COMMITTEE ON THE CHALLENGES OF MODERN SOCIETY
III
Hi
Illl
1
LA CONFERENCE
COMITE SUR LES DEFIS DE LA SOCIETE MODERNS
-------
TECHNICAL REPORT
OF THE
CONFERENCE ON LOW POLLUTION
POWER SYSTEMS DEVELOPMENT
NATO Committee on the Challenges of
Modern Society
Eindhoven, The Netherlands
February 23, 24, 25, 1971
Environmental Protection Agency
Rockville, Maryland
-------
TECHNICAL REPORT
OF THE
CONFERENCE ON LOW POLLUTION
POWER SYSTEMS DEVELOPMENT
TABLE OF CONTENTS
CHAPTER PAGE
I. Introduction 1-1
II. FEDERAL MOTOR VEHICLE EMISSION GOALS FOR II-l
CO, HC, AND NOx BASED ON DESIRED AIR
QUALITY LEVELS
Ronald Engel, Environmental Protection
Agency, USA
III. THE ADVANCE AUTOMOTIVE POWER SYSTEMS PROGRAM III-l
John H. Brogan, Environmental
Protection Agency, USA
IV. THE POTENTIAL OF THE GAS TURBINE VEHICLE IV-1
IN ALLEVIATING AIR POLLUTION
Edward S. Wright, United Aircraft, USA
V. THE STIRLING-CYCLE ENGINE V-l
R.A.J.O. van Witterveen, N. V. Philips,
The Netherlands
VI. RANKINE-CYCLE POWER SYSTEM WITH ORGANIC- VI-1
BASED FLUID AND RECIPROCATING EXPANDER FOR
LOW EMISSION AUTOMOTIVE PROPULSION
Dean T. Morgan, Thermo Electron
Corporation, USA
VII. STIRLING ENGINE ACTIVITIES AT UNITED VII-1
STIRLING (SWEDEN)
Lars G. Ortegren, United Stirling,
Sweden
-------
CHAPTER
PAGE
VIII.
IX.
X.
XI.
XII.
XIII.
XIV.
XV.
XVI.
NITROGEN OXIDE FORMATION IN THE COMBUSTION VIII-1
CHAMBER OF THE INTERNAL COMBUSTION ENGINE
AND ITS SUPPRESSION BY MEASURES FROM
COMBUSTION TECHNOLOGY
K. H. Newmann, Volkswagen, Federal
Republic of Germany
A EUROPEAN CONTRIBUTION TO LOWER VEHICLE IX-1
EXHAUST EMISSIONS
Diarmuid Downs, Ricardo
Consulting Engineers, United Kingdom
LOW EMISSIONS FROM CONTROLLED COMBUSTION FOR X-l
AUTOMOTIVE RANKINE CYCLE ENGINES
W. A. Compton, Solar Division of
International Harvester Company, USA
HYBRID HEAT ENGINE/ELECTRIC SYSTEMS STUDY XI-1
Joseph Meltzer, The Aerospace
Corporation, USA
ADVANCED TECHNIQUES IN ELECTRICAL VEHICLES XII-1
Bohers/Ducrot, Citroen Automobile
Company, France
RESEARCH AND DEVELOPMENT ON A LITHIUM- XIII-1
SULFUR BATTERY
Elton J. Cairns, Argonne National
Laboratories, USA
RESEARCH AND DEVELOPMENT PLAN OF ELECTRIC XIV-1
CAR
Shizume, Japanese Automotive
Manufacturers' Association, Japan
STUDIES BY FIAT ON THE ELECTRICALLY-DRIVEN XV-1
AUTOMOBILE
G. Brusaglino, Fiat, Italy
ELECTRICAL VEHICLES WITH FUEL CELLS: XVI-1
WHY AND HOW?
J. Beslier, Society of Peugeot
Automobiles, France
-------
ACKNOWLEDGMENTS
The United States, as pilot country on the Air Pollution Study
of the NATO Committee on the Challenges of Modern Society acknowledges
the assistance and hospitality of the Government of The Netherlands
and The Philips Company who co-hosted the Conference.
-------
1-1
Introduction
The NATO/CCMS Conference on Low Pollution Power Systems Development
was held in Eindhoven, The Netherlands, on February 23-25, 1971. The
Government of The Netherlands and Philips Company co-hosted the
Conference.
The total attendance for the meeting was just under one hundred government
and industrial representatives from twelve countries, as well as non-NATO
international organizations. The following countries were represented:
The Netherlands, Belgium, Canada, Denmark, France, Germany, Great Britain,
Italy, Turkey, Sweden, Japan, and the United States. The Organization
for Economic Cooperation and Development (OECD) and the European Economic
Community had observers in attendance.
The Conference was divided into two sessions with the first day reserved
for government representative presentations and discussions regarding
country programs and policies relative to the automotive pollution
problem. The second and third days were devoted to technical presentations
on research and development projects by both government and industry
attendees.
The "Summary of Proceedings of the Conference on Low Pollution Power Systems
Development"* was published by the U. S. Environmental Protection Agency
(EPA) in March 1971. The Summary of Proceedings document gives an overview
of the Conference and includes brief summaries of each presentation made
at the Conference by government and industry representatives.
*This document is available from the Environmental Protection Agency.
-------
1-2
This document includes the texts and figures of the technical presen-
tations that dealt with research efforts in the low pollution power
systems field. Additionally, the one technical paper dealing with the
rationale for the U. S. standards was also included because of the
extensive amount of information on the relationship between air quality
and automobile emissions contained in the paper. The U. S. companies
who participated in the conference are contractors to the Environmental
Protection Agency for low pollution power systems development. Other
discussion papers, which were general in nature relating to governmental
concern with the automobile pollution problems, were not included;
however, copies of these papers may be obtained from the participants.
The summaries are included in the Summary of Proceedings document.
-------
II-l
Chapter II
FEDERAL MOTOR VEHICLE EMISSION GOALS FOR CO, HC, AND NO
A
BASED ON DESIRED AIR QUALITY LEVELS*
by
D. S. Earth, J. C. Romanovsky,
E. A. Schuck, and N. P. Cernansky
Environmental Protection Agency,
USA
Presented at Eindhoven Conference by
Dr. Ronald Engel
Assistant Director, Bureau of Criteria and Standards, APCO
Environmental Protection Agency,
Research Triangle Park, N.C., USA
*"Journal of the Air Pollution Control Association," Vol. 20,
No. 8, August, 1970, pp. 519-523. Reprinted by permission of
the Air Pollution Control Association, Pittsburgh, Pennsylvania
USA.
-------
II-2
Federal Regulations
Background
The Federal motor vehicle air pollution control program has its origin
in the Clean Air Act, as amended in 1965.1'2 This Act borrowed heavily from
the motor vehicle air pollution control program already in effect in
California ' and reflected testimony from the automotive industry that
similar controls were feasible for application nationwide. The automotive
industry favored national standards to forestall a proliferation of regulatory
legislation at the state level. National standards, applicable to the 1968
5
model year, were promulgated on March 31, 1966. More stringent and more
equitable national standards, applicable to the 1970 model year, were
promulgated on June 4, 1968. In addition, on February 10, 1970, Secrel
Finch published an advance notice of proposed rulemaking indicating the
Department's
model years.'
Department's intent to adopt more stringent standards for the 1973 and 1975
7
Air Quality Criteria
Under the 1967 amendment to the Clean Air Act,8 the Department of
Health, Education, and Welfare is required to publish air quality criteria
Documents. With criteria for photochemical oxidants, carbon monoxide,
hydrocarbons, and nitrogen oxides either published or underway it now
becomes possible to consider future motor vehicle emission standards in
-------
terms of desired air quality consistent with information in the pertinent
air quality criteria documents. The resumes of these documents describe
what in the Secretary's judgment are the concentrations and exposure time
which cause or contribute to, or are likely to cause or contribute to,
air pollution which endangers human health or welfare.
Desired Air Quality Goals
Based on the criteria resumes and observed aerometric relationships,
Table I has been constructed to show the maximum values which would seem
to be consistent with health-related criteria. Minimum safety margin
considerations have been incorporated into these levels.
Calculating Motor Vehicle Emission Goals
Using the information in Table I, simple roll-back techniques, like
those used by California, could be used to calculate needed emission
reductions. These techniques, however, involve a number of assumptions
which may not be entirely valid. For example, inherent in the techniques
employed is the assumption that the increase in atmospheric concentrations
of primary pollutants will be directly proportional to the growth in
emissions of the contaminant. Also, the roll-back techniques employed
by California assumed a linear correlation to exist between the severity
of the manifestation, specifically oxidant index, and reactive hydro-
carbon emissions to the atmosphere. The complexities of the photochemical
alterations involving reactive hydrocarbons in the atmosphere are too
great to permit such an assumption, a^ priori, because at a minimum the
role of the oxides of nitrogen must be taken into consideration.
Clearly what is needed is to fully relate all these effects to
emissions and to predict future events in growth and required control
-------
II-4
9-11
Table I. Desired Air Quality Goals for Health Protection
Contaminant Concentration, g/m^ Average, hr
Carbon Monoxide <_ 10,000 (9 ppm) 8
Photochemical Oxidants i 125 (0.06 ppm) 1
Nitrogen Dioxide3 4190 (0.10 ppm) 1
A
Preliminary estimated value pending review of technical report on direct
health effects of nitrogen dioxide and publication of Air Quality
Criteria Document.
-------
II-5
using a comprehensive simulation model complete with modules reflecting
all input variables, meteorological variables including air transport
parameters, and most importantly-where oxidants, eye irritants, and
aerosols are concerned-the chemical kinetics describing the multiplicity
of the reactions which occur. Factual data on existing air quality are
also required in order to validate and adjust the model as necessary.
Several studies are currently investigating the generation of these
simulation modelsJ2,13 Unfortunately, none of these models is yet
ready for general application.
Relating Air Quality Goals for Nitrogen Oxides and
Nonmethane Hydrocarbons to Oxidant Goals
In the absence of a validated model which incorporates atmospheric
chemical mechanisms, a more restricted diffusion model may be employed
in which empirical relationships between primary atmospheric pollutants
and secondary reaction products are known. Such relationships have
been established between photochemical oxidants and nonmethane hydro-
carbons .11 Analogous relationships have also been developed for the
oxides of nitrogen separately as well as in combination with nonmethane
hydrocarbons. One important premise, based on experiences reflecting
the cause and effects associations observed to relate Los Angeles with
Pasadena, is that hydrocarbon emissions between the hours of 6 and 9 a.m.
result in peak oxidant concentrations 2 to 4 hours later.H This is
particularly true on the days of greatest interest-days which because
of favorable meteorology favor maximum production of oxidant.
Figure 1 includes an analysis of 3 years of data which shows that
on those several days a year when meteorological conditions were most
conducive to the formation of photochemical oxidants, nonmethane
-------
II-6
0.30
0.25
E
Q.
Q.
§ 0.20
O
LLJ
LU
O
X
0.15
0.10
0.05
LOS ANGELES
LOS ANGELES^X-TDENVER
WASHINGTON*^
" * LOS ANGELES
\ A PHILADELPHIA
LOS ANGELES
PHILADELPHIA-
PHILADELPHIA
WASHINGTON /
[_ WASHINGTOIU/pH,LADELpH|A
WASHINGTON
7
WASHINGTON, ^ ^ A AA A
A
t
AA
** ^ /^ i~A '
A
A
• A A Am
A
^ \ .
A *•
« A A A A A A AA
0 0.5 1.0 1.5 2.0 2.5
6-9 A.M. AVERAGE NONMETHANE HC CONCENT RATION, ppm C
Figure 1. Maximum daily oxidant as a function of early morning non-
methane hydrocarbons; 1966-1968 CAMP stations; May through
October 1967 for Los Angeles.11
-------
II-7
hydrocarbon concentrations of 200 jjg/m3 (0.3 ppm C) for the 3-hour
period from 6 to 9 a.m. might produce an average 1-hour photochemical
oxidant peak concentration of up to 200 pg/m3 (0.10 ppm) 2 to 4 hours
later.I1 The hydrocarbon measurements were confined to 200 ;jg/m3, or
above, because of instrumentation limitations. However, if the functional
relationships between the hydrocarbon and photochemical oxidant measure-
ments were extended to include the levels presented in Table HI as the
highest value consistent with health-related criteria, the corresponding
hydrocarbon concentration would be approximately 125 ng/m3 (0.10 ppm C).
Figure 2 presents similar data relating 6 to 9 a.m. average nitrogen
oxides concentrations to maximum daily 1-hour average oxidant concentra-
tions appearing 2 to 4 hours later. The envelope enclosing the data
presumably reflects those several days a year when meteorological
conditions were most conducive to the formation of photochemical
oxidants. On such occasions, if the functional relationships between
the oxides of nitrogen and photochemical oxidant measurements were
extended to include the level of oxidants presented in Table 1, the
corresponding nitrogen oxides concentrations would be approximately
49 jjg/m3 as N02 (0.026 ppm).
Figure 3 is a three-dimensional presentation of the same data
shown in Figures 1 and 2J1 The isopleths traced through the different
oxidant levels reflect the fact that the ratio of nonmethane hydro-
carbon to nitrogen oxides is important, along with the absolute level
of the separate independent variables, in determining the maximum
level of oxidant produced on days of favorable meteorology. Presenta-
tion of the data in this form is useful for qualitatively considering
the trade-offs between unilateral and joint control of hydrocarbons
-------
Table III. Relationship of Air Quality Goals
to Motor Vehicle Emission Goals
Emission
Carbon Monoxide
Nitrogen Oxides3
Type of
Effect
Direct
Direct
Health-Related
Air Quality Goals
10 mg/m3 (9 ppm)
190 ug/m3 (0.1 ppm)
Health-Related
Averaging Time,
hr
8
1
Emission
Goals from all
Vehicles to Achieve
Desired Air Quality,
g/mi
6.16
0.38
Control
Requirement for
All Vehicles, %
92.5
93.6
(N02) (NOe) (N02)
Hydrocarbons
Indirect 125 yg/m of oxidant
(9.06 ppm)
1
0.15
99.0
aThe NOX emission goal is identical to the N02 emission goal since in Los Angeles, the city at maximum risk,
all NO is converted to N02-
03
-------
II-
0.30
D.
Q.
).20
LU
cc
13
O
X
0.10
o
5
3
1
W - WASHINGTON
P-PHILADELPHIA
D- DENVER
• - 0.3 TO 0.9 ppm C
O-l.OTOS.OppmC
• W
-«-<•»• *I* *O »»OO O • OO
• P «O o r> O
Q •••• •Q«V»* ••Q* • • • (J «U»
• ^ O O
•o ••• ••••• o»»» 0*0 oo o
00
o
o
I
I
I
I
0 0.10 0.20 0.30
6-9 A.M. AVERAGE TOTAL I\IOX CONCENTRATION, ppm
Figure 2. Maximum daily oxidant as a function of early morning nitrogen
oxides; 1966-1968 CAMP stations.
-------
11-10
1.2
I
S1-0
z 0.8
o
UJ
< 0.6
o
z
o
2
UJ
OXIDANT, ppm
• 0.07
A 0.10
• 0.12
0.14
0.16
0.2
o>
PREDICTED LIMITS FOR 0.10
ppm OXIDANT
' PREDICTED LIMITS FOR 0.05 ppm OXIDANT
I I
I
0 0.10 0.20 0.30
6-9 A.M. AVERAGE IMOX CONCENTRATION, ppm
Figure 3. Upper limit on maximum daily 1-hour average oxidant as a
function of early morning nonmethane hydrocarbons and nitrogen
oxides; June, July, August; Philadelphia, Washington, Denver;
1966 through 1968.11
-------
11-11
and oxides of nitrogen. It must be emphasized that Figure 3 at this
time is based on relatively few points. Thus the curves can be better
defined as further data become available. In view of this, the drawing
of quantitative conclusions from this graph is not wholly justified.
Qualitative examination of Figure 3 shows that, depending on the
specific existing concentrations of oxidants, nitrogen oxides, and
nonmethane hydrocarbons, the optimum control strategy for the most rapid
reduction of oxidants may be unilateral control of hydrocarbons,
unilateral control of nitrogen oxides, or joint control of both
nitrogen oxides and hydrocarbons. It must be pointed out, however,
that all possible oxidant control strategy options are not available
to us. Recall that the postulated health-related desired air quality
for N(>2 is 190 yg/m3 (0.10 ppm) for a 1-hour average (Table I). Thus
no oxidant control strategy is acceptable which will not simultaneously
lower the N0£ values to the desired air quality. Furthermore, our
estimate of available nitrogen oxides control technology is that there
is at this time no proven or postulated device capable of lowering
nitrogen oxides emissions from internal combustion engine mobile sources
sufficiently to achieve air quality concentrations of =0.026 ppm, which
we have shown would be required to control oxidant air quality concen-
trations to 125^ig/m3 (0.06 ppm) by unilateral nitrogen oxides control.
We conclude that the most rational control strategy at this time
is to control nitrogen oxides to achieve the desired health-related air
quality for N0£ and to rely on control of nonmethane hydrocarbons to
achieve the desired oxidant air quality goal. Subsequent calculations
will be based on this approach.
-------
11-12
Again, caution must be exercised with respect to the permanency
of the recommended control approach. In particular it must be kept
in mind that all factors are not known relative to atmospheric inter-
actions. Because of this lack of information it is most difficult to
develop a comprehensive systems approach to this problem. A good
example of the need for caution was discovered when the changes in the
Los Angeles atmosphere between 1962 and 1967 were investigated.14
During this 5-year period, in the May through October months, the
6 to 9 a.m. hydrocarbons decreased by 4 percent and the nitrogen
oxides increased by 25 percent. 'Additionally the maximum daily
1-hour average oxidant decreased by 11 percent. Focusing attention
on only the decrease in maximum oxidant level, however, does not
describe all the changes. Thus this maximum oxidant decrease was
associated with an increase in the number of days on which the maximum
oxidant exceeded 0.1 ppm. The result is that a trade-off has been made,
i.e., an 11 percent decrease in the maximum for a 10 percent increase
in dosage. This latter increase in dosage, and to a large extent the
decrease in maximum, may be related to the altered ratio of HC to NOX.
Thus control approaches which suggest allowing increases in ambient
levels of nitrogen oxides are definitely subject to question. As a
result of this and other examples it is apparent that we must be always
alert to the atmospheric stiuation on a year-to-year basis. Furthermore,
we must be prepared to modify any adopted control approach as new infor-
mation becomes available.
In addition to these examples of lack of information, there are
other contributing factors which may demand changing any projected
emission values at some time in the future. During the next five
-------
11-13
years, for example, we will re-examine the Photochemical Oxidant and
Hydrocarbon Criteria documents. In the intervening years it is quite
possible that more precise measuring methods may be developed. Thus
the oxidant criteria may become ozone criteria and the present
non-methane hydrocarbon measurement may be stated in terms of reactive
hydrocarbons. Such changes will, of course, affect the emission goals
as calculated in this presentation.
Summary of Calculations Relative to Motor Vehicle Emission
Goals Designed to Produce a Health-Related
Acceptable Air Quality
Calculation of the motor vehicle emission goals for CO, HC, and
NOX requires application of two equations. The first of these equations
permits a calculation of the fractional reduction in ambient concentra-
tions necessary to achieve a specific health-related air quality.^
The equation is stated in terms of present air quality, desired air
quality, background concentrations, and the projected growth in
emissions. Mathematically stated, it has the form:
R = (GF) (PAQ) - (DAQ) m
TGF) (PAQ) - (B) u;
where R is the calculated fractional reduction required; GF is the
emission growth factor; PAQ is the present maximum air quality; DAQ
is the desired air quality; and B is the background concentration.
The second equation employs the result from Equation 1 to restate
the air quality reduction in terms of mobile emission rates. Mathe-
matically expressed, this second equation has the form:
(PER) (1-R) = DER (2)
where PER is the percent emission rate giving rise to the present air
quality (PAQ); the expression 1-R is the statement of how much of the
-------
emission can be allowed without exceeding the desired air quality.
The multiplication, i.e., (PER) (1-R) thus yields the desired emission
rate (DER) consistent with the desired air quality (DAQ).
In order that the calculated emission goals be protective of
health, certain principles must be adopted relative to the values
inserted in Equations 1 and 2. Thus, present air quality in Equation
1 must be representative of the worst situation observed in the entire
United States. For HC and NOX values to be associated with health
effects of oxidants and eye irritants, the worst situation presently
occurs in Los Angeles.16 (The 6 to 9 a.m. maximum average concentrations
6bserved during the-high oxidant potential-season^were 5.3 ppm for non-
methane HC and 0.62 ppm for NO^.) Los Angeles also had the
highest yearly 1-hour average N02 concentrations (0.69 ppm); thus, when
the health effects of N02 are being considered^air quality levels in
Los Angeles are used. With respect to CO health effects, however,
the city of Chicago, with a maximum 8-hour average of 44 ppm, represents
the .worst known case in the United States.16 In terms of growth factors,
we must again assume the maximum predicted growth in order to provide
adequate health protection. For a time period from 1967 to ten years
beyond the year of application of these motor vehicle emission goals, the
maximum predicted growth factor for mobile emissions must be used. Again,
to afford the maximum health protection, the background concentrations
used in Equation 1 must be the maximum supported by scientific inquiry.
For these purposes a value of 1 ppm was used for CO,10 0.1 ppm for HC,11
and 0.004 ppm for NOX. The lowest supportable values for the desired
air quality, as derived from the appropriate criteria documents were
used in Equation 1 and have been presented in Table I.
-------
11-15
For the present emission rate (PER), we have related derived
emission rates to a 1967 rate of 82.6 grams per miles (g/mi) for CO,
14.84 g/mi for HC, and 5.93 g/mi for NOX.17 The HC and NOX rates are
representative of the 1967 vehicle mix in Los Angeles, while the CO
rate is representative of the 1967 vehicle mix in Chicago.
The one factor which this treatment does not speak to is the
possible effect of the failure of vehicles to meet the stated emission
goals when they become standards. Thus, surveillance data from vehicles
presently in consumer use show current Federal standards are exceeded on
the average by 13 percent in terms of CO emissions and 25 percent in terms
of HC emissions.!8 This, however, is a difficult factor to take into
account because it may change in the future. If we ignore such past
failures, however, we are in essence announcing that before these emission
goals are established as standards we will have devised methods that will
force automobile manufacturers to provide control devices which will enable
vehicles in consumer use to meet required emission standards. The pro-
jected emission limitations derived here will assume that devices and
control systems in light duty vehicles in use continue to exceed the
standards by the stated percentages.
To proceed to the calculation of specific mobile emission goals
needed to achieve desired air quality goals, it now becomes necessary
to designate a target year for implementation of the required mobile
emission restrictions. Selection of a target year should be based
upon the evaluation of many important factors such as: the need to
meet the desired air quality goals at the earliest possible time;
the state-of-the-art of control technology; and future advances in
our understanding of atmospheric photochemical processes. It is
-------
11-16
beyond the scope of the paper and the competencies of the authors
to evaluate all of the above factors and their interactions in detail.
Thus, to avoid becoming embroiled in high-level policy questions which
still remain to be resolved, we have arbitrarily selected 1980 as the
target year on which to base our calculations. In the event the decision
is made to select an earlier year for implementation it will be an easy
matter to adjust the calculations accordingly. Of course, the earlier
the calculated mobile emission restrictions can be put into effect, the
earlier we will be able to achieve desired air quality goals and the less
stringent will be the necessary mobile emission limits. In accordance
with the above decision we used a 1967-1990 mobile emission growth
factor.17'19
In reviewing the required information items it will be noted that
the percentage contribution of current mobile sources to existing air
quality, the projected mobile growth rate, and the background concentra-
tions of each pollutant are the least well-defined. Thus, background
concentrations of HC are not delineated with any degree of accuracy.
The same can be said about growth rates of mobile emissions. Further-
more, the percentage contribution of mobile emissions to total emissions
must at this time be based on logic rather than known fact. In the case
of Los Angeles there seems to be general agreement that in the downtown
area the smog-season emissions of CO, HC, and NOX are mainly from mobile
sources. Consequently, the calculated emission goals were based on
the latter assumption. A similar assumption was made relative to the
Chicago data, i.e., that the bulk of the ambient CO concentrations
steam from the mobile source. If this assumption is incorrect, it
follows that those stationary sources which contribute significantly
to the downtown area concentrations of CO, HC, and NOX must be
-------
11-17
controlled to the same degree as motor vehicles. To the extent that
these sources are uncontrollable the restriction of the motor vehicle
would have to be increased.
Table ilcontains a summary of the numerical values used in cal-
culating 1980 mobile emission goals. Figures 4, 5, and 6 express the
results of the calculations for desired mobile emission rates for CO,
HC, and NOX, respectively, as a function of the desired air quality for
each pollutant without taking into account the potential deterioration
of the control measures for CO and HC. In these figures a range'of
values above and below the Table 1 values of desired air quality are
shown since the exact health-related values are still under study and
may be changed in the future. Table m contains the 1980 motor vehicle
emission goals for CO, NOX, and HC required to achieve the desired
air quality after taking presently observed CO and HC control measure
deterioration into account.
-------
Table II. Numerical Values of Parameters used in Calculating
Mobile Emission Goals9"11'16'17'19
Parameter
1967 maximum air quality values
reflated to direct health effects
1967 maximum air quality of
oxidant precursors, (6 to 9 a.m.
average)
1967 average emission rates for
all motor vehicles
Maximum background concentration
Maximum growth factor of mobile
emissions
Air Quality Goals
CO
51 mg/m3 for 8
hr, Chicago
™
82.6 g/mi
1 mg/m3
(1 ppm)
2.18
10 mg/m3 for 8
hr (9 ppm)
HC
-
3.5 mg/m3 as
CH4 (5.3 ppm C),
Los Angeles
14.84 g/mi
0.1 mg/m3 as CH4
(0.1 ppm)
2.18
125 yg/m3 of
oxidant (0.06
ppm), 1-hour
average
NOX
-
1170 yg/m3 as
N02 (0.62 ppm),
Los Angeles
5.93 g/mi
8 yg/m3 as N02
(0.004 ppm)
2.18
125 yg/m3 of
oxidant (0.06
ppm), 1-hour
average
N02
1300 yg/m3 for 1 hr,
(0.69 ppm), Los Angeles
—
-
8 yg/m3 as N02
(0.004 ppm)
2.18 I
a
190 yg/m3 for 1 hr
(0.1 ppm)
-------
11-19
11
10
10
CO
E
o>
E
d 9
o
LU
LU
E
a.
a.
8 8
LLI
O
X
00
O
X
CO
I
I
(4 5678!
TOTAL MOBILE CO EMISSION RATE, g/mi
•
Figure 4. Relationship of 8-hour average CO values to mobile emission
rates.
-------
11-20
0.070
0.05 0.10 0.15 0.20 0.25
TOTAL MOBILE HC EMISSION RATE, g/mi
0.045
0.30
Figure 5. Relationship of HC mobile emission rate to 1-hour average
oxidant.
-------
11-21
0.2 0.4 0.6 0.8 1.0
TOTAL MOBILE IMOX EMISSION RATE, g/mi
Figure 6. Relationship of 1-hour ambient IM02 levels to mobile IMOx
emission rates.
1.2
-------
11-22
III. Summary
The historical background of the development of Federal mobile emissions
standards has been presented. Based on issued and planned Criteria documents,
desired air quality goals for health protection have been set at the following
levels:
CO 4 10,000 yg/m3 (9 ppm) - 8-hour average
Photochemical oxidants 4125 yg/m3 (0.06 ppm) - 1-hour average
N0£ 4190 yg/m3 (0.10 ppm) - 1-hour average
The complexities inherent in the control of photochemical oxidants have been
discussed and a control strategy involving joint control of hydrocarbons and
nitrogen dioxides adopted. Using a modified roll-back calculation with a 1967
baseline, the following 1980 total motor vehicle emission goals by which to
achieve desired air quality goals have been derived:
CO 6.16 g/mi
Hydrocarbons 0.14 g/mi
NOX 0.40 g/mi
The calculation of future motor vehicle emission goals should be a continuous
process with new data being used as they become available.
-------
11-23
REFERENCES
1. Clean Air Act, Public Law 88-206, 88th Congress, 1st Session, December 17,
3-1-2-
1963. In: U.S. Statutes at Large, 77_:932-401 (1964J.
2. Motor Vehicle Air Pollution Control Act, Public Law 89-272, 89th Congress,
1st Session, October 20, 1965. In: U.S. Statutes at Large, 79^:992-1001
(1966).
3. "Technical Report of California Standards for Ambient Air Quality and
Motor Vehicle Exhaust," California Dept. of Public Health, Berkeley,
California, (1960).
4. "Technical Report of California Standards for Ambient Air Quality and
Motor VEhicle Exhaust. Supplement Number 1, Crankcase Emission Standard,"
State of California, Department of Public Health, Berkeley, California
(Aug. 1961).
5. "Control of Air Pollution from New Motor Vehicles and New Motor VEhicle
Engines," Federal Register (Washington), Part II, 31_(61): 5170-5238
(March 30, 1966).
6. "Control of Air Pollution from New Motor Vehicles and New Motor Vehicle
Engines," Federal Register (Washington). Part II, 33_( 108):8303-8324
(June 4, 1968).
7. "Control of Air Pollution from New Motor Vehicles and New Motor Vehicle
Engines," Federal Register (Washington), Part II, 3j[(28):2791 (February
10, 1970).
8. Air Quality Act of 1967, Public Law 90-148, 90th Congress, 1st Session,
November 21, 1967. In: U.S. Statutes at Large, 81_:485-507 (1968).
-------
11-24
9. National Air Pollution Control Administration, Air QuaUty Criteria for
Photochemical Oxidants. U.S. DHEW, PHS, EHS, Washington, D. C. (March 1970),
10. National Air Pollution Control Administration, Air Quality Criteria for
Carbon Monoxide, U.S. DHEW, PHS, EHS, Washington, D. C. (March 1970).
11. National Air Pollution Control Administration, Air Quality Criteria for
Hydrocarbons, U.S. DHEW, PHS, EHS, Washington, D. C. (March 1970).
12. Wayne, L., e_t a1_., "Modeling Photochemical Smog on a Computer for
Decision-Making," Paper presented at the APCA annual meeting, St. Louis,
Mo. (Jwne 14-18, 1970).
13. Eschenroedor, A. Q. and J. R. Martinez, "Mathematical Modeling of Photo-
chemical Smog," General Research Corp., Santa Barbara, California (December
1969).
14. Los Angeles County Air Pollution Control District Data 1962 and 1967,
Los Angeles, California
15. Larsen, R. I., "A New Mathematical Model of Air Pollutant Concentration
Averaging Time and Frequency," J. Air Pollution Control Assoc., Ijhl
(January 1969).
16. Larsen, R. I., "Relating Air Pollutant Effects to Concentration and
Control," J. Air Pollution Control Assoc., 20:4 (April 1970).
17. Kramer, R. L. and N. P. Cernansky, "Motor Vehicle Emission Rates," U.S.
DHEW, PHS, EHS, National Air Pollution Control Administration. Durham,
North Carolina (internal document).
18. Hocker, A. J., "Exhaust emissions from Privately Owned 1966-1969 California
Automobiles: A Statistical Evaluation of Surveillance Data," California
Air Resources. Laboratory, Los Angeles, Calif., Supplement to Progress
Report Number 17 (February 6, 1970).
-------
11-25
19. Landsberg, H. H., L. F. Frischman, and J. L. Fisher, Resources in
America's Future, Patterns of Requirements and Availabilities 1960-2000,
John Hopkins Press, Baltimore (1963).
-------
III-l
Chapter III
THE ADVANCED AUTOMOTIVE
POWER SYSTEMS PROGRAM
by
John J. Brogan
Director, Div. of Motor Vehicle Research and Development, APCO
Environmental Protection Agency
Ann Arbor, Michigan, USA
-------
Ill-2
Introduction
The Advanced Automotive Power Systems Program is an outgrowth
of concern by our nation, led by President Nixon, that improvement
of air quality in the United States is markedly dependent on the
elimination of the automobile as a significant source of air
pollution. In the United States, the internal combustion engine
in our automobiles contributes about half of the total pollutants
from all sources to our air environment. Figure 1 illustrates
the automobile's share of the blame in the United States. Based
on 1970 estimates, nearly 70% of the carbon monoxide from all
man-made sources comes from the automobile, also 40% of the
hydrocarbons and about 40% of the nitrogen oxides appear from
this source. Less than 1% of the particulates and less than
0.10% of the sulfur oxide emissions nationwide are due to this
source.
One result of the nationwide concern over the automobile has
been the tightening of Federal exhaust emission standards that
must be met on new automobiles. This tightening of the standards
is apparent by inspection of Figure 2. This Figure shows the
Federal exhaust emission standards through 1973 for the three
major pollutants and estimated values for 1975/76. The final
values for these latter years have not been finalized, as yet.
By model year 1975, a 90% reduction in hydrocarbons and carbon
monoxide from the 1971 levels are shown, and, by 1976, the
nitrogen oxides levels will be equivalent to a 90% reduction from
the uncontrolled levels measured for the 1971 model year vehicles.
There is considerable doubt that the conventional spark-ignition
gasoline-fueled reciprocating internal combustion engine can be
'cleaned up' adequately to meet all of these forthcoming standards.
As a result, alternate power systems which are inherently clean
when compared with the conventional engine must be considered as
potential replacements.
-------
III-3
MILLION
TONS
13O
NATIONWIDE EMISSIONS ESTIMATES
•1970-
no
100
9O
7O
6O
4O
3O
2O
1O
ALL SOURCES
•
MOTOR
VEHICLES
POLLUTANT- CO
HC
NO
PARTICULATES SO
Figure 1.
FEDERAL MOTOR VEHICLE EMISSION STANDARDS
grams/mile
1971 1972 1973 1975 1976
HC
CO
NOL
4.6
47.0
—
3.4
39.0
-
3.4
39.0
3.0
0.46
4.7
3.0
0.46
4.7
0.4-0.6
Figure 2.
-------
III-4
Bases for the Program
A panel of scientists originating in the Executive Office of the
President was formed in 1969 to evaluate industry efforts in our
country to develop alternate power systems. The conclusion
reached by the Panel was that industry was not making a serious
effort to make such developments. Lacking sufficient industry
interest, a Federal research and development program was
recommended. This program was announced by President Nixon on
February 10, 1970 (see Figure 3) and the program was underway
by July. In addition, in his February message the President
announced another Federal program to stimulate industry by
providing financial incentive to groups who independently develop
their own alternate power systems. This latter program is called
the Clean Car Incentive Program. Thus, the two parallel approaches
to develop alternate power systems appear as in Figure 4. The
total package consisting of the Research and Development Program
and the Incentive Program is called the Advanced Automotive
Power Systems (AAPS) Program.
What are the goals of the Advanced Automotive Power Systems
Program? The program is intended to provide insurance to our
nation that if a practical and virtually pollution-free power
system could be developed, this development would culminate in
a demonstrated system by 1975. Achievement of this goal can be
accomplished by demonstration of a complete power system either
developed directly under Federal government monies or, by similar
type demonstrations whose development is sponsored by industry,
perhaps as a result of stimulation brought about by the existence
of the Research and Development Program or of the Incentive
Program, or both.
In addition, the AAPS Program will provide important information
to the Administrator of the Environmental Protection Agency on
-------
III-5
EXCERPT ROM MESSAGE ON ENVIRONMENT
FEBRUARY 10,1970
u... with the goal of producing an
unconventionally powered, virtually pollution
free automobile within five yeans."
— PRESIDENT NIXON
Figure 3.
DEVELOPMENT OF ALTERNATE
POWER SYSTEMS APPROACH
TWO PARALLEL PATHS
FEDERALLY SPONSORED R/D
INCENTIVE PROGRAM FOR PRIVATE DEVELOPMENT
Figure 4.
-------
III-6
the technical feasibility of meeting the stringent 1975/76
emission standards with use of alternate power systems. Such
information aids in judgments made by the Administrator on
possible waiver requests to postpone enforcement of the
standards.
Research and Development Program
First, the research and development part of the Program will
be discussed. What are candidates for development in this
Program? There are five types of power systems that were
initially part of the Program when it began in July 1970.
These include the gas turbine, two kinds of hybrids—heat engine/
electrics and heat engine/flywheel systems—Rankine cycle power-
plants and the all-electric system. There are two additional
systems, the stratified charge engine and the Diesel engine which
are serious contenders for candidacy. The particular versions of
these last two systems under consideration are reasonably well-
developed and are being considered as candidates because of
their nearer term potential for meeting the 1975/76 emission
standards when mass produced. Thus, it is these two engines
which we expect will provide important information on whether
alternate systems can meet the standards and this information
will be available very early in the Program.
Figure 5 lists all of the candidate systems and the stratified
charge and Diesel. In general, the candidates were selected
because the technology to improve, to improvise, and to further
develop each, exists in abundance in the United States. In
addition, independent research sponsored by private industry in
the United States had been underway for some time on most systems
selected. Thus, the interest to further develop the systems was
there. Additional candidate systems may be brought into the
Program. We are continually reassessing new developments on other
systems including developments made within our country and
worldwide.
-------
III-7
It is pointed out that the Stirling cycle powerplant has been
seriously considered for candidacy since the Program work began.
The considerable strides made in the Netherlands and Sweden in
making the Stirling cycle powerplant a competitive system have
been followed closely by our staff and our Technical Advisory
Committee. However, we are not ready as yet to bring this system
into the Research and Development Program.
The all-electric is shown apart from the other candidates in
Figure 5. We do not see the complete development of the all-
electric system early enough to meet the 1975 date because the
type of batteries with adequate power and energy densities
needed to compete in performance and cost with the conventional
engine are presently in the basic research stage. The battery
systems of special interest herein are the high temperature
alkali-metal battery systems such as Lithium-Sulfur and Sodium-
Sulfur. We recognize that there are several types of batteries
in use today for all-electric automobiles and small vans.
However, in order to meet the driving requirements of the
American public, the development of completely new and more
powerful batteries is needed. As a result of these considerations
the all-electric car is not a candidate to meet the 1975 goal,
rather we see its development by 1978. So, one may ask why is
this system in the Program? It is in the Program because
eventually these new and better battery systems must be developed.
The cost of development is very high as is the risk. For
industry at the present time, the costs and risks appear to out-
weigh the ultimate advantage of having these batteries available.
By supporting the research and early development with government
monies we see a point being reached where industry will pick up
the further development of these batteries.
On what basis will any of these propulsion systems be brought
through the development phases into complete system hardware?
-------
CANDIDATE POWER SYSTEMS
HEAT ENGINE/ELECTRIC HYBRIDS
2nd GENERATION ICE's-STRATIFIED CHARGE ENGINE
DIESEL ENGINE
RANKINE CYCLE
HEAT ENGINE/FLYWHEEL HYBRIDS
ALL ELECTRIC
Figure 5.
-------
III-9
The criteria which we consider are summarized in Figure 6.
Performance requirements are defined for the system and for
individual components. Based on the designs formulated and the
tests conducted, we must have confidence that these requirements
can be met. The characteristics of the largest selling model
type automobile in the United States for year 1970 form the
basis for engine weight and volume constraints, road performance
requirements and fuel economies. Of course the exhaust emissions
must meet or do better than the emissions standards of 1975/76.
The safety characteristics of the automobile with the new engine
are another important consideration. For example, it is de-
sirable that the working fluid in the Rankine cycle engine be
non-toxic and non-flammable. The flywheel design must be such
that any structural failure of the flywheel itself not lead to
catastrophic failure outside of the confines of its casing.
The socio-economic impact noted on the slide relates to the
criteria that engine exhaust odor and noise must not be
ojectionable according to contemporary standards and that the
system cost to build in-quantity production be competitive with
the conventional engine with its exhaust treatment devices
installed on the latter. National impact is listed last, yet,
it ranks as number one on anyone's list of criteria. As one
example of how we evaluate the impact, let us consider impact
of the all-electric automobile.
Figure 7 shows the projected power generating capacity in the
United States beyond year 2000. In some of the most densely
populated areas of our country, such as New York City, we find
that the demands for electricity, at times, are not being met
by the supply. This is a city without any appreciable electric
car popultion. On a nationwide basis, if all the vehicles on
the road today were all-electrics, the total demand for
electricity would exceed the existing capacity to supply by
-------
111-10
POWER SYSTEM ACCEPTANCE CRITERIA
• SPECIFIED SYSTEM AND COMPONENT
PERFORMANCE REQUIREMENTS
• VERY LOW EXHAUST EMISSIONS
• SAFETY
• SOCIO-ECONOMIC REQUIREMENTS
NATIONAL IMPACT
Figure 6,
PROJECTED POWER GENERATING
CAPACITY IN THE UNITED STATES
O 2
ill *-
>105
£ 8
O 6
10.4;
TOTAL POWER
CAPACITY-
-NATURAL GAS
-OIL
LHYDROELECTRIC
1960 1970 1980 1990 2000 2010 2020
YEARS
Figure 7.
-------
III-ll
about 40%. In addition, the technology to control sulfur
oxides from fossil fuel burning power stations is still being
developed. Further, the nuclear source curve shown in Figure 7
is very optimistic and since drawing this Figure, one year ago,
our experience has been that as time goes on this cross-over
point will move farther to the right. Therefore, if recharge-
able alkali-metal batteries were fully developed right now and
all-electric automobiles produced in-quantity we would
aggravate our existing electrical power capacity problem and,
in addition, temporarily transfer the major sources of pollution
from a mobile to a stationary source. So then these rre some
of the factors that we consider in judging whether a candidate
powerplant should go beyond the paper study phase and into
hardware, and, of equal importance, these factors influence the
timing of certain research work and the pace of development of
some candidates.
Individual Candidates and Their Status
Now let's take a closer look at the candidates and their status.
Each candidate was at a different state of development when the
Program began, and depending on our success in achieving
technological breakthroughs, each may enter the system hardware
phase at different times during the next five years. The
first 18-month phase of the Program is intended to be a period
of evaluation wherein the systems and their components are to
be designed, the critical components bench tested and decisions
made on whether to proceed to first generation hardware.
Rankine Cycle Engine. Consider the Rankine cycle engine as
illustrated in Figure 8. In this engine, there is an external
combustor and an enclosed working fluid which is heated,
expanded to do work, then condensed to a liquid, with the fluid
being continuously recycled. There are three types of Rankine
-------
111-12
BASIC RANKINE CYCLE ENGINE
Figure 8.
WATER FLUID
S- ENTROPY
CANDIDATE RANKINE CYCLE POWER PLANTS
ORGAN 1C WORKING FLUID- RECIPROCATING EXPANDER
WATER BASE WORKING FLUID- RECIPROCATING OR A
ROTARY EXPANDER
•ORGANIC WORKING FLUID-TURBINE EXPANDER
Figure 9.
-------
111-13
systems which are presently in the design and component test
phase as shown in Figure 9. The work on the organic working
fluid reciprocating expander system is summarized in the
technical paper presented in this meeting by a representative
from Thenno-Electron Corporation where this work is being
conducted in our country. Work on the other versions of
Rankine cycle systems noted on the Figure are just beginning.
As we see them, the problem areas associated with a practical
design of the Rankine cycle appear on the next slide. The
problems appear mainly in the inefficiency of components and
complexity of the control system and, of course, with its
exhaust emissions. A representative of Solar Division of
International Harvester Company in San Diego, California, one
of three firms conducting government sponsored research and
development on improved combustors will summarize the work of
one of the high efficiency combustor systems at this meeting.
The next Figure summarizes a schedule of the planned work on
the Rankine cycle. Complete system development for the three
types mentioned earlier with parallel research on the components
is shown. We point out that the three systems contractors are
designing complete systems including all components. An integral
part of this effort is to define the design requirements that
each component must meet. For example, the temperature of the
condensate, maximum and minimum flow rates and heat release
rates are part of the condenser design requirements. The
systems contractor and the backup component research contractor
are designing a condenser to meet the identical requirements.
The condenser research contractor will design and verify in
test his condenser designs for each of three systems. The
first generation system hardware for any of the three systems
may have components made by one or more of the component research
contractors, depending on whose component design is the most
efficient with cost to produce being one of several important
-------
111-14
RANKINE CYCLE ENGINE
PROBLEMS
'CONDENSER SIZE WT.
'BOILER SIZE WT.
'CONTROL COMPLEXITY
'ENGINE EFFICIENCY
•MINIMIZE EMISSIONS
•FREEZING (water)
•LUBRICATION
'FEEDPUMP
•SEALS (non-water)
•VALVING DESIGN
Figure 10.
RANKINE CYCLE ENGINE PROGRAM
SYSTEM
DEVELOPMEIS
COMPONEN1
RESEARCH
1969
197O
System De
IT
Con
Con
Co IT
1971
sign Co
1st (
1972
mpone
1973
nt Test
1974
-3 Sys
1975
terns
generation Hardware
2nd G
rol System
denser
ibustio
n Rese
sn. Har
0—
arch (3
dware
o
1
Figure 11.
-------
111-15
elements applied in that judgment. Lastly, you can see from
Figure 11 that delivery of the first prototype engines could be
as early as 1973.
Gas Turbine. More work has been conducted by industry in our
country on the gas turbine than on any other candidate with an
unconventional engine for the automobile application. For the
gas turbine we are focusing our efforts on solving some of the
problems which have plagued past attempts to get the gas turbine
on the road. Some of these problems are noted in Figure 12.
These problems include the need for: developing manufacturing
techniques for mass producing turbine inexpensively, for
improving the part load, fuel economy, combustion research to
improve combustor design and increasing system reliability.
Actually, if we operated our automobile in the same way that
commercial jet aircraft are operated, namely, running most of
the time at constant speed, many of these problems would
disappear. This gas turbine program is oriented toward solving
problems, and if successful, is intended to stimulate industry
to apply the results of this research to their own turbine
designs. As we see it, the demonstration of the turbine in the
automobile probably will be performed by industry. It is because
industry is so close to practical hardware on this system that
our approach here emphasizes problem solving rather than a
government sponsored demonstration.
One of the big unknowns in getting the turbine on the road is
knowledge of the cost to mass produce this system; therefore,
early recognition of what are the high cost components of the
system is needed so that fruitful manufacturing research can be
properly directed. A representative of Pratt & Whitney Aircraft
from the United States will present findings on this subject in
this meeting. That firm is on the team of contractors in this
program.
-------
111-16
Hybrids. Now, for the hybrids, the heat engine/electric and the
heat engine/flywheel. The heat engine/electric hybrid consists
of a small size engine such as in a Volkswagen, and an array of
inexpensive lead-acid batteries. The manner in which these two
power sources are arranged is illustrated in Figure 13. The
same series versus parallel configurations can be applied where
the flywheel replaces the battery system. The system is designed
to extract power from the heat engine alone, or from both
sources, heat engine and battery, at the same time. We find
that the parallel hookup is best where the conventional internal
combustion engine is used as the heat engine and a series
configuration appears best where a small gas turbine is used.
In either configuration the system operates by running the heat
engine at a constant speed—say equivalent to 40 mph road speed,
and, where vehicle accelerations are required, the additional
power comes from the battery system. Above a 40 mph road speed
the engine speed slowly converges on the new steady-state
level demanded by the driver with additional power required
during slack time provided directly from the battery system.
There are several potential advantages which this hybrid concept
offers. One is that the engine speed range is relatively small
with the attendant ease of control of exhaust emissions under
such a condition. Another advantage is that the very high road
performance for a standard size American automobile, of
approximately 4,000 Ib. weight, can be demonstrated using a
relatively small and inexpensive heat engine. A basic problem
with this type system appears in its relative complexity, higher
cost and the larger system volume required with use of two power
sources. Development of improved lead-acid batteries to
accommodate the rapid charge-discharge characteristics needed
for this mode of operation is now underway in the Research and
Development Program.
In our analysis of this type hybrid many types of heat engines
have been considered by our contractors. Two types were noted
-------
111-17
OASTURBINEENCjINES
MAJOR PROBLEM AREAS
• MANUFACTURING COST
•FUEL ECONOMY
• EMISSIONS-NOXREDUCTION FOR 1980 GOALS
•RELIABILITY
Figure 12.
HYBRID SCHEMATIC
SERIES HYBRID
DRIVES WITH
BATTERY ENERGY
STORAGE
ELECTRIC
'GENERATOR
CONTROLS - ^
1
BATTERY
PARALLEL HYBRID
DRIVES WITH
BATTERY ENERGY
STORAGE
II
ELECTRIC
MOTOR/GENERATOR
Figure 13.
-------
111-18
earlier. Two firms in our country are conducting the bulk of
our work on this hybrid. One firm is the Aerospace Corporation.
A member of the Aerospace Corporation team will report on their
work at this meeting.
As stated earlier, the heat engine/flywheel system would function
in a similar manner to the heat engine/electric with the battery
replaced by a mechanical storage device, namely, the spinning
flywheel. We have progressed on this system from parametric
analysis of many practical flywheel materials such as shown in
Figure 14 and numerous practical configurations as shown in
Figure 15 to the design aid fabrication of specific flywheels
for cars.
Today we are at the point where two flywheels are now being
tested to verify predicted energy and power densities. One of
the flywheels is fiberglas and the other fabricated of 4340
steel. Work on the configuration of the key component of this
system, the transmission, will begin shortly. To give you a
feel for some characteristics of the flywheel: for a full
size family car, the flywheel is made of 4340 steel in a
constant stress configuration weighing 42 Ibs. and operating
at a maximum speed of 24,000 revolutions per minute.
All-Electric Car. Alkali-metal battery research and development
for the automobile application has been underway for more than
a year at Argonne National Laboratories. Their work was reported
on in this meeting. We anticipate that the 'proof-of-principle'
for the high temperature lithium-sulfur system will be demonstrated
on single and multiple cells within the next 18 months. This
'proof-of-principle' refers to the experimental demonstration
that the energy density and power density desired for a full-
sized battery can be obtained on an elemental cell basis. Once
this proof-of-principle has been achieved, the battery work will
move into a development phase, first with a goal of a 2 kw battery,
then a 5 kw battery and then a 20 kw battery system.
-------
FLYWHEEL MATERIALS
MATERIAL
I8N 1-400
(MARAGING STEEL)
18N1-300
(MARAGING STEEL)
4340 STEEL
1040 STEEL
1020 STEEL
CAST IRON
2021 -T8!
(ALUMINUM)
2024 -T851
(ALUMINUM)
6AL-4V
(TITANIUM)
V GLASS
S-IOI4 GLASS
DENSITY
M -
L8S/IN3
0.289
0.289
0.283
0.283
0.283n
0.280
0.103
0.100
0.160
0.092
0.067
POISSON'S
RATIO
M
0.26
0.30
0.32
0.30
0.30
0.30
0.33
0.33
0.32
0.20
0.20
ULT
TENSILE
fou)KSI
409
307
260
87
68
55
62
66
150
130
250
YIELD
TENSILE
(Fty)KSI
400
300
217
58
43
37
52
58
140
-
-
REC. WORKING
STRESS
(
-------
FLYWHEEL GEOMETRIES
PIERCED
DISC
RIM
SOLID
DISC
CONICAL
CONSTANT
STRESS
LOG OR
BAR
M
I
SHAPE
r
r
STRESS
MAP
RIMSTRESS-v
VOLUMETRIC
EFFICIENCY
(FW$MIN.WT. HOUSING)
30.4%
20%
30.4%
62%
80%
6%
Figure 15.
-------
Ill-21
Stratified Charge Engine. As mentioned earlier, the two
candidates which we are seriously considering for entry into the
Program are the stratified charge engine and the Diesel. The
stratified charge engine is a spark-ignited gasoline fueled
internal combustion engine with many hardware characteristics
of the conventional engine. Differences appear mainly in the
combustion chamber design, use of fuel injection, and in the
resulting combustion process. In one version of this engine
the fuel is injected into the cylinder as the spark appears and
therefore the fuel burns as it enters. The fuel burns initially
in an over-rich condition with fuel injectors designed to permit
fluid swirl at the top of the cylinder. The burning expands
into the lower portions of the cylinder and, overall, the
resultant combustion products are similar to those obtained using
a very lean mixture. Lean burning is very desirable from the
emissions viewpoint. The initial work on this engine was
sponsored by the U.S. Army Tank-Automotive Command in the state
of Michigan. The measured exhaust emission levels for the
stratified charge engine installed in jeeps and employing a
catalytic muffler are below the standards for hydrocarbons and
carbon monoxide set for 1975 but further work must be conducted
to reduce the nitrogen oxide emissions. Several generations
of development have been funded and if this system is brought into
the Program, the work will emphasize achievement of reductions in
nitrogen oxide emissions and then fleet testing of this well-
developed system.
Diesel Engine. The Diesel engine is not commonly used in American
made automobiles, mainly because it is heavy and costs more to
manufacture compared with the conventional Otto cycle engine. If
the Diesel is brought into the Program, emphasis will be placed
on furthering the development of a low compression ratio Diesel
with a high swirl and prechamber design. Exhaust emission levels
for hydrocarbons and carbon monoxide which are lower than the
1975 standards have been shown for this type engine without
-------
Ill-22
resorting to a catalyst. The measured nitrogen oxide levels
on the first generation engine are relatively low but are within
reaching distance of the 1976 standard. Work on this type Diesel
will concentrate on nitrogen oxides reduction, and on performance
durability and driveability testing.
Federal Clean Car Incentive Program
The goals of the Research and Development Program, the near
candidates and our approach to further development of each have
been discussed. As stated earlier, there is another important
program underway in our nation to develop a virtually pollution
free automobile using an alternate power source. This work
consists of efforts on the part of industry itself with
financial incentive provided by our government to inspire
independent work. This describes the Federal Clean Car Incentive
Program. The intent of the Program is to stimulate the private
sector to produce a virtually pollution-free automobile. The
program provides a market for large and small auto and non-auto
manufacturers who often possess new and unique approaches toward
engine designs. However, in the past, they have lacked incentive
to further independent development.
The Program description is summarized in Figure 16" After
successfully passing stringent emissions and performance testing,
first on a leased Prototype car, then on 10 purchased copies of
the prototype for Demonstration, the successful engine system
will be further tested after procurement of 300 vehicles. If the
low emissions levels are maintained and road performance satis-
factory, the car then receives certification as a low-pollution
vehicle, and this certification is significant. The certified
vehicles which make their way through the Program will be
favored by purchase in quantity for Government fleet use.
Premiums as high as 200% over the basic price normally charged
to the government are permitted for the successful developer.
-------
ELEMENTS OF THE FEDERAL INCENTIVE PLAN
PROTOTYPE
PHASE
-J.EASE 1
*
/
Meet Prototype Specs.
DEMONSTRATION
PHASE
-7EJT
FLEET TEST
PHASE
Government Procure-
ment in Quantity
i
Legislation
Low Emission
Vehicle Certification I
•TES7\
.PURCHASE
PURCHASE
10 VEHICLES.
Figure 16.
-------
111-24
The legislation to accommodate such purchases appears in the
Amendments to the Clean Air Act, recently passed by our Congress
and signed by our President. We point out that in this Program,
the developer retains all patent rights and that the total cost
to manufacture, including tooling costs plus a reasonable
profit, is provided to the developer.
The Incentive Program is interrelated with the Research and
Development Program in that, where desirable, a partially
successful candidate in the Incentive Program could receive
research funding to further improve the engine system and this
funding could come from the Research and Development Program.
Lastly, we mention that the Incentive Program is expected to
provide a valuable source of information from actual vehicles
from which to judge the capability of the industry to meet
1975/76 emission standards.
This Clean Car Incentive Program has just begun, with approximately
20 proposals from industry to enter the prototype phase received
recently, and more to come. While proposal evaluation is not
as yet complete, as we see it now, delivery of the first proto-
types of low emission cars into this Program will be made before
August of this year. Figure 17 summarizes Program status.
The Incentive Program has been planned at $20 million over a
three year period, The rate of use and extent of use of these
funds is dependent on the rate at which selected candidates
proceed through the Program. At each test stage, any given
candidate can be eliminated. The period of major costs should
be in 1972 and 1973 since a number of candidates would be
entering, or, in the demonstration test stage in 1972 and the
fleet test of one or more candidates should begin in 1973.
-------
PROGRAM STATU S
MARCH 1971 CONTRACTS (approx. 15) FOR
• Brayton Cycle Gas Turbine
-Rankine Cycle
•Heat Engine/Electric Hybrid
- Electric
* PROTOTYPE DEMOSTRATBONS IN 1971
WITH GAS TURBINES
Figure 17.
H
H
M
I
Ln
-------
111-26
World-wide Participation in Both Programs
Both Programs are open to firms residing outside of the
United States. Most contractual work in the Research and
Development Program is awarded based on competitive proposals by
industry in response to Scopes of Work which are well publicized
and then sent to firms requesting them. Multiple awards or
contracts are common to the Research and Development Program.
For example, we now have three different firms under contract
and working on three different approaches to develop high
efficiency combustors for the Rankine cycle system. There will
be four parallel approaches considered for the gas turbine
combustor design. As a step toward improving lines of communi-
cation between ourselves in the Program and industry in your
country, we intend to define mechanisms to provide the Scopes
of Work for new contracts to you and other interested industrial
firms. In addition, the Scopes of Work for all on-going
contracts, and final reports from completed contracts will be
made available to all interested parties.
We welcome your participation in both Programs. Any firm or
government representative who wishes information on participation
will be provided written material in this meeting.
Summary
I will now summarize this talk. We have defined the Research and
Development Program and the Incentive Program. Together, they
form the Advanced Automotive Power Systems Program. Their goals
are similar—to produce a virtually pollution free automobile—
their methods differ however. We feel confident that ultimate
goal will be achieved within the 5-year time frame. If the
conventional internal (Dmbustion engine can be modified adequately
by industry, so much for the better. But, it is these two
Programs which will provide the insurance to our nation in the
event that they cannot make it.
-------
111-27
Lastly, we urge that industry in your countries compete in
the Programs. We will do all that is reasonable to encourage
this activity. In this way all of us will benefit, here and
at home.
-------
IV-1
Chapter IV
THE POTENTIAL OF THE GAS-TURBINE VEHICLE
IN ALLEVIATING AIR POLLUTION
by
Edward S. Wright
Chief, Ground Systems Analysis
Research Laboratories
United Aircraft
East Hartford, Connecticut, USA
-------
IV-2
Introduction
Emissions from motor vehicles account for approximately 60 percent of
the annual total air pollutants emitted in this country (see reference
1); the three major pollutants emitted by automobiles—unburned hydro-
carbons (UHC), carbon monoxide (CO), and nitrogen oxides (NOX)—comprise
approximately 63, 93, and 46 percent, respectively, of the total of
each of these pollutants. Consequently, there has been a great deal of
interest in ways and means of reducing the emissions of Otto cycle
internal combustion gasoline engines, and of finding inherently cleaner-
burning alternatives to these engines.
It is generally believed that if the reciprocating internal combustion
engine cannot be cleaned up sufficiently in a reasonably economic manner,
the gas turbine is the most likely alternative. Its inherent emission
characteristics are generally considered to be satisfactory (presently
or potentially), and the primary issue is whether it is technically and
economically feasible. Based on several well-advertised programs of
the major automobile manufacturers in the United States and abroad, it
appears that the technical problems are surmountable, but there is, as
yet, considerable doubt concerning the economics of automobile gas
turbines.
The purpose of this paper is: (a) to summarize the emission characteristics
of automobile gas turbines, (b) to examine the potential of these engines
for satisfying automobile requirements, and (c) to assess their ability
(1) A study sponsored by the National Air Pollution Control Administration is
being conducted currently by the United Aircraft Research Laboratories,
with the purpose of examining the probable manufacturing costs of several
candidate automotive gas turbine engines, including those with simple
and regenerated cycles, and with free and single-shaft turbines. The
results, when published!, should provide an interesting data source in an
area where the available literature is scarce.
-------
IV-3
to compete with cleaned-up reciprocating internal combustion engines
by identifying their relevant technical (performance) and economic
(manufacturing-cost and fuel-consumption) characteristics. The paper
is intended to provid background information for a panel discussion at
the 1970 ASME Winter Annual Meeting,,
Emissions
Uncontrolled Emissions of Otto Cycle-Engined Automobiles--An automobile
utilizing an Otto cycle (gasoline fuel, reciprocating) engine with no
emission control devices has three sources of emission--the exhaust,
the crankcase, and the fuel supply system (fuel tank and carburetor).
Virtually all of the CO, NO , and lead, and about 60 percent of the UHC
X
come from the exhaust of the vehicle; about 30 percent of the UHC comes
from the crankcase; and the remaining 10 percent of the UHC comes from
the fuel supply system (see reference 2). In terms of quantity of emissions,
the pollutants in the exhaust, based upon a weighted average of engine
sizes, are 0.50 Ib/hr UHC, 3.5 Ib/hr CO, and 0.17 Ib/hr NO (see reference
X
3). Variations of + 20 percent or more are typical. The total emissions
from an uncontrolled Otto cycle-engined automobile have been estimated
as follows (see reference 1): unburned hydrocarbons--520 Ib/yr; carbon
monoxide--1,700 Ib/hr; and nitrogen oxides--90 Ib/yr.
Emission Standards—The first nation-wide standards for automotive emissions
were issued in 1966, and were to take effect on all 1968 model year cars
and light trucks (see reference 4). These standards are given in Table 1
-------
IV-4
in terms of concentration, and pertain to engines with 140 cubic inches
or large displacement. The standards for 1970 model year cars, also
shown in the following, tighten the 1968 standards by about 30 percent.
An important change in the standards is a change from concentration
(ppm) to a mass-per-mile basis. It will be noted that these standards
do not contain restrictions on NO or evaporations.
Table 1
Crankcase
Exhaust
Unburned Hydrocarbon
Carbon Monoxide
1968 Model Year
Zero
1970 Model Year
Zero
275 ppm 2.2 g/mi
1.5% volume fraction 23 g/mi'
Proposed future emission standards, in g/mi., are presented in Table 2,
based on data in reference 5.
Table 2. Emission Standards
Model Year
Pollutant
Unburned Hydrocarbon
Carbon Monoxide
Nitrogen Oxides
Particulates
Total
1971
2.7*
23
-_**
— **
1975
0.5
11
0.9
0.1
1980
0.25
4.7
0.4
0.03
21.8
12.5
5.4
^Includes evaporative losses from fuel supply system.
**No defined standard; estimated output for uncontrolled emission
is 5.8 g/mi of nitrogen oxides and 0.3 g/mi of particulates.
The proposed 1971 model year standards include an evaporative loss, and
the proposed 1975 model year standards include NOX and particulate require-
ments. The 1975 standards represent an approximate 80 percent reduction in
-------
IV-5
UHC, and a 50 percent reduction in CO from the 1971 standards, and an
estimated 85 percent reduction in NO from the uncontrolled automobile.
x
The proposed 1980 standards are those put forth by the President's
Environmental Advisory Board and represent reductions by about 50
percent in the 1975 standards.
Several methods of adapting the gasoline engine to meet these standards
have been proposed, and major efforts are being undertaken by the
automotive industry to meet them. However, the effectiveness of some
of the control devices has been questioned (see reference 6), and,
consequently, alternative propulsion systems have been proposed. There
is considerable evidence to suggest that the gas turbine has the potential
of low emissions without control systems.
Emissions from Gas Turbine Engines—A gas turbine does not have a component
that serves a function similar to the crankcase of the Otto-cycle engine.
Lubrication, where necessary, is usually by a method which could be
referred to as dry sump, with no chance of the fuel-air mixture venting
to the atmosphere from the oil tank.
The fuel supply system of a gas turbine consists of a fuel tank, injection
pump, and injectors; hence, there are no float chambers or other reservoirs
to allow evaporative losses such as those from the carburetor of automo^
tive reciprocating engines. Since a turbine would probably operate on
lower volatility fuels, such as kerosene or diesel oil mixed with low-
octane gasoline, the evaporative losses from the fuel tank are more easily
minimized.
-------
IV-6
Since neither the fuel system, the lubrication system, nor the fuel,
itself, (in terms of lead or sulfur content) is likely to be a source
of pollution from the vehicular gas turbine, it is evident that the
combustion process is the only significant area of concern. As described
in the following, the continuous combustion process at overall lean
fuel-air ratios leads to inherently low emissions levels of UHC, NOX,
and CO for gas turbine engines.
The concentrations of exhaust pollutants in gas turbine cannot be compared
readily with those of the Otto-cycle engine because of the disparity of
fuel-air ratios. Therefore, a method of comparison using a parameter
called the emission index (see reference 7) was devised. The emissions
index (El) is, essentially, a measure of the pounds of pollutants
emitted per pound of fuel burned; it can be calculated by using the concen-
tration of pollutant and the fuel-air ratio of the engine.
The overal1 lean fuel-air ratios and the absence of quenching on cool
walls results in complete combustion and, consequently, low levels of UHC
and CO emission. The UHC's are of two major types; the first is formed by
those hydrocarbons that pass through the combustion chamber unchanged;
the second consists of those which are produced in the low-temperature
region of the combustion process. The formation of both types, as well
as the formation of CO, takes place in the outer zones of the combustion
chamber, where the air injected to cool the liner walls has lowered the
local temperature. The emissions index is highest at idle for UHC and CO.
-------
IV-7
As the engine power output increases (i.e., the equivalence ratio
increases), combustion improves, the emissions index for UHC and CO
becomes negligible, and the absolute emission levels decrease signifi-
cantly—despite the increase in fuel consumption due to higher power
output.
As in the Otto cycle engine, the formation of NO is probably caused by
x
the oxidation of the nitrogen in air (0 + N_), and measurements to date
have indicated that it is a strong function of temperature and a weak
function of pressure, at least for practical gas turbine cycle pressures.
These measurements show NO levels distinctly lower than those expected
X
for equilibrium conditions (see reference 7), in contrast with the NOV
.A.
levels of Otto cycle engines. Furthermore, the NOX is essentially frozen
at the combustor exit, and concentrations remain the same during the
expansion process„ Among the reasons postulated for this low NOX concentra-
tion are the rapid quenching of the combustion products with the excess
air introduced into the combustor for cooling and the locally rich burning
near the fuel droplets in the primary combustion zone. Since temperatures
at the combustor exit are a function of both combustor inlet temperature
and the equivalence ratio, it has been assumed commonly that improvements
in cycle efficiency leading to higher combustor inlet temperatures (as a
result of higher pressure ratios, or of regeneration) or higher turbine
inlet temperatures will automatically lead to higher N0x emission levels,
because of the associated higher combustor temperatures. This assump-
tion is not necessarily valid„ Analytical and experimental research
conducted at the United Aircraft Research Laboratories has indicated that
-------
IV-8
the bulk of the NO formation occurs due to the strong recirculation
X
patterns characteristic of the primary combustion zones, since in straight-
through flow, the exposure of the mixture and products to high temperature
in the flame front is not of sufficient duration to allow significant NOX
formation (see reference 8). Thus, the prospect exists that further
research, which yields knowledge of how variations in primary combustion
zone geometry, residence time, combustion temperature, and fuel-air ratio
affect overall NO production, may indicate how combustor designs must
X
be modified in order to reduce the already low NO emissions level of
X
vehicular gas turbines even further.
Comparisons of Automobile Engine Types—Table 3 presents comparisons of
the emissions index (ratio of pollutant weight to fuel weight x 10-^) for
several automobile reciprocating engines and types of gas turbines, reported
from various sources and compiled in reference 7. Although these data
are not completely compatible and despite their wide variation, they can be
used as a basis of some general conclusions concerning the emission of CO
from a typical gas turbine. Using the El as a basis, the emission of CO
from a typical gas turbine is from 1 to 10 percent of an uncontrolled Otto-
cycle engine, the emission of UHC is 1 to 20 percent, and that of NO
X
from 30 to 80 percent.
Table 3. Comparison of Emission Indices
Engine/Engine Pollutant CO UHC* NO
Reciprocating (30 mph) 241 6.1 16.3
Reciprocating (cold start) 513 35 14.5
Reciprocating 407 35 13
Regenerative Gas Turbine (30 mph) 5.3 0.3 13.5
Gas Turbine (cold start) 50 1.1 9.8
Aircraft Turbojet 3.3 0.4 5.5
Aircraft Turbojet 19 25
^Expressed as hexane.
-------
IV-9
The procedure of totaling emission indices, used in this paper, is adopted
in order to provide gross comparisons between the two types of engines.
It is felt that this simplified approach has validity, since the emission
standards of Table 2 are based on reciprocating engine outputs, rather
than specifically on precise health-based criteria for particular pollu-
tants. On this El basis, the gas turbine ranges from 1 to 11 percent
of that of the untreated Otto-cycle engine.
Korth (see reference 9) and others have measured emissions from regenerated
automobile gas turbines operating in typical duty cycles. It is very
difficult to instrument mobile vehicles properly for accurate measure-
ments of the extremely low concentrations of pollutants typical of gas
turbine engines, and wide variations have been reported among several
makes of engines. Therefore, extrapolations of the test-stand measure-
ments of an aircraft gas turbine to the automobile standard duty cycle
are of interest. Although such an engine would not be applied to an
automobile, its emissions characteristics might closely parallel those
of a simple-cycle automobile engine of acceptable efficiency. For such
an engine applied to an automobile type duty cycle, an average total El
of 10.0 may be assumed which, at an assumed fuel consumption of 270 g/mi^i1
yields a total emission (UHC + CO + NOX) or 2.7 g/mi. Thus, the extrapo-
lated emissions level is approximately 50 percent of the proposed 1980
standard for the treated gasoline internal combustion engine, even though
its NO levels exceed the standard for that particular pollutant.
x
(1) About 11 miles per gallon.
-------
IV-10
It must be emphasized that this emission level is predicted for gas
turbine combustion technology of the 1960's, with no treatment, and with
no emissions criteria considered in the combustor design. Furthermore,
the ability of treated gasoline automobiles to continue to meet
emissions criteria over the lifetime of the vehicle is open to serious ques-
tion, even if costly maintenance and inspection procedures are adopted
and enforced. In the absence of perfect maintenance, if a 1980 Otto
cycle engine reverted to emissions of the level of the 1975 standard,
it would cause as much pollution as 4.6 untreated gas turbines (as
described in the foregoing); if it is reverted to 1971 standards, it
would cause as much pollution as 11.4 gas turbines; and if it reverted
to its natural untreated operating state, it would cause as much pollution
as 35 gas turbines.
The occurrence of the poor mixture/excess blowby condition commonly
observed on today's automobiles on the road would lead to even higher
levels of pollution. For the gas turbine, poor adjustment of fuel to
the combustor is unlikely to cause pollution problems., If the mixture
becomes fuel-rich, the engine simply puts out more power; if the car
were operating at maximum throttle, and the owner ignored overheat
warnings, the engine might fail ultimately, but the pollution control of
the gas turbine is fail-safe.
Alternative Power Plants—Other than the gas turbine, electric, steam,
and Stirling engines are the most frequently mentioned of low-pollution
alternative power plants to the gasoline reciprocating engine for
-------
IV-11
vehicular propulsion. Well-reported efforts have been directed by
various organizations toward investigating steam and electric alterna-
tives to the gasoline engine. These efforts nearly always confirm the
difficulties inherent with these approaches, lending support to the
strong economic pressures to continue the usage of gasoline engines for
automobiles, even though control of their pollution will involve sub-
stantial and costly modification. At present, the major automobile
manufacturers are publicly committed to the reduction of emissions from
gasoline engines as their preferred approach to the pollution problem,
but have identified the gas turbine as their first alternative.
From an emissions standpoint, gas turbine emissions are likely to be
iroughly equivalent to those emitted by a central power station producing
an equivalent amount of power for electrically powered vehicles.
Other Vehicular Applications--In terms of gross magnitude of emission of
pollutants, the automobile far exceeds all other vehicles combined, both
because of Otto cycle engine characteristics and sheer numbers„ Despite
the much lower levels of UHC, CO, and M) emissions contributed by diesel-
X
powered trucks, buses, trains, and off-highway equipment of various types,
special localized situations regarding pollution may favor the gas turbine
for these applications as well. For example, the urban diesel bus, while
theoretically a low emitter of pollutants, is severely criticized for
introducing noise and smoke to sensitive areas, and gas turbines may alleviate
these problems as well.
-------
IV-12
Conclusions Concerning Emissions—The technical literature documents
the fact that the gas turbine, at least in most of its forms, does indeed
have significantly lower emissions than the conventional Otto cycle engine
and can meet proposed future standards for UHC and CO without modification
(see references 7 and 9) . The ability of the turbine to meet future N0x
standards is not as clear, although it is hoped that the N0x emission will
be amenable to solution when sufficient research and development are per-
formed. Nonetheless, even without NOX reduction, total untreated emissions
(UHC, CO, NOX) of the gas turbine are likely to be only 50 percent of those
of the treated Otto cycle engine which meets proposed 1980 standards, and
less than 3 percent of the output of an untreated Otto cycle engine. If
all Otto cycle engines were replaced by gas turbines, the motor vehicle
contribution to pollution referred to on page 1 would probably be reduced
from 60 to 2 percent. In other words, it is likely that vehicular-caused
air pollution would be eliminated as a cause for serious environmental
concern.
Other Pertinent Characteristics
Engine Power Output—Figure 1 shows torque/speed characteristics of typical
reciprocating engines and five gas turbines of various configurations, i.e.,
single-shaft, differential with 10 percent compressor speed, single-stage
free-turbine, and multi-stage free-turbine configurations. Clearly, the
vide variations in the characteristics imply that when discussing "typical"
gas turbines, it is essential to specify the configuration. The configura-
tion most commonly considered for vehicular applications^ has been the single'
(1) The Chrysler, Rover- and General Motors automobile engines, and the GM,
Ford, and Leyland truck engines have used the free-turbine configuration.
-------
IV-13
stage free turbine, since the rising torque curve below rated speed
yields positive stability when the output shaft speed is in a fixed ratio
to the wheel speed. (As wheel speed drops due to increased resistance,
the increased torque of the engine tends to restore the system to its
original speed.)
3
o
2.8
2.4
2.0
?*-° 1.6
1.2
100% CONSTANT HORSEPOWER
MULTISTAGE FREE TURBINE
SINGLE STAGE FREE TURBINE
DIFFERENTIAL TURBINE (WITH 10%
COMPRESSOR OVERSPEED)
DIFFERENTIAL TURBINE (FOR
CONSTANT COMPRESSOR SPEED)
—, RECIPROCATING
_._. SINGLE SHAFT TURBINE
0.2 0.4 0.6 0.8 1.0 1.2
OUTPUT SPEED RAT!O N/NO
Figure 1. Engine torque speed characteristics.
The acceleration lag of gas turbine engines is due primarily to the
fact that compression ratios are more constant in reciprocating engines
(see reference 10). Nevertheless, many techniques are available to reduce
the lag to acceptable limits. These include the momentary removal of load
at some point in the power transmission train, reduction of engine shaft
mass inertia, introduction of variable geometry, or allowing momentary over-
fueling (overheat) in a fashion analogous to the accelerator pump of an
Otto cycle engine.
-------
IV-14
Because the gas turbine compresses far more air than is actually consumed
in the combustion process, it is more sensitive to ambient conditions
than its reciprocating counterparts: its power output and fuel consumption
are both affected more adversely by high temperature and altitude, and
the engine must be sized to provide adequate performance at reasonably-
adverse conditions. This fact implies that cars used extensively at
high altitudes and temperatures should be equipped with higher powered
engine options and also that certain potential "optional extras," such
as water injection, might increase the flexibility of the engine„
Maintenance, Reliability, and Life—Experience with aircraft and industrial
gas turbines has indicated that, because of their basic simplicity,
these engines can operate for thousands of hours with only minor routine
maintenance. The same can be expected for automobile gas turbines,
provided their unique requirements are adequately satisfied in the design.
One very significant requirement is the ability to withstand the thermal
stresses associated with the frequent changes in power and, where regener-
ated engines are required, the addition of cores and seals may complicate
the maintenance requirements.
Transmission Considerations--The engine torque/speed characteristics shown
in Figure 1 are unsatisfactory for vehicle propulsion without torque
multiplication. Examination of a spectrum of transmission configurations is
needed to match the wide variety of torque/speed curves characteristic of
various configurations of gas turbines to various vehicle requirements.
-------
IV-15
Ideally an infinitely variable (IV) transmission would be desirable
for use with gas turbines. In effect, the IV transmission decouples
vehicle wheel speed from engine speed and allows the engine to deliver
full power at any wheel speed (subject to traction limitations and trans-
mission efficiency), so that the torque speed characteristic of the
engine is irrelevant to the output of the transmission. The lead candi-
dates for vehicular gas turbine IV transmissions are electrical, hydro-
static, and mechanical.
The electrical system would consist of a generator (or alternator) driven
by the gas turbine, an electric motor (or motors) driving the axle or
wheels, and a control system,, Although the electrical system would yield
almost ideal power application, its manufacturing cost for automobiles
is likely to be high. Hydrostatic transmissions, consisting of displace-
ment hydrostatic pumps and motors, would require the development of
successful turbine shaft speed pumps and might have noise and safety
problems. Mechanical infinitely variable transmissions rely on rolling
contact along a variable radius, and generally suffer from Hertzian stress
problems at the contact surfaces. Therefore, they tend to result in
comparatively large and heavy installations. Nevertheless, they merit
serious consideration for automobile gas turbines.
The electrical transmissions described in the foregoing allow consideration
of hybrid automobiles and buses (see reference 11) in which power is drawn
not only from the prime mover, but also from energy storage devices, such
as batteries of flywheels. These hybrid approaches permit shutdown of the
-------
IV-16
prime mover (and thus elimination of all emissions) during operation in
emissions-sensitive areas, and also permit charging the energy system
at the most efficient and/or lowest emissions operating point of the
engine, thus reducing overall emissions.
Transmissions with a finite number of gear ratios are attractive because
of their low cost and high efficiency. For a given engine application,
the higher the basic engine torque at reduced speed, the fewer the gear
ratios needed; thus, a single-shaft engine would probably require an
impractical number of gear ratios. Hydraulic torque converters are used
in automobiles, where they are generally teamed with automatically shifting
two- or three-speed geared transmissions; they are inexpensive, conven-
ient, and adequately efficient. For a single-shaft gas turbine, such
a transmission with a minimum of 5 or 6 gear ratios (or a multiple-stage
torque converter) may furnish acceptable performance at reasonable cost.
Vehicle Performance--The prime mover and transmission must be selected to
meet some mission envelope of vehicle performance (desired accelerations,
velocities, and elapsed times for a range of payloads). Standard computa-
tional procedures (see reference 12) relate propulsion power requirements,
via transmission efficiencies, to road, grade, and air resistance; the
acceleration performance depends, additionally, on the mass and rotational
inertia of the vehicle. With appropriate allowance for installation and
accessory losses, these procedures serve to establish the engine power
requirementso Installed horsepower, as presently defined for automobiles,
(1) For a free-turbine engine, the hydraulic torque converter is probably
redundant.
-------
IV-17
can be misleading, since a typical automobile equipped with a nominal
250-hp engine will require about 30 hp to cruise continuously at 65
mph. (A truck at maximum gross weight equipped with a 250-hp engine,
on the other hand, is likely to require its full power to cruise at
that speed.) A typical automobile engine duty cycle, adapted from
information supplied by Rover (see reference 12), is reproduced in
Figure 2.
0-15 15-30 30-60 60-90 90-120120-150
POWER LEVEL (hp)
Figure 2. Typical automobile engine duty cycle.
The weight of the engine is important for establishing automobile perfor-
mance. Present engines and radiators average 16.5 percent of total car
weight (see reference 11), or approximately 660 pounds for the nominal
4,000-pound automobile. If the corresponding gas turbine weighed 160
pounds, the 500-pound saving would allow significant reductions in power
output for a given performance level, since peak power requirements for
the automobile are largely determined by the acceleration desired, and
acceleration power is directly proportional to mass. For example, the
-------
IV-18
power required at the road solely to accelerate a 3,500-pound automobile
at 5 mph per second is obviously 12..5 percent less than that required
for a 4,000-pound automobile (106.3 hp versus 121.5 hp at 50 mph).
Applying standard road and aerodynamic resistance calculations to typical
automobiles, summing power requirements, and assuming equal transmission
efficiencies, leads to the conclusion that equivalent performance at 50
mph could probably be achieved with 11.8 percent less engine output
power in the lighter vehicle (135 hp versus 153 hp) .
Fuel Consumption
Historically, regenerators have been regarded universally as essential
for practical gas turbines in automobile applications, because simple-
cycle engines have been considered unsuitable for this application on the
basis of poor full-load thermal efficiency and, even worse, part-load
efficiency. Unfortunately, the addition of the regenerator considerably
diminishes the appeal of the gas turbine engine in other respects, since
it: (a) adds complexity, cost (manufacturing, maintenance and development),
weight, and volume; (b) reduces specific power output and reliability;
and (c) complicates control and response.
Simple-cycle engine efficiency (and specific power output) can be described
conveniently by the use of en ly three parameters, i.e., pressure ratio,
turbine inlet temperature, and flow path efficiency, where the latter
represents both component efficiencies and pressure losses and is a
measure of the aerodynamic sophistication of the engine (see reference 14) .
Figure 3 shows the relationship between full-load sfc and these parameters
-------
IV-19
for a fixed flow path efficiency of 0.69 (see reference 15). As can be
seen, a wide range of combinations of pressure ratios and turbine inlet
temperatures will yield sfc's below 0.50, which has been the goal for
regenerated engines for automobiles.
0.80
1500 1700 1900 2100
TURBINE INLET TEMPERATURE,°F
Figure 3. Fuel consumption relationship for simple-
cycle engines with constant flow-path efficiency
The development of small, simple, advanced-technology components, with
high efficiencies and high pressure ratios (see reference 16), permits
reconsideration of the simple-cycle gas turbine as an automobile power
plant. Although the fuel consumption of this engine will be higher than
regenerative designs based on the same component technology,1 and possibly
slightly higher than that of competitive gasoline reciprocating engines,
this engine may be competitive for several reasons: namely (a) automo-
bile customers are not necessarily deterred by higher fuel consumptions
(power mileages), provided the other features (including price) appeal to
them; (b) the amount of fuel burned by the gasoline internal combustion
(1) Both simple- and regenerative-cycle engine concepts based on small,
simple, advanced-technology components are under study at United Air-
craft Research Laboratories as part of the aforementioned manufacturing
cost study. 5
-------
IV-20
engine, as well as the cost of the fuel may Increase as a result of
pollution control measures; (c) the gas turbine can burn lower-cost
fuels; and (d) the possibility exists of governmental fuel taxation
policy to favor the intorduction of low-pollution vehicles.
Manufacturing Cost—Although definitive production cost estimates
are highly proprietary at this time, General Motors has released data
indicating an opinion that the cost of automobile gas turbine engines
will be approximately three times that of an untreated reciprocating
engine, and from 1% to 2 times that of an eventual, pollution-treated
reciprocating engine as shown in Figure 4 (see reference 17). These
costs appear too high for serious competition with reciprocating engines,
and, therefore, methods of reducing this level of manufacturing cost would
appear mandatory in order for the gas turbine to replace reciprocating
engines in quantities sufficiently large to have a significant effect on
air pollution levels.
With regard to manufacturing cost, Eckert predicts that a proper gas turbine
design adopted to modern manufacturing techniques should result in a cost
division of 85 percent for materials and 15 percent for labor (see
reference 10). Therefore, reduction of material costs are of primary
interest in reducing manufacturing cost. Since superalloy material costs
predominate in the raw material cost of today's gas turbine engine, one
approach to reducing the manufacturing cost is to substitute ceramic materials
wherever possible for refractory superalloys in order to take advantage
of raw material costs of 5 to 40 cents per pound as opposed to $1.50 to
-------
IV-21
to $6.00 per pound; however, the lack of tensile strength in ceramics
in an impediment to this approach (see reference 18).
—
'1 80M960-NO CONTROLS
\ 1
«
E
0
5,60
UJ
Q
* 40
O
z
O
20
Z
O
ta
1 0
\
\
\
-1
M966-EXHAUST
_|
11970-IMPROVED EXHAUST
- '1970-CALIF. STD
51 IKLIN v»,
- 197X-EXHAUST REACTORS G£,\ J"" *!!1E'
,1111111111/11111" SltAM, tit.
1 | | .//////
RELATIVE POWERPLANT COST
Figure 4. Carbon monoxide reduction economics,
Another approach is to increase the specific output of the engine by
increasing pressure ratio, turbine inlet temperatures, and flow path
efficiencies, thereby reducing the size and material content of cost-
critical parts.
As stated in a previous footnote, a study is under way concerning the
probable manufacturing cost of high specific output engines. Projections
will be made concerning probable manufacturing costs of gas turbine engines
for automobiles in units of 100,000 and 1,000,000 annually, based on
material content and related costs.
-------
IV-22
Conclusions
Properly designed gas turbine engines for vehicular applications offer
significant potential for alleviation of reciprocating engine-caused
air pollution. From the emissions standpoint, this potential is repre-
sented by a predicted total emission level only 2 to 4 percent of that
characteristic of untreated gasoline reciprocating engines and only 15
to 30 percent of that of reciprocating engines meeting proposed 1975
emissions standards.
Therefore, widespread adoption of vehicular gas turbine engines has the
potential of practically eliminating vehicular air pollution as a
subject of serious environmental concern, provided that the performance,
efficiency, and manufacturing cost of these gas turbine engines are
competitive with the reciprocating engines they would replace.
Acknowledgment
The assistance of Drs. F. L. Robson and C. T. Bowman of the United Aircraft
Research Laboratories in the preparation of the emissions section is
gratefully acknowledged.
-------
IV-23
References
1. Morse, R. S., "The Automobile and Air Pollution," A Program for
Progress USGPO, December 1967.
2. Middleton, J. T., "Future Air Quality Standards and Motor Vehicle
Emissions Restrictions," National Conference on Air Pollution,
Paper A-2, December 13, 1966.
3. Maga, J. A., and Kinosian, J. R., "Motor Vehicle Emission Standards--
Present and Future," SAE Paper 660104, January 1966.
4. Anon., "Control of Air Pollution from New Motor Vehicles and New Motor
Vehicle Engines," The Federal Register, Vol. 31, No. 61, March 30, 1966.
5. Anon., "Air and Water News," December 22, 1969 and March 9, 1970.
6. Brubacker, J., and Grant, E. P., "Do Exhaust Controls Really Work?",
Second Report, SAE Paper 670689, 1967.
7. Sawyer, R. F., Teixiera, D. P., and Starkman, E. S., "Air Pollution
Characteristics of Gas Turbine Engines," Journal of Engineering for
Power, Transactions of the ASME, Vol. 91, Series A, No. 4, October
1969, pp. 290-296.
8. Marteney, P. J., "Analytical Study of the Kinetics of Formation of
Nitrogen Oxides in Hydrocarbon Air Combustion," Combustion Science
and Technology, Vol. 1, No. 6.
9. Korth, M. W., and Rose, A. H., Jr», "Emissions from a Gas Turbine
Automobile," SAE Paper 680402.
10. Eckert, B., "Has the Automobile Gas Turbine a Change?", A.T.2, Vol. 69,
No. 9, Sept.
11. Hoffman, G. A., "Hybrid Power Systems for Vehicles," Symposium on Power
Systems for Electric Vehicles, U. S. Department of Health, Education
and Welfare, National Center for Air Pollution Control, 1967.
12. Anon., "Truck Ability Prediction Procedure," SAE Handbook, Supplement
82, Society of Automotive Engineers, May 1968.
13. Penny, N., "Rover Case History of Small Gas Turbines," SAE 634A,
January 1963.
14. Wood, H. J., "A Polytropic Technique for Gas Turbine Performance
Prediction and Evaluation," SAE 660161, January 1966.
-------
IV-24
15. Kahle, G. W., and Wright, E. S., "Thermal Efficiency Versus Engine
Price—Optimizing for Industrial Vehicles," ASME Paper No. 67-6T-41,
March 1967.
16. Kenny, D. P., "A Novel Low Cost Diffuser for High Performance Centrifugal
Compressors," ASME P,aper No. 68-GT-38, March 1968.
17. Anon., "G. M. Progress of Power," G. M. Report, May 1969.
18. McLean, A. F., "The Application of Ceramics to the Small Gas Turbine,"
ASME Paper No. 70-GT-105, May 25, 1970.
-------
V-l
Chapter V
THE STIRLING-CYCLE ENGINE
by
R. A. J. 0. van Witteveen
Stirling Group
N. V. Philips
Eindhoven, The Netherlands
-------
V-2
Air Pollution is a very serious issue which is now causing considerable
concern throughout the world. The convening of this meeting demon-
strates this concern and we sincerely trust that it will not only high-
light the problem but will also contribute to the solution of air pollution,
It is hoped that the intensive Stirling engine research programme in the
Netherlands offers some positive contribution to the alleviation of the
air pollution problem and in particular, the automotive pollution.
Based on our experience with this engine, its characteristics, the
progressed state of the art and further prospects, we are convinced that
the Stirling engine should be regarded as one of the promising candidates
for alternative vehicle propulsion. It represents a very clean engine
in the heavy duty vehicle application (on and off highway), and even for
light duty vehicles.
The aim of this presentation is to provide an impression of:
(a) the nature of the work we have been doing and the resultant
expertise which has been achieved
(b) the progress made
(c) the future potential of the Stirling engine and the envisaged
applications
(d) the formulated plans for the future
It is impossible to adequately portray all the above facets within the
time scale allocated for this presentation and for this reason tomorrow
afternoon has been set aside to enable members of the conference to
visit the facilities at Philips.
-------
V-3
The initial development of the Stirling engine carried out over many
years was not specifically directed towards vehicle use. However, this
period was valuable in providing a sound basis for the later research
and development phase which commenced during the early 60's.
Fig. 1 shows a 1-98 engine undergoing test. Engines are identified
by the first digit indicating the number of cylinders and the remaining
digits the piston displacement. About 30 of the 1-98 engines have
been built used principally as test beds for various experiments. The
original rating was 10 HP but later versions were rated at 20 HP.
Maximum speed is 3,000 RPM.
Fig. 2 illustrates another engine concept namely an opposed piston type
which can be used for ultra-silent underfloor boat propulsion. A 97%
efficient wormgear reduction and coupling is built into the crankcase.
Its rating is 100 - 200 HP at 3,000 RPM.
During 1968, it was decided to design a four cylinder in-line engine which
could be used for vehicle as well as stationary applications, with a
rating or 100 - 200 HP. Another criteria was that it should be capable
of being mounted vertically or horizontally. Fig. 3 shows this engine.
The total experience on modern Stirling engines during the past ten years
has been achieved on approximately 50 engines of seven different types.
Powers range from a few tens of watts to 400 HP and the total running hours
amount to 25,000. Component testing has also been a feature of the
development programme and close to 1,000,000 hours have been logged over
the ten year period. Engines have been endurance tested for 5,000 hours
-------
V-4
Figure 1. 1-98 Engine during test,
-------
Figure 2. Four cylinder opposed piston engine.
-------
y-6
Figure 3. Philips 4-235 engine.
-------
V-7
at full load and vital running components for over 10,000 hours.
We have a total of 30 dynamometer test stands. Fig. 4 shows three of
them which are used to test the 4-235 engine.
During the past three years, a policy has been formulated which high-
ligh£s the development of engines for vehicle applications as being of
prime importance for these vehicle applications.
Emissions
Fig. 5 illustrates the external combustion system which is the basis
of the modern Stirling engine. The power piston and the displacer
are enclosed by the heater head, a composite unit comprising the
burner, heater tubes, preheater and regenerator. The system has some
excellent features.
The combustion is continuous and takes place in a space all surrounded by
hot walls of 700°C (1,300°F), so there is complete absence of quenching
effects, and combustion will be most complete. This is also accomplished
by a very good mixing in the burner itself. Furthermore, the amount of
excees air can be chosen freely, such that the design has a considerable
measure of flexibility.
Due to the pre-heating of the incoming air the flame temperature is
rather high; nevertheless, the amount of NOX formed is very low. This
is because local hot spots are avoided, and because the residence time in
the burner is very short (5 to 10 milliseconds).
-------
1
00
Figure 4. View of test cells with 4-235 engines installed,
-------
V-9
Figure 5. Heater head with piston and displacer.
STIRLING EXHAUST EMISSION (GVM)
HC
CO
NO-
30 V. REC.
1968-71 1972-75
CYCLE CYCLE
.02
1.0
.8
.16
.03
1.4
1.0
.20
1970-71
STANDARDS
2.2
23.0
1975-76 APSV
STANDARDS PROGRAM
.46 .14
4.8 620
.4-.6 .40
Figure 6. Table of emissions,
-------
V-10
Fig. 6 shows the results obtained on exhaust measurements. The
emissions on a GVM basis are calculated for the old and the new cycle test.
The standards and also the proposed standards are shown. It can be
seen that Cx, Hy and CO are extremely low. The NOX figures are valid
for a burner which is not specifically optimized for low NO , New
designs which use recirculation show measured values which lead to
numbers as low as 0.2 GVM. Recirculation offers no particular problem
because it only marginally effects engine output and efficiency.
It is also interesting to consider the comparison with the 13 mode
Californian emission cycle for heavy duty vehicles which is shown in
Fig. 7.
Other developments envisaged are for so-called suppressed flame temperature
burners which also result in extremely low NO emissions.
X
Noise and Vibrations
The comparison between a Diesel and Stirling engine of the same power
is shown in Fig. 8 and amounts to about 20 dB. The noise level from a
Stirling engine is almost inaudible at a distance of 50 metres.
Controls
Fuel control is by means of a thermostat system. Power/torque is
controlled by means of working gas pressure. A full load change in
either direction can be made within 0.3 seconds. A cold start takes less
than 20 seconds.
-------
V-ll
HEAVY DUTY VEHICLES (> 6000 IBS) Gm/BHR hr
Stirling
HC .04
NOX 2.0 (0.7)rec
CO 2.2
2.1 (0.74)rec
1973 _ 74 1975
Calif. Federal
12.5 5
40 25
Figure 7. Comparison with 13 mode California Emission Cycle.
dB
uo
130
120
110
wo
90
80
70
I
structure-borne noise
refit
sound pressure level
ref:J0002ubor
4 6 8 100 2
4 6 8 WOO 2 468 10000
FREQUENCY.cps
Figure 8. Noise level comparison.
-------
V-12
Braking
By varying the gas pressure in the working and buffer space it is
possible to obtain a braking effect up to 80% of full torque.
Overload
Engines are rated for continuous loads; short time overloading is
possible.
Multi-fuel
The single cylinder demonstration engine in Fig. 9 illustrates the
various fuels which can be used with a Stirling engine.
B.H.P./Efficiency/Torque
These characteristics are shown in Fig. 10 and 11 for various values
of working gas pressure. Note the advantageous torque increase with
decreasing speed compared with the Diesel engine.
The decision to concentrate on vehicle applications automatically poses
the question as to what precise vehicle applications are envisaged.
There is a distinction between professional or heavy duty vehicle propulsion
and light duty vehicle propulsion (passenger cars). Different require-
ments as to power-weight ratio and initial costs in relation to fuel
economy and maintenance and durability will hold. Stirling engines are
(will be) suited for both categories.
Heavy Duty Vehicle Application
The 4-235 engine shown in Fig. 12 will be mounted in a bus and is mounted
horizontally. It is initially rated at 100 HP with a capability of being
increased to 200 HP. The weight is 750 K.g. The illustration shows the
-------
Figure 9. Multi-fuel demonstration engine,
-------
V-14
OVERALL
EFFICIENCY
n
500
WOO
1500 2000
ENGINE SPEED n.rpm
t.BHP
Figure 10. Efficiency curves.
dynamometer
f torque
Jatm
250 SOO 79O
engine speed n.rpm
1000 1290 1500 1790 2OOO 2890 2100
Fieure 11. Toraue
-------
<
I
Figure 12. Philips 4-235 engine with gearbox.
-------
V-16
engine assembly with a Voith-DIWA type 502-3 gearbox. Ten of these
engines will be built for this and other applications.
Fig. 13 shows the bus with engine installed in the rear and also showing
the cooling system which necessitates an additional radiator to cater
for the increased heat load present in the Stirling cycle. Fig. 14
shows the actual bus.
It is hoped that a considerable amount of experience will be obtained
from the bus project which will provide a sound basis for future
development.
Light Duty Vehicle Application
The passenger car is undoubtably the application which commands most
attention and one of the main characteristics required is a low weight
power ratio. A different and attractive concept appears to be possible
which would bring the ratio down to as low as 2 - 3 pounds per HP.
The double acting principle is shown in Fig. 15. The arrangement is
that the hot space (expansion space) of one cylinder is connected to the
cold space (compression space) of the next cylinder with a heater, regener-
ator and cooler between. The double acting principle can be achieved by
more than one type of drive. One of the most promising of designs is
the swashplate engine shown in Fig. 16. Besides a volume and weight
advantage there is also a gain in simplicity of design and cost reduction.
Heat Pipe System
A particularly attractive heat source for the Stirling engine is obtained
-------
V-17
Figure 13. Installation of bus engine and cooling system.
-------
V-18
Figure 14. Bus powered by Philips 4-235 Stirling Engine.
-------
V-19
XPANSION SPACE
HEATER
REGENERATOR
COOLER
W—^COMPRESSION SPACE
Figure 15. Double acting Stirling cycle.
-------
Crosshead
Cihnder Compression Piston Rod / Slider Bearing
Space
Thrust
Bearing
Burner-Air Inlet
Oil Pumps
Cooler Tubes
Regenerator
Preheater
Connecting Ducts
i
M
o
Exhaust Outlet
Figure 16. Swashplate engine.
-------
V-21
by the application of heat pipes. A heat pipe is a closed volume in
which is incorporated a medium which can absorb and reject heat at a
very high rate; such a medium is sodium. By a process of evaporation
and condensing, the external heat system of the Stirling engine is
vastly improved. Fig. 17 shows a heat pipe system incorporated in a
swash plate engine. Two major advantages emanate from such a design.
(a) External and internal heat transfer can be considered
independently and the Stirling cycle can be optimized
to a still higher specific output. It is possible to use
higher speeds because heater flow losses are reduced.
(b) It is hoped that the combination with the combustor
can be made such that the combustion process and heat
transfer are partly simultaneous. The maximum flame
temperature can be reduced to 1,600°C where hardly any
NO formation occurs.
X
The development phase for such a promising combustion design has
already commenced. Engines utilizing the heat pipe system represent
engines which are commonly referred to as being of the "second generation."
Reviewing the proceeding engine types one endeavors to be realistic.
The hardware experience gained has been considerable; soon we will
acquire vehicle experience. However, the engine development stage is
only just beyond the laboratory concept. A considerable development
task particularly in the field of cost reduction and engineering lies
ahead before an economical acceptable engine is produced. A considerable
effort is also being made in the field of new fabrication methods. This
effort in advancing Stirling engine technology is almost entirely
carried out by Philips and its licensees in Germany and Sweden. Other
license agreements are being negotiated.
-------
y-22
Figure 17. Swashplate engine with heat pipe system,
-------
V-23
In Eindhoven the Stirling effort is divided into two main groups:
the Research Group Stirling Engines and the Product Group Stirling
Engines. The Research Group is primarily engaged in basic Stirling
engine research while the Product Group is engaged in engine develop-
ment for both vehicle and stationary applications.
-------
VI-1
Chapter VI
RANKINE-CYCLE POWER SYSTEM WITH ORGANIC-BASED FLUID AND
RECIPROCATING EXPANDER FOR LOW EMISSION AUTOMOTIVE PROPULSION
by
Dean T. Morgan
Special Products Staff
Thermo Electron Corporation
Waltham, Massachusetts, USA
-------
VI-2
1. BACKGROUND ON WORK LEADING TO AUTOMOTIVE SYSTEM
DESIGN
Thermo Electron Corporation has been involved in Rankine-cycle
system development since 1963; this effort has been concentrated on
completely self-contained portable powerplants in the horsepower
range of fraction to ~ 200 hp. During this period the systemhas
evolved from one of very limited practical application to one capable
of competing with internal combustion engines in a number of com-
mercial applications.
The Initial work on Rankine-cycle systems at TECO from 1963
to 1967 was concentrated on steam as a working fluid. It was realized
that steamhad some severe limitations for a low-cost, reliable system
capable of use at low ambient temperatures, but the development of
alternative working fluids had not advanced sufficiently for practical
consideration of their use. Since 1964, many new working fluid
candidates have been proposed, and sufficient thermodynamic and
physical property data have been developed for evaluating their
potential as working fluids for commercially oriented applications.
After extensive analysis of all of these fluids, thiophene was selected
in 1967 as the current state-of-the-art working fluid which best ful-
filled the system characteristics required for commercial Rankine-
cycle systems. Development of components and systems using this
fluid continued until September 1970, culminating in operation of a
complete, self-contained 5 hp package.
The primary limitation for the use of thiophene in systems to be
operated by the general public is its high flammability and toxicity...
In September 1969, TECO began laboratory testing of trifluoroethanol,
-------
VI-3
either pure or mixed with water, as a working fluid. The mixture of
*
85 mol% trifluoroethanol, 15 mol% water (Fluorinol-85) was selected
as the best composition. Fluorinol-85 has the r mo dynamic and physical
properties very similar to thiophene and, in addition, is completely
acceptable from a flammability and toxicity viewpoint for systems to
be operated by the general public. Since September 1970, testing of a
complete, self-contained 5 hp powerplant (see Figure 1) with Fluorinol-
85 working fluid has been carried out with satisfactory results. Both
thiophene and Fluorinol-85 have been operated in the same 5 hp package
(after careful flushing and cleaning when changing fluids) with prac-
tically identical power and efficiency.
Fluorinol-85, with its completely acceptable safety characteris-
tics, represents a significant advancement in Rankine-cycle technology
for automotive propulsion systems. Work on use of the TECO Rankine-
cycle system as an automotive propulsion powerplant was initiated in
June 1969 under contract to the Division of Motor Vehicle Research
and Development, Air Pollution Control Office, Environmental Pro-
tection Agency. During the first year of the program, a conceptual
design of the complete system was developed. This conceptual design,
described in Sections 2 and 3 of this report, indicated a strong potential
of the TECO approach for a practical and competitive low-emission
automotive propulsion system. The program is now in the second
year and involves conversion of the conceptual design to a more
detailed, optimized design and experimental development on several
components, specifically:
Halocarbon Products Corporation, Hackensack, N. J.
-------
1 . 3 lew Gen.eira.toir Set with Side Panels Removed .
-------
VI-5
a. Analysis and bench-testing of full-size expander intake
and exhaust valving approaches.
b. Detailed design, fabrication, and loop testing of a full-size
feedpump.
c. Fabrication of a boiler third stage section and measurement
of pressure drops and heat transfer rates.
d. Full-size rotary shaft seal testing.
In the remainder of this report, a summary description of the
component designs, system performance and characteristics, and
system packaging in an intermediate size American car are presented,
based on the APCO-sponsored studies. The component designs are
based on thiophene, since the details with Fluorinal-85 have not been
completed. The differences are generally small; where differences
occur with the Fluorinol-85, they are pointed out.
The system component sizes are based on the peak power require-
ments for an intermediate size American car to give a 0-60 mph
acceleration time of ~ 15 seconds or better and a top speed of 90-95 mph
Using these criteria, the system peak shaft horsepower is 100 shp. The
reference car size for the APCO alternative powerplant studies was
recently increased to a full-size American car. This modification
will require an increase in the system power output of approximately
20% to meet the performance goals.
2. COMPONENT DESCRIPTIONS
2. 1 System Working Fluid and Cycle Characteristics
The working fluid to be used in the system is Fluorinol-85, a:
mixture of trifluorethanol and water containing 85 mol% trifluoroethanol
-------
VI-6
and 15 mol% water. As discussed above, sufficient experience
has been obtained with this fluid so it can be considered a state-of-
the-art working fluid. In Table 1, a summary is given comparing
important cycle parameters for different trifluoroethanol-water
mixtures; the equivalent calculation is also presented for thiophene,
since this working fluid was used as the basis for the component
designs given in this report. Based on these numbers, Fluorinol-85
was selected as the optimum composition, and is being used in testing
the 5 hp system. This composition also has the minimum freezing
point of -82 °F; in Figure 2, the effect of water content on the freezing
point is illustrated. The boiler outlet temperature and pressure are
selected as 550°F and 700 psia, respectively, to maximize efficiency
and minimize the engine displacement for a given power output. This
cycle is illustrated on the pressure-enthalpy diagram of Figure 3. It
should be noted that trifluoroethanol appears to be more stable ther-
mally than thiophene, so that it maybe possible to increase the boiler
outlet temperature to 600 °F during the program with a resulting im-
provement in cycle efficiency. While Fluorinol-85 appears the opti-
mum composition at the present, future testing may indicate an alter-
nate composition is optimum.
In Table 2, the overall design point cycle characteristics for the
system are presented. The design point is based on a peak power
requirement of 100 horsepower at an engine speed of 2400 rpm. For
comparison, the design point characteristics are presented in Table 3
for thiophene, the working fluid used in the conceptual design study.
The component designs presented in Section 2 as well as the overall
system performance predictions of Section 3 are based on thiophene,
-------
VI-7
TABLE 1
CALCULATED CYCLE PARAMETERS FOR SEVERAL FLUORINOLS
AND COMPARISON WITH THIOPHENE
Release Pressure = 75 psia
Condensing Temperature = 200 °F
Subcooling T =A20°F
Expander Thermal Efficiency, rj.
ExPth
Expander Mechanical Efficiency, n
Exp,
M
Regenerator Effectiveness, r)
Pump Overall Efficiency, rj
Reg
OA
0. 8
0. 9
0. 9
0. 7
Fluid
Thiophene
Fluorinol
100
Fluorinol
85
Fluorinol
61
Fluorinol
51
Boiler
Outlet
Temperature,
°F
550
550
550
600
550
600
550
600
Boiler
Outlet
Pressure,
psia
500
400
50C
600
400
500
600
700
700
500
600
700
800
700
500
600
700
800
800
)7
Cycle,
%
1!'.. 7
K-.2
1!>. '
Ih, 0
If-. 8
If,. 2
:,(.. '••
It'.. 6
: ; ! , F>
ir . <'
I'''. !-'
Ib. v
If.. P.
1 7 „ 0
15. r
If. 7
1!'. 7
I1 . >>
1".0
QR/QB*
0.21
0.31
o.::9
0. 26
0. 24
0. 20
0. J9
0, 175
0. 2,2
0. M-
0 . Ji 3
0. 1 .1
0. 10
0. 16
0. 10
0. 087
0. 075
0. 061
0. 11
Expander CID,
in3/np
at 2000 rpm
1. 70
1. 84
1. 71
1, 58
1. 52
1.40
1.37
1,34
1. 23
1.56
1. 50
1.42
1.38
1.42
1.47
1.40
1.31
1.26
1.25
WE, Shaft
less
Wp, Shaft
Btu/lb
32.9
26, 5
23. 6
30.4 i
34. 8
36.3
37. 0
37.2 !
43. 1
44.4 |
4"-. 8
45.9
45.6
50, 2 ;
51.2 !
52. 9 i
52. 9
53. 3
58. 7
I
Regenerator heat transfer rate/Boiler heat transfer rate.
-------
VI-8
O
o
42
.£
o
0.
N
0)
4>
0
-10
-20
•30
•40
•50
•60
7O
\
s
\
X—
f
"• -.
^s,
-------
VI-9
BO 200 ZEO 40 240 ISO BOO SZO MO 39Q S*
•O SO
Figure 3. Pressure-Enthalpy Diagram and Cycle Conditions
for Fluorinol-85.
-------
VI-10
TABLE 2
DESIGN POINT SPECIFICATIONS, FLUORINOL-85
Working Fluid
Boiler Outlet Temperature
Boiler Outlet Pressure
Boiler Heat Transfer Rate
Boiler Efficiency (HHV)
Expander Design Point Intake Ratio
Expander Displacement
Expander Speed
Expander Piston Speed
Exp. Horsepower less Feedpump Power
Expander Thermal Efficiency
Expander Mechanical Efficiency
Expander Overall Efficiency
Regenerator Effectiveness
Regenerator Heat Transfer Rate
Condensing Temperature
Condensing Pressure
Subcooled Liquid Temperature
Condenser Heat Transfer Rate
Working Fluid Mass Flow Rate
Organic Volumetric Flow Rate
Feedpump Overall Efficiency
Feedpump Power
Cycle Efficiency
Overall Efficiency
Fluorinol-85
550°F
700 psia
1, 600, 000 Btu/hr
82.5%
0. 137
107 in3
2400 rpm
1000 ft/min
103.2 hp
84. 6%
91.5%
77.5%
90%
276, 000 Btu/hr
217°F
43 psia
197°F
1, 312, 000 Btu/hr
7330
11.70 gallons/minute
70. 0%
6.4 hp
16.4%
13.5%
-------
VI-11
TABLE 3
DESIGN POINT SPECIFICATIONS
THIOPHENE WORKING FLUID
Working Fluid
Boiler Outlet Temperature
Boiler Outlet Pressure
Boiler Heat Transfer Rate
Boiler Efficiency (HHV)
Expander Design Point Intake Ratio
Expander Displacement
Expander Speed
Expander Piston Speed
Expander HP less Feedpump Power
Expander Thermal Efficiency
-Expander Mechanical Efficiency
Expander Overall Efficiency
Expander IMEP
Regenerator Effectiveness
Regenerator Heat Transfer Rate
Condensing Temperature
Condensing Pressure
Subcooled Liquid Temperature
Condenser Heat Transfer Rate
Organic Mass Flow Rate
Organic Volumetric Flow Rate
Feedpump Overall Efficiency
Feedpump Power
Cycle Efficiency
Overall Efficiency
Thiophene
550°F
500 psia
J.58 x 106 Btu/hr
82. 5%
0. 137
184 in3
2000 rpm
1000 ft/min
K*7 2 hp
8-, 6%
c l , 5%
7'\ 5%
127,4 p s i
90.0%
0. 249 x 106 Btu/hr
216. 2°F
i:.r. 0 psia
%. 2°F
*. 25 x 106 Btu/hr
7377 pounds/hr
' "• I gallons/min
5><->. 7%
5.25 hp
\(,. 7%
-------
VI-12
since the details with Fluorinol-85 have not been completed. The
differences are generally small; where differences occur with the
Fluorinol-85, they are pointed out.
2. 2 Expander Design
Establishing the dimensions of a new expander design depends,
among other things, on the prediction of the indicated and mechanical
efficiencies of the expander at the design condition. Consequently, a
detailed analysis was carried out with thiophene to determine these
efficiencies as functions of piston speed and load, using measured
efficiencies obtained with the 5 hp expander on test at Thermo Electron
as a check. The performance with Fluorinol-85 should be almost
identical to the thiophene results. As a result of the analysis, it is pos-
sible to plot expander efficiency versus piston speed at various
loads. One such plot is shown in Figure 4 for an indicated mean
effective pressure (IMEP) of 125 psi. The rapid drop in efficiency
with piston speeds above 1000 ft/min occurs due to inlet valve losses.
From this analysis, a piston speed of 1000 ft/min was selected
for the design condition of 103 bhp at a vehicle speed of 95 mph. The
reduction in expander size which could be realized by selecting a
higher piston speed would probably be more than lost in boiler and
condenser size increases due to lower overall cycle efficiency.
The IMEP and BMEP are determined by the cycle design condition,
and the BMEP and the piston speed determine the piston area required
to develop the desired horsepower. A 90 °V of four cylinders was
selected as being reasonably compact without either an excessive
number of moving parts or excessive torque variation. The V design
-------
VI-13
90
80
o:
£70
o
60
50
40
£)ual Inlet Valve
Single Inlet Valve
Release Pressure - 75psia
Pi - SOOpsia
Ti = 55O°F
200 600 .1000 1400 1800 2200
Speed, ft/min
Figure 4. Overall Expander Efficiency Variation
with Piston Speed
-------
VI-14
results in a short engine for a given number of cylinders and facilitates
packaging of the system. With four cylinders, the resulting bore with
Fluorinol-85 is 3.68 inches. The mean piston speed (1000 ft/min at
design) and the engine speed are related by the expression
S = 2 LN
where S = mean piston speed, L = stroke, and N = rpm. The selected
design point speed of 2400 rpm, based on a reasonable bo re-to-stroke
ratio (1.47) and on valve train dynamics, results in a stroke of 2. 5
inches. The basic expander dimensions and specifications are given
in Table 4, with the values previously used for thiophene given for
comparison; cross-sectional views of the expander'with dimensions
given for thiophene are presented in Figures 5 and 6 with hydraulic
expander valving, slipping-clutch transmission, and feedpump
incorporated. The overall expander dimensions will be reduced
somewhat with Fluorinol-85.
All materials are identical to those now used in automotive
internal combustion engines. The Rankine-cycle expander differs
from the current automotive internal combustion engine in two very
important aspects of its design: the inlet valving and the bearing
design.
2.2.1 Variable Cut-Off Inlet Valving
The importance of having a valving system with variable cut-off
is established in Section 3. Apart from the difficulty in varying the
cut-off, the valving problem is considerably more severe than in
internal combustion engines. To avoid excessive losses, the high
-------
VI-15
TABLE 4
EXPANDER DIMENSIONS AND SPECIFICATIONS
Working Fluid
Thiophene
Fluorinol-85
Configuration
Bore
Stroke
Displacement
BMP
(feedpump work deducted)
IMEP at Design
BMEP at Design
Four Cylinders, 90 °V
4. 42 inches
3. 0 inches
184 in3
103 at 2000 rpm
127 psi
117 psi
Four Cylinders 90 °V
3.68 inches
2. 50 inches
107
103 at 2400 rpm
175 psi
160 psi
-------
VI-16
0
-J
3
Pigure 5. V-4 Expander with Hydraulically
Actuated Valves, Front View.
-------
H
I
©'
Figure"6. V-4 Expander with Hydraulically Actuated Valves, Side View.
-------
VI-18
density of the vapor at expander inlet conditions necessitates an inlet
valve comparable in diameter and lift (and therefore mass) to the
intake valve of an internal combustion engine. However, the valve
event is much shorter in the Rankine expander than in the internal
combustion engine. The design point intake ratio of 13. 7% corres-
ponds to a maximum valve event of 60°, whereas in internal com-
bustion engines the inlet valve event is on the order of 240° or more.
In a cam-operated system at a given speed, the acceleration and2
therefore, the stress level are proportional to the lift divided by
the square of the valve event; the cam stresses are much higher
at a given expander speed for the vapor expander.
One way of overcoming the problem is shown in Figure 7. In
this system two concentric inlet valves in series are driven by two
separate camshafts. Cam number 1, driving inlet valve 1, has fixed
timing with respect to the crankshaft. Cam number 2 has variable
i
timing with respect to the crankshaft, and the total valve event' is
determined by the overlap of the two valves. In this way, relatively
long cam events can be used, giving reasonable sized camshafts.
Figure 7 also shows that the mean valve opening area can be higher
with this approach than with a single valve, which should compensate
for the lower flow coefficient of the two valve system. A modifica-
tion of this concept is being set up for bench testing at TECO.
Other approaches to variable cut-off inlet valving are shown in
Figures 8 and 9- These are both hydraulic devices. A directly-
actuated hydraulic system is shown in Figure 8: A cam-operated
plunger pump operates a hydraulic column which acts on a stepped
piston on the inlet valve stem. The pump plunger is constructed with
-------
VI-19
7TZJ.C.
CRANK ANGLE
Figure 1. Two Inlet Valves in Series.
-------
VI-20
LOW PRESSURE
Figure 8. Directly Actuated Hydraulic Valve.
-------
LOW
PRESSURE
360"
OF"
Figure 9. pilot Operated Hydraulic Valve.
-------
VI-22
a helical undercut so that its angular position in its bore determines
its effective stroke, thus varying inlet valve duration. This system
is quite similar to diesel engine injection systems but has the, disad-
vantage of a relatively large power requirement.
Another hydraulic scheme is shown schematically in Figure 9.
In this system, a pump supplies high pressure oil at 700 1000 psi
to a rotary valve (shown as two valves for simplicity in Figure 9).
The rotary valve supplies the high pressure oil to alternate sides of
a piston connected to the inlet valve. Cut-off adjustment is obtained
by moving the rotary valve axially in its housing. Thermo Electron
Corporation has contracted with the British Internal Combustion
Engine Research Institute for fabrication and bench testing of a full-
size valving system based on this concept. A subcontract to American
Bosch Corporation has also been arranged under Phase II of the APCO
program at TECO for bench testing of a similar concept which uses a
solenoid-operated pilot valve in place of a rotary valve. Both of these
approaches have a low power requirement of about one horsepower for
a 4 cylinder expander.
It is thus expected that bench test results from three different
approaches will be available for selection of the expander intake
valving system to be used in the system. Phase II of the APCO
program also involves bench-testing of a full-size exhaust valve
scaled up from the exhaust valve used on the TECO 5 hp expander.
2. 2. 2 Bearing Design and Selection
The expander bearing design is strongly'influenced by the type
of transmission used* If the expander is coupled directly to the
-------
VI-23
driveshaft, as in many early steam cars, journal bearings relying
on hydrodynamic lubrication cannot be used; roller or ball bearings
must be used, because of the high bearing loads which could occur
at essentially zero rpm. On the other hand, if a conventional torque
converter were used, bearing sizes, at least on the crankshaft, can
be comparable to an internal combustion engine of the same bore,
since the peak cylinder pressures are roughly the same and the
inertia loading is lower on the Rankine expander (because of the
1000 ft/min limit on piston speed). The wrist pin bearing is more
heavily loaded than in the conventional four-stroke internal combustion
engine bebause the load on the pin never reverses; in this respect,
it is much like the wrist pin in a two-stroke engine. Any single
speed transmission can load the expander bearings fairly heavily
when the expander is idling at 360 rpm and the clutch is engaged.
At these conditions, with a maximum intake ratio of 80%, the bearing
loading is such that bearing sizes associated with a two-stroke diesel
of the same bore would be barely adequate for the Rankine expander.
A transmission with two forward speeds as well as a reduction in
the maximum intake ratio, as planned for the system, will alleviate
this situation.
It is expected that, with development, conventional journal
bearings can be used in the expander if the transmission used
permits the expander to idle at zero vehicle speed. However, to
minimize the development problems and the consequences of
momentary lubrication failure on the system, needle bearings
will be used throughout the expander for the first prototypes.
Needle bearings have been used in the 5 hp expanders tested
at Thermo Electron Corporation with excellent reliability.
-------
VI-24
2. 3 Feedpump Design
A full-size feedpump sized for the thiophene is under loop-test
at TECO as part of the Phase II of the APCO program. This pump
was designed and constructed at TECO. The primary factors con-
sidered in the selection and design of the vapor-generator ieedpump
are that the pump must be positive displacement because of the high
discharge pressure. The lubricity of either Fluorinol-85 or thiophene
is relatively poor and its liquid viscosity low. The pumping rate must
be variable from basically zero to 15 gpm for thiophene over a 800 -
2000 rpm range, and the pump must operate with low net positive
suction head (NPSH) without cavitation, since the NPSH is provided
primarily by subcooling of the liquid coming from the condenser.
The feedpump selected, illustrated in Figure 10, is a 5-cylinder
piston pump driven by a wobble plate; its characteristics are summar-
ized in Table 5. The selection of a piston pump was based on testing
of several types of positive displacement pumps, including gear and
vane-type pumps, at Thermo Electron. In general, high leakage
rates (low volumetric efficiency) and high wear rates have been
encountered for all pump types other than piston, which have given
completely satisfactory performance. With the piston pump all
bearing surfaces can be oil-lubricated.
The variable pump rate is obtained by incorporation of variable
displacement in the feedpump, permitting the pumping rate to be con-
trolled at the desired rate regardless of feedpump speed. The method
used to obtain variable displacement is to vary the effective displace-
ment of the feedpump by bypassing all or part of the pump output to
the condenser. The effective displacement is obtained by moving the
-------
BY-PASS
H
I
N5
INTAKE
Figure 10. Feedpump Cross Section.
-------
VI-26
TABLE 5
FEEDPUMP CHARACTERISTICS
Working Fluid
Pumping Rate
Number of Cylinders
Volumetric Efficiency
Overall Efficiency
Range for Maximum Pumping
Rate
Total Displacement
Bore
Stroke
Materials of Construction
Housing
Pistons and Valves
Bearings
Thiophene
15 gpm
5
90%
80%
800-2000 rpm
5. 51 in3
1. 875 in
0. 4 in
Cast Iron
Hardened Steel
Needle
Fluorinol-85
11.7 gpm
5
90%
80%
960 -2400 rpm
3,58 in3
1.51 in
0. 4 in
Cast Iron
Hardened Steel
Needle
-------
VI-27
entire cylinder block to regulate the axial position at which the bypass
port is uncovered during the discharge stroke of each of the five pistons
The bypass porting is configured so that pumping to the high pressure
exhaust occurs always from bottom dead center; this procedure mini-
mizes the discharge pressure transients and noise level of the pump.
The bypass flow is also never pumped to the high pressure, thereby
minimizing the feedpump parasitic load on the system. The control
lever for positioning the cylinder block uses a bellows rather than
a sliding seal to maintain the hermetic nature of the system.
Spring-loaded poppet suction and discharge valves are used.
The suction valve is constructed in the cylinder and is made as
large as possible to minimize the pressure loss through the valve
and the tendency for cavitation., The smaller discharge valve is
located in the cylinder head. Common suction and discharge plenums
for all five cylinders are incorporated in the housing castings. The
use of five cylinders was based on reducing pressure transients due
to the flow variation from the piston pump. These transients must
be maintained sufficiently small on the suction side of the pump that
the liquid pressure never falls below the vapor pressure of the sub-
cooled liquid. A computer analysis of the pressure transient behavior
indicated that a five-cylinder pump would be required to prevent
cavitation with 20 °F subcooling at the pump suction.
The pump drive could be either crank or wobble-plate. The
wobble-plate drive was selected because of its compactness and
easier packaging with the expander, its lower weight and vibration,
its quieter operation at higher speeds, and its more convenient
geometry for variable displacement incorporation.
-------
VI-28
2. 4 Burner-Boiler Design
The burner-boiler design has not been worked out in detail for
Fluorinol-85. In Figures 11, 12 and 13 the boiler tube bundle design
and the burner cross sections are presented, based on detailed com-
puter analysis for the thiophene fluid. The design requirements are
summarized in Table 6. Some modification will be necessary for
Fluorinol-85, due to the larger superheat requirement for Fluorinol-
85 relative to thiophene. The changes will be restricted to modifica-
tions of the tube bundle in each stage, with no change in the overall
burner-boiler envelope or in the flow paths in the boiler. The factors
considered in arriving at the boiler design, in addition to heat transfer
performance, were low materials cost, low volume, easy construction,
and low combustion side pressure drop. Maintaining a low combustion
side pressure drop is of extreme importance in meeting the develop-
ment goals of fast startup and low parasitic power loss. For startup,
the complete combustion system must be electrically driven by the
battery. To achieve fast startup, it is essential that the burner be
operated at its maximum rate. The total peak power requirement
must therefore be maintained at as low a level as possible, consistent
with the packaging envelope and heat transfer rating required. The
current design at full burning rate requires about 1 1/4 hp total to
operate the combustion air blower and fuel nozzle compressor. This
power level is practical considering the battery, alternator and motor
sizes required.
With reference to Figure 11, the combustion gases, at a tempera-
ture of ~ 3300°F, flow from the combustion chamber into the center of
the tube bundle and radially outward through the tube bundle. The flow
-------
VI-29
23 IN
Figure 11. Cross Section Through Burner-Boiler, Short Axis.
-------
VI-30
L
Figure 12. Top View of Boiler Tua3 Bundle.
-------
VI-31
Figure 13. Cross Sections Through Automotive-Size Burner.
-------
VI-32
TABLE 6
BURNER-BOILER DESIGN REQUIREMENTS
Reference Cycle Boiler Heat /
Transfer Rate 1.60 x 10 Btu/hr
Maximum Boiler Heat Transfer
Rate 1. 70 x 106 Btu/hr
Burner Design Maximum Heat ,
Release Rate (HHV) 2.06 x 10 Btu/hr
Boiler Design Efficiency (HHV) 82.5%
Turndown Ratio 15/1
-------
VI-33
path of the organic through the tube bundle is illustrated in Figure 14.
The organic first flows through stage 3, from which the combustion
gases are exhausted; this provides the lowest organic temperatures
in the boiler at the combustion gas outlet and a high boiler efficiency
without air preheat. It is important that an extremely compact and
efficient heat transfer surface be used in this stage to maximize the
boiler efficiency with acceptable pressure drop on the combustion
side. The organic next flows through the inner stage (through which
the combustion gases first flow), with a resultant high heat transfer
coefficient. Because of the high gas temperature and extended surface
on the combustion side, coupled with the high heat transfer coefficient
on the organic side, a very high heat transfer rate can be obtained in
the first stage. The organic next flows through the superheater coil
or stage 2. This stage is a bare tube coil, since the controlling
thermal resistance is on the organic side and an extended heat transfer
surface is not required on the tube.
The characteristics of the three boiler stages are given in Table
7; Figure 15 presents the calculated design point temperature and
pressure profiles through the boiler. In the last or third stage, a
matrix made of steel and copper balls brazed together and to the tube
is used. This type of extended surface provides a very high heat trans-
fer rate per unit volume and is amenable to mass production techniques.
Thermo Electron Corporation is currently making heat transfer and
pressure drop measurements on a full-size section of the boiler third
stage as part of Phase II of the APCO program. Other types of
extended surface for the boiler third stage will also be evaluated ex-
perimentally.
-------
VI-34
ORGANIC FLOW TO ENGINE
ORGANIC FLOW
TO BOILER "
STAGE NO. 3
COMBUSTION
GAS
FLOW
Figure 14. Organic Flow Path through Boiler Tube Bundle
-------
TABLE 7
BOILER DESIGN SPECIFICATIONS
Stage
No.
1
2
3
Total
^
Woo f
Transfer
Rate
Btu/hr
1. 05 x fo6
0. 33 x 106
0.36 x 106
Combustion Gas Temp.
°F
entering
3330
1670
] 114
leaving
1670
1114
467
1 1
]. 74 x 10°
—
—
i
Tubing
Length
ft.
17
35
27
79
Pressure Drop
Combustion
Side in w. c.
0. 17
0. 33
2. 26
2. 76
Organic Side
psi
—
—
__
46
be Spe c if i cations
Inner Tube ID
OD
Outer Tube ID
OD
0.930"
1. 000''
1. 125"
1.315"
F_in_ Specifications
Fins/Inch 10
Fin Thickness 0. 012"
Fin Material Copper
Fin Height 0. 356"
<
I—I
I
Ln
Matrix Specifications
Ball Size 3/32"
Ball Material 50% Carbon Steel, 50% Copper
Matrix Thickness 0.5"
Matrix Height 0.970"
(between tubes)
-------
VI-36
650
600 -
550 ^
5OO -
i- CL
3 -
Q. \fi
E £
o> a.
450
400 -
350
300 -
Tg, = 3330
1895.9
Tg3= 1190.0
Tg,, = 490.24
20 30 40 50
Tubing Lengthfrom Inlet. Feet
60 70
80
Figure 15. Design Point Boiler Temperature Profile
-------
VI-37
An intermediate heat transfer fluid (water) is used, with double
tube construction in the boiler tubes, to positively prohibit hot spots
on the organic side of the boiler. Stages 1 and 2 are connected to the
same water reservoir, which represents the high pressure side of
the boiler. The water side of stage 3 is separate and represents the
low pressure water side of the boiler.
The combustion chamber design, illustrated in Figure 13, is
based on a volumetric burning rate of 2. 8 x 10 Btu/hr/ft and is
scaled from the burner used in the 5 hp system currently under test.
The burner is constructed integral with the boiler tube bundle, as
illustrated in Figure 11. To reduce the pressure drop within the
diametrical clearance available, two identical burners operating
in parallel are used rather than one longer burner with the same
combustion chamber diameter. The pressure drop at maximum
firing rate for the two-burner setup is 1. 5" w. c.
While no pollution measurements are available on the full-scale
burner design illustrated, Thermo Electron Corporation has completed
measurements on a smaller scale burner (140,000 Btu/hr) with per-
formance characteristics similar to the burner illustrated in the
design. This burner is being used on a 5 hp system now on test at
Thermo Electron Corporation. Figures 16 and 17 present the steady
state emission levels from this burner as a function of excess air
for burning rates of 105,000 Btu/hr and 50,000 Btu/hr, respectively.
It is apparent that the emission levels are extremely low. To indicate
the transient performance of the burner, the burner was oscillated
between 50, 000 and 105, 000 Btu/hr burning rates with constant fuel-to-
air ratios maintained; CO and unburned hydrocarbon emission levels
-------
VI-38
a.
a.
C/5
z
o
(/)
UJ
200-
180-
160-
140-
120-
100-
80 -
60 -
40 -
20 -
STEADY
Q=50,OOC
FUEL JP-<
*^ *- — ^^i^
• • • •
°~' I I ! I I I
0 10 20 30 40 50 60
NO
CO
CH
EXCESS AIR (%)
Figure 16. Effect of Excess Air on Emissions, 50,000 Btu/hr
-------
200-
180-
160-
140-
120-
100-
80-
60-
40-
20-
0-
I
0
STEADY STATE DATA
Q=105,000 BTU/HR
FUEL JP-4
'X NO
i
10
I
20
30 40
EXCESS AIR (%)
I
50
i
60
OJ
VD
Figure 17. Effect of Excess Air on Emissions, 105,000 Btu/hr
-------
VI-40
were monitored continuously while a bag sample was collected through-
out the run for NO measurement at the end. The results are indicated
in Table 8. As indicated in Section 3, use of these emission concentra-
tions with the system performance gives gm/mile emission levels sig-
nificantly less than projected limits under the recently enacted Clean
Air Legislation in the U. S.
2. 5 Condenser Design
The condenser fan power represents the largest parasitic load of
the system. A very important goal in the design of the component is
to minimize the required fan power. The factors considered in the
condenser design, as well as its integration and packaging with the
system, were:
a. An efficient (low f/j ratio) extended surface should be used
on the air side of the exchanger.
b. The organic side should be circuited to provide high vapor
velocities leading to high organic side heat transfer
coefficients.
c. The frontal area of the condenser (as well as the air free
flow area) should be made as large as possible.
d. Highly efficient condenser fans should be used.
e. The system packaging and vehicle modifications should
be designed to minimize grille (intake) and engine com-
partment (exhaust) air pressure losses.
-------
VI-41
TABLE 8
TRANSIENT EMISSION DATA
FIRING
RATE
(BTU/HR)
105,000
50, 000
50, 000
50,000
50, 000
105, 000
105, 000
105,000
50,000
50,000
50,000
50, 000
105, 000
105,000
105, 000
50, 000
50, 000
50, 000
EXCESS
AIR
(%)
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
CH
X
(PPM)
6
-
5
-
4. 5
401
7
5
5
4
4. 5
-
151
-
6
-
4. 5
4
CO
(PPM)
60
-
30
25
70
1801
90
-
80
25
20
-
1000+1
-
75
75
30
15
NO
X
(PPM)
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
422
392
ELAPSED
TIME
(MIN)
0
. ?
1. 0
1. 5
2. 0
2. 5
3. 0
3. 5
4. 0
4. 5
5. 0
5. 5
6. 0
6. 5
7. 0
7 5
8. 0
9. 0
NOTES:
i
2
Short duration peak ^,10 sec.
Exhaust gas sample collected during
the 9 minute run. Then two samples
were drawn and NO analysis performed.
-------
VI-42
f. The condenser fan control should be optimized to minimize
the fan power requirement by reducing the fan power when
not required. The condenser fans are directly driven by
the propulsion engine.
The condenser core of the conceptual design is illustrated in
Figure 18 with the design point characteristics, as obtained with
the condenser computer program, given in Table 9.
The condenser core is similar to a Ford radiator with louvered
fins, except that the flattened tubes have a heavier wall (0.030" vs
0.005") and are constructed in one integral piece with partitions used
to provide the desired vapor-side flow path. Copper fins with 2. 5 mil
thickness are used, and the tubing is made of carbon steel rather
than brass as in the Ford radiator. The frontal area used in the design
represents the maximum practical area for an intermediate size
American car with some rework of the front end frame and grille.
It is expected that improvements can be made to the condenser
design illustrated here. Use of a strip fin rather than the louvered
fin results in an improved f/j ratio and will reduce the fan power
required for a given condenser physical size. If the fluid compati-
bility properties are satisfactory, use of an all-aluminum condenser
will significantly reduce the condenser weight and will permit easier
construction.
2. 6 Regenerator Design
The regenerator design conditions are illustrated in Figure 19
and the regenerator design in Figure 20. While the design is based
on thiophene, the requirements are very similar to those for
-------
3 0"
— 501L
H
I ==
ZO. 6'
Figure 18. Condenser Design.
-------
VI-44
TABLE 9
CONDENSER DESIGN POINT CHARACTERISTICS
Heat Rejection Rate
(20 °F Superheat Entering ,
20 °F Subcooling Leaving) 1.25 x 10 Btu/hr
Length 50 in.
Height 20. 6 in.
Depth 3.0 in.
Frontal Area 7. 15 ft
Condenser Inlet Pressure 25 psia
Design Ambient Air Temperature 95 °F
Air Pressure Drop 3.45 in. we
Ideal Fan Power 7. 91 hp
Air Flow Rate 63, 000 Ib/hr
-------
Vapor
25 psia, 348°F
(550-A P) psia, 285°F
1
Q = 249,000 Btu/hr
(25-A P) psia, 230°F
Liquid
550 psia, 199°F
H
I
Ln
Liquid Flow Rate = Vapor Flow Rate = 7377 Ibs/hr
Figure 19. Regenerator Design Point Requirements.
-------
-775-
<
H
I
Figure 20. Regenerator Design.
-------
VI-47
Fluorinol-85 and only minor design changes are expected. The design
is based on obtaining a compact regenerator with geometry suitable for
packaging directly above the expander in the expander compartment
and with low pressure drop on both the liquid and vapor sides. On the
vapor side, a brazed ball matrix extended surface with 1/16" ball
diameter is used. The exchanger is divided into four parallel liquid
circuits; in each circuit, the vapor passes through four separate
stages, permitting the exchanger to approach a pure counterflow
exchanger.
*
The regenerator is also designed to function as an oil separator.
The low vapor velocities coupled with the ball matrix should provide
very efficient separation of the lubricating oil droplets from the ex-
haust vapor from the expander. A drain line returns the collected oil
back to the expander crankcase.
2. 7 Automatic Transmission
An automatic transmission will be used which permits the main
propulsion expander to idle at zero vehicle speed and thereby drive
the system and vehicle accessories. The use of a transmission also
permits matching very closely the driving characteristics of current
automobiles powered by the I/C engine with three-speed automatic
transmission. The Dana Corporation, Toledo, Ohio, as part of Phase
II of the APCO program, is designing a two-speed slipping clutch
transmission based on the type of transmission they develop and
produce for off-the-road vehicles.
Patent Application Pending
-------
VI-48
2. 8 Controls
The startup and control system is designed to provide completely
automatic startup and operation of the system. The functions required
of the driver are identical to those now required in current, conventional
cars with automatic transmission. The controls are assembled from con-
ventional components with the exception of the burner fuel-air control,
which is currently under development at TECO.
The primary control problem in the system is control of the burning
rate and of the pumping rate to the monotube vapor generator to main-
tain boiler outlet pressure and the temperature within specified limits
over any type of transient encountered by the system. The control
system is based on operating the boiler as close to quasi-steady state
as possible over all transients; both burning rate and pumping rate are
maintained as closely as possible to the values corresponding to the
instantaneous vapor flow rate from the regenerator.
A schematic layout of the principal control components is presented
in Figure 21. The feedpump control, used to control the pumping rate to
maintain boiler outlet pressure, is simplified by the fact that the organic
flow rate at any expander rpm is approximately linear with intake ratio.
The feedpump and expander valving are thus directly operated as a unit
by the accelerator pedal; a vernier control working from the boiler outlet
pressure is used to reduce deviations from the design point and to eliminate
any unbalance in the system automatically. A mechanical governor is used
to limit the maximum intake ratio as a function of rpm and to govern
expander speed at idle.
In order to maintain the burning rate at a value corresponding to
the organic flow rate into the boiler, the burner control uses an orifice
in the organic line to sense almost instantaneously any changes in the
-------
Accumulator for
Constant Fuel
Pressure
Pressure By
pass Valve
-------
VI-50
organic flow rate. This signal, along with a similar signal from an
orifice in the fuel line, is used with a diaphragm controller to provide
the proper fuel and air flow rates. If necessary, the fuel-to-air
ratio can easily be varied as a function of turndown to minimize
pollutant emissions at any burning rate. A vernier control operating
from the boiler outlet temperature is used to reduce deviations from
the design point value; at low power levels, when the organic flow is
low, this temperature control becomes the primary control on the
burner. The fuel control valve cross section is presented in Figure 22
and the air control valve cross section is presented in Figure 23.
3. SYSTEM PERFORMANCE AND PACKAGING
3. 1 Performance
The important decisions •which have a strong influence on the
overall system performance and cost relative to the internal com-
bustion engine are the method of driving accessories at zero vehicle
speed and the choice between constant intake ratio expander valving
with throttle valve control and variable intake ratio expander valving.
With respect to the first decision, two alternatives are feasible. An
auxiliary5 constant speed expander can be used to drive all accessories,
permitting the expander to be coupled directly to the driveshaft (through
a simple geat transmission for forward, reverse, neutral, park control).
A more complex transmission can also be used, permitting the main
propulsion expander to idle at zero vehicl/fe speed so that all accessories
canbe drivendirectly by the main propulsion expander; this transmission
is still simpler than that required for the I/C engine-driven system,
-------
'-"-"••~\ \ \--~~
<3o; iff, ,25. \ -J5j
\ i29'. \ V2Z) A \ 27, \ fcl
\-T\T \ \ \ \ f
fir*)
,-««
\ (^V. ^)
\ \^-\ ^ \ ^
<
M
t
Figure 22. Fuel Control Valve Assembly,
-------
ft*
I
Ul
K>
- Up*.Kea?n
Figure 23. Combustion Air Control.
-------
VI-53
It has been the conclusion of this study that the use of the main
propulsion expander to drive all accessories is preferable. Even though
the accessories must be larger when driven by the variable-speed main
propulsion expander, the accessory designs are identical and the same
number of parts must be processed; very little cost differential exists
between the different sized components. Any cost reductions due to
smaller accessory components are more than counterbalanced by the
requirement for an additional expander of 1 5 20 hp to handle short term
accessory peak loads, with governor-throttle valve control. The system
with all accessories driven by the main propulsion expander, therefore,
seems preferable in terms of simplicity, cost, and packaging; this
system has been selected as the optimum approach.
With respect to expander valving, a detailed analysis with
computer modeling of the expander and boiler performance for all
operating conditions has been carried out, comparing a system with
constant intake ratio expander valving and throttle valve control and
a system with variable intake ratio expander valving. The results
are summarized in the performance maps presented in Figures 24
and 25 for constant IR and variable IR (IR = 0. 29) systems,
max
respectively. Comparing these two performance maps for the same
boiler and expander sizes, the following conclusions can be made:
a. The system with variable IR valving has a peak efficiency
of 18. 5% versus 15. 0% for the system with constant IR
valving. This 20-25% improvement in efficiency (or mpg)
occurs over a large power-speed region including the region
-------
Ln
0
0
200 400 600 800 1000 1200 1400 1600 I8OO 2OOO 2200
Engine RPM
Figure 24. Performance Map with 184 CID Expander and
Maximum Intake Ratio of 0. 29.
-------
VI-55
o>
*
o
I
no
100
80
60
40
20
0
Full Power, Variable IP
of O 8 Max
14%
0 200 400 600 800
1000 1200 1400
Engine RPM
1600 1800 2000 2200
Figure 25. Performance Map with 184 CID Expander
and Constant Intake Ratio of 0. 137.
-------
VI-56
of 20 - 40 hp and 600 - 1200 rpm, where the system would
operate most of the time.
b. The peak power for the equivalent-sized expander and boiler
is much greater over most of the expander speed range for
variable IR valving than for constant IR valving. Calculations
of 0 - 60 mph wide-open-throttle times indicate a 50% increase
for constant IR valving relative to variable IR valving.
For these two reasons, there exists an extremely strong incentive for
incorporation of variable intake ratio valving in the expander; this
type of valving is used in the reference expander design presented
earlier.
In Figure 25, the dotted line'to the left of the plot represents the
peak system power if a maximum intake ratio of 0.8 (the maximum
practical) rather than 0.29 were used. (IR) = 0.29 was selected
max
as optimum for the folio-wing reasons;
a. As evident from Figure 25, very little performance is lost
by decreasing (IR) from 0. 8 to 0. 29.
max
b. The required feedpump displacement is decreased by a
factor of 2. 7.
c. The low vehicle speed condenser rejection rate is reduced,
permitting greater utilization of ram air.
d. The wide-open-throttle system efficiency at expander speeds
to 800 rpm is increased.
With reference to Figure 25, the part-load performance of the
system is an extremely important consideration. Thus, while the
design point efficiency, defined by the peak power requirements
-------
VI-57
for acceleration, is 13 7%, the efficiency increases under part-load
conditions where an automobile normally operates The increase
with the variable IR valving occurs because part-load operation is
obtained by reducing the IR below the design point value of 0,137,
providing a more efficient expansion in the expander, reduction of the
condenser pressure under part-load operation also occurs, again
leading to a more efficient cycle. The region of high efficiency
(> 17%) is broad, that is, a high efficiency is obtained over a broad
range of engine power and speed. Thus, while the peak thermal
efficiency of the Rankine-cycle system (18.5%) is much less than
that of the I/C automotive engine (~30%), the average efficiencies
for typical consumer driving cycles are much closer, The Ford
Motor Company has calculated, using the performance map of Figure
25, with (IR) = 0. 8, the fuel economies for different driving
max
conditions of the Rankine-cycle system with single-speed slipping
clutch transmission installed in an intermediate size American car,
and compared them directly with the fuel economy calculated for the
same body with 302-2V engine and three-speed automatic transmission.
The results are summarized in Table 10 for both steady speed and
dynamic driving cycle operation. For the customer average driving
cycle, the mpg for the 302 CID I/C engine is 15.7, versus 12,7 mpg
for the Rankine-cycle system. For steady speed, zero grade operation,
the Rankine-cycle engine has better fuel economy than the I/C engine
up to 50 mph and poorer fuel economy above 50 mph. While calculations
have not yet been made, it is expected that use of a two-speed trans-
mission as planned will improve the city fuel economy significantly,
resulting in an improved customer average. This change, coupled
with a potential for increasing the peak cycle temperature to 600 °F,
-------
TABLE 10
VEHICLE ECONOMY PROJECTIONS
THERMO ELECTRON CORPORATION
RANKINE-CYCLE ENGINE
VEHICLE: Intermediate size American car, Wheelbase - 116 in.
TIRES: 7.75 x 14, Rolling Radius - 1. 08 ft. , Rev/Mile - 778
(IR) = 0.
max
Ford Production Engines
(1969 Model)
302-2V 3-Speed Automatic
250- IV 3-Speed Automatic
Rankine- Cycle Engines
182-CID Clutch Drive
Single Speed
182-CID Torque Converter,
1. 88 Ratio Speed- Up Gear
182-CID Torque Converter
2. 77 Ratio Speed-Up Gear
Idling
Speed
rpm
500
600
300
300
300
Fuel Flow
Ibs/hr
3. 75
2. 68
2. 00
2. 00
2. 00
City
mpg
13. 3
12. 9
9.6
10. 1
10. 3
Fuel Economy Computer Program PB1213
Suburban
mpg
18.0
19.6
15. 8
15.0
15.2
Customer
Average
mpg
15. 7
16.3
12. 7
12. 5
12. 8
Steady Speed mph
30
mpg
27.3
27.0
33. 1
32.4
32. 3
40
mpg
23. 3
25. 4
25. 1
24. 6
24. 8
50
mpg
20.4
22.9
20.4
19. 7
19. 5
60
mpg
18.0
20. 1
16. 7
16. 0
15. 7
70
mpg
16 2
I 7. 6
13. 4
13.0
12. 9
30-70 mph
mpg
21. 0
22. 6
21.7
21.2
21.0
I
Ui
CXI
-------
VI-59
could ultimately result in better fuel economy than the 1969 302-2V I/C
engine. It should also be noted that the fuel economy of the I/C engines
decreases year by year as additional pollution controls are required.
The acceleration performance of the vehicle with a given power-
plant is strongly dependent on the transmission used.
Performance characteristics of the Rankine-cycle system in an
intermediate size American car are presented in .Figures 26 through
29 and Tables 11 and 12 for a torque converter transmission and
single-speed clutch transmission with IR = 0. 8 and for single
max
and two-speed clutch transmissions with IR = 0. 29. These cal-
max
culations were prepared by the Ford Motor Company and the Dana
Corporation, using existing computer programs, for the Rankine -
cycle system with different types of transmissions and for the Ford
production 302-2V engine with three speed transmission. The basic
input to these calculations was the performance maps presented in
Figures 24 and 25.
The most important conclusions from these calculations are:
a. The Rankine-cycle system with 184 CID expander (thiophene
working fluid) should be capable of providing 0-60 mph
acceleration times of less than 15.0 seconds, taken as the
criterion for acceptable performance. The acceleration
performance is fairly dependent on the type of transmission
used. A maximum level grade vehicle speed of 90 100 mph
should be attainable, irrespective of type of transmission
used.
-------
VI-60
1800
1600
1400
1200
en
.0
LJ
0)
o
D
1000
800
600
400
200
0
TWO SPEED CLUTCH
2.5/1 Ratio to ISOOEngineRPM Downshift
I/ I After Downshift
2.0/1 Ratio to 1800 Engine RPM Downshift
I/I After Downshift
Single Speed Clutch
1
0 20
Figure 26.
40 60 80 100
Vehicle Speed, MPH
120
Comparison of Tractive Effort for Single and Two
Speed Clutch Transmission.
-------
VI-61
2200
2000
1800
1600
1400
UJ
_>
O
O
1200
1000
800
600
400
200
TWO SPEED CLUTCH
2. 5/1 Ratio to 1800 Engine rpm
downshift, 1/1 after downshift.
] I
,. Intermediate size American car, 302-2V
Engine, Three Speed Transmission
Two Speed
Clutch, 2. 5/1
Ratio to 1800 Engine
—rpm downshift, 1/1
after downshift.
3420 Eng. RPM
I
20 4O 60 80
Vehicle Speed, MPH
100
120
140
Figure 27- Comparisons of Tractive Effort for an Intermediate Size
American car, powered by 302-2V Engine and by Rankine
Cycle System with 134 CID Expander, IR = 0.29
max '
with Dana Two Speed Clutch Transmission.
-------
VT-6?.
CO
,0
H
u
nJ
*H
H
2200
2000
1800
1600
1400
1200
1000
800
600
400
200
Intermediate Size American Car, 302-2V
Engine, Three Speed Transmission
.0/1 Ratio to 1800 Engine Rpm
1/1 After Downshift
20 40 60 80 100
Vehicle Speed, Mph
120
140
Figure 28. Comparison of Tractive Effort for an Intermediate Size
American Car powered by 302-2V Engine and by Rankine
Cycle system with 184 CID Expander, IR = 0.29, wit.i
Dana two-speed clutch transmission.
-------
VI-63
2200
2000
1800
1600
v>
O
5t
UJ
0)
o
O
1400
1200
1000
800
600
400
200
-Intermediate Size American Car, 302-2V—
Engine, Three Soeed Transmission
^Maximum Torque Capacityof System
Rankine Cycle, 184 CID, IRm(^0.8
Single Speed Clutch Transmission
Rankine Cycle, l84CID,IRma=0.8
Toraue Converter Transmission
D Converter)
Low Speed Torque
Used in Performance
Calculations
3420 Eng.RPM
2000 Eng. RPM
I
I
0
Figure 29.
20 40 60 80 100
Vehicle Speed.MPH
120
140
Comparison of Tractive Effort for an Intermeliate Size
American Car Powered by 302-2V Engine and by Rankine-
Cycle System with 184 CID Expander, IRmax = 0. 8, with
Single Speed Clutch Transmission or with Torque Converter
Transmission.
-------
TABLE 11
VEHICLE PERFORMANCE PROJECTIONS
THERMO ELECTRON CORPORATION
RANKINE CYCLE ENGINE
VEHICLE: Intermediate Size American Car, Wheelbase - 116 in.
TIRES: 7.75 x 14, Rolling Radius - 1.08 ft. , Rev/Mile - 778
(IR) =0.
max
Ford Production Engines
302-2V 3-Speed Automatic
11-1/4 Dia. Converter
250- IV 3 -Speed Automatic
11-1/4 Dia. Converter
Rankine Cycle Engines
182-CID Clutch Drive, Single Speed
200- CID Clutch Drive, Single Speed
220 -CID Clutch Drive, Single Speed
240- CID Clutch Drive, Single Speed
182-CID 1. 88 Ratio Speed-Up Gear
12-5/16 Dia. Converter
182-CID 2. 77 Ratio Speed-Up Gear
12-5/16 Dia. Converter
Transmission
Gear Ratio
2.46, 1.46, 1.00
2.46, 1.46, 1.00
None
None
None
None
None
None
Axle
Ratio
2. 79
2.79
1.64
1.64
1.64
1.64
3.06
4. 50
N/V
36. 2
36.2
21. 3
21. 3
21. 3
21. 3
21. 3
21. 3-
Car
Weight
Ibs.
3539
3522
3539
3539
3539
3539
3539
3539
Performance - Computer Program PB1 1 1 1
0 -4 Sec.
ft.
89. 9
78. 2
70. 0
77. 2
86.0
94.4
74. 2
79. 8
0 - 10 Sec.
ft.
469.2
405. 7
407. 9
442.0
480.4
516. 7
412. 2
414. 6
0-60 mph
sec.
11. 9
15.6
14.2
12. 7
11.4
10. 3
14.4
14. 3
0 - 1/4 Mile
sec.
18. 8
20. 6
20. 3
19. 5
18. 7
18. 0
20. 1
20. 1
Passing at
50 mph
sec.
9.72
11. 59
10.30
9. 74
9.22
8. 79
10. 97
10. 85
<
H
I
-P-
-------
VI-65
TABLE 12
VEHICLE PERFORMANCE PROJECTIONS,
IR = 0.29
max
54 ft # Subtracted from Full Torque Curve at All Speeds for Accessories
Transmission
Single Speed
Two Speed
Two Speed
Two Speed
Two Speed
Two Speed
Gear
Ratio
1/1
2.5/1
1/1
2.5/1
1/1
z.oo/:
1/1
1. 75/1
1/1
1.50/1
1/1
Engine Downshift
Speed
-
1800
1800
1800
1800
1800
0-60 mph
Acceleration Time
Seconds
19.6
14.4
14.5
14.6
14.9
15.5
Gradability
17.0%
48. 9%
42.5%
37.2%
31.9%
26. 7%
-------
VI-66
b. The Rankine-cycle system with two-speed clutch transmission
and IR = 0. 29 provides a close approximation to the tractive
max
effort delivered by the 302-2V internal combustion engine with
three-speed automatic transmission.
c. The Rankine-cycle system with two-speed clutch transmission
and IR = 0. 29 provides a gradability of 49% with 54 ft-lbs
max
of torque subtracted for driving accessories.
3. 2 System Weight and Estimated Cost
In Table 13, the weight of the Rankine-cycle system reference
design based on thiophene (957 Ibs total) is compared with thg.t of
the 302 CID engine with three-speed transmission (806 Ibs total). Sub-
stantial weight reductions may be expected in the Rankine-cycle system
with development. For example, use of an all-aluminum condenser
would reduce the weight of this component by approximately 50 Ibs.
Use of Fluorinal-85 will also permit reducing the expander disp'lace-
3 3
ment from 184 in to 107 in , with a resulting reduction in the ex-
pander weight.
Using the conceptual design drawings, cost estimators from the
Ford Motor Company have prepared a preliminary large-volume
production manufacturing cost estimate for the system. The cost
was quoted relative to the 1970 302-2V with the result:
Cost, Rankine Cycle System _ +0.31
Cost, I/C System -0.15
The system manufacturing cost estimate will be updated as the devel-
opment continues.
-------
VI-67
TABLE 13
TABULATION OF TOTAL SYSTEM WEIGHT
AND COMPARISON WITH 302-2V INTERNAL COMBUSTION
SYSTEM WITH 3-SPEED TRANSMISSION
Expander Assembly
Feedpump
Expander Subsystem
Transmission
Burner- Boiler
Regenerator
Condenser
Radiator
Controls, Fans, Accessory
Drives and other Miscellan-
eous Components
Total
Rankine
Reference
Design
220
45
265 Ibs
135 Ibs
273 Ibs
54 Ibs
115 Ibs
75 Ibs
957 Ibs
302 in3 V-8
with 3 -speed
Automatic
479 Ibs
159 Ibs
54 Ibs
114 Ibs
806 Ibs
-------
VI-68
3. 3 System Packaging
For the conceptual design, the system, was packaged in the engine
compartment of an intermediate size American car; a full-size, com-
plete mockup of the system was constructed in the engine compartment
of this car with excellent results. In Figure 30, a photograph of the
mockup is presented and in Figure 31, sketches illustrating the expander
transmission, burner-boiler, regenerator, and condenser locations
in the system are illustrated. The only change required in the engine
compartment in packaging the system was modification of the frame
and fender panels at the very front of the car to facilitate placement
of the condenser.
3. 4 Emission Projections for Rankine-Cycle Automotive Propulsion
System
Using the burner emission levels obtained from the burner devel-
oped at Thermo Electron Corporation for a 5 hp Rankine-cycle cur-
rently on test, projections have been made of the emission level of
unburned hydrocarbons, CO, and NO from a Rankine-cycle automotive
propulsion system using 10 mpg fuel economy; the results are pre-
sented in Table 14, and compared with projected Federal standards
for 1975 and 1980;* presented also are results for an uncontrolled
I/C engine and the IIEC targets for emission control for the I/C
engine. The Rankine-cycle system emission levels are lower than
the projected 1980 standards by a factor of 5 for unburned hydro-
carbons, by a factor of 13 for carbon monoxide emission, and by
a factor of 1.6 for NO emission.
The projected Federal standards in Table 14 are those existing
prior to the passage of the Clean Air Legislation in the U. S.
-------
<
I—I
I
Figure 30. Photograph of Rankine-Cycle System Mockup In Intermediate
Size American Car Engine Compartment.
-------
VI-70
r-.-l.--L
\
\
,/
'
\ VJ 1'
x i-4^--4
i
''^- -
n ^
•
n
i
\. I
n l L
V J
FtCOPUMP
Figure 31. Position of Major Components for 100 hp Rankine-Cycle
Powerplant in an Intermediate Size American Car Engine
Comoartment.
-------
TABLE 14
EMISSION COMPARISON OF RANKINE CYCLE WITH INTERNAL COMBUSTION ENGINE
System
Emission HC
t
| Level CO
I gms/mi NO
Projected
1975/1976
Standards
0. 46
4. 7
0. 4
Uncontrolled
I. C.
Engine
5. 5
24
5 5
IIEC Targets
for
I, C. Engine
0. 82
* 7. 1
0.68
TECO Projections
for Rankine- Cycle
Propulsion System
0. 05
0, 35
0 25
H
I
-------
Vl-72
3. 5 Major Conclusions
a. The Rankine-cycle system offers a very strong potential for
minimum emissions of particulates, unburned hydrocarbons,
nitric oxide, and carbon monoxide.
b. The use of Fluorinol-85 as a working fluid, with a moderate
maximum cycle temperature of 550 °F, permits a significant
cost reduction relative to the equivalent steam system. This
reduction may permit the Rankine-cycle system to be com-
petitive costwise with the equivalent I/C system, particularly
since the stricter emission level standards will require sig-
nificant cost increases in the I/C system.
c. An organic-based working fluid, Rankine-cycle system
approximately equivalent in performance to a 302 CID
I/C engine with 3-speed automatic transmission can be
packaged in current automotive engine compartments
with only minor modifications required in the sheet
metal and frame.
d. The system development and design is sufficiently advanced
to insure a high probability of meeting the design goals and
performance estimates presented in this report.
-------
VII-1
Chapter VII
STIRLING ENGINE ACTIVITIES AT UNITED STIRLING (SWEDEN)
by
Lars G. Ortegren
Vehicle Applications
K. B. United Stirling (Sweden) AB & CO.
Malmo, Sweden
-------
VII-2
Introduction
The development of Stirling engines in Sweden is taken care of by
KB United Stirling (Sweden) AB and Company. The main object of
the company is the development and adaptation for production of
Stirling engines. At present, about 70 engineers are engaged in the
Swedish Stirling engine project.
General information on the Stirling engine principles, its properties
and potentials is given in a brochure, copies of which are available
at the conference. The most important characteristics of this engine
are its very clean exhausts and silent operation.
A major part of United Stirling's resources is being used for the
development of a 200 hp four cylinder in-line engine, suitable for a
number of applications, but especially for operation in areas where the
environmental properties are of great importance. Generally speaking,
this engine will have outer dimensions allowing the exchange of most
existing diesel engines in city buses, mining vehicles, boats and
stationary installations.
The 4-615 Engine Program
The 200 hp engine to be produced by United Stirling is named 4-615 (four
O
cylinders, 615 cmj of swept volume per cylinder). A development program
is in progress, starting with prototype 4-615A engines running in labora-
tories during 1971. The manufacturing and testing of a number of
gradually more advanced prototype engines is scheduled for the following
-------
VII-3
years. Series deliveries are planned to commence in 1976, preceded
by earlier pilot production.
A very important background for this development is an extensive use
of value analysis technique in order to approach a production version
with minimum outer dimensions and manufacturing cost.
The consecutive versions will be gradually more simplified and adapted
to series production methods. Further increased realiability and ease
of maintenance are other major objects of work. The engine principle,
with heat generation and working cycle separated from each other,
facilitates independent parallel development of components, e.g.,
burners, preheaters, control components, etc.
The 4-615 engine is a four cylinder, displacer type engine, with rhombic
drive mechanism. Nominal power output is 147 kW (200 hp) at 2,400 rpm
at the main output shaft. When choosing design parameters, efficiency has
been given a high priority, which, for the rather conservative values of
heater tube temperature and mean cycle pressure, has resulted in a rather
slow running engine with a relatively large swept volume. Thus the crank-
shafts turn at 1,550 rpm at nominal power output, and a speed increaser is
used to adapt the output shaft speed to existing automotive transmission.
One important result of value analysis work applied to the 4-615 engine is
a considerably reduced size, as compared to the present state of art.
The estimated dry weight of this engine is 900 -1,000 kg (2,000 - 2,200
pounds^ .
-------
VII-4
A brief preliminary specification of this engine is as follows:
Rated gross power 200 hp at 2,400 rpm
Maximum gross torque 80 kpm (580 Ibft) at 700 rpm
Maximum brake efficiency in
vehicle application 37%
Installation dimensions
Length 1,200 mm (48 inches)
Width 500 mm (22 inches)
Height 1,100 mm (44 inches)
Fuels Diesel oil, kerosene
Optional fuels LPG, LNG
Traction Application Projects
The installation and testing of engines in vehicles has always been
considered a very important part of engine development. This is
particularly true in the case of Stirling engine development, because
of the inherent relations between engine performance (power, efficiency)
and properties of the external cooling system.
City bus propulsion is considered one of the most attractive fields of
application for United Stirling's engine type 4-615. To make possible a
near-future commercial realization of Stirling engine powered buses, a two-
phase vehicle development project has been started.
For basic test and evaluation, a Philips type 4-235 engine will, as Phase I,
be installed in a test vehicle during 1971. In this phase, theoretical
calculations and considerations concerning vehicle performance will be
verified. A practically useful theory for matching engines, transmissions
and cooling systems will be established. Performance is scheduled to be
demonstrated during 1972.
-------
VII-5
Phase II, which is planned to be completed during 1973, consists of the
development of a city bus, using a 4-615 Stirling engine prototype. In
this phase, bus and propulsion systems will be mutually matched in order
to achieve a near-optimum solution with regard to driving characteristics,
space utilization, system noise level, passenger compartment heating,
etc. The development will be performed in close cooperation with a bus
manufacturer. Experience feedback to production engine design is a
very important part of this phase.
As a result of this two-phase development project, production of city
buses with exceptionally good environmental properties will be made
possible. Besides the very low exhaust emission level described earlier
in this paper, a remarkably low noise level will be the result.
Referring to the standardized measuring distance of 7.5 metres (25 feet)
from vehicle center line, a noise level below 70 dB(A) can be expected
under most urban operation conditions.
-------
VIII-1
Chapter VIII
NITROGEN OXIDE FORMATION IN THE CO~MBUSTION CHAMBER OF THE INTERNAL COMBUSTION
ENGINE AND ITS SUPPRESSION BY MEASURES FROM COMBUSTION TECHNOLOGY
by
W. E. Earnhardt
Research Department 2
Volkswagenwerk AG
Wolfsburg, FRG
Presented at Eindhoven Conference by
K. H. Newmann
Research Development Group
Volkswagenwerk AG
Wolfsburg, FRG
Translated for EPA by SCITRAN
(Scientific Translation Service)
Santa Barbara, California, USA
-------
VIII-2
Summary
Production and decomposition of nitric oxide in the internal
combustion engine was investigated, with consideration of the
kinetics of the elementary chemical reactions. A simple
mathematical model of nitric oxide formation in the engine
was developed. It consists essentially of an appropriate
reaction mechanism for the combustion process in the engine
and a thermodynamic analysis of the combustion process.
Conclusions obtained from this model concept were then tested
on a one-cylinder engine using such measures of combustion
technology as stratified charge operation to influence the
combustion process so that nitric oxide formation is
inhibited because of too low temperature, in spite of the
presence of oxygen. In this way the combustion is controlled
so that it is still possible to oxidize carbon monoxide and
hydrocarbons almost completely.
-------
VIII-3
Contents
1. Introduction
2. Model of Nitric Oxide Formation in the Engine
2.1 Reaction Mechanisms for Nitric Oxide Formation
2.2 Model of Flame Propagation and Thermodynamic Analysis
of the Combustion Process
2.3 Calculation of the Kinetics of Nitric Oxide Formation
3. Combustion Technology Measures for Suppression of Nitric
Oxide
4. Summary
5. Bibliography
-------
VIII-4
1. Introduction
Investigation of non-equilibrium processes of the
processes occurring in the combustion chamber of the internal
combustion engine has for some time been of major scientific
interest. In view of the drastic measures against air pollution
by automobile exhaust which have been required by legislators,
it is pressingly necessary to find ways to decrease the
injurious materials carbon monoxide (CO), nitric oxide (NO) and
unburned hydrocarbons (CH) contained in the exhaust gas, in
order to attain more complete combustion and thus a cleaner
exhaust gas.
In the following report, the thermal formation of nitric
oxide within an internal combustion engine is studied theoretically
because of its particular importance for environmental protection.
Thus a model of automobile nitric oxide formation was developed,
and its transferability to a real engine was studied. On the
basis of the conclusions obtained from that, it was attempted
to control the combustion process through measures from
combustion technology so that the NO concentration would remain
as small as possible, and also that the oxidation processes
necessary for a reduction of the HC and CO would be completed.
We shall not consider here the potential for NO reduction
by exhaust gas recirculation or by use of so-called "diluents"
such as C02? t^O, He, Ar, No, etc; by use of reduction catalysts;
by choice of unconventional fuels, or by addition of certain
additives to promote combustion. In this respect, see
G. H. Meguerian [11].
-------
VIII-5
2. Model of nitric oxide formation in the engine
2.1 Reaction mechanisms for nitric oxide formation
Thermal nitric oxide formation can be represented by the
gross reaction
N + 0 z=± 2NO
Such a reaction consists of a number of elementary reactions
which, considered microscopically, describe the collisions
between molecules, radicals, or atoms, through which new
elements are formed. The elementary reactions make up the
chemical transformations actually going on, while the gross
reaction reflects only the total result of these elementary
reactions, without considering the intermediate products.
On the basis of a literature study, the thermal nitric
oxide formation in combustion processes can be characterized,
for example, by the reaction mechanisms shown in Figure 1.
Reaction mechanism I was stated recently by H. K. Newhall and
S. N. Shahed [l], and mechanism II by L. S. Caretto and co-
workers [2]. Equations (2) to (5) of the first mechanism in
Figure 1, to be sure, are contained in the reaction scheme
suggested by K. Vetter [3] as early as 1945, while Equations
(3) to (8) of the second mechanism had been stated by H. K.
Newhall and E. S. Starkman [4j as well as by G. A. Lavoie and
co-workers [5]. Since both calculation and experiment have
shown [6, 7] that the nitrogen monoxide, NO, is the only
nitrogen oxide of significance in the engine, those elementary
reactions in these mechanisms which contain nitrogen dioxide,
NCL, or nitrous oxide, NoO, as participants can be neglected,
without causing any large error.
W. Bernhardt [8] showed recently that with respect to
nitric oxide formation in the engine, the reactions
-------
VIII-6
N + O + M^NO +M
and OH + N F^ NO + H
are of secondary importance. See also [9]. Thus two equations
remain in each of the reaction mechanisms shown in Figure 1.
These are well known under the name of the Zeldovich mechanisms [10]
If one compares the velocity constants in the Zeldovich
mechanism, using the kinetic reaction data of K. L. Wary and
I. D. Teare [12], it appears that the reaction
N2 + 0 ?± NO + W
is clearly the rate-determining reaction for thermal nitric
oxide formation because of its high activation energy. See
also Figure 4.
-------
Zeldovich mechanisms
N
^ 0 + M
02+N
N2+0
I02 + 0
^
-
-
NO
NO
NO
NO
+ M
.0
+ N
•>• 0?
fy
\
N2 +
02 +
OH +
N2 +
N7 +
0
N
N
02
OH
^
^
N
N
N
*
N-
0 +
0 *
0 +
? 0 +
>0 +
N
0
H
0
N20+0
N2+ 0+M
NO +0+M
NO * NO
N20 +M
N2 0 +0
N+0 + M
A/2 + 0 v- M
NO +NO
NO +M
Research 2
BE-70-01
M
I
The triple collision partner M can be any particle from the reaction
volume; it acts only to remove energy.
Figure 1. Reaction mechanisms for the formation of nitric oxide in engines.
-------
VIII-8
2.2 Model of flame propagation and thermodynamic analysis of
the combustion process
In the model of flame propagation used in this work it is
assumed initially that during the combustion in the combustion
chamber there are areas with quite different temperatures. The
region already passed through by the flame front contains a burned
mixture of high temperature. The region not yet reached by the
flame contains fresh gas of relatively low temperature. It is
assumed, however, that in both regions it is possible to have
intensive mixing and thus a more rapid decay of the temperature
differences, and that the energy exchange from the unburned
region to the burned region occurs so rapidly that the fresh
gas temperature suddenly rises to the higher temperature of
the exhaust gas.
So far it has not been possible to establish generally
valid analytical relations for the flame advance within an
internal combustion engine. It is assumed, however, that
the model concept used here sufficiently describes the processes,
especially of the actual pressure curve while the combustion
process goes on. Along with the pressure, which is taken as
constant everywhere in the combustion chamber, the volume V
is given as a function of time, from the well-known kinematic
equation for piston motion. According to this model concept,
the total volume is made up of the volume of the fresh gas and
the volume of the burned gas.
The temperature in the unburned region is calculated assuming
an isentropic change of state for an ideal gas. This is a crude
approximation, because the prerequisite for application of
equilibrium thermodynamics is not given. Because of the lesser
effect of temperature in the unburned gas on the following
calculations, we can proceed in that way here.
-------
VIII-9
The temperature in the burned region can be determined
by the time course of conversion of the chemical energy into the
internal energy of the combustion gases. For this purpose, of
course, a thermodynamic analysis of the processes in the engine
must previously have been performed. In this analysis it must
be considered that the working fluid changes during the process
because of the chemical reaction. Figure 2 shows the result of
such an analysis for a VW Type 3 engine (3010 rpm, air ratio
A = 0.93, full load operation).
Extensive descriptions of the procedure for determining
the energy transformations in an internal combustion engine
can be found in W. Hinze [13]and W. Bernhardt [14]. For that
reason we will not discuss it extensively here. Figure 3 shows
the time course of the temperature in the unburned and burned
regions of the combustion chamber for the combustion process
being studied. The figure also shows the mean temperature
in the working gas, as would be given from an adiabatic mixing
process. In the temperature determination it was assumed that
the state behavior of the reacting substances is given by the
equation of state for ideal gases.
-------
1TJ
H-
OP
c
H
0)
o> m
3 3
X
n
p t^
P
H-, 3
C Oq
3 CD
O
r+ H-
H. 3
O
3 S3
3" CD
CD i-i
3
>-t P
CD I—1
P
n n
rt O
H' 9
O CT
3 C
3 o
CD 3
CD
C
CD
7.5
J
1.25
1.0
0.75
05
0.25
0
internal energy of the combustion
gases
CD
3
M-H
0
reaction time
research 2
BE-70-03
-------
OQ
c
CD
OJ
H H
CD H-
OQ 3
h" CD
O
3 n
C/1 O
C
O i-j
i-h W
CD
rt
^ O
CD Hi
n rt
o tr
3 CD
C rt
C/l CD
rt 3
O CD
O rt-
tr1 C
P H
3 CD
CT
n> H'
H'
I-H
H-i
CD
H
CD
3
H-
3000-
K
2600-
CD
tn
3
2 2200-
region
1—I
H
I
1,6
3,2
4,8 .6,4 8,0 9,6
11,2
12,8 74,4 76,0 ms 17,6
reaction time
research 2
BE-70 -02
-------
VIII-12
2.3 Calculation of the kinetics of nitric oxide formation
The model of flame propagation which has been described was
used to calculate the kinetics of nitric oxide formation. The
temperatures and volumes or densities calculated by means of this
model are particularly needed. For the Zeldovich mechanism,
the time change of the NO concentration can be written in the
form presented in Figure 4 for isochoric processes, in which the
density remains constant.
Problems arise in the application of this differential
equation, however, because of the determination of the time
course of the concentrations of the participants in the reaction,
especially of the 0 atoms. Starting from the known initial
concentrations of Oo and N~ immediately before the beginning
of the combustion process, the concentrations of Oo, 0 ^ and N
are calculated approximately for the temperatures in the different
regions of the combustion chamber by use of the equilibrium
constant, considering the dissociation and the fuel conversion
which has already occurred. With the known concentrations
"0^ ' ^2 ' 0 and "N » known density, and known temperature,
the differential equation in Figure 4 can be solved by means of
the method of Runge-Kutta. Thus the NO formation can be followed
analytically in the unburned as well as in the burned region of
the combustion chamber.
As an example, Figure 5 shows the curve for the NO concen-
trations in the two regions existing, according to the model
concept, with a VW Type 3 engine at 1000 rpm, ignition advanced
to 12° before TDC, and an air ratio of % = 0.93. The
nitric oxide formation is shown graphically, starting at the time of
ignition ( t = 0 ), during the combustion and expansion processes.
o
The nitric oxide concentration is shown in moles per cm.
As was to be expected from studies carried out previously
by means of equilibrium theory, the nitric oxide formation in
the fresh gas is negligible in comparison to that in the burned
-------
Oq
c
CD
4^.
n
(a
I—"
n
H'
O
O
l-h
O
in
O
l-h
O
l-i
3
PJ
rt
H'
O
OO
1
06
O
0
O'
O'
0'
00
O'
CD
O
/
/
/
7
X
/
^—
7.6 3.2 4.8 6,4 8.0 9.6 77.2 72.8 74.4 1t
reaction time ^
$.0 77.6
ms
^uP BE-70-OU
research 2
-------
time change of the NO concentration for the Zeldovich mechanism:
OQ
C
H
CD
H
H'
3
CD
O
O
e
l-t
O
Hi
2
O
n
o
o
CD
fo
H-
H'
O
d(5
NO _
in which O is the density,Ojis the number of moles of component i per
mass, t is the time, and k. is the velocity constant of reaction j:
1
2
3
A
reaction
N2 + 0 — NO + N
NO + N — N2 + 0
02 + N — NO * 0
NO + 0 — 02 * N
velocity constant CfTT^
' mol • sec
7- 10 13 exp C-75000/ R T]
1,55- ]013
15,3 • 109 T exp (-7080 /
3,2 • 109 T exp [-39100/
R-T]
R-T]
T in K , R in ccU/(mol- Kj
research 2
BE-71-01
-------
VIII-15
region because of the lower temperatures in the fresh gas
region ( < 1000 °K). The nitric oxide concentration in the fresh
gas is so slight that it is barely detectable in Figure 5. It
is a characteristic of the thermal nitric oxide formation in an
Otto engine that it becomes perceptible only quite late,
after about a third of the total combustion period has passed.
[This applies only for the average nitric oxide concentration,
referred to the total combustion volume, which we are considering
here.] The maximum of the (average) NO concentration occurs only
when two thirds of the combustion period has passed. After the
maximum has been reached, there is hardly any detectable change
in the nitric oxide concentration during the expansion process.
This result is in distinct agreement with the experimental studies
of H. K. Newhall and E. S. Starkman [4, 16], who recorded the
monochromatic emission of NO by means of a monochromator directly
during the expansion, and with the reaction kinetic calculations
of H. K. Newhall [17]. On comparison of the engine nitric oxide
concentrations in the exhaust gas, as predicted by this, model of
engine nitric oxide formation, with the values measured for an
actual engine, it can be seen that this model is valid only
for nearly stoichiometric mexture ratios and for the region X > 1.
The reason is that in the rich region ( X < 1 ) the effect of
the fuel is not considered. This has also not been done in
other models just recently proposed [18, 19].
According to G. H. Meguerian [11], with a fuel excess in
the Zeldovich mechanism, the oxygen and nitrogen radicals react
with the unburned or partially burned fuel molecules. This
leads to lower nitric oxide concentrations, and the NO values for
15% fuel excess are about 30% below the values obtained for the
stoichiometric mixture. See also [19J. In the thin region
( X > 1 ) the predicted NO values agree well with those measured
in the engine exhaust, because the flame velocity decreases with
increasing air ratio, and this effect is considered in the
flame propagation model.
-------
VIII-16
3. Combustion Technology Measures for Suppression of Nitric Oxide
In the calculation of the kinetics of nitric oxide formation
it appears that temperatures above 2600 °K are necessary for
formation of NO in the burned region of the mixture. See Figures
3 and 5. By contrast, the CH and CO oxidation processes go on
even at considerably lower temperatures. If it were possible to
control the combustion process in the engine so that the temperature
of the combustion gas would remain below about 2600 °K, the
nitric oxide formation would be inhibited. This would certainly
lead to lower NO concentrations in the exhaust gas. Another
important conclusion from the model of engine nitric oxide
formation is that NO is formed only quite late in large concen-
trations in the burned mixture. Since the flame front of the
burned region has already passed, we speak here of "post-flame
reactions".
On the basis of knowledge obtained from the model concept
of NO formation, an engine combustion process was tested with
a CLR one-cylinder engine in research at the Volkswagenwerke AG. ^
The process appeared to be able to decrease simultaneously the
CH and CO concentrations as well as the NO concentration. This
amounts essentially to use of a pre-combustion chamber, in which
a nearly stoichiometric or even a rich mixture is first ignited
and burned. The design of the combustion chamber is such that,
after passage through the pre-combustion chamber, the flame front
advances into the main combustion chamber. There it meets the
load-controlled main mixture at relatively low temperatures.
(The main combustion chamber can even contain completely pure air.)
In this way the hot combustion gases are rapidly lowered in
temperature, so that in spite of excess oxygen only small amounts
of NO can be formed. The temperature which results is still high
enough, though, to allow rapid oxidation of carbon monoxide and
hydrocarbons. A pre-combustion chamber arrangement which is
similar in principle was suggested recently by Newhall [20].
These investigations were performed by Graduate Engineer
I. Geiger.
-------
VIII-17
Figure 6.
Combustion chamber with pre-combustion chamber,
y**\/s\jp*!*jf*vu^
'
„ ___ , ____„_. ^
" ' ' '' '
.
Figure 7.
Combustion in a combustion chamber with a pre-
c ombus t ion chamber .
Figure 8. Combustion in a combustion chamber with a pre-
combustion chamber.
-------
VIII-18
Figure 6 shows the arrangement of the pre -combust ion chamber
and the main combustion chamber with a VW Type 3 cylinder head.
In these studies, the ratio of pre -combust ion chamber volume to
main combustion chamber volume was between 1:10 and 3:10. With
this pre-combustion chamber arrangement, very slight exhaust
gas emissions were measured with a CLR one-cylinder engine
at 2000 rpm under full load operation with optimum ignition
timing. The emissions were as low as 40 ppm NO, 40 ppm CH
(measured as hexane) and 0.1% CO by volume at air ratios of
X = 2 to X = 3.
In conclusion it should be mentioned that the combustion pro-
cesses in the combustion chamber could be filmed by use of a
high-speed camera (6,000 pictures /second) so as to test
whether the combustion technology measures applied satisfied
the information obtained from the model of engine NO formation.
Figure 7 shows 64 pictures from a single combustion process.
They were made at 2000 rpm, ignition at 30° before TDC, effective
2
mean pressure P = 2.5 kp/cm , an air ratio of A = 2.5 in
the main combustion chamber and an over-all air ratio ^n~G = 1.6,
6c;>
using the high-speed camera. In this case we measured 170 ppm NO.
This is more than an order of magnitude less than in conventional
combustion chambers.
Figure 8 shows an enlargement of one photograph from
Figure 7.
4. Summary
The nitric oxide concentration in the exhaust gas can be
predicted for stoichiometric and fuel-poor mixtures with the
simplified model of engine nitric oxide formation described here.
It is shown that, along with the velocity constants of the reaction
N + 0 i=i NO + N
-------
VIII-19
the temperature in the region passed by the flame front is the
principal quantity affecting the prediction.
Based on conclusions from the model, and with basic
changes in the previous combustion processes, such as stratified
charge operation, exhaust gas emission measurements on a CLR
one-cylinder engine with a modified VW Type 3 cylinder head
showed NO at 5 to 10% and CH at 10 to 25% of the levels observed
with conventional combustion processes. Pre-combustion chamber
arrangements, combined with stratified charge operation could
prove to be potential solutions for significant reduction of
the injurious materials in automobile exhaust gas. In spite of
the good results obtained so far, it is doubtful whether this
method alone will be suitable to attain the drastic reduction
of emissions required by American legislators after 1975,
with retention of good specific power and economy.
-------
VIII-20
REFERENCES
1. Newhall, H. K. and S. M. Shahed. Kinetics of Nitric Oxide
Formation in High Pressure Flames. Paper presented at
Thirteenth Symposium (International) on Combustion, Salt Lake
City, Utah, August 1970.
2. Caretto, L. S., L. J. Muzio, R. F. Sawyer and E. S. Starkman.
The Role of Kinetics in Engine Emission of Nitric Oxide.
Paper presented at the AICHE-Meeting, Denver, Colorado,
August 1970.
3. Vetter, K. Kinetics of Thermal Decomposition and Formation of
Nitric Oxide. Z. f. Elektrochemie, Vol. 53, 1949, pp. 369-80.
4. Newhall, H. K. and E. S. Starkman. Direct Spectroscopic Determina
tion of Nitric Oxide in Reciprocating Engine Cylinders. SAE-
Paper No. 670, 122, 1967.
5. Lavoie, G. A., J. B. Heywood and J. C. Keck. Experimental and
Theoretical Study of Nitric Oxide Formation in Internal
Combustion Engines. Combustion Sci. § Technol., Vol. 1, 1970,
pp. 316-26 .
6. Wimmer, D. B. and L. A. MacReynolds. Nitrogen Oxides and Engine
Combustion. SAE Trans., Vol. 70, 1962.
7. Campau, R. M. and J. C. Neerman. Continuous Mass Spectrometric
Determinations of Nitric Oxide in Automobile Exhaust. SAE-
Trans., Paper 660 116, Vol. 75, 1967.
8. Bernhardt, W. Studies of the Nonequilibrium Processes of the
Reactions Occurring in the Combustion Chamber of the Internal
Combustion^Engine. Lecture at the VDI Thermodynamics Col-
loquium, Wiirzburg, 5 October 1970.
9. Campbell, I. M. Chemical Mechanismus Relevant to the Production
and Emission of Nitric Oxide and Carbon Monoxide from Com-
bustion Engines. Lecture 7 in: A short Course on Fundamen-
tals of Engine Exhaust Pollution, University of Leeds,
September 1970.
10. Zeldovich, Ya. B. The Oxidation of Nitrogen in Combustion
Explosions, Acta Physocochimica URSS, Vol. 21, 1946, pp. 577-
628.
11. Meguerian, G. H. Nitrogen Oxide-Formation, Suppression, and
Catalytic Reduction. Paper PD 23. Presented at the World Oil
Congress in Moscow, June 1971.
12. Wary, K. C, and J. D. Teare. Shock-Tube Study of the Kinetics of
Nitric Oxide at High Temperatures. J. Phys. Chem., Vol. 36,
1962, pp. 2582-96.
-------
VIII-21
13. Hinze. W. Procedures for Thermodynamic Evaluation o£ Test
Stand Results for Internal Combustion Engines Studied.
Dissertation, Dresden Technical University, 1956.
14. Bernhardt, W. Thermodynamic Evaluation of Test Stand Results
to Determine Laws of Energy Conversion in the Engine. VW
Report T 327, 11 June, 1969 (unpublished).
15. Bernhardt, W. Formation of Nitric Oxide in Internal Combustion
Engines. Brief Report from Research 2, 18 December 1969
(unpublished) .
16. — . Control of Oxides of Nitrogen. Fifth Status Report of VW
of America to California Air Resources Board, Los Angeles,
VW Report V/70, (unpublished).
17. Starkman, E. S. Formation of Exhaust Emission in the Combustion
Chamber. XIII. Congress of FISITA, Bruxelles, Belgium,
Paper No. 15. 3. D. , June 1970.
Newhall, H. K. Kinetics of Engine -Generated Nitrogen Oxides and
Carbon Monoxide. Twelfth Symposium (International) on Combus-
tion, The Combustion Institute, Pittsburgh, Pennsylvania,
1969, pp. 603-13.
19. Heywood, J. B., S. M. Mathews and B. Aven. Predictions of
Nitric Oxide Concentrations in a Spark-Ignition Engine
Compared with Exhaust Measurements. SAE -Paper 71001, -Vol. 11,
15 January 1971, Detroit, Mich.
20. Muzio, L. J., E. S. Starkman and L. S. Caretto. The Effect of
Temperature Variations in the Engine Combustion Chamber on
Formation and Emission of Nitrogen Oxides. SAE -Paper 710158,
Detroit, Michigan, Vol. 11, 15 January 1971.
21. Newhall, H. K. and I. A. El-Messiri. A Combustion Chamber
Concept for Control of Engine Exhaust Air Pollutants Emissions
Combustion and Flame, Vol. 14, 1970, pp. 155-58.
18
-------
IX-1
Chapter IX
A EUROPEAN CONTRIBUTION TO LOWER VEHICLE EXHAUST EMISSIONS
by
Diarmuid Downs
Managing Director,
Ricardo and Company
Shoreham-by-Sea,
Sx., United Kingdom
-------
IX-2
I do not propose to discuss whether the present limits and
proposed future limits for road vehicle exhaust emissions are
sensible economically or well founded scientifically. I will
only remark in passing that the NQx regulations will be the most
difficult to meet and will have the greatest effect on the economics
of operation of motor vehicles. It behoves us to be sure,
therefore, before imposing them, that they are really essential
medically and environmentally, and that we have our priorities
right in relation to other atmospheric pollutants.
In this Note, I have deliberately confined myself to two
areas of interest and concern to the automobile engineer
1. The small European automobile and its problems
in meeting future US legislative requirements.
2. The contribution which combustion chamber design
can make to reducing exhaust emissions from the diesel engine.
-------
IX-3
The European Automobile and the American Market.
As Consulting Engineers engaged in design, development and
research on internal combustion engines, we have been very active
in assisting European car manufacturers to meet the U.S. Federal
and Californian State exhaust emission limits. Our experience is
probably unique in the range of vehicle type we have studied and
the variety of approach we have used to meet individual requirements.
To meet the U.S. Regulations up to and including those imposed
for the 1971 model year, we have used in the main what has come
to be known as the Cleaner Air Package (CAP) approach, involving
a tightening of production tolerances and detailed adjustments to
the carburettors and ignition settings over the load and speed
range, combined with the use of such devices as intake air heaters
and manifold air depression limiters. The art is to reach the
required exhaust emission levels without unacceptable loss of
driveability. This is harder to achieve with the small European
car than with the larger and generally more powerful American car.
For .this reason, a number of European car manufacturers have adopted
petrol injection in place of carburation, which, because of the
more precise metering and delivery of the fuel over the operating
range including the transients, enables the present exhaust emission
limits to be attained without sacrifice of driveability; in fact,
in most cases, with driveability enhanced. For the 1972 model
year, it seems probable that further refinement of the CAP approach,
particularly if associated with petrol injection, will enable the
lower limits associated with the change of test procedure, to be
attained. In some cases, however, it may be necessary to use
manifold air oxidation, a device which has been used intermittently
-------
IX-A
since the early days of pollution control, but which, because of
the cost of the air pump and the problems posed by its installation
and drive, has generally been abandoned wherever possible.
We have been working for some time with our eye on 1975 and
Table I summarizes the present position. The first line gives the
limits which we thought we were going to have to meet prior to the
signing of the Nixon/Muskie bill last December. The second line
gives our present target, necessarily somewhat vague in regard to
NOx. Some typical results obtained with small European cars fitted
with catalytic afterburners are shown in the second half of the
Table. It can be seen that, ignoring NOX for the moment, the
pre-Muskie limits for HC and CO have almost been reached with
copper/chromium catalyst and could probably be still further
reduced with more development work. With the platinum catalyst,
even the limits required by the Nixon/fauskie bill can be met in
regard to HC and CO. When the attempt is made to reach the NO
limits as well, by the use of Exhaust Gas Recirculation (EGR)
combined with a platinum catalytic afterburner, the NO is certainly
X
reduced, but the HC and CO are then too high. These results were
obtained with an unleaded fuel and it should be emphasized that
the performance of these catalytic systems in respect of endurance
has not yet been fully assessed.
It would seem that it is going to be very difficult, if not
impossible, to reach the figures required for 1975 by the use of
EGR, without unacceptable loss of driveability. For this reason,
we believe that a double catalytic system holds out the best hope
for a European car, with a first stage reducing the NOX and a
second stage oxidising the HC and CO. We have no results to
report on such a system as yet.
-------
IX-5
TABLE 1
AUTOMOBILE EXHAUST EMISSIONS
(1972 FEDERAL TEST PROCEDURE)
HC CO NOX
1975 U.S. FEDERAL LIMITS ~ ~
1. Pre Nixon/fouskie 0.5 11.0 0.9
2. Post Nixon/kuskie 0.^6 4.8 0.4-0.6
RESULTS WITH SMALL EUROPEAN CARS
1. Cu/Cr Oxidation Catalyst 0.6 8.5 5-0
2. Pt Oxidation Catalyst 0.3 3-0 5-0
3. Pt Catalyst + EGR 1.7 10.0 1.0
-------
IX-6
Although the US legislative requirements between 1972 and 1975
are not entirely clear, it would appear that we shall have to meet
low limits for NOX in California in 1974, As the figure would not
appear to be nearly as low as that proposed for 1975, however, it
could probably be met by the use of a small amount of EGR combined
with a catalytic afterburner, without too much loss of driveability.
The Federal limits for 1975 could probably be met in regard to CO
and HC by an oxidation catalyst system. It is going to be very
much'more difficult to meet the 1975 limits for NOX, and we in
Europe would certainly hope that the application of these limits
would be postponed until 1976, as is provided for in the Nixon/Muskie
bill.
Exhaust Emissions from the Diesel Engine.
Attempts to lower the exhaust emissions from the spark-ignition
engine by changes in combustion chamber design have been disappointingly
ineffective. The same is not true of the diesel engine where the
swirl chamber system, long recognized as giving a cleaner exhaust
in regard to smoke than the direct injection system, is now shown
to be superior also in regard to gaseous exhaust emissions, particularly
nitrogen oxide. This is illustrated by comparative tests carried
out on a single-cylinder engine of 1600cc (96 cu.in.) capacity
fitted with a) a Ricardo Comet V swirl chamber combustion system
and b) a Direct Injection combustion system. Fig. 1 shows
comparative figures for carbon monoxide emissions over the load and
speed range of this particular engine. The CO emissions are very
low in both cases, in comparison with the figures which would be
obtained from a spark-ignition engine, but, even so, those from
the swirl chamber system are lower than those from the direct
-------
IX-7
EMISSIONS OF CARBON MONOXIDE FROM TWO
I2O X I4O mm. DIESEL ENGINES AT OPTIMUM
INJECTION TIM1MG5.
^ COMET ~JT COMBUSTION CHAMBER
B. DIRECT INJECTION COMBUSTION CHAMBER
DRG. No D. 21653
DATE:- 4- 11-70
Plfi. I
600 ; 8QO : looo
1600 ! I80O 200O : ,
-------
injection system. The unburned hydrocarbon figures shown in Fig.2,
although they are low in comparison with those from an untreated
gasoline engine, give no cause for complacency when we remember the
improvements which have been made in the latter unit in recent years.
Here again the swirl chamber engine gives lower hydrocarbon emissions
than the direct injection unit. Fig. 3 shows the nitrogen oxide
figures and here it can be seen that the emissions from the swirl
chamber unit are less than half those from the direct injection unit.
An additional advantage of the swirl chamber system is that the
injection timing may be retarded from the optimum with a big reduction
in NOX, and incidentally a reduction in noise, but with only a small
change in the smoke-limited power, as shown on Fig. 4- A
corresponding change in the injection timing of the direct injection
engine, although it also reduces the nitrogen oxide concentration
considerably, at the same time reduces the smoke-limited power
appreciably. There are very good technical reasons for the better
performance of the swirl chamber system in regard to nitrogen oxide
production. The high rates of fuel/air mixing and therefore of
combustion of this system enables retarded injection timings to be
used, vrlth a lowering of nitric oxide concentration as noted above,
without the severe effect on smoke-limited power experienced with
the slower burning Direct Injection unit. Also, with the swirl
chamber system, all the fuel is initally injected into only half
the air and only afterwards is this rich fuel/air mixture mixed
with the rest of the air in the cylinder and burning completed.
This results in a lower temperature/volume/time integral for the
the swirl chamber in comparison with the direct injection system
and this is an important parameter due to the strong temperature
-------
TX-Q
EMISSIONS OF UNBURNED HYDROCARBONS FROM
TWO I2O X I4O mm. DIESEL- ENGINES AT OPTIMUM
INJECTION TIMINGS
DRG. No. D. 21654
DATE:- 4 -II- 70
FIG.2
_A^ COMET T COMBUSTION CHAMBER
_B_ DIRECT INJECTION COMBUSTION CHAMBER
EMISSIONS QUOTED IN ppm. CARBON
oo I ieoo 2000
-------
IX-10
EMISSIONS OF" OXIDES OF NITROGEN FROM TWO
I2O X I4O mm. DIESEL ENGINES AT OPTIMUM
INJECTION TIMINGS
_A_ COMET TT COMBUSTION CHAMBER
_B_ DIRECT INJECTION COMBUST/ON CHAMBER
EMISSIONS QUOTED IN ppm. NO
DRG No. D 21655
DATE- 4-11-70
FIG 3
;: izoo : : 1400 ;
RPM
-------
TX-ll.
EMISSIONS OF OXIDES OF NITROGEN FgQM
TWO 120 x I4O mm. DIESEL ENGINES WITH
INJECTION TIMING RETARDED BV A-° FROM OPTIMUM
k_ COMET T COMBUSTION CHAMBER
B DIRECT INJECTION COMBUSTION CHAMBER
EMISSIONS QUOTED IN ppm NO
DRG. No. D. 21656
DATE - 4- II - 70
FIG 4
-------
IX-12
dependence of nitric oxide formation.
In the United States, the State of California Air Resources
Board has proposed regulations for exhaust emissions from heavy
duty diesel engines, as set out in Table 2. The second half of
the Table gives results derived from the single-cylinder engine
tests, just described. It can be seen that CO is no problem, but
that the standard Direct Injection engine is marginal for 1973 and
that even the swirl chamber combustion system will not meet the
1975 limits in standard form. It should be mentioned that these
single-cylinder engine test results are generally in accord with
those obtained from a wide variety of multi-cylinder units.
With retardation of injection timing and possibly irith other
minor modifications to the engine, it should be possible to get down
to 5 gms/hp/h with the swirl-chamber engine, but it is going to
be very difficult to do it with the Direct Injection system.
Using EGR with the swirl chamber engine should enable figures
of 3 gms/hp/h to be achieved, and somewhat similar or slightly
higher values should be obtained with water injection.
Neither EGR nor water injection are very attractive solutions,
however, because of their possible effect on engine durability in
addition to their influence on performance. A catalytic method
of dealing with NOX, as is proposed for the spark-ignition engine,
would appear to be denied us on the diesel, as no one has yet
developed a catalytic method of eliminating nitrogen oxide which
will work in the oxidising atmosphere almost always present in the
diesel engine exhaust. To this extent nitric oxide poses a much
-------
IX-13
DIESEL ENGINE EXHAUST EMISSIONS
(gma/hp.h.)
CO HC + N02
CALIFORNIA PROPOSALS ~
1973 40 16
1975 25 5
EUROPEAN DIESEL ENGINES
1. Direct Injection 2.6 13-4-
2. Ricardo Comet Swirl Chamber 3-0 7.3
3. Ricardo Comet 4° Retard 5-5 5-9
4. Ricardo Comet 4° Retard
20% EGR 8.2 3-0
5. Ricardo Comet Optimum Timing
2:1 Water/Fuel Injection 4.0 3-4
-------
IX-14
more difficult problem for the diesel engineer than for the gasoline
engineer and brings me back to the question I asked right at the
beginning. Are we really sure that it is necessary to achieve such
low limits of nitric oxide in the engine exhaust?
-------
X-l
Chapter X
LOW EMISSIONS FROM CONTROLLED COMBUSTION
FOR AUTOMOTIVE RANKINE CYCLE ENGINES
by
W. A. Compton, J. R. Shekleton, T. E. Duffy,
and R. T. LeCren
Solar Division of International Harvester Company
San Diego, California, USA
Presented at Einchoven Conference by
W. A. Compton
Assistant Director-Research
Solar Division of International Harvester Company
San Diego, California, USA
-------
X-2
ABSTRACT
Rankine cycle engines have a high potential of meeting the emission levels
established by the 1970 amendment to the Federal Clean Air Act for the 1975-76 auto-
mobile. This paper discusses a Solar research and development program sponsored
by EPA/APCO which demonstrates a full scale 2 million BTU per hour working model
of a Rankine cycle engine combustor and controls which can surpass the emission goals
established.
Special features of the combustor are the unique methods of precisely control-
ling both the fuel and air to provide optimum flame performance at any engine power
level. This paper discusses the special requirements of the Rankine cycle engine and
shows why the very wide range of fuel flow required necessitates use of special tech-
niques in fuel atomization, fuel and air control, and aerodynamics. Sufficient discus-
sion is included to show the design methods that are necessary, in general, to achieve
low emissions in continuous flow combustion systems. Emphasis is placed on the
importance of interfacing a combustion system with other engine parts if a successful
low emission, wide turndown ratio combustor working model is to be achieved. Suf-
ficient discussion on combustion kinetics is included to advise on approaches necessary
to minimize NO formation in external combustion systems while maintaining high
efficiency and low CO and unburned hydrocarbons.
-------
X-3
I. INTRODUCTION
The Rankine cycle engine has a high potential of meeting the established 1975-
76 emission levels when installed in a family car, thus eliminating an atmosphere of
undesirable fumes now commonly contributed to the internal combustion engine. The
major portion of such an effort must, however, be devoted to a system to develop a
full-scale (2 million BTU per hour) prototype combustor system to demonstrate that
the desired low emission levels can be met with the proper operating performance.
Solar has addressed itself to these problems by drawing on its many years of
gas turbine experience, studies of combustion kinetics, and applying a novel fuel
atomization method. In addition, a precise air-fuel ratio control concept is necessary.
The solution lay in continuously modulating the fuel flow to match engine power demands
and to similarly control the air-flow to the combustor so that optimum combustion
could be obtained under all power level conditions. Such a method of control lay outside
the state-of-the-art of present day methods and special techniques had to be applied to
the solution of this problem.
Throughout the design, automotive features have been emphasized. Low
emission, compactness, high response, low cost potential, and high efficiency have
been major considerations in the design selection process. The system described in
this report includes all necessary controls to supply and regulate both fuel and air.
High response rates are obtained while continuously maintaining an optimum air-fuel
ratio with the lowest emissions. An axial blower is used to supply the required air to
the combustor. Air-fuel ratio and fuel rates are controlled by a single power demand
lever. The effects of air temperature, pressure leakage, flow, speed, and fuel back
pressure are automatically compensated by simple mechanical control systems.
H. 1975-76 EMISSION LEVELS FOR MOBILE ENGINES
The goals set for this program were those for a six-passenger automobile
vehicle with a maximum test weight of 4600 pounds, tested for emissions in accordance
with the procedure outlined in the November 10, 1970 Federal Register (Ref. 10). Values
are:
• Hydrocarbons* 1.65 mg/gram fuel (0.46 gram/mile)
• Carbon Monoxide 16.25 " " " (4.7 " " )
• Oxides of Nitrogen** 1.38 " " " (0.4 " " )
• Particulates 0.1 " " " (0.03 " " )
* Total hydrocarbons (using 1972 measurements procedures) plus
total aldehydes. Aldehydes to be 0.16 mg/gram fuel maximum.
** Computed as NO2-
-------
X-4
The task of meeting the goals are graphically displayed in Figure 1, comparing
typical emission levels for light duty vehicles of 1960 and 1970 to those published as
requirements for 1975-76 in the 1970 amendment to the Federal Clean Air Act.
III. PERFORMANCE GOALS
The basic performance goal of the demonstration system is to develop a full
scale (2 x 10^ BTU/hr) prototype combustion system which could be capable of achiev-
ing the emission levels when such a system is integrated into Rankine cycle engines
installed in an automobile. In order to have the widest possible technical significance
( to a potential family of yet undefined engines), extreme boundary limits were placed
upon the performance goals. These limits appear in the power range, transient
response, parasitic power, and volume goals. The importance of transient response
was emphasized because previously reported combustor and engine tests with external
combustors indicated that on-off cycling and rapid transients were a major source of
emissions. The following performance goals were used in synthesizing the combustion
system design:
• Heat release: 2 x 106 to 2 x 104 BTU/hr in a maximum of 1.33
cubic feet combustor volume. The unit shall be able to control the
time average heat release at any point from 2 x 10^ to 2 x 10
BTU/HR with low emissions. This represents a 100 to 1 ratio of
heat release between maximum and minimum.
• Fuels: Diesel No. 1, Jet A, or Kerosene
• Rapid transients in firing rates without severe degradation of
emission performance. A goal of 50 percent power change per
second or 1 to 100 percent power change in 2 seconds has been
established to allow interface with fast response flash vapor
generators.
• Minimum volume consistent with automotive research goals.
Weight factors were considered, but prototype construction
practices were used in the early research model.
• Parasitic power losses of less than 2 horsepower without vaporizer.
• Air density variations caused by altitude ranges between zero and
5000 feet, and temperature variations due to variations from 0 to
130° F will be accounted for in the design.
-------
X-5
NO2-mg/g
1960
ITYP)
1
1970
(TYP)
-25-
— J
0-
3-
1971
-10-
41.S
1*4
HC-
20 40 «0 10
nng/g
CO-mg/g
I960
(TYP)
290
FIGURE 1. COMPARING TYPICAL LIGHT DUTY VEHICLE EMISSION
LEVELS TO 1975-76 LEVELS
• Rapid startup to full power in less than three seconds.
• High Reliability and low cost for automotive applications shall be
inherent in the design approaches.
IV. COMBUSTION SYSTEM DESIGN
4.1 REACTION KINETICS
Three rules must be obeyed to burn liquid hydrocarbons quickly and efficiently
(Ref. 1, 2, 3, and 4):
1) The fuel must be rapidly evaporated
-------
X-6
2) The fuel and air must be rapidly mixed.
3) The rate of chemical reaction must be maximized
The rules are satisfied when:
_ a) Fuel droplet size is small
Evaporation
b) Droplet to air relative velocity is high
c) Air is injected at high velocity through a large number
Mixing of holes into a small combustor
d) A large number of fuel injection points are used
Reaction e) The air-fuel ratio is stoichiometric
In practice, unless heat losses are involved, it is necessary to add additional air in
order to avoid chemical dissociation losses.
With variations, these rules are the basis for the design of most combustors
used in cars, gas turbines, power plants, and home heating systems. Highly efficient
combustion can be obtained, free of emissions of smoke, carbon monoxide or fuel.
Unfortunately, this design method can result in high emissions of nitric oxide (NO).
The rate of formation of NO has been found accurately represented by
dNO 14 -67,000/T
——= 6.62 x 10 e
dt
T
1/2
(Ref. 5)
where NO, N£, and 02 are concentrations of nitrogen, oxygen, and nitric
oxide in mole fractions
t is time in seconds
T is temperature, ° K
P is pressure in atmospheres.
-------
X-7
Time can be assumed infinite after one second and the resultant equilibrium values
of nitric oxide are shown in Figure 2. Peak NO emissions occur at slightly greater
than stoichiometric air and fall rapidly when excess fuel is present. The greater the
time spent at top temperature at any air-fuel, the greater is the amount of NO formed
as shown in Figure 3.
In a reducing atomsphere, the lack of excess oxygen permits the use of
catalysts to break down any nitric oxide that is formed and so in an automobile, pre-
sents additional flexibility in control (Ref. 6). In combustion systems using excess
oxygen, catalysts are not practical and the only method of control is to limit the initial
NO formation. The rule governing such a combustor design is:
1) Reaction time must be small in those parts of the combustor where
air fuel is near stoichiometric.
Therefore, combining the rules for efficient and fast combustion with
minimum NO formation, we would have a combustor of small size,
having a high pressure loss, a multiplicity of fuel injection points and
an air-fuel distribution as follows:
a) A well mixed primary fuel-air zone having excess fuel, of
sufficient volume to permit maximum reaction of fuel
b) A secondary fuel-air zone where excess air is rapidly and
uniformly added and having no more volume than is necessary
to assure completion of the combustion reaction
c) A tertiary fuel-air zone (when needed) where any required
excess air is added and where no reaction occurs.
Such rules would apply in gas turbines, Rankine cycle engines, and any other combus-
tion system where air-fuels much greater than stoichiometric are involved. In
Rankine cycle external combustors operating near stoichiometric, some additional
control is permitted because heat losses to the boiler walls can bring down flame tem-
peratures substantially. Often, in such applications, inlet air preheat is used which
has the reverse effect. In gas turbines, because wide operating conditions result in
large changes of air-fuel, there is only one unique point (usually a high power condition)
where the rules can be ideally maintained. A penalty in increased emissions at other
operating conditions is ordinarily accepted and this is the cause of high hydrocarbon
emissions near airports because of ground taxi or start up. Power plants in autos do
not have any unique operating point and typically have to start frequently and undergo
rapid power level changes. In a gasoline engine, it is therefore necessary to precisely
control the air-fuel ratio by means of a carburetor. It is also considered essential in
Rankine engines used in an auto that the airflow into the combustor be controlled so
that the rules for minimum emission are held at any fuel flow.
-------
10
FIGURE 2.
10
AIR/FUEL RATIO
EQUILIBRIUM CONCENTRATIONS BY VOLUME OF CARBON
MONOXIDE AND NITRIC OXIDE AS A FUNCTION OF AIR/
FUEL RATIO
-------
X-9
10.000 i-
i.ooo U
o
10 12 14 16 18 20 22 24
100 U
FIGURE 3. EMISSION OF NITRIC OXIDE AS A FUNCTION OF AIR FUEL
RATIO AND TIME
-------
X-10
4. 2 AIR AND FUEL CONTROL SYSTEM DESIGN APPROACH
The control system was designed to regulate the flow of fuel and air to set
power outputs as a function of an input command from vaporizer control. A fuel ratio
range of 100 to 1 has been established as a design goal for this demonstration system.
Since present state-of-the-art controls have difficulties providing turndown ratios of
more than 15 to 1, this particular requirement has been recognized as a critical design
area requiring unique solutions. In order to optimize the low emission characteristics
of the combustor, it is essential that throughout the entire power range (1 to 100%), an
optimum air-fuel ratio be maintained. An optimum envelope in which the air-fuel
ratio should be centered across the entire power range is discussed in detail in Section
5 and forms the basis for the control.
By directly mechanically coupling a fuel metering valve and the large diam-
eter air metering plate (Fig. 4), the fuel flow area and the air flow area can be kept
in an exact ratio correspondence. Once the area ratios have been fixed, it is only
necessary to regulate pressure drops in order to maintain the desired weight flow
ratios. Figure 5 shows the overall mechanical arrangement of the demonstration com-
bustor system.
For low pressure rise, air compression can be neglected and the weight
flow relationships can be written as:
(1)
Wf = Kf Af j4Pf (2)
where W = mass air flow into combustor
a
A = air metering valve area
P = upstream pressure (blower discharge)
P = downstream pressure
LJ
T = air temperature at metering valve
W = fuel mass flow
AP = pressure drop across fuel valve.
-------
BYPASS VALVE
*t.o
FUEL METfRMG
VALVE 1.0 to IM LB/HH
Af REGULATOR
L_.
FUEL
PUMP
•Ul UK 1
:~i i~
Q]
30 GPH
LV^I
^g P ^msu^
,1 O _-<- A
OK IF ICE
An
-^ Of
p\ '.\\vsn:
L-. . r-
Iw.V.VX.' v
AT,
P /T COMPENSATION
VALVE
- kfcV"*""^-
3EALED CATrTY
FIGURE 4. CONTROL SYSTEM SCHEMATIC
-------
FIGURE 5. DEMONSTRATION SYSTEM
-------
X-13
If the inlet pressure (altitude) and temperature are constant, a simple relationship
can be written for the air fuel ratio:
(3)
To ensure the air-fuel ratio remains a function of flow area, only the pres-
sure drop across the fuel valve will be controlled proportional to the pressure drop
across the air metering ports. A APf regulator valve performs this function by a
force balance across diaphragms. Operation of this component is illustrated in Fig-
ure 4. PI pressure is connected (through a density compensator) to the bottom side
of diaphragm A]_. Downstream pressure (P2) is connected to the opposite side of the
diaphragm. Fuel system pressure is regulated by a flapper valve that recirculates
excess fuel back to the tank. A small diaphragm on the fuel side balances the pressure
across the air metering ports A^ against the fuel pressure. If the air pressure c rop
increases (causing Wa to increase), the force across the system becomes unbalanced
and the fuel pressure regulator moves up; reducing the by-pass flow, and thus increas-
ing the fuel pressure and its mass flow across the fuel metering valve. Since the force
balance must be maintained across the APf regulator, we have:
E Forces = Ag (APf - PJ =\(PI~ Pg)
and where AP is large compared to P
i £
t
where P = P (times) ambient air temperature and pressure
compensation factor
A! -
we have AP = ~—(P -P) (4)
t A2 1 2
Thus equation (3) can be rewritten as
' P2) A2
_ 4 _ z
= constant
Blower speed changes due to voltage or load variations, blower efficiency
reductions due to fouling or wear, and by-pass valve leakage variations are auto-
matically compensated by maintaining the AP as a function of the air valves
pressure drop.
-------
X-14
It can be seen from equation (1) that the air flow is a function of
Thus, an altitude and ambient air temperature correction is applied by adding a
TI compensator (Fig. 4). Blower discharge pressure (P^) is admitted to the AP regu-
lator by a contoured needle valve. This valve is positioned by a force balance across
a diaphragm having P^, T^ on one side and cavity sealed with air at standard conditions
Thus, as the air temperature increases, the sealed air will expand and move the valve
upwards. This action will increase the pressure drop across the contour valve and
thus reduce PI to a corrected value of PI on the bottom side diaphragm A. As PI
decreases, the fuel pressure will drop, producing the desired fuel flow reduction as
the ambient air temperature increases.
Air to the combustor is metered across a series of twelve ports cut into two
plates (Fig. 6). An input lever rotates one plate with respect to the stationary backup
plate. The metering port areas are caused to open or close as the input power lever
rotates matched ports in each plate. By maintaining a known relationship to power
lever position and metering port area, the port contour can be arranged to provide
near linear control as a function of power lever position and voltage with the AP regu-
lator compensating for differences in the range of 1 to 2.5 inches of
The capability of the fuel regulator to control fuel pressure levels propor-
tional to differential metering valve pressures of 1 to 2. 5 inches of H2O is vital to
obtain a wide flow range control with high response that maintains an accurate air fuel
ratio from 1 to 100 percent power demands. Initial analysis indicated that the use of
motor voltage (speed) regulation could not meet the range accurately or the stringent
transients improved by the Federal driving standards. Inertial lags of the motor would
keep it out of synchronization with the fuel flow for an unacceptable percentage of the
standard driving cycle. By maintaining the fuel pressure proportional to the AP across
the air metering plates, the demonstration system allows limited voltage control (to
reduce parasitic losses and noise at normal driving speeds) with precise air-fuel ratio
control under transient conditions. If the by-pass valve is maintained closed for the
power range from 100 to 70 percent, the air flow into the combustor can be accurately
controlled by proportionally lowering the voltage to the blower motor while the regu-
lator compensates for the reduced pressure differential across the air metering valve
at lower speeds. Since a mechanical linkage maintains a known ratio of air valve area
to fuel valve area, the regulator will automatically maintain the air fuel ratio as the
power is reduced. A limitation on this sytem is the lower differential pressure signal
that can maintain accurate fuel pressure regulation. Analysis and test has indicated
that 1. 0-inch of H2O is a reasonable lower limit. Using this value, the part load
-------
X-15
MrmiNG onmcra
TTP. (11) PLCS,
TAN MOTOR
COOLING FINB
FIGURE 6. SWmL ELIMINATION AND PRESSURE EQUALIZATION
BAFFLE ARRANGEMENT
power demands of the combustor system can be established. Although the combustor
requirements are moderate (1.25 HP for the optimum configuration and 2. 3 HP for
the demonstration system), addition of a typical boiler configuration can make parasitic
power losses at part loads unacceptable. If a boiler of the type described in Reference
(11) was used, an additional 2.4 horsepower would be required. Total parasitic power
levels would then be as high as 3. 65 horsepower. If the system were to require this
high power level from full power down to idle condition, a highly undesirable condition
would exist. Lowering the input voltage across a voltage range that allows accurate
AP compensation by the regulator will eliminate the motor inertial speed lags since
the air fuel ratio changes are continually maintained by the regulator. Power reduc-
tion by operating at 1. 0 inch of ^O can be estimated by simple calculations based on
fan laws for a series wound motor. For voltage changes of approximately 2 to 1, the
following relationships give reasonably accurate results.
For a given voltage change V to V
flow Q: Q/QO = V/Vo
1 A L £
-/
pressure p Pi/p2 = (Vi/V2)
fan BHP is: BHP /BHP0 = (V /V
-------
X-16
By maintaining the bypass valve closed and reducing the voltage to the motor,
the flow through the metering valve will drop as a function of both the motor voltage
and area change of the metering valve. Pressure will drop as a function of the square
of flow. By using 1. 0-inch as the lower limit of adequate control, the bypass valve
can be maintained in a closed position until the voltage is reduced to 1/2 or 0. 7077-^
Since power is approximately a cubic function, we will obtain (0.707)3 (BHP), or a
65 percent reduction in power. With the optimum system and a relatively high air
side boiler pressure drop, a parasitic power loss reduction of approximately 2.4 horse-
power. Total power with this high pressure drop boiler would be approximately 1.25
horsepower at vehicle power demands below 35 percent. One difficulty with this
approach is that it is more difficult to make the system linear. However, since this
is not a driver input command linearity, it is of secondary importance if the correct
air-fuel ratios can be maintained.
Fuel must be accurately mete red from 109 to 1. 0 pounds per hour. At low
fuel rates (1. 0 pounds per hour), the flow is approximately two drops per second.
Standard valves do not have sufficient linear range to accommodate these severe
requirements. Additionally, the valve should not be sensitive to temperature induced
fuel viscosity changes. A new approach to this problem has been taken by the applica-
tion of a dual slotted shear valve (Fig. 7) consisting of two flat (ground and lapped)
plates with matched contour slots 90 degrees to each other. At the intersection of the
two slots, a square orifice is formed whose area is a function of the relative position
of the top movable plate. The square shape (and thus the discharge coefficient) can
be maintained constant throughout the entire 100 to 1 area ratio. Since the plates are
in contact, fuel will flow only through the slot in each plate and not between the plates,
thereby reducing the clearance leakage path to the microfinish of the contacting sur-
faces. Figure 8 shows the assembled fuel valve of the demonstration system. A drain
groove is provided between the upstream pressure and the metered outlet fuel passage,
thereby reducing the potential leakage pressure to the level of the frictional flow drop
to the spin cup. Since this is normally less than 0.5 psid, resulting leakage of metered
fuel into the drain system will probably be negligible.
The size of the orifice slots used is a trade-off between four factors.
• Fabrication capabilities - requires large dimensions
• Contamination - requires large dimensions
• Backpressure sensitivity - requires high pressure and thus, small sizes
• Temperature sensitivity - it is desired that changes in fuel temperature
have little effect of the coefficient of discharge. This factor requires
a high Reynolds number and thereby small orifice dimensions.
-------
X-17
10 : 1 SLOT
DRAIN TO TANK
MOVABLE PLATE —
FLOW AREA SHADED
-ROM Fl'El , ( \
REGULATOR (J \
( ri
LJ 4
DENED AND LAPPED __/
VR PLATE SURFACES
h'r
1
V
SPRING
/ POWER LEVER
/ POSITION INPUT
10 PSIG u \j, ^ y^^\
-^^- ' ' ' '^ ' ', ', \\ \
<• - f -^B^l ^V ^^-^ /
J t *. ». ^ J X,_ j*
I /f^»^» ^**"™ "
^ 1
DHA1,N
0 PSIG ^J n«^- TO SPIN CUP
FIGURE 7. FUEL METERING VALVE CONCEPT FOR 100:1 TURNDOWN
4.3 FUEL INJECTION, ATOMIZATION AND IGNITION
The requirements of the fuel injector are:
1) Small droplet size
2) Reliable and low cost
3) Multiple injection points
4) Precise spray angle.
In detail, these requirements entail:
1) Small droplet size must be obtained with kerosene at any fuel flow between
1.0 and 109 pounds per hour with fuel viscosity, varying with ambient
temperature, ranging from less than 1 to over 16 centistokes.
-------
X-1S
FIGURE 8. FUEL METERING VALVE ASSEMBLY
2) The injector must be of proven design, unsophisticated, not requiring
a high fuel pressure pump or other complex or costly auxiliaries.
Fuel orifices must be large to prevent contamination.
3) The object of multiple injection is to provide the maximum interface
of fuel and air and hence provide fastest and most uniform mixing so
important for optimum air-fuel ratio and fastest reaction rate. Pack-
aging in an auto requires, at this stage, maximum design flexibility.
If the combustor were mounted horizontally, being at least 10 inches
in diameter, the manifold head effect would require, for good fuel
distribution, a fuel pressure of over 10, 000 psi, and is impractical.
The mixing must therefore be mainly achieved by good design of air
injection and providing the maximum surface of fuel spray to the air.
4) Precise spray angle is necessary to control fuel air mixing and also
to obtain ignition by assuring that the spark is adjacent to but not
smothered by the fuel droplets. Ignition is easiest at low velocities
and hence should be done at 1.0 pound per hour. (An ignition failure
at 109 pounds per hour would be a dangerous fire risk and serious
polluter.)
-------
X-19
There are four principal classes of fuel injectors (1) fuel pressure, (2) air
assisted, (3) rotary, and (4) vaporizer (Ref. ). The first three operate on the same
principle; i.e., the fuel is presented to the air at a high relative velocity which shatters
the fuel into small droplets. Vaporizers rely on heat input, usually from the flame, to
vaporize the fuel. A design study, adumbrated below, indicates the rotary injector to
be the best choice.
1) Pressure injectors cannot atomize fuel flows as low as 1 pound per hour
except when nonviscous fuels are used. Orifice size would be minute and
fuel pressures above 250,000 psi are needed. Spray angle is extremely
sensitive to viscosity.
2) Air assist injectors use high velocity air to atomize the fuel. There
are two types, one using high air volume and low pressure (8-inch water)
(Ref. 2). The other using high air pressure (several psi) and low volume.
Some of these injectors use such high air pressures that sonic velocities
are reached (Ref. 7). The high air volume system is not practical, be-
cause at low fuel flows, the air flow required would be far too much for
combustion. The high pressure system injects air at such high velocities
(400 ft/sec) that at low fuel flows, where combustor air velocities are of the
order of 1 ft/sec, a flame could not be stabilized. In addition, the spray
angle is sensitive to viscosity, especially at low fuel flows and, coupled
with the high air velocities, makes reliable ignition doubtful. Unless high
fuel pressure is used at high fuel flows (with resultant small orifice sizes)
air flow has to be high and substantial power is required. Both types of
injectors need power absorbing and costly auxiliaries and involve the use
of small air and fuel passages.
3) A rotary atomizer does not need auxiliaries as it can be mounted directly
onto the fan required to deliver combustion air. No fuel pressure or fuel
orifices are needed as the fuel is passed through a large tube (3/16-in.
diameter) that passes through the center of the fan motor shaft, and trick-
led onto the cup surfaces. The fuel adheres to the cup wall and is ejected
from the outer lip of the cup at a velocity of about 40 feet/second. The
power required to drive the cup is negligible. Fuel spray angle is extremely
precise and independent of fuel flow or viscosity (no ignition failures have
occurred in tests to date).
4) A vaporizer requires an auxiliary source of heat for ignition and initial
flame propagation. The response is slow and the heat losses involved at
1 pound per hour would prevent combustion unless auxiliary heating were
continuously supplied. The large variation in heat release makes it
-------
X-20
most difficult to prevent overtemperature of the vaporizer at high flow
rates while maintaining sufficient heat at low flows. It is sensitive to
fuel volatility, hence requires a more precisely refined fuel type than
the rotary injector and flame performance will be significantly influenced
by ambient day temperature (in a cold start it will smoke). These objec-
tions apart, a vaporizer is a desirable system. This is because the
absence of fuel droplets avoids evaporation delay and allows a more
uniform mixing of fuel and air.
4. 4 COMBUSTOR DESIGN
The combustor must be small in size in order to fit in an auto and also to
limit combustion time. The use of high combustion velocities requires high air injec-
tion velocities for proper mixing and hence high pressure loss. A limit of auxiliary
fan power prevented a pressure loss of more than 8 inches of water being available.
This dictated the use of a relatively large combustor (Fig. 5 and 9). The combustor
was a scaled up version of a similar rotating cup combustor used in a Solar 10 KW
gas turbine under development. Construction techniques are typically as used in gas
turbines, except that the best available oxidation resistant material was used (Hastelloy
X). This permitted operation of highest possible wall temperatures and minimized
film cooling. Film cooling, by its very function, implies low rates of mixing (Ref. 9)
and breaks a fundamental rule in mixing of fuel and air in that it must be good.
Air was arranged to enter the combustor through various ports, the air-fuel
ratio calculated, and computer runs made of the resultant emissions. Figures 10 and
11 are typical results. Combustion time at full fuel flow is nominally 0. 01 second,
and this increases to 1.0 second at minimum fuel flow. The resultant increase in NO
emissions at low flows is clearly seen in Figure 11.
V. DEVELOPMENT TESTING
5.1 COMBUSTOR TESTING
Initially, the fan was not used and air was supplied by a remotely located air
compressor that supplied air via a plenum to the combustor. This guaranteed uniform
air distribution. For convenience of test, the cup was driven by a small electric motor
independent of the fan (Fig. 12).
After development of the combustor and cup, it was possible to maintain an
efficient flame over a range of fuel flows from over 109 pounds per hour to less than
0. 3 pounds per hour, provided that the air-fuel ratio was kept within reasonable
limits (Fig. 13 and 14). A range of air-fuels were tested and Figures 15 and 16 show
the emissions of CO and NO at the optimum air-fuel together with the results of a 10
-------
X-21
EXHAUST STACK & LOCATION
FOR BOILER
FIGURE 9. SIDE VIEW OF ROTATING CUP COMBUSTOR ASSEMBLY
percent error in air-fuel. Figure 17 shows the optimum air-fuel required for mini-
mum emissions. During the testing, hydrocarbon emissions were monitored, at
optimum air-fuels, below background levels. This is not surprising in view of the
close proximity of the test facility to the San Diego airport (200 yards). The results
are noteworthy in that:
1) The NO emissions fall off sharply at low fuel flow and the CO emissions
increase. This is due to the high heat losses from the flame at these
conditions.
2) The large increase in emissions in the mid-fuel flow range when any
deviation from optimum air-fuel occurs. This was found caused by
poor fuel atomization that occurred in this region and was subsequently
eliminated by design changes.
3) The large variations in optimum air-fuel as a function of fuel flow. At
high fuel flows, this is partly attributable to the poor fuel atomization
which varied as a function of fuel flow and due to the changes in fuel-air
-------
X-22
FIGURE 10. AIR/FUEL RATIO ALONG COMBUSTOR FROM
COMPUTER ANALYSIS
DISTANCE - INCHES
FIGURE 11. NO» CO CONTENT FROM COMPUTER ANALYSIS
-------
X-23
ROTATING CUP
ROTATING
CUP, MOTOR
FIGURE 12. ROTATING CUP AND MOTOR ASSEMBLY
FIGURE 13. FLAME AT 109 POUNDS/HOUR FUEL FLOW
-------
X-24
FIGURE 14. FLAME AT 0.25 POUND/HOUR FUEL FLOW
mixing that inevitably occur. At low fuel flows, the most efficient
flame resulted when the primary zone was leaned out with excess
air. This shortened the flame, reduced heat losses, and hence pro-
vided more efficiency.
4) The increase in CO as a function of fuel flow. A not unexpected result
in view of the time dependence of the reaction. By extension of the
secondary combustor zone, the CO emissions can be reduced and, pro-
vided good fuel-air mixing has been obtained, NO emissions theoretically
would not increase significantly.
Tests were done with a water cooled heat exchanger mounted on the rear of
the combustor and designed to simulate the vaporizer that, in the final design, must
be used. NO emissions were significantly reduced, CO increased slightly, but hydro-
carbon emissions increased an order of magnitude. It is concluded that heat losses
to the vaporizer must therefore be minimized by shielding it from the flame.
Tests were then done using the fan and control system previously shown in
Figure 4. The initial combustion characteristics were totally different and emissions
unacceptably high. Two main features were noted. At high air flows, circumferential
air maldistributions were so high that raw fuel escaped from the exhaust. At low air
flows a pronounced swirl occurred in the flame. Air distribution circumferentially
was good but axially considerably different from the rig tests. Aerodynamic analysis
-------
X-25
u
U
17
16
15
14
13
12
11
10
I
FIGURE 15.
16.25 - PROPOSED LIMITS 1975-76
Wa nnn = 2835 LB/HR AIR
100%
10 20
30 40
50
AIR FLOW,
60
Wa
Wa"
70 80
±10% FROM OPTIMUM
AIR-FUEL
OPTIMUM AIR-FUEL
j
90 100
100%
EMISSIONS OF CARBON MONOXIDE AS A FUNCTION OF COMBUSTOR
AIR FLOW, AT OPTIMUM Am FUEL FOR MINIMUM EMISSIONS AND
ALSO WITH A ±10% DEVIATION OF AIR FUEL FROM OPTIMUM
-------
JX-26
u
D
fo
o
O1.8
1.6
1.4
X
O 1.2
O
s
0.6
0.4
PROPOSED LIMITS OF EMISSION
± 10% FROM OPTIMUM
AIR-FUEL
OPTIMUM
AER-FUEL
Wa „ 2835 LB/HR OF AIR
jLOu /o
j
10
20
30
40
50
AIR FLOW,
60 70
Wa
Wftioo%
80
90 100
FIGURE 16.
EMISSIONS OF NITRIC OXIDE AS A FUNCTION OF COMBUSTOR AIR
FLOW, AT OPTIMUM AIR FUEL FOR MINIMUM EMISSIONS AND
ALSO WITH A ±10% DEVIATION OF Am FUEL FROM OPTIMUM
indicated these two problems were caused by two design features in the fan design.
The fan used was a highly loaded solid vortex design typical of those used in aircraft
where minimum size and weight is needed (Ref. 8). The velocities in it were high
and to supply air to the combustor, substantial diffusion was required. Diffusion is
unstable and results in circumferential maldistributions. At low air flows, the pres-
ence of slight swirl from the fan results in the establishment of a free vortex in the
combustor of relatively high velocity (the fan runs continuously at full speed). Because
the combustor pressure drop is low (0. 0008-inch water at 1. 0 pounds per hour of fuel)
this has a significant effect on the discharge coefficients of the air metering holes with
consequent axial maldistributions. Aerodynamic problems of this sort are the bane
of combustion design. In gas turbines, not only is efficiency reduced but serious hot
spots are caused. The much wider range of air flows of the Rankine combustor makes
-------
X-27
50
60
70
80
90 100
Wa
FIGURE 17.
Waioo%
AIR-FUEL RATIO FOR MINIMUM EMISSIONS AS A FUNCTION
OF COMBUSTOR Affi FLOW
the problem of precise air-fuel control more critical and the tests indicate it to be a
leading problem if emissions are to be kept low and vaporizer life to be long.
The expedient of introducing a large plenum between fan and combustor
(Fig. 18) provided a temporary solution and rig and fan tests were found to aerody-
namically duplicate each other. This was confirmed by exhaust temperature traverses
of both rig and fan air supplies and which were in close agreement (Fig. 19). The
profile of temperature is indicative of inadequate mixing. Apart from detracting
from vaporizer reliability, it is bound to raise emissions because optimum air-fuels
are not maintained.
-------
X-28
FIGURE 18. FAN AND COMBUSTOR ASSEMBLY UTILIZING LONG MIXING DUCT
5. Or
4.0
a
X
u
5 3.0
CO
s
BJ
U
J
8
2.0
1.0
0.0
FAN AIR SUPPLY
RIG AIR SUPPLY
1400 1600 1800 2000 2200 2400 2600 2800
BOILER INLET TEMPERATURE, °F
FIGURE 19. REPEATABILITY OF RADIAL PROFILE OF TEMPERATURE INTO
BOILER USING BOTH FAN AND RIG AIR SUPPLIES
-------
X-29
As the plenum involved a duct of 36 inches in length, it is not acceptable for
use in an auto. The solution lies in a fan designed specifically for the combustor, of
larger diameter and lower velocity. The air control system would retain the present
features and it is doubtful whether any other air control system, such as a butterfly
valve, would be satisfactory because of the asymmetry of air that would result.
After further development, endurance tests indicated the most significant
problem to be carbon build up in the combustor walls. Temperature and emission
measurements, as well as visual flame observation indicate that improvements in
air-fuel mixing are needed; and it is expected that further development can provide
substantial performance improvements.
5.2 CONTROL SYSTEM DEVELOPMENT TESTS
Control system development tests have been completed on all major compo-
nents. Final fuel metering valve performance tests have demonstrated the gain to be
61 pounds per hour per inch stroke with a linearity of ±2.5 percent across a range of
3 to 115 pounds per hour (Fig. 20). Good repeatability and flow control to 0. 5 pounds
per hour indicates the valve has a dynamic range greater than 200 to 1 (Fig. 21). Fuel
pressure regulator valve (Fig. 22) development tests have extended its range and
improved accuracy. An important goal of extending the control range to 1. 0 inch of
H20 input actuator pressure differential has been achieved. Blower motor power reduc-
tions of as much as 65 percent can now be compensated with the fuel regulator. Correct
fuel flow can be maintained within ±4 percent with actuator input differential pressures
ranging from 1. 0 to 2. 5 inches of H2O (Fig. 23).
Air metering valve problems caused by high dynamic head, swirl, and turning
losses by use of an off-the-shelf blower, required high pressure loss baffles and flow
straighteners to be installed. As a consequence, the blower motor was required to be
operated approximately 6 volts above its normal input. Additionally, combustor tests
with the air valve showed that potential aerodynamic problems existed that would require
considerable development effort to minimize volume. Analysis of these problems
indicated that the best solution would be to specially design a matched fan-motor com-
bination to the functional and geometric requirements of the combustor. It was estab-
lished that the demonstration system should utilize a simple dump plenum to make the
fan more aerodynamically compatible with the air valve and combustor. Thus, a 36-
inch extension between the fan and air valve was incorporated into the demonstration
valve. Figure 24 shows typical performance on bench tests of the air metering valve.
-------
SLOPE - 61 PPH/INCH
PRESSURE DIFFERENTIAL: 10 PSID
FUEL: JP-5
TEMPERATURE: 76'F
O DATA RECORDED 12/11/70
• DATA RECORDED 12/15/70
0.1 U.2
0.4 0.5
0.7 0. s
1.0 1.1 1.2 i.'J
FUEL VALVE POSITION (INCHES)
FIGURE 20. FUEL METERING VALVE FINAL WEIGHT FLOW CALIBRATION
U)
o
-------
2-31
10.0
a.o
8.0
7.0
8
I
OS
H
O,
6.0
H
3.0
2.0
1.0
SLOPE - 64.2 PPH/IN.-
SLOPE • 61 PPH/IN.
PRESSURE DIFFERENTIAL: 10 PStt)
FUEL: JP-5
TEMPERATURE: 76 *F
O DATA RECORDED 12/11/70
• DATA RECORDED 12/15/70
LINEARIZATION REFERENCE POINT (1.0 PPH AT 0.011 INCHES)
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
FUEL VALVE POSITION (INCHES)
FIGURE 21. FUEL METERING VALVE FROM 0.5 TO 10 POUNDS PER HOUR
(FINAL WEIGHT FLOW CALIBRATION)
-------
X-32
fyssssaas^^ -
FIGURE 22. FUEL REGULATOR ASSEMBLED WITH Pl - P2 ACTUATOR
5. 3 EMISSION MONITORING
All emission data presented was taken with Beckman Model 315A, CO, CO2,
and NO analyzers along with Beckman Model 402 Hydrocarbon Analyzer. The emis-
sion monitoring system is designed in accordance with existing vehicular specifications
as defined by Federal Standards, the Automobile Manufacturing Association and the
State of California. It was selected for use in this program because its wide spread
acceptance permits direct comparison of results with other investigators' work. The
CO, CO2, and NO analyzer uses NDIR cells selected for the ranges required for the
program. Table I lists the ranges, repeatability and interference factors that are
characteristic of this measurement system. In order to achieve a 0-150 ppm range
for NO, an exceptionally long (41 inches) NDIR cell is incorporated for this important
parameter. Figure 25 shows the control panel and strip chart data readout employed
with these instruments. Sampling was accomplished with a 1/8-inch diameter cooled
probe for CO, CO2, and NO to prevent reaction in the sampling line. Hydrocarbons
were sampled with a line maintained at 350° F by means of electric heaters.
VL CONCLUSIONS
The 1975-76 emission levels for automobiles are feasible when using a Rankine
engine combustor. Package size and power requirements of the combustor need not be
excessive, the components used are not complex, are capable of low cost mass produc-
tion and, as involve proven concepts, high reliability is attainable. The fuels used
need no critical refining capabilities, and are much safer and less expensive than
gasoline. They can be instantly burned in the coldest weather and do not have any
-------
• 11X1% FLOW
A *C* FLOW
• It FLOW
1.1
1.4 l.j 1.6 1." i.S Lit 2.0
CONT&UL ACTUATOR INPUT PRESSUREt>lFFER£NTUL (INCHES OF WATEB)
2.1
2.3
U)
1.5
FIGURE 23. REGULATOR CALIBRATION WITH 0.008 INCH FLAT RUBBER DIAPHRAGM
-------
100
— I
B
70
r,o
40
CALCULATED
ACTUAL I /
TEST WITH 2.25 WIDE
BY-PASS VALVE
(BAND ON ORIFICE
PLATE SIDE)
j i
Ki
0 ^- 1
0 10 20
7% FtOW@ 7. 5% PLP
«Me>KE VELOCITY
5% FLOW @ 5.9% PLP :CALIBRATIONS
2%FLOW (§!3.8% PLP | | j I
r>0 (JO 70 80 90 100
30 40
POWER LEVEL POSITION (PLP) %
FIGURE 24. FLOW CONTROL PERFORMANCE OF AIR METERING VALVE
TABLE I
INSTRUMENTATION ACCURACY USED FOR DATA ANALYSIS
Constituent
Gas
CO
C°2
NO
H C
x y
Ranges
0-1000 ppm by Vol.
0-2.5% by Vol.
0-250 ppm by Vol.
0-16% by Vol.
0-5% by Vol.
0-1000 ppm by Vol.
0-150 ppm by Vol.
0-5 ppm
16 ranges through
0-250,000 ppm
Repeatability
% of full Scale
1.0
1.0
1.0
1.0
1.0
±1.0
±2.0
<±5
Interference
Interfering
Gas
co2
H20
CO
co2
Mole
Percent
5.0
3.3
5.0
5.0
—
Extraneous
Response
3.4 ppm
1. 1 ppm
1 ppm
3 ppm
---
Calibration
Gas Accuracy
±r,%
±5%
±5%
±5%
-------
X-35
FIGURE 25. TOTAL VIEW OF BECKMAN GAS ANALYZER INCLUDING THE
FLAME IONIZATION DETECTOR
warm-up emissions as occurs on current engines. They need no special additives for
combustion control, nor will the current automotive evaporative control systems be
necessary.
Development is necessary, especially in regard to integration with other
engine components because they can affect both the shape of final performance of the
combustor and particularly, the critical aerodynamic interface necessary for low
emissions. The most critical area is in air-fuel control. To achieve the best possible
performance (and the results, while below current goals, could be substantially im-
proved) requires aerothermodynamic sophistications more typical of advanced gas
turbine combustors than conventional heating systems.
Further development of the combustor components would yield substantial
benefits in both size and power requirements (Fig. 26). However, if an economic
vaporizer of reasonable size was used (Ref. 11), the horsepower requirements of the
combustor would increase from 1. 25 to 3. 65. By developing the inherent variable fan
speed capability of the fuel air control system, the parasitic losses at part loads could
be maintained at approximately 1 horsepower. Typically, maximum power is required
only transiently and normal power output is 35 percent or less than maximum. This
suggests that the design should be optimized for the normal engine load demands of 35
percent power or less.
-------
If the motor speed was reduced to supply the air required for 35 percent
power, the total power required of combustor and vaporizer would be reduced from
3.65 to 1.25 horsepower, and full fan power would only be required in occasional
engine power demands above 35 percent where fan speed would be increased. Such
a design concept provides considerable potential for combustor improvements as it
permits the use of higher combustor pressure losses and a smaller combustor.
-4 — 31-
N
/
1.16 FTJ
COMBUSTOR
PLUS CASE
Tc
'r
••
^n r~"
UP j
OTOR| !
-1 i
12.7f, DIA
(J. M FT3
AIR VALVE
|_ ,
...
FAN
MOTOR
FAN
'
1
7 1
^ r
nw
-3
PANCAKE
DC MOTOR
FUEL INLETI
FAN RLADE
FAN STATOII
CONTROL SECTION
TOTAL VOLUME 1 89 FT"
DEMONSTRATION SYSTEM
OPTIMUM SYSTEM
3
Total Volume (ft )
Length (Inch)
Diameter (Inch)
3
Combustor Volume (ft )
Combustor Diameter (Inch)
Horsepower
Combustor Loss (Inch Water)
Diffuser Loss (Inch Water)
Metering Loss (Inch Water)
Overall Pressure Loss (Inch Water)
Fan Efficiency (%)
Motor Efficiency (%)
Overall Efficiency (%)
Motor Speed (rpm)
Demonstration
System
Design
1.89
31.0
13.00
0.687
11.00
2.30
8.00
3.00
2.00
13.00
75.0
75.0
56.2
14000
Optimum
System
Design
1.09
14.25
13.00
0.687
12.00
1.25
6.0
0.5
1.5
8.0
85.0
75.0
68.0
7000
Spec
1.33
2.00
Note: Ignition, Fuel and Air Regulators not included.
FIGURE 26. TWO FAN SKETCHES - PRESENT AND OPTIMUM
-------
X-37
VII. REFERENCES
1. Lefebvre, A. H., "Theoretical Aspects of Gas Turbine Combustion
Performance". Co A Report Aero No. 163, The College of Aeronautics,
Cranfield, England (1966).
2. Lefebvre, A. H. and Miller, D., "The Development of An Air Blast
Atomizer for Gas Turbine Application". Co A Report Aero No. 193,
The College of Aeronautics, Cranfield, England (1966).
3. Spalding, D. P., "Performance Criteria of Gas Turbine Combustion
Chambers". Bunhill Publications, Ltd., London, England (1956).
4. Hottel, H. C., Williams, G. C, and Miles, G. A., "Mixedness in the
Well Stirred Reactor". Eleventh Symposium on Combustion, 1967,
L. C. Card 55-9170.
5. Caretto, L. S., Sawyer, R. F. and Starkman, E. S., "The Formation of
Nitric Oxide in Combustion Processes". Central States Section/Combustion
Institute (1968).
6. Yolles, S., Wise, H. and Berriman, L. P., "Study of Catalytic Control
of Exhaust Emissions for Auto Cycle Engines". Stanford Research
Institute (1970).
7. Hawthorne, W. R. and Olson, W. T., "Design and Performance of Gas
Turbine Power Plants". Princeton University Press, 1960, L. C.
Card 58-5027.
8. De Kovats, A. and Desmur, G., "Pumps, Fans and Compressors".
Translated by R. S. Eaton, M. A. Blackie and Sons, Ltd., London,
England (1958).
9. Clarke, J. S. and Jackson, S. R., "General Considerations in the Design
of Combustion Chambers for Aircraft and Industrial Gas Turbines".
Joseph Lucas, Ltd., Burnley, England (1962).
10. Federal Register, Vol. 35, No. 136, July 15, 1970 and No. 219, Nov. 10, 1970.
11. Strack, W. C., "Condensers and Boilers for Steam Powered Cars;
A Parametric Analysis of Their Size, Weight and Required Fan Power".
NASA TND-5813, Lewis Research Center, Cleveland, Ohio (1970).
-------
XI-1
Chapter XI
HYBRID HEAT ENGINE/ELECTRIC SYSTEMS STUDY
by
Joseph Meltzer
Director of Pollution and Resources Programs,
The Aerospace Corporation
El Segundo, California, USA
-------
XI-2
This briefing summarizes the results of a comprehensive broadbased
study aimed at determining the feasibility of using a hybrid heat engine/
electric propulsion system as a means of reducing exhaust emissions from
street-operated vehicles. In this hybrid concept, the source of power is a
combination of heat engine and batteries (in essence, the heat engine
supplies steady state power and the batteries supply transient power
demands). The study examined—for several classes of vehicles--many
types of heat engines, batteries, and other major components, as well as
several design configurations. Following a review of the associated tech-
nologies, hybrid performance, exhaust emissions, and major component
requirements were determined. Based on these results, recommendations
can be formulated to ensure the development of critical powertrain com-
ponents for an early demonstration of prototype vehicles.
In the propulsion of the hybrid heat engine/electric vehicle, the
ultimate source of all energy to be expended is the heat engine. The key to
success in reducing exhaust emissions is good part-load and full load
efficiency of powertrain components, and the ability to restrict operational
requirements £or the heat engine to those of supplying road load power and
(in conjunction with a generator) recharging advanced high power/high
energy batteries that supply acceleration power. With this idea in mind,
the study was tailored to examine six classes of vehicles: the 4000-lb family
car, 1700-lb commuter car, low- and high-speed postal/delivery van, and
low- and high-speed intracity bus. For each class of vehicle, five engines
were included in the powertrain: spark ignition, compression ignition, gas
turbine, Rankine cycle, and Stirling cycle. Lead-acid, nickel-cadmium,
and nickel-zinc batteries were studied for adequacy in supplying acceleration
power to each vehicle. Also, a wide range of ac and dc motors, generators,
and power conditioning and control systems were evaluated for performance
efficiency, weight, simplicity, and cost.
-------
XI-3
Throughout the study, the following ground rules prevailed:
• The hybrid vehicle should match the conventional
automotive vehicle in acceleration, speed, gradeability,
curb weight, and powertrain weight.
• External recharge of the battery should not be required.
This requirement was simulated in computations by
requiring that the heat engine-driven generator recharge
the battery to the original state-of-charge prior to the
end of a selected emission driving cycle.
• The battery is to discharge only when the vehicle is
undergoing acceleration, not on a smooth grade or at
cruise conditions.
• The heat engine is to supply steady road load power
and is not required to undergo rapid acceleration.
• Only design concepts compatible with near term
(1972-1975) prototype vehicle development are to be
considered.
The following set of charts summarizes the study content, parameters
examined, and the results. Some of the results are highlighted in the text
below for the family and commuter car.
• Only the spark ignition internal combustion engine and
the gas turbine engines can be practically packaged
into the hybrid heat engine/electric vehicle with the
performance specified in the study.
• All hybrids examined showed marked emission reduc-
tions over current conventional vehicles.
• If currently available technology--not including catalytic
converters--is used, no version of the family car could
meet 1975 emission standards (HC = 0.46 grams/mile,
CO = 4.79 grams/mile and NO- =0.4 grams/mile).
• If advanced technology--including the catalytic converter
for the internal combustion engine--is used, all versions
but the diesel could meet 1975 standards (except for the
minor NO2 excess for the spark ignition family car
version).
• Commuter car emissions are less than one-half of those
for the family car and with advanced technology easily
meet the 1975 standards. (The commuter car weighs
only 1700 Ib and has reduced acceleration and maximum
cruise speed capabilities.)
-------
XI-4
Emissions are approximately 10 and 15 percent lower
for the parallel powertrain configuration as compared
to the series configuration in the family and commuter
cars, respectively.
Study results are based on hot start data. Incorporation
of cold start effects would still allow the advanced tech-
nology versions of some hybrid vehicles to meet
1975 standards.
Regenerative braking has essentially no effect on
emissions.
Motor and engine part-load characteristics and motor
efficiency are extremely important in determining
emissions.
Improved lead-acid batteries could be used in near
term hybrids. Increased energy density and power
density capabilities are needed. Battery lifetime and
improved charge acceptance are the most critical areas
requiring improvement. Nickel-zinc looks promising
for the post-1975 period.
Battery charge acceptance characteristics play an
extremely important role in determining resultant
vehicle exhaust emissions.
Hybrid fuel consumption is about the same as for the
current conventional car.
For the hybrid using a spark ignition engine, high-
production costs would range from 1.4-1. 6 times
todays car.
-------
XI-5
PURPOSE OF STUDY
TO DETERMINE :
• FEASIBILITY OF HEAT ENGINE/ELECTRIC HYBRID AUTOMOTIVE VEHICLES
• POTENTIAL REDUCTION IN AUTOMOTIVE EXHAUST EMISSIONS
• MOST PROMISING DESIGN CONCEPT IN EACH VEHICLE CLASS:
• FULL-SIZE FAMILY CAR
• SMALL COMMUTER CAR
• DELIVERY AND POSTAL VAN
• CITY BUS
TO RECOMMEND '
• TECHNICAL DEVELOPMENT PLAN FOR CRITICAL COMPONENTS
• TECHNICAL DEVELOPMENT PLAN TO ASSURE PRODUCTION VEHICLE
IN 1975 -1980 PERIOD
• SCHEDULE
• RESOURCES ALLOCATION
• MILESTONES (e.g. TEST BEDS, PROTOTYPES )
THE HYBRID CONCEPT
• POWER FOR PROPULSION IS SUPPLIED BY TWO SOURCES : HEAT
ENGINE AND BATTERIES
• POWER FOR ACCELERATION IS SUPPLIED BY THE BATTERIES -
CRUISE POWER IS SUPPLIED BY THE HEAT ENGINE
• HEAT ENGINE SIZE CAN BE REDUCED
• HEAT ENGINE OPERATES OVER RESTRICTED RPM RANGE
• ENGINE DESIGN AND OPERATION CAN BE OPTIMIZED TO REDUCE
EXHAUST EMISSIONS
• ALLOWS INTERMEDIATE STEP BETWEEN CURRENT INTERNAL
COMBUSTION AND PRACTICAL ALL-ELECTRIC VEHICLE OF 1985-1990
-------
XI-6
ADVANTAGES OF OPERATING HEAT ENGINE IN HYBRID VEHICLE
• RAPID ENGINE ACCELERATION REQUIREMENT IS REMOVED
• EXPECT NO STUMBLE FROM LEAN ENGINE OPERATION
• EXPECT IMPROVED EXHAUST EMISSIONS
• RPM AND LOAD RANGE ARE RESTRICTED
• EXPECT IMPROVED COMBUSTION AND FUEL CONSUMPTION
• EXPECT IMPROVED EXHAUST EMISSIONS
• REDUCE DESIGN REQUIREMENTS FOR CATALYTIC CONVERTER
SCHEMATIC OF HYBRID CARS
SERIES CONFIGURATION
PARALLEL CONFIGURATION
HEAT
ENGINE
L
GENERATOR
CONTROL
SYSTEM
t
BATTERIES
GEARING
MOTOR
, WHFFl
-------
XI-7
VEHICLE COMPONENT ARRAY
ENGINES
• I.C. SPARK
• DIESEL
• GAS TURBINE
• RANKINE CYCLE
• STIRLING CYCLE
BATTERIES
• LEAD-ACID
• NICKEL-CADMIUM
• NICKEL-ZINC
MOTORS
• A.C. INDUCTION
• D.C. SHUNT WOUND/
EXTERNALLY EXCITED
• D.C. COMPOUND
• D.C. SERIES
• D.C.BRUSHLESS
GENERATORS
• DC
• A.C. (ALTERNATOR)
POWER CONDITIONING AND CONTROL
• SILICON CONTROLLED RECTIFIERS
• INVERTERS
• SOLID STATE INTEGRATED CIRCUITS
• CYCLOCONVERTER
• RELAYS/SWITCHES
• RESISTORS/INDUCTORS
HYBRID VEHICLE SPECIFICATIONS
FAMILY
CAR
V MAX (miles/hr) 80
V GRADE AT GRADE
(miles/hr, AT %) 40 AT 12
RANGE (miles) 200
LOADED WEiGHT(lb) 4,000
COMMUTER
CAR
70
33 AT 12
50
1,700
INTRA - CITY
BUS
40
6 AT 20
200
30,000
DELIVERY /POSTAL
VAN
40
8 AT 20
60
7,000
ASSIGNED POWER
TRAIN WEIGHT (Ib) 1,500
ASSIGNED POWER
TRAIN VOLUME (ft
28
600
16
6,000
175
I ,700
42
ACCELERATION
EQUAL TO CONTEMPORARY AUTOMOTIVE VEHICLE
-------
XI-8
DRIVING CYCLES FOR EMISSION COMPARISONS
60
•&
FAMILY AND E
COMMUTER £
9s
CAR
7.5 miles
INTRA-CITY
BUS
DELIVERY
VAN
60
•s.
»
£
200 400 600 800
TIME , sec
0 1 miles
1 | |
0 10 20 30 40 50
TIME, sec
60
0 2 miles
20
40 60
TIME, sec
80 100
1000 1200 1400
DESIGN DRIVING CYCLE, 4000-lb FAMILY CAR
MAXIMUM ACCELERATION
- HIGH SPEED CRUISE
- HIGH SPEED CRUISE
FOR RANGE
12% GRADE
10,330 10,340
TIME, »c
11,061 11,075
-------
XI-9
POWER REQUIREMENTS FOR 4000-Ib FAMILY
CAR AT MAXIMUM ACCELERATION
( AT THE POWER CONVERTER OUTPUT )
POWER TO ACHIEVE
GRAOEABILITY
REQUIREMENT
ROILING RESISTANCE PLUS
AERODYNAMIC DRAG
40 50 60
VELOCITY, mph
SERIES CONFIGURATION - VARIATION OF HEAT ENGINE
POWER WITH VEHICLE SPEED
LEVEL ROAD
POWER
DELIVERED
EXCESS TO BATTERY
AND/OR ENERGY
DUMP CIRCUIT-,
MINIMUM
ALLOWABLE
POWER-
POWER REQUIRED FOR
STEADY ROAD LOAD
VEHICLE SPEED, mph
-------
XI-10
ELECTRICAL CONTROL SCHEMATIC, SERIES CONFIGURATION
HEAT f
ENGINE V
T
I
i
I
L.
\_ ALTERNATOR^.
1 RECTIFIER
1 t
i i i
1 ' If
1 L-
| j
BATTERY
VOLTAGE | i
. . J. _,. .
1 "
1
i
- CONTROL SYSTEM
POWER
* LEVEL
T
I FOOT
i PEDAL
* 1
-, T 1 1
If ' '
_ J |
VOLTAGE J
, 1
RPM IQlIf
V WHEELS
L/
MECHANICAL POWER
^^- ELECTRICAL POWER
SENSING OR CONTROL
WEIGHT COMPARISON FOR ELECTRIC GENERATORS
10-
i i T i i i i r
I10
I0l L
DC-
10 I02
CONTINUOUS RATED POWER, kW
-------
XI-11
WEIGHT COMPARISON FOR ELECTRIC MOTORS
1400
1200
1000
. 800
J
600
400
200
DC MOTOR + SCR CONTROLLER
DC MOTOR
AC MOTOR + INVERTER
+ COOLING SYSTEM
FAMILY CAR
I A I
0 50 100 150 200
CONTINUOUS RATED POWER, hp
250
HYBRID VEHICLE PERFORMANCE EVALUATION
DIGITAL COMPUTER PROGRAM
CASE INPUT
HFAT ENGINES EMISSIONS
VEHICLE WEIGHT
VEHICLE VELOCITY HISTORY
BATTERY CAPACITY
BATTERY RECHARGE EFFICIENCY
REGENERATIVE BRAKING EFFICIENCY
TRANSMISSION EFFICIENCY
GEAR TRAIN EFFICIENCY
GEAR RATIO
MINIMUM GENERATOR CURRENT
ACCOUNTABILITY
HEAT ENGINE POWER
AUXILIARY POWER
GENERATOR EFFICIENCY
SYSTEM VOLTAGE
MOTOR EFFICIENCY
MOTOR SPEED
MOTOR TORQUE
VEHICLE ACCELERATION
VEHICLE VELOCITY
TRACTIVE EFFORT
STEADY ROAD HORSEPOWER
VEHICLE KINETIC ENERGY
ROAD GRADE
AERODYNAMIC DRAG
TIRE PRESSURE
TIRE ROLLING RADIUS
BATTERY CHARACTERISTICS
(INCLUDING MAX 8 MIN
ALLOWABLE VOLTAGE)
OUTPUT
VEHICLE EMISSIONS
BATTERY STATE-OF-CHARGE
BATTERY CURRENT
BATTERY VOLTAGE
GENERATOR CURRENT
MOTOR CURRENT
ENERGY TO ROAD
ENERGY TO BATTERY
ENERGY DISSIPATED
TIME
DISTANCE TRAVELLED
-------
XI-12
BATTERY DISCHARGE CHARACTERISTICS
DURING DHEW URBAN DRIVING CYCLE
4000-lb FAMILY CAR
GENERATOR OUTPUT =38Amp
38 AH LEAD-ACID BATTERY
IUO
»* 99
V 98
o
97
96
I
200
400
600 800
ELAPSED TIME, sec
1000
1200
1400
LEAD-ACID BATTERY DEVELOPMENT GOALS
4000-lb FAMILY CAR
DHEW EMISSION
DRIVING CYCLE
38 AH BATTERY OPERATING
FOR 1370 sec
7.5 VEHICLE MILES WITH
73 BATTERY CHARGE /
DISCHARGE CYCLES AND
1.34 AH DEPTH OF
DISCHARGE
DESIGN DRIVING CYCLE
38 AH BATTERY DELIVERED
462 Amp
1
HYBRID VEHICLE-FAMILY
CAR DEVELOPMENT GOALS
38 AH BATTERY AT LESS THAN
5% DEPTH OF DISCHARGE TO
DELIVER UP TO 500 Amp
5000 hr OF OPERATION
AND 100,000 VEHICLE
MILES WITH 975,000
CHARGE / DISCHARGE CYCLES
-------
XI-13
COMPARISON OF ENERGY/POWER DENSITY CHARACTERISTICS
OF LEAD-ACID AND NICKEL-ZINC BATTERIES
WITH DESIGN GOALS
POWER DENSITY,Watt/lb
300
200
100
ADVANCED LEAD-ACID
BATTERY DESIGN GOALS
4000-lb FAMILY CAR
1500-Ib AVAIL ABLE
POWERTRAIN WEIGHT
395-lb BATTERIES
10% DEPTH OF DISCHANGE
SU LEAD-ACID
NICKEL-ZINC
10 20
ENERGY DENSITY, Walt-hr/lb
30
CYCLE LIFE OF LEAD-ACID BATTERIES
I06
10-
10'
I I I r
GOULD-NATIONAL
\ BIPOLAR (1968)
\
\
\
EAGLE-PICHER
MILK TRUCK
SIMULATION
N ^-STATE-OF-ART
\
HYBRID
GOALS —
SLI
REQUIREMENT
(SAE TEST)
PRESENT
CAPABILITY ESB
ESB
PROJECTION
20 40 60 80
DEPTH OF DISCHARGE, percent
100
-------
XI-14
CHARACTERISTICS OF SECONDARY BATTERIES
CHARACTERISTIC BATTERY TYPE
RELATIVE COST*
DEMONSTRATED CAPABILITY**
POWER DENSITY, W/lb
ENERGY DENSITY, W-hr/lb
CYCLE LIFE AT DEPTH OF DISCHARGE 137,400 at 6.7% 34,000 at 25% 190 at 100%
* BASED ON ACTIVE MATERIAL COST
**POWER AND ENERGY DENSITIES NOT DEMONSTRATED SIMULTANEOUSLY
LEAD-ACID
1.0
328
23.3
NICKEL-CADMIUM
II. 1
450
18
NICKEL-ZINC
2.5
180
22
COMPARISON OF HEAT ENGINE CHARACTERISTICS
(89HP ENGINE)
GASOLINE
DIESEL
GAS TURBINE
RANKINE CYCLE
STIRLING CYCLE
WEIGHT
NUMBER
320
756
294
802
1090
VOLUME
FT3
11.4
17.3
9.1
12.7
21.6
SFC
*IBHP-HR
.50
40
.57
.85
42
-------
XI-15
EFFECT OF AVAILABLE POWERTRAIN WEIGHT ON
BATTERY DESIGN GOALS
4000-lb FAMILY CAR
1000
800
600
400
200
PEAK DEMAND 93.2 kW
DIESEL
.STIRLING
1000 1400 1800 3200
AVAILABLE POWERTRAIN WEIGHT, Ib
PEAK DEMAND = 0.41 kW-hr AND
BATTERY DISCHARGED 10% OF CAPACITY
£ 100
I 80
£ 60
£ 40
§ 20
DIESEL
STIRLING
§ 0-
8 1000
1400 1800 2200
ft/AILABLE POWERTRAIN WEIGHT, Ib
EFFECT OF BATTERY RECHARGE EFFICIENCY
ON BATTERY DESIGN GOALS
4000-lb FAMILY CAR
FINAL STATE-OF-CHARGE INITIAL STATE-OF-CHARGE
AT IJB =0.7, PEAK ENERGY DENSITY = 10.4 W-hr/lb
1.15
1.10
NORMALIZED I 05
BATTERY
PEAK ENERGY
DENSITY 1-00
(W-hr/lb)/
(W-hr/lb)vro% 0.95
0.90
0.85
50
60 / 70 80 90
"RECHARGE EFFICIENCY (T/B),%
-------
XI-16
INSTALLED BATTERY REQUIREMENTS AND
PROJECTED BATTERY CAPABILITIES
T
VEHICLE
WEIGHT = 4000 Ib
NICKEL-ZINC
BATTERY
INSTALLED
BATTERY REQUIREMENTS -,
FAMILY CAR
S. I. ENGINE
SERIES CONFIGURATION
0 10 20 30 40 50
MAXIMUM INSTALLED ENERGY DENSITY, W-hr/lb
TYPICAL SPARK IGNITION ENGINE EXHAUST EMISSIONS
vs AIR/FUEL RATIO (GASOLINE)
_ CURRENT
SPARK IGNITION
ENGINE A/F RANGE
14 16
AIR/FUEL RATIO
22
-------
XI-17
SPARK IGNITION ENGINES (GASOLINE) HYDROCARBON
EMISSION, STEADY STATE DESIGN LOAD
I I I
I
I I I I
PROJECTED TECHNOLOGY-LEAN
A/F = I9
A/F=I5 =
PROJECTED TECHNOLOGY
A/F= 22, CATALYST, RECIRCULATION
i i i
10 I02
DESIGN BRAKE HORSEPOWER
10'
HEAT ENGINE EXHAUST EMISSIONS, HYDROCARBON -
STEADY STATE DESIGN LOAD, LARGE ENGINES (>50hp)
I.U
0.8
fo.6
O-
CD
1 —
§.
^0.4
0.2
Q
i i i i i
i— I i 1 STATE-OF-THE
—
—
-
—
i
2!
u_
—
-------
XI-18
HYBRID (HEAT ENGINE/ELECTRIC) VEHICLE EMISSIONS SUMMARY
4000-Ib FAMILY CAR/DHEW CYCLE*
CARBON MONOXIDE EMISSIONS
8.0 r
5.0
„ 4.0
8
2.0
1.0
>
-
-
-
-
£
1 — .
*AIR CONDITIONER NOT ON. HEAT ENGINE OPERATED
CONTINUOUSLY AT 18.4 Bhp (20.8% OF DESIGN
POWER LEVEL OF 88.8 Bhp )
\- 1975 STANDARD
Q CURRENT TECHNOLOGY
Sk
-
-------
XI-19
HYBRID (HEAT ENGINE/ELECTRIC) VEHICLE EMISSIONS SUMMARY
4000-lb FAMILY CAR/DHEW CYCLE*
HYDROCARBON EMISSIONS
1.15
0.8
2 0.6
e
0.4
0.2
7
I,
I 1 CURRENT
1 ' TECHNOLOGY
PROJECTED
TECHNOLOGY
* AIR CONDITIONER NOT ON. HEAT ENGINE OPERATED
CONTINUOUSLY AT 18.4 Bhp(20.8% OF DESIGN
POWER LEVEL OF 88.8 Bhp )
1975 STANDARD
Ik
rh
S.I. ENGINE
DIESEL GAS TURBINE RANKINE
STIRLING
VEHICLE EMISSION COMPARISON
CONVENTIONAL OPERATION vs HYBRID OPERATION
SPARK-IGNITION ENGINE
50
40
30
EMISSION LEVEL,
grams/mile
20
10
1970
CONVENTIONAL
s S.I. ENGINE
- 8
8
7— MOHF HHF
1 ' CYCL
* . •
CONVENTIONAL
S.I. ENGINE
(A/F=I5-I6) + RECIRC.
CONVENTIONAL
S 1. ENGINE LEAN
(VARIABLE A/F) OPERATION,
' 3 ' NO RECIRC. miurFn
o IA/F-IQ1 ADVANCED
n ~ ' .''3| - TECHNOLOGY
° PLUS
o^o ^ ^A/F= 22+ CAT.+ RECIRC.,
W
_§
CONVENTIONAL HYBRID VEHICLE /DHEW CYCLE
VEHICLE (4000-lb FAMILY CAR)
-------
XI-20
COMPARATIVE EMISSION LEVELS OF THE FAMILY AND COMMUTER CAR
120
110
100
80
0
5 70
z
2 60
CO
£ 50
&40
i —
cc
if 20
10
—
__
u
~ •
•
1
•
~L
r
P
^
^
^
%
//
'/A
I
1975 STANDARD
—
1975 STANDARDS
fa HC -
r
%
%
///
I
'//
'//
//
'//
CO -
N02-
046 gm/mi
4.7 gm/mi
0.4 gm/mi
ADVANCED TECHNOLOGY ~~
WRALLEL CONFIGURATION
t
S.I. GAS TURBINE S.I
Sj
i :
-
GAS TURBINE
«^^y
FAMILY CAR
COMMUTER CAR
EFFECT OF BATTERY RECHARGE EFFICIENCY ON N02 EMISSIONS
FAMILY CAR-SERIES CONFIGURATION-PROJECTED TECHNOLOGY
STIRLING
RANKINE
1C. ENGINE
S.I. ENGINE
50 60 70 80
RECHARGE EFFICIENCY!^),'
90
100
-------
JXI-21
EFFECT OF VEHICLE WEIGHT ON N02 EMISSIONS
FAMILY CAR/DHEW CYCLE-PARALLEL CONFIGURATION-CURRENT TECHNOLOGY
1.31 r
1.2
§1
Q "-- in
UJ —. '-U
2*
0.9
-GAS TURBINE
I.C. ENGINE
RANKINE
STIRLING
'DIESEL
0.8
4000 4100 4200 4300 4400 4500 4600 4700 4800
VEHICLE WEIGHT, Ib
EFFECT OF BATTERY CAPACITY AND TYPE ON HC, CO, AND N02 EMISSIONS
FAMILY CAR/DHEW CYCLE-SERIES CONFIGURATION
e. u
ATIO,
5 STANDARDS
(T>
D EMISSION F
/(gm/milelij-
D —
D ro
kl y
^ -^
Z §. 04
°c
1 1 1 1 1
-
r \^
- A^_
1975 STANDARDS
HC -0.46gm/mile
_ CO -4.7 gm/mile
-
1 , 1
10 20 30
1 ' ' 1 ' 1 '
S.I. ENGINE/PROJECTED TECHNOLOGY
r BASELINE BATTERY CAPACITY USED
FOR FAMILY CAR EMISSION
CALCULATIONS
I
IIC
i Ni-Zn BATTERY
0 Ni-Cd BATTERY
Y 1 , ,1,1,
40 50 60 70 8C
BATTERY CAPACITY, Amp-hr
-------
XI-22
EFFECT OF DRIVE MOTOR EFFICIENCY ON N02 EMISSIONS
FAMILY CAR/DHEW CYCLE-PARALLEL CONFIGURATION-CURRENT TECHNOLOGY
,„-??
1.40
120
¥ I. 1.10
" 1.00
0.90
0.80
DIESEL
/STIRLING - DIESEL
RANKINE
I.C. ENGINE
GAS TURBINE
40 50 60 70 80
MOTOR EFFICIENCY (ijj,
90
100
HYBRID VEHICLE FUEL CONSUMPTION AND
PROJECTED PRODUCTION COSTS
FUEL CONSUMPTION
VEHICLE
COMMUTER CAR
FAMILY CAR
LOW SPEED VAN
HIGH SPEED VAN
LOW SPEED BUS
HIGH SPEED BUS
SERIES CONFIG.
(mi /go I)
26
I I
3.75
4
1.25
I 50
PARALLEL CONFIG
(ml/pal)
30.5
12.5
HIGH-PRODUCTION COSTS COMPARED TO CONVENTIONAL CAR
VEHICLE
CURRENT CONVENTIONAL CAR
HYBRID CAR
SPARK IGNITION
DIESEL
GAS TURBINE
RANKINE
STIRLING
RELATIVE COSTS
I
1.4 -1.6
1.5 -1.7
-1.6
2+
2.25+
-------
XI-23
SUMMARY
• HYBRID HEAT ENGINE/ELECTRIC FAMILY CAR
• SHOWS MARKED REDUCTION IN EMISSIONS OVER TODAY'S CONVENTIONAL
HEAT ENGINE-DRIVEN AUTOMOBILE
• FOR NEAR-TERM DEVELOPMENT, THE I.C. SPARK IGNITION ENGINES
OFFER MEANS FOR MEETING 1975 STANDARDS BY PERMITTING
LEAN OPERATION
• FOR THE FUTURE, THE GAS TURBINE CAN EXCEED 1975 STANDARDS
FOR ALL EMISSIONS.
• FOR THE FUTURE, RANKINE CYCLE AND STIRLING CYCLE CAN FAR
EXCEED 1975 STANDARDS, BUT WEIGHT AND VOLUME CONSIDERATIONS
MAY LIMIT THEIR USE. EXTENSIVE DEVELOPMENT ACTIVITY IS REQUIRED
• SIGNIFICANT IMPROVEMENTS IN ELECTRIC MOTOR PERFORMANCE
APPEAR TO BE READILY ACHIEVABLE
• LEAD-ACID BATTERY TECHNOLOGY IS AVAILABLE (BUT NOT BEING
PRODUCED) WHICH CAN SATISFY HYBRID NEEDS FOR POWER AND ENERGY
DENSITY REQUIREMENTS BUT MARKED IMPROVEMENT IS NEEDED IN CYCLE
LIFE AND IN CHARGE ACCEPTANCE CHARACTERISTICS
• HYBRID HEAT ENGINE/ELECTRIC FAMILY CAR
• PARALLEL vs SERIES POWERTRAIN CONFIGURATION
• PARALLEL HAS 10-15% LOWER EMISSION
• THE PARALLEL CONFIGURATION IS A SLIGHTLY MORE COMPLEX SYSTEM
WITH LESS DESIGN FLEXIBILITY
• BATTERY RECHARGE EFFICIENCY
• SMALL EFFECT ON EMISSIONS SHOWN OVER A REALISTIC RANGE
OF EFFICIENCY
• REGENERATIVE BREAKING EFECTS
• SLIGHTLY DECREASES DEPTH OF DISCHARGE
• NEGLIGIBLE EFFECTS ON VEHICLE EMISSIONS
• TOTAL VEHICLE WEIGHT
• BATTERY DESIGN GOALS ARE VERY SENSITIVE TO POWER SYSTEM
WEIGHT ALLOCATION
• COLD START EFFECTS
• FOR HC AND CO INCREASE FROM 15 TO 50% ( /VERAGES ABOUT 30%)
• FOR NOX SLIGHT DECREASES
• HYBRID HEAT ENGINE/ELECTRIC COMMUTER CAR
• LOW-WEIGHT REDUCED-PERFORMANCE CAR CAN REDUCE EMISSIONS
BY A FACTOR OF 2.5 COMPARED TO THE HYBRID FAMILY CAR
• HYBRID HEAT ENGINE /ELECTRIC BUS AND VAN
• DIESEL AND TURBINE ENGINES LOOK ATTRACTIVE FOR BUS AND
VAN APPLICATIONS WITH MINIMUM DEVELOPMENT EXPENDITURES
• POTENTIAL IMPROVEMENTS IN HYBRID BUSES AND VANS COULD
LEAD TO SIGNIFICANT REDUCTION IN N02 AND CO
• VEHICLE EMISSIONS DATA ARE REQUIRED OVER REALISTIC
DRIVING CYCLES ON CURRENT BUSES AND VANS BEFORE
COMPARISON OF THE HYBRID MODE CAN BE ASSESSED
• BATTERY POWER DENSITY AND ENERGY DENSITY SHOULD EASILY
BE MET FOR THE HYBRID BUS APPLICATION
-------
XII-1
Chapter XII
ADVANCED TECHNIQUES IN ELECTRICAL VEHICLES
Bohers/Ducrot
Engineer, Research and Development Division
Citroen Automobile Company/Electricity Section
Nanterre, France
Translated for EPA by SCITRAN
(Scientific Translation Service)
Santa Barbara, California, USA
-------
XII-2
Ladies and gentlemen, my name is Pol Ducrot. I am an
engineer with the Research and Development Division of the
Citroen Automobile Company — Electricity Section.
Let me, in turn, thank the Philips Company for inviting me
here. Thanks to them, I shall report to you on the advanced
studies we are carrying out in the field of electric vehicles.
In many fields, from the old drive to the recent SM, Citroen
has been in the vanguard— for example, in the area of vehicle-
ground communications or oleopneumatic suspension for vehicles
of a more modest type.
Thus, it is not surprising that the Company is engaged in
significant research in the area of electrical drive particularly,
One of the systems we are studying will be described here.
This study is drawn from a research program carried out in
close cooperation with the Compagnie Prancaise de Raffinage and
the Battelle Institute of Geneva.
As we all know, an electric drive system must include a
source of electrical energy, together with a unit to convert 'it
into mechanical energy in a form adapted to the flexibility
requirements which characterize drive.
-------
XII-3
Unfortunately, nobody has yet found an ideal energy source:
quiet, nonpolluting, inexpensive, economical to run, very light,
and... so simple to design in sketches of future vehicles.
One must look for compromises and optimizations to reduce
the often unpleasant disparity between the dream and reality.
This is the purpose of the first part of the presentation,
dealing with a cell supplied with air and liquid or gaseous
hydrocarbons.
Optimization of the Fuel Cell
This is a solid electrolyte cell, the principle of which
is well-known. We know that the working temperature may be
selected between 800 and 1000° C, which enables us to avoid usinj
precious catalysts like platinum and to use such inexpensive
fuels as hydrocarbons, since conversion becomes quite simple.
The specific performance obtained in this way is quite
interesting. The inconvenient aspects of this high temperature
should also be mentioned.
The choice of materials is more limited, and the technology
is more difficult. The electrolyte thickness should be small,
both to reduce the internal voltage drop in the electrolyte and
to minimize the mass to be raised to a high temperature, and
thus the energy required for the proper temperature.
In parallel with the laboratory tests, an optimization
calculation by computer was carried out beginning with the
different mathematical models to take into account the effects
of the numerous parameters of the cell, which include, for
instance :
-------
XII-4
— arrangement of the converter (exterior or integrated)
— cell dimensions
— maximal and minimal electrolyte temperatures
— activation polarization
— average partial pressure of oxygen in the cathode
compartment
— blast-engine output
— heat flux released by parasitic combustion
Photo 1
Each element has a current-voltage characteristic depending
on its location in the cell.
Here are the typical characteristics of an element:
Rp . JE designates the voltage drop in the electrolyte related to
its ionic conductibility.
AEA designates the activation polarizations
AEG designates the concentration polarizations
Photo 2
The design shows here the cell with an external converter to
be used in models 1 and 2.
Photo 3
Here in Model I, the current drainage occurs perpendicularly
to the active faces. We shall term it "cells with transverse
current".
-------
XII-5
Photo
In Model II, current drainage occurs parallel to the active
faces and the elements are arranged in series by a suitable
incorporated arrangement. We shall call it: "cell with internal
series arrangement".
Photo 5.
Here we see the beginning of Fortran writing of the program
relative to the latter model supplied by computer.
Photo 6
Here, by way of example, is one of the curve networks
supplied by calculation of Model I with a transverse current.
It gives the net output of the cell core as a function of the
power-volume ratio for 3 values of the electrolyte thickness.
These curves correspond to a cell center which is completely
heat-insulated. The preheating of the air is done without loss
by heat recovery.
Photo 7
Here is the same network for Model II with internal series
arrangement.
Photo 8
Here is a universal optimization curve. The curve in a heavy
line is the envelope of the different special curves represented
by the dotted lines.
-------
XII-6
This curve shows the compromise between the power-volume
ratio and the net output of the cell core, taking into account
the blast-engine.
Electrical Drive System
After this first series of photos, we shall look at a
special system of electrical drive by an asynchronous, three-
phase motor with casing, supplied by a continuous current source
and the result of bench tests carried out in 1968, beginning with
a rectangular or sinusoidal modulation drive.
There is no need to emphasize the strength and low price
which characterize the asynchronous motor with casing which can
be produced in a large range of velocities and outputs.
On the other hand, its adaption to drive starting with a
fuel cell requires an "undulator" designed especially for that
use .
Photo 9
This is a schematic diagram of the universal control. We
can single out:
—— the undulator bar control.
—• the mechanisms for guiding the bar control. They will
receive, regardless of the type of control used, orders
for voltage V and frequency fs, and will transform these
signals into properly implemented signals in time to
obtain the desired voltage amplitude and frequency in the
stator. These signals will drive the bar control circuits
-------
XII-7
— operational circuits for supplying V and. fs references.
These circuits send out the two parameters: the flux and
current for the motor.
We then have to select an undulator pattern suited for this
application. Two things are certain:
— It must have "two" return diodes, sometimes called "free
wheel diodes."
— It cannot be of the "series" type where the commutation
capacity is placed in series and which does not have the
desired operational flexibility.
Even limiting ourselves to extinction by condenser discharge,
several extinction circuits can be imagined.
Photo 10
Since some order was needed, they were classified according
to the following criteria:
— the extinction class, termed A or B depending on whether
the commutation inductance is or is not traversed by the
motor current. This makes the extinction circuit either
heavily or slightly dependent on the main current.
— extinction systems, which may be either direct or al-
ternating.
The table you see illustrates this classification.
At the top, the continuous direct current extinction systems
No. 1: universal. All thyristors are cut off simultaneously
No. 2 and 3' individual or semi-individual.
-------
XII-8
At the bottom: "alternating current" extinction systems.
No. 4 and 5: complementary extinction systems — very
economical — where lighting one group causes
the extinction of the other group.
No. 6 and 7: individual or semi-individual extinction
systems, which are much more complex.
Semi-individual alternating current extinction
of class B was chosen for our tests, because
it had the required flexibility for sinusoidal
modulation allowing smaller harmonic content
in the current wave.
Photo 11
This picture is taken from our sinusoidal control patent
and shows the control ensemble schematically.
To simplify the design, the thyristor protection circuits
in dV and dl , , , ,
— — have not been shown.
dt dt
We see three control groups U, V and W, with for each:
— Tl and T2 main thyristors
— free wheel diodes Dl and D2
— extinction thyristors SI and S2
and an inductance/extinction capacity circuit.
Photo 12
Here are the output voltages of the undulator with rectangular
modulation. The form factor T/TM is 50% here.
-------
XII-9
UR - Us shows the composite voltage between phases.
Photo 13
This shows the law governing the change of the cut-off
frequency as a function of output frequency.
The cut-off frequency is a whole multiple of 6 fs. When
starting, fs = 10 Hz, K = 16, fM = 960 Hz.
We see that K successively takes on the values 16, 8J 4, 2,
1, so that fM never exceeds 2 Khz.
Photo 14
Here we see the theoretical behavior of the current in a
main thyristor with sinusoidal modulations at cos = 1
Photo 15
This photo is also taken from the same Citroen patent. To
obtain the correct cut—off sequence, a special sinusoidal
modulation system was perfected.
Here is the principle employed:
Above, diagram A shows a sinusoidal voltage wave 1 with
frequency fo.
At Instants tl, t2, t3, following periodically at intervals
of T = 1/fe, we measure the instantaneous value of the sinusoidal
voltage with respect to the reference level 3, a voltage shown by
iengths al, a2 , a3 . . .
-------
XII-10
Thus, a series of asymmetrical triangular signals is
generated as shown in diagram B.
We see that their steep fronts bl, b2, b3 occur at instants
tl, t2, t3-
— their heights are proportional to the instantaneous values
al, a.2, a3 of sinusoidal voltage 1 measured at these
sampling times tl, t2, t3-
— the oblique fronts cl, c2, c3 all have the same given
slope p.
— in diagram C the triangular signals are transformed into
rectangular impulses, each with a width equal to the
duration between the rigid front of the corresponding
triangular signal and the moment when the oblique front
reaches the reference level visible in diagram B.
Thus, impulse II has a width equal to the period dl.
So, we see that the impulses I are synchronous with the
sampling with frequency fe and are modulated in width in accor-
dance with the instantaneous value of sinusoidal voltage 1 at
those times, since the gradient provides a voltage-time linear
conversion.
If the reference level is identified with the zero level of
the triangular signals, the width d of the impulses is directly
proportional to the instantaneous value of sinusoidal voltage
measured with respect to level 3.
These impulses I are used to control turning on and off the
thyristors inserted in the group commuting an undulator phase..
-------
XII-11
One of these thyristors is kept on as long as impulse I
lasts, and it is turned off by forced extinction during the
interval T - d.
e
Clearly, we need only sample the same sinusoidal voltage 1
at other times offset by 2£ to obtain the impulses needed to
control two other groups.
A closer examination of the system would show that a second
order harmonic would appear with an amplitude proportional to
frequency fo; however, a simple and effective correction was
introduced by modulating the reference level.
Photo 16
Here we see a simplified diagram of the sampling and
amplitude-time conversion apparatus.
— For each impulse f, the sampling condenser C is charged
very quickly by the push-button p.
— It is discharged permanently through the constant current
source lo.
—- The comparator enables us to obtain a logic signal
defining the commutation sequence for an undulator phase.
One should note that this signal is the image of the phase
output voltage.
Photo 17
Here is the drive curve which shows the behavior of the motor
coupling as a function of velocity.
-------
XII-12
For this motor, a 3.6 Mkgs starting coupling was used for
a nominal output of 10 KW with a maximum velocity of 12,000 rpm.
Photo 18
Here is the output measured for rectangular and sinusoidal
control.
We see that in the average velocity range the rectangular
command system yields a greater output for the laws of cut-off
frequency used. This can be explained by the unequal losses in
the extinction circuits.
We have also noted increased fatigue in the sinusoidal
control extinction circuits —- including the extinction con-
densers — when we attempt to obtain a starting coupling.
For the sake of completeness, I should point out the slightly
greater complexity of the sinusoidal control electronics, with
little effect on price.
On the other hand, I should stress the coupling modulations
derived from pulsating parasitic couplings with rectangular
modulation.
In starting, their relative amplitude is on the order of
20%3 and the frequency is 60 Hz.
Consequently, we must anticipate a transmission to avoid
mechanical fatigue, not to cause resonances and to retain the
qualities of comfort and silence so highly valued in electric
vehicles.
-------
XII-13
I shall conclude this presentation with one small detail:
The continuous current source used In these tests was not
from a fuel cell.
Thank you, Mr. President, I am prepared to answer questions,
to the extent that they are not too compromising — I think you
will understand.
Company address:
Societe des Automobiles Citroen
qua! A. Citroen
Paris XV, France
Research Division:
Pol Ducrot
Service Recherches
Societe des Automobiles Citroen
1 rue P. Millet
92 Nanterre, Prance
-------
P lane h e
Fig. 16. Electiic
characteristics of an
element. ,-
conversion '
compartment
anode
Figure ^. Principle of
: the heart of a cell n:
with external conversion
joxes
conducting band
fuel distribution
channels
inter-plaaue
connectors
membrane
active
surface
porous
support
connectors
Fig. 1. anode box model 1
braces
fuel distribution
^channels [
connecting band
[connected with
the metal parts
—^ of the box
connectors for
internal series
connection
electrically insulated
from porous support
inter-plaque connectors
**
Fig. 10. anode box, Module II
Ph.'I
ph.a
-^ Ph,3
'Appendix 5 FORTRAN listing of program II
Legend for Ph. 6 and Ph. 7
1 — activation polarization: 0.10 V
2 — electrolyte temperature: 800 °C
3 — cathode thickness (In.O ): 50u
Z, J
Ph.
cell with transferse current
exterior converter
cell with internal series
connection
exterior converter!
yield
calculated points^, ^7
closest to optimum JF(!)J
maximum current *~—-
through electrolyte '^.
combustion yield at -":
the center of core
power
Ph. &
Ph , 7
Ph. <3
volume ratio
-------
, 7 ',
ondulator V
supply control^., j motor
In
traction_
braking:
conversion
=p
transducers
o o
3 -H
G 4->
•H O
O JJ
class A
total
semi-
indiv.
In class B
no known example FIG.7
no known example
-Q complementary diagonal
impossible
Ph. 6
Fig. 2.3 Change in cutoff
frequency f as a function of
output frequency fq.
•Values of k 16, 8, 4, 2, 1
Ph.1.
Ph.
-J
Ph. -13
1968
-{current
: voltage
-^- couple
-^velocity
output
voltages
at the
ondulator
for rectangular
modulation with Fm = 12 Fs and T/T =
TT"' n'Lr,^ n voltage delivered by 50%
U , U_ and U '
R s r ondulator phases'
U^, Ugcomposite voltage" x"l
U neutral motor potential
K
U -U .U neutral-phase voltage
— R n phR ;
Fig. 1.7. theoretical
curve for current in a main,
thyristor for sinusoidal
modulation for a charge
with
FIG. t
Ph. -t 5"
couple-yield (m kg) /(%)[:-|:pr July 1968L
^pittBi
Ph.
Fig. 1.2. Traction curve
numerical values;
trigger
Icrn
constant
current
source
Fig. 12. Simplified
diagram of a sampling
and amplitude |
time conversion device
•^j-, •- i--fyield_with
'ield-half c-^~^e jsinusoidal ph A 6>
•. :.•;".''~T "• r-|Ll_: modulation
/;iFig. 3.1. total yields
..TilL; !X; r.:_itotal chargeJIii-'
li-l-i i-!-'. ~> ^-|:-;-:-i-f-i-f-H-rl couple delopped
—r T ] i . : 1^^, - - \- , I c r *•
fxl
H
M
I
M
Ln
ph -ir
-------
XIII-1
Chapter XIII
RESEARCH AND DEVELOPMENT ON A LITHIUM-SULFUR BATTERY
by
Elton J. Cairns,
Section Head, Energy Conversion Section,
Argonne National Laboratories
Argonne, Illinois, USA
-------
XIII-2
Introduction
The Air Pollution Control Office of the Environmental Protection Agency
(EPA) has initiated a broad program for the development of low-emissions
vehicles. A number of alternatives to the internal combustion engine
can be considered to be candidates for use as power plants in such
vehicles. Some alternatives are listed in Slide 1 (copies of the
slides are attached). The Argonne National Laboratory, under the
sponsorship of EPA, is pursuing the development of lithium/sulfur
batteries for all-electric vehicles. Some of the goals that ANL has set
for itself (in keeping with the EPA goals) are shown in Slide 2. The goal
of 220 W-hr/kg cannot be met by any conventional battery, hence the interest
in the lithium/sulfur system, which uses a liquid lithium anode, a liquid
sulfur cathode, and a molten salt electrolyte containing lithium halides,
and operates at 375°C.
Though this program is still in the laboratory stage, some interesting
results have been obtained which are worthy of review. The manner in
which the lithium/sulfur cell operates during discharge is indicated in
Slide 3. It is particularly important to provide for the removal of
the product la^S from the reaction site, and to supply more sulfur and
electrons for reaction, without losing any material to the surroundings„
A major portion of our effort is centered around this process.
Experimental Results
An example of a small lithium/sulfur laboratory cell is shown in Slide 4.
Voltage-current density and voltage-capacity density curves for such a
-------
XIII-3
SOME ALTERNATIVES TO THE INTERNAL COMBUSTION ENGINE
1. BRAYTON CYCLE (GAS TURBINE)
2. RANKINE CYCLE (STEAM ENGINE)
3. STIRLING CYCLE
4. HYBIRD: HEAT ENGINE PLUS BATTERIES
5. ALL-ELECTRIC: SECONDARY BATTERIES
EPA PROGRAM
HIGH SPECIFIC ENERGY Li/S BATTERY
FOR ELECTRIC AUTOMOBILES
GOALS: 220 W-hr/kg
220 W/kg
1000 CYCLES
$10/kWhr
LOW COST, HIGH SPECIFIC ENERGY
-------
ANODE-FELTMETAL
CONTAINING Li
CATHODE-FELTMETAL
CONTAINING S
ELECTROLYTE: e.g. LiBr-RbSr
CELL REACTIONS
CELL: Li/Lit ELECTROLYTE )/S (+ Li)
ANODE: Li0-*- Li+4- e~
CATHODE: 2Li+ +2e" + S°-»Li2$
OVERALL: 2Li°+s°->-Li2s
-------
XIII-5
ANODE LEADS
CATHODE LEAD
Nb PLUNGER
Cu GASKET
Nb CATHODE HOUSING
CATHODE CURRENT
COLLECTOR
CONTAINING
SULFUR
S.S. ANODE CUP
ALUMINA CRUCIBLE
ELECTROLYTE
LiBr-RbBr
ANODE FELTMETAL
CONTAINING LITHIUM
-------
XIII-6
cell are shown in Slides 5 and 6. The capacity density of Slide 6 is
too small for the goals of Slide 2. Therefore, efforts have been made to
improve the capacity density by modifying the structure of the cathode
current collector, as shown by some examples in Slide 7. The voltage-
capacity density curves for constant-current operation of lithium/
sulfur cells with a comb, laminated, and enclosed laminated structures
are shown in Slides 8, 9, and 10, respectively. The laminated cathode
has yielded the highest capacity densities, and the enclosed laminated
structure has shown the best cycle life. Another long-lived cathode is
the so-called reservoir/cathode, which is comprised of a disk cathode
with a space above it filled with sulfur, and an electrolyte-wetted layer
of porous material below the disk current collector, to prevent the escape
of sulfur.
Some of the important performance parameters for various cathodes are
summarized in the table of Slide 11. The 220 W-hr/kg goal of Slide 2
o
corresponds to about 1.2 A-hr/cm , or 0.4 A-hr/gm. The cycle lives of
the enclosed laminated and reservoir cathodes are within a factor of 2-3
of the goal. It can be seen that some further improvements in the cathodes
are necessary before serious scale-up and engineering work can be carried
out.
In addition to the experiments with laboratory cells, several other areas
of investigation are being pursued, as indicated in Slide 12. The phase
equilibrium investigations have as their objective the identification of
electrolytes and additives to sulfur which minimize the solubility of sulfur-
-------
CURRENT COLLECTORS
STAINLESS STEEL FELT
POROSITY PORE SIZE
O 80% 240/A
D 80% 29 M
A 40% l9i
2.5
Li/LiF-LiCI- Lil/Li in S
ANODE AREA = 2.6cm2
CATHODE AREA =0.7 cm2
INTERELECTRODE DISTANCE = 0.3cm
CELL TEMPERATURE » 375°C
SHORT-TIME DATA
A
D
21012345
CURRENT DENSITY, A/cm2
8
0
-------
3.0
Li/Li Br- Rb Br/Li in S
2.5
LJ
CD
<
h-
o 1.5
\
CATHODE AREA
ANODE AREA
INTERELECTRODE
DISTANCE
= 0.7 cm2
= 2.6cm2
= I .Ocm
LU
O
.0
0.5
CURRENT COLLECTOR
POROSITY
PORE SIZE
TEMPERATURE
CURRENT DENSITY
S.S FELT
= 80%
= 29/im
= 375°C
= 0.33 A/cm2
0
0
0.05 O.I 0.15 0.20
CAPACITY DENSITY, A-hr/cm2
0.25
-------
DISK
COMB
LAMINATED
ENCLOSED LAMINATES
CATHODE CURRENT COLLECTOR STRUCTURES
-------
3.0
0.26 A/cm2
RECHARGE \ 2
T
Li/ Li Br- RbBr/Li in S
0.53 A/cm2 RECHARGE
0.5
0.0
0
0.53 A/cm2
DISCHARGE
i
ANODE AREA 2.92 cm2
CATHODE AREA 1.89 cm2
INTERELECTRODE DISTANCE I cm
TEMPERATURE 395°C
CATHODE CURRENT COLLECTOR
GRAPHITE, 1.4/APORE SIZE, 63% POROSITY
THEORETICAL CAPACITY 1.45 A-hr/cm2
I I
O.I
0.2
0.3
0.4
CAPACITY DENSITY, A-hr/cm'
0
10 20
PERCENT OF THEORETICAL CAPACITY DENSITY
I
M
O
ON
30
-------
3.0
2.5
0.45 A/cm2 RECHARGE I
Li/LiBr -RbBr/Li in S
ANODE AREA 2.7cm2
CATHODE AREA 0.96 cm2
NTERELECTRODE DISTANCE
TEMPERATURE 390°C
THEORETICAL CAPACITY
DENSITY 2.06 A-hr/cm2
0.72 A/cm2
RECHARGE 2
04 A/cm2 DISCHARGE 3
0.31 A/cm
DISCHARGE
0.52 A/cm2
DISCHARGE I
CATHODE CURRENT COLLECTOR
4 SULFUR ELEMENTS 1.6mm THICK
80% POROSITY 30yu- PORE SIZE
5 ELECTROLYTE ELEMENTS 0.45mm THICK
83% POROSITY 25/z PORE SIZE
I
0
0.
0.2 0.3 0.4
CAPACITY DENSITY, A-hr/cm2
0.5
0.6
I
I
0
10 20
PERCENT OF THEORETICAL CAPACITY DENSITY
-------
UJ
o
>
o
I I I
0.18 A/cm CHARGE - 34
0.2 A/cm" CHARGE -17
0.2 A/cm' CHARGE-2
0.2 A/cm2 DISCHARGE-2
.0
0.5f
0
Li/LiCI-Lil-KI /Li in S
ANODE AREA 2.5 cm
CATHODE AREA 2.53 A/cm
INTERELECTRODE DISTANCE I cm
TEMPERATURE 380°C
CATHODE CURRENT COLLECTOR
3 GRAPHITE ELEMENTS
63 % POROSITY 1.4^. PORE SIZE
MOLYBDENUM FOAM CASING
THEORETICAL CAPACITY DENSITY 0335 A-hr/cm2
0.2 A/cm DISCHARGE-34
0.2 A/cm DISCHARGE- 17
0
0.05 O.I
CAPACITY DENSITY, A-hr/cm;
0.15
0
10 20 30 40
PERCENT OF THEORETICAL CAPACITY DENSITY
50
-------
XIII-13
CATHODE
TYPICAL PERFORMANCE CHARACTERISTICS
FOR VARIOUS CATHODE STRUCTURES
W/cm2x A-hr/cm2** A-hr/cm3 A-hr/gm CYCLES LIFETIME
hr
DISK
COMB
LAMINATED
ENCLOSED
LAMINATED
RESERVOIR
3
3.5
4
4
2
n TM^MOTT1
0.2
0.4
0.5
0.2
0.36*
V 1 s*m TMTT
0.6
0.3
0.5
0.15
0.15
?-DTTT ffTDnn
0.21
0.15
0.17
0.05
0.05
17 n T C T A Ml
<5
<2
<10
419
>400
"'T?
<10
<5
<20
588
>500
**ONE-HOUR RATE
OTHER AREAS OF INVESTIGATION
1. PHASE EQUILIBRIA: Li2S-S-LiX
2. INTERFACIAL PHENOMENA
3. CATHODE MATERIALS STUDIES
4. SOLID ELECTROLYTES
5. MATERIALS EVALUATION
-------
XIII-14
bearing species in the electrolyte. A minimum solubility is desired in
order to minimize the rate of loss of sulfur from the cathode. Some
typical results are shown in the form of a pseudo-ternary phase diagram
in Slide 13. Here, the extent of the phase marked LS is to be minimized.
Interfacial phenomena are also important in retaining sulfur. The cathode
current collector should be well-wetted by sulfur, but not by electrolyte.
The wetting properties of various solids by sulfur and by electrolytes
are being studied. Cathode materials (i.e., sulfur, plus various additives
such as phosphorus) which have high electronic conductivities, low
viscosities, low vapor pressures, and other desirable properties are being
sought. Investigations of the suitability of solid lithium-ion conductors
for use as electrolytes has recently begun. Materials stability continues
to be an important area of investigation. An up-to-date summary of the
corrosion rates of a number of candidate materials of construction in
lithium-sulfur mixtures and in pure lithium is given by Slides 14 and 15,
respectively.
Electric Vehicle Performance Calculations
In order to evaluate the potential performance of an all-electric vehicle
powered by a lithium/sulfur battery, some computer calculations have been
performed for an electric automobile having the characteristics shown in
Slide 16. The power requirements for this vehicle were calculated from
the equations given in Slide 17, using the values of the constants
shown in Slide 18, The driving profiles assumed for the purposes of the
-------
A
D
O
0
THREE PHASES (quench)
TWO PHASES (quench)
PHASE BOUNDARY
DTA
Li2S-S-(LiBr-RbBr)
PSEUDO TERNARY
SYSTEM 360 °C
LiBr-RbBr
-------
XIII-16
CORROSION BY 20 % Li~S MIXTURE 375°C
INCONELJ702)
2RK65 SS
ZIRCALOY-2
347 SS
HASTELLOY-X
ALUMINUM
205 SS
AVE. RATE, I00-300hr
MAX. RATE, 620 hr
AVE. RATE, 620 hr
012345
CORROSION RATE.rnm/yr
6
-------
XIII-17
CORROSION BY MOLTEN LITHIUM AT 375° C
BeO°
Th02
AIN
BN
LiAI02
MgOb
BeOC
BeOd
(
! 1 1 1 ill! ; } \ \
3
^ TEST DURATION 1000-1200 hr
P
z
'//A
///////\
/////////A
//////////.
'////////////////////\ \/ / / / //
a. HOT-PRESSED, HIGH-PURITY
b. SINGLE CRYSTAL
c. RECRYSTALLIZED GRADE
d. COMMERCIAL GRADE
lilt till n 1 1
D 0.5 1.0 12
CORROSION RATE, mm/yr
1 1 «\
20 40
CORROSION RATE, mils/yr
480
-------
XIII-18
ELECTRIC VEHICLE CHARACTERISTICS
CURB WEIGHT
PAYLOAD WEIGHT
BATTERY WEIGHT
SPECIFIC ENERGY (4hr RATE)
SPECIFIC POWER (Ihr RATE)
ACCESSORY POWER
AIR CONDITIONING
POWER STEERING
TOTAL
1588 kg
227 kg
397 kg
220 W-hr/kg
220 W/kg
230 W
4400 W
1900 W
6530 W
(3500 lb)
(500 lb)
(875 lb)
(100 W-hr/lb)
(100 W/lb)
POWER REQUIREMENTS EQUATIONS
"P "P
1.1 Ra)
Pr = V(Rr + R^ + Rg
V2
Rr = Tc x W
Rw = fA CD Af
R = W sin e
o
R = W dV
a g ar
-------
XIII-19
VALUES OF CONSTANTS
ROLLING RESISTANCE COEFF. (T ) 0.0175
AIR DRAG COEFFICIENT (CD) 0.35
FRONTAL AREA (A£) 2 . 32m2
TRANSMISSION EFFICIENCY (E -E ) 0.82
^ m ej
TOTAL ACCEL. /LINEAR ACCEL, 1.1
"6
AIR DENSITY/ gc
1.25 x 10
-------
XIII-20
calculations are presented in Slide 19. For each profile the range was
calculated for no accessory power for 230 W (lights, heater, etc.) and
for 6530 W (air conditioning and power steering added).
The laboratory results were put into the electric vehicle calculations
in the form of E = f (q, i) equations, where E is cell voltage, q is
capacity density, and i is current density. Slide 20 shows a set of E - q
plots for various i, drawn by the computer. These equations, together
with those of Slide 17 and the driving profiles, were combined with a
cell and battery design similar to those shown on Slides 21 and 22 to
yield the results shown in Slide 23. The ranges shown are less than
the EPA goal of 322 km (200 mi); hence, there is a need for improvement
in the capacity density of the laboratory cells and/or the design of the
battery. These results do show, however, that electric vehicles based on
lithium/sulfur batteries can be expected to have ranges of 150-250 km, if
25% of the curb weight of the vehicle is alotted for batteries.
Acknowledgement
This paper represents a summary of the work of the many people listed in
Slide 24. They are the ones to whom credit should go for the accomplish-
ments described above. I also wish to thank the Air Pollution Control
Office of the Environmental Protection Agency for support of this program.
-------
0
URBAN DRIVING PROFILE
SUBURBAN DRIVING PROFILE
i
NO
CROSS-COUNTRY DRIVING PROFILE
O\J
60
40
20
0
—
i i i i i i i i i i i , ,
300 600 900 1,200 1,500
TIME, SEC
1,100
-------
Li / LiBr -RbBr/ Li in S
375°C
EMPIRICAL FIT
CURRENT DENSITY
A/cm2
A 0.095
A 0.094
V 0.091
• 0.049
0.40 —
0.20
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10
CAPACITY DENSITY, A-hr/cm2
i
M
N>
-------
SEAL CLAMP
&•:••••-; SULFUR , X-.'. ,•%
ELECTROLYTE
H
I
NJ
1;33cm
ELECTRICAL INSULATOR
(HIGH TEMR PLASTIC)
LITHIUM/SULFUR CELL CONCEPTUAL DESIGN
-------
XIII-24
56 cm
THERMAL
INSULATION
METAL
CONTAINER
POWER TERMINALS
95 cm
67 CELLS
PER STACK
COPPER BAR
LITHIUM / SULFUR BATTERY
CONCEPTUAL DESIGN
-------
XIII-25
300
2501-
E 200
LU
O
< 150
LL)
_l
U
£ 100
50
0
X
ACCESSORY LOAD
BO 6530 W
230 W
mm o w
150
100
50
URBAN
DRIVING
PROFILE
32 Km/hr
(20mph)
SUBURBAN
DRIVING
PROFILE
67 Km/hr
(41 mph )
CROSS-COUNTRY
DRIVING
PROFILE
106 Km/hr
(66 mph )
0
-------
XIII-26
EPA PROGRAM
R. K. STEUNENBERG
J. P. ACKERMAN
B. A. FEAY
M. L. KYLE
H. SHIMOTAKE
R. RUBISCHKO (GOULD)
D. M. GRUEN (CHM)
A. J. ZIELEN (CHM)
T. W. LATIMER (MSD)
J. N. MLJM)Y (MSD)
D. E. WALKER (EBR-II)
R, M, YONCO
J. R. PAVLIK
F. J. MARTINO
-------
XIV-1
Chapter XIV
RESEARCH AND DEVELOPMENT PLAN OF ELECTRIC CAR
by
Mr. Shizume
Japanese Automotive Manufacturers' Association
Japan
-------
XIV-2
1. Fundamental Plan
(a) We intend to develop a "new type electric car" for the purpose
of use in the town from 1971 to 1975. New type electric cars include
the following variations:
(1) Small scale passenger car.
(2) Small scale cargo car.
(3) Medium scale passenger car.
(4) Medium scale cargo car.
(5) Bus
The development cost is estimated about thirteen million dollars.
We develop it by concentrating the R & D abilities of national
organizations, private automobile companies, universities and so on.
Our main purpose to develop it is to protect the environment (air
pollution caused by exhaust gas or noise caused by engines).
(b) We will develop it according to the following procedure:
(1) We will develop the experimental electric car by pursuing the
optimal structure and performance to accomplish this purpose. We
will make the primary experimental electric car and test its
performance and safety, etc. for the first three years. Then we
will make the complete experimental electric car and not only
test its performance and safety but also judge it from the view
of technique, social needs or economy for the following two
years. The performance data shown later is the performance of
primary electric cars which we will develop for the first three
years.
(2) We will research and develop several components used to the
experimental electric car such as Pb-PbC^ battery, new type
-------
XIV-3
battery, new type motor (thyristor motor, wheel motor, etc.)
controller, plastic body and so on. We will use this result
to develop the experimental electric car.
(3) On the other hand we will study the utilization system of
electric cars from the view of city traffic system, standardiza-
tion, optimal total energy system and so on in order to use it
well and promote the popularity.
Append ix
National Research and Development Program
(1) Object
From the standpoint of national interest, many fields of industrial
technology need urgent research and development which requires a
great deal of expense and a long-term period for their fulfillment.
It is demanded of the Government to promote such important research
and development positively by making plans, bearing research expenses
and by organizing research abilities.
In answer to these acute needs the National Research and Development
Program System was established in 1966.
(2) Method of Selection
Each project is selected from the projects proposed by governmental
organizations, national laboratories or private companies under the
following criteria.
-------
XIV-4
(a) The project is very important and immediately needed to
achieve leveling up of industrial construction, effective
exploitation of natural resources or preventing public nuisance.
(b) Its techniques are advancing and have effective spin-offs.
(c) Large amount of financial resources over a long period are
needed to research and develop the techniques and, these, in
addition, are subject to heavy risks.
(d) The target of R & D of the techniques must be set and also
the technical method to achieve it must be forecasted.
(e) The R & D of the techniques requires the concentration of the
R & D abilities of national organizations, private companies,
universities and so on.
Now the National R & D Program System (NRDPS--known as Ogata-project--
in Japanese) includes the R & D of "Magneto-hydrodynamics generator,"
"Large-scale digital computer," "Desulfurization process," "New
process for olefin production," "Sea water desalting and by-product
recover," and "Remotely controlled undersea oil drilling rig."
From April 1971, "Electric car," "Jet engine used to civil aircraft,"
and "Pattern recognition system" will be starting to develop.
(3) Method of Management—A special organization for managing the
System has been established in The Agency of Industrial Science and
Technology (AIST). The organization manages each of six projects.
The following five headings has the responsibility of the management.
(a) Selection of R & D project.
(b) Planning the actual programme.
(c) Selection of sub-contractors.
(d) Operation of the System.
(e) Management of the achievements.
Another organization, such as the advisory committee or evaluating
committee, assists the organization in managing the System.
-------
2. Research and Development Plan of Electric Car
Establishment of Judging
Standard 6t judgement
of Electric Car
i
Development of Experimental
Electric Car
Development of Several
Conpoents
Decelopment of Charging
System
Study of Utilization System
Development Cost
(Million Dollars)
1971
Study of .
Standard
raent of
Equipmen
Design 1 >
Design 1 *
Fundamental
| System Stud
1.25
1972 1973 1974 />, 1975
Tudijing Establishment Develoocent of
Develop- — * of Judging — ^ Test > Testing Equip- — » Final 1
testing Bshod ~/^ ment Test |
: Study & Esta- A 'V
blishment of
Judging Standard
Trial Production T Production
of Primarv Exoeri- s[ _, . ~l -. r .,,,,, ,
. . _,J . ' ^ Design > of Final Expei'l-
mental Electric e- ^ mcntal Elcctric
Car ^ * **
A. v>Qr
1 t*—
Trial Pi'uducLlou ^Design S ^ ^
. J
Studyl (Application Study] *-•
y |— ^Application Study] 'Trial Production}
*\ *, i
^[System SLuuy| >-
3.64 3.97 2.56 2.0
Remarks
Nations. l Laboratory
Car Maker
Private Company
& National
Laboratory
Private Company
National Organizatior
13.42
(TOTAL)
I
Ln
-------
3. Specification & Performance of the Experimental Electric Car
(The End of 1973)
Passenger + Payload (kg.)
Total Weight (kg.)
Maxmum Speed (km/h)
Range (km)
Acceleration
Ability (0 - 30 km/h)
(Sec)
Climbing Ability
(Speed of 6 degree slope)
(km/h)
Cargo Car
Small Scale
2 + 200
1,100
70
130 - 150
5
40
Medium Scale
2 + 1,000
3,500
70
180 - 200
5
40
Passenger Car 1 Bus
Small Scale
4 or 2 + 100
1,000
80
130 - 150
4
40
Medium Scale
5 or 3 + 300
2,000
80
180 - 200
3
40
Large Scale
60 - 80 persons
15,000
60
230 - 250
8
40
I
ON
-------
XV-1
Chapter XV
STUDIES BY FIAT ON THE ELECTRICALLY-DRIVEN AUTOMOBILE
by
G. Brusaglino
Chief, Electrical Research Department
Fiat Research and Development
Turin, Italy
-------
XV-2
Electrically-driven automobiles have a long history at
Fiat. As early as the last world war, some electric cars were
built on chassis and bodies of that period.
Within the framework of research aimed at reducing air
polution, in 1961 Fiat began new studies on the possibility of
using electrical drive in the light of current technological
possibilities.
For the first experiments, we used vehicles provided with
a clutch and gearbox, namely: a Fiat 200 van and a Fiat 1100
automobile.
In orienting these studies, we took into consideration the
various possible drive systems with an eye to defining the best
applications for use in automobiles, especially urban areas. We
found that such systems must satisfy the following conditions:
1. High overall output of the whole drive system, i.e., of
the motor and its control system under all vehicle operating
conditions, especially in stop-and-go city driving.
-------
XV-3
2. Capability of quick response in traffic. Thus, the
following were deemed necessary:
— fast acceleration;
— automatic braking effect when accelerator is released
(similar to effect in thermal motor) with possibility of
later energy recovery.
— possibility of velocity control and coupling with no
discontinuity or with a discontinuity contained within
limits judged acceptable for thermal motor vehicles.
3. Availability of high specific outputs even at small rpm
rates to allow good motor operation.
This characteristic enables us to bring into play certain
laws of motion using appropriate control to reduce acceleration
time, taking into account passenger comfort.
These conditions must be accompanied by feasibility of each
element in the system, low price, and inexpensive maintenance.
In the drive system of the vehicle mentioned above,we used a
compound excitation motor where one of the coils can be supplied
independently. The armature has no dissipation element: velocity
regulation depends only on variation of the current in the in-
dependently-supplied coil.
This system results in variable motor velocities in a con-
tinuous fashion over a range of values which, through special
design features, has proven sufficiently broad. Moreover, this
system has, naturally, the possibility of dynamic braking by
energy recovery with a braking effect similar to that of a piston
engine. In fact, under over-excitation conditions with respect
-------
XV-4
to the velocity imposed by the car, the motor shifts to operating
as a generator with no discontinuity.
The energy expended for regulation is practically negligible
and the output of the drive system, identified with that of the
motor, assumes values close to 95% under normal operating condi-
tions .
The output in the transitory regime may reach the same
values by similarly setting the de-excitation gradient in
correlation with velocity.
In this system, the motor velocity must not drop below a
certain regime corresponding to maximum excitation; consequently,
the installation must be completed with a gearbox and transmission.
Nonetheless, the system has the advantages of durability, economy
and high overall output, which were the bases of the research
orientation.
We next attempted to simplify the operation of experimental
cars by eliminating the transmission and gearbox.
For that purpose, we added a nondissipating regulation system
to the armature circuit to permit the motor regime to vary from
zero velocity to the velocity when regulation of the independent
field begins.
A system of this type was used on a car derived from the
Fiat 850 model.
The characteristics of the car are as follows:
— Car Fiat 850 (with structural
modifications)
-------
XV-5
— Total unloaded weight 1025 kg
— Batteries lead, special type with
plastic vat
— Battery weight 320 kg
— Battery voltage 96 V
— Motor Fiat, 6 poles, compound
excitation
— Nominal output peak 21 CV
— Maximum peak output 45 CV
— Maximum velocity 72 km/h
— Acceleration from 0 to 50 km/h in 8 seconds
— Range at 60 km/h on
level road 65 km
In order to ascertain more realistically the performance
of the regulation system developed for a small light car es-
pecially suited for use in town, a Fiat 500 car was transformed to
electrical drive.
To simulate the performance which could be supplied by these
future, high energy density, light-weight batteries, the car was
fitted with a limited number of lead batteries.
The characteristics of the car are as follows:
— Car Fiat 500 (with structural
modifications)
— Total unloaded weight 730 kg
— Battery weight 160 kg
— Battery voltage 96 V
— Motor Fiat, 6 poles, compound
excitation
— Nominal output 21 CV
-------
XV-6
— Maximum peak output 45 CV
— Maximum velocity 80 km/h
The operational characteristics of this car, aside from its
range which was not significant in this study, were judged satis-
factory. The car was easy to drive in city traffic, due to the
outstanding pick-up.
We should point out that these cars were produced especially
for the purpose of testing certain drive systems. These systems,
moreover, allow us simply to increase the performance to the
point of making them similar to those of traditional vehicles for
use not only in the city.
We are at present producing a vehicle fitted with a drive
system with an asynchronous motor fed by batteries through a
static converter (produced in collaboration with Philips Co. of
Milan).
This experience was regarded as a prelude to the use of such
a system in buses.
In particular, we plan the production of hybrid buses with
batteries recharged by an electric generating group, which could
be brought about by a gas turbine.
The drive systems tested so far use motors of almost entirely
conventional types, even if they are designed with particular
features to make best use of certain characteristics.
The relative control modes consequently conform to the need
to adapt the characteristics of the motor to the requirements of
driving the vehicle — elements which can not always be reconciled,
-------
XV-7
At this point, it was regarded as opportune to approach the
problem of a drive system as a whole anew and to revise the
motor design by bringing it closer to the requirements of a
vehicle through more rational regulation, i.e., considering the
structure of the motor as intimately connected to the system
of control.
Current Fiat research programs on electrical vehicles are
based on these principles.
-------
XVI-1
Chapter XVI
ELECTRICAL VEHICLES WITH FUEL CELLS: WHY AND HOW?
by
J. Beslier
Chief of Electrical Equipment Research
Peugeot Automobile Company
Paris, France
-------
XVI-2
In order not to impair road traffic, the electric
vehicle should have a performance close to that of other
vehicles on the road. It must have sufficient speed,
good acceleration ability, and a normal action radius.
But strengthening the car structure to ensure safety
in case of impact makes the chassis heavier and will
make the construction of small vehicles more and more
difficult. All this leads to a vehicle having medium-
sized dimensions, equipped with a powerful generator,
with large energy capacity. Only the fuel cell meets
the requirements of the problem.
The associated Companies PEUGEOT-ALSTHOM have signed
an important contract for research with ESSO-USA, which
should lead to a methanol-air cell that would be suit-
able for the contemplated vehicle.
In an electric vehicle, the energy consumed by the motor
comes from a generator which is to the electric motor what the
boiler is to the steam engine. To answer the problem posed,
which Is the reduction of pollution, one needs a nonpolluting
generator and a silent motor transmission unit.
-------
XVI-3
What Vehicle to Make?
Before examining each of these points, we must define the
main characteristics of the desired vehicle in order to select
the best adapted elements. In fact, we may attempt to produce a
vehicle answering a very special list of specifications (delivery
vehicle, taxi, vehicle for city use, small public transporta-
tion vehicle) or, on the other hand, we may try to convert the
standard vehicles presently used which are included in the various
possible categories. We shall not deal here with heavy weights
and other large specialized vehicles which are not customarily
produced by the Peugeot Company.
Very diverse considerations will influence the choice. They
deal with industrial requirements, technical capacities, market
demand and the norms imposed by regulations.
Effects of Regulations
As far as the latter are concerned, it is undeniable that
they play a large role in the work of the Research Section.
Whether it is a question of passive or active safety, nuisance
caused by the vehicle or its operational features, all these
problems have caused significant chassis changes both with regard
to structure and equipment.
Improved safety in case of collision involves strengthening
the structure. To withstand front, back, and side collisions,
while providing survival space, there must be more and more space
for the structural elements which are indispensable for protection
of the main body. The latter will tend to increase to permit the
installation of protective devices for the passengers. Another
-------
XVI-4
consequence of this strengthening is a substantial increase in
chassis weight.
All this leads us to believe that it will become increasingly
difficult to reduce vehicle dimensions.
Market Demand
It is undeniable that the vehicle best corresponding to the
demands of European consumers is the average-sized vehicle.
Very small cars have had only ephemeral or limited success. They
remain relatively expensive, especially when the price is compared
with their very limited space, performance; and comfort capacities.
By a very small vehicle, we mean vehicles designed to carry two
persons, with a minimum of exterior accoutrements.
The average-sized vehicle owes its success to the fact that,
for a price many can afford, It offers a good compromise allowing
practical and comfortable use in diverse business or family
circumstances.
This success does not seem to be diminishing, and it leads
us to think quite naturally that a new vehicle will have a greater
chance of market penetration if it corresponds to the needs of the
consumer —that is, if it is in the average-sized bracket.
In dus t r ial_ Re q_u ir e men t s
As far as industrial installations are concerned, it is
apparent that the replacement of the internal combustion motor
by an electric generator-motor unit will involve very complex
and costly reconversion problems.
-------
XVI-5
For these reasons, it appears inevitable that the classical
vehicle will continue to be produced in parallel. The new
investments required will be very high^ and traditional equipment
will have to be written off.
The last factor coming into play will be purely technical in
nature: what motor, and especially what generator, will be used?
The Motor
Many studies have been made around the world on developing
electrical motors suited for automobile propulsion. This is a
very special problem, because the conditions of use are quite
different from those for industrial motors and locomotive engines,
For the latter, the weight and dimensions are not as important
as in the automobile, since longevity must be considerable:
2000,000 km represents the yearly use of a locomotive. On the
other hand, the automobile motor must be able to be mass-produced
and must be light and compact. Finally, it must be suitable for
the various operating regimes encountered by a vehicle used in
different circumstances and climates, by demanding or careless
customers.
Motors can be classed into two large categories, according
to whether they operate on direct or alternating current. The
development of semiconductors enables us to improve flexibility
and transmission output and to contemplate the production of DC
motors with no collector or supplying AC motors from a DC source.
Finally, it is possible (and may be of safety interest) to use
the motor for braking. Finally, let us note that the motor out-
put can be divided to activate two motorized axles or four
motorized wheels.
-------
XVI-6
The subject of motors is quite vast. Much research remains
to be done before we can draw up a complete, technical, and
economical balance sheet for the unit comprised of the motor and
the electronic power circuits.
The Generator
The sources of electrical energy, or generators, fall into
two categories: storage batteries or secondary generators which
only restore the energy which they store in electrochemical form
and cells or primary generators, which produce electrical energy
directly from fuel.
Storage batteries are well known, especially the lead storage
battery whose low mass-energy can be improved only slightly. Many
other types of storage batteries have been studied or are being
tested. It appears that real progress can be made, but at the
cost of hitherto undetermined mechanisms. Improvements are also
possible in the area of recharging, the duration of which is a
significant handicap. The latter is still encumbered by the
inevitable need for a complicated infrastructure connecting the
users with a distribution network, using a counter for tabulating
the energy absorbed.
Cells are also the object of many studies, and various fuels
have been investigated. Aside from dangerous, difficult to handle
products, or costly products, there are a number of possibilities.
The most interesting may be methanol combined with atmospheric
air: it is a relatively easily obtained fuel, with an
acceptable cost, which can be distributed by existing service
stations with no new infrastructure. However, if this product
can be used commercially with no major difficulties, its use to
produce electrical energy proves difficult: it is a stable body
-------
XVI-7
which liberates ions reluctantly even in the presence of noble
catalysts.
The latter must be disregarded due to their rarity. The
search for an effective catalyst which is cheap, easily prepared
and long-lasting is a big problem. Nor must one underestimate
the difficulties relative to technology of the cell itself and
its equipment.
Work is progressing, however, and various laboratories have
announced positive results with small cells which are still bulky
and expensive. However, their great merit lies in showing that
this path is open.
Vehicle Characteristics
The preceding conditions enable us to define the principal
characteristics of the vehicle.
We have seen that, because of safety requirements, it will
be very difficult if not impossible to make a very small vehicle.
We have also seen that this vehicle will be relatively heavy.
To guarantee adequate commercial success, it will have to
have a performance approximating that of comparable traditional
vehicles, or at least, not much lower if the users are to be
satisfied with it. This point is likewise important for traffic
flow since, as we know, traffic flow is better, the more homo-
geneous it is as far as acceleration and velocity capabilities are
concerned.
We can see that the vehicle with a classical lead storage
battery is far from meeting these conditions because of its low
-------
XVI-8
capabilities. The considerable mass of the required batteries
limits the acceleration, and the low mass-energy makes it
necessary to limit the velocity. The mass of the batteries
itself requires a heavier structure.
The problem of the small vehicle with a storage-battery is
well-known by Peugeot, who produced approximately ^400 during the
years of extreme restriction, 19^1-^3- It was the VLV, a light,
two-seater, town vehicle with lead batteries and, thus, of very
limited performance like all vehicles of the type. With this
little car, approximately 70 km could be covered with a peak
velocity of 36 km/h. Its total weight was 365 kg, of which 160
kg was batteries. Restored and modernized, it is certain that
its performance could be increased somewhat. Velocity could be
raised to 55 to 60 km/h perhaps, but only by retaining the old
type of chassis which does not conform to safety norms.
VLV PEUGEOT
The new batteries being tested will certainly be much better
than the present ones, but, short of significant progress, their
weight will still limit the possibilities of the vehicles using
t hem.
-------
XVI-9
To obtain the minimal performance defined above, it is
necessary to have a generator providing at least 450 wh/kg. Let
us recall that lead storage batteries have difficulty in reaching
40 wh/kg. Only a fuel cell appears capable of giving an adequate
energy density to ensure velocity, acceleration and action
radius.
One should note that the weight of a storage battery is
clearly proportional to the stored energy, meaning that the
weight increases rapidly with the required action radius. In the
case of a fuel cell, this is not the case because the weight of
the cell is a function of the power it supplies, and the action
radius is a function of the capacity of the fuel tank. With
sufficiently powerful fuels, as in the case of methanol, the
consumption in kg per kW/h is low relative to the overall weight
of the vehicle.
Since the fuel cell is the only solution leading to the
production of a vehicle answering the problem posed, we had to
find this cell. The studies carried out in this field by the
Alsthom Company interested Peugeot, since they were performed
in accordance with a method taking into consideration technological
problems from the outset. It is not a question of only performing
laboratory tests which cannot be transposed to industrial uses.
The development in parallel of two techniques is doubtless more
difficult, and perhaps slower, but only on the surface because
it eliminates the final necessity of spending a great deal of
time exploring paths which may not be practical. The two companies
collaborated to study a vehicle with a cell. Alsthom, brought to
Peugeot its experience in the field of electrical motors and
control circuits. This Association has been working since the
end of 1967, and Interesting results have been obtained.
-------
XVI-10
Above an experimental vehicle is shown. It is a small car
capable of transporting 11 passengers plus the driver with 175 kg
of luggage. The velocity is 95 km/h.
The vehicle is derived from a type J? Peugeot, and is to be
used for electrical transmission tests.
-------
XVI-11
It is equipped currently as a hybrid with a controlled
ignition leading to a constant velocity of the electric generator
working in parallel with the batteries. The generator assembly is
placed in the rear of the vehicle, and it is anticipated that it
can easily be replaced by a fuel cell.
Even in its present form, the performance of the vehicle is
comparable to that of traditional vehicles of its category.
In the special field of cells, the level of development
obtained with hydrazine cells has interested ESSO-USA, who has
just signed an important research contract with the Alsthom-
Peugeot Association to develop the methoanol-air cell.
The French Government has recognized the value of the program
and is aiding the Alsthom-Peugeot Association by loans supplied
through the General Delegation for Scientific and Technical
Research.
We have seen the reasons leading Peugeot to develop research
on an average-sized electrical vehicle. This type of vehicle
satisfies the requirements of reducing nuisances due to atmos-
pheric and noise pollution, and at the same time answers a
natural consumer demand.
For commercial success, it is also important that the
proposed vehicle need not require too many modifications
in the driving habits of the users. They should be able to
switch from one vehicle to the other without disorientat ion or
any special constraints. The problem of recharging storage
batteries, which requires time and special fixed installations,
would always be an unpleasant constraint.
-------
XVI-12
In conclusion, we feel that we should seek to produce a
vehicle which, for the user, will differ very little from
vehicles currently in use. The electric vehicle with a fuel
cell is a solution, but it is still far in the future.
------- |