stage and relies on the high radiant cooling.
SUMMARY
Based on a review of the state.of-the.art of catalytic
combustion concepts, an assessment has been made of
their applicability to gas- and oil-fired home heaters and
386
-------
~ CONVENTIONAL BURNER
FUEL---
COMBIJ~;T I ON
CHM1BI:R
'-------,
I
1ST PASS
HEAT
EXCHANGER
OPTIONAL
CATALYST PLUG
WATER COOLED
SHELLS
- 2ND PASS HEAT
EXCHANG!:R
I: II
If
, II
SECT! ON
Figure 7. Commercial and industrial boiler design # 1, firetube boiler.
commercial and industrial boilers. Preliminary con.
ceptual designs have been made, but these designs are
limited due to the basic lack of knowledge of the
catalyst characteristics, wash coat, support, and bed
depth required for an oxidation reaction to go to com.
pletion.
Because of the need for high system efficiency (to
conserve fuel) and low operating temperature « 2,800°
F to prevent NOx formation), the system concepts of
flue gas recirculation, two-stage combustion, or direct
bed heat removal must be examined along with the
catalytic combustion process to achieve the desired goal.
There appear to be no major technical obstacles to pro-
ducing a catalytic combustion space heater or boiler at
this time.
Because of formidable problems associated with oil.
fired equipment, including oil vaporization, startup, and
air preheating, catalytic retrofit of oil-fired units is not
recommended. In fact, retrofit for gas-fired systems also
exhibits special problems, such as limited furnace
volume, low heat transfer coefficient, and cyclic nature
of operation, and therefore redesign represents the more
viable solution for catalytic combustion applications.
The basic lack of information concerning the
catalyst bed is a significant hindrance to the designer,
and a program aimed at obtaining this basic desi~ln data
through catalyst screening tests and small-scale systmTl
concept evaluation tests is strongly recommended.
387
-------
Coolant Q.Jt .....
Coolant In --
~
III
<:I
"-
I
-------
REFERENCES
1. H. Davy, "Some New Experiments and Observations
on the Combustion of Gaseous Mixtures," J. Davy,
ed., The Collected Works of Sir Humphrey Davy,
Vol. VI: Miscellaneous Papers and Researches, Part
I, "On the Safety Lamp for Preventing Explosions
in Mines, Houses Lighted by Gas, Spirit Warehouses,
or Magazines in Ships, etc., With Some Researches
on Flame," Part II, "Papers Published in the Philos-
ophical Transactions and in the Journal of Science
and the Arts, on the Fire Damp, the Safety Lamp,
and on Flame;" Smith, Elder, and Co., Cornhill,
London, 1840, pp. 81-88.
2. D. B. Spalding, "Heat Transfer from Chemically
Reacting Gases," Modern Developments in Heat
Transfer, W. Ibele, ed., Academic Press, N.Y., 1963,
pp.131-139.
3. A. L~'tken and H. Holst, Opfindelsernes 80g, Vol. 4
(1914), Nordisk Forlag, K0benhavn, pp. 131-139.
4. W. A. Bone, "Surface Combustion and Its Industrial
Applications," Engineering, Vol. 91, April 14, 1911,
pp. 487-489.
5. W. A. Bone, "Surface Combustion," Engineering,
Vol. 93, May 10, 1912, pp. 632.634.
6. W. H. Walker, W. K. Lewis, W. H. McAdams, and E.
R. Gilliland, Principles of Chemical Engineering, 3rd
Edition, McGraw-Hili Book Company, Inc., New
York, 1937, pp. 202-204.
7. R. E. Thompson, D. W. Pershing, and E. E. Berkau,
"Catalytic Combustion - A Pollution-Free Means of
Energy Conversion," Environmental Protection
Technology Series, EPA-650/2-73018, August 1973.
8. G. J. K. Acres, A. J. Bird, and P. J. Davidson,
"Recent Developments in Platinum Metal Catalyst
Systems," The Chemical Engineer, March 1974.
9. G. J. Barnes and R. L. Klimisch, "Initial Oxidation
Act ivity of Noble Metal Automotive Exhaust
Catalysts," SAE Paper 730570, presented at the
Automobile En!Jineering Meeting, Detroit, Michigan,
May 14-18, 1973.
10. National Materials Advisory Board, "A Semi-Delphi
Exercise on Substitute Catalysts for Platinum in
Automobile Emission Control Devices and Petro-
leum Refining," NMAB-314, March 1974.
11. W. S. Blazowski and G. E. Bresowar, "Preliminary
Study of the Catalytic Combustor Concept as
Applied to Aircraft Gas Turbines," Technical Re-
port AFAPL TR-74-32, Air Force Aero Propulsion
Laboratory, May 1974.
12. David N. Anderson, R. R. Tacina, and T. S. Mroz,
"Performance of a Catalytic Reactor at Simulated
Gas Turbine Combustor Operating Conditions,"
NASA TM X-71747.
13. F. B. Wampler, D. W. Clark, and F. A. Gaines,
"Catalytic Combustion of C) HI! on Pt Coated
Monolith," presented at Western States Section
Meeting, Combustion Institute, Northridge,
California, October 21, 1974.
14. J. Houseman and D. J. Cerini, "On-Board Hydrogen
Generator for a Partial Hydrogen I njection Internal
Combustion Engine," SAE Paper 740600, presented
at West Coast Meeting, Anaheim, California, August
1974.
15. J. Houseman and F. W. Hoehn, "A Two-Charge
Engine Concept: Hydrogen Enrichment," SAE
. Paper 741169, presented at I nternational Stratified
Charge Engine Conference, Troy, Michigan,
October-November 1974.
16. J. C. Sharer and J. B. Pangborn, "Utilization of
Hydrogen as an Appliance Fuel," presented at the
Hydrogen Economy Miami Energy (THEME) Con-
ference, Miami Beach, Florida, March 18-20, 1974.
17. W. V. Roessler et aI., "Investigation of Surface Com-
bustion Concepts for NOx Control in Utility Boilers
and Stationary Gas Turbines," Environmental Pro-
tection Technology Series, EPA-650/2-73-014,
August 1973.
389
-------
HEAT PIPE APPI,.IANCES
James F. Rice* and Edward F. Searightt
Abstract
Recent awareness of the extent of energy shortages
in this country has increased. the recognition of the ne-
cessity of des.'gning appliances which arc capable of pro-
viding signifi,:ant reductions in energy consumption.
This should, however, be accomplished without sacrific-
ing the ecolofJical objective of reducing emission of toxic
gases or vapc rs. The heat pipe appliances discussed in
this paper acc, Jmplish these objectives.
, The heat pipe concept has been in existence since
the early 7940's, though the technical development of
heat pipes as they are designed today, did not begin until
7963, Since then, this heat transfer device has received
considerable iiHerest for a wide range of applications.
At Thermo Electron, under the sponsorship of
Southern CaliFornia Gas Company, heat pipes have been
combined witll forced combustion and jet impingement
heat transfer to produce a group of gas-fired residential
and commerc:al appliances. These appliances utilize the
isothermal ch./facteristics of heat pipes together with the
inherent high efficiency and low emissions of forced
combustion s {stems to provide improved performance
compared to contemporary equipment. Included in
these appliances are a commercial griddle, an oven for
reconstitution of frozen foods, a deep fat fryer, and a
wa ter heater.
Typical f,1St data for these appliances show carbon
monoxide levl11s of 70 to 700 ppm and total oxides of
nitrogen concHntration of 5 to 20 ppm.
Cooking I'fliciency for the oven was improved from
less than 42 percent for conventional equipment to 54
percent. For the water heater, both operating and stand-
by losses wem reduced with the combustion efficiency
increased from 70 percent to over 80 percent. Similar
improvements were accomplished for the other appli-
ances. '
These appliances illustrate that heat pipes can be
applied in useful and practical designs to provid~ prod-
ucts with s'ignYicant advantages over conventional appli-
ances, including improvements in efficiency and
emissions, while providing uniformity of temperature
and better ten, perature control.
*Research Project Manager. Sc.,lthern California Gas Com-
pany, Los Angel\~s. California.
tManager, Clim:lte Control and Appliances. Thermo Elec-
tron Corporation. Waltham, Massachusetts.
INTRODUCTION
Recently a number of new heat transfer and com-
hustion techniques have been applied to the design of
commercial and residential appliances. Many of these
approaches had been in existence for some time, but had
been applied primarily to space age problems. In this
paper, four gas-fired appliances are described which uti-
lize heat pipes, powered combustion, and jet impinge-
ment heat transfer techniques to improve efficiency and
performance while producing low emissions of carbon
monoxide and oxides of nitrogen in the flue gases.
I ncluded are an isothermal griddle, a deep fat fryer, a
food reconstitution oven, and a storage water heater.
These products were developed at Thermo Electron
Corporation under the sponsorship of Southern
California Gas Company.
This paper reviews the basic technology involved in
gas-fired powered combustion, jet impingement heat
transfer, and heat pipes as applied to water heating and
cooking appliances. It illustrates that through proper en-
gineering techniques, these technologies can be
combined to produce appliance designs which are effi-
cient utilizers of energy and provide superior operating
characteristics.
HEAT PIPES
The heat pipe, which is a spinoff from the United
States space/nuclear technology program with significant
applications in the civilian market, is a device for trans-
ferring heat from point to point with no moving parts. It
can possess an effective thermal conductivity thousands
of times greater than that of most conductive metals
such as aluminum or copper. Because of this high effec-
tive'thermal conductivity, it can replace many conven-
tional fluid heat transfer systems which require electro-
mechanical pumps for fluid return,
The operation is fundamentally simple and easily
understood. There is, however, technical complexity in
the design and processing of reliable high-performance
heat pipes. The technical work started in the early,
1940's, and a patent, which is now in the public domain,
was issued to R. S. Gaugler in 1944 (patent No.
2,350,348). The technical development as we know it
now, however, did not begin until 1963, when G. M.
Grover of Los Alamos Scientific Laboratory independ-
ently hit on a similar device and coined the name "heat
pipe" to describe it.
390
-------
Figure 1 is a schematic drawing of a heat pipe. I t is a
Jollow chamber (not necessarily cylindrical) whose walls
include capillary structure, which assists in moving liquid
from one point to another and functions much as the
wick in an oil lamp. The chamber, which has been evac-
uated and then sealed, contains a substance (water, or-
ganic compounds, or liquid metals) which is in liquid-
vapor equilibrium at the heat pipe operating temper-
ature. When heat is applied to one end of the heat pipe,
liquid vaporizes, absorbing large quantities of energy and
raising the pressure at that end of the pipe. The resultant
high pressure drives the vapor to the end of the heat pipe
that is cooled, where the vapor condenses and releases
the heat absorbed at the heat input end of the heat pipe.
Thus, heat is carried from one end of the heat pipe to
the other via vapor. The condensed liquid is then
returned to the heat input end of the heat pipe by the
capillary forces developed in the wick structure. The
cycle is then repeated. The result of this process is that
large quantities of heat can be transported from one
place to another with very little decrease in temperature,
and the system is self-contained, thus requiring no
external equipment.
JET IMPINGEMENT
Jet impingement heat transfer techniques have been
'incorporated in a wide variety of applications, ranging
from cooling a thermionic diode to air heating in a hot
air furnace. Relatively large heat transfer rates are attain-
. able with jet impingement when compared with nonim-
. pinging flows. The jet impingement concept is also
attractive from tt'!e design standpoint, allowing flex-
ibility in control of surface heat transfer.
I n jet impingement heat transfer, gas flows through
a series of holes in an orifice plate, creating many jets.
These jets of gas strike the target and the momentum in
the jet reduces the thermal boundary layer on the target
plate, thereby increasing the heat transfer coefficient.
The momentum of the jet striking the target plate plays
an important role in reducing the thermal boundary
layer thickness. Small hole diameters will increase the
momentum of the jet at the exit, but jet spreading can
significantly reduce the jet momentum normal to the
target surface if the target is far from the orifice plate.
Closely spaced, small-diameter jets will provide many
local areas of. boundary layer reductions, but if the jets
are too close to each other, one jet may interfere with
the development of the other. If there are too many jets
in line with the direction of the outlet flow, the momen-
tum of the "spent gas" (i.e., outlet flow) tends to bend
the last rows of jets, reducing the momentum compo-
nent normal to the plate. Thus, the optimum heat
transfer design requires judicious iterations of orifice
diameter, orifice-to-orifice spacing, and orifice plate-to-
target plate distance. These jet impingement parameters
are shown pictorially in figure 2.
I n the design of appliances, it is desirable to mini-
HEAT IN
HEAT OUT
Figure 1. Heat pipe.
391
-------
" :\
? Xn
, ''''\. '\..'-
,~ "
/ '
Xn' ~
/ / ",
A'
t t
FLOW DIRECTION
.---ttEA1 TRANSFER SURFACE
~ ".', , , ," :"- "', " "" ".(., " ' " " " " " , " 'f '." '-
r f r r t in
, "A : mZl~o~'~;RJc;;~'~ ! w~ ,;,
ORIFICE PLATE
X n 'SPACING BETWEEN HOLES ON
SQUARE ARRAY OF HOLES
00 'DIAMETER OF HOLES
Zn : DISTANCE BETWEEN ORIFICE PLATE
AND HEAT TRANSFER SURFACE
Figum 2. Jet impingement parameters.
mize pressure dr9P at a given flow rate while maintaining
the highest possible heat transfer coefficient.
, Because ;)f its apparent advantages, an experimental
pro!lram was undertaken to investi~late thl! heat transfer
characteristic:; of jet impingement heat transfer. The
experimental study yielded some very promising results
with heat transfer coefficients of over 40 Btu/hr ft 0 F
for very modl!rate pressure drop.
POWERED COMBUSTION
To fully take advantage of potential improvements
in efficiency and size reduction for gas,fired appliances
made possibh! by use of heat pipe and jet impingement'
technologies,3 third technique is needed: powered com-
bustion. This also permits control of gaseous emissions"
within satisfactory limits.
Atmosph!ric burners have been used in almost all
residential and commercial gc~-fired appliance designs
used in cont.!mporary equipmer,t. The required com-
bustion volume has been determined by the energy avail-
able at the bl mer for mixing air with gas to permit the
combustion reaction. The pressure drop available in the
heat transfer system has been limited to the buoyance
force available in the column of flue gases. This results in
relatively low velocity in the gas stream and, therefore,
low heat transfer coefficients between the flue gases and
the transfer surface. The lower pressure drops available
also determine the diameter of vent needed, since again,
only buoyancy forces are available to perform this
function. The use of powered combustion provides pre-
cise control of the fuel-air ratio through the use of pre-
mix and a zero governor. This then permits a high heat
release rate per unit volume of combustion chamber.
Since buoyancy forces are not important, only small
diameter vent lines are needed and can run up, down, or
horizontally. Heat losses due to natural draft through
the appliance vent
-------
32"
35"
14'
1
~
'~._--u_- ~
~.~.~ .
,
-
......
JI:
l
r:
THERMO ELECTRON
ISOTHERMAL GRIDDLE
i
. i .
,), .
Figure 3. Gas-fired heat pipe griddle.
393
-------
with a series of jet impingement stages for the purpose
of heating the ,ower surface of the heat pipe so that the
flue products can be power vented. The combustion and
temperature control system consists of an off-on pres-
sure switch, which, sensing the heat pipe pressure, acti-
vates the combustion blower, the gas solenoid, and the
ignition system. The gas flow to the burner is propor-
tional to the motor speed to yield a constant gas/air
ratio by allowing the gas to be drawn through a zero
governor and f,xed orifice into the combustion blower.
In addition to f:he primary thermostat, there is an over-
pressure switch which cuts the main power as the evap-
orator vapor pressure approaches one atmosphere.
Table 1 is a comparison between the operation of
the heat pipe griddle and that of a conventional gas-fired
griddle of the same size. The results show that the heat
pipe griddle heats up in one-fourth the time, its effi-
ciency is twice as high, it consumes one-third of the fuel
on standby operation, its surface temperature variation
is an order of magnitude lower, and its cabinet surface
temperatures in all places are lower than those of the
conventional griddle.
Table 1. Performance comparison of heat pipe
griddle versus conventional griddle
Heat pipe
'griddle'
Con ven t i on a 1
griddle
Initial g.~s input, Btu/nr
Cont ro 1 griS input, Btu/hr
Time to reach 400° F, minute
Sys tern ef-Fi ci en cy, percen t
Steady s trite gas
consumption, Btu/hr
Control type
Control accuracy, of
Control tE!mperature
variation, of
Surface tE~mpera ture
variation,OFa
Cabinet tE'mperature, of
Exhaust tE!mperature, of
Excess 02' percent
Exhaust CO, percent
Exhaust C02' percent
aNot including the corners of the unit.
83,000
35,000
85,000
32,000
7
26
81
40
12,000
high-1ow/off-on
32,000
high-low/continuous
:t 30
:t 1 °
:t 5
:t 25
:t 2.5
:t 25
240
350
430
580
5.5
18
0.001
0.002
8.7
1.5
394
-------
A temperature profile of the heat pipe griddle sur-
face was measured while the unit operated at a control
setting of 345 0 F. The results of this measurement have
been plotted in figure 4. The plot shows that the entire
surface is essentially constant except at the extreme
corners, which are lower by 150 F.
These results dramatically establish the feasibility of
using the heat pipe principle to produce efficient, iso-
thermal cooking equipment.
HEAT PIPE OVEN
The frozen food reconstitution oven, shown in
figure 5, consists of a heat pipe and a sealed gas-fired
combustion system. The heat pipe forms the internal
cavity of the oven, which measures 36 inches wide x 24
inches deep x 12-3/4 inches high and contains three
shelves spaced 4 inches apart. The entire oven cavity is
surrounded lJy 2 inches of fibrous insulation to minimize
heat leakage. This produces overall dimensions of 40
inches wide by 28 ,inches deep by 25 inches high. The
front of t~e oven is sealed by two hinged, insulated
345°F
doors which swing out of the way to allow access to the
oven.
The combustion system is similar but not identical
to the combustion system used on the isothermal
griddle. It consists of a forced combustion burner and jet
impingement heat transfer stages and is capable of firing
rates up to 90,000 Btu/hr and combustion efficiencies of
80 percent. The overall efficiency of heating food is 54
percent as compared to less than 45 percent for conven.
tional ovens.
The fundamental principle of operation of the oven
consists initially of transferring heat from a gas-fired
combustion system to the heat pipe evaporator. The
working fluid in the evaporator is vaporized and travels
to the condenser surfaces. The vaporized working fluid is
condensed, then returned by gravity and capillary action
to the evaporator. Heat is transferred to the food by
radiation, conduction, and convection. This system is
capable of maintaining a constant and uniform tempera-
ture in the oven cavity over a range of 2000 F to 4500 F.
By employing the heat pipe principle in its design,
this oven possesses the following advantages:
340
345
340
Figure 4. Surface temperature distribution of heat pipe g'riddle,
36 in. x 24 in.
395
-------
\'Ii
--- !I'I'
Figure 5. Photograph of heat pipe oven.
396
-------
Temperature uniformity independent of the food's
position in the oven;
Maintenance of high humidity content;
Elimination of the large blower used in convective
ovens; and
Reconstitution time of frozen foods independent of
position in the.oven.
These advantages improve the overall quality of the
food, particularly when large quantities of food must be
prepared.
1.
2.
3.
4.
HEAT PIPE DEEP FAT FRYER
The heat pipe deep fat fryer is illustrated in figure 6.
The appliance consists of a heat pipe fry tank, a sealed
combustion chamber, and gas controls, all contained in
package measuring 15 Inches wide x 28 inches deep x 14
inches high. The fry tank, measuring 14 inches wide x 15
inches deep x 8-1/2 inches high, contains an integral heat
pipe covering its entire bottom and back. The heat pipe
contains wicking on the evaporator surface to distribute
the fluid uniformly over the entire evaporator area. The
bottom of the fry tank is pitched slightly to allow for
drainage at the front of the unit. A cold zone for the
accumulation of food particles is provide~ at the front
of the tank. The use of the heat pipe eliminates the need
for fire tubes in the fry tank, which in turn reduces the
required height of the tank and facilitates cleaning. The
heat pipe is capable of providing closely controlled and
uniformly distributed temperatures in the fry tank; these
features enhance the quality of the cooking and the..
longevity of the cooking oil.
The combustion chamber consists of a premixed
gas-fired burner and jet impingement heat transfer stages
contained in a sealed insulated housing which surrounds
the entire heat pipe section of the fry tank. .
The performance of the heat pipe fryer was as
follows:
Frying Performance
Input from fuel. . . . . . . . . . . . . 65,600 Btu/hr
Efficiency, based on
flue analysis. . . . . . . . . . . . . . . . . 77 percent
Energy requ irement,
AGA method. . . . . . . . . . . . . . . . 900 Btu/lb
Frying rate, raw-to-done
potatoes. . . . . . . . . . . . . . . . . . . . 84 Ib/hr
This unit provides a gas-fired deep fat fryer which is:
1.
2.
3.
as compact as comparable I!lectric units,
more efficient than current gas-fired units,
much lower in coc:,ing oil breakdown than either
gas or electric units,
4.
5.
easy to clean and maintain, and
economical to operate.
HEAT PIPE WATER HEATER
Contemporary gas-fired storage water heaters are
normally constructed with glass-lined steel tanks pro-
tected from corrosion by a sacrificial magnesium anode.'
The residential models include a flue, usually from 3 to
4 inches in diameter, extending through the tank and
terminating in a vent hood connection at the top. The
flue and bottom head assembly are glass-lined in a
separate operation from the shell and top head assembly,
and then the two subassemblies are welded together. A
combustion chamber is attached to the bottom of this
tank, the tank is insulated, an outer enameled steel jack-
et is added, and an atmospheric-type gas burner and con-
trol system are installed. Commercial heaters differ in
that they either use a number of internal flues (as many
as 12) or pass the flue gases around the outside of the
tank in a "floater" design in order to accomplish the
higher heat transfer required in commercial installations.
While designs of this type have provided reasonable
performance for many years, they are subject to certain
problems, as summarized in table 2.
Because the glass coating cannot cover all of the
steel surfaces in the assembled tank, some bare areas
occur which would corrode rapidly if they were not pro-
tected. For this reason, a magnesium anod~ is installed in
the tank to provide additional corrosion protection. The
anode has limited life, however, espec!ally in some
waters, so this, together with the solubility of the glass
coating in water, determines the life of the heater.
The anode itself may create a problem since, in
some waters, it promotes a type of chemical reaction
which can cause odor and taste problems in the water.
Scale formations can create problems in both com-
mercial and residential designs. In commercial desigm,
because of the high heat flux required, actual burnouts
or thermal stress failures can occur 'in waters with high
dissolved solid concentrations if the heater is not cleaned
frequently. In residential designs, the scale buildup is
more apt to result in an objectionable "rumbling" noise.
Both the center flue and floater desi!Jns result in a
continuous heat loss' from the stored water when the
burner is not operating. The flue is at essentially storetl
water temperature and promotes a thermal circulation of
air through the flue and out the vent system. This can
result in a service efficiency of less than 40 percent, even
though the flue gas losses during burning are less than 30
percent.
397
-------
--~
~-
III
III
~
III
rt..
.,~
III
I
III"
. -,
:d
II
Figure 6. Photograph of heat pipe deep fat fryer.
398
-------
Table 2. Advantages of heat pipe water heaters.
Con venti ona 1
des i gn
problem Cause Solution
1. Corrosion Solubility of tank With heat pipe design, heat
coating and anode life source is separated from
1 imits tank li.fe. tank construction, so
different materials can be
used with longer life
th an s tee 1 .
2. Rumb 1 ing Scale accumulation causes More uni form temperature
bubble formation which reduces scale formation.
shakes heater.. Bottom head heat transfer
is not used, so heat is
not transferred through
heavy scale deposits.
3. Eff1 ciency Heat losses from exposed No flue passageways exist
flue surface result in in the heater, so service
high standby losses efficiency will approach
causing service efficiency recovery effi ci encr {approx-
of less than 40 percent in imate1y 70 percent.
some cases.
4. Large vent Con venti ona1 gravi ty Power combustion system
system combustion system requires requires no chimney, and
large diameter vent and uses a small di ameter vent
chi mney. line.
5. High surface Conventional burner and With forced combustion, a
temperature combustion system makes simple chamber can confine
containment of heat the heat of combustion
difficult in combustion eliminating hot jackets~
chamber area.
6. Taste and In some waters, use of With nonstee1" tank
odor magnesium anodes promotes construction, an anode is
a chemical reaction which not requi red.
causes unpleasant odor
and taste.
. The use of gravity vent systems requires connection
to a chimney, as well as a large diameter flue pipe, and
large vent hoods, especially in the case of high input
commercial designs.
Surface temperatures of the heater jacket in the
combustion chambE!r area are difficult to restrict to
reasonable levels.
In the proposed design, heat is transferred from the
products of combustion to the water by a heat pipe.
. This permits separation of the primary heat source from
399
-------
the tank construction, thereby permitting greater flexi-
bility in selection of tank materials. Molded, reinforced
plastic matE'rials can be considered which are not subject
to thermal or chemical degradation by the water; thus,
heater life f:an be improved and use of anodes can be
eliminated.
Since the heat pipe design can result in elimination
of flue passageways in the water heater, the attendant
standby los~es caused by air flow through the heater can
be eliminated. Service efficiency can be improved from
40 percent (0 values near 70 percent. In addition, the
use of a heat pipe for transferring heat from the flue
gases to the water should reduce scale formation and
permit easier cleaning.
HEAT
PIPE
BURNER
f:VAPORATOR
GAS-
The use of a forced combustion jet impingemen~
heat transfer system will eliminate the requirement for a
conventional chimney. Since the heat of combustion can
be confined to a simple chamber in this system, the
jacket temperature problems can be eliminated.
--'Figure 7 illustrates the heat pipe-water heater
co ncept. The heater shown consists of a plastic,
40-gallon storage tank with a conventional plastic cold
water dip tube, a drain valve, and an immersion thermo-
stat. The tank is coated with a "foamed in place" insula-
tion which also serves as the heater jacket. The tank rests
on a cylindrical base which contains the combustion
system and heat pipe assembly.
The combustion system consists of a 40.000-Btu/hr
..-----............
...
...
,
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CONDENSER
INSULATION
.
FLUE
Figure 7. Heat pipe storage water heater.
400
-------
Ifarced drillt hurner which directs hot flue products
IICroSS thlJ evaporator surfacH of u heat piplJ. The flue
gases are then collected and exhausted through a 1-1/2
inch-diameter vent pipe.
FORCED COMBUSTION EMISSIONS
An important consideration in any new appliance
design is the level of emissions in the flue gases. Of par-
ticular interest is the concentration of oxides of nitro-
gen. Table 3 gives NOx, NO, and N02 levels for three of
the designs discussed in this paper as a function of
oxygen level in the flue gases. The recorded NOx levels
iII'l) distinctly 10wl)! Ihan for convl)IIIiollal appliancl!s
using atmosphuric hurnurs. Typical carbon mOl1oxicfH
levels for these appliances range from 10 to 100 ppm.
SUMMARY
This paper has used four new gas-fired appliance
designs, a water heater, a griddle, an oven, and a fryer, to
illustrate the ability of heat pipes, combined with new
heat transfer techniques, to provide substantial improve-
ments in efficiency and emissions, as well as perform-
ance features in practical equipment designs.
Table 3. Oxides of nitrogen emission from forced
combustion appliances
Heat pipe oven Heat pipe fryer
Gas input, Btu/hr 86,000 88,000 67,000
Exhaust 02' ,6.5 3.2 6.5
percent
Exhaust C02' 8.0 9.9 8.0
percent
NO , ppm 17.5 104 13.5
x
NO, ppm 16.5 100 13
N02' ppm 1.0 4 0.5
Water heater
Gas input, Btu/hr 38,800 36 , 700 38,600 38,700 38,800 38,800 38,400
Exhaust 02' 3.3 4 4.8 5 5.2 5.5 6
percent
Exhaust C02' 9.2 8.9 8.5 8,3 8.2 7.9 7.6
pe rcen t
NO~, ppm 82 21 16.5 12 10.0 8.0 5.8
NO, ppm 75 18.5 15 10.5 9.5 7.5 5
N02' ppm 7 2.5 1.5 1.5 0.5 0.5 0.8
401
-------
402
-------
6 November 1975
Session VI:
NEW INDUSTRIAL PROCESSES
G. Ray Smithson*
Session Chairman
* Assistant Manager, Energy/Environmental Programs Office, Battelle Columbus Laboratories, Columbus, Ohio.
403
-------
404
-------
.consumed in the United States. Figure 1 gives a break-
down of energy production and consumption during
1971. While current figures may differ somewhat from
those indicated, the situation probably remains essenti-
ally unchanged on a relative basis.
Looking first at the electric power generation sec-
tor, we see that significant amounts of natural gas and
oil are consumed for generation of electricity. From an
energy policy point of view, this is not defensible and, in
fact, efforts are being. made to encourage use of coal
(with stack-gas scrubbing). This sector represents a large
market for coal-derived fuel gases and there is presently
a major research program directed toward utilizing such
fuels in combined gas turbine-steam turbine power
cycles.
Because of the properties of coal-derived gases (to
be discussed), it does nut appear that their use in the
residel1tial/commercia.1 sector would be attractive. The
use of natural gas and oil in this sector is justifiable.
Similarly, no application of coal-derived gases in the
transportation sector is apparent.
. It is the industrial sector of the economy that is the
major consumer of energy in the United States. Industry
consumes about 18 percent of the oil, 36 percent of the
coal, and 45 percent of the natural gas utilized in the
United States each year. Only about 14 percent of the
total energy input goes to nonfuel uses, i.e., feed stocks
(ref. 1); approximately 50 percent is used for steam gen-
eration for process or space heating and about 34 per-
cent is used in direct heat applications (ref. 2).
Fully 93 percent of the natural gas used by industry
is burned in process furnaces or .in boilers. In some in-
dustries, natural gas is consumed wastefully; frequently,
only 5 to 25 percent of the heating value of the fuel is
utilized in a process (refs. 3,4).
It is the authors' belief that the proper use of
natural gas is primarily within the residential/commercial
sector. In the discussion to follow, it will be shown that
the technology exists to produce substitute fuels from
coal that will be suitable for most industrial purposes.
The utilization of coal-derived gases would permit dis-
placement of natural gas from industrial to more effici-
U.S. ENERGY FLOW. DIAGRAM (1971)
NA TURAL GAS 10.438
OIL 5.391 INDUSTRY 2.329
22.623 HYDRO 2.798
r----------
COAL 4.465 I
I I I NUCLEAR 0.405
TOTAL=20.294 I r ----------
I I
~ ~
Nf.TURAL GAS 7.346 NATURAL GAS 4.125
-
RESIDENTIAL ELECTRIC
OIL 6.545 AND POWER
COMMERCIAL 3.160 OIL 2.417
GENERA TION
17.441
COAL 0.390 COAL 7.698
.
I TOTAL-14.281 I I TOTAL=17.443 I
NA TURAL GAS 0.825
TRANSPOR-
TATION LOSS 11.936
OIL 16.139 0.018
COAL 0.007 16.969 I 5.507 I
I TOTAL-16.971
i ;~ ..053 I
~
...
FLOWS IN QUADRILLION BTU
DATA F"ROM DUPREE AND WEST,
"U.S. ENERGY THROUGH THE YEAR 2000"
U.S. DEPT. OF" THE INTERIOR (1972)
Figure 1. U.S. energy flow diagram (1971).
406
-------
PROPERTIES OF INDUSTRIAL FUEL GASES
MANUFACTURED FROM COAL USING
COMMERCIALLY PROVEN TECHNOLOGY
James I. Joubert, Ph.D., and Daniel Bienstock*
Abstract
Gaseous fuds have been manufactured from coal for
well over a ceniury. Although coal gasification was once
widely emploYld in the United States, the advent of
inexpensive natllral gas and petroleum in the 1920's and
1930's virtually eliminated such processes.
Industry, which consumes about 45 percent of our
natural gas production, is now faced with curtailments in
supplies of this fuel. Fuel gases of low (150 Btulft3) and
medium (300 l1tulft3) heating value could be substi-
tuted for natural gas in many industrial applications.
Such fuel can bl] manufactured from coal using existing,
well-proven ted'nology.
This paper discusses the properties of coal-derived
fuel gases relev;mt to combustion processes. General re-
lations are presented that show that the combustion pro-
perties are mai,1ly a function of the fuel gas heating
value, and are largely independent of the process and
coal used for production of the fuel gas. .
Various tv,'Jes of commercial coal gasifiers, gas
cleanup systemi, and gasification costs are discussed
briefly. Approaches to utilization of an integrated gasifi-
cation system, .IS well as problems to be anticipated in
retrofit applicat,'ons, are also considered.
Introduction
A great dl!al has been written about America's
energy situatiol within the last 2 years. Numerous
articles have ap~'eared discussing alternate energy sources
such as solar, geothermal, oil shale, synthetic natural gas,
and oil from coal. Unfortunately, it appears now that
more than a d,!cade will pass before any of these ap-
proaches could make a significant contribution to meet-
ing our energy mquirements.
The question remains, then: What can be done
now, with existing technology and resources, that will
begin to make an impact on our Nation's energy prob-
lems in the near term?
First, let w; examine the situation with respect to
our fossil fuel resources. It is well known by now that
. James I. Jc ubert is an Assistant Research Supervisor, and
Daniel Bienstock is a Research Supervisor with the Energy
Research and Development Administration, Pittsburgh Energy
Research Center, ~ittsburgh, Pennsylvania.
our domestic reserves of oil and natural gas are dwindl-
ing and that we are becoming increasingly dependent on
foreign sources, particularly in the case of oil. It is likely
that suitable economic incentives could alter the situa-
tion somewhat, but this is speculative at this point.
It is also well known that coal represents our largest
indigenous fuel reserve. Estimates of the amount of coal
economically recoverable indicate that this fuel would
last 300 to 800 years even if consumed at a rate equal to
our present and anticipated future total energy demand.
At the moment, coal contributes only about 18 percent
of the energy utilized in the United States annually.
The reasons. why coal contributes such a small
amount to our energy needs are fairly evident. It is im-
practical for use in the transportation sector (i.e., the
automobile). The cleanliness and convenience of gas and
oil have largely displaced coal from domestic and com-
mercial use. For the same reasons, gas and oil are pre-
ferred by industry. In addition, because of its ash and
sulfur content, coal is simply not suitable for certain
manufacturing operations. With increasingly more strict
pollution regulations, there has been a trend toward dis-
placement of coal from electric power generation facili-
ties. .
There are, therefore, three main reasons why coal
occupies its present position in the energy picture: (1) it
is a solid fuel, and somewhat difficult to handle; (2) it is
a dirty fuel-much of our coal has a high sulfur and ash
content; (3) the more convenient fossil fuels, natural gas
and oil, have been available at lower or reasonably com-
petitive rates.
However, the situation has been changing rapidly
with respect to item (3) listed above. If coal could be
converted to a clean gaseous fuel, the problems associ-
ated with the direct use of coal would be largely elimi-
nated.
It is the purpose of this paper to discuss the major
aspects of utilizing clean fuel gases manufactured from
coal using existing technology. By clean fuel gases it is
meant gases with heating values ranging from about 150
to 300 Btu/ft3. This type of fuel gas is not to be con-
fused with "high-Btu" gas made from coal, (1,000
Btu/fe ), which presently is the subject of intensive re-
search.
To determine whether utilization of the fuel gases
of the type described could contribute significantly to
meeting our energy needs, let us examine how energy is
405
-------
.consumed in the United States. Figure 1 gives a break-
down of energy production and consumption during
1971. While current figures may differ somewhat from
those indicated, the situation probably remains essenti-
ally unchanged on a relative basis.
Looking first at the electric power generation sec-
tor, we see that significant amounts of natural gas and
oil are consumed for generation of electricity. From an
energy policy point of view, this is not defensible and, in
fact, efforts are being made to encourage use of coal
(with stack-gas scrubbing). This sector represents a large
market for coal-derived fuel gases and there is presently
a major research program directed toward utilizing such
fuels in combined gas turbine-steam turbine power
cycles.
Because of the properties of coal-derived gases (to
be discussed), it does nGt appear that their use in the
residential/commercial sector would be attractive. The
use of natural gas and oil in this sector is justifiable.
Similarly, no application of coal-derived gases in the
transportation sector is apparent.
It is the industrial sector of the economy that is the
major consumer of energy in the United States. Industry
consumes about 18 percent of the oil, 36 percent of the
coal, and 45 percent of the natural gas utilized in the
United States each year. Only about 14 percent of the
total energy input goes to nonfuel uses, i.e., feed stocks
(ref. 1); approximately 50 percent is used for steam gen-
eration for process or space heating and about 34 per-
cent is used in direct heat applications (ref. 2).
Fully 93 percent of the natural gas used by industry
is burned in process furnaces or in boilers. In some in-
dustries, natural gas is consumed wastefully; frequently,
only 5 to 25 percent of the heating value of the fuel is
utilized in a process (refs. 3,4).
It is the authors' belief that the proper use of
natural gas is primarily within the residential/commercial
sector. In the discussion to follow, it will be shown that
the technology exists to produce substitute fuels from
coal that will be suitable for most industrial purposes.
The utilization of coal-derived gases would permit dis-
placement of natural gas from industrial to more effici-
U.S. ENERGY FLOW DIAGRAM (1971)
NA TURAL GAS 10.438
OIL 5.391 INDUSTRY 2.329
22.623 HYDRO 2.798
r----------
COAL 4.465 I
I I I NUCLEAR 0.405
TOTAL-20.294 I r ----------
I I
H
Nf.TURAL GAS 7.346 NATURAL GAS 4.125
RESIDENTIAL ELECTRIC
OIL 6.5415 AND POWER
COMMERCIAL 3.160 OIL 2.417
GENERATION
17.441
COAL 0.390 COAL 7.698
.
I TOTAL-14.281 I I TOTAL=17.443 I
NATURAL GAS 0.8215
TRANSPOR-
TATION LOSS 11.936
OIL 16.139 0.018
COAL 0.007 16.989 15.1507 I
-
I TOTAL 16.971
i .~..053 I
...
...
F"LOWS IN QUADRILLION BTU
DATA F"ROM DUPREE AND WEST,
"U.S. ENERGY THROUGH THE YEAR 2000"
U.S. DEPT. OF" THE INTERIOR (1972)
Figure 1. U.S. energy flow diagram (1971).
406
-------
~nt end uses, as well as provide an alternate fuel for
industries presently experiencing curtailments in natural
gas supplies.
History of Coal Gasification
Despite the publicity generated by various novel
coal gasification schemes in recent years, the technology
is hardly a new one. Coal has been gasified, using one
technique or another, for nearly 175 years.
A list of the various types of fuel gases that have
been manufactured from coal is given in table 1. The
compositions and heating values indicated are only
typical and may vary somewhat depending on the coal
and the equipment used to manufacture the gas.
Coal gas, or retort gas, was made by heating coal in
the absence of air. It appears that the principal involved
was discovered as early as 1609 (ref. 5). This was the
first type of fuel gas manufactured for public distribu-
tion, with gas light companies being established in
London in 1807 and in Baltimore in 1816. The process
involved, however, was not very efficient from the point
of view of gas production; most of the coal was con-
verted to coke.
Coke oven gas results from the distillation of coal in
byproduct coke ovens and is quite similar to retort coal
gas. Again, the main product of the coke oven is coke,
but with larger yields of other products than in retort
processes. I n the 1930's and 1940's, coke oven gas repre-
sented a significant portion of the manufactured gas dis-
tributed by gas utilities (ref. 6). Today, coke oven gas is
consumed largely within the steel industry.
Blue water gas was produced by reaction of steam
with incandescent carbon. The process was intermittent,
consisting of periods in which the fuel was heated with
air, and gasmaking periods during which steam was
passed through the hot fuel bed. Blue water gas was used
largely as a raw material in the production of car-
buretted water gas.
Carburetted water gas was made by enriching blue
water gas with the gaseous decomposition products of
various oils. This raised the heating value to 530 to 550
Btufft3, which met the minimum requirements for dis-
Table 1. Coal-derived fuel gases
Percent by volume Heating
C02 02 CO H2 CH4 Ill. a N2 valueb(HHV)
Btu/SCF
Coal gas (retort) 2.4 0.8 7.4 48.0 27.1 3.0 11. 3 542
Coke oven gas 2.0 0.3 5.5 51. 9 32.3 3.2 4.8 569
Blue water gas 5.4 0.7 37.0 47.3 1.3 0.0 8.3 287
Carburetted water 5.5 0.0 32.5 37.0 12.0 9.0 4.0 540
gas
Producer gas 4.5 0.6 27.0 14.0 3.0 0.0 50.9 163
(air/steam)
(low-Btu gas)
Synthesis gas 12.8 0.3 53.3 30.9 0.4 0.0 2.3 275
(02/steam)
(medium-Btu gas)
a I 11 urni nants: C2H4' C3H8' C3H6' etc.
bSCF at 600 F, 1. 0 a tm.
407
-------
tribution. Thi~ was the only commercial process employ-
ed in this cour try with the sole purpose of producing gas
for distribution by gas utilities.
It is the utilization of producer gas and synthesis gas
that is the mail subject of this paper.
Producer aas, made by the steam/air gasification of
coal, is genereilly recognized as the cheapest form of
coal-derived fuel. Typical heating values range from 130
to 180 Btu/fe (low-Btu gas). *
If pure oxygen is used in place of air, synthesis gas,
or medium-Btll gas results with heating values ranging
from 250 to ~IOO Btu/ft3. Unlike low-Btu gas, this fuel
has not been produced commercially in the United
States, although it has been used for a number of years
in other coun':ries, mainly as a feedstock in chemical
plants. Its production on a large scale was not feasible
until the adven t of tonnage oxygen plants.
Unlike th£' processes involved in the manufacture of
other coal-dt1rived gases, production of low- and
medium-Btu gCls is continuous and efficient. Virtually all
of the combustible material in the coal can be converted
to gaseous fuel.
It appears that the first commercial gas producer
was constructed by Bischof in Germany, in 1839 (ref.
7). Units were subsequently built by Ebelman in France
(840), and by Ekman in Sweden (1845).
The first large-scale use of gas producers was
brought about by the Siemens brothers in Germany,
who demonstrated their combined gas producer/blast
furnace in 1861. By 1881, applications of producer gas
included firing in small industrial furnaces, and power
generation by neans of gas engines.
The many advantages of a gaseous fuel over a solid
fuel, such as c:>al or wood. caused rapid acceptance of
gas producers i, many manufacturing operations. By the
mid-1920's, ab.)ut 150 companies were engaged in the
design and supply of producer gas equipment (ref. 7).
During that pHiod, approximately 11.000 units were
operating in thl1 United States alone (ref. 8); these were
employed principally in the steel, glass, ceramics, lime,
metallurgical, and chemical industries. About 1,000 gas
producers were being used for power generation at that
time.
It appears that utilization of producer gas reached
its peak in 19:m (ref. 9). Natural gas and petroleum
already were being used to a large ext~nt in industry; the
costs and incre,lsing availability of these fuels left little
incentive for Cldditional investmp.nt in producer gas
facilities. Coke-r)ven gas was useu IOcreasingly within the
steel industry (t'1e major user of ga~ producers).
"The .terms low-Btu gas and medium-Btu gas are not well
defined, but are U! ed hore to .facilitate discussion.
It is estimated that by 1936 about 2,600 gas pro-
ducers remained in continuous operation (m!. 9). By
1948, the number had dwindled to fewer than 2,000;
about three-fourths of these were considered obsolete
(ref. 6).
Today, there are only a few gas producers remaining
in the United States. These are used intermittently or are
kept in standby condition.
Fortunately, because of the lesser availability of oil
and natural gas in other countries, development of gas
producers has continued over the past several decades. A
number of modern systems are now considered commer-
cially available. in the United States. A brief description
of each is given in the next section.
Current Commercial Gasification Technology
A. Operational Aspects of Commerical Coal Gasifiers
The coal gasification systems presently commerci-
ally available may be categorized according to the three
basic ways in which the coal is processed: (1) fixed-bed;
(2) fluid-bed; and (3) entrained-flow. While many varia-
tions are possible, some general comments can be made
regarding these basic types that may aid in judging the
suitability of a given gasifier for a particular application.
1. Fixed.bed gasifiers
Fixed-bed gasifiers are generally recognized as the
most efficient devices for converting a solid fuel into a I
combustible gas. In a fixed-bed gasifier, the fuel flows
countercurrent to the gasifying medium, which pro-
motes efficient transfer of chemical energy from the fuel
to the final product gas. An extremely high carbon con-
version is usually achieved due to the long residence time
of fuel in the reaction vessel. Because of the relatively
low gas velocities in the system, the carryover of solids is
generally small. Oft-gas temperatures range from
8000 -1,2000 F, which should lead to lower efficiency
losses than in other gasifier types if it is necessary to
cool the gas prior to cleanup.
A criticism of fixed-bed gasifiers has been that they
are not able to gasify caking coals. While this may be
generally true, it appears that if a fixed-bed unit is
equipped with a suitable stirring device such coals can be
successfully processed. Usually, carefully sized fuels
must be used with a minimum of fines. The ash fusion
temperature of the fuel imposes an upper limit on
maximum bed temperature, which may preclude the
processing of fuels of low reactivity.
Usually tar is produced in the lower temperature
regions near the top of fixed-bed units; this could cause
problems in downstream components.
Because of the generally small throughputs possible
with fixed-bed gasifiers, these units are most suitable for
small fuel consumers or in situations where multiple
units are deemed desirable.
408
-------
2. Fluid-bed gasifiers
The principal advantages of fluid-bed gasification of
coal are claimed to be: the ability to gasify coals with a
broad range of particle sizes; a high rate of gasification
per unit cross sectional area; operability over a wide
range of throughputs without a significant loss in
efficiency; and good temperature control due to particle
mixing within the bed.
On the other hand, because the exit gas is essentially
in thermal equilibrium with the solids in the bed at gasi-
fication temperature (1,500°.2,000° FI. there is a con-
siderable loss of thermal energy from the gasifier proper.
This energy must be recovered in downstream compo-
nents to maintain reasonable efficiency. Compared to
fixed beds, fluid beds exhibit fairly high carryover of
combustible solids that must be recovered and recycled
to the gasifier or utilized in some other fashion. Removal
of ash from a fluid bed, without loss of fuel, is also a
difficult task. Tar formation mayor may not be a prob-
lem, depending on the fuel and the operating conditions.
Finally, experience with fluid-bed gasifiers now being
developed for various high-Btu gasification processes has
shown that coals with free-swelling indices greater than 2
generally cannot be processed without pretreatment or
dilution with inert solids.
3. Entrained.flow gasifiers
Entrained-flow gasifiers can utilize virtually any
grade of coal. Unlike fixed-bed and fluid-bed units, no
tars are produced since entrained-flow gasifiers are ideal-
ly operated at high temperatures (in the slagging region).
The overall gas production rate per unit volume of an
entrained-flow system is substantially higher than for
fluid-bed or fixed-bed gasifiers.
The major drawback of entrained-flow gasifiers is
due to the high gas-flow rates and short residence times
of particles in the reactor. Since part of the coal charged
is carried out of the reaction space, it must be mechani.
cally recovered and recycled to improve carbon conver.
sion. Because of high exit-gas temperatures
(2,500°-3,000° FI. heat recovery is imperative; however,
heat recovery, as well as solid collection, may be compli-
cated by the presence of molten ash particles in the exit
gas stream.
B. Commerical Coal Gasifiers
1. Koppers-Totzek
The Koppers-Totzek is an entrained-flow type gasifi-
er marketed by the Koppers Co., Inc., of Pittsburgh.
Gasification is carried out at high temperature using
oxygen and steam, and about 50 percent of the coal ash
exits the gasifier as molten slag. Operation is essentially
at atmospheric pressure or slightly above.
Units have been built with outputs ranging from 36
million Btu/hr to 234 million Btu/hr; larger units, with
outputs of 560 million Btu/hr, are under construction.
Turndown capability is about 60 percent of maximum
output.
The first commercial Koppers-Totzek was con-
structed in 1952. Since then, 48 units have been built or
are under construction in various parts of the world
largely for production of synthesis gas for ammonia
manufacture.
2. Lurgi
The American Lurgi Corporation (New York)
markets a fixed-bed gasifier that operates at pressures
ranging from 20 to 35 atmospheres. Maximum output of
a 12-ft-diameter unit (largest size made) is about 230
million Btu/hr with a turndown capability of 50 percent.
The gasification agent may be either air-steam or
oxygen-steam. Ash is removed through a revolving grate
at the bottom of the unit.
The first commercial installation of Lurgi gasifiers
was in Germany in 1938. Since then, 58 units have been
built to supply town gas and synthesis gas.
3. Wellman-Galusha
McDowell-Wellman Engineering Company of Cleve-
land produces Wellman-Galusha fixed-bed gasifiers. Unit
sizes range from 6.5 ft in diameter to 10ft in diameter.
Maximum gas production rates of 75 million Btu/hr are
possible in an agitator-equipped 10-ft unit; a turndown
capability of 7 percent of maximum output is claimed.
Operation of the Wellman-Galusha is similar to that
of a Lurgi except that the former is an atmospheric
pressure unit. Like the Lurgi, air-steam or oxygen-steam
may be used.
The first commercial agitator-type Wellman-Galusha
was built in this country in 1952. There are several
plants in the United States presently operating such
units or keeping them on standby in the event of natural
gas cutbacks.
4. Winkler
The Winkler is an atmospheric pressure, fluid-bed
gasifier marketed by Davy Power Gas, Inc., of Lakeland,
Florida. Units have been built ranging from 3 to 18 ft in
diameter. Maximum output from an 18-ft unit is about
500 million Btu/hr of fuel gas, with a turndown of 50
percent. Either air-steam or oxygen-steam may be used
as the gasification agent.
The first commercial Winkler gasifier was built in
Germany in 1926 and since then, 36 units have been
constructed in various locations throughout the world.
5. Riley-Morgan
The Riley Stoker Co. (Worcester, Massachusetts) is
developing a modified version (equipped with a stirrer)
of the Morgan fixed-bed producer that was widely used
409
-------
natural gas. Most such processes involve some type of
liquid sorption. followed by regeneration of the sorbent
to release H2 S for conversion to elemental sulfur in a
Claus plant.
Many of those processes arc not selective for H2 S
but also remolte substantial amounts of CO2, which is
not necessary, ilnd which may not be desirable since this
increases the required sorbent flow rates. Large amounts
of CO2 in the off-gas may also complicate Claus plant
operation. In ajdition, some of the processes operate at
below ambient temperatures, which would require
further cooling of the fuel gas. Others operate at eleva-
ted pressures, which would require compression of the
fuel gas.
A brief de!cription of a number of commercial pro-
cesses that should be considered for fuel gas desulfuriza.
tion is given below.
1. Amine Scrubbing Processes
Processes based on scrubbing with various types of
amine solutions have been used for several decades.
Modern versions include the Sulfinol, Alkazid, Sulfiban,
and Econaminl: processes. Main application of amine
scrubbing has teen for sweetening of sour natural gases
and treatment c.f refinery gases.
All of the amine processes remove, in addition to
H2 S, varying ar,ounts of CO2, Temperature of operation
ranges from ambient to about 1000 F. The acid gases
react chemical", with the sorbent to form loosely bound
salts. The spen': sorbent is usually regenerated by pres-
sure reduction, and steam heating in a stripping vessel.
. 2. Ph¥sical Absorption Processes
Physical absorption processes were developed pri-
marily to purif ( synthesis gases made by gasifying vari-
ous fuels, including coal. Examples of commercial proc-
esses presently in use and the absorbent employed
are: Rectisol (methanol), Purisol (N-methyl-2-pyrrol-
idone). Selexol (dimethyl ether polyethylene glycol),
and Fluor (proJ:ylene carbonate).
These proc!!sses remove both H2 S and CO2 and, in
some cases. NH 3. HCN, COS, tars. and higher hydrocar-
bons. All opel ate at temperatures near ambient or
below. Regener,ltion of spent sorbent consists of depres-
surization and heating.
3. Alkali ~)Grubbing Processes
Alkali scrubbing processes involve chemical reac-
tions between ilcid gases and various alkaline salts dis-
solved in water,
The Benfield and Catacarb processes employ potas-
sium carbonate in solution wit~ ;j catalyst and can oper.
ate at temperatues up to 2500 r . Both processes remove
large amounts of CO2 in addition to H2 S. They operate
best at elevated pressures (ca. 300 psig). and regenera-
tion consists of :,ressure reduction and heating.
The Shell ,lot K3 P04 process can also operate at
temperatures up to 2500 F. KJ P04 is more soluble in
water than K2 C03 resulting in lower liquor flow rates.
However, large amounts of steam are still required for
regeneration of sorbent.
The Vacuum Carbonate process employs Na2 C03
and is similar to the now obsolete Seaboard process.
Regeneration consists of heating under a vacuum of
about 25 in. Hg.
4. Direct Oxidation to Elemental Sulfur
Direct oxidation processes convert H2 S removed
from the gas stream directly to elemental sulfur without
the need for a Claus plant; they are selective for H2 S-a
desirable characteristic. Commercially available systems
include the Stretford, Giammarco-Vetrocoke. and
T akahax processes.
The Stretford process operates over a wide range of
pressures and in a temperature range of 500-1000 F.
Applications have included treatment of town gas (from
petroleum), coke-oven gas, and producer gas.
The process consists of scrubbing the gas with an
alkaline solution containing a vanadium salt along with
anthraquinone-disulfonic acid (ADA). H2S is converted
to elemental sulfur by the vanadic salt. which itself is
reduced to the vanadous form. The reduced liquor is fed
to oxidizers where the vanadium is restored to its vana-
die form via a redox mechanism with ADA. Air is blown
through the liquor to reoxidize the ADA and elemental
sulfur is separated by froth flotation. The froth is either
filtered or centrifuged, washed, and melted to produce
high-quality sulfur. Systems are available, if necessary, to
provide pretreatment of the raw fuel gas to remove trou-
blesome impurities, as well as purify the scrubbing
liquor.
The Giammarco-Vetrocoke process employs a scrub-
bing solution of sodium arsenates and arsenites. Absorp-
tion of H2 S occurs at temperatures in the range of
500 -2120 F and at pressures ranging from atmospheric to
75 atm. Regeneration of sorbent takes place in an air-
blown oxidizing column at 1000 F and atmospheric pres'
sure, simultaneously producing elemental sulfur.
The Takahax process has been employed exclusively
in Japan for treatment of coke oven gas, chemical plant
waste gases, and sewage fermentation gases. The absorb-
i ng solution contains sodium, lA-naphthaquinone,
2-sulfonate as catalyst. and sodium carbonate. Absorbed
Hz S, in the form of sodium bisulfide, is oxidized by the
catalyst to precipitate elemental sulfur. The solution is
regenerated by contacting with air. Both absorption and
r!!generation occur under ambient conditions,
Economics of Coal Gasification
A detailed discussion of the costs involved in pro-
ducing a clean fuel gas from coal is beyond the scope of
this paper. To the authors' knowledge, no independent
410
-------
~rior to 1940. Two unit sizes of 12-ft and 18-ft in
diameter will be offered; predicted output of the larger
unit is about 350 million Btu/hr, when oxygen-blown.
Maximum operating pressure will be 40 inches of water.
Present company plans call for extensive testing of a
'prototype prior to marketing.
6. Kellogg
The Kellogg Co. (Houston) is prepared to market a
stirred, fixed-bed gasifier that operates at pressures rang-
ing from 10 to 20 psig; it is otherwise similar to a
Wellman-Galusha. Unit sizes range from 6 ft to 12 ft in
diameter. Maximum output of the largest unit is about
150 million Btu/hr with a turndown to 25 percent. The
gasifier may be either air-blown or oxygen-blown.
7. Wilputte
The Wilputte Corp. of Murray Hill, N.J., is offering
a version of the Chapman gas producer that was once
popular in the United States. Several such units are still
operating at a site in eastern Tennessee.
The Wilputte is an atomospheric-pressure, fixed-bed
gasifier with a maximum internal diameter of 10 ft.4 in.
The unit is. equipped with a rotating grate for ash dis.
charge, and a rotating rabble arm to break up coal
agglomerates at the top of the bed. Maximum output of
an air-blown unit is about 25 million Btu/hr. Oxygen
may also be used.
8. Woodall-Duckham/Gas Integral
The WD/GI gasifier is a two-stage fixed-bed unit
offered by Woodall-Duckham (U.S.A.) Ltd. of Pitts-
burgh. Coal is distilled in a retort section, driving off
water vapor, oil and tar, and volatiles. Offtake tempera-
ture of the gas from this section is about 2500 F. The gas
(referred to as top gas) is treated downstream to remove
tar and oil.
Coke produced in the retort is gasified in a lower
section as in conventional single-stage gasifiers. Gas pro-
duced in the lower section (clear gas) exits at 1,2000 F,
passes through' a cyclone, and is combined with the top
gas, resulting in a mixture of raw fuel gas at 7000 F.
Maximum output from a standard 12-ft-diameter
unit is about 80 million Btu/hr. A turndown to 25 per-
cent of maximum is possible. WD/GI gasifiers have been
in service for over 30 years in Europe, South Africa, and
Australia. Approximately 100 units have been con-
structed.
9. Applied Technology
Both a single-stage and a two-stage gasifier are avail-
able through Applied Technology Corp. of Pittsburgh.
Applied Technology is under license to the British firm
Wellman-Incandescent, Ltd.
The two-stage system is very similar to the WD/GI
gasifier described above. The single-stage unit is
equipped with an eccentric revolving grate for ash dis-
charge, but has no stirring device. The generator sections
of both types are identical. Unit sizes range from 4.5 to
12 ft in diameter. The output of 12-ft-diameter units is
claimed to be 97 million Btu/hr. Twenty-two units have
been built in South Africa since 1964. Four more are
presently under construction.
Fuel Gas Cleanup
Regardless of the process used to make fuel gas
from coal, it is likely that the gas will have to be cleaned
to meet environmental standards, end-use requirements,
or both. The major impurities present in these gases will
be hydrogen sulfide, particulates, and in the case of
fixed-bed processes, possibly condensable tars and oils.
Compared to stack gas scrubbing to remove sulfur
dioxide from the products of complete combustion,
there are several advantages in desulfurizing raw, coal-
derived fuel gas. The sulfur compounds, mainly hydro-
gen sulfide, are present in higher concentrations-the gas
has not been diluted with the final combustion air. The
volume of gas to be treated is only a fraction of the
volume of the final combustion products; hence, smaller
equipment sizes can be used. Finally well-proven tech-
nology exists for removal of H2 S from gases as com-
pared to S02 stack gas processes, which are still ex-
periencing technical difficulties.
All commercially available sulfur removal systems
. 0
operate at temperatures below about 250 F. Off-gas
temperatures from fixed bed gasifiers are generally in the
range of 8000-1,2000 F, and may be much higher for
fluid-bed or entrained-flow gasifiers. Thus, the gas will
have to be cooled substantially prior to final cleanup.
This can be accomplished in a waste-heat boiler- The
steam so generated can be added to the distribution
mains of the industrial complex, or used, in part, in the
gasification process.
In some cases, the raw fuel gas will contain tars and
oils that will begin to condense from the gas in th.e waste
heat recovery system. Tars removed at this point could
possibly be reinjected into the gasifier, burned as boiler
fuel, or sold as a byproduct.
Gas temperature after heat recovery will be in the
neighborhood of 3500-4000 F. The gas can be cooled
further by water scrubbing, which will remove any
remaining tar and particulates. Venturi scrubbers will
provide more than adequate separation of such impuri-
ties. Water used for scrubbing will be filtered (to remove
condensed tars and particulates), cooled, and recircu-
lated to the scrubber.
Sulfur in the raw fuel gas will be present chiefly as
hydrogen sulfide. Numerous processes have been devel-
oped for removing H2S from various gases including
coke oven gas, refinery gas, synthesis gas, and sour
411
-------
analysis of t:,e various commercial processes has, as yet,
been published.
Inasmuch as estimates of capital and operating costs
relJorted by manufacturers are not on a common basis,
reporting of specific figures here without detailed expla-
nation would be unfair, and perhaps meaningless. How-
evp.r. to give .3n approximate indication of costs involved,
the ranges 0'; important cost factors as given in the man.
ufacturers' rrcent literature are discussed.
Installed costs for gasification equipment. excluding
sulfur remova~, range from $8,000 to $23,000 per 106
Btu/hr of fuel gas. with an average of $15,300. These
figures, in sOlne casp-s. include particulate removal and/or
integral was1'e heat boilers. Installed costs, including
desulfurization equipment. range from $36,000-$51,000
pe,. 106 Btu/hr. with an average of $45,200. The invest-
ment in sulfur removal equipment could, therefore.
easily be the largest cost associated with a coal gasifica-
tion facility.
Manufac',:urers' estimates of operating costs are
givp.n in terms of dollars/106 Btu of fuel gas. However,
these figures are highly sensitive to the values assumed
for coal price!:.
To obtai.1 a somewhat clearer comparison, produc-
tion costs w£re converted to incremental costs per 106
Btu of fuel g.IS, relative to the cost of the coal gasified.
On this basis. incremental costs of clean fuel gas ranged
from $1.10 tD $1.51 per 106 Btu. with an average of
$1.24. Thus. for coal costing $30 per ton (12.500
Btu/lb), this would result in a clean fuel gas cost of
about $2.44 ICeI' 106 Btu.
No attenlpt has been made here to distinguish
between prOCl~SS costs based solely on air-steam gasifica-
tion or solely on oxygen-steam gasification. Suffice it to
say that the costs for producing either a desulfurized
low-Btu gas or a medium-Btu gas fall within the ranges
quoted above. !t is not apparent from the available data
as to which fu::1 could be produced more cheaply.
While thE' initial reaction may be that the use of
oxygen in a .}Bsification system would lead to higher
production ccsts, the opposite may be true in certain
situations. Wlti!e the investment in an onsite oxygen
plant is significant. power requirements for driving the
compressors lT~ay be met. in whole or in part. by use of
waste heat frcm the gasifier. Use of oxygen in place of
air should increase the maximum throughput of a given
gasifier by 50 to 100 percent; hence. smaller. or fewer.
gasifiers are nl!eded for a given fuel requirement. like-
wise, smaller desulfurization eq<.Jipment is necessary.
Finally, the cost savings to bE' realized in utilizing the
higher quality medium-Btu gas. r.::ther than low-Btu gas,
may significartly shift tne economics in favor of an
oxygen-based system (see Properties of Coal-Derived
Fue/ Gases, end of C,).
Properties of Coal-Derived Fue/ Gases
A. Chemical Composition
The chemical compositions of a number of different
fuel gases manufactured from various coals (and coke)
are given in table 2. These gases were produced by vari-
ous commercial gasifiers operating at near-atmospheric.
pressure. Since a purpose of this study was to draw some
general conclusions about combustion properties of
coal-derived fuels, the gases indicated were selected at
random from data reported in the literature.
The main combustible constituents in these gases.
CO and H2. result from the chemical reactions* involved
in the gasification process:
C + O2 = CO2 (1)
C + COz = 2CO (2)
C + Hz 0 = CO + Hz (3)
CO + H20 = CO2 + Hz. (4)
Reaction (1) is exothermic and supplies the heat
necessary for the actual gasification reactions (2) and
(3), which are endothermic. Reaction (4) is the so-called
water gas shift reaction. . .
The trace amounts of methane found in these gases
probably result. from devolatilization of the coal. Pro-:
duction of methane by other chemical reactions is gener-
ally insignificant at low pressures.
The heating values of the fuel gases are related to
the amount of nitrogen in the gases; the latter is. in turn,
related to the degree of oxygen enrichment of the gasify-
ing agent. The data plotted in figure 2 (taken from ref.
11) show that there is a fairly uniform increase in fuel
heating value. with increasing oxygen level in the oxi-
dant stream.
The presence of fairly large amounts of carbon
monoxide in coal-derived gases probably makes these
fuels undesirable for residential or commercial use. How-
ever. industry has had considerable experience in the
handling of toxic gases; hence, the use of a CO-rich fuel
should present no special problems.
B. Material Flows Involved in Combustion of Coal-
Derived Fuel Gases
The quantities of fuel, air, and combustion products
involved in the combustion of coal-derived gases were
evaluated. All computations were based on the assump-
tion that the various fuels are burned with their stoichio-
metric amounts of air. of the following composition:
. A more detailed discussion is given in reference 10.
412
-------
Percent
N2
O2
CO2
Ar
H20
77.10
20.69
0.03
0.92
1.26
It is of interest to compare the values calculated for
the coal-derived fuels with the corresponding quantities
for methane, carbon monoxide, and hydrogen. Hence,
data for these latter three fuels, for the case of stoichio-
metric combustion, are given in table 3. For all intents
and purposes, methane may be assumed to be equivalent
to natural gas.
The relationship between fuel heating value and
volume of fuel required per million Btu is shown in
figure 3. This relationship is purely mathematical in
nature, and is given by
SCF fuel = 106/HHV.
(5)
Thus, in order to maintain a given thermal input to
a process, approximately three times as much medium-
Btu gas (ca. 300 Btu/SCF) as natural gas is required; for
a low-Btu gas (ca. 150 Btu/SCF), the ratio increases to
more than six.
The increased fuel flow rate required for low- and
medium-Btu gases has serious implications when retrofit-
ting of existing installations is considered. Since pressure
losses in piping and equipment are approximately pro-
portional to the square of the fluid velocity, it is readily
seen that losses incurred in utilizing the coal-derived
gases in an existing system can be substantial.
Problems associated with distribution of the fuel
might be overcome by either enlarging the fuel mains or
pressurizing the system. To maintain similar velocities in
mains presently delivering natural gas at a given pressure,
an increase in diameter by a factor of about 1.8 would
be necessary for medium-Btu gas; for low-Btu gas, the
factor is about 2.6. Such increases in duct size might not
be feasible in certain installations because of space limi-
Table 2. Fuel gases produced by steam/air, steam/enriched-air,
or steam/oxygen gasification of coal and coke
Percent by volume
Heating valuea
Fuel Gas No. CO H2 CH4 C02 N2 02 (HHV) Btu/SCF
1 22.0 14.0 1.0 7.0 56.0 0.0 126.1
2 31.0 9.3 0.7 3.6 55.4 0.0 136.7
3 37.8 11.3 0.4 10.8 39.5 0.2 162.0
4 28.6 15.0 2.7 3.4 50.3 0.0 167.7
5 44.6 22.6 0.4 10.8 21.0 0.6 220.5
6 48.8 29.1 0.4 12.0 9.0 0.8 255.0
7 53.3 30.9 0.4 12.8 2.3 0.3 275.3
8 48.2 35.3 1.8 13.8 0.9 0.0 287.3
9 56.6 37.3 0.0 5.0 1.1 0.0 302.7
a
SCF at 60° F, 1.0 atm.
413
-------
JOO
...
... i!50
It)
'\.
:;)
....
~
iii
~
~
::; i!00
:;)
Q
o
~
II.
...
o
100
i!O
(')
'10 60 ao
O~VG~~ I~ OI(IDA~T STR~AM, PERCENT
100
Figure 2. Effect of oxygen level on heating
value of product gas (gasification of coke).
16000
111000
::;)
...
en
~ 12000
()
-
.,J
~ 10000
~
~ 8000
~
::;)
"-
"-
()
"-
v
II)
6000
'4000
2000
o
. 0 cOO '400 600 800 1000
HHV OF FUEL GAS. BTU / SCF
Figure 3. Volume of fuel per million Btu vs.
heating value of fuel gas.
Table 3. Stoichiometric combustion data from CO, H2, and CH4
tations. Use of pressurized mains could solve this prob-
lem, but may not be desirable because of safety consid-
erations.
Use of 10w.Btu gas in burners originally designed for
natural gas (and/or oil) is probo>h!y not possible. It has
been reported, however, that a medium.Btu fuel could
be burned in utility boilers with only minor modifica-
tions to the existing equipment (ref. 12).
An analysis of problems associated with combustion
--
Fuel H,~ating Value (HHV) SCF Air/ SCF/million Btu
Btu/SCF SCF Fuel Fuel Air Comb. prod.
----
CO 321 2.42 3115 7527. 9279
H2 324 2.42 3084 7453 9090
CH4 1010 9.67 990 9570 10560
of coal-derived gases using various conventional indus-
trial burners is given in reference 18. The overall conclu-
sion of the authors was that there should be no difficul-
ty in burning medium-8tu gases. However, flame stabili-
ty problems may be encountered in combustion of low-
8tu gases.
In the past, a number of different types of burners
have been successfully used to burn low-Btu gas (refs.
7,8). Undoubtedly, such burners can be improved to
414
-------
~eet modern standards. Again, the real problem is one
~f space limitations that may occur in existing systems;
room may not be available to accomodate the larger
burner diameters necessary for low-Btu gas firing at a
desired thermal input.
The relationship between fuel heating value and
volume of air required for stoichiometric combustion of
coal-derived gases is given in figure 4. It may be seen that
a linear correlation exists; the data were fitted by least-
squares to obtain the expression
SCF AIR/SCF FUEL =
7.38 x 10-3 HHV + 0.035.
(6)
An expression relating the volume of air per million
Btu of fuel fired is obtained by simply combining equa-
tions (5) and (6):
SCF AIR/(106 Btu) =
7,380 + 35,OOO/HHV.
Equation (7) is plotted in figure 5 along with the
data pertaining to the various fuel gases given in table 2.
2.5
'07H'
I/)
~ 2.0
o
-J
'"
::)
...
...
o
...
....
I/)
"'-
~
<0[
...
o
10.
IJ
I/)
o
II
1.5
o
II
o
1.0
o
0.5
100
200 ~oo
HHV OF Fun GAS BTU/ SCF
.,00
Fig4re 4. Volume of air required to burn, 1 ft3
of fuel gas vs. heating value of fuel gas, 100
percent stoichiometric air.
~
o
,:J
-J
~
/
~ 7500
<0[
10.
o
10.
....
I/)
o
o
CO
o
o 0 .
04-H2
o
o
7000
100
200
~OO
.,00
HHV OF FUEL GAS. BTU/SCF
Figure 5. Volume of air per million Btu of fuel.
vs. heating value of fuel gas, 100 percent
stoichiometric air.
(7)
Figure 5 indicates that, for all of the gases consid-
ered, the air requirements per million Btu of fuel fall in
the range of 7,400 to 7,900 SCF. Since the air required
for burning a million Btu of natural gas is about 9,600
SCF, no problems are anticipated in regard to air supply
when substituting a coal-derived gas for natural gas in an
existing system.
The variation in volume of combustion products
resulting from burning different fuel gases is depicted in
figure 6. Volumes associated with 10w.Btu gases below
200 Btu/SCF are substantially greater than that resulting
from natural gas (CH4) combustion. This is due largely
to the much greater quantities of fuel required when
burning the low heating-value gases as compared to
natural gas; as was pointed out, the air requirement is
actually less for the coal-derived gases.
The increased flow of combustion products from
the low-Btu gases causes retrofit problems similar to
those enumerated earlier regarding fuel flow rates. In
existing installations, there will be greater velocities and
draft losses than when burning natural gas at a given
firing rate. For gases above about 200 Btu/SCF, there
appears to be no problem.
As the heating value of the fuel gas decreases, the
ratio of the amount of combustion products to the
amount of combustion air increases. This has implica-
tions in regard to downstream heat recovery if only air
preheating is employed; the incoming air can only ab-
sorb a fraction of the sensible heat of the combustion
products and, hence, overall process efficiency suffers. A
much larger percentage of the energy in the flue gas can
be recovered if both air and fuel are preheated, and this
415
-------
/6000
::)
ti)
~
o
.,J
.,J
~
"
~
u
::)
o
o
ex
Cl
:t
o
t::
U)
::)
~
~ 10000
o
v
"-
o
IA.
v
U)
8000
o
CH..
~
"
. "
.~. CO ".".///
. ".
..... -
""".......---......"'-
a
200
600
800
1100
1000
HHV OF FUEL GAS, BTU / SCF
Figure 6. Volume of combustion products per million Btu of fuel V5.
heating value of fuel gas, 100 percent stoichiometric air.
can be done in suitably designed systems. However, the
(actual) volume increase due to heating of the fuel may
exacerbate pmblems related to fuel flow to the burners.
C. Combustion and Heat Transfer Properties
An evalu3tion of several quantities related to the
combustion af\d heat transfer characteristics of the fuel
gases listed in table 2 was conducted. In some cases, all
of the gases were considered. In other cases, only select-
ed examples were analyzed.
Figure 7 shows the relationship between the calcu-
lated adiabatir; flame temperature of the fuel gases and
their heating Halues. For the purposes of these computa-
tions, it was assumed that each of the gases was burned
with its stoichiometric amount of air, with both fuel and
air initially at 77° F (298.15 K). The calculations were
carried out using the PERC-Multiphase Chemical Equi-
librium Progr,Jm developed a' the Pittsburgh Energy
Research Cent~r (ref. 131. *
The nearly linear relationship between flame tem-
perature and heating value is not surprising. The main
combustible constituents in these gases are H2 and CO,
which exhibit nearly identical heating values and flame
temperatures. As we go from low. to medium.Btu gases,
we are simply approaching a mixture of H2 and CO.
The data presented in figure 7 were fitted by least
squares, yielding the expression:
T flame = 4.20 HHV + 2,464.
(8)
This equation should provide a reasonable first
approximation to the flame temperature of any coal-
derived gas ranging in heating value from 100 to 300
Btu/SCF, provided that the CH4 content is not much
greater than the percentages indicated in table 2.
It should be noted that the flame temperature for
natural gas (CH41. with a heating value of 1,010
Btu/SCF, is 3,517° F. This is some 2000 F below that
calculated for fuel no. 9, table 2, with a heating value of
about 303 Btu/SCF. The reason for this apparently I,
anomalous situation is that it requires a considerably
*The auth ),s acknowledge the aid of Francis E. Spencer
and James .C, Fendrie of the Energy Conversion section, Pitts.
burgh Energy Rnsearch Center, in making these computations.
416
-------
~800
co
Hr
0"'" ~600
'"
It
;:)
...
-------
Q
oq
o
-J 0.8
o
I-
Q
\oj
It
It
\oj
~ 0.6
C:
~
~
M£DIVM-BTV GAS (JOJ BTV/.>C
...
:)
It
~ 0."
..J
oq
~
Ck
\oj
:t
0-
I&. 0.2
o
C:
()
l-
V
~
I&. 0.0
1000
2000 3000
FLU£ GAS T£MP£RATUR£, of
'1000
Figure 9. Altailable heat vs. flue gas temperature,
100 percent stoichiometric air.
heat transfer to the load and the same exit temperature
o .
of 2,000 F. about 15 percent more Btu of natural gas
are required than of medium-Btu gas. Under the same
conditions, a :)out 33 percent more Btu's of low-Btu gas
are needed. I:or a flue gas temperature of 2,500° F, 26
percent more methane and 64 percent more low-Btu gas
are required than medium-Btu gas.
The above analysis considers only the thermody-
namics of the heat transfer process, and offers no insight
regarding the actual rates of heat transfer involved. In
order to gain such insight, the rates of heat transfer
achieved when burning various fuels under idealized con-
ditions were evaluated. The conditions selected were as
follows; (1) all fuels were burned with their stoichio- .
metric amounts of air, with no preheat of either air or
fuel; (2) combustion took place in a well-stirred, cubical
furnace bounded on all sides by a black surface main-
tained at 1,000° F; (3) the size of the furnace was such
that nonluminous radiation was the major mode of heat
transfer, convective heat transfer being negligible (a
me:m beam I;mgth of 10.0 ft w:!s chosen, which corre-
sponds to an actual furnace s;ze of about 17.5 ft on a
side); (4) there were no heat losses from the system.
Emissivites and absorptivities were evaluated using
the methods described in reference 14. The emissivities
corresponding to several different coal-derived gases and
methane are displayed in figure 10 as a function of tem-
perature. It should be noted that the values of emissivity
associated with the low-Btu gases are substantially less
than those for methane combustion products. On the
other hand, the emissivity curves for the medium-Btu
gases fall quite close to that for methane.
The values of absorptivity were found to fall in the
range 0.4-0.5 in all cases considered and were a weak
function of temperature. Hence, the absorptivities had
only a minor affect on the calculated heat transfer rates.
Figure 11 depicts the efficiency of radiant heat
transfer as a function of furnace load for a low-Btu gas, a
medium-Btu gas, methane. and No.2 fuel oil. This plot
is analogous to the plot of available heat curves (figure
9) discussed previously, but with one principal differ-
ence. In order to achieve a given heat transfer rate in the
idealized furnace, it is necessary for gases with low emis-
sivity to maintain a higher temperature in the furnace
than for gases with high emissivity. As a result, the lesser
radiating gases exit the furnace at a higher temperature
and their utilization efficiency declines.
In comparing figure 11 with figure 9, one sees that
the relative positions of the methane and medium-Btu
gas curves remain essentially unchanged because of the
similar emissivities involved in the calculations. The
0.4
OXYGCN- BLOtIIN{
GASIFI£R
M£:THAN£
I 303 BTU/ SCF
275 BTU/5CF
0.3
)0.
...
~
Vi
III
i
141
0:2
AIR-BLOVIN (
GASIFI£R
0./
1500
2000 2500
T£MP£RATUR£, of
3000
Figure 10. Emissivities of combustion products
of coal-derived fuel gases and methane. Pres-
sure = 1.0 atm; mean beam light = 10.0 ft.
100 percent stoichiometric air.
418
-------
0.6
iiI
~
(;I 0."
"-
Q
<{
o
...J
(;I
0.2
o
50
100
-,
QLOAD' BTU/HR )( 10
Figure 11. Efficiency of radiant heat transfer
vs. thermal load, 100 percent stoichiometric
air.
curve for No.2 oil is now higher than the rest because
the emissivities used were corrected to allow for lumi-
nosity, as suggested by reference 15.
The curve for the low-Btu gas, however, now falls
substantially lower than the others. This is a direct result
of the much lower emissivities associated with combus-
tion of this fuel. To maintain a heat transfer rate of 65 x
106 Btu/hr (about midrange in figure 11), it would be
necessary to use 227 percent more Btu of low-Btu gas
than of medium-Btu gas.
Figure 12 shows the general trend in utilization effi-
ciency of various gaseous fuels as a function of fuel heat-
ing value, at fixed furnace load. While the region be-
tween 300 and 1,000 Btu/SCF was not examined, the
broken line is probably a fair representation of the
actual situation. The curve shows clearly that,' for the
case considered here, a medium-Btu gas of about 300
Btu/SCF is nearly the optimal gaseous fuel in terms of
heat transfer properties. This figure also indicates that a
coal-derived fuel gas of about 220 Btu/SCFwould exhib-
it heat transfer characteristics identical to those of nat-
ural gas. In view of figure 2, such a fuel could be manu-
factured using enriched air containing about 50 percent
oxygen; this is equivalent to an oxygen/air ratio of 0.6
SCF 02/SCF air.
The principal conclusions to be drawn from this
analysis are: (1) coal-derived .fuel gases with heating
values in the range 200 to 300 Btu/SCF should provide
utilization efficiencies equal to, or better than, that of
natural gas; (2) the heat-transfer properties of gases
below about 200 Btu/SCF are such that for some appli-
0.6
~
cr 0.4
'\
a
~
...
cr
(. " - - - - - - - - - - - - - - CIf,
----~
0.2 -
o
.
ZOO
I
I
I
I
1000
.
400 600
HHV OF FUEL GAS, BTU/SCF
800
Figure 12. Efficiency of radiant heat transfer
vs. heating value of fuel gas, 100 percent
stoichiometric air QLOAD = 50x 106 Btu/hr.
cations it may be more economical, in terms of greatly
reduced fuel consumption, to use a medium-Btu gas; this.
would appear to be particularly true in high-temperature
processes that do not employ heat recovery.
Discussion
It appears, from the preceding analysis, that a clean
fuel, suitable for industrial use, can be manufactured
from coal using existing, well-proven technology.
Medium-Btu gases, made with oxygen or oxygen-
enriched air, should exhibit combustion and heat-
transfer properties equivalent to or better than those of
natural gas; their use in existing facilities would require
relatively minor modifications of equipment. Low-Btu
gases, resulting from steam/air gasification of coal, could
also be used, but somewhat less effectively; more exten-
sive modifications of existing equipment would be neces-
sary. In a properly designed system, however, even the
low-Btu fuels would provide satisfactory performance.
A question remains concerning the optimum
approach to utilizing coal gasification technology. A
large company may find it advantageous to construct its
own gasification/cleanup facility for onsite use. The
byproducts of gasification-steam, and possibly oxygen
and nitrogen used in the process (and often needed in
manufacturing operations)-could be absorbed by the
operator or sold to nearby plants.
A very attractive arrangement would be an. indus-
trial park served by a central gasification facility. Fuel
, gas, process steam, and steam for space heating could be
distributed to a number of closely situated plants. Cost
419
-------
of the opera:ion would then be shared by the resident
industries.
It appears that medium-Btu gases can be transmitted
economically over considerable distances (refs.
16,19,20). This presents the possibility of a mine-mouth
gasification plant distributing fuel gas to industries with-
in a radius of perhaps 50 to 100 miles, much as mer-
chant gas pic nts did in the past. Such an arrangement
offers the economic advantages of large-scale operation.
It has be,m pointed out by Mitsak and Kamody (ref.
16) that if al of natural gas consumed by industry were
displaced by an alternate fuel, the rate of discovery of
new reserves (under present economic conditions) is
almost sufficient to meet the needs of residential and
commercial users. Assuming an annual population
growth of 1.1 percent, the existing natural gas supply
system would be adequate for about 80 years.
While t01al conversion of industry to coal-derived
gasHs is probably not practicable, let us examine the situ-
ation if such were the case. Total annual consumption of
natural gas b't industry is roughly equivalent to about
550 million 10ns of coal coverted to low- and medium-
Btu fuel gases. This is nearly equal to the amount of coal
presently produced annually (600-650 million tons). It is
believed, however, that doubling of the annual output of
coal could bo effected, with some difficulty, by 1985
(ref. 17).
The gasifcation hardware required in this hypothet-
ical situation would probably amount to 2,000 to
12,000 units of present design, depending on the sizes
utilized. It should be remembered that, at one time,
11,000 gasification units were in operation in the United
States, althou~h these were of smaller capacity and were
without desul':urization facilities.
While thl! magnitudes of capital and resources in-
volved in a t01al conversion of industry to coal-derived
fuels appear immense, such a possibility is probably not
beyond our rational capability under the proper set of
circumstances The point to be made, however, is that,
realistically, we have the technology and resources avail-
able to effect a very substantial displacement of natural
gas from the illdustrial sector. The freeing of this fuel for
more appropriate end uses could extend its use many
decades beyolld the point of extinction now predicted.
Conclusions
Any industry, faced with curtailments in its supply
of natural gas, would do well to seriously consider the
use of coal-dErived fuel gases. ~uch fuels can be manu-
factured using currently available technology.
Medium.f;tu gases (':J. 300' Btu/SCF) from coal
could be substituted for natural gas in many applications
without major changes in equipment. Low-Btu gases (ca. I
150 Btu/SCF) could be effectively utilized in properly
designed new installations.
The cost of coal-derived fuel gases will undoubtedly
be greater than what American industry is accustomed
to paying for natural gas and oil. However, cost may be
of secondary importance when the choice is between a
more expensive, readily available fuel or no fuel at all.
An early acceptance of coal gasification technology
by industry could have a profound impact on our energy
supply situation in the years ahead. Large amounts of
natural gas would be freed solely for residential and
commercial use. which could conceivably extend the
availability of this fuel well into the next century.
REFERENCES
1. W. G. Dupree, Jr., and J. A. West, "United States
Energy through the Year 2000", U.S. Department
of the Interior, December 1972.
2. D. W. Locklin, H. H. Krause, A. A. Putnam, E. L.
Kropp, W. T. Reid, and M. A. Duf,fy, "Design
Trends and Operating Problems in Combustion
Modification of Industrial Boilers", report ROAP
No. 21 ADG.83, prepared for the Environmental
Protection Agency, April 1974.
3. R. H. Essenhigh, "Optimum Efficiencies of Fur-
naces: Setting the Targets," Conference on Fuel
Efficiency in Industry, Pennsylvania State Universi-
ty, April 8-12, 1974.
4. A. S. Powell, "Energy Use Efficiency Costs and
Values," Conference on Fuel Efficiency in Industry,
Pennsylvania State University, April 8-12, 1974.
5. L. Shnidman, ed., "Gaseous Fuels," second edition,
American Gas Association. New York, 1954.
6. J. F. Foster, and R. J. Lund, ed., "Economics of
Fuel Gas from Coal," McGraw-Hili Book Co., New
York, 1950,
7. N. E. Rambush, "Modern Gas Producers," Benn
Brothers Limited, London, 1923.
8. R. T. Haslam, and R. P. Russell, "Fuels and their
Combustion," McGraw-Hili Book Co.. New York,
1926.
9. B. J. C. van der Hoeven, in Chemistry o( Coal Utili-
zation, John Wiley and Sons, Inc., New York. 1945.
10. H.H. Lowry, ed., Chemistry o( Coal Utilization,
supplementary volume, John Wiley and Sons, Inc.,
New York, 1963.
11. R. F. Mitchell, "Oxygen-Steam Producer Blast,"
presentation to the Chemical Engineering Section at
the Annual Conference of the Chemical Institute of
Canada, June 25, 1946.
420
-------
~ 2. A. F rendberg, "Performance Characteristics of
Existing Utility Boilers when Fired with Low-Btu
Gas," Proceedings of the Conference on Power Gen-
eration-Clean Fuels Today, Electric Power Research
Institute, Monterey, California, April 8-10, 1974.
13. F. E. Spencer, Jr., A. A. Orning, and D. Bienstock,
"Applications of a High-Temperature Chemical
Equilibrium Program to Problems in Coal Combus-
tion," presentation at the American Institute of
Chemical Engineers Meeting, New York, November
29, 1972.
14. W. H. McAdams, Heat Transmission, third edition,
McGraw-Hili Book Co., New York, 1954.
15. H. C. Hottel, and A. F. Sarofim, Radiative Transfer,
McGraw-Hili Book Co., New York, 1967.
16. D. M. Mitsak, and J. F. Kamody, "Koppers-
Totzek: Take a Long Hard Look," Symposium on
Coal Gasification, Liquefaction, and Utilization,
University of Pittsburgh, August 5-7, 1975.
17. R. E. Bailey, "Coal as a Key to U.S. Energy Poli-
cies,". Symposium on Coal Gasification, liquefac-
tion, and Utilization, University of Pittsburgh,
August 5-7,1975.
18. D. G. Ball, G. Smithson, R. Engdahl, and A. Put-
nam, "Study of Potential Problems and Optimum
Opportunities in Retrofitting Industrial Processes to
Low and Intermediate Energy Gas from Coal," re-
port ROAP No. 21 ADD-30, prepared for the Envi-
ronmental Protection Agency, May 1974.
19. National Research Council, Evaluation of Coal Gasi-
fication Technology. Part /I, Low- and Inter-
mediate-Btu Fuel Gases, 1973.
20. T. F. Edgar, and J. T. Richardson, "Resources and
Utilization of Texas Lignite," report to the Gove-
rnor's Energy Advisory Council, State of Texas,
November 15, 1974.
421
-------
THE EPA R&D PROGRAM IN WASTES-AS-FUEL:
AN OVERVIEW FOCUSING ON PROCESS ENVIRONMENTAL/ENERGY IMPACTS
George L. Huffman*
Abstract
The USEI'A is conducting a broad-based program in
research and d~!felopment in the area of waste utilization
as alternative /uel sources. Major waste-as-fuel technolo-
gies and meth ods are being explored to determine their
environmental and energy impacts, and to develop tech-
nology to control undesirable environmental impacts.
Major process4~ include: use of waste as a supplemen-
tary fuel in 16 rye pulverized-coal boilers, in other types
of coal-fired boilers, in oil-fired boilers, and in industrial
and smaller i.1stitutional boilers; direct conversion of
wastes to elecl'ricity via a gas turbine; and the pyrolytic
conversion of wastes to fuel gas and fuel oil. A descrip-
tion of these .orDcesses is presented, as is the status of
those R&D plojects which encompass them. The envi.
ronmental aspfJcts of these processes are summarized,
along with prc;ections of their energy conservation attri-
butes.
INTRODUCTION
Solid wa!;te is generated from most all activi-
ties: from the farm to the feedlot; from the home' to
businesses; frcm mineral extraction operations to all
types of indus:rial manufacturing operations. The use of
solid waste as ,) fuel for energy generation has been prac-
ticed for some time in Europe. Its practice in this coun-
try. however, is essentially just beginning. I n the United
States, numerous processes are under development
which convert the energy contained in solid waste into
steam, oil, gas, and electricity.
In this paper, the amount and availability of the
heat and enemy content of the Nation's combustible
. solid waste wil] be briefly discussed. A number of proc-
esses under EPA development to convert otherwise-
polluting solid waste into usable energy forms will also
be considered, O1long with a brief summary of the envi-
ronmental characteristics of such conversion processes.
The discussion will begin with a review of EPA's legisla-
tive authorities in the waste-as-fuel area, and will con-
clude with an \)verview of EPA's current environmental
protection R&I> program in the area.
.Chief. Fuels Technology Subp'')gram, Environmental Pro-
tection Agency, Industrial Environmen+al Research Laboratory,
Cincinnati, Ohio.
E.PA'S LEGISLATIVE MANDATES
IN WASTES-AS-FUEL
The administrator of EPA is mandated under the
1965 Solid Waste Disposal Act, as amended by the 1970
Resource Recovery Act (PL 91-512), to perform re-
search and demonstrations to develop and to apply new
and improved methods for processing and recovering
both materials and energy from solid wastes. The act
also requires the study of adverse health and welfare
effects of the release into the environment of materials
present in solid wastes, as well as methods of eliminating
such effects. The act authorizes EPA to issue contracts
and grants for research and/or demonstrations of tech-
nologies in the waste-as-fuel area to enable appropriate
response to these mandates.
The administrator of EPA is mandated under the
Clean Air Act and the Federal Water Pollution Control
Act to establish performance standards for existing and
new sources of pollution based on available technology.
Also, specific provisions within each of these acts require
the administrator to periodically review existing guide-
lines and revise information regarding best available con-
trol technologies, the costs of achieving standards, and
other associated impacts. Under both acts, the adminis-
trator is charged with establishing a national research
and development program for the prevention and con-
trol of pollution. The Clean Air Act authorizes EPA to
do research on characterization of the air pollution prob-
lem associated with a new or existing source, such as a
waste-conversion-to-fuel facility, and authorizes R&D on
control technology aimed at lessening the environmental
impact. The Federal Water Pollution Control Act and
the Solid Waste Disposal Act authorize EPA to ~onduct
R&D aimed at lessening the groundwater pollution
potential associated with the disposal of solid wastes
onto the land-wastes whose energy value is literally
thrown away.
WASTES AVAILABLE FOR FUEL RECOVERY:
WHAT ENERGY IMPACT IS POSSIBLE
It has been estimated that about 4.5 billion tons
(4.1 trillion kilograms) of municipal, industrial, mineral,
and agricultural solid wastes were produced in this coun-
try in 1970 (ref. 1). Of this amount, about 13 percent
422
-------
~as been estimated to represent the dry, combustible
'raction. Table 1 shows the detailed breakdown of this
13 percent that represents the country's dry, combusti-
ble solid waste stream. This table summarizes data pre-
sented in a 1972 report to the Environmental Protection.
Agency by the International Research and Technology
Corporation (ref. 2). It shows that about 1,135 billion
pounds (517 billion kilograms) or about 570 million
tons of dry comblJstible solid waste were discarded in
1970 (refs. 2,3). By way of comparison, a Bureau of
Mines estimate arrived at a figure of 880 million tons
(800 billion kilograms) as the dry, organic waste genera-
Table 1. Quantity and fuel value of dry combustible
solid waste discarded in 1970 (refs. 2,3)
Waste source
Combustibles
discarded
(billion of pounds)
. Fuel value in quad-
rillions of Btu's
. (Btu x 1015)
275.5
2.177
Refuse and sewage--urban
generated waste
Household and municipal
Sewage solids
Commercial and institu-
tional
Manufacturing plant
waste
Demolition
Manufacturing and process-
ing waste
Wood related wastes
(SIC 24xx - 27xx)
Textiles and fabric
wastes (SIC 22xx - 23xx)
Nonfabric synthetic
materials
Food processing
(SIC 20xx)
Miscellaneous manufac-
turing (SIC 21,31)
Agricultural wastes
Animal wastes
( SIC 01 3x - 014)
Crop wastes
(SIC Ollx - 012x)
Forest and logging
residues
(SIC 19xx - 39xx)
Total combustible wastes
and fuel value
168.4
13.8(13f.)
62.1
1.348
0.083
0.496
23.5
7.7
0.188
0.062
54.5
0.436
51.3
0.6
0.411
0.005
0.9
0.011
0.008
1.5
0.2
0.001
805.1
5.913
413.4
340.0
2.779
2.720
51. 7
0.414
1 ,135. 1
8.526
Note - 1 pound = 0.4536 kilograms; 1 Btu = 252 calories.
423
-------
tion rate for 1971 (ref. 4). This latter figure represents
an energy equivalent of about 1.1 billion barrels (175
million cubic meters) of oil per year, or about 70 per-
cent of the oil .mported that year (refs. 4.5).
The fuel value associated with the 1970 waste
stream depictej in table 1 represents a yearly waste of
8.5 quadrillion Btu's (refs. 2,3). This is about 12 percent
of the total e 1ergy requirement of the United States,
which approximates some 72 quadrillion Btu's/year
(refs. 5,7). Nct all of this combustible solid waste is
readily availab e for use as a fuel, however. Although
urban.generated waste may qualify as a fuel because it is
available in rather concentrated form at appropriate
locations, othe' kinds of wastes from processing, manu-
facturing, or a!lricultural operations are often generated
in widely dispersed areas. One study (ref. 3) found that
only about 1.5 to 3.0 percent of the Nation's total ener-
gy requirement Gould be supplied by our solid waste due
to its geographical dispersion, assuming only minim;!1
waste transportation could be justified {the former
amount if 500-ton/day plants are economically justifia-
ble, the latter amount if 100-ton/day plants are worth
installing). The amount of waste transportation costs
that can be borne has increased substantially, however,
now that energy prices for conventional fuels have more
than quadrupled. Consequently, it may now well be that
substantially more than 1 to 3 percent of our energy
requirement ought to come from discarded waste materi-
als.
Table 2 is intended to put into perspective what 3
percent of the national energy requirement represents.
Three percent is about twice what we used in 1968 to
lig~t our homes and offices, and about 20 percent more
than what we used that year to air condition them (refs.
7,8). Another way of looking at what an "extra" 3 per-
cent in energy (from solid waste) might mean to the
country is the following: One study (ref. 9) estimated
recently that there would be an energy penalty of about
Table 2. Major end-uses for energy, 1968 data (refs. 7,8)
Type of use
Transportation (fuel; excludes lubes
and greases)
Space heating (residential, commercial)
Process steama (industrial)
Direct heata (industrial)
Electric drive (industrial)
Feedstocks, raw materials (commercial,
industrial, transportation)
Water heating (residential, commercial)
Air conditioning (residential, commercial)
Refrigeration (residential, commercial)
Lighting (residential, commercial)
Cooking (residential, commercial)
Electrolytic processes (industrial)
Total
Percent of total
U.S. energy
consumption
24.9
17.9
16.7
11.5
7.9
5.5
4.0
2.5
2.2
1.5
1.3
1.2
97.1
a
Includes some use for space heating, probably
enough to bring total space heating to about 20 percent.
424
-------
. percent associated with providing the pollution control
'measures of complete S02 and particulate control at
stationary sources, secondary level sewage treatment for
the whole country, waste heat dissipation with cooling
towers at power plants, and appropriate sanitary landfil-
ling of all of our municipal solid waste. Thus, energy
recovery from solid waste could more than offset the
energy required to operate these four significant pollu-
tion control processes. .
MAJOR WASTES-AS-FUEL PROCESSES
UNDER LONG-TERM EPA DEVELOPMENT:
ENVIRONMENTAL CHARACTERISTICS
EPA has been involved in furthering the develop-
ment of major waste-as-fuel technologies in response to
its legislative mandate to do R&D on methods which
dispose of waste materials in an environmentally accept-
able manner. Certainly one of the best "disposal" op-
tions available (best economically and environmentally)
is to treat the waste in such a fashion as to enable the
recovery of its energy and material resource values. That
is, why should we necessarily face the difficult environ-
mental control problems associated with the "throw-
away" options of improper land disposal or air-polluting
incineration of high-Btu wastes? Conceptually at least,
~oes it not make more sense rather to recover the energy
and economic value of these high-Btu wastes and to ef-
fect such recovery in an environmentally acceptable
fashion?
As early as 1967, EPA (actually its DHEW predeces-
sor) began development of several waste-as-fuel proc-
esses. Two of the major process options that began to be
researched at that time were the CPU-400 System and
the Horner and Shifrin Fuel Recovery Process (St.
Louis). Table 3 illustrates some of the salient features of
these two processes, along with similar data for two
other more recently initiated developmental efforts (the
Garrett and Monsanto Pyrolysis Processes). Each of
these processes will be briefly described-followed by a
description of some of the R&D projects begun within
the last year in the EPA Wastes-As-Fuel program.
The CPU400 System
Figure '1 represents a schematic of the CPU-400 Pro-
cess (refs. 3,10,11). This process uses prepared solid
waste to fire a conventional gas turbine for the produc-
tion of electricity. The solid waste preparation stage is
one in which the refuse is shredded, air classified, and
placed in a storage vessel prior to use. The combustion
system initially consisted of a fluidized bed combustor
and three stages of cyclonic inertial separators to remove
particulates from the gas prior to its expansion through
the turbine. The turbine compressor is used to provide
Table 3. Some resource recovery processes under development by EPA
EPA's grantee
(contractor)
Process
U.S.
tons per day (TPD)
of solid waste (SW)
trea ted
Baltimore, Md.
, (Monsanto Enviro-Chem)
San Diego County, Cal.
(Garrett Research)
St. Louis, Mo.
(Horner and Shifrin
and Union Electric)
(Combustion Power Co.,
Menlo Park, Cal.)
SW pyrolysis to generate
steam & char (LANDGARD)
SW pyrolysis to fuel oil
and char
1,000
200
300-600
SW combustion along with
coal at a steam power
plant
100
Direct SW combustion
conversion to electricity
(CPU-400)
425
-------
SOLID WASTE
600 TPD
SOLID
WASTE
RECEIVING
& STORAGE
r
SHREDDER
L
LIQUID WASTE
44,000 GPO
LIQUID
WASTE
RECEIVING
& STORAGE
I AIR
E.ASSIFIER
HEAVY MATER IAL
~AGNETIC
EPARATOR
FERROUS METALS
35 TPD
TO MATERIAL RECOVERY
10 TPD
EXHAUST
COMPRESSED AIR
WASTE
HEAT
BOILER
ELECTRICITY
3MW
9MW
~
ENERATOR
! I
ASH
90 TPD
Figure 1. CPU-400 system.
combustion a r, and the whole system operates at about
4 atmospherE s pressure. Solid waste is fed into the
fluidized bed combustor through two rotary air-lock
feeder valves that discharge into a pneumatic transport
system.
This solie waste disposal/resource recovery process
has been undm development through an EPA research
contract with Combustion Power Company of Menlo
Park, California, since 1967. A 100-ton/day
(90,700-kilognm/day) pilot plant employing a 1,000 kW
Ruston Horn!;by turbine has been developed and is
currently undngoing testing: Tests have demonstrated
the system's a.)ility to burn solid waste while generating
full-power ou':put under complete automatic control.
Unfortunately, however, tests have also demonstrated
that extensive deposits on the first-stage turbine stators
develop after Cinly a few hours of operation. The inertial
separators wem simply ineffective, permitting too many
sticky aluminum and other particles to enter the turbine
and result in the deposition problem. The separators
were also inen~ctive from an :>!r oollution control point
of view (ref. 12). As can be SL im from table 4, which
shows the gaseous and particulatt: emission characteris-
tics of the first configuration of the CPU-400 pilot plant,
the particulate loading in the turbine's exhaust reached a
high of 1.6 Ib/106 Btu (0.7 grams/106 joules) (ref. 13).
EPA emission requirements for particulates are 0.1
Ib/106 Btu or 0.043 g/106 J.
To solve the turbine deposition/erosion problem.
and to develop an effective air pollution control tech-
nology for high-temperature/high-pressure applications,
a granular filter has been designed and recently installed
at the EPA CPU-400 pilot plant. Figure 2 is a sketch of
the granular filter design. A continually recirculating,
moving bed of A 1203 balls having a mean diameter of
0.077 inches (1.95 mm) will remove the particulate from
the turbine-bound gas stream. The startup of this filter is
scheduled for November 1975.
The St. Louis Waste-As-Supplementary Fuel Process
A schematic of the Horner and Shifrin Fuel Re-
covery Process is shown in figure 3 (refs. 3, 10, 11, 14,
15). Municipal solid waste is prepared for firing in a
coal.burning utility boiler. The solid waste provides 10
to 20 percent of the fuel requirement and the coal
supplies the balance. The solid waste is shredded, and
the nonburnables are removed by air classification. The
solid waste is then conveyed from a storage bin to a
I
426
-------
Table 4. CPU-400 pilot plant emission
characteristicsa
Flue gas constituent
Exhaust
concentration
02
C02
CO
CH
x
502
NOx
HCl
Particulate
16.1%
5.2%
<30 ppm
2 ppm
<35 ppm
100 ppm
100 ppm
1 .6 1 b/l 06 Stub
aOuring high-pressure, municipal
solid waste-burning operation.
bUsing the ineffective inertial sep-
arators; iie., before installation of
the granular filter.
rotary air-lock feeder and then pneumatically injected
into the boiler. This system has been under development
by Horner and Shifrin and others for the EPA, for the
Union Electric Company, and for the City of St. Louis
since 1968.
Numerous hours of system testing have been
.'ccumulated to date, and the effect of solid waste on
slagging, corrosion, precipitator performance, and ash
handling requirements have been and are being evalu-
ated. The need for an air classifier was established early
when it was discovered that the glass and metals in the
feed material were causing serious wear problems in the
pneumatic transport system and in the rotary feed
valves. Once an air classifier was incorporated into the
system, plant performance improved significantly. The
technical and economic worth of the process concept
has been well established; its overall success has caused a
significant increase in plans to utilize this technology
across the country.
. To define the environmental characteristics of this
cofiring (with coal) option, EPA has been sponsoring
tests being conducted by the Midwest Aesearch Institute
(MAl). Union Electric (UE) has also been carrying out
Iparallel tests of its own. The cofiring of solid waste had
CLEAN GAS
.-.
AStI~
nUIDIZED
BED'
t
Figure 2. CPU-400 pilot plant granular filter.
427
-------
SOlU> WASTE
15( 10 TPD
BOLIO
WASTE
RECEIVING
&
STORAGE
RESIDUE
250 TPD
MAGNETIC
METALS
9
MAGNETIC METALS
100 TPD
ELECTRICITY
STANDARD FUEL
SUSPENSION
FIRED
BOilER
SANITARY
LANDFILL
RESIDUE
Figure 3. Horner and Shifrin Fuel Recovery Process.
no discernibl ~ effect on S02, S03, NOx' or Hg +
emissions (refs. 16,17); however. coals with different
sulfur contents were used in the coal-only and the coal/
solid waste fii-ing operations. Earlier estimates had pre-
dicted a 13 percent reduction in 502 emissions, had
solid waste bee" fired with the same coal to provide 20
percent of thl! boiler heat input at UE's Meramec station
(ref. 15). Em ssions of C1- increased moderately when
solid waste W35 cofired with the coal (refs. 16,18). The
average C1- cnncentration in the stack gas was 17 per-
cent higher when solid waste was cofired. (When coal-
only was burn~d, the concentration was 320 ppm.)
ParticulaB emissions, however, have been consider-
ably more troJblesome. Figure 4 presents a comparison
of electrostatic precipitator (ESP) efficiencies obtained
using the mean values of inlet and outlet grain loadings
from the indil/idual UE and EPA/MRI test series run in
1973 (ref. 161. No significant differences in ESP effici-
ency as a fur,ction of fuel mixture were noted in the
EPA/MRI test> done at that time. However, efficiencies
calculated from the somewhat earlier UE tests did show
a marked dep~ndence on fuel mixture-in all UE cases,
lower ESP eff ciencies resulted from cofiring waste with
coal. In tests I:onducted in May ;)f 1'975, EPA/M R I data
tended more -:0 corroborate the '!arlier UE results than
the earlier Ef'A/MRI fir.Jings, though the new data
points may nelt be representative (refs. 18,191. To sub-
stantiate the worth and validity of these more recent
EPA/MR I tests, additional analyses are planned. .
Two factors that do influence the collectability of
particulate matter in ESP's are known to have some
effect during waste cofiring operations. Increased gas
flow rates through an ESP and increased particle resisti-
vity tend to reduce ESP efficiency- Gas floW rates are
increased when waste is cofired with coal (at a constant
power generation rate). Similarly, particle resistivities
have been found to be about three times higher during
waste cofiring tests than during coal-only tests (refs.
18,191.
The Garrett Pyrolysis Process
In late 1972, work began under the joint sponsor-
ship of EPA and San Diego County on the Garrett
Pyrolysis Process. Figure 5 is a flow chart of this process
(refs. 3,10,11,20), The process employs two stages of
shredding, an air classification operation, and a drying
operation to produce a minus 14- to 24-mesh fuel for
the pyrolysis reactor. The heat required for the pyrolysis
is obtained from the combustion of the pyrolysis off-gas
and a fraction of the char produced. This heat is trans-
ferred by means of the heat exchanger. From the
pyrolysis unit, the gases are exhausted to a cyclone,
which removes the char, and then are scrubbed to re-
move the product, a highly oxygenated fuel oil, and
428
-------
'DO
II;
III
U
a:
III
CI.
t
Ii!:
III
~ 95
IlL.
III
a:
~
~
U
III
a:
..
EPAIMRI
6 COAL
. COALtREFUSE
U.E.
o COAL
" COAL... REFUSE
U.E. COAL ONLY (OCT. 1973)
--
- - .....r U.E. COAL
"" ONLY TEST
NOV. 1973
U.E.COAL'" REFUSE
. (NOV. 1973)
ID
80
90
100
110
120
130
14D
150
JO
GROSS GENERATION, mfgawalts
Figure 4. Variation of ESP efficiency with changes in fuel and boiler load. .
other solids and liquids. The clean gas has a heating value
of approximately 550 Btu's per cubic foot (4.93 x 106
cal/m3) and is recycled to the pyrolysis reactor. A
4-ton/day (3,640-kg/day) plant has been tested. A
200-ton/day (182,000.kg/day) plant is currently under
construction in San Diego.
The projected environmental characteristics of the
San Diego plant include the following: (1) the amount
of waste otherwise going to landfill will be reduced by
over 75 percent; (2) each ton of waste treated will pro-
duce the energy equivalent of 27 gallons of No.6 fuel
oil; (3) when 100 percent pyrolytic oil was burned as a
fuel, only 290 ppm of S02 in the stack gas resulted
(though NOx and HC1 emissions may well be higher
than when No.6 fuel oil is burned); and (4) much of the
water used (and formed) by the process will be recycled
within the process for glass recovery and for char
quenching before receiving final treatment (ref. 20).
The Monsanto Landgard System
The Monsanto Landgard Process is shown schema-
tically in figure 6 (refs. 3,10,11,21). It employs a rotary
kiln pyrolyzer into which shredded solid waste is fed at
one end and air and fuel at the other end. Hot fuel
products of combustion contact the solid waste and
pyrolysis gases and char are formed. The char and solid
waste inerts exit from the kiln at one end into a quench
~ank while the pyrolysis gas exits from the other end
into an afterburner. The gases from the afterburner are
used for steam production and then sent through a wet
scrubber and exhaust fan. A 50-ton/day (45,500-kg/day)
pilot plant has been operated by Monsanto in St. Louis,
and a 1,000-ton/day (910,000-kg/day) system (suppor-
ted partially by an EPA demonstration grant) has
recently begun operation in Baltimore.
Groundbreaking ceremonies for the City of
Baltimore plant took place in early 1973, and startup
commenced in mid-1975. The Baltimore Gas and
Electric Company (BG&E) is purchasing the 200,000
Ib/hour (91,000 kg/hour) of low-pressure steam pro-
duced by the plant (for use in heating and cooling down-
town buildings). The projected environmental character-
istics of the Baltimore plant include: (1) the savings to
BG&E of 15 million gallons of oil per year otherwise
needed to generate the steam; (2) the disposal by the
city of much of its solid waste at lower particulate
emissions than is possible using the city incinerator; and
(3) expected S02 and NOx emissions of less than 100
and 50 ppm respectively. Groundwater protection may
be required to prevent char leachate from entering it, if
the char product is not sold/used but disposed of on the
land (ref. 21).
429
-------
SOLD WASTE
1000 TPD
SOLID
WASTE
RECEIVING
& STORAGE
CLEAN
WATER
112 TPD
J\IR
CLASSIFIER
HEAVY
MATERIAL
MAGNETIC
SEPARATION
GAS
PURIFICATION
GAS
WATER
TREATMENT
DISPOSAL OR
MATERIAL RECOVERY
115 TPD
RECYCLE CHAR CHAR 51 TPD
+ OIL
PRODUCT 228 TPD
CHAR
FERROllS METALS
6"' TPD
Figure 5. Garrett Pyrolysis Process.
SOLD WASTE
1:100 TPD
SOL.O
W A~)TE
RECEIVING
& STORAGE
STEAM
200,000 LB/HR
EXHAUST
QUENCH
TANK
WATER
FLOATS
CHAR
200 TPD
2500 BTU/LB
MAGNETIC
SEPARATCN
FERROUS
METALS
GLASSY
AGGREGATE
Figure 6. Monsanto Langard System.
430
-------
THE CURRENT EPA R&D PROGRAM
IN WASTES-AS-FUEL
I n addition to the four major processes just describ-
ed, EPA is actively engaged in furthering the develop-
ment of, and in environmentally assessing, several other
wastes-as-fuel process options. A broad R&D program (a
new thrust) in this area hanecently been initiated with-
in EPA's Office of Research and Development (ref. 22).
This research program will be summarized below with
the aim of providing some insight into program ration-
ale.
The current EPA program can be subdivided into
four major areas of emphasis. These areas are: (1)
Assessment R&D (environmental, technical, economic);
(2) Waste Cofiring with Coal, Oil, or Industrial Waste;
(3) Waste Cocombustion with Sewage Sludge; and (4)
Pyrolysis and Bioconversion Processes.
Table 5 lists those project areas that are illustrative
of work ongoing in the first subdivision, Assessment
R&D. Work here features environmental, technological,
or economic assessments of waste-as-fuel options. Re-
search in "Assessments" includes a study to acquire
pollutant characterization data for the competing waste-
as-fuel processes as well as to develop environmental
assessment criteria and proper sampling/analytical tech-
Iniques. With these data, R&D will begin on assessing the
technology available to control air pollution from these
processes and on developing needed control technology.
Work is also underway, by the Ralph M. Parsons Com-
Table 5. Assessment R&D (environmental,
technical, and economic)
Pollutant characterizations
Assessment criteria
Sampling/analytical methodology
Air pollution control technology as-
sessment/development
Process evaluations (parsons)
Waste surveys
Technical assistance (10 cities)
Waste preprocessing
Side-by-side evaluation
S-O-T-A evaluation
(NCRR)
Table 6; Waste cofiring with coal, oil,
or industrial waste (technical and
environmental evaluations)
With coal
St. Louis/Union Electric
Battelle/City of Columbus
Comparative test burns of cofiring
in tangentially-fired versus stok-
er-fired boilers
Densified waste cofired in an in-
dustrial-sized stoker-fired boil-
er
With oi 1
Refuse cofiring with oil
With industrial waste
.Use of refuse in bagasse boilers:
feasibility study (Hawaii)
pany of Los Angeles, on performing a comparative tech-
nical and economic evaluation of the developing waste-
as-fuel processes. Several waste surveys will be con-
ducted by other organizations in an attempt to better
define the energy values of selected waste streams, such
, as those from industry, from agriculture, and from de-
molition operations. To spur the coming together of
waste producers (e.g., the cities) and waste users (like
the power industry). EPA has awarded approximately 10
grants to selected municipalities to allow for appropriate
planning of waste-as-fuel utilization or conversion facili-
ties (done under the "Technical Assistance" charter
shown on table 5). It is also shown under Assessment
R&D on table 5 that work will be done in waste pre-
processing. An EPA grant has been awarded to the
National Center for Resource Recovery to enable a side-
by-side evaluation of existing shredders, air classifiers,
etc., to define operating characteristics and efficiency. A
state-of-the-art evaluation of other preprocessing tech-
niques for grinding and separating the inorganic and
combustible waste fractions will also be performed.
Table 6 illustrates R&D to be done in the second
subdivision of the waste-as-fuel program, in Waste
Cofiring with Coal, Oil, or Industrial Waste. Work in this
431
-------
area includes finalization of the work at St. Louis (sef!
"Major Wastl'-As-Fuel Processes"). which will center on
the identific ition of potentially hazardous emissions
associated wi th waste cofiring with coal. R&D will con-
tinue by Battelle at a small Columbus, Ohio, boiler in
which varyinu ratios of coal-to-refuse will be run, Partic-
ular emphasis here will be on defining the corrosion rates
associated with cofiring. At another site, a comprehen-
sive el!,aluation will be made of cofiring waste with coal
in tangentially fired boilers versus stoker-fired boilers.
The evaluatie>n will compare process and environmental
characteristic:; of each option. Yet another coal/waste
cofiring option will be tested: that of densifying (e.g.,
pelletizing) the waste and injecting along with lump coal
into an induS1rial-sized stoker-fired boiler.
Table 6 also lists two other options that will be
investigated: (1) the combustible fraction of municipal
solid waste will be doubleground and cofired along with
fuel oil in e"isting oil-fired boilers (technical and envi-
ronmental feilsibility will thus be determined); and (2) a
feasibility study is underway in Hawaii to determine the
value of cofiring solid waste along with bagasse in
industrial bagasse boilers.
Table 7 ndicates projects recently initiated in the
waste-as-fuel area, known as Waste Cocombustion with
Table 7. Waste cocombustion with sewage
sludge (technical and environmental
evalua'tions) ,
St. Pilul/Seneca treatment plant
sludge incinerator
(Waste ~upplies the,heat to inciner-
ate the sludge in existing multiple-
hearth furnace, thereby backing-out
conventional fuels that are in short
supply and accomplishing the dispos-
al of t~u wastes instead of one.)
CPU-400: Power recovery while dis-
posing of solid waste and sewage
sludge
(Sludge is injected into the four-
atmosphere fluid bed along with
solid waste, the off-qases from
which pass through a granular filter
for air pollution control and are
then expanded through a gas turbine.)
Sewage Sludge. The first of these projects centers on the I
utilization of various municipal and industrial wastes
(e.g., refuse, tires, wood chips). and/or low-sulfur coal,
to combust or incinerate sewage sludge in an existing,
full-scale multiple hearth sludge incinerator in St. Paul,
Minnesota. The second project shown involves using the
CPU-400 pilot plant (see "Major Waste-As-Fuel Pro-
cesses") to test the process and environmental character-
istics associated with cocombusting solid waste and
sewage sludge in a high-pressure fluidized bed, the off-
gases from which are cleaned upstream of their expan-
sion through a gas turbine for power recovery.
Table 8 lists those R&D projects EPA has underway
in the waste-as-fuel areas of Pyrolysis and Bioconversion.
The most comprehensive of these is the recently begun
research at the Energy Resources Company (ERCO) in
Cambridge, Massachusetts. ERCO, with assistance from
MIT, will conduct bench and pilot-scale (500-lb/hr or
230-kg/hr) tests on the pyrolytic conversion of mixed
waste to fuel. Work at this environmental test facility
will feature the pyrolysis of various mixtures of waste
(e.g., refuse, industrial wastes, agricultural wastes) at
varying residence times (space velocities) and operating
temperatures to define what product ratios (of fuel gas
to fuel oil to char) are possible and what environmental
Table 8. Pyrolysis and bioconversion R&D
(technical and environmental evalua-
tions)
Pyrolysis
Pyrolytic conversion of mixed waste
to fuel (ERCO)
Pyrolytic conversion of solid waste
to polymer gasoline (the Navy at
China Lake, California)
Portable pyrolysis of agricultural
waste to transportable fuels (Geor-
gia Institute of Technology)
Bioconversion
Enzymatic hydrolysis of waste cel-
lulose to glucose to an ethyl alco-
hol fuel (the Army at Natick, Mas-
sachuse~ts)
432
-------
.,nscquencl!s may result. The China lake Naval
"~eapons Center is conducting bench-scale research for
EPA on upgrading (via polymerization reactions) pyro-
lytic fuel oil/gas to gasoline. Another R&D effort under-
way in the waste pyrolysis area is that which is being
done at the Georgia Institute of Technology. It centers
on the concept of disposing of agricultural wastes by
having them processed in a portable van containing a
pyrolysis reactor that converts the otherwise wasted
material to transportable fuels, such as fuel oil or char.
Regarding the last project listed on table 8, the Army's
Natick laboratories are carrying out bench-scale re-
search on converting waste cellulose to an ethyl alcohol
fuel by using enzymatic hydrolysis of cellulose to glu-
cose as an intermediate step.
SUMMARY
The 570 million tons of dry, combustible solid
waste discarded each year represent an energy loss of
about 12 percent of the total U.S. energy requirement.
One-fourth of this amount, or the energy equivalent of
about 3 percent, exists in lots large enough to feed
energy recovery plants that are sized to handle 100 tons
of solid waste per day.
Following a discussion of EPA's legislative mandates
'n the wastes-as-fuel area of research, four major pro-
cesses that have been under long-term EPA development
were described, as were the environmental characteristics
of each. These processes are: the CPU-400 System, the
St. Louis Waste-As-Supplementary Fuel Process, the
Garrett Pyrolysis Process, and the Monsanto Landgard
(Pyrolysis) System.
The current EPA R&D program in wastes-as-fuel
was then summarized. Environmental research in this
area centers on: (1) Technical, Economic, and Environ-
mental Assessments; (2) Waste Cofiring with Coal, Oil,
or Industrial Waste; (3) Waste Cocombustion with
Sewage Sludge; and (4) Pyrolysis and Bioconversion Pro-
cesses.
REFERENCES
1. U.S. Environmental Protection Agency, Office of
Solid Waste Management Programs, "First Report to
Congress: Resource Recovery and Source Reduc-
tion," February 22, 1973, p. 2.
2. International Research and Technology Corpora-
tion, Problems and Opportunities in the Manage-
ment of Combustible Solid Waste, Final Report on
Contract No. 68-03-0060 to the U.S. Environmental
Protection Agency, Solid and Hazardous Waste Re-
search Laboratory, NERC-Cincinnati, 1972,509 p.
3. R. A. Chapman, "Solid Waste as a Fuel tor Power
Generation," Proceedings of the 1973 Washington
State University Thermal Power Conference,
October 3-5, 1973.
4. L. L. Anderson, "Energy Potential from Organic
Wastes: A Review of Quantities and Sources," U.S.
Bureau of Mines Information Circular 8549, 1972.
5. National Commission on Materials Policy, Material
Needs and the Environment Today and Tomorrow,
final report, June 1973.
6. G. M. Fair, J. D., Geyer, and D. A. Okum, "Water
Purification and Wastewater Treatment and Dis-
posal," Water and Wastewater Engineering, Vol. 2,
J. Wiley and Sons, New York, 1968, pp. 36-6 to
36-8.
7. T. W. Bendixen and G. L. Huffman, "Impact of
Environmental Control Technologies on the Energy
Crisis," Newsletter by the U.S. Environmental Pro-
tection Agency's Cincinnati National Environmental
Research Center, January 11, 1974.
8. Stanford Research Institute for the Office of
Science and Technology, "Patterns of Energy Con-
sumption in the United States," January 1972
(available from NTIS as PB-212-7761.
9. E. Hirst, "The Energy Cost of Pollution Control,"
Environment, Vol. 14, No.8 (October 1973), pp.
37-44.
10. W. E. Franklin, D. Bendersky, L. J. Shannon, and
W. R. Park, "Resource Recovery: Catalogue of Pro-
cesses," Council on Environmental Quality, 1973,
151 p. (available from NTIS as PB-214-148).
11. A. W. Breidenbach and G. L. Huffman, "Thermal
Degradation of Solids/Sludges," Proceedings of the
Philadelphia 66th Annual Meeting of the American
Institute of Chemical Engineers, November 15,
1973 (in press).
12. R. A. Chapman and G. L. Huffman, "Solid-Fuel-
Fired Gas Turbine Pilot Plant Testing Status," dis-
cussion paper for United Nations'/Economic
Commission for Europe's Second Seminar on Desul-
furization of Fuels and Combustion Gases, Washing-
ton, D.C., November 11-20, 1975.
13. R. A. Chapman, "Development of a Solid-Waste-
Fired Gas Turbine System," presented at the First
International Conference on Conversion of Refuse
to Energy, Montreux, Switzerland, November 4,
1975.
14. Horner and Shifrin, Incorporated, "Solid Waste as
Fuel for Power Plants," for the U.S. Environmental
Protection Agency's Solid Waste Management
Office, 1973 (available from NTIS as PB-220-316).
15. R. A. Lowe, "Energy Recovery from Waste: Solid
Waste as Supplementary Fuel in Power Plant
433
-------
Boilers," ~)econd Interim Report (SW-36d.ii) by the
USEPA, 1973.
16. J. D. Kil1roe, L. J. Shannon, and M. P. Schrag,
"Emissior.s from the Suspension Firing of Municipal
Solid Waste and Pulverized Coal," presented at the
68th Annual Air Pollution Control Association
Meeting, EJoston, June 15-20, 1975.
17. L. J. Shannon et aI., "St. Louis-Union Electric Re-
fuse Firing Demonstration Air Pollution Test Re-
port," Rllport EPA-650/2-74-073, U.S. Environ-
mental Pr:>1ection Agency, Washington, D.C., May
1975.
18. J. D. Kil!lroe, L. J. Shannon, and P. G. Gorman,
"Emissions and Energy Conversion from Refuse
Processing and Mixed Fuel Boiler Firing," presented
at the Fin,t International Conference on Conversion
of Refuse to Ener!Jy, Montreux, Switzerlalld,
November 1975.
19. Midwest Research Institute for the USEPA, "Test
and Evaluation Program for St. Louis-Union Electric
Refuse Fuel Project," Monthly Report No.4 on
EPA Contract No. 68-02-1871, June 1975.
20. Steven J. Levy, "San Diego County Demonstrates
Pyrolysis of Solid Waste," USEPA Report
SW-80d.2, 1975.
21. David B. Sussman, "Baltimore Demonstrates Gas
Pyrolysis: Resource Recovery from Solid Waste,"
USEPA Report SW-75d.i, 1975.
22. USEPA's Office of Energy Research, "EPA Wastes-
As-Fuel Research, Development, and Demonstration
Program Plan," April 1975.
434
-------
PRELIMINARY ASSESSMENT OF THE ROLE OF ENERGY
STORAGE AND IMPLICATED TECHNOLOGIES FOR
ENERGY CONSERVATION IN INDUSTRY
Donald R. Glenn*
Abstract
Energy storage is but one possible means forreduc-
ing the consumption of fuel and/or electricity within the
industrial plant A twofold approach is underway to
identify candidate industries where heat storage can
measurably improve energy efficiency and produce a
concomitant improvement in operating economics. This
paper presents the methodologies selected to identify
candidate industries and processes, the nature of avail-
able information, and how additional needed data on
energy utilization dynamics are being accomplished.
Federal Energy Administration and the Energy
Research and Development Administration have recog-
nized a needed focus on current energy-related factors
describing the industrial sector and programs to explore
applicable conservation techniques. In one such pro-
gram, the role of energy storage, particularly thermal
energy, is synthetized into process designs for a first
approximation of cost-benefit and energy conservation
~mpacts afforded by inclusion of this function. Examples
of near-term planned or feasible energy storage schemes
are shown. With some 40 percent of the total energy
demand related to industrial requirements including
space heat, process and direct heat, and boiler heat for
steam, the full spectrum of temperature and energy
levels are represented. Matching in-plant available excess
or reject energy to other compatible needs separated by
a time and/or magnitude function, can represent a real
opportunity for an energy storage application. Signifi-
cant design challenges for accomplishing requisite heat/
energy exchange and storage in operational industrial
energy systems are recognized and discussed.
Fifteen years ago a policy decision was made to put
a man on the moon by the end of the 1960's, and it was
done. During that same decade a second major policy
was formulated to deal with our wastes and the resulting
degradation of our environment, and it is slowly being
done. Most recently we have recognized the abusive
practices employed in consuming our once abundant
energy resources. We have yet to make significant com-
mitments towards a firm policy to see that this crisis is
alleviated. Aggregated forces united in purpose are what
.Program Manager, Tharmal Energy Storage, General Elec-
'ric Company. Space Division, Philadelphia, Pennsylvania.
make things happen; the moon in 9 years with a work
force of 200,000 people; the environment in ? years
with a force of millions; and energy in ?? years with
multimillions needed for assured success. A visible tar-
get, easily perceived by all involved, is the major impetus
in planning conquests and achieving them. The environ-
mental and energy situations are so nonaggregated that a
single approach with a manageable supporting force
cannot effect a measurable penetration toward achieving
the needed objectives. No one approach is going to
resolve singly our worsening energy posture. Recognizing
this, we can explore one permutation of the energy
problem represented by the industrial sector where some.
40 percent of our yearly energy supplies are expended.
Energy management is a term that will become a
way of life for industry once the several components
comprising a corporate energy system are defined and
their interactions better understood. As a definition,
energy management is a purposeful program of defining
realistic energy demands, planning interactive processes
and facilities to achieve highest practicable efficiencies in
their operations and therefore the lowest energy con-
sumption profiles consistent with each product:s value
and market demand, and their implementation.
Overview of Energy and the Industrial Sector
Unlike its nearest energy-consuming relative, the
electric utility industry has not formulated the sophisti-
cated data and statistical records that are the starting
point for effective energy management. While the rea-
sons may be obvious, nevertheless it takes these kinds of
record keeping to objectively formulate conservation
potentials and assess their impacts. The author has been
engaged in a study aimed at determining applications
and developing conceptual designs of thermal energy
storage systems responsive to these applications. To this
end, a review of industrial energy consumption was con.
ducted with the express goal of citing regions and indus-
try within these regions expected to be hard pressed
energy-wise because of energy shortages and/or high
costs. A second thrust involved a review of basic process
designs within each major industry in a more classic engi-
neering attack on energy inefficiencies. The major
constraint in this second effort is the dearth of available
data describing the dynamics of process operations. The
Federal Energy Administration (FEA) funded several
studies to describe energy conditions in major industrial
435
-------
groupings. While each provided insights never amassed
previously or the processes and factors involved, none
explored real. time energy flows as impacted by through-
put, process design, shift operations, seasonal and day-
of.week vari" bil ity, and the effect of each on energy
consumption and efficiency. Within the bounds of this
situation, en,ngy storage is one approach of many
needed within the industrial energy system to more
effectively utilize industry's main ingredient-energy.
Energy Sourcas for Industry
Project I,dependence studies have forecast energy
demands ($11/bbl oil with conservation) through to
1985 by sector (table 1). and for the industrial sector by
fuel/energy source (table 2).
Gradual changes in the source of industrial energy
have taken place over the years. Coal use has been stead.
ily declining in recent years due to pol!ution require-
ments and th~ ready availability of natural gas and oil.
This trend is projected to reverse due to, among other
factors, limited natural gas supply and greatly increased I
petroleum prices. The historic uradual incmase in I!h!c.
trical consumption is projected to ucctderate as tht!
United States increases its nuclear generation capability.
A regional split of these fuels for 1971 shows the
characteristics of large uses of fuels contiguous to that
region (table 3).
As an example, coal is heavily consumed where it is
produced, due in part to high transportation costs. Iron
and steel, the largest industrial coal consumers, have
located at their primary energy source rather than their
raw material source. The top four coal-producing census
regions are responsible for about 91 percent of U.S. coal
production and in turn consume about the same percent.
age. Similar scenarios exist for natural gas and oil.
Although industrial consumption of electricity
represents only 10 percent of industry energy, certain
industries consume a considerably higher percentage.
Primary aluminum is the most notable example, requir.
ing over half its energy as electricity. This factor, com-
Table 1. Net distrubed energy - quadrillion Btu
Sector 1968 1970 1972 1977a 1980~ 1985a
Residential/commercial 15.35 16 . 72 17.94 18.24 19.66 22.59
Indus tri a 1 21.09 22 . 32 22.94 24.12 25.51 27.93
Transportati on 15.54 16.69 18.07 17. 71 18.16 19.13
Total 52.00 55~74 58.95 60.07 63.33 69.65
aproje cted.
Table 2. Source of industrial energy - percent
1968 1970 1972 19 na 1980a 1985 a
COf.l 24.3 22.4 18.7 19.8 21.4 21.9
Petrol eum 22.9 23.0 24.9 28.5 28.8 28.8
Natural gas 44.0 45.6 46.4 39.6 37.1 35.4
E 1 E'ctri ci ty 8.8 9.0 10.0 12. 1 12.7 13.9
rota 1 100.0 100.0 100.0 100.0 100.0 100.0
('Projected.
436
-------
Table 3. Regional industrial energy sources - percent
Na tura 1
Census re g i on Coal Petroleum gas Electricity
New Engl and 1.9 58.9 17.6 21.6 100.0
Middle Atlantic 41. 9 18.7 25.9 13.5 100.0
East North Central 41.2 8.1 38.8 11.9 100.0
West North Central 12.8 11.0 64.5 11.8 100.0
South Atlantic 30.1 18.1 34.8 17.0 100.0
East South Central 29.7 8.3 40.3 21.7 100.0
West South Central .4 14.7 80.3 4.6 100.0
Mountain 13. 1 12.4 62.9 11.5 . 100.0
Pacifi c 4.5 16.2 59.4 19.9 100.0
bined with a very high energy intensity of about 75
million Btu per ton of product, leads aluminum plants to
areas having cheap electric power. Over one half of the
industry's electricity is generated by hydroelectric power
plants. On a broad basis, an alternate energy concept
such as energy storage must be economically (and tech-
nically) attractive to enough of the industrial sector to
establish a viable market for the device and thereby
significantly impact total industrial energy cons~mption.
The changing energy source mix will impose severe
economic stresses on all regions with the degree of sever-
ity varying in certain regions and energy-consuming sec-
tors. The economic analyses presented here analyzed the
input/output inventory of energy by fuel types in each
region, reduced to per capita consumption and applied
to the demand projections for 1985 (in this case). Short-
ages/surpluses are allocated according to end-use priority
and that growth rates remain even by fuel type. Alloca-
tions are made on the basis of percentages of shortages.
Resulting demands then are as shown in table 4.
Each consuming sector in each region will have con-
siderably increased expenditures for energy in 1985
from the base year. Using constant dollars to factor out
inflationary effects, the expenditure increases can be
broken out into three categories:
1. I ncreases due to increased energy use alone,
2. Increases due to price increase alone, .
3. I ncreases due to energy mix change alone.
This procedure, analogous to variance analysis in
accounting, can assist greatly in determining economic
stress areas. Item 3 is the prime indicator of stress, since
energy shortages and capital expenditures must occur.,
Energy. price projections become necessary at this
point. A matrix of regional and sector prices was derived
that was consistent within itself and with the Project
. Independence projections. These prices and 1971 actual
prices were applied to consumption data to yield total
industrial energy expenditures (table 5).
Analysis of the 1971 to 1985 expenditure increase
yields the following data (table 6).
The Middle and South Atlantic and West North
Central regions are the likely stress areas in terms of fuel
mix changes. An even more compelling stress is the
prospect of energy supply curtailment. The industrial
sector will feel the heaviest impact of the gas shortage
under present curtailment schedules. The economic
effects of curtailments can quickly make economic
effects of high energy costs seem insignificant.
There are certain industries in which natural gas is
. .
the only fossil fuel that can provide process heat and still
maintain product purity. Glass is a prime example, with
many northeastern glass producers facing almost certain
shutdowns in coming months. Since the regenerative
furnace most commonly used in glass manufacture
requires about 2 weeks to stabilize in temperature, even
a 1- or 2-day fuel curtailment would involve a substantial
loss of production.
Oil (and coal) can use several means of distribution
including tank cars, trucks and barges so that "black
marketing" and "leakages" buffer absolute controls of
437
-------
Table 4. Projected 1985 industrial energy source 1012 Btu
-_:S::-
Regi on Consumpti on Coal Oil Gas Electri city
New [ng1 and 494 12 303 28 152
Middle Atlantic 3,265 1,258 1,186 289 532
East North Central 6,033 3,180 1,290 773 790
West North Central 1 , 441 175 721 325 220
South Atlantic 2,774 695 1,203 311 565
East South Central 1 , 832 557 37 265 523
West South Central 8,324 36 1,138 6,729 421
Mountain 1,268 132 193 801 142
Pacifi c 2,500 76 1 , 45 7 442 525
Total 27,932 6, 121 7,528 9,963 3,870
Table 5. Industrial energy expenditures - millions $2
New England
Middle Atlantic
East North Central
West North Central
South Atlantic
East South Central
West South Central
Mountai n
Pacific
1971
$ 648
3,344
4,730
956
. 2,086
1,354
2,841
671
1 , 709
1985
$ 1, 752
9, 102
12 , 1 70
4,544
.7,182
4,125
13,444
2,601
5,974
aSourcp:
analysis.
1971, Census of Manufactures; 1985, GE
438
-------
Table 6. Expenditure increase - 1971 to 1985
Tota 1 increased Due to Due to Due to
expendi tures vo1u~ p ri ce mix change
New Eng1 and $ 1, 104 285 825 -6
Mi dd1e At1 anti c 5,758 1,077 3,663 1,018
East North Central 7,440 2, 100 6,214 -874
West North Central 3,588 336 1 ,448 1 ,804
South Atlantic 5,096 952 3,537 607
East South Central 2, 771 439 2,226 106
West South Cen tra 1 10,603 1 , 288 9,281 34
Mountai n 1,929 336 1 ,444 149
Pacific 4,265 1 ,024 3,330 -89
oil apportionments. Therefore, there is greater flexibility
of overcoming oil shortages, though at higher costs, to
secure needed supplies, where gas users have little
recourse of curtailed supplies.
Energy Storage
Functional types of energy storage compatibilities
are shown (tables 7 through 11 and figs~ 1 through 3).
Of these, water and rocks have been most widely used
for thermal storage, particularly in the European residen-
tial/commercial sector. Energy conservation through
off-peak electrical usage was successfully implemented in
Europe using thermal storage for room heating. Batteries
have found use predominantly as backup to primary
power in case of outages. Other than for spinning re-
serve, flywheel technology has not been economical.
Technical breakthroughs in materials developments and/
or structural design are needed. Each of the other
methods requires some complex or erratic interaction
with current systems to mature a design approach com-
patible with in situ installations, codes, and institutional
barriers. Nevertheless, as the value of energy commod-
ities continues upward, newer, more challenging con-
cepts will fall into favor for a bona fide benefit, meas-
ured as cost or fuel savings. An indication of the ap-
proximate availability of various advanced energy stor-
age methods is shown in figure 3.
Selection of a particular form or product of energy
storage must consider minimizing conversion steps, as
from 20 percent to 60 percent efficiency losses are com-
mon for each conversion. Conversely, the stored form
must satisfy the system requirements; for example, if a
3-month time lag is required between charge and dis-
charge of a stored energy form, a thermal s~orage tech-
nique would most probably be eliminated in favor of a
chemical storage method such as hydrogen or methane.
So, there are criteria defined not only by the total
energy system but also by user needs. Each application
can, and will, have several possible solutions depending
on what optimizations are required.
1. Thermal Storage Technologies and Status. . The cur-
rent ERDA-sponsored program, being carried out at the
GE-Space Division has reviewed the status of thermal
storage technologies and devices. A survey of available
literature was conducted to compile a listing of devices
applicable to thermal energy storage. Table 12 lists the
results of this survey and includes results from both pri-
vate and government-sponsored research projects, as
published in the literature.
2. Chemical Energy Storage has focused about hydrogen
and methane reactive concepts and battery develop-
ments. Hydrogen storage forms have been tried as a
liquid, compressed gas, and in hydrided metals. Liquid
hydrogen at cryogenic temperatures has been handled,
stored, and safely transported as part of manned space
flight programs among other uses. While several acci-
dents and mishaps have occurred (ref. 21. there is little
doubt that industrial capability can be easily adapted to
overcome these situations. The nature of H2 storage
being cyclical, embrittlement due to stress fatigue needs
further R&D.
439
-------
=..-=r .-=.~
Table 7. Energy storage methods
Energy density (Btuif-~3)-=
Methods
14
Hydroelectric (pumped water @ 100 ft head)
Steam
15 PSI @ 2120 F
130 PS I @ 3470 F
500 PSI @ 4670 F
15 PSI @ 2120 F
130 PSI @ 3470 F
500 PSI @ 4670 F
Water
Hydrogen
Gas
15 psi @ 600 F
1,000 psi @ 60° F
Liquid 15 psi @ -425
Hydride (M92Ni; or FeTi)
Ammonia
Methdnol
Gaso line
Battl~ries
Flywheels
Compl'essed gas
Cryo!lenic magnetic fields
Sensible heat materials} Low
PhasE' change materials
Sensible heat materials }w h
Phase change materials 19
temp.
temp.
--
Hydrided metals are excellent storage materials in
that high ener!ly densities are attainable with the hydro-
gen in a neutral condition and the reactivity controlled
by low temperatures, -800 F to charge and 2000 F to
discharge an ron-titanium hydride bed. Hydrides for
storage have n,)t yet been demonstrated on a large scale,
however, a near.term applicatiun :~ entirely feasible tech-
nically. Brookhaven National Laboratories have been
researching and experimer~ing with metal hydrides for
some years. Iron-titanium alloys, among other alloys,
40
340
1 ,270
9,000
16,000 (see figure 1)
21,000
. 280
18,500
200,000
250,000
340,000
430,000
830,000
See table 11
See fi gure 2
See table 8
See table 9
See table 10
have the potential for forming hydrides. The Brook.
haven researchers performed an in itial evaluation of
three methods of hydrogen storage, as LH2' compressed
gas, and in hydride form. The results were based on a
large-scale system and showed hydrides, compressed
hydrogen, and liquid Hz ranked in that order.
Methane is employed in two chemical heat pipe
designs, the first uses carbon monoxide, reacting to
water in the (Adam-Eve) process, while the second
employs carbon dioxide, forming hydrogen and carbon
440
-------
Table 8. Thermal energy storage materials
Melting Hea t of Heat capacity
po i nt, fu sian, Btu/ft3 a
Phase change materials OF Btu/1b
Sa It hydrates
K2HPO 4' 6H20 52-56 47 4,900
Ca(N03)2.4H20 117 66 7,650
MgC12.6H20 239 71 6,940
Na2S04.1/2NH4C1.1/2NaC1.10H20 55 78 7,200
Na2S203.5H20 113-120 90 9,200
Na2S0 4' 101120 90 108 9,900
Organi c waxes
C14 paraffin 35-40 71 3,420
C16 paraffin 58-65 86 4, 190,
1-decano1 40-45 89 4,590
Pl16 paraffin 116 90 4,380
Iron are
Scrap iron
Speci fi cheat,
B tu/ 1 b
1.0
0.21
0.20
0.15
0.115
Heat capacity,
Btu/ft3 a
1,250
610
640
980
1,100
Sensible heat materials
Water
Concrete
Stone
aBased on 20° F temperature change.
441
-------
Table 9. Properties of selected "prime" thermal energy storage materials utilizing sensible heat
Materi a 1
MgO
~
~
N
~..,,+ non,,;+., QTIII~.3 Iv,n4,..
..--- --"-''''J -.""",,,,, ~v I
HS-2,OOO°F
OT-1,OOO°F
6.7
5.3
5.3
8.5
9.0
7.5
12.0
9.0
4.8
6.8
9.2
HS-1.500°F
OT-l.OOO°F
3.3
2.7
2.7'
4.2
4.5
3.7
6.0
4.5
2.4
3.4
4.6
4. 1
5.4
HS-1.000°F
OT - 200°F
3. 7
2.9
4.7
5.0
4. 1
6.6
5.0
2.6
3.7
5.0
6.4
8.6
8.7
24.0
The rma 1
conducti vi ty
Btu/hr-ft of
CaO
BeO
B
A1203
CrB2
Fe
SiO
FeD
Fe3C
liF
75NaF /25MgF 2
69liF/31BeF2
Sb203
Sb2D3
liOH
Mg(N03)2-6H20
Glycerol
*
HS - Heat source temperature; OT - Operating temperature.
The values for the Heat source (HS) and operating temperatures (OT) were arbitrarily chosen to
reflect present day needs. Any other values of (HS) and (OT) will have little or no influence
on materials selection. -
H5-500°F
OT -200°F
1.9
1-10
1.6
2.5
2.7
2.3
3.6
0.5-3
5-30
5-20
1-10
1-10
2.7
1.5
2.0
2.8
2.4
3.3
40- 70
0.1-1
1-10
1-10
0.1-10
0.5-3
3.3
9.0
6.9
0.5-3
0.5-3
0.5-3
3.3
2.8
1.6
0.5-3
0.5-3
O. 1-1
Con ta i "-men t
Mild steels
Mild steels
Mild steels
Mild steels
Mild steels
Mil d s tee 1 s
Mild steels
Mild steels
Mil d s tee 1 s
Mild steels
300 series SS
300 series SS
300 series SS
Mi 1 d s tee 1 s
Mild steels
300 se ri es S5
Mi 1 d s tee 1 s
Mi 1 d s tee 1 s
-------
Table 10. Properties of selected "prime" thermal energy storage materials
utilizing the latent heats
Heat..
Operatin!: lIent" (1) (2) (3) Dcnsit~ Thermal
Temperature Storage AHf AHs AHv 13 1'1' :r' Conductivity
Hange, 0 F Material Temp. ° I' BTUllh I3TU/lb BTU/lh (x 10 :' BTU/hr-ft-O I' Cortainnwnt
SI 25HO 710 10.3 47.0
Be 2310 520 f>.O !)4. 0
2:iOO M!:F2 2320 402 7.7 :100 Ser;es SS
CU20 2250 1(1!) f>.4 MilcI Steels
to Fe30 2230 12.1 5.7 1.5 Mild Steels
CuF., 21!)O IH3 5.2 4.0 300 Series SS
2000 AIF:; 21UO lfi50 2U." 300 Serics SS
1.11 2140 550 13.!)
Mo03 2110 412 12.0
I'hMo04 1950 1~7.:' 5.7
CuCI2 IU:IO 750 15.7 :HlO Ser les SS
2000 NiCI2 IHHO ~5fi 5.7 :100 Sel'les 5S
MnF2 IH:W 147~ 3fi.:, :100 Series SS
to Na!' I~~o :\22 ;1, Ii 1.5 3011 Series SS
NICI2 17H5 745 Hi.:1 300 Series SS
I(;()O VO 17.10 404 ].1.0
Gc 1715 202 r..7 13.0
ScCl3 1720 22f> 5.3 300 Series SS
VCI2 1f>70 filiO 12.7 300 Series SS
\.IF 15% 400 5.7 0.3 300 Scrips SS
75NaF/25M!:F2 1530 275 4.9 300 Series 5S
CoCO:! 1515 22A ~.!) 1.0 ~lIlcl Steels
WOO VB"2 14711 3A3 1 o. ~)
CcI 1410 3A2 20." 34.0
to 50LlF/5I1M!:F~ 13RO :12.' :'.:\ :100 Series SS
\.ill 1290 1150 5. fi 2.6 321 Type SS
1200 TICt2 1290 70H 1:\. H 300 Series SS
Se 1265 50:! 15.0 32.0
r.0I.iF/4I1NaF 1205 3.12 5.3 300 Series SS
GeU 1310 1020 2H.1i
.\liLiF/4.INa F/I0MRF2 1174 364 5.9 300 Series SS
ZrF4 1110 439 12.1 300 Series SS
1200 Fel2 1100 310 10.3 300 Series SS
KC!04 9HO 53s 7.8
tu TU2 895 289 9.0
ToO.} H50 fif>5 23.H
>Hlil LlOIi 840 IHf> 3.0
511!.10H/50LlF 800 220 :1.,1 :WO Series SS
TiCI:I HOO 490 H.3 :1110 Series SS
1!°207 fiH5 r.70 2;-).2
HIIO Na2U2 f>80 135 2.4
Se02 600 339 8..1
to Ite207 565 5~1 2.24
1'11'4 540 332 5.H :WII Series SS
400 LlN03 486 159 2.4
AICI:\ 378 120 1. H2 :100 Series SS
AICi:\ 360 196 2.UH 300 Series SS
GeFI 310 440 A.5
.100 112U 212 9r.l r..O 0.:1 Mild Steels
tu Ct'ITolo\\ Eutectic 13fi 39.1 2.2
LI 1010:1 - :11120 A6 128 1.8
0 VF!) f>7 148 2.3
1120 32 144 0.9 1.4 ~lilcI Steels
.Basod on meltIng puint, sublimation point or boiling point.
"A materials heat density is bosed nn its lotent heat multiplied by Its mass density.
(I) I.atent Beat of Fusion.
(2)l.atcnt Beat of Suhlimation.
PIl.atent Boat of Vaporlzatinn.
Blanks in the last two columns of Table K-7 inclicate no dnta is avallnble.
443
-------
Table 11. Review of current battery technologies3
=--
Pe rforman ce Cycle Projected
System Wh/lb W/lb 1 i fe cost Problems
Lead-aci d 10 20- 30 1,500 $ 80/kWh
Ni cke l-i ron 25 50 ? $lOO/kWh Gassing, maintenance
effi ciency
Nickel-zinc 30 150 200-400 Same as Life
Lead-aci d?
Metal - gas battery performance
Performance
Sys tem Wh/lb W/lb Problems
Iron-air 40-50 10-20 Ca thode co rros i on, 1 i fe
Zinc-air 40-50 10-20 Life, cost
Ni eke l-hydrogen 30- 40 ? Volume, 1 ife
Zinc-oxygen 50-60 10-30 Li fe, cos t
Cadmi um-oxygen 30-40 ? Li fe, cost
Zinc-ch lori de 50-75 40 -60 Life
Alkali metal - high-temperature battery performance
Performance Cycle
System Wh/lb . W/lb 1 i fe Prob lems
SOdium-sulfur 80- 1 00 80-100 200-2000 Life, costs
(beta alumina)
Sodium-sulfur 80-100 80-400 100+ Life, materials
(glass) stabil ity
Lithium-sulfur 100 >100 2000 Materi a 15 corro-
sion, costs
Lith; um chlorine 50 » 1 00 100 Life
444
-------
100
10.0
I/)
III
.J
I
..
i3
w
~
1.0
1000
!r
I
w
~IOO
~
w
a:
~
10
I
30
5
10 15 20
ENERGY DENSITY - 103 BTU/FT3
25
Figure 1. Energy density of water vs. pressures.
o
10
103
ENERGY POTENTIAL' WATT -HRS'LB
Figure 2. Flywheel energy capacity.
445
-------
PRESENT 1985 2000
(1975)
PUMPED-HYDRO X .
COMPRESSED AIR OR STEAM , X .
SENSIBLE ~'EAT STORAGE X ~
(H20, Oil RCCKS)
LEAD-ACID BATTERIES X .
ADVANCED UATTERIES X .
LATENT HEAT STORAGE X ~
FL YWHEEl) X .
HYDROGH X .
CHEMICAL :;IORAGE X ~
SUPERCONIJUCTING MAGNETIC X .
STORAGE
Figure 3. Availability of energy storage.
monoxide (GE HyCO). This latter concept. being ex-
plored by ER[lA and the author's company follows the
reaction:
HEAT
CH4 + CO2 ~ 2H2 + 2CO
with the sponHneous absorption of a large endothermic
level of reactbn across a heat exchanger at 7000 to
1.000° C.
The systern is a closed-loop scheme, allowing the
endothermic rEaction to occur near/at the heat source
and the exothl!rmic reaction at the end use point via a
catalytic reactor. Gases used are easily handled by cur-
rent hardware. A schematic of this system i.s shown (fig.
4).
Batteries rI'present the only perfected method avail-
able for bulk e~ectric storage. They have been deployed
in many fixed and mobile applications. Their major
limitations are weight, capacity, and life. Their very
nature of operation requires that modularity (multi-
plicity) be a dlisign parameter for any specific applica-
tion. i.e.. number of plates, p!p.ctrolyte Quantity/cell.
number of cell!. etc. Table 12 r ~viewed current battery
technology and future potential battery materials con-
figurations with commensurate estimates of performance
and cost projections. It is generally believed by devel-
opers of advanced electrochemical (battery) systems that
a decade will be needed to prove the stated potentials of
the leading near-term concepts. thus limiting electrical
storage capability to present lead-acid based hardware.
3. Mechanical Storage-Flywheels and Compressed Air.
Mechanical energy storage includes pumped-hydro.
steam, flywheels, and compressed air. Pumped-~ydro
systems have been developed for both underground and
conventional uses. Siting considerations preclude prac-
tical application in an industrial setting for anyone
industry.
Flywheels have the inate ability to provide"mechani-
cal and electrical energy. Using appropriate linkages.
differentials. rotation. or oscillatory (piston) power can
be derived. By inclusion of coils and inductive pickoffs,
a.c. or d.c. voltage can be generated. Small flywheels
have been placed in cars, small buses, and commuter
trains, indicating the power and portability intrinsic to
flywheel usages. Flywheels have been used for emer-
gency power backup aircraft carrier catapljlts, and to
store wind power-generator output (in the U.S.S.R.).
Basic limitations facing use of the hardware are related
to materials and structural design. Most known flywheels
446
-------
Table 12. Thermal energy storage devices
~
~
-..J
Class
(Latent/ Economic
Device Sens.) Principle of Operation Application Considerations Materials/Components Reference
Energy Storage L Utilizes latent heat of paraffin The construction of the - Paraffins group C10H22 to A
Panel hydrocarbons contained in an storage panel will per- C35H72 which display la-
aluminum honeycomb struc- mit its integration with tent heat in the range of
ture. (phase chan!\; temps be- a wall or roof of a 60 to 120 Btu/lb and melt-
tween 70 and 130 F) building utilizing solar ing temps between 70 to
energy 1300 F
Eutectic L Uses crystals of sodium Heating-Cooling of - Na2 HP04, . 12H20 seeds. A
Reservoir phosphate to prevent super- buildings
cooling of a sodium phos- Container of 40 to 80 gal
phate solution at the bottom volume cylinder less than
of a stratified tank with the 28" dia.
upper portion at maximum
temperature
Groundwater L Stores high quality byprod- Steam turbines - Water A
Basin uct heat by injection of
heated water into a ground-
water aquifer and is later
withdrawn
Concrete L Thermal energy is stored as Power station Initial cost of $50 Concrete structure 300' x A
Container latent heat of fusion of a salt million-1>nly 10% of 300' x 50' deep, contain-
the capital invest- ing a salt
ment required for a
1000 MW nuclear
power station
Hydrogen Storage L A metal hydride is placed on Solar house heating Container thickness Metal Hydrides such as B
Container a roof in a sealed f\a t metal for higher dissocia- LaNis, SmCosH2.s
container connected to a hy- tion pressures can FeTiH1.o
drogen storage tank. When become too large for
the hydride is heated, hydro- practical use
gen gas is dissociated and
stored in a basement reservoir
(2000 F)
Heat Pipe L Heat addition melts LiF. Solar collector thermal - LiF inside a sealed tube C
Heat extraction solidifies power system cluster
LiF
Heat Storage S Stores hot water in a Stores heat energy from $60/KWe for reser- 2 x 10Sm3 tank for a D
Reservoir stratified tank a solar pond voir and $2.50 m3 10MWe system
for excavation and
sealing
-------
Table 12. Thermal energy storage devices (con.)
~
~
CIO
Class
(Latent/ Economic
Device Sens.) Principle of Operation Application Co nsfderat ions Materials/Components Reference
I I
Therm.Bank* L Stores heat via sensible and Stores e lectrici ty in the $1.50 per thousand Welded steel vessels filled E
*(C&W latent state in a mixture of form of heat. For do- Btu for a 0.42 million with anhydrous sodium
Tradename) inorganic chemicals e&.lled mestic heating and can Btu unit when manu- hydroxide
Thermokeep. be applied to commer. fact ured in large
(350-600° F melting) cial and industrial space quantities
(250-900° F operating) heating. Intended to
shave peaks off electric
demand profile
0-
rES Regenerator L Air passes thru the device con- Suitable for use with hot - Corrugated plate fins sand- F
taining phase change materials, air heat distribution wiched between pans filled
discharging or receiving heat system with TES phase change
material
Imbedded L The heat transfer surface is a The unit is suitable for More expansive than Aluminum sheets spaced G
Refrigerant stack of 42 horizontal alum i- use in air conditioning sensible heat storage one inch apart with TES
Module num sheets spaced one inch cooling storage systems or conventional refrig - phase change materials
apart with imbedded refriger- in space heating systems eration unit. Lower sandwiched in between. -
ant passages. The refrigerant with suitable choices of space requirement Aluminum sheets contain
is in direct thermal contact material justifies further refrigeran t passages
with the phase change TES development
material
Steam Accu- L Stores heat in the form of Possible use in solar- $100/m3 for standard Standard pressure vessels H
rnulator steam for off-peak use thermal electric power vesse I
generation
Brick or Gravel S Air enters a rock imbedded Hot air heating systems $1.00/1000 Btu Rocks, bricks, stones im- I
Bin cavity to store or extract heat to reduce peaks in bedded in an enclosed
to/from ceramic like materials electric demand cavity
Shell and Tube L A certain percentage of tubes Domestic space heating - Shell and tube type ex- J
TES in the tube bundle of a shell systems changer with phase change
and tube type heat exchanger materials filled in certain
are filled with TES material tubes
Heat Storage L Extracts heat from heat stor. Liquid vaporization or - Storage medium composed '0 K
Heat Exchanger age container of a phase change rapid extraction of heat ofalkali metal hydroxide.
material which solidifies at the from storage Eccentrically wound coil
lower end of a heat storage cy.
cle, comprising a conduit tra-
versing the heat storage material
and connecting an inlet and out.
let via an open ascending coil
having turnings ofat least two
different diameters forming a
multilevel alternating distri-
butlon of larger and smaller
diameter turnings
-------
Table 12. Thermal energy storage devices (con.)
NOTES FOR TABLE 12
A. Smithsonian Science Information Exchange, Inc.,
Research I nformation Package 1 B31-Energy Stor-
age, April 1975.
B. G. G. Libowitz, Met1J1 Hydrides for Thermal Energy
Storage, Allied Chemical Corp., Morristown, N.J.,
Paper #749025 presented at IECEC Conference,
August 1974.
C. R. Richter, Solar Collector Thermal Power System,
V 0 I u me III, Xerox Corporation/Electro-Optical
Systems.
D. A. F. Clark, J. A. Day, W. C. Dickinson, L. F.
Wouters, The Shallow Solar Pond Energy Conver-
sion System: An Analysis of a ConceptJ.Jal 10-MWe
Plant, Lawrence Livermore Laboratory, January 25,
1974.
E. Design, ConstrUction, and Testing of a Pilot Model
Heat Storage Unit Based on the Heat-of-Fusion
System, Final Report, Comstock and Wescott, Inc.,
July 1963.
F. Solar Heating and Cooling of Buildings, Phase 0,
Feasibility and Planning Study Volume, Final
Report, General Electric Company, Document No.
74SD4219, May 1974.
G. H. Yeh, Conservation and Better Utilization of Elec-
trical Power by Means of Thermal Energy Storage
and Solar Heating, Phase III Report, University of
Pennsylvania, March 1973.
H. Solar Thermal Electric Power Systems, System
Studies and Economic Evaluations, Final Report
Volume 2, Colorado State University and Westing-
house Electric Corp., November 1974.
I. Conservation and Better Utilization of Fuel, Univer-
sity of Pennsylvania.
J. M. Altman, Conservation and Better Utilization of
Electric Power by Means of Thermal Energy Storage
and Solar Heating, University of Pennsylvania,
October 1, 1971.
K. U.S. Patent 3,452,720, "Heat Storage Heat Ex-
changer," W. T. Lawrence, to Hooker Chemical
Corp., Niagara Falls, N.Y., July 1, 1969.
are of steel construction, yielding relatively poor energy
storage/pound of material (fig. 2). Although, there have
been significant improvements in steel flywheel energy
density capacity and efficiency, the hazard problem is
serious enough to result in derating to a fraction of their
theoretical maximum performance (usually 50 to 60 per-
cent derated). The next generation of hardware will
utilize new structural design techniques and aerospace-
developed material composites. With tensile strengths of
300,000 psi and a low weight/ma,ss ratio, the safety
aspects are radically improved. Among candidate mate-
rials are Kevlar, fiberglass, bulk glass, wood, and carbon
products. Rotational speeds, to produce -30 wh/lb are
below 3,600 RPM, well within bearing, lubrication, and
seal state-of-the-art.
The flywheel has a unique characteristic in that its
power density can be tailored to rotational speed and/or
demand levels and duration. The flywheel is an ideal
absorber of off-peak power where in an in-plant genera-
tion installation, thermal needs can exceed electrical
load demands. Electrical generation can be continued to
produce the needed heat,with the excess power fed into
the flywheel reserves for later use in any power profile.
A typical system configuration utilizing flywheel kinetic
storage is shown (fig. 5).
Compressed air storage for power is commonly en-
visioned for power tools, pneumatic lifts, and for engine
starting. Man-made gas storage vessels (as opposed to
caverns or abandoned mines) are less demanding of siting
restrictions relative to environmental impact; however,
they do require substantial volumes (hence above-ground
appurtenances). The primary use for this storage scheme
would be for augmenting peak power periods of say 2
hours with a recharge rate of about 8 hours. The trade-
off here involves sizing of compressor flow to recharge,
stored pressure, tank size and, of course, peak demands.
A schematic of a compressed air storage system is shown
(fig. 6).
I n essence, all mechanical storage schemes are best
suited to offsetting peak-load periods of relatively short
duration (-4 hours). The main advantage is, being able
to use off-peak power hence lowest cost power. Massive
use of this concept allows upwards of 50 percent reduc-
tions in power generation equipment design and com-
mensurate savings in transmission and distribution net-
works. '
Applicable Data of Industrial Energy Usages
The energy data base for each major using sector is
nowhere more nonaggregated and/or nonstandardized
than for industrial consumption. Residential/Commer-
cial energy usage is largely monitored as a byproduct of
utility loads tracking for purposes of anticipating load
449
-------
[}1~ ~ ~::~~:<
HIGH GRADE
HEAT
PIPELINES
TURBI NE
CONVERSION
TO ELECTRI CI TY
PRIMARY
ENERGY
SOURCE
INPUT
CATALYST
REACTOR
P-V WORK
SYNTHESIZER
AND PUMPS
OUTPUT
CATAL YTI C
CONVERTER
LOWER GRADE
HEAT
Figure 4. Schematic of a closed-loop system.
l
FUEL] ELECTRON! CALLY
r-- CONTROLLE D SWI TCHI NG, r+-
STORAGE OR USE OF
KINETIC ENERGY
60 Hz AUXI LI ARY FL YWHEEL KI NETI C
GENERATOR GENERATOR/MOTOR STORAGE PM MOTOR
- - OR GENERATOR,
ENGINE: - SUPPLYING - INTERFACE WITH DEPEND' NG ON
51 TE ELECTRI CAL KINETI C STORAGE
PO WE R NOMI NAL 800 Hz MODE. NOMI NAL
SPEED 25,000 RPM
1- 60 Hz OPERATI NG POWER
PEAK LOAD DEMAND SI GNAL
FOR 51 TE APPLI CATIONS
REJECTED HEAT TO
SI TE APPU CA TIONS
Figure 5. Onsite power generation with kinetic storage for handling peak loads
(typical basic configurations).
450.
-------
HEATER (OPTIONAL)
WORK
TURSI NE
(EXPANDER)
MOTOR-
GENERATOR
STORAGE
COMPRESSOR
Figure 6. Compressed air storage-based energy systems.
patterns and peaking trends on their systems. As primary
residential and commercial energy use is tied to space-
heating and hot water functions (about 75 percent).
climate and season playa major role in regional con-
~umption. The industrial plant uses only about 5 percent
~f its energy demand for these functions (for employee
support). Production functions utilize the overwhelming
amounts of energy. These effects are easily seen in a
graph of energy consumption of each sector (fig. 7).
Conspicuous for the residential and commercial sectors
are the winter and summer peaks (heat and cooling)
essentially absent from the industrial plot. With utilities
providing most of the energy used in the residential and
commercial sectors there are sufficient (though impre-
cise) methods to derive energy use patterns. Industry
purchases large amounts of coal, gas, oil, and electricity
and internally apportions their uses to the unique fea-
tures comprising their individual process energy systems
design. These uses and their resulting energy demands
are shrouded in secrecy to protect competitive positions
or proprietary processes.
Rough estimates have been made of energy use in
certain industries. Percentages for each generic form are
shown (table 13). This provides only a surface view of
industrial requirements. More data are clearly necessary
to identify potential energy conservation candidates. A
major reason for the lack of in-depth data is that, until
relatively recently, energy use was not the political and
technical stewpot that marks current concerns, hence
ac;:countability was haphazard or nonexistent because,
~nergy sources were reasonably cheap, inexhaustible,
-. . .-
and readily available. For these reasons, no major statis-
tical records were kept in a form allowing system anal-
ysis. Known data sources for industrial energy are tabu-
lated (table 14).
While not purposely attempting to project an overly
pessimistic picture, a data base, even if available, would
catalog processes and energy flow patterns numbering in
the thousands. Details of how each energy source is used
and its conversions, such as from oil to steam to elec-
tricity, condensing and flash evaporation functions,
further camouflage and defy uniform approaches to
developing industrywide, summary statistics useful to
research and engineering firms for synthesizing energy
storage system applications. Thus, the current roster of
government-sponsored energy-conservation studies in-
cludes tasks covering surveys, questionnaire mailings,
interviews, and the like to overcome the dearth of infor-
mation and acquire needed data for R&D priority-setting
efforts. In many of the interviews conducted thus far in
the ERDA-sponsored thermal energy storage feasibility
study now underway, many respondents indicate that
(1) the data are not available, (2) no program exists to
decipher their plants energy uses, or (3) no dedicated
energy analysis personnel are employed and that the job
is relegated to a catch-all individual, i.e., industrial or
plant engineer. Unfortunately, until a centralized data-
reporting system is designed and formulated to protect
proprietary information and/or production strategies
linkable to energy use patterns, analyzing energy uses
dynamics by other than company personnel is at best an
intelligent guess and not likely to result in new equip-
451
-------
80
United State.
80
75
-------
70
.,
.0 -----
------_._- ----------
tUI ---------
---_...-_.._-- ------
~
~ !l0 _.
X
f. 4~ ._-----
C
~
.J 40
¥
..
~ 35
Z
3 30
.J
cD
-~
20
-----'--'.--'.'-
t5 ----
. ----.
10 ----
II -----
o
o
.li..Ll
LLLULU
II
J
s
o
s
o
14
J
14
J
18.11
19118
11170
---.--
------- -. 75
-------
70
_.+----
. -._-_.~---
---- ----- 6~
-.-..- . -.-----.- .....-----
.._-. -.. GO
...-- --.0.__'
....-. 55
(II
a:
.. ----- ~o ;)
o
X
45 ~
0(
:t
.- 40 ~
"
.
3!\ 0
(II
",1
o
30 :;
,I
iii
-------'---"'-- - -.- ...--------- ---
-------- ;'0
.-- ---.-.
--- 15
-...---
--------- 10
----
.---- --- ---- 5
5
II I I I I II
o 14 J 5
J
!i
o
o
s
o
o
""
J
""
11171
11172
1973
Figure 7. Industrial, commercial, and residential demand--historical.
ment designs due to uncertainties of performance re-
qu irements.
Industry Thermal Demands and TES
Several TES applications have been identified in the
ongoing ERDA-GE TES study. These applications are
independent of feasibility studies of technology require.
ments. A revit!w of industrial processes is performed to
develop TES potential not to judge the risks, invest-
ments, or enurgy/cost factors. The volatile nature of
energy flow precludes ruling on the ultimate merits of
these and other applications yet to be conceived. The
final recommendations will, however, rank TES applica-
tions in a relctive matrix of priorities considering eco-
nomics, energ( impact, and StCHe-of.the.art aspects of
. each scenario.
Energy demand r:atterns (hour-by-hour load
profiles) are a critical determinant of the feasibility and
desirability of thermal energy storage systems. Load
profiles are required both to size thermal energy systems
and to quantify operating cost savings associated with
the resulting load curve. One of the biggest obstacles to a
systematic identification of the potential for increased
use of thermal storage systems is the lack of organized
data. There is no lack of raw data; rolls of strip charts
and file cabinets filled with circular charts are not in a
form directly useable for systems studies. To develop
"typical" daily profiles for different time periods
(monthly, seasonal, etc.) and regions of the United
States, it is necessary to analyze the raw data, filter out
the special cases, and identify typical energy demand
patterns. Probably because load profile information is of
little use in industry except for in-plant generation
studies and troubleshooting exercises, very little has
been done in the way of data analysis and categoriza-
tion. By way of examples, several industrial settings are
presented below considering steam demands. Steam
production consumes some 30 percent of industrial
452
-------
Table 13. Industrial energy use by function percent distribution of industrial energy
Industry
Paper and allied
products
(SIC 26)
Chemicals and
a'll ied products
(SIC 28)
Chlorine/causti c
Pet rochemi ca 1
Petroleum and coal
products
(SIC 29) Reentry
Primary metal
(SIC-33) Steel mill
Food and kindred
products
(SIC-20)
Stone, cl ay, and
glass
(SIC 32)
Process
steam
76-89
36-46
53
29
35
28
18
40-80
1
Di rect
process
Space Purchased In-plant Mechanical
heat electricity generation drive
6-9
5-6
6
1-2
6
45-50
5
6
4
4
61
31
5
16
5
60-90
66
87-96
61
3
2
8
20
3
1
4
1
2
1
0-6
12
12
35 - 45
1
86-93
7
7
1
0-5
Sources: 1. General Electric Engineering Consultanting Services -C.M. Young;
space heating was not separately accounted for. 2. General Electric Industrial
Sales Division. As indicated in reference 3, specific plant types were consid-
ered. The energy use levels reflect the operation of new plants being evaluated
and installed for operation in the late 1970's and mid-1980's. Space heating
was not separately accounted for. 3. Westinghouse Corporation. As indicated
in reference 4, a detailed study of all industries was performed. Direct proc-
ess heat includes mechanical drive and all energy uses above 4000 F.
453
-------
Table 14.
Industrial process energy bibliography
1. .-\ Study of Process Energy Requirements in the Chemical Industry:
American Gas Association. Arlington, \"a., Catalog ='0. C20006
2.
Jnd\l~tri3.1 Ene:-;;y S~::c:o ':of :~e P~a..;::ic::. artd Ru.bber Industries, SIC's
2S2 and 30. Foster:O;. Snell, Inc. for FEA, :O;TIS :\"0. PB236211,
~Iay 197-1
3. Industrial Energy Study of the Petroleum Refining Industry, Sobo!tca
and Co., Inc.: for FEA, ~TIS :0;0. PB235671, )Iay 197-1
4. Industrial Energy Study of the Drug )Ianufactu ring Industr ies:
Versar, Inc.: for FEA t:. S. Dept. of Commerce, XTIS ~o. PB235994,
Sept. 197-1
.:s.
-~ Study of Process Energy Requirements in the Paper and Pulp
Industry, American Bas Association: Catalog =C~0003.
--\rlingron, Va.
'3. .-\ Study of the Xon-Ferrous ~Ietals Industry, American Gas Asso-
ciation, Catalog :\0. C20010, Arlington, Va.
7. Barrell, K.C; "The )Ianufacture of Portland CemeM," Series of
articles in Cement. Lime, Gra\'el. January, June - .-\ugust, 1971
~
~
S. "Carbonated Be\'erages" in ECT 1st 00., Vo!. 3 pp. 112-124, by
W.T. )!iUer and J.F. Hale, :O;ehi Corporation
9. ~1.B. Jacobs, Manufacture and Analysis of Carbonated Beverages,
Chemical Publishing Co., ~e\\' York, pp. 89-109, 2Zl-228.
Richard L. HaU, Food Techno!. 15,22-26 (Dec. 1961): Food Process.
22,27-28 (Aug. 1961): Whar~ ~e\\' in Food and Drug Research S,
(July 1961).
10. Reiberna/Halcro\\'. R: l'.S. Energy and Fuel Demand to 1985: A
C'omposite Projection by rser within PAD Districts. l"nh"ersity
of IlIinois, Report :;;0. CAC108R, ~Iay 197-1
11. Thermo Electron Corp., Waltham, ~Iass. Solar Energy for Process
Steam Generation Jerry P. Da\'is 25 :0;0\'. 19n 40 p ref Sponsored
by r\SF, Washington, D.C..
12. The 1973 Fuel and Electrical Energy Requirements of Selected
Mineral Industries Acth'ities. Bureau of ~Iines. \\'ashington. D.C.
Dh'. of Ferrous Metals.
13. Research on the Application of Solar Energy to the Food Drying
Industry Progress Report, California Polytechnic State Cnkersit}',
San Luis Obispo. 1 JuJ. - 30 Sep. 1974
14. The Impact of Energy Shortages on the Iron and Steel Industries.
Booz-Allen, and Hamilton, Inc., Bethesda, ~Id. Final Report,
Aug. 1974
15. "Industrial Energy Study of the Concrete, Gypsum and Plaster Pro-
ducts Industries," Stanford Research Institute, for FEA, :;;TIS
No. PB-237833, August 1974
16. Bobo. D.L. et ai, ..:\ Sur':ey af Fuel a;1d Energy IrnoTr.-:ation Sources,"
Volumes I & II, ~Itre Corp. :\"TIS :\"0. PB-197 396 7. :O;o\'emhpr 1 Q70
17. "Data Base for L'te Indus,rial Energy Study of the Industrial Chemicals
Group," IR and T, for FED, NTIS :0;0. PB-~379~5, September, 1974
19. Bullard, C.\V. and Herend:Jm. R.A., "Energy l"se in t.~e Commercial and
,Industrial Sections of !he l".S. Economy, 1963." l"niy. of nlinois,
t"rbana, :\"0\'. 1973
19. Institute of Gas Technolov.y, Efficient t"se of Fuels in Process and :lIanu-
facturing Industries, S:"1":1PQsi::r.1 sponsored b~' IGT, Chicago, nlinois,
AprillG-19,197-1
20. Institute of Gas Technology, Efficient l"se of Fuels in the ~tetal1urgical
Industries, S~~po5ium sponsored by IGT, Chicago, illinois, December
9-13, 1974
21. "Study of Industrial l'ses of Energ:y", IGT, for EPA, ~TIS :0;0. 'PB-237
215, July 1974
22.
"Potential for Energy Conservation in ~one Selected Industries - The
Data Base." Gordian Associates, for FEA. June 1974
23. 1972 Census of ~lanufact1Jrers Inde.x Special Report Series, Dept. of
Commerce GPD, Washington, D.C.
24. EEl Historical Statistical Yearbook (to 19701 plus Al1nual Issues (to 1973),
Edison Electric Institute, ~ew York City .
25. Liu Y.H. and Jordan, R.C. Long Term A\'erage Performance of Flat Plate
Solar Ellergy ColJectors, Journal of Solar Energ)', Vol. 7, :-10. 1963
26. Powell J.C., etal; Dynamic Con\'ersion of Solar Generated Heat to Elec-
tricity, r\ASA CR-134 72';, August 1974
27. Energy Consumption in ~1anuiacturing: The Conference Board (Cambridge,
Mass.), Ballinger Publishing Co. 197-1,520.00
28. Energy Prices, 1960-1973, Foster -~ssociates, Inc. (Cambridge, Mass.)
Ballinger Publishing Co. - 19i4
29. Intra-Industry Capability to Substitute Fuels: Science Communication.
Inc., ~lcLean, Va., October 1974, ~TlS 1(0. PB-237 605/1WE
-------
Sr.i 1000 LBS/HIt
ELEC t:W
BLR 1 BLR 2 BLR 3 BU 4
400 PSIC
4.
~I ctJ
4.
l~v -0
MRV PRV H-M
I 8
I
~ ~
C1I MAKEUP
C1I
9 ~
PRoe. . S1M
I l2~. l
-8 DS 0.1 .
-L-.- _.J
ctv AUX BLQ'.l DN
5 PSIG ~
20 OZ. PSIG
NORTH AV!.
, 138. ; OLD PLANT
GEN. 24.2
PURCR. 38.0
CASE OW~ COST 2227 $/HR TOTAL S'IH 594 TOTAL ELEC. 62.2
Figure 8. Typical winter steam and electric loads--GE Schenectady plants;
-------
energy demand and represents a major component of
process enerqv impact where thermal storage can be
effective.
Steam Demand Patterns at the General Electric
Schenectady Plant
General Electric's Schenectady plant is a large
industrial site where several company divisions are locat-
ed. The work ~orce consists of a mix of manufacturing,
engineering, product development, research and develop-
ment, etc. Principal products include gas turbines, steam
turbine-generntors, electric motors, and other electrical
equipment.
Although the GE Schenectady plant is probably not
a typical industrial facility, an energy-use-data-gathering
effort was un.:fertaken. One of the reasons was to estab-
lish problems encountered in collecting and analyzing
raw data while another was to test the hypothesis that
most industri,ll loads are flat. The Schenectady plant is a
three-shift opE!ration 5 days a week.
The GE Schenectady plant is actually an example of
in-plant generJtion where about two-thirds of the elec-
trical energy is purchased from Niagara Mohawk and
one-third is !Jlmerated on site. The peak electrical de-
mand at the plant is about 60 MW. All of the plant's
process heat and space-heating requirements are met
with onsite boilers. A simplified diagram showing ther-
mal loads and flows is shown (fig. 8). As illustrated, the
thermal energ'l source is actually four oil-fired boilers,
two rated at 300,000 Ib/hr and two at 200,000 Ib/hr. As
indicated, there are four primary steam lines, 400 psig,
160 psig, 5 P! ig, and 1.25 psig. Although some of the
400 psig steam is used for process heat, most of it is used
for power ger:eration. Intermediate-pressure steam for
process and low-pressure steam for space heating are
made available by using automatic extraction turbines.
Although it vlOuld have been desirable to determine
. process-heatin£ loads at the various pressure levels, the
required data-:)rocessing effort made this impractical.
Because of thn number of meters at various end use
points, calculation of total loads means aggregation of
many hourly d~mands from raw circular chart data. Not
all of the raw dai:a are stored at one central source.
Total hourly steam demands (400 psig) for the
Schenectady plant were calculated by adding the out-
puts of the four oil-fired boilers. Hourly steam demands
for a summer day, winter day, and spring day are shown
(fig. 9). It is important to recognize that these demand
profiles are not necessarily t~'~;,:,::1 profiles, since no
sampling and dilta analyses were tJerformed. One of the
problems is thE' fact that tntal demand information is
not readily available; this was developed by summing
steam output levels for each boiler. The demand profiles
are interesting in that they show significant hour-by-
hour and seasonal variations.
It is difficult to explain the specific load variations.
Again a significant data analysis effort, correlating steam
charts to power charts, is necessary to establish some of
the cause-effect relationships.
Energy Demand Patterns for a Film-Manufacturing Plant
Typical daily profiles of process steam and kilowatt
loads for a large photographic film-manufacturing plant
were tabulated. Steam demands for typical factory days
and for weekend days for each of the four seasons are
plotted (fig. 10). This particular manufacturing opera-
tion has reduced second and third shifts and weekend
production.
As indicated (fig. 9), the hourly load variations are
in the 17 to 28 percent range. The variation of loads
with season is relatively small. As previously indicated, a
review of the literature demonstrated that hour-by-hour
industrial process heat demand information has not been
widely documented. The most complete source of in-
dustrial information happens to be a book on the subject
of steam storage published in German in 1933 and
revised in 1970 (ref. 3). The subject matter of the book
deals exclusively with the design and operation of
industrial heat accumulators which like district heat are
rather common in Europe but seldom found in the
United States. Goldstern presents rather detailed discus-
sions of steam storage installations in several industries.
Much of the steam demand profile information is
reproduced (figs. 11 through 15).
Goldstern indicates that historically the interest in
steam accumulators focused on large accumulators,
especially in the electric utility industry. As steam pres-
sures increased, however, the application of accumu-
lators became more and more difficult, and they became
regarded as obsolescent. The recent fuel shortage has
given a new impetus to steam storage. Today, however,
the question is "How small can the accumulator be, and
at what minimum pressure can it still perform its func-
tion 7"
Thermal Energy Storage has also been examined for.
possible applicability to industries and functions other
than solely for steam uses. A review of these industries
or functions foll~ws.
In-plant Generation. Industry will be turning to
methods of providing its own electricity with increasing
fervor with the continued increase in fuel and electricity
costs. Applicability for inplant generation (IPG) is predi-
cated on an industry having significant heat require-
ments to capitalize on the recoverable heat rejected by a
power turbine-generator. The obvious advantage is that
456
-------
.00
lOCATION MIDOH AtlANTIC R(!~tON
100
1 AuGUST
1 JANUARY
) ...PAll
~
"2
~ 400
~
~
'lAN'OP1RAIIOHS JSHI.'S..0AYS
'00
SYStfM 'DURnll fIRtO.DIURS
''''.XIC.JI<. 110'11114'0011 lbo"',
OUT""-J1 400 PSI Sli AM fOR
HfCIHt('", 'HUIM..t lOADS
..1----L
. 00
Ill! IHIt:t.I~dHAllnN 'JINf'lANT
'1 ,,.UMCtiASiD
"'All. UI(IRK 110'"
KtAIiNG DlGRIt 1000
.1.- .1COOL'.NG I J.~~L..L. J
! . .5 . , . . 10
I
Q
o
Q
-1- -L-L-...L -1- _~_-L - I
I , . . , . , .
I
"
NOI)N
.1 -'
,
Figure 9.
Total steam demand: GE Schenectady plant,
location: Middle Atlantic region.
W((KDAY
WHltiNO
I JANUARY
).vAIL
]-JULY
4OCTOILIR
.
.
c
o
...
~ '00
"2
~
~
~
..~ "
, ......_._-_.-...'...'>;;;:~::~:_;:,;"g~~>?,._..._.._,
" .
...
1'1 ANT {)PiAAIION fUll FlRSI SHlf'
AtOtlUU SfCONOfTHIAOfWl(lI:IND SHU 15
H(ATING DIQRIE( DAVS-eaao
Hf. A IIN'" OH.RH DAVS-.xm
COUlLN{iOfGfUt DAVS'IJOO
_L_.I
,
I --L_I-L-.1..-..o -1- I
. . , . . III If I'
"
,
L .-1-.-1
. , .
I
00
NOON
Figure 10.
Film manufacture - SIC 3861, location:
Mountain region.
457
-------
It
...",,,,
" II U
Figure 11. Steam consumption of a sugar
factory.
1D~ '
.:"~-~- ~~-i-- -! '!--l -
: --:= 'J~~ - i~ij~!r~i~ :1,- J~~r,- ~r" - ~~-
~ 6 ,If j'-I-)-+1.j--f- .+-
r n fii \ !IV - li',!,!,~ i -. J.-s~~=;'L_- :.,
. J I~ Ii: i - \J !'iJiU11U! Ill~~ _J ( I
, - -I' ~-F--'i- -f -(~J{II .--
: - ~n-I-t=-I---I~- !~ -
4 , 6 70 II ,. 16 ;8 h, 13
Timll 01 day
Figure 1 :!. Steam consumption of a textile
\/\Iorks in winter ~nd summer.
458
I) .- -I\~
- \ -) - -\: -\
S
tOft/h,
10
]1$
t
~
£
: 10
C;;
s
,
I
IIJ
U 18 ZZ
Time 01 d.y
1 h, G
Figure 13. Steam consumption of a paper-pulp
factory.
JO
ton/hr
I
I 1-
.J
1- - - - -- --
j _J- .,- .- I .-
,-- - ~ u
I I. L
. I .- -. -
- . - i T-; -:- r
I
-- 1 --, -:v - t '~-: I --
I
- --. - -1 i fi ~ -t r~~
I-jlt -'I- i
~Lc-- -,li- I i
~
I ~
--
- - I
- --.-
16
14
1Z
10
~ 18
'II
2'6
'b
e"
..
~ 1/
II)
10
8
G
.
z
Os I 7 8
J 10 71 11 1J ,. 15 16hr 11
Time of day .
Figure 14. Steam consumption of a carpet
factory with laundry.
-------
, 1D
n/hr
,
-
"
..
E I
..
..
"5
~
.. .
E
..
..
;;;
Z
D 1D
11
11
V 7f
Timil 0' dil"
11
76 hr 71
Figure 15. Steam consumption of a
carpet factory.
with lPG, a plant can offset its purchased electrical
power and recover the exhaust heat for steam or other
heat processes previously supplied by a boiler steam
system, still required but with an added energy use.
Major constraints to IPG uses are the independently
varying electrical and/or heat loads causing mismatches
,[I availability. Utilizing heat storage as Dart of a solar
SOLAR THERMAL
SYSTEM
FUEL
BOILER
ACCUMULATOR
FEEDWATER
PUMP
thermal system, heat load following is possible and
boiler fuel consumption is offset by the waste energy
when available or augmented by a solar-thermal system
(fig. 16).
Thermal storage for IPG can be applied at either of
two points in an advanced system. Turbine exhausts,
usually designed for process steam uses, are fed to the
TES at temperatures of up to say 5000 to 6000 when
plant steam demands are down. For a system using a
solar-thermal energy system, or where an alternate waste
heat source exists, these heats can be stored and used to
supplement boiler power to the turbine if high tempera-
ture recovery is planned, or as a feedwater preheater.
One such system is now under investigation by GE and a
major oil company. A hot oil (-5000 to 6000 F) is the
heat store used to preheat a staged extraction turbine,
recovering exhaust heats to maintain oil temperatures.
For lPG, thermal storage provides load following or.
matching where sudden heat demands exceed IPG nomi-
nal heat output limits.
Bricks and ceramics-Bricks are relatively' simple-
0:byproducts to manufacture; however, large amounts of
. 0 0
heat over the range of. from 100 to 1,900 Fare reo
quired. There are two types of processes, both following
ELECTRIC
GENERATOR
THERMAL
ENERGY
STORAGE
Figure 16. Input generation with solar-thermal system.
469
STEAM TO
PROCESS
RETURN
-------
similar func1ions but with different operating schemes.
The modern continuous (or tunnel) kiln is automated to
dry, fire, an(1 then cool the bricks in a present schedule
(timing cycln). The bricks enter the kiln (fig. 17) and
undergo high air flows and slowly increasing heat (100°
to 400° F) to drive off the moisture contained. The
timing is extended to allow the entrained water to
evaporate slowly, so as not to form steam within the
brick, hence forming voids, which weaken the structural
strength reqllired. This takes several hours, after which
the heat rate is increased at lower air flows to 1,890° F,
and bricks arc "soaked" until the desired hardness is
achieved. Cooldown proceeds more quickly than the
heating porti:m until the shipment exits the kiln. Clean
heat must be used during the heating cycle to prevent
impurities from penetrating the soft bricks and causing
adverse cher~,ical reactions leading to cracking and
imperfectiom of the final product.
The secc.nd method uses a periodic kiln (fig. 18).
Temperatures are as described for the continuous kiln;
however, the time cycle is significantly longer due to the
types of bril:ks produced with this method. As the
figure indicates, a 85- to 90-hour heatup period is
required. A specific temperature rise-time relationship is
followed unti I the 1,800° F level is reached. A 4-hour
"soak" insur£s that all bricks are uniformly heated be-
forE! cooling i! initiated.
STACK
31,000 CFM t
.3S0-S00°F
1:
I :
1'------ -----,
.L~_- ------, I
I '
I I
I I
--
COOLER
Stored heat may be applicable for the controlled
heating phase required to dry the bricks prior to kiln I
firing up to the cure temperatures. High gas flow rates
(18,000-31,000 CFM) are initially applied to carry the
high weight of removed moisture from the bricks,
stacked in roughly 6-ft x 6-ft x 6-ft cubes on dollies.
TES potentials abound in brickworks plants, partic-
ularly where periodic kilns are used. As shown (fig. 19).
kilns operate over a wide temperature range and have
equally wide heat flow requirements. Brick and ceramic
drying and cure (firing) cycles are time-temperature
critical to dry residual moisture before using the prod-
uct. Recovery of convective and conductive heat losses
by a TES can supplant dryer heat cycles and "bias" the
rigid heat profiles in lieu of excess air controls, an
energy-wasting method of thermal control. One such
scheme is diagrammed (fig. 20).
Iron and Steel
The several stages of ironmaking and steelmaking
have differing process heat requirements. In addition,
there are several batch-type operations identified.
Coke Processing This is a batch process where coke
ovens operate in -20.minute cycles until a supply of
coke is processed for the day's steel foundry operations.
The American coke ovens quench the baked coke and
TO DRY~R
, 300-400°F
,- - - - - - -, '- - -'
I I
I
I I
L,,~'- - - -.-I~~I
t
1B90°F
--
COOLER
INLET
AIR
INLET
AIR
. COMBUSTION GASES USED TO PREHEAT INCOMING LOAD
. REMAINDER OF HEAT CONTENT OF COMBUSTION GASES TO STACK
. (CANNOT BE USED IN DRYER eECAUSE OF IMPURITIES)
. COOLING AIR IS USED FOR DRYING PROCESS
Figure 17- Continuous kiln.
460
-------
PERIOOIC
KILU
lOAD
BRICKS
~--~J:L
~
TO
STACK
FIRING
AT COOl.! NG
1800°F
4 HOUR
SOAK
lm
)0 DRHR
85-90 HOURS
EXHAUST AIR TO SlACK
Figure 18. Periodic kiln.
terminate burning. The Russian process (fig. 21) uses the
hot coke to create steam for plant use, thereby recover-
ing the sensible heat. A further refinement would be to
place the hot coke into a sealed receiver, where through
oxygen starvation the combustion reaction will termi-
nate, a,nd, being situated on a TES (fig. 22), the heat is
recovered for preheating the next coke oven charge.
Blast Furnace. In blast furnace operations, several
potentials for heat recovery exist for a TES. I none
phase, the slag is periodically removed from the continu-
ously fed furnace. The slag heat content can be stored
for use to preheat the scrap charge, annealing ovens,
maintain BOF temperatures between charges, or preheat
coke before entering the coke ovens. The second use
considers the effects of injecting oxygen (- 6 percent)
into the blast furnace to increase temperatures to 3,000°
F. Frequently, the furnace temperature overruns and is
controlled by excess air. A TES can act as a thermal
capacitor absorbing the energy in excess of 3,000° F
. a latent heat material reactive at this level as a
ring fluid (fig. 23).
Basic Oxygen Furnace (BOF). This equipment is
charged with molten metal and scrap. Heat is raised to
3,000° F using oxygen. Waste gases are exhausted. Be-
tween charges, furnace heat levels are maintained
through an auxiliary combustion heating system (fig.
24). A TES in conjunction with recovery of t~e wasted
exhaust (3,000° F) can store the recovered heat and be
recycled to maintain BOF temperatures between charges
or to preheat the scrap, thereby lowering the oxygen
needed per charge.
Electric Furnace. In addition to the blast and basic
oxygen furnace, the electric furnace is used for periodic
high-quality steel. A batch function, these devices accept
scrap and other steel charge. Preheating this scrap or
steel charge to 930° F has been estimated to reduce
energy use 18 percent, electrode wear 24 percent, and
increase throughput about 20 percent. If a preheat of
1,800° F is affected, overall energy saved approaches 45
percent. A TES concept for this operation is shown (fig. .
2~. .
461
-------
40000
BONE CHINA IN
PERIODIC KILN_A""'"
.......
100CO
UJ
a:
-------
SETTING FIRING TO COOLING DRAWING
TEMPERATURE WARE
yyy yyy
TES UNIT
. PERIODIC KILN'" 20% EFFICIENCY
. RECOVERY TO TES FROM KILN AND FLUE GASES
. TES ABSORBS ENERGY DURING KILN COOLING
. TES PREHEATS KILN AFTER SETTING
Figure 20. TES periodic kiln concept.
RUSS IAN DRV roOT.INC
roKE
OVENS
~
HOT roKE
11800 F)
n::;
.::r::~:~
:~. ::::
~:: ~::::
... ',',
o
TO OI.A::1' I~IRNA<:E
Ii' NEI':I\EI\ IN.:TAN!'!.\'
11)1'
(,.AS
(J40 i')
-.
COKE (3700n TO
BLAST PUHNACE HAW
MATEHIAL .BIN
N2' ro, ro2
:;TIWI
ROIum
C:VCT.ONI':
SEPARATOR
[-- O~~TR~~J
Figure 21. Russian dry cooling.
463
-------
J-DT AIR
I!OT mKR
(lROO F)
(J1KE
(\VRN~
t--
~.
TI~~ TANK
IN~UT.ATION
IIRAT AVA II.AJI I.E FOR flATCIl
PROCE~~E~ SUCII A~ SCRAP
PRE-IW.ATING. ANN".AI.ING.
n. (\. ". TEMPr.RATURr. MA IN-
TENANCE. ETC.
() BY STARVIN(; 01'1' 02 TO COKE CIIARI;I,- COKE BURNING
CEASES WIIILE AT TilE SAllE TIME CIIARGES n:s TANK.
Figure 22. Variation of Russian dry cooling method.
..
TO AIR STOVES
VARIABLE HEAT INPUT
TES roILS
LINING B. F.
WALLS
SLAG
TO TES
FOR SCRAP
PRE- HEAT
Figure 23. Blast furnace with O2 injection.
464
TES
T
-------
HEAT FROM
OTHER PROCESSES
WASTE
GASES
.
. HE..<\T
B.O.F.
B IITWEEN
CHA RG ES
.
.
.
.
. '--
---
- =- -- .
---
-
.
- .
.
TES
B.O.F.
Figure 24. Basic oxygen furnace/TES application.
Jl:AGN~
I~
~
Yi(. .
t~ J :
PRE-HEATED SCRAP
SCRAP PILE
PIC. OF IRON
I
PRE-HEATER
CHAMB ER
- -
-_..- - --
--- - - -
---_.-
ELECTRIC FURNACE
TES
HEAT FROM
OORE QUENCHER,
STAG, ETC.
Figure 25. Electric furnace/TES application.
465
-------
nONllS flAS' IIDUCID '0 I-IN, Sill, THIN) 4-IN., AND nOllD
r.~:~::,. . . , -... ,
\" . ~ """ - - '...... I. u..,- u(.... ..'111.1
\ X -> ~~:K.~;~ ~~t:~. " "'-/."".' fl'" ':"~~~:;::J'
f.~.. -_::. ----.": :,'~' .//.;-'11 11'~ i E-
..-,..- , ,./ ~.., ',! ' .. ~I ;':¥~I
~t~,i;:~:~:~~;~:t;....~. ~,~, '~:'..' r,~:t~(~~:"~'!.~~~:'
SOMP.l' !MES, RAW MATERIALS
ARE PRE-Immn wrrn KIT.N
WASTE "'~AT.
nRY PROCESS
LESS ENERGY INTENSIVE
lAW MAnllALS All alOUND TO POWDII AND IUNDID
,:",..~~~:,~, ' III" -." ,~::,.,~::.~:~~;, .'~~ ~: ,;:: I:' r-~Tl
't i~i 0Fi j I; )'Y'/ .:.:::. :: . II ,\
'::.;,;:'~;:,~-l,,~:~:,~.:;:r ,:~::,~::,,~:~:~,,: .~ ..~;~
.U'IOUII"C\tOtO ."".'006 "'U 1-.' .,',"10 ..."
..,......" ",0.
rT'~l
:1 I' ': f
II ,
': ': , ,
~ n ".' "" I
"'0..... ...
.o.'u..", "0"."1
I WET PROCESS PRF.OOMINANT IN US .
L:'AW MAUIIALS All GIOUND, MIUD WITN WAUl '0, fOIM SLU"', AND IUNDID ,
6;;:;, ~~ .~, ~:'.~'.,:'~.~ );~';'~'::~:::i;~,~: ~~,.~
."",..',""-.,'-,',,,,;../0' - ",' .~ bj"..\ '~-.., :, --- :
. ... .,,, ...~ .,'''\'1'0 ...11. .-"..... .......
UI."""'''OOotIO
MnnERN KILNS WASTE ('.AS
PRE-"'~T RAW MATERIA!.
nEFORE ENTERING KILN.
IUININO CNANGIS lAW Mil CHIMICALU IN'O CIMIIn CLINK II
't:'.I'o. CAI.CINATIDN
I ~ '
I IP~;:" :;':'~;:A.I,~~',~"'.";"I,'...,
Uil."::::-- '-
, :;~ . .
I.~- - -~' :_.: - ~, ~LJL
.0" ... ~ ~~~, ,.' ........,/t'
,.. '... ., n ..~ ,,"..
';;;:. :! n I -. .~
.... ':: r ~ I I 1
.' ,"', . _t,. .... '?I
""~"'-' I :. "~,,!...., l I. ,..:....".",' 1=.
- :,'" '~: :: .~.:.::.,-:.~~:~~...:JJ
1.._" '1><'..' .~ ..'---:-.. .--
(I'"'' ..., 4'"",- C""""
'...........~
CLINIII WITN G'PSUM ADDID IS GIOUND INTO POR'LAND (IMIN' AND INIPPID
.~.
..
UUII'....
" .0,,"101
" .-
~d1U ," " :12
. .:; I .. ~"'~.'~~".~~,'''.:'.,,;b
'.~ . - -- 'L
.., .' '1 ~
-;:~~;:2:~r;", ':'::""~~":"'-'" ~';,,!~
I J . l'l;
, j"
, ,'\
~-
""",
:.: 'j ..,- --2-'1-
, ,t
\t' ',../
:~: :.' :(],..,., . .a,-
I' :-r H I~}]
,'.
""',
,.-,..,
"'-
...
fI\oI,1
..".
,..
~.
...
..c...,...
-,.,...
Figure26. Cement process.
466
--!.
~
-------
Coal and natural gas are major sources of fuel for
'kiln firing with electricity being used for grinding and
mixing operatiqns. Kiln energy utilization is given below
for two examples of the dry process:
Typical heat recovery techniques of modern plants
include:
1. Preheating of incoming raw material with exit
gases,
2. Preheating of combustion air with clinker,
3. Drying of raw materials with hot kiln gases.
The average kiln age of the 10 largest cement com-
panies is approximately 20 years, and many kilns are
over 40 years old.
Preheater systems have a disadvantage that alkali
metal compounds, normally removed in the hot exit
gases are condensed on the cold raw mix and reintro-
duced into the kiln. A cycle is thus set up that can
increase alkali metal concentration in the processed
clinker. To overcome this, gas bypass systems are used
which divert 5 to 10 percent of the gas with the alkali
components in vapor. Cooling of this gas from 2,000° F
to 850° F condenses these contaminants, and this gas is
then used to dry raw materials. The loss of 5 percent of
the gas to bypass represents an extra heat consumption
of about 130 MBtu/ton. The processing of limestone and
gravel into clinker offers several possible TES applica-
~tions. At the head of the plant, heat processes form the
clinker, which is subsequently ground and processed into
the final product. Slaking creates heat in the ground
clinker not now recovered. The gas heat re-
covered/stored in a TES, can supply drying heat for in-
coming raw charge (wet or dry process) or preheat com-
bustion air. These applications would most likely benefit
older plants devoid of integrated heat recovery equip-
ment.
The clinker cooling pit can be integral to a TES with
the resulting recovered heat supplied for any of the al-
ready mentioned functions. For those more modern
systems which have pre heaters to heat incoming raw
charges to the kiln and where gases must be periodically
bypassed to reduce alkali metal concentration, the by-
passed gases can be used to charge a TES, which would
then be able to supply additional heat for charge pre-
heating and/or raw material drying.
Paper manufacture
Paper manufacture is accomplished in two major
steps:
1.
ess,)
2. Refining of pulp into finished paper (a mech-
~nical process).
, The basic schematic is shown (fig. 27) with approxi-
Production of pulp (primarily a chemical proc-
mate energy levels as indicated in figure 28.
The pulp and paper industry generates two by-
products that are used as fuel for the production of
steam: bark and black liquor. Bark must be removed
from the logs because of its low fiber value, and it is
usually burned in bark boilers. Black liquor is the spent
cooking liquor from the digester, a solution of inorganic
chemicals (essentially sodium sulfide and caustic soda)
and organic matter (lignin). The chemicals are recovered
by evaporation, and the organic matter is burned in the
recovery furnace producing thereby a substantial share
of the mill's steam requirements.
There are several functions that appear to be com.
patible with basic TES operations capability, the batch
digestion function, smelt discharge, and in a paper break-
ing sequence within the paper mill. An advanced concept
using solar energy is shown in' figure 29. I n addition,
several unit processes have potential.
Batch Digester. Wood chips that have been digested
with liquor are "blown" into blow tanks from the digest-
er to reduce the pressure from 150 psi and also to aid in
breaking the chips into pulp. This material is at 340° F.
This is followed by cooling and washing. Also steam is
liberated during this "blow." Use of this energy-steam
liberation and pulp cooling charges a TES. This TES
might then be used to preheat an incoming charge, there-
by reducing process steam requirements. The restriction
is that most new digesters are continuous processers.
Smelt Discharge. Melted char containing recoverable
inorganic chemicals is continuously discharged from the
black liquor heat-recovery furnace. This smelt is then
quenched, crushed, and recycled to recover useable
chemicals for the digestion process. The char melts at
about 1,800° F (1,000° C). Energy from the smelt cool-
ing and/or energy from steam liberated during the
quenching operation could be used to charge a TES
which would be used for process steam, electrical genera-
tion, digester input preheating, etc. The restriction is
that smelt withdrawal is a continuous process.
Paper Breaking in Mill. When the process paper
breaks in a paper machine, a TES could collect the
energy normally going to the dryer rolls. This could then
be used to preheat the rolls prior to restarting, preheat
the digester charge, etc.
Aluminum
Aluminum production may be broken down into
three steps:
1. Ore mining and preparation, .
2. Alumina preparation (Bayer process),
3. Aluminum smelting (Hall process):
467
-------
'9"
~~ @:
:c , ~-.J
. ~ ~UPPl£"E"'TAA" ,....L.. """
.;~t ~U~B~U!l _\1 t :~~
;~ '. ---Y' ~IJ,~
..... g :' TU'-60 .--.....,;;:.
u I -L.. t t ...J-."-- Be 'C. e ,Cue. I
! . I p~ocn~ ' ~!., UN" --,
h jJ STfA~ +; ~. P.fC~P~"O~
ii.. ----, . ~~
"'u SUPPL£IoI[froirUl't I I .~' , ..
~~ FUEL. ~.LL ~ - - - .
~ .....ou...r L-
~ ~ 13' ,L-" ~'~i~:;g:::~:
\:!) :--.
(ttl..""ET ~fL T O'~~L""""G ,
T ."'. ",' GoRE EN
lICuO~
I llJ @ : C'U~TICllE~
~I
I: 1-
! tFUfLj I
t ~ c.e_J
~ C",PP!~S
~,<#-u ,I I I I
...;~ I
I. cxr / I---r--"'---I I
LCX;S'Y RAIL. Q ". t I
Tauc«. oa WA TEl :
.1:0-
en
CO
Q
CoCCo
....
..."
o~
@~;
9 :u
WI.F'IE~
PAPER U ~ U
....
Q~
...~
~~
.~
~
...
...
..
vD-o
@
PAPER
MACHINE
-(!)
OIGUTU'
...
z
. ~
oj
SHIPMENT
AS PULP
WHITf LIQUOR
~ Ij\~
... ~.
... ...
IE ..
1. BARK DRIED TO 50% MOISTURE CONTENT.
2. BARK USED AS FUEL FOR STEAM.RAISING
3. ;i...;'Cr: LiGuvn i;.u r i5 5;:n~.,.i:u uNlv wALL
OF BOILER WHERE FURTHER WATER REMOVAL
AND FINALLY COMBUSTION TAKES PLACE.
4. SMELT TAPPED FROM FURNACE, aUENCHED
AND GROUND TO RECOVER CHEMICALS, WHICH
AFTER FURTHER PREPARATION CAN BE
RECYCLED INTO THE COOKING PROCESS
5. STEAM COOKED AT HIGH TEMP & PRESSURE
6. DIGESTERS
TEMP = 340 F. PRESSURE = 150 PSI,
HEAT FROM STEAM
BATCH PROCESS & CONTINUOUS PROCESS
7 PULP IS PASSED BETWEEN TWO RUBBING
. SUR FACED.
8. MECHANICAL WATER REMOVAL (ROLLERS &
PR ESSES)
DRYERS - LARGE DIAMETER HOLLOW ROLLS,
HEATED WITH STEAM
CALENDAR MACHINES - COMPRESS PAPER TO
GIVE IT A SMOOTH FINISH. COLORING MAY
BE ADDED IN WHICH CASE PAPER ISWETTED
UP TO 35%. AFTER THAT. IS DRYED IN
AFTER.DRYERS SIMILAR TO INITIAL DRYING
MACHINES.
9. TRIMMING & CUTTING OPERATIONS. NG
SURFACED.
Figure 27. Pulp-paper manufacturing operation.
-------
129 PAPER HOT
TURBINES WATER
~
0)
to 4 LIQUOR
TANKS
9 11 67
Ev APS
MISC ANO PRECIP
LANCING MISC
29 ElEC CO NT
POWE R DIGESTER I
SOURCE STEAM- 1000 Ib/hr
Watson, J.H.. Continous Digesters:
Their Effect on a Steam Power System,"
Tappi, May 1968 Vol. 51, No.5.
5
208
284
NO.2 6990 - - 7292 NO.3
6000kW 6000 kW
- 8818 ELECTRIC
UTILITY
POWER 483 5427 NO.1
PLANT - PAPER
MACH
1504 5770 NO.2
WATER PAPE R
MACH
346 3686 NO.3
CAUSTIC - PAPER
MACH
RECOVERY 468 579
. BAGPAK
WOOO 606 1350 - PULP
ROOM -- MILL
EVAPS 390 1128
-- WASHERS
CaNT 16 1347- CONT
DIGESTER . DIGESTER
. 74
23
171
elECTRIC POWER-kW
Figure 28. Steam and electric power flow diagram in integrated paper mill.
-------
SOLAR COLLECTOR
ARRAY.
rt
THERMAL
STORAGE
SYSTEM
~
-..J
o
TEMPERATURE
PRIMARY
LOOP
(A) NaK
(B) SYNTHETIC
OIL
EXP
TANK
P
8000 F
750° F
4560 F
STEAM
4500 psi
HEAT TRANSFER DIAGRAM
(A)
HEAT TRANSFERRED
STEAM
GENERATOR
(A) SUPERHEATED
STEAM
(B) SATURATED
STEAM
r----..,
I FIRED
SUPERHEATED I
~ONl~_J
HEATED
FEEDWATER
SUP~R~E,l\.TED
STEAM
TO
TURBINES
~
5500 F
-~MINOL55
4500 F 4560 F
STEAM
450 psi
TEMPERATURE
HEAT TRANSFER DIAGRAM
(B)
HEAT TRANSFERRED
Figure 29. Solar energy systems for pulp-paper industry steam generation.
-------
Only 15 percent of the requ ired baux ite ore is ob-
'ained domestically. Also, since the domestic supplies
that are available must be shipped long distances, it is
usual practice to dry the ore at the mine to reduce ship-
ping costs. Therefore, even though a rotary kiln at
1,100° F is used for drying, the waste energy is not'
available for later processes.
I n the Bayer process, ground bauxite is digested
with a caustic soda solution in steam-jacketed containers
at 300° F and 50 to 70 psi for 208 hours. The sodium
aluminate is cooled and pumped to precipitators where
alumina hydrate crystals are formed. The crystals are
then heated in kilns at 1,800° to 2,200° F to produce
alumina.
The Hall process is an electrolytic reduction process
where alumina is continuously dissolved in molten
cryolite in a reduction cell. Aluminum is liberated and
sinks to the cathode while oxygen goes to the anode,
which it attacks, burning to CO and then to CO2 with
contact to air. The low-voltage direct current cell oper-
ates at about 1,800° F.
Two anode systems are used, prebaked electrodes
(fig. 30) and continuous Soderberg eletrode (fig. 31).
The Soderberg system uses an anode which is baked by
the reaction heat of the cell.
The finished aluminum is tapped off from the cell,
ir siphoned out, and then poured into preheated cruci-
pies. From there it is either cast into ingots or alloyed
for further processing.
Though more modern plants are highly engineered
for energy efficiency, several possible TES applications
are recognized.
Bauxite Drying. Steam driven off during the drying
step could be used to char,ge a TES and then later used
for further processing (assuming plant is located near the
mine). preheating of incoming charge, or space heating,
as required. After ore has been dryed, the heat content
of the still hot ore could be used to charge a TES which
would then be used for the same functions as above.
Bayer Process. Energy is captured from steam given
off in rotary kiln and is used to provide steam for digest-
er, to generate electricity, and/or to use in the carbon
electrode baking process. This process may be employed
to capture energy from hot alumina, which has the same
uses as for Bauxite Drying.
Auxiliary Processes. These are:
1. Evaporation of spent liquor to recover energy
from vapor driven off,
2. Red Mud Kiln for the recovery of energy from
~Oling process.
Hall Process. This process is for the cooling of cast-
ings and ingots. This energy is routed to TES, which
could then be used to preheat crucibles.
Prebaked Electrodes. These are used to:
1. Recover energy from kiln baking in the form of
steam,
2.
Recover energy from kiln cooling.
Petroleum
Petroleum refining utilizes most of the raw energy
as direct heat (combustion operations) followed by
extensive reclamation and sense of waste heat in a cas-
cading series of thermal process functions. Common to
all refineries are the process temperatures required for
cracking, distill ing, and reforming functions to produce
the derivative petrochemical products of crude oil.
Unfortunately, process and equipment designs are as
numerous (individualized) as there are U.S. refineries
.1247 in 1972). A general schematic of the most preva.
lent process is shown (fig. 32). Refineries are located
near or on sites where oil availability is abundant. little
electricity is required in the process, but natural gas for
butane and some reformers is increasingly harder to
obtain. Temperatures and pressures for major energy-
consuming processes are shown on the above diagram
(fig. 32). Operations (called "crude runs") are con.
tinuous for as long as certain fuel stocks (grade) are
being produced. The need to produce, say, jet fuel after
a run of heating oil requires a plant shutdown and redi-
recting of certain crude process parameters, crackers,
and distillation operations to restart the refining for the
new product. As there are many different refining de-
signs, additional analysis of unit processes is needed to
develop any potentials for solar-thermal applications;
however, large amounts of steam and middle-high tem-
perature (600° to 1,000° F) heat processes are character-
istic of the process train. .
Automobiles
Auto production involves many energy-consuming
processes including casting, forging, machining, molding,
planting, paint baking, etc. The metals included are steel,
copper, aluminum, and cast iron; nonmetals include
rubber, plastics, vinyl, and textiles. A total energy
accountability would show the total energy consumed
from raw materials to final assembly to be as shown
(table 15) using 1974 basis and considering three sizes
and weights of cars.
The operations implicit in the manufacture of the
many products assembled into a modern automobile
must be investigated in that a wide variety of heat func-
tions are utilized. The summary of industry groups
471
-------
,...
J.
'8\1.,11
'I11III:
~
....
I'.)
'00& .,-
TI=y. =~
r=.:;;;S?;':;:;:~'n
..-' - .
....1 iI'...': "..I',i "'"
""V I
"":.': 8 \i'"
!. '----------1 ~I~~~
~':~. ~ \'i : r- --., ~:~..lj=-=-::i --'1-- - - - -- -- ------ ---,
'L, J., I I t j;~ hO'lo~" I
"".T~{i~i-r?)~,]'''''~L:')-(:~1~:'l : ~.:~:; !~~ :
.,..,., ~ ~ II L-3~ ---.bd I I V. I
- I I I
~ ----------------------.1 I LillI I
I ""!I'O. I '1.' I
I . I I
I .----.-- ---r' I
I I"""OU,.~. I
i . f'~C:":~"'"';t~~;~:: ! !
: ~~ ::~ 1 r== ,,,,..' 1 :
-- -c.-,...,:...t.tJII\~".,501C'" . ..==~~ 'U~'.'( I I
.:::-~-j !;~~-~-~ ,....1(. ,,8!n .., I "I,.TU I
~....(-L. ':'."' I ...~ tAu. I
. .(0 ""0 I I .(~~.[. ,Io-ac, . .',,10. I
~-i'-R1.-p---~ : r- - - ~
r- ---- !~'::' ~.~._- -----.- - ~!r---. t I
i ~"'. I I
0'''''''' I I I
I ~= =1 ~~ I
~ -:--.-. .~; ~ In' I "."..,... : I
~-:: ! ..;: .:,;,,:; -r--:~ ".". : I
'h:.!.~>-I II 1 j ! ~ .... ~::-:.-...J ~ J
Si~d-!::-;;::!''';~:::'; ~il I
"....~..I . . i I
r ~ "j" """ .::::::..: I
'::e' ~':::::~~..... &~;::., R....~--...pJ.
F ::::.-.. .=:=:- :J-~ '.,--- ~=-, ,'w',., .'On "'. I OR I
.:---1 ":;", :"; ,"I i 8-;... i -'-"~;7~ - ~ --.::.::"~ - -.J - ~ - - -----l
"~!I.' 10 810..(1'011 ""''''1
I STORAGE AND - ORE GRINDI~G or.- BAUXITE
-
SHIPPING - CALCINATION - HINING
.
I
ROTARY KIL~
1100" F
HALL-HERCULT
PROCESS
SODERBORG CONTINUOUS
ELEGrRODE METHOD
HIGH VOLTAGE
POWER SUPPLY
A~DE
BUS BARS
CARBON
PASTE
GAS
THERHAL I:-ISUlATION
_-.J
,
AUJMINUM TO
IDLDING FURNACES
AND CAST INGS
CATHODE
BUS BARS
}! ILLS
SODERBORG PREPARED ELEGrRODE METHOD
PREBAKED ELEcrRODES
FROM PETROLLEUH COKE
. Figure 30. Alumina reduction cell Soderborg system.
-------
COKE MIXED WITH
PITCH, OR OIL & TAR
STEAM-JACKETED 4
KNEADING
GRINDING
.
CALCINED
. (RED HEAT) ~
FOR 3 - 4 HOURS
COOLED
PETROLEUM COKE
.
CRUSHED
.
MOLDS OR
PRESSES
.,.
AIR
DRY
FINISHED
ELECTRODES
.
1880 2500° F
FURNACE OR
KILN BAKING
48 HOURS
(SLOW HEATING TO
AVOID CRACKING)
4
(OLDER KILNS NOT EOUIPPED FOR ENERGY RECOVERY, HOWEVER
MORE MODERN KILNS ARE NET PRODUCERS OF ENERGY [AS STEAM])
Figure 31. Prebaked electrode process.
t:r.RI):~n)r.
i'; 1 nr.ATIN(: 011.
AVtATtnrJ
I(!-:(:UIAK
'2 1IF..~TINr. Otl.
l'IH''IlllN
tPt:
, ~I ,t" 6 Fl1F:1. 0IT.
1,'111-:1. CAS
ROAO on
AsrnAT TS
II',.
CATAlyrr,:
HIA\'Y
1:-",:11' nTI
(~TAI'y.['tl:
I.A',I'I!:NE
1~'\I'AI i"I" C:
,:\,,;1.1: ntl
I.tcm STKAHarr
RUN c:AS
(20n-400" n
IIEAVY STRAlr.Hf
I{UN C:A~:
(I~OO-6')Ou n
111STll,IATE';
(I,OO.(,'iO F)
TurrEn CRtli)E
(WIO.I <;oo~ F)
(:R, lIIJI~
oIl.
IIEA'f
IOO,uno UTI!
HARREl. of I;HIIDE
Figure 32. Sample petroleum industry.
473
-------
Table 15. Automobile material breakdown and energy for manufacture
107..-4. nl'9_~_......I- ~'_"-..i 1:;74 i-'iymouth Satellite 1974 Plymouth Fury ill
- - . .. ""J 4"'.V""'''U. ¥ CI..l.J.C11.H..
4-Dr Sedan, 225 cm engine, 4-Dr Sedan, 318 cm engine, 4-Dr Sedan, 360 cm engine,
Auto trans. pwr steer,pwr Auto trans, pwr steer, pwr Auto trans, pwr steer. pwr
brakes, radio & heater brakes, radio & heater brakes, radio & heater
Materials Group Curb Wt 3111 lb Curb Wt 3697 lb Curb Wt 4315 lb.
Energy
BTUs/lb Lbs/ BTU/Car bbls crude Lbs/ BTT.:/Car bbl s crude Lbs/ BTU/Car bbls crude
(Table I) Car (Thousands) oil/Car Car (Thousands) oil/Car Car (Thousands) oil/Car
Plain Carbon Steel 21,000 1630 34,230 6.113 1970 41,370 7.388 2248 47,210 0.430
Stainless Steel 34,000 9 306 .055 10 340 .061 13 442 .079
Alloy Steel 22,300 78 1,739 .311 91 2,029 .362 111 2,475 .442
Galvanized Steel 21,500 63 1,355 .242 75 1,613 .288 95 2,043 .365
Aluminized Steel 21,500 29 624 .111 34 731 .131 42 903 .161
Cast Iron 10,300 436 4,491 .802 470 4,841 .864 505 5,202 .929
Modular Iron 14,050 24 337 .060 82 1,152 .206 95 1,335 .228
~ Malleable Iron 15,500 41 636 .114 68 1,054 .188 70 1,085 .194
..... Aluminwn, rolled-drawn 110,000 14 1,540 .275 20 2,200 .393 10 1,100 .196
~
Aluminllm, cast 10,000 56 560 .100 66 660 .118 72 720 .129
Copper & copper alloys 65,700 26 1,708 .305 31 2,037 .364 30 1,971 .352
Zinc, cast 45,500 28 1,274 .228 17 774 .138 67 3,049 .344
Lead 22,000 35 770 .138 35 770 .138 31 682 .122
Body Solder 22,000 6 132 .024 6 132 .024 7 154 .026
Glass 13,000 79 1,027 .183 88 1,144 .204 110 1,430 .255
R ubbe r 36,900 183 6,753 1.206 221 8,155 1. 456 220 8,148 1.455
Plastics 25,000 77 1,925 .344 80 2,000 .357 147 3,675 .656
Soft Trim 7,000 76 532 .095 81 567 .101 95 665 .119
Sound Deadeners & sealers 7,000 52 364 .065 51 357 .064 87 609 .108
Paint & protective dip 7,000 23 161 .029 26 182 .033 26 182 .033
Fluids and lubricants 7,000 146 1,022 .183 175 1,225 .219 234 1,638 .293
TOT-~L~ 538,250 3111 51,486 10.980 3697 73,333 13.095 4315 84,718 15.128
Presentation by Doran K. Samples, Vehicle Emissions Planning,
Product Planning and De\"elopment Office, Chrysler Corporation,.
at Energy Seminar conducted under the auspices of Institute of
Sciences and Technology, T.:nil"ersity of Michigan on Friday,
August 23, 1974 in Tra\"erse City, Michigan.
-------
.hose products support the overall manufacture and
~ssembly of end items, such as autos, represents the five
largest energy users. Consequently, only the assembly
and finishing operations are unique to this manufac-
turing process. These represent 26 percent of the total
energy required, a significant portion. Included in the 26
percent figure are plating, paint-baking, various corro-
sion-preventing dipping operations, and support systems.
A further investigation of these operations is warranted
for possible TES applications.
Conclusions
The potentials for TES in industry can be categoriz-
ed as basically relating to steam systems with the highly
variable demands and heat supply for process functions
spanning the entire spectrum of industrial temperatures.
Major challenges lie in materials compatibility of TES
media and containment, heat exchange, and
monitoring/controlling the system with respect to the
heat capacity of the store and dynamics of heat flow on
a demand basis.
Several industrial applications are apparent, a few
having nearer-term potentials, most requiring new tech-
nical innovations both to current process equipment and
in creating means for capturing,. exchanging, and storing
the vast amounts of high-temperature reject heats of
~heavy industry.
ACKNOWLEDGMENTS
The author wishes to thank the members of the GE
Thermal Energy Storage project team, particularly
Robert McCarthy, John Schelkopf and Tom Gresko, and
Dr. Phillip Lowe and C. J. Swet, Thermal Energy Storage
Program Management of the Energy Research and
Development Administration, Washington, D.C., for
their guidance and support on the basic program. The
maj9r portion of the materials reported in this paper are
the result of an ongoing program funded by ERDA (con-
tract No. E(11-1)-2558).
REFERENCES
1. H. J. Schwartz, "Batteries for Storage of Wind-
Generated Energy," NASA Lewis RC, Workshop
Proceedings, NSF /RA/W-73-006, December 1973.
2. P. M. Ordin, "Review of Hydrogen Accidents and
Incidents in NASA Operations," Paper No. 749036,
Proceedings of the 9th IECEC Conference, ASME,
New York, 1974.
3. Walter Goldstern, Steam Storage Installations,
Second Revised and Enlarged German Edition,
Pergamon Press, New York, 1970.
475
-------
ENERGY CONSERVATION BY THE RENOVATION
OF HIGH-TEMPERATURE TEXTILE WASTEWATER
Craig A. Brandon, Ph.D.*
Abstract
Hyperfiltrcttion has previously been shown qffective
in the renovation of textile dyeing and finishing waste-
waters for diMct recycle at full process temperature.
Direct recycle .lchieves conservation of resources, which
is an ideal form of pollution abatement.
The objecrille of the research program at Clemson
University is tc assess the potential impact of the appli-
cation of high-temperature hyperfiltration technology
on energy c0n59rvation and environmental preservation.
This objective includes the determination of the practi-
cality of this application, i.e., the cost effectiveness.
The larges.f single use of energy in the textile in-
dustry is for dyeing and finishing-about 4.1 x 1rf'
gallons of fuel oil per day (equivalent). The potential
energy conservcltion benefit to be realized by full utiliza-
tion of direct f/'Cycle at process temperature is estimated
to be about 2;: 106 gallons of fuel oil per day (equiva-
lent).
The descr~Jtion of the high-temperature hyperfil-
tration researcf. program is presented by way of a brief
discussion of preliminary results for two examples. One
example of a latch-type process and one example of a
continous procl~ss is presented. These two categories of
processes each represent about half the production in
the textile indw:tTY.
I NTROD UCTION
Hyperfiltration has been shown effective in the ren-
ovation of textil2 dyeing and finishing wastewaters for
direct: recycle ilt full process temperatures (ref. 1). Of
course, such direct recycle, while achieving conservation, .
is also an ideal form of pollution abatement.
The objective of the research at Clemson
Universityt is to assess the potential impact of the appli-
cation of high temperature hyperfiltration technology
on energy consmvation and environmental preservation.
Our objectives include the determination of the practi-
cality of this ap:>lication, i.e., the cost effectiveness.
. Associate Professor. Department ')f Mechanical Engineer-
ing, Clemson University. Clem'::.n, South Carolina.
tHesearch sponsored by EPA Grant No. R803875.
The textile industry utilizes about 2.8 percent of
the industrial energy consumed in the United States (ref.
2). About 70 percent of the energy used in the textile
industry is provided by fossil fuels. The largest single use
of energy is in dyeing and finishing; about 4.1 x 106
gallons of fuel oil per day, or its equivalent.
. Without conservation, the energy use in the industry
will increase by at least 30 percent in the next 4 years.
By way of reference, this will be about 10 to 15 percent
of the flow of the Alaskan pipeline.
The effluent from the wet processing of textiles
constitutes more than a minor industrial waste stream. A
recent study (ref. 3) concludes that over the next decade
the daily discharge of textile wastes will be nearly con-
stant at about 1 billion gallons per day. The constancy is
a consequence of anticipated improvements in finishing
processes and not in a fixed fiber useage. The potential
pollutional load on the environment of the textile waste
stream is considerable, since the biological oxygen de-
mand is about 600 mg/I. About three million pounds of
dissolved solids are included in the daily waste stream. In
addition, the textile effluent may produce an esthet-
ically unacceptable coloration of the receiving waters.
This paper presents a description of the high-
temperature hyperfiltration research program by way of
a brief discussion of preliminary results for two exam-
ples. One example of a batch-type process and one
example of a continuous process are presented. These
two categories of processes each represent about half the
production in the textile industry. The results presented
are preliminary since this program was initiated on July
20, 1975.
ECONOMIC POTENTIAL OF REUSE
The use of dynamic membrane hyperfiltration to
effect energy conservation by renovation of hot process
effluents for direct recycle was suggested in 1971 (ref.
4). The waste streams, i.e., process effluents, are in fact
relatively dilute mixtures of residual chemicals in the
originally "pure" process water. The water and the
chemicals each are an operating cost of the textile proc-
ess. The treatment of the effluent is an increasingly sig-
nificant additional cost as discharge requirements
become more stringent.
The economic potential of reuse of the constituents
in the composite waste stream from a textile dyeing
476
-------
IIIlant is illustrated by the following values for the con-
~nts of a 2 x 106 gallons/day (7,570 m3/day) waste
stream:
Constituent
Chemicals @ 4.4 ct/kg
(@2 ct/pound)
Water @ 13 ct/m
(@ 50 ct/1,OOO gallons)
Heat @ 1.9 ct/1 07 J
(@$ 2/106 Btu)
Value ($/day)
500
1,000
857
Total 2,357
These values are directly proportional to the volume of
the waste stream. The total dollar value for the billion
gallon (3.8 x 106 m3) daily discharge estimated for the
textile industry is nearly $1.2 x 106/day. In the current
era of rapidly rising prices, this may be a conservative
estimate.
Full recovery of the potential value of the solids
undoubtedly presents the greatest difficulties. However,
if only the NaCI salt is reused, e.g., with acid and direct
dyeing operations, the wasted chemicals in the effluent
represent a significant economic loss.
The economic potential of reuse is enhanced for
direct recycle of the effluent from individual processes.
In particular, the energy conservation achieved by direct
~ecYcie at full process temperature is optimum, and the
~ecovery of chemicals may be less difficult, when only a
simple concentration is required.
Before presenting the preliminary results for the
two examples, a brief description of the hyperfiltration
process with .dynamic membranes, and of two related
research programs will be given for completeness.
HYPERHL TRATION
Hyperfiltration, or reverse osmosis, is a separation
process (ref. 5) involving the filtering of aqueous solu-
tions by membranes capable of removing not only sus-
pe n ded particles but also substantial fractions of
dissolved impurities, including organic and inorganic
material. The process is illustrated schematically in fig-
ure 1. Application of high pressure to the feed solution
causes purified product water to pass through the mem-
brane. The remaining feed becomes a concentrated solu-
tion.
Dynamic membranes, which are formed by deposi-
tion of materials from the circulation feed stream onto a
suitable porous support, may be formed from waste con.
stituents themselves or from certain polyelectrolytes
ref. 6). The previous dye waste tests (ref. 7) were made
roth with self.formed membranes and preformed dual.
layer hydrous Zr(lV) oxide.polyacrylate membranes
(ref. 8). The preformed membranes, which are usually
superior in both solute rejection and purified water pro-
duction rate, will be used with the process effluents in
the present program.
The performance of dynamic membranes is charac-
terized by high product flux, of the order of 50 to 100
gallons per day per square foot and good solute rejec-
tion, of the order of 70 to 95 percent.
In addition, dynamic membranes have the character-
istic of a stable positive performance coefficient with
temperature. This capacity to treat high-temperature
wastewater is not possessed by present-day commercial
cellulose acetate and polymide membranes.
The present program is related to two other projects
that will be described briefly.
RELATED PROGRAMS
Pilot Plant Demonstration
In 1972 the potential savings were recognized by
EPA (Grant No. S800929) and industry. A pilot plant
demonstration of hyperfiltration treatment of composite
wastewater from a typical textile dyeing and finishing
plant was sponsored at La F ranee I ndustries, a division
of Riegel Textiles (ref. 9).
Even though 32 becks were exhausted to a common
drain, both the purified product water and the concen-
trated residue were recycled in 18 full-scale production
dyeings that produced 1,000 m of furniture and uphol-
stery fabrics (ref. 10). Even though it was thought that
such a mixture may only be useful for dyeing black,
eight standard shades were processed and sold as parts of
"first quality" commercial orders. I n table 1, the
amount of dyes that would have been used to achieve
the standard shade is compared to the amount of dye
. used in the test dyeings. On the average, 16 percent
savings in dyes were observed.
Mobile Demonstration
As a result of this pilot plant, a mobile laboratory
demonstration project was initiated (EPA Grant No.
S802973 to the South Carolina Textile Manufacturers
Association!. Table 2 indicates some of the character-
istics of the 12 different plants involved in this program.
The major types of fibers and processes were included in
this survey program. The plant size is indicated by th()
discharge that ranged from 2,700 mJ /day (0.7 MGD) to
45,400 m3/day (12 MGD). The range of chemical char-
acteristics and pollutional loading is indicated. The
flow-weighted average solids loading, 1,900 mg/I, indica-
ted in table 2, is over three times the industrywide aver-
age of 600 mg/I given by Porter (ref. 3) earlier.
This mobile demonstration was necessary because
field test evaluations of membranes are essential to ob-
tain meaningful design data. It is also important that the
477
-------
Pressure Vessel
Feed, (
-
. Brio
,f -
- Membrane
Porous Membrane Support
11' 1
e
Product, Cw '
Figure 1. Schematic diagram 01 hyper1i1tration membrane.
Table 1. Reuse dyeings 01 cotton velour with concentrate water a
Amount of dyes
Number of dyes Jpercent of fabric weightt
Weight of fabric Reuse Standard Reuse Standard
(1 b) (kg) Shade test formula test formula
48 22 Tan 3 4 0.412 0.438
97 44 Blue-green 2 2 10.480 0.932
82 37 Light blue 1 3 0.055 O. 189
85 39 Rose 3 3 0.233 0.302
82 37 Violet 2 2 O. 1 30 O. 130
197 90 Tan 3 4 0.305 0.438
153 70 Rose 3 3 0.310 O. 189
100 45 Blue-green 2 3 0.485 0.632
52 24 BUi~~t-orange 3 3 1.450 1 .940
157 71 Gold 4 3 0.990 1 .033
a S800929.
EPA Grant No.
478
-------
Table 2. Characteristics of textile effluents in mobile hyperfiltration projecta
Fibers processed (percent)---- Waste discharge characteristics
Cottons Synthetics Wool Flow Temperature pH COD TDS
106 GPD 103 m3jd of °C (mgj,Q,) (mgj,Q,)
X X 3.0 11.4 107 42 9.8 1284 na
X 0.7 2.7 100 38 4.0 na 430
X X 2.0 7.6 110 43 11. 0 2400 3600
X X 8.3 31.4 99 38 11.4 na 1850
~
-.J X X 4.8 18~2 88 31 11.0 890 980
co
X X 12.0 45.4 120 49 11.0 1780 1750
X X X 2.0 7.6 91 33 7.3 670 1130
X X X 6.0 22.7 81 27 11 .6 1033 2580
X X 1.4 5.3 84 29 11. 3 na na
X X 6..0 22.7 95 35 12.4 1200 3000
X X 1.0 3.8 130 54 4.0 700 800
X X 2.4 9. 1 140 60 7.0 450 350
aEPA Grant No. 5802973.
-------
cooperative dforts with the industrial personnel involve
the evaluaticn of recycle in their own laboratories and
production facilities.
The organization of the mobile demonstration pro-
gram is shown in figure 2. The diagram indicates the
cooperation ;Jetween government agencies and industry;
between equipment manufacturers and users; and the
central role of the university and trade association. This
is the same organization for the current energy conserva-
tion program except that the EPA grant is directly to
Clemson Uniuersity.
The pilo': plant and mobile laboratory programs ap-
proached the recovery problem as indicated in figure 3.
Thu composite wastewater from the entire plant was
processed by hyperfiltration to achieve recovery. How-
ever, since tt e objective is to separate the mixture for
recycle, this separation should be achieved before any
unnecessary rnjxing occurs. Hence, the approach for the
current resea.'ch is to use a hyperfilter for each unit
process as indicated in figure 4.
It is typical in the textile industry that the higher
temperature (Jffluents are the more polluted with resid-
ual chemicals. The preparation, the dyeing, and the fin-
ishing of fabr cs often is carried out at high temperature.
The initial discharge of spent process fluid and the initial
wash water are the most concentrated in chemicals. The
maximum return of energy and chemicals is from treat-
ment of these hot initial process effluents.
It is interesting to recall that the performance of
dynamic membranes is improved at high temperature.
Thus, the unusual situation occurs that the maximum
potential recovery coincides when the membrane costs
are minimum.
Conseque:,~ly, the current program will evaluate the
energy conservation potential by hyperfiltration of the
high-temperature effluents of unit textile processes for
direct recycle.
ENERGY CONSERVATION PROGRAM
The objectives of this program are to demonstrate
the direct recycle of hot textile process waters renovated
by high-temperature hyperfiltration. The energy conser-
vation and engineering performance details will be
assessed for selected unit manufacturing processes.
The objectives will be met by detailed evaluations of
selected manufacturing processes in current production
operations. The evaluations will be carried out in the
following steps:
1. Plant visits for gathering information on details of
energy, water, and chemical discharges;
2. Assessment of conservation potential based on field
data and laboratory hyperfiltration of wastewater
samples;
3. Onsite demonstration of the performance of hyper-
filtration on selected manufacturing processes, one
batch and one continuous, to obtain engineering
design data; and
4. Modification of the manufacturing process will be
considered and evaluated within the textile plant
facilities. .
The results of the assessments for these selected
manufacturing processes will provide the engineering
data needed for a full-scale demonstration. Ultimately,
the demonstration may include the modifications to the I
textile processes and the application of advanced treat-
ment technology to approach zero discharge.
The following discussion of the two examples in-
cludes only the description of the process and the assess-
ment of the conservation potential.
Batch Process
The batch process selected for this study invoives
the piece dyeing of upholstery fabric in large open
kettles, or atmospheric becks, figure 5. The piece of
HYPERFILTRATION FOR ENERGY CONSERVATION
TEx'rILE FINISHING
PLANTS
EPA
T
CLEMSON UNIVERSITY
I \
OAK RIDGE NATLONAL LABORATORY
TEXTILE TRADE
ASSOCIATIONS
MEMBRANE MANUFACTURERS
Figure 2. Organization of cooperative textile research.
480
-------
PLANT
ENERGY
CHEMICALS
IIIIIIII------
WATER
WASTE TREATMENT
SLUDGE
PRODUCT
Figure 3. Schematic diagram of renovation of plant composite wastewater.
ENERG'1:
I
.WATER I
. ----- --1 CONCENTRATE
I
- -+DRAIN
+ .
CHEMICALS
Figure 4. Schematic diagram of renovation of unit process effluent.
UNIT
PROCESS
PRODUCT
481
-------
BECK VENT
THROUGH ROOF
DAMPER
DOOR
ROTATING
FRAME
DOOR
COLD
WATER
SUPPLY
[;;OVE R FLOW
TO
DRAIN
WATER
CLOTH
I
BAFFLE
-STEAM
PORT
Figure 5. Batch process atmospheric dye beck.
cloth ("-50 m 10 600 m in length) is caused to circulate
through the ba':h of chemicals by rotation of the oblong
frame in the center of the machi'~e.
The finishing operation may contain three periods
of high-temperature operation. The cloth is introduced
into the beck after the initial charge of cold water is
heated to about 31° C (90° F). The water is further
heated, by the submerged introduction of steam, to 49°
C (120° F). Chemicals are added to scour the cloth (a
cleaning step required for approximately 30 percent of
the plant production). and the temperature is raised to
74° C (165° F). After about 45 minutes. the scour is
482
-------
~erminated by the introduction of cold process water.
,,"he cooling is accomplished by the dilution and over-
flow of the chemical bath.
The resulting cold water bath is heated to 49° C
(120° F) prior to the addition of dyeing chemicals. The
temperature is then elevated and maintained in the range
of 9f to 96° C (195° to 205° F) for 4 to 12 hours. The
shade being dyed is a principal factor in determining the
length of the exposure at temperature. After the desired
shade is obtained, the cooling cycle is achieved by the
dilution and overflow with co!~ process water.
Significant amounts of energy are utilized to heat
the cold process water during the scouring and dyeing
operations. In this plant about 200 million Btu/day are
effectively discarded as hot water. Interestingly, in this
case it wa~ determined that an even larger quantity of
energy is lost as vapor discharged by the forced draft
ventilation system. (Attention has been given to mini-
mizing the volume of this draft by closing the access
doors to the beck.)
The potential for conservation through hyperfiltra-
tion is being assessed. By modification of the manufac-
turing process, it may be possible to carry out the entire
operation at about 74° C (165° F). Recycled hyperfiltra-
tion permeate can be utilized to wash the cloth. It thus
°
~WOUld be necessary only to heat the dye bath from 74
(165° F) to 91° C (195° F).
Chemical recovery may be achieved by recycle of
the concentrate. In particular, it may be possible to re-
cycle the scouring chemicals. The salt used with direct
dyes also is a potentially practical recycle (ref. 9).
Continuous Process .
The continuous process is shown diagramatically in
figure 6. This process continuously prepares cloth for
subsequent finishing, i.e., dyeing or printing. A series <;>f
chemical baths are each followed by high-temperature
washing to "clean" the cloth. The incremental energy
discharged is over 250 million Btu/day at a cost of
approximately $500/day.
However, the chemical consumption in this process
is over $5,000/day. Both the recycle of hot water and
chemicals will be investigated. The potential for some
reduction in chemical usage by partial recycle of concen-
trate may result in a significant conservation of materials
and reduction in pollutionalloading.
SUMMARY
These examples illustrate the characteristics of the
two basic categories of textile wet finishing operations,
i.e., batch and continuous processes. The detailed assess-
~ment of energy and material conservation for five spec if-
'C cases will be considered in detail. Estimates of the
total impacts for the textile industry have been made.
Potential Energy Benefits
I n general, the dyeing and finishing processes are
often such that approximately 25 percent of the water is
discharged at quite elevated temperatures. The remaining
75 percent is fairly cool wash water. The potential ener-
gy benefit to be realized by full utilization of direct
recycle at process temperature was estimated assuming
that the hot water is only 100° F above the available
cold water supply. Often it is much more. The savings
for the textile industry is over 2 million gallons of fuel
oil per day, or its equivalent.
Potential Environmenml Benefits
The direct recycle of waste chemicals and hot water
(energy) at the point source obviously is the best possi-
ble pollution abatement technique. Though thermal pol-
lution is not often discussed as a major problem for the
textile industry, the treatment of high-temperature in-
dustrial wastewater is a difficult problem, even when the
plants are associated with municipal treatment facilities.
Hence, the elimination of heat and materials by recycle
has a definite impact on the need and expense of any
downstream treatment.
Previous research has concentrated on renovation
and reuse of composite wastewater. This program will
evaluate hyperfiltration at point sources in order to ef-
fect energy conservation through direct recycle of hot
process water. Implementation of point source renova-
tions for direct recycle of heat and chemicals will cer-
tainly enhance the treatment of the resulting composite
wastes relieved of often refractory materials.
The expected results will provide the engineering
data needed for a full-scale demonstration of high tem-
perature hyperfiltration. The results of the energy sur-
vey, the laboratory, and the on site tests will permit the
design for complete treatment to approach closed cycle
operation.
The combination of conservation, energy, water,
and chemical, and pollution abatement simultaneously
achieved by membranes may make the cost of this tech-
nology practical.
REFERENCES
1. C. A. Brandon, J. J. Porter, and T. N. Sargent,
"Reuse of Total Composite Wastewater Renovated
by Hyperfiltration in Textile Dyeing Operations",
Proceedings of the National Conference on Manage-
ment and Disposal of Residues, Washington, D.C.,
February 1975. .
2. "Blueprint for Energy Conservation", staff report,
Textile Industries, February 1975.
3. J. J. Porter et. al. "The State of the Art of Textile
Waste Treatment:' Final Report for FWQA Project
No. 12090-ECS, 1970.
483
-------
CHEMICALS
GREIGE
a..aTH -
WASHER
DE SIZE
SR'URATOR
40 GA.\ 1800F
CLOTlf
PREPARED
F~
FINISHING
PEROXI[£
WA90fER
40 GPM, 2000 F
CHEMICALS
DESIZE
WASHER
CAUSTIC
StU"URATOR
30 GPM. 2000F
CHEMICALS
PEROXIDE
!WURA~
CAUSTIC
WASI€R
40 GPM, 200° F
Figure 6. Continuous process preparation range.
4. C. A. Brondon, "Dynamic Membrane Hyperfiltra-
tion - Kev to Reuse of Textile Dye Waste," ASME
Paper No. 71-TEX-4, 1971.
5. J. S. John;on, L. Dresner, and K. A. Kraus, "Princi-
ples of Desalination," K. S. Spiegler, ed., Academic
Press, New York, 1966.
6. K. A. Kraus, A. J. Shor, and J. S. Johnson, Jr,
Desalinatic'n, Vol. 2, 1967, p. 243.
7. C. Aurich, C. A. Brandon, J. S. Johnson, Jr., R. E.
Minturn, P. H. Wadia, and K. Turner, Journal of
Water Pol/ution Control Federation, Vol. 44, 1972.
p. 1545.
8. .1. S. John!;on, R. E. Minturn, and P. Wadia, Electro-
analytical Chemistry and Interfacial Electrochem-
istry, Vol. 79, 1972, p. 1232.
9. C. A. Brandon, J. J. Porter, "Demonstration of
Hyperfiltration for Complete Renovation of Textile
Finishing Plant Wastewater," Final Report for EPA
Grant No. 5800929.
10. C. A. Brandon and J. J. Porter, "Complete Recycle
of Composite Textile Dyeing and Finishing Waste-
water Renovated by Hyperfiltration," presented at
the Environmental Symposium Textile Technol-
ogy/Ecology Interface, American Association of
Textile Chemists and Colorists, May 1975.
484
-------
NEW PROCESSES AND ENERGY CONSERVATION
IN THE PRIMARY METALS INDUSTRY
R. H. Cherry, Jr., Ph.D.*
Abstract
This paper reports results derived from three recent
studies of energy use patterns and requirements in the
primary metals industries. These studies were aimed at
first categorizing energy and fuel use patterns in 83
metallurgical and related processing areas and, then,
evaluating the theoretical potential for energy conserva-
tion in seven basic industries.
This paper focuses on some of the important results
obtained for aluminum and copper in terms of:
1. energy requirements,
2 the potential for energy savings both with and
without extensive process modifications, and
3. 'developing or emerging process technology.
INTRODUCTION
Battelle-Columbus has been engaged in the assess-
ment and evaluation of process-related environmental,
energy, and resource allocation problems for more than
20 years. This paper is based on results derived from
selected portions of three recent studies which, in com-
bination, were 'arrived at by first categorizing energy and
fuel-use patterns in 83 metallurgical and related process-
ing areas, and then evaluating the theoretical potential
for energy conservation in seven basic industries. These
studies are described (in chronological order) very brief-
ly in the following paragraphs.
Study of the Energy and Fuel-Use Patterns in the Non-
ferrous Metals Industries
This study is one of a series conducted for the Fed-
eral Energy Administration (FEA) and the Department
of Commerce covering a number of energy-intensive in-
dustries (ref. 1). The objective of the program was to
develop a preliminary data base in a short time, describ-
ing the general patterns of fuel use in the nonferrous
metals industry.
For the major SIC's, the information employed in
the compilation of the data contained in this report was
obtained in part from the literature, but primarily from
.Manager, Applied Metallurgy Section, Metallurgy Depart-
,ment, Battelle Columbus Laboratories, Columbus, Ohio.
direct contact with industry representatives. In some
cases, excellent responses were received in a relatively
short time. In general, smaller companies were found'to
be in a better position to respond than were large com-
panies.
This study of the patterns of energy use in the non-
ferrous metals industry included specifically 10 SIC in-
dustries categories:
3331 [[[ Primary Copper
3332 [[[Primary Lead
3333[[[ Primary Zinc
3334 .........................:........................Primary Aluminum
3339........................... Primary Nonferrous Metals, n.e.c.
3341 ..................................Secondary Nonferrous Metals
3351 ................................... Copper Rolling and Drawing
3352 ..............................Aluminum Rolling and Drawing
3356............................. Nonferrous Rolling and Drawing
3357............... Nonferrous Wire, Drawing, and Insulating
The level of effort applied to each SIC was determined
to a degree on the basis of the total energy use in the
industry. The primary nonferrous metals, n.e.c., non-
ferrous rolling and drawing, and nonferrous wire re-
ceived secondary emphasis on this basis. For example,
the analysis of the primary nonferrous metals, n.e.c.,
industry was restricted to silicon, magnesium, and tita-
nium, the major components of the SIC in terms of
production. Data for these metals were obtained entirely
from the literature.
It must be strongly emphasized that the results and
conclusions of this study were based on a very limited
sample of each industry and must be considered as pre-
liminary.
Because of the complexity and breadth of the sub.
ject of this paper, we somehow must limit the oral and
written presentations to something that is manageable.
This study served to provide background for the
broader, more complex Bureau of Mines Study
(described later). Table 1 gives the estimated total
energy consumed for selected SIC's.
Some modest additional background can be injected
by appealing to table 2, in which is presented one esti-
mate of the patterns of fuel usage in selected SIC's.
Not surprisingly, this pattern reflects what we know
-------
Table 1. Estimated total energy consumed and unit energy factors
in the nonferrous metals industries by SIC
a
Total energy
1012 Btu
SIC Industry 1971 1973 1974
3331 Primary copper 107.3
3334 Primary b 520.0 . 590.0 639.0
a 1 umi n urn
3351 Copper, rolling and 58.2 60.2 63.4
drawi n gC
3352 Aluminum, rolling 120.3 147.2 178.9
and drawing
3357 Drawing and insu- 45.9 43.6 38.4
1ating of wire
aExpressed as fossi1-fue1-equivalent energy.
bVa1ues for aluminum assume that 35 percent of electrical
energy is hydroelectric.
CData are for only nickel rolling and drawing.
Evaluation of the Theoretical Potential for Energy Con-
servation in Seven Basic Industries (ref. 2).
The objectives of this study were to determine the
minimum thee;retical energy requirements in seven basic
industries through a thermodynamic analysis of the pro-
cesses emploYEd in each industry. The study includes the
steel, copper, nluminum, glass, synthetic rubber, selected
plastics, and p,3per industries. Results of the calculations
for these sever industries include the minimum theoreti-
cal energy, the efficiency of selected unit processes, and
the offect of c'~rtain process changes on the energy use.
A computer model was developed to perform the
necessary calculations. The model performs the custom-
ary energy balances based on the first law of thermo-
dynamics. Availability losses are calculated according to
the second law of thermodynamics. The model is general
and can be applied to an entir: ;~dustrial process or to
any of the unit operations withil. the process. The model
identifies large energy or availability losses. The model
can be operated in an interactive mode so that the net
fuel savings. VI hich could be expected to result from
changes in operating conditions and procedures or even
major process changes, can be computed and displayed
in real time. The model is a powerful tool for identifying
those operations in which significant fuel savings can be
realized through straightforward modifications. In addi-
tion, the model will show where the potential for energy
conservation is limited by the irreversibilities inherent in
the basic processes employed. In the latter case, differ-
ent, more reversible processes would have to be employ-
ed to further reduce energy consumption. Such a defini-
tion of the nature of the various availability losses is
invaluable in directing the attention of process engineers
to those operations that offer the greatest potential for
overall fuel savings.
Energy-Use Patterns in Metallurgical and Nonmetallic
Mineral Processing (ref. 3).
Eighty-three commodities were selected as being im-
portant basic industrial materials. The detailed energy-
use appraisal being performed for these commodjties
should be of particular value in assessing the national
486
-------
Table 2. Patterns of fuel use by fuel type by SIC, 1973
~~..a:~.. c.,~ --..#:o'.~~~-~-.:c::-==-~--a:-~'::::":"'';'==-....~~-=-=r t,;;.-:-:.::.--
Percent of total energy consumption
Type of fuel 3334 3339 3351 3352 3357
Propane, butane >1 2 >1 2
Distillates >1 1 >1
Re sid ua 1 oil 9 7 22
Gasol ine >1 >1 >1
Lubri cants >1 1 1 >1
Mi sce 11 aneous
petroleum pr~ducts >1 4 11
Coal >1 22 3
Natural gas 3 22 33 46 12
a 6 19
Coke
Electrical energy 91 37 52 41 54
aFor primary a 1 umi n urn th is; s petroleum coke.
~ .
. pattern of energy consumption. This study differs from
the usual energy analysis in that it includes estimated
energy requirements for mining and beneficiation, con-
sumable raw materials, transportation, and fuels and
electrical energy.
The analysis of the energy-use patterns for each of
the 83 commodities consists of the following compon-
ents:
1. Concise characterizations of the commodities based
on 1973 data;
2. Brief descriptions of the major production pro-
cesses;
3. Flowsheets, which indicate the major raw materials
and unit processes;
4. Estimated energy values for the various unit opera-
tions;
5. Notes and references, which clarify important fact-
ors and indicate the sources used to develop the
tabular data.
The most important facets of commodity character-
ization may be summarized as follows:
1. The magnitude of imports of ores and concentrates
is given. Such imports are of particular significance
with regard to energy because, in the mining and
beneficiation of ores in which the desirable mineral
is present in very small amounts (e.g., nickel ores), a
fairly large amount of energy is consumed in pro-
ducing a' net ton of concentrates. The level of such.
imports also indicates the degree of U.S. depen-
dence on foreign sources for basic raw materials.
2. Imports of partially or completely processed pri-
mary product are shown to indicate the extent of
reliance on foreign sources, which represent an
. energy import.
3. Qualitative measures are made to indicate the level
of byproduct and coproduct recovery, secondary re-
covery (from scrap). and the importance of com-
plementary operations.
4. The amount of U.S. production and consumption
for 1973 is shown. The consumption tonnage
applies only to primary product and does not in-
clude consumption of product from secondary
(scrap) recovery.
5. The total energy required per net ton of primary
product or products is given.
6. The total energy required for 1973 U.S. consump-
tion of primary product or products is shown.
The methodology used to obtain reliable data was a
broad-based analysis and evaluation of information avail-
able from the following sources:
487
-------
1. Previous nonproprietary BCL studies on various
metals, r1inerals, and basic industrial commodities;
2. BCL staff who had pertinent industrial experience
in production or research;
3. A wide I'ariety of published information related to
industriai process descriptions, materials, energy re-
quirements, and statistical data;
4. Wherever possible, energy data for a specific item
from sevE'ral different sources;
5. Discussions with various trade associations for data,
when available;
6. DiscussiMS concerning energy consumption with
representdtives of companies producing the various
primary rroducts;
7. Review, where possible, of the energy analysis by
industry representatives to obtain comments and
suggestions;
8. Reviews and discussions with Bureau of Mines staff
associated with the study.
The results of the estimation of the total energy
requirements (in trillion Btu) for the U.S. consumption
of selected commodities permit the ranking as shown in
table 3.
The desi~cnation H following the estimated energy
requirement indicates that the particular commodity was
prejudged to have high significance to this study; the
designation I indicates an intermediate level of signifi-
cance.
Table 3. Ranking by total energy required for
U.S. consumption (trillion Btu)
* Iron + S tee 1 ( All )
*Aluminum (Ingot) .
Port 1 an d Cenen t
Nitrogen (Ammonia)
*Copper (t,11)
Glass (Containers)
Chlorine (Gas, Liquid)
Calcium (Quicklime)
Phosphor~s (Elemental, Acid)
Zinc (Metal)
Titanium (Sponge + Oxide)
Oxygen (Gas, Liquid)
~1anganese (Mn, Fe-Mn)
Ceramics (Common Brlr~)
Nitrogen (Gas, Liquid)
MagnesiulT (Metal) .
3,770 H
1 ,480 H
688 H
586 H
236 H
216 H
199 H
182 H
147 H
92H
78 I
73 I
66 I
62 H
48 H
42 I
Table 4. Ranking (continued) by total energy
required for U.S. consumption (trillion Btu)
Sodium (Na, NaC1)
N i cke 1 ( All)
Sil icon (Ferro-)
Chromium (Ferro-)
Potassium (Chloride)
Cl ays(A 11)
Lead (Refined)
Gypsum (Calcined)
Refractories (Basic Brick)
Refractories (Fireclay Brick)
Uranium (Oxide)
Fluorine (Fluorspar)
Molybdenum (Oxide)
Sulfur (Acid)
39
36
35
35
33
31
24
19
18
7
7
7
6
1
Table 4 lends some sense of perspective in that the
range of estimated energy requirements is displayed;
metal commodities have been emphasized.
It is obvious from table 5 that iron and steel, alumi-
num, and copper head the list of primary metal com-
modities in terms of energy consumption. As you can
see, the specific energy requirement per ton of steel is
quite low; the process has been judged (elsewhere) to be
relatively efficient. The high energy consumption derives
from the fact that the U.S. steel industry in 1973 poured
the equivalent of 128 million net tons of steel slab. The
manufacture of aluminum ingot represents the highest
specific energy requirement.
This presentation focuses on copper and aluminum,
each of which is treated as a unit in the following pages.
The iron and steel industry, while enormously impor-
tant, will be treated in subsequent publications.
Studies of the sort reported here require the aggres-
sive support of a number of talented people. The author
must consider himself a spokesman for this group. While
the author has made (and continues to make) contribu-
tions, it is mandatory that the following key Battelle
staff members be recognized in terms of the content of
this paper (A, B, and C denote department affiliation):
Phone
Ext.
Mr. E. S. Bartlett (A) ...............2873
Mr. H. N. Conkle (B) ...............2485
Mr. D. C. Drennen (A) .............. 1861
Dr. J. E. Flinn (B) . . . . . . . . . . . . .1615
(Project Manager, ref. 1)
488
-------
Table 5. Energy requirements for the three largest
primary metal consumers
Commodity (product)
Energy
requi red
T ota 1 u. S.
consumption,
trillion Btu
Per net ton,
mi 11 ion Btu
Iron + steel
(Steel slabs)
(Gray iron castings)
(Steel castings)
A1 uminum
( Ingot)
Copper
(Cerren t coppe r)
(Refined copper)
24
34
42
3,350
366
54
3 , 770
244
1 ,408
87
112
15
221
236
Mr. R. W. Hale (C) ................2151
(Project Manager, ref. 3)
Dr. E. H. Hall (B) .,.......
(Project Manager, ref. 2)
Mr. J. B. Hallowell (A) ..............1195
. Dr. W. T. Hanna (B) . . . .3376
Mr. A: O. Hoffman (A) ..............1889
Mr. H. W. Lownie (A) . . . . . . . . . . . . . . .2882
Mr. D. J. Maykuth (A) ..............2234
Dr. E. J. Mezey (B) .,.............. 1448
Dr.C.Mobley(A) ............ .3215
Mr. R. J. Nekervis (A) . . . . . . . . . . . . . . .2522
Dr. L. D. Reed (8) ................3509
Mr. J. Varga, Jr. (A) . . . . . . . . . . . . . . . .2941
Dr. D. N. Williams (A) . . . . . . . . . . . . . . . 1597
. . . .1888
A = Metallurgy Department (Mr. F. C. Holden, Manager)
B = Energy and Environmental Processes Department
(Dr. E. W. Ungar, Manager)
C = Resource Management and Economic Analysis
Department (Dr. N. L. Drobney, Manager)
The author acknowledges for Battelle-Columbus the
very extensive contract support provided by the U.S.
Bureau of Mines, the U.S. Department of Commerce,
and the U.S. Federal Energy Administration.
COPPER
Estimated Energy Requirements for the Production of
Refined Copper
Table 6 summarizes the production of refined
copper. The production of copper is shown schematic-
ally in figure 1. The estimated energy requirements
(million Btu per ton) are given for each operation.
The commercially important copper ores are sul-
fides of copper, copper-molybdenum, copper-zinc, and
copper-lead-zinc. Most ores contain. small quantities of
silver, gold, platinum, selenium, and tellurium, which are
recovered as byproducts. Mine(:l ore containing less than
0.3 percent copper is deposited in waste rock dumps,
from which some copper is recovered by dump leaching.
In this study, estimates were based on data for eight
open-pit mines. The ore grades ranged from 0.55 percent
copper to 0.92 percent copper, and averaged 0.7 percent
copper.
In the large open-pit mines, the source of most
domestic copper ore, stripping ratios (ratio of waste rock
to ore) vary from about 1: 1 to as high as 12: 1. Much of
the energy expended in mining is consumed in the exca-
vation and hauling of overburden. In one mine, about 40
million Btu per ton of copper was expended to mine ore
489
-------
Table 6. The production of refined co'pper
Imports
Ore and matte
Buster and black
Refined
Byproduct recovery
Secondary recovery (1972)
Complementary operations
U.S. production
U.S. consumption
Energy requi red
pe r net ton
for U.S. consumption
42,880 net tons (Cu)
154,101
201 , 511
Silver, gold, platinum,
selenium, tellurium,
molybdenum, sulfuric
acid
465,000 net tons
Production of sulfuric
acid is primary means
of controlling sulfur
emissions; acid used
for leaching and sale
1.87 million net tons
1.97 million net tons
112 million Btu
221 trillion Btu
which averaged 0.7 percent copper at a stripping ratio of
12:1. In anott er mine, 9 million Btu per ton of copper
was expended to mine ore which averaged 0.55 percent
copper, but where the stripping ratio was 2.5:1.
Explosives consumption was taken as 0.25 pounds per
ton of mater al mined, according to information ob-
tained from two copper producers.
The ener!lY required for truck haulage of ore and
overburden and for electric railroad handling was based
on published information (ref. 4). These values for haul-
ing represent ilbout 15 percent of the estimated energy
consumption in mining. Estimates made by one com-
pany indicatec that haulage accounts for about 50 per-
cent of the total energy consumed in open-pit mining.
(ref. 5).
The ore from the mines ,is sent to the concentrators,
which include crushing. grinding, and flotation opera-
tions. Kellogg has made a quaititative analysis of the
energy required for the milling and flotation of various
types and grades of the ores of copper, molybdenum,
zinc, lead, and iron (ref. 6). His analysis was based on
data published by the Bureau of Mines (ref. 7). Kellogg
confirmed that the most significant fact in energy con-
sumption is the grade of ore. Over the years, as, the
grades of all kinds of ores have decreased, more energy
has been needed to extract the mineral values from
them. In this study, milling data from three copper com-
panies were analyzed. The energy, consumed varied from
38 to 56 million Btu per net ton of copper for an
average of 26 million Btu per net ton of copper. This
amounts to 60 percent of the average total energy (42
million Btu) used at the concentrator.
The concentrate, which contains about 25 percent
copper, is sent to the smelter charge preparation plant,
with cement copper produced by leaching and precipita-
tion operations. In most smelters, this mixture is furnace
,dried before it is charged into reverberatory furnaces.
The energy consumed, 24.645 million Btu, is the average
of reverberatory furnaces in three different smelters.
Two calcine-charge reverberatory furnaces in other
smelters were analyzed and found to consume an average
of 13.9 million Btu. Roasting operations in three smelt-
490
-------
,...------~
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I 1
I I
I I
I :
I I
I
I I
I I
I CONCENTRATOR I
L___----_J
OllUTE CQPI'ER
SOLUTION
TAiliNGS
CONCENTRATe
SMEL TER
-
--- ---- -------,
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
I
I
I
- - - - - - - - - .- - - J
r---------------------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
L--- -------------- -------
WASTE GAS
TO STACK
SULFURIC
ACIO
ANOOES
GOLD
$IL VER
S(LINIUM
PLATINUM
,.----------- ---
I REFINERY
I-
I
I
I
I
L__---""';--------
---------
CA THQOes
REFINERY
SHAPES
REFINEO COPPER
112.292
Figure 1.
Production of refined copper by conventional mining
and smelting processes.
491
-------
crs cOllsullwd !In !lverage of 1.84 million Btu. Exclusive
of roasting or drying operations, the calcine reverberato-
ry furnaces consumed about 10 million Btu less than the
green-charge reverberatory furnaces. Many other factors,
including copper content of the concentrate, heat loss
from the furnace, and operating variables, affect the
energy consJmed in green or calcine reverberatory
furnace oper~ tions.
In some )Iants, the concentrate is roasted to convert
some of the i;ulfides to oxides and to eliminate some of
the sulfur. III roasting, most of the required energy is
derived from the burning of the sulfur in the concen-
trate. Roastej concentrate, or calcine, is then charged
into the reverberatory furnace. In addition to concen-
trates and coment copper, the reverberatory furnace
charge includl!S silica rock, lime rock, and ore as fluxing
agents, and copper-bearing dust collected from the elec-
trostatic preci pitators.
Waste hea1 from reverberatory furnace flue gas is
recovered in mcuperators and waste heat boilers. The
steam generat ~d is used in the smelter power plant or is
consumed in other plant operations. Up to about 50
percent of th~ fuel energy consumed by the reverbera-
tory furnace i~ recovered.
Matte, a mixture of copper sulfide and iron sulfide,
which contains from 30 to 60 percent copper, is tapped
from the revel beratory furnace into ladles for transfer to
the converten,. Waste slag from the reverberatory fur-
nace contains very little copper and is sent to the slag
dump.
The convmters are cylindrical furnaces into which
compressed ai' is blown through tuyeres below the sur-
face of the liquid matte. The sulfur and iron in the matte
are oxidized, I ~aving an impure blister copper containing
between 98 ar.d 99 percent copper. Most of the heat in
the converter operations is provided by the oxidation of
the sulfur and iron. A little fuel is needed to keep the
converters hot between charges. Electrical energy or
process steam is consumed by the blowers, which pro-
vide air. Slag from the converters is recycled to the rever-
beratory furna:cs. Converter off-gases containing about
4 percent sulf J, dioxide are cooled, cleaned of dust,
then sent to thli sulfuric acid plant.
Blister co~per is fire refined at the smelter in the
anode furnaces, which may be reverberatory furnaces or
cylindrical furnaces similar to the converters. Air is
blown through the molten blister copper to complete
the oxidation of iron and to oxidize minor impurities
such as nickel. After the final oxidation step, the slag is
removed and re turned to the rev"rberatory furnace. The
oxygen content of the copper is ~educed by bubbling
reformed natu-al gas thlvugh the melt (poling). A
200.ton heat ill the anode furnace may requim 1.5 hours
for blowing and slag skimming, and 3 more hours for I
poling. This operation consumes significant quantities of
petroleum fuel, natural gas for poling, and electricity for
air blowing. Copper anodes are cast directly from the
anode furnaces into coated copper molds.
The anodes, which contain 99+ percent copper, are
shipped to the refinery. It was estimated that about
750,000 tons per year of anodes and blister copper are
shipped 2,000 miles from smelters in Western States to
refineries on the East Coast. This amounts to 910 net
ton-miles per ton of product, based on 1.65 million net
tons of blister and anode production in 1972.
The major operation of the refinery is the electro-
lytic refining of anode copper to cathode copper, which
contains 99.8+ percent copper. The tankhouse is a large
consumer of electrical energy for electrolysis and of
natural gas or oil for heating the copper solution. The
finished cathodes are sold or are melted and cast into
refinery shapes such as wirebar, cakes, or continuously
cast wirerod. Silver, gold, platinum, and selenium, the
principal byproducts of the copper refinery, are recover-
ed from the slimes that drop from the anode into the
cell during electrolysis.
Energy consumption figures shown in the flowsheet
take into account the energy required for existing pollu-
tion control processes and equipment. The primary
pollution source in the copper industry is sulfur dioxide
generated in roasters, reverberatory furnaces, and smelt-
ers. Multihearth roasters, used extensively in the past,
and conventional reverberatory furnaces are not easily
adapted to the recovery of sulfur or sulfuric acid because
the effluent gases contain small amounts of sulfur
dioxide. Gases from converters contain enough sulfur
dioxide (4 to 6 percent) to permit recovery as sulfuric
acid. Some smelters have installed fluidized-bed roasters,
in which the effluent gases contain 12 percent or more
of sulfur dioxide and for which acid plants can be oper-
ated economically. Other companies have discarded the
conventional roasters or are using them for the drying of
concentrate that is charged into green reverberatory fur-
naces.
Gases from reverberatory furnaces contain about 2
percent sulfur dioxide. These gases may be passed
through electrostatic precipitators before being dis-.
charged to the atmosphere through high stacks. The in-
dustry is using a monitoring system to measure the sul-
fur dioxide concentrations at ground level at various dis-
tances up to 20 miles from the stack. When the ground-
level concentration of sulfur oxides reaches an unaccept-
able level, the reverberatory furnaces are temporarily
shut down. This method of control, referred to as "sea
492
-------
laptaining:' has been successful in meeting current air
~uality regulations, but creates a loss in operating time
of about 10 to 15 percent.
Meaningful data on energy needed for controlling
sulfur dioxide emissions are not generally available from
the copper industry. One company provided estimates
for its mining, milling, and smelting operations at various
plant locations. The data do not cover individual process
operations. The tabulation shown below presents esti-
mates of additional energy that would have to be
expended at one smelter (reverberatory, converter, and
anode casting operations) if the plant were equipped
with controls under the present regulations:
Emission control energy,
Year percent of total smelter energy
1972....... .1.1
1973 . . . . . . . . . . . . . . . . 1.4
1974 . . . . . . . . . . . . . . . . 1.6
1975 . . . . . . . . . . . . . . . 13.0
1976 . . . . . . . . . . . . . . . 17.0
The final estimate of the energy requirement for
refined copper was derived almost entirely from data
obtained from questionnaires sent to copper producers
and from site visits to selected companies. Considerable
judgment had to be exercised in the necessary interpreta-
~ion of this information. The estimated 112 million Btu
er net ton of copper is higher than most published
estimates; it should be noted that this value includes the
energy value for cement copper. One company, which'
provided what appears to be the most carefully. analyzed
energy data available to Battelle, estimated its total
energy consumption to be 125 million Btu per net ton.
Potential Energy Savings
In the conventional copper smelter, the major
energy-consuming operations are the concentrator and
the smelter, which require approximately 38 percent and
34 percent of the total energy, respectively. More than
half of the energy used in the concentrator is used for
fine grinding of the ore prior to flotation. In the smelter,
more than half the required energy is used for reverbera-
tory smelting, or about 20 percent of the total energy
used for the production of refined copper.
Potential Energy Savings in Grinding Operations
The actual energy consumed in grinding depends
very much on the nature of the ore. Principal variables
are the ore hardness and the degree of grinding necessary
for substantially complete liberation of the valuable
components. If the energy consumption is expressed
relative to the tonnage of refined copper product, fac-
.ors such as ore grade and recovery in subsequent proc-
essing steps are to be included in the analysis. Therefore,
a wide variation in grinding energy per ton of final pro-
duct is to be expected.
In commercial practice, the grinding process is not
completed in a single pass. A portion of the mill product
is recirculated continually. This circulating load typically
is 200 percent to 300 percent of the feed. The optimum
circulating load depends on a number of items such as
the grindability of the ore, the cost of electrical energy,
and the capital investment in the mill; it must be deter-
mined for each case.
Once the optimum circulating load has been deter-
mined, every effort should be made to operate a mill
under conditions as close to the optimum as possible,
which requires that adjustments be made almost contin-
ually because of variations in feed load and characteris-
tics and because of wear of the grinding media. Until
recently, it was difficult to exercise close control be-
cause of lack of suitable instrumentation. Presently,
several instruments are on the market that rapidly deter-
mine the particle size distributions in the various streams
in the mill circuit. These instruments can provide the
feedback signals necessary for computerized control of
the circulating load at the optimum level. Energy savings
typically of the order of 10 to 15 percent of the energy
presently used for fine grinding have been published (ref.
8).
There are indications that energy can be saved by
the installation of rubber liners in the mills. Although
results were not impressive when the liners first became
available several years ago, it appears now that, with
some minor changes in design, energy savings of the
order of 10 percent are readily obtainable (ref. 9).
Another new development that is claimed to reduce
energy consumption is the use of fluid energy mills (ref.
10). These devices submit ore particles to tensi Ie forces
rather than to the compressive forces used in con-
ventional grinding. Although the tensile strength of rock
minerals is known to be much less than the correspond-
ing compressive strength, net overall energy savings
evidently have not been demonstrated to potential users.
One significant problem area is the wear of components
in contact with high-velocity solid particles. Until energy
savings are clearly demonstrated, no recommendations
can be made. Further developments deserve some atten-
tion.
Potential Energy Savings in Smelting and Refining Oper-
ations (Without Major Process Changes)
Developments in copper smelting during the last
two decades do not reflect a special concern for energy
savings. Motivating forces have been the need for effec-
493
-------
tive and least expensive air pollution control systems and
for overall cost reductions through process simplifica-
tion. As a result, two general trends have emerged.
In existing smelters, roasting operations have been
reduced or phzsed out in favor of wet charging of rever-
beratory fur.laces. This has reduced operating and
maintenance CDstS at the expense of higher fuel con-
sumption. In other cases, flash smelting furnaces have
been installe(l, again eliminating roasters, but also re-
placing the rI'verberatory furnace with a more efficient
unit. Predictably, smelters of this design show the lowest
energy consumption in existing copper smelters. The
Outokumpu flash smelting process is being installed in
New Mexico.
In recent years, much emphasis has been placed on
development of a continuous smelting process to con-
serve energy iII~d to reduce air pollution. Presently, two
continuous smelters are being constructed, one in
Canada using the Mitsubishi process and one in Utah
using the Noranda process. The use of new smelting
processes represents major process changes, which are
discussed later.
One relatively new development that has not been
implemented ':0 a significant degree is the electrorefining
of copper at 1igh-current densities using a new cell de-
sign that may permit savings in capital investment. The
energy requir,~d in refining at high-current density is
significantly greater than for copper produced conven-
tionally.
Individual companies may achieve significant energy
savings by rep acement of inefficient haulage equipment
by newer items of larger unit size, a situation that is not
typical for thu industry. Similarly, individual refineries
may benefit from the use of titanium cathodes. As a
rule, these kin:ls of improvements have been implemen-
ted gradually by most producers.
There appears to be little additional opportunity to
conserve ener~ y in conventional processing. One large
copper companv, after making a detailed study of exist-
. ing plants, corcluded that an energy savings of about 1
percent may b~ possible without installing a new smelt-
ing process or other improvements requiring large capital
expenditures (ref. 11).
Potential Ener9Y Savings With Major Process Changes
Based on (1) the energy requirements to produce
one net ton of refined copper by conventional means
and the total annual energy associated with the U.S.
consumption of refined copper; (2) the complexity of
the conventional production prr:;esses in terms of the
number of unit operations and/o;' chemical reactions
associated with refined cOtJper production; and (3) the
availability of known alternative energy-saving produc-
tion processes for producing refined copper, it is con-
sidered highly desirable to develop energy-saving replace-
ment technologies (where feasible) for the manufacture
of copper.
Energy-saving replacement and/or alternative proc-
esses are generally directed toward replacement of the
smelting steps. Several operations that offer an opportu-
nity to save energy relative to the conventional smelting
procedures include (1) Outokumpu-type flash smelting,
(2) continuous smelting, such as the Noranda, the
WORCRA, the Q-S, and the Mitsubishi processes, (3)
autogenous smelting, and (4) blast furnace smelting.
The type of flash smelting developed by the
Outokumpu Smelter in Finland and a variation deveiop-
ed by International Nickel Company in Canada provide a
significant fuel savings in comparison with reverberatory
smelting. Flotation concentrates, with flux and pre-
heated ore, are injected into a hot chamber where the
flash burning of sulfides in suspension furnishes the
energy needed for smelting. This reduces the need for
auxiliary fuel input and also reduces the volume of the
combustion products. Off-gas is produced with the
Outokumpu Flash Smelter, which contains 8 to 14 per-
cent sulfur dioxide.
The International Nickel variation uses 95 percent
oxygen in place of air. A more concentrated gas is pro-
duced, which contains 70 to 80 percent sulfur dioxide.
Kellogg has indicated that the Outokumpu-type flash
smelter requires only about 40 percent of the energy
used in a typical green-charge reverberatory furnace (ref.
12). The Inco-oxygen Flash Smelter is anticipated to
provide even greater net energy savings relative to the
conventional reverberatory smelter. Besides improved
utilization of energy and diminished gas volume, flash
smelting practices are considered to yield a higher grade
matte than reverberatory smelting. High-copper-content
slags are produced in flash smelters and additional
energy would be required to reclaim the copper from the
high-copper slag.
Combining the conventional reverberatory-converter
operation into a single unit should provide an energy
savings. A number of these continuous smelting proc-
esses have been developed to the pilot-plant stage and a
few to the semi commercial stage. Several of the continu-
ous smelting processes are de~cribed.
A process developed in Canada by Noranda Mines,
Ltd., uses a single, long (about 70 fee!), combined smelt-
ing-converting unit (refs. 13,14). The unit consists of a
horizontal, cylindrical furnace having a central depressed
area for copper collection and a round hearth at one end
for slag removal. A burner heats the smelting end where
concentrates and flux are charged. Air or an air-oxygen
mix is introduced through tuyeres along the base of the
494
-------
turnace to oxidize the matte that is formed. Reducing-
'gas conditions are maintained at the opposite end of the
furnace for slag reduction and settling. The gas, which
leaves near the center of the furnace, is of sufficient
sulfur dioxide content for the subsequent production of
sulfuric acid. After testing a 100-ton-per-day pilot opera-
tion, Noranda has built an 800-ton-per-day commercial
operation. Kennecott has also announced plans to install
a Noranda-type continuous smelter. Whil~ detailed
energy requirements are not known for the Noranda-
type continuous smelter, this process should allow an
energy savings over the conventional smelting process.
The WORCRA continuous smelting process has
undergone pilot testing over the last few years (refs.
13,15). This process combines smelting, converting, and
slag cleaning in one operation within a long stationary
furnace. Molten <;opper is removed from one end of the
unit, slag is removed at the other end, and oft-gasses are
removed as a single stream for subsequent processing. In
place of the tuyeres used in the Noranda process, air
lances enter from the top of the furnace. An analysis by
Kellogg of the operating data for the WORCRA furnace
at the Port Kembla pilot plant did not indicate a signifi-
cant energy savings relative to conventional reverberato-
ry smelting (ref. 12).
~ Named after the inventors Paul Queneau and
I'.:einhardt Schumann, the Q-S process is also a multi-
stage progressive converter operation that combines con-
tinuous smelting and converting in one furnace. Sulfide
concentrates are flash-smelted with oxygen, which also is
introduced through submerged tuyeres to effect the pro-
duction of copper. A slag scavenging operation is also
part of the process. Pilot-plant investigations of the Q-S
process are currently in the planning stage; no estimate
of the energy requirements is available (ref. 16).
A process developed by Mitsubishi Metals Corpora-
tion of Japan differs from the Noranda and WORCRA
continuous processes in that not all of the processing is
done in a single unit (refs. 17-19). The concentrates are
smelted in one furnace, from which the slag and matte
flow continuously through a slag-cleaning furnace into a
converting furnace equipped with overhead air lances.
The process features countercurrent flow as matte moves
from the smelter furnace to the converter, while the
converter slag is returned to the smelter.
The Mitsubishi process has been scaled to a
1,500-ton blister copper per month semicommercial
plant which started operation in November 1971.
Steady-state operation and control of the Mitsubishi
process is achieved by on-line, feed-forward computer
control of the inputs of air and flux to the converter
",rnace. Assuming that long-term, steady-state operating
"nditions can be maintained, the Mitsubishi process
may make possible significant reductions in energy con-
sumption relative to the conventional smelting practices.
Autogenous smelting work has been conducte9 on a
laboratory scale by the United States Bureau of Mines
(ref. 20). Like the Noranda and WORCRA processes, the
Bureau of Mines' autogenous smelting feasures continu-
ous operation of a single unit to produce copper directly
from concentrate. The furnace combines flash smelting
with converting by means of an oxygen lance immersed
through the slag into the matte. While this process
appears to offer the potential for energy saving relative
to the conventional smelting processes, no further devel-
opment of the process appears to be underway.
The Momoda blast furnace developed by Sumitomo
Metal Mining Company is currently used by two copper
smelters in Japan (ref. 21). Concentrates are charged to
the blast furnace with other copper-bearing materials as
a stiff plasticized mass containing 10 to 15 percent
water. The energy requirement for smelting a ton of
charge in a Momoda blast furnace is given as only about
28 percent of the energy required for reverberatory
smelting with a wet charge; thus there appears to be a
substantial energy savings. While the blast furnace smelt-
ing of wet concentrate may provide a lower energy route
to copper production than the conventional smelting
processes, a detailed analysis of the blast furnace opera-
tion is required before the magnitude of the energy
savings is known.
Copper may also be produced by hydrometallurgical
processes such as Anaconda's Arbiter process, the
Sherritt-Gordon process, the Cymet process, and the
Duval process (refs. 22-24). While there are few precise
data available on the fuel and power requirements, these
hydrometallurgical processes do not offer any potential
for energy conservation relative to conventional smelting
processes. Kellogg has estimated that the hydrometal-
lurgical processes for producing 'copper require at least
twice the energy associated with the conventional pyro-
metallurgical smelting processes (ref. 12).
ALUMINUM
Table 7 summarizes production of aluminum ingot
from bauxite by the Bayer-Hall process.
Estimated Energy Requirements for the Production of
Aluminum
A simplified flowsheet for the primary production
of aluminum by the Bayer-Hall route is shown in figure
2, which includes the manufacturing unit operations for
carbon anodes and cathodes. Estimated energy consump-
tion is noted for each unit operation. Aluminum produc-
tion may be subdivided into four major steps: mining,
495
-------
CARBON ANODE
MANUFACTURE
20.83
CARBON
CATHooe .
MANUFACTURE
1.21
LIME
0.85
CAUSTIC SOOA
4.5
~
~
SEED
ALUMINA
TRIHYDRATE
CRUSHING. WASHING.
AND SCREENING
0.22
RECLAIMED
NaOH
SPENT LIQUOR
RECOVERY
1.56
ALUMINA
FLUORSPAR
0.01
ELECTROL YSIS
168.00
ALUMINUM INGOT
243.90
Figure 2. Simplified flowsheet for the production of aluminum from
bauxite by the Bayer-Hall route.
496
-------
Table 7. Production of aluminum ingot from bauxite by the
Bayer-Hall process
Imports
Bauxite
I n go t
Byproducts
Secondary recovery
Complementary operations
U.S. production (primary)
U.S. apparent consumption
Energy required:
Per net ton
Total U.S.
85 percent of consumption
11 percent of consumption
Not important (Gallium)
1.23 million net tons
Important (electrode
making, 10 percent of energy)
4.53 million net tons
5.77 million net tons
2 . 44 mill i on B t u
1408 trillion Btu
~handling, and drying of bauxite at the mine site, trans-
portation of ore to the bauxite refinery, refining of the
bauxite into alumina, and smelting the alumina into
aluminum. .
Most bauxite is surface mined. It is then crushed,
ground, and kiln dried to remove excess moisture.
Energy requirements for mining have been estimated
from the literature on the basis that one net ton of
aluminum is received from 4.7 tons of bauxite, the 1974
U.S. industry experience (refs. 25,26,28). The processed
ore is transported to the bauxite refinery, which requires
an estimated average energy consumption nearly as large
as that for the actual mining operations (refs. 27,29).
Estimates of the energy required for the transportation
of imported bauxite were based on 1973 statistics;
imports from each of the exporting countries were
proportionately distributed 00 a ton-mile basis.
Purified alumina is manufactured from bauxite ore
in the Bayer process. The ore is digested in caustic soda
at elevated temperature and pressure. The insoluble
impurities are filtered from the resulting solution of
sodium aluminate. The alumina trihydrate, which precip-
itates from the cooled solution, is then filtered, washed,
and calcined in kilns to drive off water of hydration to
yield chemically pure alumina.
~ At the smelter, the alumina is dissolved in the
,molten cryolite electrolyte in the Hall-type cells. Molten
aluminum deposits on the carbon lining at the bottom of
the cell, which acts as the cathode. As the aluminum
builds up in depth, it is siphoned off. Oxygen in solution
combines with the carbon anode at the top of the cell
and is released as carbon dioxide. Estimates of the
energy requirements for the reduction of alumina to
aluminum in the Hall cell are based primarily upon infor-
mation published by the Bureau of Mines and on indus-
try contacts (refs. 26,28,33). Estimates of the energy
requirements for the production of carbon anodes and
cathodes are based on contacts with and publications by
the aluminum and carbon industries (refs. 28,31,32).
The estimates of energy usage do take into account
present knowledge of energy expenditures for pollution
control. The most obvious source of emissions to the
atmosphere in the aluminum smelter is the gas and
fumes from the reduction cells, which contain carbon
dioxide, carbon monoxide, sulfur oxides, and fluoride
particulates. In plants using Soderberg anodes, the emis-
sions contain tars and oils. Flue gas from the anode bake
plant in plants using pre baked anodes contains all of
these constituents, since anode stubs are recycled. Dust
is formed if solid-pitch-handling systems are used. A
final source of pollution is the chlorine-nitrogen-carbon
monoxide degassing of molten aluminum in a batch
operation carried out in holding-alloying furnaces.
The efficiency of the reduction-cell hooding is a
497
-------
factor which determines the pollution control measures
used. Where high-efficiency hooding is possible, air
pollution cor trol devices may be applied to the primary
pot gas. Lower efficiency hooding (because of cell de-
sign) necessitates the treatment of pot-room (secondary)
air. In the ca ie of primary control of the pot line, either
dry or wet g" factor of 10,500 Btu
per kWh. (Thi:; factor is used urllformly for all commod.
ities included in the master study to assure a common
basis.) For aLJrninum. another perspective should be
noted. Almost 40 percent of the electricity consumed by
the aluminum industry is generated by hydropower. Thel
AI uminum Association compiled and supplied the
following figures on actual electric power sources in U.S.
aluminum smelters (1974). '
~ Percent
of
swwl.Y.
Conversion
factor
8tl] per kWh
Hydropower
Purchased thermal
. power
Self-generated
thermal power
Total
39.9
41.4
3,413
10,500
18.7
12,370
100.0
8.022
Using the Aluminum Association factor of 8,022
Btu per kWh would lower the estimated energy required
per net ton of primary product from 244 million Btu to
204 million Btu and the total energy required for U.S.
consumption from 1,408 trillion to 1.177 trillion Btu.
Potential Energy Savings for Alumina Reduction Cells
Without Major Process Changes
There are several areas iri the Bayer-Hall process.
particularly in the reduction of aluminum in the Hall
cell, where modifications in techniques lead to a reduc-
tion in energy requirements. Unfortunately. these
measures markedly increase the capital cost. Indeed, the
magnitude of these additional costs is such as to override
the benefits of more efficient operation. Except in cases
of plant expansion, such expenditures appear to be more
than can be absorbed in the near future by the primary
aluminum industry.
Modern reduction cells operate on an electrode
potential of 4.5 volts with current flows up to 100.000
amperes. The Faraday (electrochemical) efficiency is 90
percent. However. the power efficiency is only about 45
percent, owing principally to voltage drops caused by
high anode and electrolyte resistances.
The high, and apparently inefficient, energy con-
sumption in the Hall cell has caused a continuing effort
to reduce power requirements, and it has borne fruit.
Whereas the U.S. industry average for power consump-
tion is around 8.0 kWh/lb of aluminum. the more effi-
cient U.S. plants produce aluminum at around 6.5
kWh/lb. The Pechiney plant of Auzat, France. produces
33,000 annual net tons from 140 pots using prebaked
anodes operating at 90 KA with a specific power con-
sumption of 5.9 kWh/lb (ref. 34). In this connection, it
should be noted that. owing to high energy costs, pro-
grams to reduce energy consumption started 10 years
earlier in Europe than in the United States.
These reductions in Hall cell energy requirements
498
-------
have been achieved primarily by the following interre-
lated modifications (refs. 35-37):
1. Installation of larger capacity Hall cells, up to
100,000 amperes, to achieve lower heat losses by
reducing the exposed cell surface area per unit of
output;
2. Reduction of current density in these larger cells to
accommodate the increased anode size and to
improve power efficiency;
3. Reduction of the anode and cathode resistances by
more efficient lead arrangements which include
increasing bus bar cross sections and redesigning the
cathode section;
4. Redesign of the bus bar layout to maintain a stable
metal/flux interface by minimizing electromagnetic-
ally induced turbulence which tends to become in-
creasingly severe at cell sizes above 85,000 amperes
capacity.
Purely operational procedures have been adopted to
improve efficiency; among these are:
1. Improved control of the interpolar gap at the
minimum at which stable operation is maintained;
2. Improved control of temperature as near to the
liquids of the flux as possible;
3. Improvements in systems for supplying alumina so
as to maintain a maximum amount in solution in
the electrolyte;
4. Improvements in the systems for supplying fluoride
salts to the electrolyte;
5. Improvements in anode quality;
6. Raising the conductivity of the electrolyte with
additions such as lithium and magnesium fluoride,
which allow higher current densities without more
rapid consumption of the anode.
. Making additions to the electrolyte bath raises
operating costs. Other recommended measures in the
listing above raise capital costs to a marked degree, and
the economic justification depends on the costs and
availability of power (ref. 38). These modifications will
be incorporated into three U.S. plant expansions that are
under construction, as well as in two planned expansions
and a new hydroelectric plant on which environmental
permits are being sought (ref. 39). Together these will
raise U.S. capacity 12 percent over the next few years
and will reduce the average U.S. Hall cell electrical
energy requirement by a few percent. (The new Alcoa
aluminum chloride prototype plant under construction
is not included in this summarization.) ,
It should be noted that the new construction will
use prebake anode systems rather than Soderberg anode
systems, which have much higher voltage losses and thus
~onsume much more energy.
There is still another condition that affects energy
usage in Hall cell technology that is based on the cost
and availability of energy. In some hydroelectric plants,
particularly some in the Pacific Northwest where power
costs have averaged 2.0 and 2.5 mils per kWh, output is
maximized with some loss in the energy efficiency. In
other regions where energy costs are 6 to 7 mils per kWh
and up, operation at maximum energy efficiency is pre-
ferred, especially in periods of slack demand (ref. 40).
Potential Energy Savings for Bayer Processing Without
Major Process Changes
The energy consumed in producing alumina from
bauxite is around 17.5 percent of the total energy re-
quired to produce aluminum. The process is essentially a
large heat exchanger with hot-end digestion of bauxite
and cold-end precipitation of alumina. Energy consumed
in the process is primarily process heat (90 percent) and
electrical energy (10 percent). Over 90 percent of the
total energy introduced into the process becomes waste
heat. Accordingly, regenerative heating should be em-
ployed wherever possible, and heat losses in existing
equipment should be kept as low as practicable (ref. 41).
Other changes in Bayer processing which would re-
sult in energy savings are conversions which would re-
quire considerable capital costs. One of these is a conver-
sion of the process to produce fine, floury alumina
rather than the coarse, sandy alumina specified in the
United States (refs. 41,42). If European plants, which
produce fine alumina, were to switch processes so as to
produce the coarse, sandy alumina specified in the
United States, productivity would drop 35 percent while
the consumption of energy per unit of output would
increase by 30 percent (ref. 42).
One area where conversion of existing equipment is
economic is the replacement of rotary kilns, used for
calcining the alumina, by fluidized-bed calciners. The
fluidized-bed calciner uses 30 percent less energy per
unit of output than rotary kilns (ref. 43). The industry is
in the process of converting.
Potential Energy Savings by Process Replacement
The estimated energy required by the U.S. alumi-
num industry to produce one net ton of aluminum ingot
is 244 million Btu. Based on recent Battelle work, the
theoretical minimum energy required to convert bauxite
to aluminum ingot is 28.3 million Btu per net ton (ref.
2). The efficiency associated with U.S. aluminum ingot
production is slightly less than 30 percent on the basis of
3,413 Btu per kWh, or 14.2 percent if the electrical
input is taken at 10,400 Btu per kWh.
The efficiencies of the major energy-consuming
Bayer and Hall processes are less than those of many
other major unit operations in the steel, copper, and
499
-------
glass industries.. Based on the energy requirement and
the efficiencies of the major unit operations for alumi-
num prodUl:tion, it is considered highly desirable to
develop enel gy-saving, replacement technologies (where
feasible) for the manufacture of aluminum.
Alternative processes for the production of alumi-
num have bI!en reviewed and described in a survey by
Ing and Zee:leder and in a recent paper by Peacey and
Davenport ~refs. 44,45). These alternative processes
typically fal: within three categories: replacements for
the Bayer process, replacements for the Hall process
which uses Bayer-process aluminum as the primary input
material, amI processes which replace both the Bayer
and Hall processes. Alternative processes of these three
types are disc ussed below.
Alternatives to the Bayer Process
Of the Clluminum-bearing materials, bauxite is the
richest in aluminum. Untreated bauxite typically con-
tains 55 to 65 percent alumina so that about 4 pounds
of bauxite al e processed per pound of aluminum pro-
duced. Typic)lIy, 8 to 9 pounds of clay or kaolin are
required per pound of aluminum. The use of a leaner ore
than bauxite implies that more ore must be treated and
that more enl!rgy is required to produce a pure alumina.
Because baux.te suitable for the Bayer process is not as
available as kaolin or clays, numerous attempts have
been made to develop alternative processes. Wet-type
replacements ':0 the Bayer process do not appear to offer
any major energy savings.
Probably the best known electrothermal process is
the Pedersen :Jrocess, wherein bauxite, coke, limestone.
and iron ore are melted in an electric-arc furnace to
produce pig iron and a calcium aluminate slag. The slag,
typically containing 30 to 50 percent alumina and 5 to
10 percent si.ica, is treated to produce alumina. The
energy assign ~d the Pedersen process alumina will
depend on thf credits assigned the pig iron and furnace
gases produced in the electric-arc furnace; the calcium
aluminate sla~ is essentially a coproduct. The overall
energy requir.~ments for alumina produced via the
Pedersen proc(~ss or other electrothermal methods are
generally great,3r than the Bayer process energy require-
ment if the electrical energy input to the furnace is
taken at 10,400 Btu per kWh.
While the Bayer process appears to be relatively
inefficient, alternative processes fl)r producing pure
alumina do no'; offer significant energy-saving potential.
*The effect MlneSS value of tll" L.dyer and Hall processes are
comparable to tt'ose of the major unit operations in the copper
and glass industr ies and generally less than those of the major
unit operations in the steel industry. Effectiveness values are
essentially equal to efficiencies for aluminum ingot production.
Replacement of the Hall Proce$$: The Alcoa Process
In the Alcoa process. Bayer alumina is chlorinated'
to produce a gaseous mixture of aluminum chloride,
carbon dioxide, and carbon monoxide. The aluminum
chloride is separated from the gases by condensation in a
fluidized bed of solid aluminum chloride particles. The
solid aluminum chloride is introduced continuously into
an electrolytic cell containing a fused chloride electro-
lyte at about 1,290° F (700° C). Liquid aluminum metal
is produced at the cathode. Gaseous chlorine, which
recycled to the chlorination plant, is formed at the
anode.
A simplified flowsheet for the Alcoa smelting
process is shown in figure 3 next page. Alcoa has utilized
two electrolytic cell designs: a monopolar cell similar to
the conventional Hall cell, and a bipolar cell containing
four bipolar electrodes. Both cell arrangements are
started to offer reduced power consumption relative to
the standard Hall cells, due largely to the greater electri-
cal conductivity of the chloride electrolyte. Peacey and
Dayenport estimate a 20 percent reduction in electrical
energy requirement for the Alcoa smelting process rela-
tive to the best Bayer-Hall system (ref. 45). (Alcoa has
stated that the Alcoa smelting process could reduce
energy requirements by 30 percent relative to the best
Hall cell operation.) The more efficient U.S. aluminum
producing plants use about 6.5 kWh/lb of aluminum in
the Hall cell. A 20 percent reduction (a reduction of 1.3
kWh/lb) would result in a specific energy consumption
of 5.2 kWh/lb of aluminum. On this basis, the energy
required per net ton of aluminum would be lowered
from 244 million Btu per ton to about 185 million Btu
per ton. Peacey and Davenport indicate that the Alcoa
smelting process would consume about 20 percent less
carbon than the best Hall cell operation. They estimate
that a 100,OOO-metric-ton-per-year Alcoa smelting proc-
ess plant requires 5 percent less fixed capital investment,
but 10 percent greater direct operating cost than a
comparable size Bayer-Hall system.
Alternatives to the Bayer~Ha" Processes
Several processes to produce aluminum have been
proposed which would replace both the Bayer and Hall
processes. Of the methods considered, two processes
may provide an energy savings relative to the traditional
aluminum production route. These are (1) electrolysis of
aluminum sulfide and (2) the reduction of aluminum
trichloride with manganese (the Toth Process),
Electrolysis of aluminum sulfide rather than
alumina would provide several advantages: electrolysis at
° 0
a lower bath temperature (1,560 F rather than 1,740
F in a Hall cell); a greater solubility of the aluminum
sulfide in the electrolyte than is possible with alumina in
500
-------
CHLORINATION
REACTOR
REACTOR
AI203
COKE
1290-
1650 F
CI2
CO/C02
CYCLONE
CONDENSER
160 F
AICI3(S)
MOLTEN
NaCI- LiCI - AICI3
ELECTROLYSIS
1290 F
Aluminum
Figure 3. Schematic flowsheet of the Alcoa aluminum chloride process.
cryolite; and a lower consumption of anode material
because depolarization of oxygen and its subsequent
reaction with the anode would be eliminated. No esti-
mate is available for the potential energy savings with
this process. There are, however, some technical prob-
lems with the preparation of pure aluminum sulfide and
the purity of the resulting aluminum.
The Toth process, which has received much public-
ity recently, is based on the exchange reaction between
manganese metal and aluminum trichloride to give
manganese dichloride and aluminum metal. A simplified
flowsheet for the proposed Toth process is shown in
figure 4. Calcined clay and coke are chlorinated at
1,700° F with a mixture of chlorine and silicon tetra-
chloride; this technology is well known at present. The
volatile chlorides are condensed from the carbon
monoxide to yield liquid aluminum chloride. The liquid
aluminum chloride would be reduced metallic aluminum
and a manganese aluminum chloride salt mixture by
reaction with manganese. metal. Aluminum chloride is
separated from the fused salt by evaporation and subse-
quent condensation. The remaining solid manganese
chloride is oxidized to produce manganese sesquioxide,
which is then reduced to manganese metal in a conven-
tional manganese blast furnace.
A major question concerning the purity of the
bl ast-furnace-regenerated manganese and its overall
effectiveness in reducing aluminum chloride must be
resolved by difficult and expensive experimental devel-
opment work.
Peacey and Davenport estimate the Toth process
would require about 1 kWh/lb of aluminum (ref. 45).
The associated operation of a manganese blast furnace
with the high carbon requirement per ton of aluminum
501
-------
AI
AIC'3(Q) POWDER
FeCI3 ;TiC14
AICI3 ALUMINIUM CYCLONE
GENERATOR SEPARATORS/
PUR'I FI ER
570 F 2oops; EVAPORATORS
SiCI4 Mn
BLAST MnCI2
FURNACE OXIDISER +02
1110 F
CI2 COKE
Figure 4. Schematic flowsheet for the proposed Toth process.
may make t1e energy requirements associated with the
Toth Proces~ at least equal to those of the conventional
Bayer-Hall processes.
Other processes which are currently being consid- '
ered to replace the Bayer.Hail processes, such as the
Alcan proce:iS and the Monochloride process, are not
considered energy conserving relative to the best Bayer-
Hall technology available. An evaluation and comparison
of probable capital and operating costs, electric power
requirements, and carbon requirements of the Alcoa,
Alcan, Monochloride, and Toth processes with the best
available Ba\'er-Hall technology as reported by Peacey
and Davenpo.1 are given in table 8, (ref. 45).
Because Df the large energy requirement to produce
a ton of aluminum and the relatively low efficiency of
the Bayer.Hail process, it is highly desirable to develop
alternative processes; the Alcoa process appears particu-
larly promising in terms of technical feasibility and
possible ener!ly savings. Other processes which may offer
future energv savings will require extensive continuing
process development and evaluation, or are not currently
acceptable to the industry because of reduced purity of
the aluminum product or economic considerations. In
the reasonabl'l near future, it undoubtedly will be advan.
tageous to the aluminum industry to be able economi-
cally to prOCESS low-grade domestic ores in order to be-
come less dep.mdent on the supply of foreign bauxite.
REFFRENCES
1. Study of the Energy- and Fuel-Use Patterns in the
Nonferrous Metals Industries, Contract No.
14-Q1-QOOl-1658 (Federal; Energy Administration
with Battelle-Columbus), December 31,1974. Com-
pleted. (Project Officer: Mr. Bernard Haffner, U.S.
Department of Commerce; Project Manager: Dr.
James E. Flinn, Battelle-Columbus).
2. Evaluation of the Theoretical Potential for Energy
Conservation in Seven Basic Industries, Contract
No. 14-Q1-Q001.1880 (Federal Energy Administra-
tion with Battelle-Columbus), August 15. 1975.
(Project Officer: Mr. Thomas Grass, Federal Energy
Administration; Project Manager: Dr. Elton H. Hall,
Battelle-Columbus) .
3. Energy Use Patterns in Metallurgical and Non-
metallic Mineral Processing, Contract No. 50144093
(U.S. Bureau of Mines with Battelle-Columbus), in
progress. This paper is based on interim reports for
Phases 1 through 9, 1975. (Project Officer: Mr.
Ralph C. Kirby, U.S. Bureau of Mines; Project
Manager: .Mr. Raymond W. Hale, Battelle-
Columbus).
4. Transportation Systems Center, Energy Statistics: A
Supplement to the Summary of National Trans-
portation Statistics, September 1973, NTIS No.
PB.225-231.
5. Private communication.
6. H. H. Kellogg, Energy Consumption in Flotation
Benefication, unpublished, Columbia University,
New York, N.Y., July 1973.
7. J. L. Morning, and G.Greenspoon, "Flotation
Trends," Minerals Yearbook, U.S. Bureau of Mines,
1970.
502
-------
Table 8. Evaluation and comparison of alternative aluminum smelting
processes with 8ayer-Hall technology (110,000 net tons per year)
Processes
Bayer- Mon 0-
Hall Alcoa A1can ch 10ri dea Toth
Fixed capital
investment, $
per net ton A1 1 ,870 1 ,775 1 ,030 1,120 1,120
Direct operating
costs, $ per net
t on A 1 385 424 385 327 577
Electrical energy,
kWh per lb Al 6.35 5.08 11.58 6.86 0.98
Carbon requi re-
ment, 1 b C per
net ton A1 1,090 870 2,395 1 , 850 10,090
aBauxite is taken as starting material.
8. A. J. lynch, and R. l. Wiegel, "Experiences With a
Computer-Controlled Pilot-Scale Grinding Circuit,"
Mining Congress Journal, October 1972, pp. 49-56.
9. Unpublished work (Battelle's Columbus laborato-
ries).
10. L. White, "Shock-Shatter Process May Challenge
Conventional Milling Technique," Engineering and
Mining Journal, December 1972, pp. 76-77.
11. Private communication.
12. H. Kellogg, "New Copper Extraction Processes," .
Journal of Metals, August 1974, p. 21.
13. F. C. Price, "Copper Technology on the Move,"
Special Edition, Joint Issue of Chemical Engineering
and Engineering and Mining Journal, Special
Section, March 1973, pp. RR-DDC.
14. N. J. Themelis, et aI., "The Noranda Process,"
Journal of Metals, April 1972, pp. 25-32.
15. "What's Happening in Copper Metallurgy," Engi-
neering and Mining Journal, February 1973, pp.
75-79.
16. "Form Consortium to Exploit New Q.S Process,"
Journal of Metals, March 1974, p. 12.
~p. R. S. Shomakar, "Minerals Processing in 1973,"
Mining Congress Journal, February 1974, pp. 24-29.
18. "Mitsubishi's Continuous Copper Smelting Process
Goes on Stream," Engineering and Mining Journal,
August 1972, pp. 66-68. .
19. T. Suzuki and T. Nagano, "Development of New
Continuous Copper Smelting Processes," Tokyo
Meeting of AIME, May 27, 1972.
20. R. B. Worthington, "Autogenous Smelting of
Copper Sulfide Concentrate," U.S. Bureau of Mines,
Report of Investigation 7705, 1973.
21. "Copper Smelting Today: The State of the Art,"
Special Edition, Joint Issue of Chemical Engineer-
ing, and Engineering and Mining Journal, Special
Section, March 1973, pp. p-z.
22. N. Arbiter, "Anaconda's Ammonia leach Process;'
Dallas Meeting of AIME, February 1974.
23. P. R. Krusi et aI., "Cymet Process-Hydrometallurgi-
cal Conversion of Base Metal Sulfides to Base Met-
als," CIM Transactions, Vol. 76 (1973). pp. 93-99.
24. "Duval Cl.aims Development of Pollution-Free
Hydrometallurgical Copper Refining Process,"
Engineering and Mining Journal, September 1970,
p. 171.
25. Peter F. Chapman, The Energy Costs of Producing
Copper and Aluminum from Primary Sources, Open
503
-------
University Energy Research Group Report, No.
ERG 001, August 1973 (revised December 1973).
26. W. L. Faith, D. B. Keyes, and R. L. Clark, Industrial
Chemica/~, Third Edition, John Wiley & Sons, New
York, N.Y., 1965.
27. National Coal Association, Bituminous Coal Data,
1972 edit on, p. 121.
28. Private cm'respondence and communications with
aluminum industry representatives.
29. Bureau oj: Mines, U.S. Department of the Interior,
"Commodity Data Summaries, 1974," Appendix I
to Mining and Minerals Policy, Third Annual Report
of the Secretary of the Interior under the Mining
and Minerals Policy Act of 1970.
30. Merton C Flemings, Kenneth B. Higbie, and Donald
J.' McPherson, "Report of Conference on Energy
Conservation and Recycling in the Aluminum In-
dustry,' summary of conference held at
Massachw;etts Institute of Technology, June 18-20,
1974, cosponsored by the Center for Materials
Science and Engineering, M.I.T., the U.S. Bureau of
Mines, and the Aluminum Association, p. 9.
31. S. W. Martin and H. W. Nelson, Carbon Technology
Related to Aluminum Production, Great lakes
Carbon Corporation, Morton Grove, Illinois, 1955,
p.49.
32. Private communication with major porducer of
aluminulTI and a major producer of carbon products.
33. Bureau of Mines. U.S. Department of the Interior,
"Alumim from Domestic Sources; A Miniplant Pro-
ject to Evaluate Alumina Recovery Processes,"
AppendiJl A, Commodity Statement on Aluminum,
January 1974, p. A-20. ,
34. Jean-Pierre Dugois, "An Example of Power Saving
Technolo~y: The Auzat Pechiney Plant," preprint,
104th AIME Annual Meeting featuring Energy
Technolo;Jy, Americana Hotel, New York, N.Y.,
February 17, 1975.
35. F. .R. A. Smith, "Aluminum Reduction and Refin-
ing," Me~a/~ and. Materials, Vol. 8, No.3 (March',
1974), p~}. 183-186.
36. Ya. A. B~rshstein, and L. A. Kaluzhskii, "Current
State in the Production and Development of Alumi-
num Electrolysis Techniques in Developing and
Capitalist Countries," Tsventynye Metal/y, Vol. 44,
No. 11 (November 1971), pp. 89-94.
37. Robert A. lewIs, "Trends in Aluminum Cell De-
sign," Chemical Engineering Progress, Vol. 54, No.5
(May 1970), pp. 78-82.
38. United Nations, "Studies in the Economics of In-
d\,Jstry; 2. Preinvestment Data for the Aluminum In-
dustry," New York, N.Y., 1966,54 pp.
39. "Aluminum Industry Expanding Capacity," Metals
Week, June 22, 1974, pp. 6-7.
40. M. F. Elliot-Jones, "Aluminum-SIC 3324 and
3352," Chapter 31 in Energy Consumption in
Manufacturing; The Conference Board in coopera-
,tion with The National Science Foundation,
Ballinger Publishing Company, Cambridge,
Massachusetts, p. 535.
41. Morton C. Flemings, Kenneth B. Higbie, and Donald
J. McPherson, "Report on Conference of Energy
Conservation and Recycling in the Aluminum In-
dustrY," conference cosponsored by M IT Center for
Materials Science and Engineering and the U.S.
Bureau of Mines, with the cooperation of the
Aluminum Association, Massachusetts Institute of
Techology, June 18-20, 1974, pp. 7-11.
42. K. Bielfeldt, F. Kampf, and G. Winkhaus, "Heat
Consumption in the Production of Alumina," pre-
print, 104th AIME Annual Meeting featuring'
Energy Technology, Americana Hotel, New York,
N.Y., February 19, 1975, paper no. A75-78, 18 pp.
43. Allen C. Sheldon, "Energy Use and Conservation in
Aluminum Production," preprint in TMS-AIME
Symposium, Energy Use and Conservation in the
Metal Industry, 104th AIME Annual Meeting,
Americana Hotel, New York, N.Y., February 18,
1975, p. 159.
44. Ing and A. Zeerleder, "Attempts to Improve Alumi-
num Reduction Since Heroult and Hall," Journal of
'Institute of Metals, Vol. 83 (1954-1955), pp.
321-328.
45. J. Peacey, and W. Davenport, "Evaluation of Alter-
native Methods of Aluminum Production," Journal
of Metals, July 1974, pp. 24-28.
504
-------
ENERGY CONSERVING INDUSTRIAL CHANGES
AND THEIR ENVIRONMENTAL IMPACT
Herbert S. Skovronek, Ph.D. *
Abstract
Industry uses about 40 percent of the Nation's
energy supplies for space heating and process needs.
Economic need coupled with technological skill will lead
to the development of new energy-conserving processes
or practices over the coming years as one means of con-
serving this costly resource. An EPA funded study has
been awarded to Arthur D. Little, Inc., (ADL) to assess
the environmental impact of such innovations and to
determine whether control technology is or will be avail-
able to prevent increased environmental insult from such
new processes and practices. A methodology has been
developed and used for the selection of the most signifi-
cant industrial segments. Major processes likely to be
implemented by these industries over the coming 15
years as part of the energy conservation strategy have
been identified and work is underway to complete the
pollution assessments.
Industry has been estimated to consume approxi-
mately 40 percent of the Nation's energy supplies an-
nually, equ ivalent to about 23 quads (10' 5 Btu), for
space heating and cooling, feedstocks, process heat and
other miscellaneous internal uses (table 1). The fraction
used for each purpose varies grossly from industry to
industry and, undoubtedly, from plant to plant within
anyone industry. Approximately a 10 percent reduce
tion-across the board-could be achieved almost im-
mediately by such means as housekeeping (thermostat
control, elimination of heat losses, repair of steam leaks,
etc.) and improving insulation and other forms of heat
transfer control on process equipment.
Further improvements require a more thorough
examination of each industry and its practices and proc-
esses. These changes will occur more slowly and prob-
ably will be more capital intensive. Nevertheless, another
10 to 20 percent reduction would probably be achiev-
able. The precise level would be determined for each
industry by the availability of fuel forms, their cost, and
the ability to tradeoff increased cost for continued
productivity. It is in this area where more sophisticated
*Herbert S. Skovronek is with the U.S. Environmental Pro-
tection Agency's Industrial Environmental Research Laboratory,
Cincinnati, and is located at the Raritan Arsenal, Edison, N.J.
08817.
Table 1. I ndustrial energy use patterns,
1968 ( ref. 1) .
Percent
Direct combustion
29.0
Process steam
44.7
Direct electric heating
Motors, lights, electrolysis
1.3
25.9
changes such as recovery of heat currently being wasted
in the form of off-gases, hot water, or steam can occur
through the use of heat exchangers, recuperators, and
recycle or "cascade" use within a plant. In other cases,
energy value is being lost through inefficient use of raw
materials, incomplete recovery, or unnecessary disposal.
A final alternative exists for cutting energy needs
still further. This includes relatively complete changes
from current manufacturing practices or processes which
will come about either because of shortages or feared
shortages of a specific fuel form, increased price of
petroleum products needed for energy or as feedstock,
or simply by the introduction of new, more energy-
efficient technology. .
Although the latter two classes of innovation, waste
heat reuse and new practices and processes, offer signifi-
cant potential for an improved energy picture for in-
dustry, they introduce another unanswered question
requiring simultaneous attention: that of pollution con-
trol. Drastic changes in industrial manufacturing prac-
tices can be expected-at least in' some cases-to bring
about changes in air, water, or solid waste discharges
from those we are currently accustomed to dealing with.
For example,if an industry or a single plant changes
from natural gas to coal, the pollution generated will be
much greater and much more harmful, both at the plant
site and, incrementally, at the source of the energy
form-the mine. Of necessity, industries will be under-
going changes in their fuel mix over the coming years
505
-------
(table 2). MallY similar situations can be expected as side
effects of industrial efforts to maintain productivity.
Although the nature of the pollution which will result
from such ct'anges can be estimated at this time, it is
important thilt the environment,!1 impacts of all energy-
conserving approaches be considered early in order to
(1) encouragl! those processes which are also environ-
mentally mo~.t attractive, (2) discourage or delay intro-
duction of [)rocesses for which only inadequate or
economically impractical control technology exists, and
(3) in the worst case, avoid the introduction of radically
new pollutan':s, the effects of which are as yet unknown.
Those instanc:es where energy conservation efforts will
also benefit the environment (many of the waste heat
reuse scheme:; can reduce thermal discharges; solid waste
incineration "or boiler heating can reduce solid waste
volume) ShOlld be supported and accelerated. In other
cases, even where completely new processes replace cur-
rent schemes, there may be little impact on the pollution
generated. Tt.ese processes will succeed or fail on purely
economic rec!sons. Those energy-conserving changes in
industrial practices and manufacturing processes where
the pollutior from the new approach presents major
roadblocks t:> implementation should require parallel
development of adequate control technology at accept-
able cost. The final type of changes, those where the
pollution sirr:ply is not sufficiently well identified to
allow a sound decision to be made, will require addi-
tional research to establish both acceptable and safe pol-
lutant discha -p;e levels and appropriate control technol-
Ogy.
In any case, the environmental impact or the envi- I
ronmental consequences of any altered technology
should play an important role, both for the government
and for industry, in assessing the desirabil ity of new
energy-conserving technology. It would be both socially
and economically unwise to allow an energy-conserving
process to be developed and implemented only for it to
face severe and/or costly environmental restrictions or to
consume the very energy saved by the process change in
add-on pollution control.
In wrestling with this subject some months ago,
EPA decided that a companion program was needed for
the work underway by FEA and that anticipated by
ERDA in the industrial energy conservation area. The
intent of this program would be first to independently
identify major changes in industrial processes or
practices which would be likely to come about over the
next generation or so in industry's effort to reduce its
energy intensity.* After establishing that the selected
processes would conserve energy or energy forms in
short supply, as complete as possible an assessment of
the environmental impacts-whether positive or nega-
tive-would be undertaken to establish whether the proc-
ess changes could be expected to be free of more serious
environmental consequences than the current practices.
In order to accelerate those programs with the greatest
energy and environmental benefits, it was recognized
that FEA, ERDA, DOC, and EPA would have to work
-Energy required per unit of produCt.
Table 2. Selected industrial energy sources, 1971, by percent (ref. 2)
Steel Paper
COill 64.5 13.1
Fuel oils 8.4 30.4
Natural gas 20.6 23.8
pw'chased electricity 4.2 11.6
Other 2.6 33.1a
Cement
(ref. 3)
Glass
32.0
0.0
12.0
5.6
37.0
77 .2
17.1
19.0
0.0
---
iiResidues used as in-plant energy sources.
506
-------
~ogether ~md, where necessary, support those tradeoffs
~n both energy conservation and environmental protec-
tion and regulation which seemed to yield the greatest
national benefit.
In one complex sentence which became the title,
the entire intent of the project as it evolved can be
summed up: "Assess the environmental impacts of
anticipated energy conserving manufacturing process
options and the adequacy of available control technolo-
gy," Considerable soul searching was devoted to the
question of whether this project was best handled by
separate contracts for each industry or a single contract
covering all (and "all" referred to an arbitrary dozen)
industries. It was concluded that for management pur-
poses and consistency in approach a single contract was
preferable. Naturally, the contractor could use as many
subcontractors or cocontractors as he wished. The proj-
ect was initiated in June 1975 with award to Arthur D.
Little, I nc., of Cambridge, Massachusetts, as Contract
68-03-2198 (ref. 2).
The initial task of the project, development of a
methodology for selecting the 12 "best" industry seg-
ments based on energy use, potential changes expected,
the scale of the environmental impact of the industry,
and the anticipated environmental effects of the innova-
tions expected to be introduced by the industry, has
Inow been completed. After much additional soul search-
ing, discussion with the EPA project officer and con-
sideration of some of the' auxiliary goals of the project
for ERDA, FEA, and DOC, it was agreed by all parties
that while sufficient quantitative information concerning
the energy uses of most industries existed, the potential
for change and the environmental impact of those antici-
pated changes could only be evaluated in advance on a
more subjective basis, taking advantage of the combined
knowledge of EPA and ADL staff members. (Input from
industrial representatives was also to be factored into the
project as time and OMB clearance permitted.) Com-
bining these factors of energy intensity (see table 3 for
some examples). potential for change, and environ-
mental impact, a ranking of industrial segments was
derived by ADL, which, after discussion with EPA and
FEA, was accepted as a reasonable basis for the indepth
assessment studies. There were few surprises on this list
(tables 4, 5) although anyone else could certainly make a
case for some changes, depending on definition of terms
or the level of aggregation/disaggregation used within an
industrial segment. Admittedly, a degree of second
guessing also contributed to the final selections through
our suggestions of how Standard I ndustrial Classification
(code) (SIC) segments might best be aggregated, i.e., 2-,
3-, or 4-digit SIC codes. An attempt was made not only
to select the most energy-intensive segments but to
touch upon industries of different technological or
business character so that the results of the study would
establish a broad pattern that could be applied to future
analysis of other segments.
Certain other ground rules also were imposed by
Table 3. Estimated energy intensivity - selected industrial segments,
1971, (ref. 2)
Energy per
Total energy Energy/Unit value .added
Industry - (SIC) (quads) (1 06Btu/ton) (l 06Btu/$) (U$T
Steel (3312) 1.68 19.0 0.16 11.2
Petroleum refining (291l) 1.68 0.4 0.36 12.8
Primary aluminum (3334) 0.59 173.0 0.72 26.6
Alkalies/chlorine (2812) 0.24 0.52 24.5
Cement (324) 0.52 6-8.0 0.45 21.0
Glass containers (322l) 0.17 18.2 0.12' 7.0
507
-------
Table 4. Industrial energy consumption,
1968, (ref. 4)
Energy consumed
Industry segment SIC quads Percent of industrial
Primar.v metal s 33 5.3 21.2
Chemi Cd 1 s and all i ed products 28 4.9 19.6
Petroll~um refining and related
industries 29 2.8 11.2
Food and kindred products 20 1.3 5.2
Paper c:nd allied products 26 1.3 5.2
Stone, cl ay, glass, concrete 32 1.2 4.8
All others 8.0 32.0
Industry total 25;0
EPA to avoid overlap with other programs of EPA or
other agencins. For example, as a general rule, fuel
switching wa!,; excluded where it was used to raise steam
or generate electricity; where it affected the process, as
in direct firing for cement or glass manufacturing, it was
accepted as a possible approach to be considered. In
general, the industrial power plant was totally excluded
from the study so that attention could be devoted to
changes in the battery limits of the manufacturing opera.
tions. Similar Iy, although we recognized that many
opportunities exist for "over-the.fence" conservation of
energy, these were excluded as being unique and site
specific. Even internal waste heat utilization was ex-
cluded, where it contributed no apparent impact on the
process. Whem direct contact, such as in preheating of
feed materials as has been suggested in the glass industry,
could be anticipated to change the pollutant discharges,
such modifications became acceptable candidates.
A number of peripheral questions were soon raised.
Although change in feedstock was included as one
"process change" to consider. ;~ was decided not to car-
ry out full aSSE'ssments o,f the e'~;:ect of increased use of
scrap materials slich as iron, aluminum, glass, and paper
even though it was recognized that energy consumption
might be reduced by as much as 95 percent relative to
that required for virgin materials. This decision was
based on a judgment that the plant limits for both
energy and environmental factors were reasonably well
known and ~hat the impact of scrap use was more socio.
economical or institutional in character rather than tech-
nical and revolved about collection, transportation,
acceptability of different products, as well as regulatory
restrictions. It was agreed, however, that situations
where such socioeconomical, institutional, or regulatory
influences controlled the likelihood of implementation
rather than technological barriers, these would be identi-
fied and discussed in qualitative terms but left for pos-
sible future study under a different set of project guide-
lines.
As shown in table 5, at least the first five industry
segments would seem to deserve attention if only be-
cause of their gross energy use and their overall import.
ance to society and its energy needs. Even a small change
in the practices of these industries could, if widely
implemented, have a significant impact on national
energy use, but could concurrently have a marked envi-
ronmental impact if precautions were not taken. Inclu-
sion of these industries should come as no surprise. The
remainder of the industries selected presented something
more of a problem. Here it was necessary to consider
508
-------
Table 5. Industrial energy use rankings, 1971 (ref. 2)
Industry (SIC)
Energy use.
Percent of total industria1b
1015 Btu/yr
*B1ast furnaces and steel mills
(3312)
*Petro1eum refining
( 2911)
*Paper and allied products
(26)
Food and kindred products
(20)
01efins
(2869)
*Ammonia
(287)
*A1uminum. primary
(3334)
*Texti1e mill products
(22)
*Cement. hydraulic
(324 ).
*G1ass c
( )
*A1ka1i and chlorine
(2812)
Motor vehicle parts
(3714 )
E1ectrometa11urgica1 processes
( 3313 )
Gray iron foundries
(3321)
*Phosphorus and phosphoric acid
(2819)
(3274)
*Primary copper
(3331)
*Ferti1izers
( 2871)
Lime
7.3
7.3
1.68
1.68
6.8
5.5
1.59
1.27
0.98a
4.2
2.7
0.63a
2.6
2.3
0.59
0.54
2.2
0.52
.0.31
1.3
1.0
0.24
0.19
0.8
0.6
0.14
0.13
0.6
0.4
0.10a
0.4
0.3
0.10
0.08
0.3
0.08
aADL estimates.
bBased on a 23.1-quad industrial
fossil fuel to electric power.
c3211. 3221. 3229.3296.
*Included in ADL study.
usage; includes looses in conversion of
509
-------
more subjectively the energy and environmental impact
potential of specific changes within each industry and
possible ma;or shifts in the nature, economics, or even
the siting 0': an industry. Admittedly, as noted earlier,
other choices might have been equally appropriate.
The contractor has now gone on. to identify the
specific processes or practices which seemed appropriate
within each industrial segment. Here again, considerable
soul searching and bending of the original ground rules
took place in an effort to generate a product useful to
other agencies as well as to EPA and one which would
touch on different conceptual approaches to energy con-
servation and/or environmental protection.
Tables £i-9 present some examples of the technology
innovations dentified by ADL as being considered with-
in the diffemnt industrial segments. Final selections will
be made by ADL after discussion with EPA and indus-
trial representatives. In this way we hope to insure that
the selected processes are reasonable and a realistic pre-
diction of ",hat will happen-assuming no other major
perturbation; are placed on the system.
ADL work is now proceeding in three directions.
Staff membms are collecting the available energy and
material balcnce information for the selected processes.
They are generating as much information as possible
concerning 1 he potential pollutants and their levels-
including new or different pollutants. Finally, they are
identifying peripheral factors that should be considered~
such as the importance of transportation, shifts in in-
dustry siting, trends from localized pollution (in-plant
power generation) to centralized pollution (central
power plants), etc. Another problem recognized is that
as a process becomes more "blue-sky," adequate assess-
ment of energy, economics, environmental impact, and
the probability of succeeding become less definitive.
Since full studies for such processes would be a rather
academic exercise and it is unlikely that such completely
new processes would achieve widespread use in the next
15 years or so, ADL has chosen to concentrate on proc-
esses which are more likely to playa commercial role
over the next 15 years.
The program has been formulated so that the results
for different industrial segments will be completed at
intervals over the next 3 or 4 months (December-March).
Shortly thereafter (May 1976), ADL will issue a sum-
mary report in which certain general, broadly applicable
new technology will be explored and general conclusions
will be presented. Of particular value to EPA, areas will
be highlighted where additional R&D is needed to more
soundly identify and quantify the pollution of new proc-
esses or to accelerate the development of generic pollu-
tion control technology to support the most beneficial,
or at least the minimally injurious processes from both
the energy and the environmental points of view. As I
Table 6. Anticipated change: steel industry - blast furnaces,
(SIC 3312) (ref. 5)
Annual fuel consumption 1.7 quads (1968)
Options
Impact
Environment
Energy
Est. 1014Btu recoverable
as high-pressure steam
Reduced coke use (2-3
percent)
Est. 1013 Btu recoverable
Coke dry quenching
BOF Off-gas collection
Reduced quench water
discharge
Reduced coke dust
Reduced thermal dis-
charge
Reduced particulates
Reduced fossil fuel
burning
510
-------
Table 7. Anticipated change: container glass industry
(SIC 3312)
Annual fuel consumption 0.155 quads (1971),
75 percent as natural gas
=.::.=--.--..
Options
Energy
Impact
Environment
Coal direct fired
Conserve natural gas
Increased SO , NO
. x x
particulates
Reduced NOx (lower
temperature), SOx'
particulates
(scrubbing)
Batch preheati n.g
20-35 percent total sav-
ings potential; equiv-
alent to 0.03-0.05
quads
Table 8. Anticipated change: textile mill products
(SIC 22)
Annual fuel consumption 0.5 quads (1971)
30 percent electric, 70 percent fossil
Option
Energy
Environment
Al from A1C13 (Alcoa
smelting)
~30 percent reduction in
electrical power needs
Undefined
Closed system should
reduce pollution
511
-------
Table 9. Anticipated change: aluminum (and alumina)
(SIC 3334) (refs. 4,6)
Annual fuel consumption 0.59 quads (1971)
Est. 90 percent electric, 10 percent fossil
Option
Energy
Reduced wastewater
Solvent finishing and
and dyeing
Impact
Environment
Up to 50 percent reduction
in steam requirements
Reduced natural gas needs
for drying
Reduced power plant.
discharges
Increased solvent
emissions
noted earlier, it can be expected that certain process
options will be identified where energy conservation and
pollution abatement efforts are joined to give an overall
advantageous package. Others will occur where the
known hazards of the new processes, or the cost or
energy requil'ements of appropriate control technology
will be so !treat as to overwhelm the energy savings
achieved in 1 he process. Still others will have relatively
unchanged ellvironmental impact. And finally, there will
be certain processes where insufficient information
exists to definitively place the process in one of the
three foregoing classes but where energy saving is an
incentive to develop cost- and energy-effective pollution
control techn ology.
Yet ancther conceptual "grouping" exists, one
which could fit into any of the classes just noted. This
includes processes whi<;h offer such large energy savings
that tradeoff:; with the level of environmental protection
must be considered; thus, modifications of applicable
environmental regulation may have to be considered in
order NOT t) hinder the development and widespread
implementation of such processes..ADL will attempt to
classify theSt process options, identify the additional
knowledge or technology needed, and point out where,
in their opin on, such regulatory tradeoffs are possible
and the bene"its they offer. It will, however, remain the
responsibility of the EPA, work.mg together with other
interested Fejeral agencies (FEA, ERDA, DOC, etc,) to
make the fin 11 determina'~;ons of how best to meet our
Nation's enerqy demands safely.
In closing I would like first to apologize for not
having more quantitative data at this time. Secondly, as
the EPA project officer for this project, my thanks to
Mr, Haley and Dr, Kusik of ADL and the staff now
devoted to this project, as well as to the scientists of
EPA and other Federal agencies who are so willingly I
participating in the development and review of this
program.
REfERENCES
1. Elias P. Gyftopoulos, Thomas J. Lazaridis, and
ThomasF. Widmer, "Potential Fuel Effectiveness in
Industry," a report to the Energy Policy Project of
the Ford Foundation, Ballinger Publishing Co.,
Cambridge, Mass., 1974.
2. Arthur D. Little, "I ndustry Priority Report," draft,
USEPA Contract 68-03-2198.
3. Robert D. MacLean, "Energy Use and Conservation
in the U.S. Portland Cement Industry:' testimony
presented at the U.S. Senate Committee on Com-
merce Public Hearings on Energy Waste in Industrial
and Commercial Activities.
4. I nterTechnology Corp., "Energy Conservation,"
USEPA Contract No. 68-01-0578.
5. Battelle Columbus Laboratories, "Potential for
Energy Conservation in the Steel Industry," FEA
Contract No. CO-04.57874.00, May 30. 1975.
6. Battelle Columbus Laboratories, "Study of the
Energy and Fuel Use Patterns in the Non-ferrous
Metals Industries," December 31, 1974. .
512
-------
TECHNICAL REPORT DATA
(Please rcad illSlrlletiol/s 01/ the rCl'efse before completing)
1. REPORT NO. /2. 3. RECIPIENT'S ACCESSIOr+NO.
IIPA-600/2 -76-212
LE AND SUBTITLE 5. REPORT DATE
~posium on Environment and Energy Conservation August 1976
(November 1975, Denver, Colorado)' . 6. PERFORMING ORGANIZATION COOE
7. AUTHOR(S) 8. PERFORMING ORGANIZATION REPORT NO.
Franklin A. Ayer (Compiler) ERDA 47
9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT NO.
Research Triangle Institute lAB013' ROAP 21BBZ-012
P. O. Box 12194 11. CONTRACT/GRANT NO.
Research Triangle Park, NC 27709 68-02-1325, Task 29
12. SPONSOR.ING AGENCY NAME AND ADDRESS * 13. TYPE OF REPORT AID PERljD COVERED
EPA, Office of Research and Development Proceedings; 3 75-6 76
Industrial Environmental Research Laboratory 14. SPONSORING AGENCY CODE
Research Triangle Park, NC 27711 EPA-ORD
15. SUPPLEMENTARY NOTES (*) ERDA cosponsored this symposium. ERDA project officer was
K. D. DeVine; EPA project officer was W. B. Steen.
16. ABSTRACT The principal objective of this symposium was to identify the environmental
benefits and threats of alternative energy conservation systems and to compare the
environmental impacts of energy conservation strategies. It was also an objective of
the symposium to anticipate and publicize environmental impacts so as to: facilitate
IIIII'per consideration of these impacts in the assignment of funding priorities for
~rgy conservation programs under the national energy research and development
program; and begin work in a timely manner to solve future environmental problems.
17. KEY WORDS AND DOCUMENT ANALYSIS
--
a. DESCRIPTORS b.IDENTIFIERS10PEN ENDED TERMS C. COSATI Field/Group
~---
Pollution, Conservation, Ecology, Energy Pollution Control 13B 06F
Research Management, Environments Energy Conservation 05A
Electric Power Plants, Gas Turbb1es Environmental Impact lOB 13F
Electric Power Transmission Staged Combustion 09C
Magnetohydrodynamics, Thermodynamics Geothermal Energy 201 20M
Resources, Transportation, Heat RecoveryWastes-as-Fuel 15E
~STRIBUTION STATEMENT 19. SECURITY CLASS (This Report) 21. NO. OF PAGES
Unclass Hied 524
~imited 20. SECURITY CLASS (Tllispage) 22. PRICE
Unclassified
EPA Form 2220-1 (9-73)
513
-------
514
-------
|