&EPA
United States
Environmental Protection
Agency
Industrial Environmental Research EPA-600/7-80-017a
Laboratory January 1980
Research Triangle Park NC 27711
Advanced Combustion
Systems for Stationary
Gas Turbine Engines:
Volume I. Review and
Preliminary Evaluation
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Mention of trade names or commercial products does not con-
stitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-80-017a
January 1980
Advanced Combustion Systems
for Stationary Gas Turbine Engines:
Volume I. Review and Preliminary Evaluation
by
S.A. Mosier and P.M. Pierce
Pratt and Whitney Aircraft Group
United Technologies Corporation
P.O. Box 2691
West Palm Beach, Florida 33402
Contract No. 68-02-2136
Program Element No. INE829
EPA Project Officer: W.S. Lanier
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
FOREWORD
This report was prepared by the Government Products Division of the Pratt & Whitney
Aircraft Group (P&WA) of United Technologies Corporation under EPA Contract
No. 68-02-2136, "Advanced Combustion Systems for Stationary Gas Turbine Engines." It is
Volume I of the final report which encompasses work associated with the accomplishment of
Phase I of the subject contract from 12 December 1975 to 12 September 1976. The originator's
report number is FR-11405.
Contract 68-02-2136 was sponsored by the Industrial Environmental Research Laboratory
of the Environmental Protection Agency (EPA), Research Triangle Park, North Carolina
under the technical supervision of Mr. W. S. Lanier.
The authors wish to acknowledge the valuable contributions made to this program by
Mr. W. S. Lanier, whose skillful management and insight have been a key factor in the success
of the program.
The Pratt & Whitney Aircraft Program Manager is Mr. Robert M. Pierce; the Deputy
Program Manager is Mr. Clifford E. Smith. Mr. Stanley A. Mosier is Technology Manager for
Fuels and Emissions Programs at the Government Products Division of Pratt & Whitney
Aircraft Group.
Special recognition is due Mr. E. R. Robertson of the Component Design and Integration
Group, who was responsible for all drafting, hardware fabrication, and data processing
activities. The skillful assistance of Mr. R. J. Mador of the Combustion T&R Group,
Commercial Products Division, in the programing and development of the analytical com-
bustor model is also acknowledged.
Hi/ iv
-------
TABLE OF CONTENTS
Section
I INTRODUCTION
II PHASE I REVIEW AND PRELIMINARY EVALUATION
2.1 Objective [[[ 4
2.2 Approach [[[ 4
2.3 Task 1 Stationary Gas Turbine Engine Duty Cycle Review .............. 5
2.4 Task 2 Candidate Design Compilation Summary ................................ 26
-------
LIST OF ILLUSTRATIONS
Figure
1
2
3
4
5
6
7
8
9
10
11
General Relationship of Program Phases and Tasks
Distribution of Stationary Gas Turbine Engine International Sales
Distribution of Domestic Electrical Utility Units on Line in 1973
Average Power Settings of Domestic Electrical Utility Units on Line in 1973
Peak-Power Demand for Domestic Electrical Utility Units on Line in 1973
(60 Minute Duration or Longer)
Turbine Inlet Temperature Distribution for All Units Sold Since 1970
Pressure Ratio Distribution for All Units Sold Since 1970
Relative Severity Factor Distribution for Engines Surveyed
Representative Part-Power Characteristics for Free-Turbine Gas Generators
Representative Part-Power Characteristics for Direct-Drive Gas Generators
Distribution of Power For All Stationary Gas Turbine Units Sold Since 31
December 1970 (World-Wide Sales)
Page
3
7
10
11
13
14
15
17
18
19
21
12 Projected Net Plant Heat Rate For Simple Cycle Units (ISO Conditions;
No. 2 Fuel Oil; Peak Rating) 22
13 Projected Net Plant Heat Rate For Combined Cycle Unit (ISO Conditions;
No. 2 Fuel Oil; Base Ratings) (Unfired Boilers) 23
14 Projected Simple Cycle Plant Performance Characteristics (ISO Conditions;
Peak Rating; Distillate Fuel Oil) 24
15 Projected Combined Cycle Plant Performance Characteristics (ISO Condi-
tions; Peak Rating; Distillate Fuel Oil) 25
16 Projected Relative Severity Factor for Non-Nitrogenous Liquid Fuel 26
VI
-------
LIST OF TABLES
Table
I
II
III
IV
V
VI
VII
VIII
IX
Emission Goals
Average Powerplant Size Distribution
Stationary Gas Turbine Engine Operating Conditions
Stationary Gas Turbine Engine Utilization Rates
Dominant Modes of Operation
List and Brief Description of Combustor Concepts
Fuels and Constituents in the Model
Baseline Operating Conditions
Candidate Concept Classification
Page
4
6
8
8
12
26
32
35
37
Vll
-------
SUMMARY
This report describes an exploratory development program to identify, evaluate, and
demonstrate dry techniques for significantly reducing production of NO, from thermal and
fuel-bound sources in burners of stationary gas turbine engines.
In Phase I, duty cycle analyses were conducted to identify current and projected
dominant operating modes and requirements of stationary gas turbine engines. These analyses
indicate that the propensity for NO, to be generated in combustors of stationary gas turbine
engines will increase significantly in the future as compression ratios and turbine inlet
temperatures are increased to improve thermal efficiency and net plant heat rate. In ten years,
uncontrolled thermal NO. generation is predicted to double over today's levels; in 20 years, the
factor is predicted to triple. These predictions are based on an assumption that current fuel
characteristics will prevail into the future. Uncontrolled emissions would increase by an even
greater factor if future fuels contain significant amounts of chemically bound nitrogen.
An extensive survey was made of candidate combustor design concepts and an analytical
study was accomplished from which those concepts considered to have significant potential for
reducing production of NO, were identified. The initial compilation of 26 design concepts
included many variations of basic strategies such as fuel-rich combustion, ultralean combus-
tion, heat removal, fuel prevaporization, and fuel-air premixing. An assessment of the
NO.-control effectiveness of each concept was made using a combustor streamtube computer
code. The code employs a modular approach in the prediction of combustion emissions (NO.,
CO, and unburned hydrocarbons), with submodels for the internal flow field, physical
combustion (including droplet vaporization and droplet burning), hydrocarbon
thermochemistry, and NOi kinetics.
The results of the computer studies were drawn upon to select a group of concepts for
experimental screening in a bench scale combustor test rig. These experiments were carried
out under Phase II of the program, described in Volume II of this report series.
viu
-------
SECTION I
INTRODUCTION
On an overall basis, the stationary gas turbine engines used in industrial and utility
applications make a small contribution to air pollution. Less than 2.5 percent of the total
oxides of nitrogen (NO,) emitted from stationary sources domestically in 1972, for example,
were attributed to the gas turbine (Reference 1). Although this amount represents a small
contribution to the deterioration of air quality on a gross scale, it can be a cause of significant
local concern, especially in the vicinity of engine installations where the background pollution
level from other sources might already be objectionably high. Therefore, it is necessary that
means be developed for reducing the concentrations of undesirable exhaust emissions, particu-
larly NO,, from stationary gas turbine engines.
Pollutants produced in the combustors of gas turbine engines include both particulates
and gaseous species. The particulates of most concern are the soot-like matter generated in
fuel-rich regions of the combustion chamber, which are discharged from the engine as a visible
plume. The invisible gaseous species of concern consist principally of unburned hydrocarbons
(UHC), carbon nonoxygenated derivatives of the carbon-based fuels burned in the combustor.
The variety of hydrocarbon compounds identified in gas turbine engine exhaust gas has been
reported to be significant (Reference 2). The major species of the total oxides of nitrogen are
nitric oxide (NO) and nitrogen dioxide (NO2). Of the two, the concentration of NO has been
reported to that of N02 (Reference 3).
In recent years, a significant effort has been directed to reducing visible emissions from
gas turbine engines. Although this effort commenced to accommodate aesthetic arguments
regarding smoke and haze, it also contributed to reducing the carcinogens both comprising and
associated with the emitted soot. Unfortunately, techniques implemented to reduce the
production of soot generally served to increase production of the oxides of nitrogen from
nitrogen and oxygen in the air (thermal NO,). When techniques were incorporated to reduce
soot, concentrations of thermal NO, were generally increased.
The soot-NO, tradeoff effect has also been encountered during implementation of
techniques to reduce the concentration of the objectionable emissions UHC and CO. Both are
nonequilibrium byproducts of the reactions between engine fuel and air; under ideal
thermodynamic conditions neither should be present in combustor exhaust gas. However, when
the engine is operated at low-power (idle) conditions, the inlet air temperature and pressure
and the overall fuel-air ratio are low; therefore, the reaction temperature and rates for the fuel
breakdown and consumption processes are either very low or kinetically frozen. Consequently,
thermodynamic equilibrium is not achieved in the combustor and unacceptable quantities of
CO and UHC species appear in the exhaust gas. Techniques implemented to enhance the
consumption of UHC and CO, such as locally increasing the fuel-air-ratio to raise the adiabatic
flame temperature, often have resulted in an increase in the rate of production of thermal NO,.
An additional problem is encountered when fuel is burned that contains chemically
bound nitrogen. Conditions under which thermal NO, is produced generally enhance prod-
uction of NO, from the reaction of fuel-bound nitrogen and oxygen (fuel-bound NO,).
However, the leanburning techniques commonly implemented to reduce the production of
thermal NO, are ineffective in reducing the production of fuel-bound NO,.
It has been suggested by some that priority be given to developing means for reducing the
production only of thermal NO,. This position might be satisfactory if operators of stationary
gas turbine engines could be guaranteed a never-ending supply of high-quality fuel having
-------
negligible bound-nitrogen content. Unfortunately, there is the real possibility that as the
energy shortage intensifies, fuels offered for gas turbine engine use will include petroleum-base
crude oil and heavy distillates and both liquid and gaseous derivatives of oil shale and coal; the
bound-nitrogen content of these may not necessarily be insignificant. Therefore, it is essential
that technology be developed for reducing the production of thermal and fuel-bound NO, in
gas turbine engine combustors while maintaining soot, UHC, and CO concentrations within
environmentally acceptable limits.
In this exploratory development program, the overall goal was to develop technology to
significantly reduce the total production of NO, when liquid fuels and gaseous low-Btu fuels
are burned with air in combustors of stationary gas turbine engines. The effort was ac-
complished in four interrelated phases that comprised a total of 12 tasks, as shown in Figure 1.
This report documents the work performed under Phase I of Contract 68-02-2136. The
work accomplished under the remaining phases is documented separately in Volume II,
Volume III, and Volume IV.
An Appendix is included for conversion from commonly used English units to SI units.
-------
Task 1
Stationary Gas
Turbine Duty
Cycle Review
1
Task 4
Design
Selection
Task 2
Candidate
Design
Compilation
Task 3
Analytical
Screening
Task 1
Screening
Experiments
Task 3
Program
Goal Review
Task 2
Model
Update
Task 1
Combustor
Design
Task 2
Test Plan
Preparation
Task 1
Combustor
Installation
Task 2
Performance
Testing
Task 3
Analysis and
Reporting of
Test Results
Phase I
Review
Phase II
Bench - Scale
Evaluation
Phase III
Combustor Rig
Preparation
Phase IV
Verification Testing
Figure 1. General Relationship of Program Phases and Tasks
-------
SECTION II
PHASE I REVIEW AND PRELIMINARY EVALUATION
2.1 OBJECTIVE
The overall objective of Phase I was to review and conduct a preliminary evaluation of
various gas turbine combustor designs that could potentially be capable of meeting the
emissions goals shown in Table I. These designs were to incorporate dry techniques only and
were to address NO, emission emanating from both thermal and fuel-bound sources.
TABLE I
EMISSION GOALS
Fuel Characteristics Exhaust Gas Characteristics
Type Nitrogen Content NO,* CO
No. 2 Fuel Oil, Trace 50 ppmv 100 ppmv
Low-Btu Gas Trace 50 ppmv 100 ppmv
No. 2 Fuel Oil Up to 0.5% 100 ppmv 100 ppmv
"Corrected to 15% O2
Wet techniques for NO, control usually imply the use of water or steam injected into the
combustion process. This type of technique has been shown to aid in the reduction of thermal
NO, by virtue of lowering combustion temperatures. The injection of ammonia or NO into the
engine exhaust have also been considered as wet techniques. Dry techniques for NO, control
are methods not requiring the use of water, steam, or other substance. These techniques
involve controlling the combustion process itself in ways to reduce pollutant emissions.
2.2 APPROACH
The Phase I analytical effort was accomplished in four complementary tasks. In the first
task, engine duty-cycle analyses were conducted to identify current and projected dominant
modes and requirements of stationary gas turbine engines used in a variety of applications.
This information was needed for two purposes: to serve as an additional constraint against
which the potential of candidate N0,-control concepts might be evaluated and to provide
guidelines or critical criteria for the design of stationary gas turbine engine burners.
In the second task, potential N0,-control techniques were identified that might be
incorporated into the design of stationary gas turbine engine burners. In this effort, two
compilations were prepared. The first was a detailed list of N0,-control concepts and the
second was a collection of emission data on existing combustor designs. This information was
needed to serve as a data bank from which the most promising N0,-reduction concepts could
be selected for experimental evaluation and eventual implementation to a full-scale stationary
gas turbine engine burner.
In the third task, the effectiveness of the N0,-control concepts identified earlier was
estimated using analytical modeling techniques. Concepts selected in the previous task for
their N0,-reduction potential were translated into preliminary, bench-scale designs and were
simulated using a computer model. Predictions of emission concentrations for these designs
were generated using the model and parametric studies of variables that were considered to be
of importance in influencing the rate of production of objectionable emissions.
-------
In the fourth, and final, task of Phase I, candidate NO.-control techniques were selected
for experimental evaluation in Phase II. The selection was made from information generated in
the preceding three tasks of Phase I.
2.3 TASK 1 STATIONARY GAS TURBINE ENGINE DUTY CYCLE REVIEW
Engine duty cycle analyses were conducted to identify current and projected dominant
modes and requirements of stationary gas turbine engines used in a variety of applications. To
accomplish this task, a survey of existing and planned gas turbine engine installations was
conducted to determine their current and projected application profiles. The results of this
study range from the more obvious conclusion that electrical utilities have dominated, and will
continue to dominate, the stationary gas turbine engine market to the more subtle conclusion
that the dominant mode of operation is, by far, derated-power running. As far as the problems
of exhaust emission control are concerned, trends in thermodynamic cycle characteristics of
stationary gas turbine engines will make the design of combustion systems increasingly more
difficult to meet stringent emission requirements in the future.
2.3.1 Scope
In this study, stationary gas turbine engines (SGTE) are considered to be those that
operate at fixed sites. Mobile gas turbine powered generating units, which can be moved from
site to site, are generally included because they operate only when set up at a fixed location;
however, marine and other transportation applications are excluded. Using this definition the
SGTE market can be divided into four main categories: electrical utilities, pipeline trans-
mission, industrial, and emergency/stand-by power.
In the electrical utility category, gas turbine engines are used to drive generators that
produce electricity at ratings in the range from 15 to 100 megawatts per unit. Historically,
these units have been used to supplement plant-generating capacity during short periods of
peak electrical demand, resulting in a total operating time of less than 1500 hours/unit/year.
However, increasing thermal efficiencies, combined with long lead times associated with
bringing a conventional steam plant on line, are pushing these gas turbine units into higher
utilization.
In pipeline transmission applications, gas turbine engines producing up to 25,000 horse-
power are used as compressor drives. These engines are generally remotely operated and,
unlike their counterparts in the electrical utility industry, are run continuously.
In the industrial segment of the SGTE market, gas turbine engines are used in a wide
variety of processes that require heat or power. In addition to supplying electrical or
mechanical power, high-pressure air can be extracted from the compressor; and, with the
addition of a waste-heat boiler, process steam can be produced using the sensible heat in the
turbine exhaust gas. Engine operation in the industrial segment is similar to that in pipeline
applications with the possible exception that industrial engines are subject to down-time in
other plant systems and will, therefore, experience more shutdown and restart cycles.
The emergency and stand-by power SGTE units are operated by institutions that are
particularly vulnerable in the event of power failures. Hospitals, communication networks, and
the military operate relatively small SGTE units in the event of mainline power interruption.
When compared to the other SGTE market segments, both generating capacity and annual
utilization of stationary gas turbine engines in this category are extremely low.
-------
2.3.2 Power Distribution
The relative depth of the preceding four market segments was assessed to identify
dominant modes of engine operation. This assessment was needed in order to direct re-
duced-emission combustor designs toward the proper portion or portions of engine duty cycles.
To accomplish this effort, a survey was conducted encompassing all reported gas turbine
engines sold internationally during the five-year period beginning 31 December 1970 and
ending 31 December 1975. The survey was limited to this five-year period for several reasons.
First, the availability of accurate, published information on recent SGTE sales was greatest
during this period; second, during this period, gas turbine engines significantly intruded into
diesel-dominated pipeline and emergency power markets; and, third, by addressing engine
sales made during this period, the assessment could be slanted away from the older,
less-sought-after models of stationary gas turbine engines.
The results of the power distribution survey are summarized in the histogram of Figure 2
in terms of purchased power generating capacity and number of units. The electrical utilities,
having purchased over 80% of the total power generating capacity, represent the dominant
market sector. Pipeline and industrial sector power generating capacities are nearly the same
at 9 and 7 /<, respectively, of the total. Finally, the purchased power generating capacity for
the emergency and standby markets is the least of the four major sectors; accounting for less
than one percent of the total.
Figure 2 also depicts the purchased power distribution for the four major sectors in terms
of SGTE units. The term "units" refers to the total number of powerplants. For pipeline and
emergency applications, the number of units is indicative of the number of engines purchased.
However, for electrical utility and, to a lesser extent, industrial applications, the number of
units is not necessarily indicative of the number of engines purchased. In these cases, the
buyer purchases a block of power that might consist of more than one gas generator. In Figure
2, the histogram represents approximately 28,000 megawatts of power distributed among 1,275
installations.
From the information presented in Figure 2, an estimate of the average size of the
powerplant purchased in each major sector can be made. The results are shown in Table II.
TABLE H
AVERAGE POWERPLANT SIZE DISTRIBUTION
Market Sector Average Power/Unit
Electrical Utilities 50 MW
Pipelines 5.6 MW (7,500 hp)
Industrial 10 MW
Emergency/Stand-By 1.8 MW (2,500 hp)
Because military operators have been included in the emergency and standby category,
the average power per unit is somewhat inflated. Private operators in this sector rarely exceed
2,000 horsepower
-------
100
80
60
3
o
i
s.
40
20
£
CO
9
c
'c-
(0
- ti-
ed
I
. c.
I
0)
9
a
E
Electrical
Utilities
Pipelines
Industrial
Processes
Emergency
and
Stand-By
Figure 2. Distribution of Stationary Gas Turbine Engine International Sales
-------
2.3.3 Stationary Engine Operation
The operation of a stationary gas turbine engine can be described in terms of engine duty
cycles. Duty cycles are generated from information on engine operating conditions and
utilization rate. The five engine operating conditions that are generally referred to are
identified in Table III in terms of design-rated power. The three utilization rates that are also
generally referenced are shown in Table IV; the terms usage, range, and average usage refer to
the time during a given year of service in which the engine is connected to load. Combinations
of ratings from Table HI and rates from Table IV define stationary gas turbine engine duty
cycles.
TABLEm
STATIONARY GAS TURBINE ENGINE
OPERATING CONDITIONS
Rating Percent of Design-Rated Power
Maximum Capability 115
Reserve Peak 105
Peak 100
Electrical Baseload 90
Baseload 85
TABLE IV
STATIONARY GAS TURBINE ENGINE UTILIZATION RATES
Usage Range Average Usage
Rate (hr/yr) (hr/yr)
Peaking 0-2000 100
Intermediate 2000-6000 4000
" Baseload 6000-8000 7000
For two of the four major market sectors, the duty cycle is simple and should remain
unchanged in future applications.The simplest is the emergency and standby unit; purchased
to meet a specific power requirement, these engines will always be operated at their rated
capacity. These units can be automatically started by loss of mainline power with no need for
throttle-type control.
The second of the simple duty cycles is characteristic of pipeline pumpers. Whereas the
emergency/standby units represent the ultimate peaking application, the pipelines are the
epitome of baseload operations; once installed, these units are operated continuously. Engine
time-between-overhauls (TBO) is generally impossible to assess because, partly due to the
remoteness of most installations, maintenance is being accomplished more and more on an
"on-condition basis." Under this philosophy, periodic monitoring of engine health leads to a
yes or no decision to perform some maintenance function. More often than not, engine failure,
either actual or imminent, is the criterion for shutting down.
In terms of power settings, the gas pumpers operate at baseload or less. With the
utilization rate and maintenance philosophy characteristic of this market, it is easy to
understand the life-prolonging advantages of derated operation, even at the expense of
somewhat increased fuel consumption.
The industrial-process sector resembles the pipeline market in utilization and power
setting; however, the number of shutdown and restart cycles is significantly higher. Because
these units interact with the operation of the entire industrial complex, they are subject to
8
-------
down-time in other plant systems. Published information is scarcest on this segment of the
SGTE market, but the number of cycles per year is considered to be considerably less than
that required of a peaking plant in the electric utilities category. In summary, like the gas
pumper, the dominant mode of operation of these SGTE units should be classified as baseload
and continuous.
As shown in Figure 2, the electric utilities sector operate an overwhelming 83% of the
total SGTE generating capacity purchased during the survey period. Prior to 1970, the gap
between the utilities and the other three major applications was even wider; indications are
that the trend will continue in the future. In addition to accounting for the largest share of
capacity, the modes of operation in the electrical utility sector are the most complex. The
utilities must meet the constantly changing demand for power and they must do so, for the
most part, within Government-regulated profit margins. Fortunately, the rather public nature
of this sector has led to a considerable amount of published information on their operational
characteristics. In fact, numerous scenarios have been written on the theory of pow-
er-generation mix and where the gas turbine engine should fit into the baseload, intermediate,
and peaking modes of the utility market.
To determine how gas turbine engines are actually operated in practice, a second survey
was conducted that encompassed all SGTE units in the electrical utilities sector that were on
line domestically in 1973. At the time of this survey, comprehensive data on SGTE operations
in the post-1975 years was not available.
Figure 3 is a histogram developed from this survey that shows the distribution of
domestic electrical-utility SGTE units on line in 1973 as a function of annual utilization.
Encompassed in this survey are over 1,000 domestic installations whose combined generating
capacity was over 29,000 megawatts. Seventy-two percent of these installations operated as
peaking units; they had an annual utilization of less than 2,000 hours; 23% of the units were
operated as intermediate loaders; and the remaining 5% functioned as baseload plants.
Although the dominant mode was obviously one of peaking, the electrical utility units, unlike
their counterparts in the other three major sectors, are far from fixed in their operating
characteristics. The peaking units are comprised primarily of simple-cycle machines while the
intermediate and baseload operations are a mixture of combined-cycle generators and older
simple-cycle units that were pressed into extended service at derated power settings. The
latter is primarily a stopgap solution to delays encountered in construction of conventional
baseload plants.
Whereas the utilization rate distribution for the electrical utility SGTE units shown in
Figure 3 is primarily one of peaking, the dominant operating condition or power level, was
found to be less than baseload, as shown in Figure 4; the units were run mostly at less than
85% of their design-rated capacity. Approximately 89% of all engines in the survey were
operated at an average power setting of baseload or lower; less than four percent of the engines
operated above a purely peaking mode, i.e., design-rated power or greater. Therefore, since the
histogram in Figure 4 represents all engines in the field surveyed, a single gas turbine engine
designed to typify operation in the electrical utilities sector could be considered to have a duty
cycle defined by the distribution shown. In other words, the bulk of its operation, 89%, would
be at baseload power or less and only 4 % of its operation would be at rated power or more. In
effect, then, the ordinate of Figure 4 can be considered to represent the percent of total
operating hours for the average engine.
-------
50
*
0
CO
.1
I
2
(0
2
o
£
8
«
0.
u
O
ro
O
-1
O
72% of Total
Installations
I I I
Average Utilization = 18.5%
(1620 hr/yr)
10 T 20 30 40 50 60 70 80 90 100
A Percent of Annual Utilization
Figure 3. Distribution of Domestic Electrical Utility Units on Line in 1973
10
-------
*:*»
22
20
18
16
CO
g
1 14
S
In
c
1 12
o
§ 10
<5
Q.
8
6
4
2
0
A
B
E
P
R
M
Overall Average Po\
Level = 63% of Ra
Capacity
Baseload
Electrical Baseload
Peak
Reserve Peak
Maximum Capabilit
wer
ted
y
0 10 20 30 40 50 601 70 80
i 70 80 i 90 100 i 110 i
120
A B E P R M
Average Power Level, % of Rated Capacity
Figure 4. Average Power Settings of Domestic Electrical Utility Units on Line
in 1973
11
-------
Figure 5 is a histogram that describes the peak-demand requirements of the SGTE units
in the 1973 survey. Forty-two percent of the installations were never required to operate above
their design-rated power; however, 20% were required to run at the maximum-capability
rating to meet peak-power demands. Now, recognizing that Figure 5 represents a distribution
of demand, and not supply, and assuming that even the best engine is not going to exceed 120
percent of rated power, Figure 5 indicates that approximately 10% of the installations in the
survey were unable to meet the peak demand: resulting in brownouts. Consequently, this
shows that although the dominant power setting is baseload, engines purchased for the electric
utilities sector will be required to operate at maximum-capability rating.
In summary, then, the dominant modes of operation for stationary gas turbine engines in
the four major-use sectors are shown in Table V.
2.3.4 Thermodynamic Cycle Characteristics
Investigation of the thermodynamic-cycle characteristics of the stationary gas turbine
engines sold during the survey period revealed that turbine inlet temperatures and overall
pressure ratios were distributed as shown in Figures 6 and 7. Units constituting over 80% of
the total power capacity involved were designed to operate at turbine inlet temperatures of
2000°F or less and at overall pressure ratios of 11:1 or less. An "average" engine, designed to
operate at these values of turbine inlet temperature (2000°) and overall pressure ratio (11:1)
was considered to be representative of current units. This "average" engine was selected to
serve as the baseline unit in a study to appraise the relative difficulty in resolving the problem
of thermal-NO, emission production in the burners of future stationary gas turbine engines.
2.3.5 NO, Production Considerations
A severity factor was developed (Appendix I) to serve as a measure of the NO,-
production capacity of gas turbine engine combustors. Intrinsic in its formulation are overall
pressure ratio and turbine inlet temperature for the engine of interest. From this, a relative
severity factor was then formulated (Appendix I) that provided a measure of the NO,
production capacity of the engine of interest relative to that of the baseline engine. The higher
the severity factor and the relative severity factor, the more difficult the NO.-control problem.
TABLE V
DOMINANT MODES OF OPERATION
Application
Sector
Emergency/
Standby
Pipelines
Industrial
Electric
Utilities
Utilization
Rate
100 hr/yr
Continuous Duty
Continuous Duty
1620 hr/yr
Duty
Cycle
Intermittent at maximum power
Baseload
Baseload
Predominantly baseload or less
but required to have maximum
power capability
Comments
Cycles dependent on availability
or mainline power
"On-condition" cycle require-
ments
Cyclic requirements tied to other
plant systems
Approximately 400 cycles/year;
distribution of power setting de-
fined by Figure 4
12
-------
£O
*)A
99
9fi
1 D
W 10
O
oi
"55
c
03
B 14
B
c 10
« i«=
o
0)
Q.
10
8
6
4
2
0
'
B Baseload
E Electrical
Baseload
P Peak
R Reserve Peak
I M Maximum Capability
60 70 80 & 90 100 1 110 i 120 130 140 150 160
BE PR M
Peak Demand on Plant, % Rated Capacity
Figure 5. Peak-Power Demand for Domestic Electrical Utility Units on Line in
1973 (60 Minute Duration or Longer)
13
-------
s
o
o>
0)
CL
60
50
40
30
20
10
1400 1500 1600 1700 1800 1900 2000 2100 2200
Turbine Inlet Temperature, °F
Figure 6. Turbine Inlet Temperature Distribution for All Units Sold Since
1970
14
-------
70
60
50
£ 40
o
"o
c
1 30
Q.
20
10
7 9 11 13
Overall Pressure Ratio
15 17
19
Figure 7. Pressure Ratio Distribution for All Units Sold Since 1970
15
-------
Figure 8 shows the relative severity factor distribution for the stationary gas turbine
engines included in the utilization survey. Over 80% of the units comprising the power
capacity purchased during the survey period were designed to operate with a severity factor
equal to or less than that of the baseline engine. Approximately 19% of the units surveyed
were designed to operate with a more severe NO, problem than the baseline engine. However,
less than 0.1% of the units were designed to operate with a relative severity factor of 1.4 or
greater. It should be noted that the severity and relative severity factors were developed
considering only thermally formed NO,. Fuels containing significant nitrogen concentrations
present further NO, forming potential beyond that assessed in this study. In a gas turbine with
an uncontrolled combustion environment, a large portion of the fuel-bound nitrogen is
converted to NO, in the combustion process.
2.3.6 Configurations! Influences on Emission Control
Four types of stationary gas turbine engine configurations were investigated briefly to
determine the influence of their characteristics on two of the principal thermodynamic
variables influencing the production of NO, in combustors, viz. pressure level and turbine inlet
temperature. The four engine types considered were as follows:
a. Single-Spool/Direct Drive (SSDD), in which all turbomachinery is con-
tained on a single shaft that is directly connected to the load
b. Dual-Spool/Direct Drive (DSDD), in which the rotating hardware is
divided between two spools, a low-pressure shaft and a high-pressure
shaft; the low rotor is directly connected to the load
c. Single-Spool/Free Turbine (SSFT), in which all of the turbomachinery is
contained on a single shaft, as with the SSDD, but the shaft is not
mechanically connected to the load; power is supplied by a free turbine
d. Dual-Spool/Free Turbine (DSFT), in which two shafts are used, as with
the DSDD, but as with the SSFT, the power is supplied by a free turbine.
Although each of these engine arrangements offers certain advantages and disadvantages
for the stationary gas turbine user, no attempt has been made to compare the overall net
desirability of each. An attempt has been made, however, to provide an indication of the
relative NO.-production potential of the four gas turbine engine configurations predicated
upon the overall pressure ratio and turbine inlet temperature associated with each type of
machine.
As discussed earlier, the dominant operating condition or power level for electrical-utility
gas turbine engines on-line domestically in 1973 was well less than baseload (Figure 4).
Consequently, an estimate was made of the part-power, or derated, pressure ratio-turbine inlet
temperature characteristics for the four principal engine types. Figure 9 shows predicted
representative variations in these variables with power level for single and double-spool free
turbines, and Figure 10 shows predicted and measured (Reference 4) variations for direct-drive
machines. In Figure 9, both turbine inlet temperature and pressure ratio can be seen to
decrease with reducing power level because engine speed, air flowrate, and fuel-air ratio are
also decreasing. However, as shown in Figure 10, although the turbine inlet temperature for a
direct-drive machine decreases as power level is reduced to engine start conditions, due to
decreasing fuel-air ratio, the compressor pressure ratio decreases little. This trend is the result
of a direct-drive machine operating at constant speed and air flowrate over its power range.
16
-------
60
50
40
s
o
t 30
o
20
10
*Less Than 0.1%
Above 1.4
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
t
Baseline
Relative Severity Factor for NOX Emissions
Figure 8. Relative Severity Factor Distribution for Engines Surveyed
17
-------
CO
CC
2
0> O)
2
Q.
E
o
O
120
100
80
60
40
20
NSingle Spool (SSFT)
Dual Spool (DSFT)
I
2
o
Q. C
E 9>
o w
50
20
40 60 80
Power Level, % of Design
100
120
Figure 9. Representative Part-Power Characteristics for Free-Turbine Gas
Generators
18
-------
120
100
SFT for Reference
110
40 60 80
Power Level, % of Design
Figure 10. Representative Part-Power Characteristics for Direct-Drive Gas
Generators
19
-------
Because the combustion chamber pressure at a given power level can be significantly
higher for the direct-drive engine than for a free-turbine engine, there is a greater propensity
for NO, to be generated in the direct-drive system than in the free turbine. Of course, the
trends shown in Figures 9 and 10, are presented in terms of "percent of design level" and not
in terms of absolute pressure. Consequently, it is certainly possible that a low-pressure
direct-drive machine might be operating at an overall value of pressure that is lower than, say,
a free turbine at 50% power. However, as discussed in the following section, the trend in
future stationary gas turbine engines is toward higher pressure ratio and higher turbine inlet
temperature operation to achieve increased thermal efficiency.
2.3.7 Projections
A series of projections have been made as a result of the work accomplished in this task.
First of all, predicated upon the distribution of stationary gas turbine engine-provided power
for all units sold since 31 December 1970, shown in Figure 11, four general conclusions can be
made:
a. There will be an ever-widening gap between the electrical utilities and the
other market sectors.
b. There will be an ultimate saturation of pipeline and industrial-directed
units.
c. There will be an increasing use of combined-cycle plants.
d. The positive slope of simple-cycle purchases indicates a continuing need
for low-heat-rate, basic gas turbine engines for continued peaking and
intermediate applications.
From an energy conservation viewpoint, the trend in gas turbine performance will be in
the direction toward achieving reductions in the net plant heat rate. Figure 12 is an estimate of
the rate of reduction in heat rate with time for simple-cycle machines. The trend shown is
predicated upon published information on units ordered through 1978, and upon estimates of
heat rates for units incorporating four major technological improvements incorporated into
production machines through 1995. Figure 13 is a companion plot to Figure 12 for com-
bined-cycle stationary gas turbine engines. Again the trend shown has been established by
published information and logical projections regarding the implementation of major techno-
logical improvements into production machines.
Performance characteristics for future simple and combined-cycle units are shown in
Figures 14 and 15, respectively. Pressure ratios and turbine inlet temperatures for engines
incorporating the projected major technological improvements identified in Figures 12 and 13
are identified. The trend in both simple and combined cycle machines is toward ever
increasing thermal efficiencies and specific output power as a result of increasing pressure
ratios and turbine inlet temperatures.
Figure 16 summarizes the preceding information on future engine operating character-
istics and duty cycles in terms of the NO.-control burden. As shown, the relative severity
factor for future stationary gas turbine engines is predicted to increase significantly. The
datum upon which the curve was based is a 10,000 horsepower gas pumper that was estimated
to currently have a severity factor of 1.8. However, even if a datum engine were selected that
currently had a relative severity factor of unity, in ten years the relative severity factor could
increase by a factor of nearly 1.6; in twenty years the factor could increase by a factor up to
approximately 2.8.
20
-------
It is important to emphasize that these predictions are based on an assumption that
current fuel characteristics will prevail into the future. Uncontrolled emissions would increase
by an even greater amount if future fuels contain significant quantities of chemically bound
nitrogen.
24
20
16
I
"5
in
in
s
(0
D
Q.
s
3
3
O
12
-4
Electrical
Utilities
Combined
Cycle
Pipeline and
Industrial-
Emergency and
Standby
71 72 73 74 75 76 77 78 79
Year of Installation (as Planned)
80
81
82
Figure 11. Distribution of Power For All Stationary Gas Turbine Units Sold
Since 31 December 1970 (World-Wide Sales)
21
-------
to
to
IB
16
X
d. 14
w
C.
i
x.
3
0
£ 12
(D
o O5
35 &
Q3
-Q5 Q7
U ^»
r- Convec
/
\ ***""*'
tive Air-Cooled /
r- Precook
/
B *.^
Q5 GE
Q7 GE
R1 We
~2 We
1 [J3 We
Q5 We
A Tur
O Air<
iirfoils
d Cooling Air
r- Water C
/ Vanes
/
3 J
Frame 5
Frame 7
stinghouse 191
stinghouse 251
stinghouse 352 ~
stinghouse 501
bodyne Type II
:raft Derivative
xjled
y-Water Cooling
' All Airfoils
M
mm
65
70
75
80 85
Year of Order
90
95
100
Figure 12. Projected Net Plant Heat Rate For Simple Cycle Units (ISO Conditions; No. 2 Fuel Oil; Peak Rating)
-------
to
Oi
11
10
>
X
r" 9
.c
1
X.
3
m
£ 8
as
DC
1
§
Q. '
0>
6
5
2
'^f
V V
^V 40
Xj)5
Precoo
Convective
| Airfoils
1
A
^
^^ ^"*«»
ed Cooling Air -
Air-Cooled
__
B
^..^
r
Wa
Ct
Vs
ter
x>led
nes
§GE Stag 100
GE Stag 200
GE Stag 300
§GE Stag 400
GE Stag 500
Westinghouse Pace
A Turbodyne Mega-Pak-CC
Q Aircraft Derivative
Gasification Plants
Lead to Significant Use
of Combined Cycle
Baseload Plants
C <>>-
"T
1 Water
* Cooling
All
Airfoils -
D "~"
T
65
70
75
80
85
90
95
100
Year of Order
Figure 13. Projected Net Plant Heat Rate For Combined Cycle Unit (ISO Conditions; No. 2 Fuel Oil; Base Ratings) (Unfired Boilers)
-------
to
2
"o
is
CO
Compressor
"""" Pressure
D. Ratios
_\20
Convective Air-cooled
Airfoils (1980)
Precooled Turbine
Cooling Air (1985)
Water-Cooled
Vanes (1990)
D Water Cooling All
Airfoils (1995)
1
Turbine Inlet
Temperature, °F
180 220
Specific Output Power, kw/lb/sec
240
300
Figure 14. Projected Simple Cycle Plant Performance Characteristics (ISO Conditions; Peak Rating; Distillate Fuel Oil)
-------
60
to
56
>
I
u
c
2
I
52
44
Compressor Pressure Ratio
32
Turbine Inlet
Temperature, °F
40
100
2.'?00
A Corrective, Air-Cooled Airfoils (1980)
B Precooled Turbine Cooling Air (1985)
C Water-Cooled Vanes (1990)
D Water Cooling All Airfoils (1995)
140
180 220 260
Specific Output Power, KW-hr/sec
300
340
380
Figure 15. Projected Combined Cycle Plant Performance Characteristics (ISO Conditions; Peak Rating; Distillate Fuel Oil)
-------
s
o
o
x
o
1970
1975
1995
2000
1980 1985 1990
Year of Order
Figure 16. Projected Relative Severity Factor for Non-Nitrogenous Liquid Fuel
2.4 TASK 2 CANDIDATE DESIGN COMPILATION SUMMARY
Twenty-six concepts were identified that offer the potential for reducing the production
of NOX from thermal and fuel-bound sources in stationary gas turbine engines. These concepts
were selected from two compilations generated from the research literature: a detailed list of
N0,-control concepts and a collection of emission data on existing low-NO, combustor designs.
The compilations served as a data bank from which NO.-reduction concepts were examined.
Although the concepts chosen involve a limited number of basic NO.-control strategies and
could be considered to be subsets of a major category or of one of the other concepts identified,
they represent different approaches to implementing those strategies. Table VI is a list of
these combustor concepts and a brief description of each.
TABLE VI
LIST AND BRIEF DESCRIPTION OF COMBUSTOR CONCEPTS
Concept No. Title and Description
1 Low-Intensity Flame
Extended length flame jet, fuel rich, mixes slowly with surrounding air.
Bound nitrogen NO, reduced under fuel rich conditions within flame jet.
2 Premixing Catalytic Burner
Catalyst preceded by premixing/preburning module. In low power preburn-
ing mode, fuel is partially burned and mixed with air before entering
catalyst, thereby ensuring uniform high temperature mixture for efficient
operation of catalyst.
26
-------
TABLE VI
LIST AND BRIEF DESCRIPTION OF COMBUSTOR CONCEPTS (Continued)
Concept No. Title and Description
3 Superlean With Heat Recirculation
Premix tube air preheated indirectly in liner convective cooling passages or
by other means to improve fuel vaporization and widen flammability limits.
i Lean burning for low thermal NO,.
4 Superlean With Preburner
Premix tube air preheated directly by preburner to improve fuel vapor-
ization and widen flammability limits. Lean burning for low thermal NO,.
5 Heat Removal
Coolant tubes inside the combustor reduce temperature of rich burning
mixture before excess air is added for CO oxidation and final dilution.
6 Quench Reheat
Main burning zone is rich, resulting in low flame temperature. This mixture
is rapidly quenched to a very lean equivalence ratio, causing excessive
formation of CO but very little NO,. In a reheat zone, effluent from a pilot
burner heats the mixture to an intermediate temperature for CO consump-
tion.
7 Staged Centertube Burner
An axially staged burner configuration with swirl mixing. Concentric center-
tubes of different lengths determine the axial fuel-air distribution. As an
experimental device the configuration allows easy variation of burning and
mixing zone lengths. Both rich and lean air schedules were considered.
8 Exhaust Gas Recirculation
Gas abstracted near the end of the primary zone, and mixed with fresh air
in passages leading to the premix tube. Heat lost from the mixture in the
passages (to surrounding inlet air) causes a reduction in flame temperature.
9 Hydrogen Enrichment
Hydrogen injected along with fuel results in lower lean flammability limit of
primary zone mixture.
10 Surface Combustion
Flame stabilized in contact with surface of porous plate flameholder.
Coolant tubes imbedded in plate remove heat from flame.
27
-------
TABLE VI
LIST AND BRIEF DESCRIPTION OF COMBUSTOR CONCEPTS (Continued)
Concept No. Title and Description
11 Distributed Flame
Perforated plate flameholder produces many small flames, each stabilized
separately, eliminating large scale recirculation and reducing residence
times.
12 Ceramic Liner
Wall quenching of flame diminished by the elimination of film cooling air
and a higher allowable wall temperature.
13 External Combustion
Combustor located outside the gas turbine engine. Geometrical constraints
and residence time limitations of on-board combustors are eliminated.
14 Boost-Air Dilution
Dilution air injected at higher pressure than other burner airflow to achieve
higher mixing rate and reduce lag time in reaching desired equivalence
ratios. Compressor or other means of achieving pressure differential
reJquired.
15 Artificial Excitation
Vibrational excitation of burning gases in the combustor to increase reaction
rates, allowing residence times to be reduced. Method of excitation may be
acoustic, electronic, or other means.
16 Extended Injector
Perforated plate flameholder with tubular extension pieces. By varying the
number and length of tubes, their routing, and discharge points, mixture
and temperature profile can be controlled.
17 Pebble Bed
An external burning concept with a vertical discharge low velocity com-
bustor. Ceramic (or other material) pebbles are fed in near the exit, fall
through the flame, and remove heat. They are collected and recycled
through a heat exchanger (where pebbles are cooled by inlet air) back into
the combustor.
18 Coanda Flame
Flameholder using coanda wall attachment effect. High velocity fuel-air
mixture discharges through ring nozzle onto surface of concial nosepiece and
entrains flow from surrounding environment. Method of setting up low
intensity flame.
28
-------
TABLE VI
LIST AND BRIEF DESCRIPTION OF COMBUSTOR CONCEPTS (Continued)
Concept No. Title and Description
19 Electric Assist Nozzles
Atomization of liquid fuel enhanced by an applied electric field, with
dispersion of charged droplets which are further guided, vaporized and
mixed with air under the influence of electric fields.
20 Virtual Staging
Burning zone expands in volume and elongates as combustor loading in-
creased (from idle to max. power). Flamefront grows into additional com-
bustion airflow needed at max. power, thereby providing automatic or
virtual staging.
21 Engine Inlet Fuel Injection
Vaporization and premixing of liquid fuel to very lean equivalence ratio for
reaction in a catalyst or flameless combustor. Achieved by introducing fuel
into engine inlet.
22 Flameless Combustion
Large volume burner operating at very lean equivalence ratios. Consumes
fuel by low-temperature long-residence-time flameless reactions.
23 Air Staging
Combustor airflow distribution controlled by variable geometry to maintain
desired equivalence ratios in burning zones and elsewhere over the range of
engine operating points.
24 Fuel Staging
Multiple fuel injection points provide variable fuel distribution and set up
successive zones of desired equivalence ratios.
25 Vorbix
Acronym Vortex burning and mixing. Swirling air jets cause high rate of
mixing in main burning zone. Pilot burner used for rapid vaporization of
main fuel and controlled autoignition of resultant mixture. Lean burning for
low NO,.
26 Fuel Air Premixing
Fuel injected into airstream prior to combustion zone to produce a uniform
fuel-air mixture and reduce spread in localized equivalence ratios before
burning begins.
Note: Concepts 27, 28, and 29 were conceived during the bench-scale evaluation program of Phase II.
29
-------
2.5 TASK 3 ANALYTICAL SCREENING
In this task, estimates of the degree of NO, control attainable with concepts identified in
the preceding task were made using analytical methods. The principal analytical tool employed
to accomplish this work was a streamtube combustor model formulated and developed
previously under Air Force sponsorship and then modified to accommodate the needs of this
program.
2.5.1 Introduction
To identify and rank the effectiveness of NO.-control techniques, a number of approaches
can be considered. One might be to conduct a comprehensive experimental combustor program
to evaluate many design concepts and proposed modifications intended to reduce pollutant
concentrations without incurring system performance degradation. This approach is costly and
time-consuming. Another approach might be to formulate a generalized analytical combustor
model that can realistically describe the coupled physical and chemical processes occurring
within a combustor and predict species concentrations and distributions throughout the
reaction chamber of interest. A model that can provide these predictions as a function of
combustor geometry, aerothermodynamics, and general operating conditions would be a vital
analytical tool permitting the combustion engineer to estimate the impact of these variables on
the production of pollutants.
A gas turbine engine combustor model that could accomplish the preceding with a high
degree of accuracy would indeed be Utopian. There are a few, such as the streamtube model
being used in this program, that are capable of serving as design tools insofar as providing
general guidance and direction (Reference 5). None currently exist, however, that can accom-
modate, except in a very cursory manner, effects of minor variations in designs. These minor
variations can render an ostensibly noneffective combustor configuration successful. Conse-
quently, this type of design finesse is still obtained through carefully planned experiments.
An informative review on the general subject of mathematical modeling of pollutant
formation was presented in Reference 6. Although the presentation on gas-turbine combustor
modeling was relatively modest, particularly in light of the widespread analytical work and
funding that have been devoted to the subject in recent years, the general topic of modeling
was aptly described. The philosophy that "the modeler seeks reasonable mathematical sim-
ulations of actual phenomena" was incorporated into the approach taken in accomplishing the
work of this task. Our objective was to predict trends, and to evaluate parametrically, within
the constraints of the available simulations of physical and chemical phenomena, the effects of
minor variations in combustor design concepts.
The principal analytical tool used to accomplish this task was a previously formulated
streamtube combustor model. The steamtube model is described in the following sections.
2.5.2 Description of the Streamtube Model
A computerized procedure that was developed under Air Force sponsorship for predicting
chemical species concentrations and distributions within and from gas turbine engine com-
bustors was the principal analytical tool used in this study. The prediction program in-
corporates several separate mathematical models (modules) to simulate the aerodynamic and
thermochemical processes characteristic of combustion in a gas turbine main burner. These
modules were uniquely combined to provide a realistic, yet practical, engineering tool for
predicting concentrations of pollutants from gas turbine combustors. The objectives of the
current program, however, required, in part, analyses that were beyond the scope of the
existing emissions prediction computer program. Consequently, the inclusion of several mod-
ifications was necessary to meet program objectives.
30
-------
2.5.2.1 The Original Model
The primary objective of the work under which the original model was formulated was to
develop an engineering tool to assist in the design and development of low-emission combustor
hardware. For this reason, it was appropriate that the analysis be directly related to details of
the combustor geometry, fuel injection system, and engine operating conditions. Development
of individual submodels and their combination in modular fashion allowed the requisite detail
to be incorporated in a tractable mathematical analysis. A necessary constraint on the degree
of complexity, however, was that the resulting analytical model be practical in terms of
computer time required for routine engineering use.
The modeling approach taken was to formulate mathematical treatments for the princi-
pal physical and chemical mechanisms considered to influence the combustion process, and to
integrate these mechanisms through a sequence of thermodynamic states obtained from the
coupling of these mechanisms with the physical combustor flow field. The simultaneous
solution of the combustion rate mechanisms and the flow field equations provided the gas
temperature, flow velocity, and chemical-species concentrations as a function of position
within the combustor which in turn influence subsequent combustion. The principal elements
of the analysis were a combustor internal flow field model, a physical combustion model, a
treatment of hydrocarbon-air chemical kinetics, and a NO, kinetics model. A further descrip-
tion of the original model may be found in References 7 through 11.
2.5.2.2 Model Modifications
The streamtube model had been formulated and developed to address conventional gas
turbine engine combustors. The original model was incapable of handling several items
necessary for the analysis of many of the combustion concepts identified in the program.
These items included: the ability to use fuels other than JP-5 (i.e., No. 2 fuel oil); fuel-bound
nitrogen capability; the removal or addition of heat from the combustion process; provisions
for the injection of secondary fuel; and the ability of the user to control the aerodynamics of
dilution airflow and the rate of raw fuel consumption for specific applications. Consequently,
modifications to the original model to accommodate these items were formulated and in-
corporated into the streamtube model. In order to Accomplish the program in a timely manner,
predictions of pollutant emissions from combustors incorporating a variety of control concepts
were also made using the revised model in varying stages of completion.
a. Expanded Fuel Capability
Modification to the original physical chemistry and hydrocarbon thermochemistry models
was made to treat composite fuels up to 20 constituents and No. 2 fuel oil. A complete list of
the fuels and constituents of the updated model is given in Table VII.
b. Fuel-Bound Nitrogen
The original streamtube model did not address the production of nitrogen oxides
resulting from the reaction of oxygen in the air with nitrogenous species chemically bound in
the fuel. A means was sought to add this capability without major revisions to the model.
Because of its capacity to treat a complex phenomenon in a basically simple manner, with good
results, the correlation of Fenimore (Reference 12) was adopted.
31
-------
TABLE VII
FUELS AND CONSTITUENTS IN THE
MODEL
Component ID Index
1
2
3
4
5
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
28
Component
JP-4
JP-5
No. 2 Fuel Oil
Iso Octane
Methanol
H2O
CO
H2
CO2
02
N2
AR
H
O
OH
NO
N
C
CH4
Natural gas, dry
NH3
It was assumed that oxides of nitrogen from fuel-bound nitrogen are formed as raw fuel
vapor is transformed either to the unburned hydrocarbon intermediary or to products of
combustion. In this manner the fuel-bound nitrogen is incrementally converted, along with the
unburned fuel vapor to which it is bound, and consumed by reaction with oxygen. The fraction
of the bound nitrogen converted to nitric oxide was determined by Fenimore's correlation.
= 1 exp
^\.
where
(__[NLJlINO]_
v
[N] = concentration of fuel-bound nitrogen, fully converted to NO, as
ppmv
[NO] = concentration of NO, in ppmv, actually formed from the fuel-bound
nitrogen
X = parameter characteristic of the flame and independent of the
fuel-bound nitrogen concentration
The correlation was evaluated at the local level of hydroxyl concentration. Fenimore had
found that the parameter X, when made proportional to the equilibrium hydroxyl concentra-
tion in the flame,
X = 2 - 3 [OH]
best explained the experimental results. The fuel-bound nitrogen not converted to NO is
considered to be unavailable for further oxidation and is assumed to be transformed into
diatomic nitrogen.
32
-------
c. Heat Removal and Heat Addition
The original streamtube model treated aerodynamic and combustion phenomena as
occurring within an adiabatic combustion chamber. Heat extraction and heat addition were
not considered. In this program the capability for predicting emission concentrations under
nonadiabatic conditions was needed in the evaluation of several design concepts.
To provide this capability the equilibrium chemistry calculations performed in the deck
by the Brinkley routines were altered to consider heat addition and heat removal in computing
the total chemical system enthalpy. Total system energy is the basis to which the iteration of
equilibrium concentrations of the combustion products and the mixture temperature is made
to converge.
For configurations in which heat was to be removed in distributed fashion over a given
segment of burner length, it was necessary that the equilibrium-product concentrations and
temperature reflect the decline with length in system enthalpy. In the original model
equilibrium-product concentration and temperature data were computed and stored for later
retrieval prior to initiating the lengthwise integration procedure. The capability for continuous
heat removal required that provision be made for halting the integration procedure in order to
update the stored data tables. This feature was incorporated. Under heat removal or heat
addition conditions, the system enthalpy was adjusted incrementally and new data tables were
generated. The rate at which the enthalpy was permitted to change was assumed to be
constant over the length of the system wherein heat was being transferred. A fixed interval' of
0.1 in. was chosen and found to cause only a slight distortion in meeting conservation criteria.
A consequence of the heat transfer option was a notable increase in computer running time.
Increases by a factor of two to four were common, depending upon the combustor length over
which heat was transferred.
d. Secondary Fuel Injection
The original streamtube model allowed liquid fuel to be injected into the primary zone as
droplets, or as a partially-to-fully vaporized component premixed with air. Combustor con-
figurations characterized by a second fuel injection station elsewhere in the burner was not
accommodated.
A capability was incorporated for injecting liquid or gaseous fuel at a second location
downstream of the primary injection station. Distribution of the secondary fuel into the
streamtubes was established as a user's option. Vaporization and combustion of the secondary
fuel then proceeded according to the precepts of the physical chemistry models discussed
earlier.
e. Additional Phenomenological Modifications
To accommodate the needs of the program, several other submodels in the streamtube
combustor procedure were modified and partially rearranged. These revisions were made to
provide the increased flexibility needed in simulating the aerodynamic and combustion
processes called for in certain design concepts.
In the original model, the allocation of penetration air into the streamtubes was assumed
to occur at a programmed rate determined by the user. The rate selected is done so with
particular consideration being given to the combinatorial effects of the turbulent exchange of
cooling air. To increase the flexibility of the computer program, means were incorporated to
allow an arbitrary distribution of the dilution air into the streamtubes to be specified as an
input item by the user. In addition, the turbulent-exchange model was dismantled and revised
33
-------
to permit specification of the rate at which cooling streamtube air is mixed into the adjacent
streamtube as an input item by the user.
Finally, provision was made for adjusting the rates at which raw fuel vapor was consumed
in the turbulent flame system. A multiplicative factor was applied to the expression for
turbulent flame speed that can be altered as an option by the user.
2.5.3 Chronological Development
As stated, the analytical screening effort involving the application of the streamtube
model to generate emissions predictions was structured as two distinct tasks: first, streamtube
model revisions to provide expanded capability; and second, the emissions predictions them-
selves. Several of the concepts considered in this program called for extensive revisions to the
existing models for their treatment. Due to the requirement that preliminary predictions be
completed and the results made available to the selection process early in Task 3, only those
model revisions which could be incorporated expeditiously were implemented prior to the
initial predictions. Upon the completion of these predictions, the more extensive model
modifications were incorporated, followed by further updated predictions. The nature of the
model refinements proved fairly extensive and consequently the modeling tasks occupied a
significant portion of the analytical screening effort.
t,r. The preliminary concepts proposed early in the program for reduced NO, emissions
included several features which could not be treated analytically with the original streamtube
model. These included heat removal, fuel staging, catalytic combustion, variations in primary
zone stabilization systems, and dilution air injection and distribution systems requiring a more
flexible treatment than that incorporated in the original models. In addition, the need to
assess NO, production in the combustion of low Btu gaseous fuels required an extensive
revision of the thermochemistry model, originally limited to JP-type fuels. Still further, a
capability to predict the increased contribution to burner exit plane NO, from the combustion
of liquid fuels containing bound nitrogen was required.
The model revisions cited represented varying degrees of effort with regard to implemen-
tation. With this in mind, the chronological order of the analytical screening effort proceeded
as follows:
1. The early model revisions: extension of the streamtube model to provide
secondary fuel staging capability; dismantling of the dilution injection
model and the turbulent exchange model; and provision for reducing total
system enthalpy as a crude method of simulating discrete heat removal.
2. The "once over lightly" simulation of each concept and pertinent para-
metric evaluations were then completed in order to lay down initial
guidelines for the concept selection task. A preliminary one-step approx-
imation of the effect of fuel-bound nitrogen was also completed.
3. The model revision work was resumed and final predictions were obtained
for the heat removal concept and the fuel-bound nitrogen parametric
study.
4. The streamtube model was expanded to treat a variety of liquid fuels,
including No. 2 fuel oil and various composite fuels.
34
-------
2.5.4. Concept Simulation
The streamtube combustor model was used to estimate the exhaust emission character-
istics of the first twelve candidate gas turbine engine burner concepts listed in Table VI.
Individual descriptions of the concepts were presented in this table. A further description of
the concepts may be found in Volume II of this report series.
The candidate design concepts were represented by various configurations of a baseline
five-inch diameter, can-type combustor of arbitrary length, and were examined at the same
operating conditions using, whenever possible, common assumptions. The operating conditions
are listed in Table VIII. Common assumptions included the use of JP-5 to simulate No. 2 fuel
oil, and the means of treating injected fuel and air. The use of JP-5 fuel as a substitute for No.
2 fuel oil was validated later in the modeling effort. Liquid fuel was assumed to be injected
through a fuel-air premixing tube in which a fraction of the fuel is vaporized and mixed with
air. The fraction was varied over the range from 0.25 to unity. In each case the mixture
equivalence ratio assumed for recirculation zone species and temperature was the primary-zone
equivalence ratio. In addition, dilution air was assumed to be uniformly apportioned among
the stream tubes; for those situations in which cooling air was provided, only 5% of the dilution
air was injected into the cooling streamtube. For configurations in which secondary fuel is
injected, the fuel was also assumed to be uniformly apportioned among the streamtubes;
however, the degree of vaporization of the secondary fuel depended upon the concept being
simulated.
TABLE VIII
BASELINE OPERATING CONDITIONS
Total Air Flowrate 3 Ibm/sec
Inlet Temperature 800°F
Inlet Pressure 212 psia
Burner Pressure Drop (AP/Pt) 3%
Total Fuel Flowrate (JP-5) 0.063 Ibm/sec
Overall Fuel-Air Ratio 0.021
Fuel Nozzle Pressure Drop 125 psi
2.5.5 Conclusions
Twelve configurations, (of the 26 identified) embodying a variety of approaches for
reducing the production of nitrogen oxides in stationary gas turbine engine combustors were
examined using the streamtube combustor model. The use of liquid JP-5 fuel, which was
considered to be representative of No. 2 fuel oil, was assumed in the analytical studies, and
later was shown to give nearly identical results as No. 2 fuel oil.
While the computer studies were considered to be a valuable tool, the results of the
studies were not taken as conclusive evidence for any particular concept. The results, however,
were used as an indication of expected trends, tradeoffs, and areas of a particular concept that
might have required special attention. The results were used for general guidance in the
bench-scale experimentation conducted during Phase II.
General conclusions derived from the cases examined in the analytical study can be
summarized as follows:
1. To reduce the production of NO, in the primary zone, rich burning, in the
equivalence ratio range between 1.2 and 1.4 is preferred over superlean
burning. Although the combustion process in the primary zone is in-
complete as a result, the reaction can be directed to completion by
carefully injecting and mixing dilution air.
35
-------
2. Elevating the temperature of the primary-zone air supply to achieve more
complete fuel vaporization is counterproductive under fuel-lean operating
conditions, generally enhancing the production of NO,.
3. The production of NO, can be effectively curtailed by removing heat in the
primary zone.
4. When rich burning is incorporated in the primary zone, dilution air must
be rapidly injected and mixed to curtail further production of NO, while
also enhancing oxidation of carbon monoxide and unburned hydrocarbons.
Main-stream temperature levels in the range from 2000 to 2800° F, for
example, are most desirable.
5. Reducing NO, production is enhanced by prevaporizing liquid fuel and
mixing the fuel with air prior to entering the combustion chamber.
6. To curtail production of NO., recirculation-zone length should be min-
imized; a perforated-plate flameholder is a potential means for ac-
complishing this.
7. Significant reductions in NO, production can be achieved by reducing or
eliminating injection of cooling air into the main combustion stream; this is
particularly true for the injection of air into the primary zone during fuel
rich operation.
8. A promising means for reducing production of NO, is to incorporate fuel
staging in which the secondary fuel is uniformly distributed throughout the
main stream.
In Task 4 of Phase I, the concepts were divided into three categories: initial primary
concepts, secondary concepts, and concepts which should be dropped from experimental
screening for various reasons. Table IX shows a list of the 26 concepts and the classification
assigned to each based on the activities of Phase I.
The five initial primary concepts selected were judged to have the greatest apparent
potential for NO, control based on the information on hand at the end of Phase I. These
concepts were selected for definitive testing in the first cycle of screening experiments during
Phase II. The secondary concepts were from those remaining candidates that had apparent
potential but required further substantiation. The concepts which were eliminated from
further consideration were ones which were judged to be either outside the scope of the
investigation or were thought to be extremely difficult to implement as a result of practical
constraints.
36
-------
TABLE IX
CANDIDATE CONCEPT CLASSIFICATION
Initial Classification
Concept No. Title Primary Secondary Dropped
I Low-Intensity Flame
2 Premizing Catalytic Burner
3 Superlean With Heat Recirculation
4 Superlean With Preburner
5 Heat Removal
6 Quench Reheat
7 Staged Centertube Burner
8 Exhaust Gas Recirculation
9 Hydrogen Enrichment
10 Surface Combustion
11 Distributed Flame
12 Ceramic Liner
13 External Combustion
14 Boost-Air Dilution
15 Artificial Excitation
16 Extended Injector
17 Pebble Bed
18 Coanda Flame
19 Electric Assist Nozzles
20 Virtual Staging
21 Engine Inlet Fuel Injection
22 Flameless Combustion
23 Air Staging
24 Fuel Staging
25 Vorbix
26 Fuel Air Premixing
37
-------
REFERENCES
1. Lachapelle, D.G., J.S. Bowen, and R.D. Stern, "Overview of Environmental Protection
Agency's NO, Control Technology for Stationary Combustion Sources," Presented at 67th
Annual Meeting of the AIChE, December 4, 1974.
2. Conkle, J. P., W. W. Lackey, and R. L. Miller, "Hydrocarbon Constituents of T-56
Combustor Exhaust," Bioenvironmental Analysis Branch, Environmental Sciences
Division, USAF School of Aerospace Medicine, AFSC, Brooks AFB, Texas, Progress
Report No. SAM-TR-75-8, April 1975.
3. AGARD Conference Proceedings, Number 125, "Atmospheric Polution by Aircraft En-
gines" AGARD-CP-125, September 1973.
4. "The Combustion of Shale Derived Marine Diesel Fuel at Marine Gas Turbine Engine
Conditions," by M.C. Hardin, presented at the AIAA Symposium on Alternate Fuel
Resources, 25-27 March 1976 at Santa Maria, CA.
5. Mellor, A.M., "Gas Turbine Engine Pollution," Prog. Energy Combust. Sci. 1, 1976.
6. Caretto, L.S.,"Mathematical Modeling of Pollutant Formation," Prog. Energy Combust.
Sci 1, 1976.
7. Mosier, S.A. and R. Roberts,"Low-Power Turbopropulsion Combustor Exhaust
Emissions, Volume I: Theoretical Formulation and Design Assessment,"
AFAPL-TR-73-36, June 1973. Sensitivity of emissions to operating variables.
8. Mosier, S.A. and R. Roberts,"Low-Power Turbopropulsion Combustor Exhaust
Emissions, Volume II: Demonstration and Total Emission Analysis and Prediction,"
AFAPL-TR-73-36, June 1973.
9. Mosier, S.A. and R. Roberts, "Low-Power Turbopropulsion Combustor Exhaust
Emissions, Volume III: Analysis," AFAPL-TR-73-36, June 1973.
10. Mador, R.J., "User's Manual General Emissions Prediction Computer Program,"
P&WA Report 4929, February 1974.
11. Mador, R.J. and R. Roberts, "A Pollutant Emissions Prediction Model for Gas Turbine
Combustors," AIAA Paper No. 74-1113, 10th Propulsion Specialists Conference, San
Diego, California, October 1974.
12. Fenimore, C.P., "Formation of Nitric Oxide from Fuel Nitrogen in Ethylene Flames,"
Combustion and Flame 19, 1972.
38
-------
APPENDIX
CONVERSION OF ENGLISH UNITS TO SI UNITS
English Unit
British Thermal Unit (Btu)
Foot (ft)
Pounds Force (lb,)
Pounds Mass (lbm)
Pounds per Square Inch (psi)
Temperature (°F)
SI
Unit
joule
meter
Newton
Kilogram
Kg/cm2
°C
Multiply
English
Unit By
1055.87
0.3048
4.448
0.4536
0.0703
(5/9) (F-32)
39
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-80-017a
2.
4. TITLE AND SUBTITLE Advanced Combustion Systems for
Stationary Gas Turbine Engines: Volume 1. Review
and Preliminary Evaluation
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
January 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
S.A. Mosier and R. M. Pierce
FR-11405
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Pratt and Whitney Aircraft Group
United Technologies Corporation
P.O. Box 2691
West Palm Beach, Florida 33402
10. PROGRAM ELEMENT NO.
INE829
11. CONTRACT/GRANT NO.
68-02-2136
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
3. TYPE OF REPORT AND PER
Final; 12/75 - 9/76
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES JERL-RTP project officer is W.S
2432.
Lanier, Mail Drop 65, 919/541-
16. ABSTRACT
The reports describe an exploratory development program to identify, eval
uate, and demonstrate dry techniques for significantly reducing NOx from thermal
and fuel-bound sources in stationary gas turbine engines. Volume 1 covers Phase I
of the four-phase effort. In Phase I, duty cycles were analyzed to identify current
and projected dominant operating modes and requirements of stationary gas turbine
engines. These analyses indicate that as compression ratios and turbine inlet tem-
peratures are increased to improve thermal efficiency, uncontrolled NOx emissions
can be expected to double in 10 years and triple in 20 years. An extensive survey
was made of candidate combustor concepts, and an analytical study was made from
which those concepts considered to have significant potential for reducing production
of NOx were identified. An initial compilation of 26 combustor design concepts was
assembled, indicating potential for controlling NOx from clean fuels and/or fuels
containing significant amounts of bound nitrogen. Computer simulations of these com-
bustor concepts aided in prioritizing the designs prior to experimental screening in
a bench-scale combustor test rig. The experiments were carried out under Phase n
and are described in Volume 2.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Gas Turbine Engines
Stationary Engines
Nitrogen Oxides
Combustion
Combustion Chambers
Mathematical Models
Pollution Control
Stationary Sources
Combustor Design
Dry Controls
Duty Cycles
13B
2 IE
2 IK
07B
2 IB
12A
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
45
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
40
------- |