M

THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS

345 E. 47 St., New York, N.Y. 10017 82-GT-255
The Society shall not be responsible for statements or opinions advanced in papers or In
li discussion at meetings of the Society or of its Divisions or Sections, or printed in its
publications. Discussion is printed only if the paper is published to an ASME Journal.
^I Released for general publication upon presentation. Full credit should be given to ASME,
the Technical Division, and the author(s). Papers are available from ASME for nine months
after the meeting.
Printed in USA.
Copyright © 1982 by ASME

R. A. Battista
(Formerly) Engineer—Combustion
Low Heating Value Fuel Burning
Capabilities of General Electric
Development,
Gas Turbine Division,
General Electric Company,
Schenectady, NY
Industrial Gas Turbines
R. P. Pandalai
Engineer, The advent of higher fuel prices and operational costs associated with energy
Aircraft Engine Business Group, production have had a profound impact on the application of the gas turbine as a
General Electric Company, prime mover. This paper describes an extensive program carried out by the General
Evendale, OH Electric Company to demonstrate the ability of both the heavy-duty and aircraft-
derivative gas turbines to operate satisfactorily while burning a wide range of lower
M. B. Hilt heating value fuels, typical of industrial process or gas conditioning plants.
Analytical predictions of flammability limit changes due to fuel composition
Manager—Combustion Aerothermal Design, variations, and combustion inlet air temperature effects are compared with small
Gas Turbine Division,
scale atmospheric burner tests. Finally, full scale single burner and sector test
General Electric Company,
Schenectady, NY results are presented which demonstrate the capability of present generation
combustion systems to operate on lower heating value fuels.

N()MFNr:T.ATTTRF. I. Introduction

The advent of higher fuel prices and

Symbol Definition operational costs associated with energy
production have had a profound impact on the
Al Standard LM2500 Gas Fuel Nozzle application of the gas turbine as a prime
A3 LM2500 Low BTU Fuel Nozzle mover. There is an increased interest in the
A4 LM2500 Low BTU Fuel Nozzle use of alternate gaseous fuels for application
f Fuel - Air Ratio in both the Heavy Duty and the LM2500 aircraft
LHV Lower Heating Value, BTU/SCF derivative industrial gas turbines. This paper
P3 Compressor Discharge Total Pressure, describes an extensive program carried out by
PSIA the General Electric Company to demonstrate the
PF Pattern Factor ability of these turbines to operate
R Specific Gas Constant satisfactorily while burning a wide range of
T3 Compressor Discharge Total low heating value fuels.
Temperature or Combustor Inlet
Temperature These fuels, typical of industrial process
T 95 Combustor Exit Average Temperature or gas conditioning plants, are generally high
T Turbine Inlet Total Temperature in inerts such as carbon dioxide or nitrogen
T F Combustor Exit Maximum Temperature and low in combustibles such as methane,
^a x Combustor Air Flow Rate, lb/sec ethane, and propane. Because of process
AP Combustor Pressure Drop, PSIA differences, it is necessary to design a
Equivalence Ratio combustor that will tolerate heating value
variations due to combustible to inert
CONVERSION FACTORS composition ratio changes and still provide
acceptable operation over an entire load range
BTU/SF at 60 ° F and 14.73 PSIA Dry x 9.377 = of the turbine.
Cal/m at 0 ° C and 760 mm HG Dry
° °
t C = (t F - 32)/1.8 Initially analytical predictions of
flammability limit changes due to fuel
composition variation and combustion inlet air
temperature effects are compared with small
scale atmospheric burner tests. Finally, full
scale single burner and combustor sector test
results are presented which demonstrate the
capability of present production combustion
systems to operate on low heating value fuels.

The analytical predictions and small scale

burner tests provide technical information
which is common to both types of turbines and
Contributed by the Gas Turbine Division of the ASME. will be presented first. The full scale test
 results are specific to the Heavy Duty and dependence of NOx formation, carbon
LM2500 turbines and will be presented monoxide would produce over 3 times as
individually. much NOx as natural gas in a given
combustor design.
Part 1 will include a description of the
Heavy Duty combustion system and test results The flame temperature of the gas is
and part 2 will present similar information for an indication of the ability of the
the LM2500 combustor. combustion reactions to go to completion
and thus an indication of expected overall
II. Low BTU Fuel Combustion Considerations combustion efficiency for a given
combustion system.
The principal characteristics of gaseous
fuels which concern the combustion system The overall flame temperature has
designer are (1) lower heating value, (2) flame some affect upon liner metal temperatures,
temperature, (3) flammability limits, and (4) although for gaseous fuels the flame will
reaction rates. generally be non-luminous and thus have
considerably lower emissivities than
A. Lower Heating Value liquid fuels with luminous flames.

Low heating value impacts the C. Flammability Limits

combustion system in several ways. First,
heating value determines the sizes of the For stable combustion to be sustained
gas supply manifold and pipes to each in a gas turbine combustor, a
individual fuel nozzle. The gas nozzle recirculating region is maintained in the
injection holes must be sized to head end of the combustor. This is
accommodate the fuel flow while achieved through the use of a swirler on
maintaining acceptable pressure drops at the fuel nozzle coupled with careful
full flow conditions as well as minimum design of the the combustor cap and liner
flows. Too high a pressure ratio can lead mixing air hole pattern. Inside the
to hot spots on the combustion liner wall recirculation zone, a gradient of fuel-air
as well as requiring excessive supply ratios exists. As gas turbine load and
pressures upstream of the fuel control hence fuel flow change, the range of
valve. On the other hand if the pressure fuel-air ratio also varies. The wider the
ratio across the fuel nozzle is too low, flammability limits of a given fuel, the
combustion instabilities in the form of more flexible it will be for off-design
blowout or combustion driven pressure operation such as startup and low load
oscillations may result. operation.

Second, the heating value of the fuel The Gas Turbine Division has
impacts the mixing requirements in the determined empirically that a ratio of
combustor. As the heating value upper to lower flammability limit of 2.2
decreases, the proportion of fuel to air to 1 (on a volume basis) or greater is
increases, requiring that large quantities sufficient for operation of the gas
of fuel must be mixed with air to provide turbine from start-up to full load. This,
acceptable combustion. of course, implies satisfactory operation
when using a standard combustion system.
Third, putting large quantities of Special combustion systems have been
fuel in the combustor reaction zone tends designed to extend the range of operation
to decrease the pressure drop across the on low BTU fuels. In any case, a careful
combustion liner at a given design look at the flammability limits will allow
condition. Thus the combustor designer the combustion system designer to evaluate
must compromise between very high pressure a given special fuel and make comparisons
drops when operating on conventional fuels with known combustion system performance.
and low pressure drops when running on low
BTU fuels. This situation is generally D. Reaction Rate
not a severe problem.
In the combustion of any hydrocarbon
B. Flame Temperature fuel, a number of intermediate reactions
take place during the burning of the fuel.
The adiabatic stoichiometric flame The oxidation of carbon monoxide is one of
temperature of the gas fuel is an the slowest reactions to be completed.
important parameter when considering since the combustion process inside a gas
combustion effects. Although this turbine combustor must take place rapidly,
temperature is not reached in an actual any decrease in reaction rate will lead to
combustor, where heat loss occurs, it decreased combustion efficiency, i.e.
serves as a useful basis of comparison of excessive CO or unburned hydrocarbons
fuels. Flame temperature is not (UHC) in the exhaust. Therefore the
necessarily related to the heating value combustion system designer must evaluate a
of the fuel. For example carbon monoxide given fuel to insure that the combustor is
has about 30% of the heating value of designed to keep the reaction temperature
natural gas; however, the adiabatic as high as possible. Also he must allow
stoichiometric flame temperature is 268 0 F adequate residence time in the hot
higher. Because of the strong temperature products so that the reactions will go to

K
 completion. The first of these criteria LHV
W.I. = _ (BTU/SCF) ( o R) -
is determined primarily by the Sp.gr TFUEL
stoichiometry of the reaction zone and the
latter by overall physical size of the The fuel nozzle area factor is the ratio of
combustion system. WOBBE indices for a natural gas (1025 BTU/SCF)
to the gas in question. This gives an
III. Analytical Techniques indication of the size of gas metering holes
needed for a given fuel, compared to that used
Several gaseous fuel mixtures were for natural gas burning machines.
considered during the course of this program,
ranging in lower heating valve from 211 to 955 INLET AIR

BTU/SCF. The composition of these fuels are COMPONENT MOL. WT VOL% WT. %

listed in Table 1. For the particular machine AR 39.95 0.97 1.34

application required, the 425 BTU/SCF gas was N2 28.02 77.25 74.99

CO2 44.01 0. 0.
selected as the prime fuel although other tests
H 2O 18.02 1.00 0.63
were conducted to demonstrate flexibility. 02 32.00 20.78 23.04

MAIR 28.86 LBM/LB-MOLE

FUEL

TABLE 1 FUEL GAS ANALYSIS NAME MOL. WT VOL. PCT. WT.PCT. LHV

Volume CO 2 44.01 87.31 90.72 0.

Percent I II III IV V VI VII CH 4 16.04 1.47 0.56 13.40

.47 65.00 C 2H6 30.07 8.95 6.35 145.19

Nitrogen - - - - -
C 3 11 8 44.09 2.27 2.36 52.71
Carbon Dioxide 77.00 12.64 75.69 85.9 87.30 78.16
Methane 4.88 74.13 2.45 1.85 1.47 2.55 35.00
Ethane 10.69 6.49 2.76 5.75 8.95 8.20
BTU/MOLE BTU/SCF BTU/LBM KCAL/NM3
Propane 6.40 3.42 12.09 4.85 2.27 9.45 LHV 80020. 211.31 1889.36 1879.96

.37 .49 .39 .59 HHV 87374. 230.73 2062.99 2052.72

I-Butane - -
N-Butane .59 1.17 .52 .65 - .94
0. WT.FRAC.LIQUID 0. MWT. OF LIO. FUEL
I-Pentane .03 .27 .04 - - .05
18550. LHV LIO. (BTU/LB M) 211.3 LHV GAS (BTU/SC F)
N-Pentane .04 .47 .05 1.00 - .06 80.0 TEMP.LIO.FUEL (F) 0. WT.RATIO(HVLIO/HVTOTAL)
0.2075 SP.HEAT.GAS FUEL 1.2919 CP/CV-GAS
Hexane Plus - .37 .01 - - .004 -4.0 TEMP.GAS.FUEL (F) 42.35 MWT.GAs
H 2 S (PPM vol) 36 8 56 - 45 30 1.4675 SP.GRAY.GAS FUEL 6.7944 F-NOZ.AREA FAC.-GAS
8.17 WOBBE INDEX 0.456 MOLES OF 0 2

LHV Btu/Scf 397 955 475 279 211 425 319 0.66858 TOTAL FUEL/AIR 2.195 MOLES OF AIR
0.45559 F-A BY VOL 0.31299 FUEL FRACTION
f STOICH .346 .079 .296 .518 .688 .333 245 FLAMMABILITY LIMITS IN DRY AIR

T STOICH I'I 3331 3835 3432 3100 2919 3388 3484

VOLUME % F/A BY WT.
T FUEL 70 98 80 70 400 250 70 TEMP UPPER LOWER U/L UPPER LOWER U/L

0 Lean 121 .59 .532 .577 .632 .655 .588 .592 5. 3205
. 23.965 1.34 0.69 0461 1.50
105. 31.97 23.033 1.39 0.69 0.438 1.57
,b up/0 lean 131 2.46 3.23 2.73 1.87 1.57 2.54 2.20
305. 31.87 21.170 1.51 0.68 0.393 1.74
605. 31.82 19.374 1.73 0.68 0.329 2.07
(1) At 600°F air temp. 1005. 31.87 14.647 2.18 0.68 0.251 2.73

(2) Equivalence ratio at lean flammability limit, 1 ATM, 100°F mixture temp DILUENT ASSUMED TO BE CO 2 ICO 2 +2/3H 2 O) = 0.87309
DILUENT ASSUMED TO BE N 2 IN 2 + H 2 0/3) - 0.
(3) Ratio of upper to lower flammability limits, 1 ATM, 10o°F
INERT/FUEL = 6.879

Equilibrium flame temperature and STOICH. CONDITIONS AT REF. TEMPS,

PREF TIN DELT TFLAME FUEL/AIR MOLWT H.LIO.FUEL
flammability limits for the various gases were 150.0 600. 2211. 2811. 0.6687 3238 0.
calculated as a first step in determining their SP.HT . CP/CV EOU.RAT ENTHALPY

suitability as gas turbine fuels. The flame 0.3565 1.2117 1.0000 -774.99

temperatures as well as equilibrium combustion PRODUCTS MP - 32.33

products were calculated using a computer AR CO CO2 H2O NO N2 OH

program which contains the NASA Equilibrium 0.6515 01158 34.6485 12.5647 0.0188 51.9225 0.0194
0 2 H2
Program (Ref. 1) as a subroutine. Also 0.0512 0.0109
included in the program is a subroutine which
calculates flammability limits using Bureau of
FIG. 1 COMPUTER ANALYSIS OF GASEOUS FUELS
Mines data [(Bulletins 503 and 627) (Ref. 2 &
3)]. The program approximates the variations In addition to listing the fuel
in flammability limits with addition of N 2 , compositions, Table 1 summarizes the
CO2 , and H ? 0. Also included in the calculation calculations made for the various gases. As
is the effec
t reactant temperature (fuel + Table 1 indicates, increasing the heating value
air mixture) on flammability limits. These serves to improve matters by increasing the
relationships are also taken from the Bureau of flame temperature as well as widening the
Mines reports. flammability limits. Here, the flammability
limits are given in terms of equivalence ratio,
A typical printout is shown in Figure 1. 0 defined as
The output is, for the most part,
self-explanatory. The WOBBE Index (W.I.) is actual fuel-air ratio
defined hereby 0 -
stoichiometric fuel-air ratio

3
 Thus, values of 4' < 1 represent fuel lean IV. Atmospheric Bench Scale Tests
conditions. In terms of equivalence ratio, the
lean limit is relatively insensitive to heating The investigation of lean blow-out (B/0)
value as shown in Figure 2. However, the limits was continued using an atmospheric test
fuel-air ratio at the lean limit varies rig. This rig is a turbulent flow axisymetric
considerably. The rich limits decrease diffusion flame burner consisting of a fuel
considerably as LHV decreases, until for LHV injector and air supply in a well insulated
less than about 200 BTU/SCF, the calculations cylinder. Fuel is injected in both radial and
indicate the low BTU gas mixtures would be axial jets and stabilized at the fuel nozzle.
non-flammable ("knee" of curve). The 211 BTU The air and fuel were electrically preheated to
fuel is therefore very close to the minimum the various desired inlet conditions. The
heating value. Increasing the fuel-air mixture products of combustion leaving the flame zone
temperature, Figure 3, does serve to widen the are contained in an insulated chimney. The rig
flammability limits but appears to have little was designed to allow for gas samples
effect on the "knee".
extraction at various axial locations along the
chimney centerline. Using the laboratory on
2.0 1 ATM; 80°F line emission analysis it was possible to
determine concentrations of CO, CO , 02, and
N,/CH y
unburned hydrocarbons in addition 1?o meas ured
CO 2 ,CH 4 ,C 2 H 6 , C3H8
F gas temperatures. A schematic of the burner is
shown in Fig. 4. Lean B/0 limits were
determined for the 211 BTU/SCF fuel gas at
pQ \^ COZ/CHQ ° °
100 F and at 500 F and a 211 BTU/SCF fuel
composed of CO 2 and C 3 H $ onl'. The air
temperature was varied from 450 F to 950 F. °
The results are shown in Figure 5. As both the
z I 1.0 \
m l calculations and experimental results indicate,
U , the CO 2 /C 3 H 8 mixture would present a slightly
>o , more severe test of a combustor than the high
w^ heating value gases since the lean B/0 limit is
greater than the comparable 211 BTU fuel. The
65% N 2 35%CH 4 effect of fuel temperature on B/O equivalence
MS3002 S/C
ratios may be somewhat higher than expected.

GAS SAMPLING PROBE

0L I I I I I TOP
1000 800 600 400 200 0
LHV — BTU/SCF

NSULATION
FIG. 2 CALCULATED FLAMMABILITY LIMITS OF VARIOUS GAS MIXTURES

2.0

I
U
m0
U)
ca
m

1.0

NSULATION
0
1000 800 600 400 200 0
LHV — BTU/SCF

FIG. 3 EFFECT OF MIXTURE TEMPERATURE ON FLAMMABILITY LIMITS OF AIF
VARIOUS GASES

FUEL

4 FIG. 4 ATMOSPHERIC BURNER

 In a well mixed system, one would not measure inlet and discharge temperatures,
expect the fuel temperature to have much effect pressure drops, metal temperatures, emissions,
1.0r and dynamic pressures.

OPEN SYMBOLS — B/O

0.8

211 BTU 500F

211 BTU 100F

x 0.6 CALC., CO 2/C 3 H 8 100F
cI 0
a
0.4
s
211 BTU
100F
0
211 BTU
0.2
500E
_CO2/C3H9
211 BTU, 100F

The LM2500 engine uses an annular

0.0 - ' 1
200 400 600 800 1000 1200 combustor with thirty (3) fuel nozzle cups,
AIR TEMP — F therefore for the test program conducted in the
Gas Turbine Development Laboratory, a three cup
FIG. 5 ATMOSPHERIC BURNER TEST RESULTS LEAN BLOWOUT sector test rig was installed in the
Laboratory. Figure 7 shows the installation of
the rig.
near the blow-out limit since the equivalence
ratio, 0, is considerably less than one. Thus
the air temperature will have more pronounced
effect on the mixture temperature than the fuel
and should influence the flammability limits to
a greater extent.

However, in a diffusion flame, where the

reactions occur at near stoichiometric
conditions, one would expect the fuel
temperature to have a greater effect on
flammability limits, particularly in the case
of low heating value fuels where equivalence
ratios in the reaction zone of the combustor
approach 1.0 (or greater for some fuels). Even
though it was recognized that uncertainties
existed in the data, the results were
encouraging in that a stable flame was achieved
at equivalence ratios below those calculated
from atmospheric flammability data.

V. Laboratory Facilities FIGURE 7 LM2500 THREE CUP SECTOR TEST RIG INSTALLATION (GTD,
SCHENECTADY)

The General Electric Gas Turbine Division

Development Laboratory is a multimillion dollar The laboratory has various fuel supplies
facility dedicated to the advancement of Heavy including propane, methane, distillate and
Duty gas turbines. The laboratory has a residual fuels. In addition it is possible to
process air supply capable of delivering up to blend combustible gases with nitrogen to
50 lbs. of air per second. For heavy duty simulate low and medium BTU coal derived fuels
combustor development, a single burner test rig from either an air blown or an oxygen blown
is used which can be likened to taking a "pie gasifier. To perform the combustion testing
shaped" cut from the gas turbine combustion utilizing a blended gaseous fuel consisting of
system and putting a pressure vessel around it. Co , CH, C 2 H 6 and C H 8 required facilities
The test rig is configured aerodynamically to moDification. Carbon dio xide was trucked to
simulate the flow passages within the gas the laboratory in liquid form. The liquid CO 2
turbine using contoured baffle plates. Figure was then piped into the blended gas system,
6 shows a typical single combustor test stand vaporized using the existing N vaporizor and
located in the Gas Turbine Development delivered to the system at 275 psig. A mixture
laboratory. Instrumentation is placed of methane and ethane was brought to the
strategically in and around the test rig to laboratory in tube trailers in gaseous form.

5
 These gases were then mixed with gaseous The compressor discharge air flows
propane using the existing low BTU gas blending around and cools the transition piece, and
system. then enters the combustion liner through
various diluting, cooling, and reaction
VI. Part 1. Heavy Duty Test Program air passages. Fuel injected into the
combustion reaction zone (roughly one and
Description of Combustion Systems one half diameters in length) reacts with
a portion of the available air entering
The General Electric Company produces a through the combustion holes.
variety of gas turbine models to satisfy Recirculating flow patterns of air and
customer requirements. The basic gas burning gases maintain flame stability.
turbine used for industrial and utility The hot combustion products are then mixed
power generation is of the simple cycle with secondary air entering through
design, i.e. air is compressed, fuel downstream rows of dilution holes to lower
burned, and the products of combustion are the average temperature and establish the
expanded through a turbine. To improve temperature profile of the gas before it
upon overall cycle efficiency, enters the turbine.
regenerative cycle machines route the
compressor discharge air through an Heat is transferred from the flame
exhaust gas-air heat exchanger called a and hot gases to the liner walls via
regenerator to transfer waste heat energy radiation and convection. The liner wall
to the air prior to combustion. The is cooled by external convection and by
combined cycle power plant uses the waste the thin film of air which is continually
exhaust heat to make steam in a boiler. passed through the cooling louvers (or
This steam can then be used in a process slots) to bathe its inner surface.
or to drive a turbine. Although there are
many possible combinations and The reverse flow design permits
configurations for the gas turbine combustor sizes large enough to optimize
combustion system, the description that combustion rates commensurate with heavy
follows will serve to define the typical fuels. This capability provides the
Heavy Duty combustor and its operational necessary flexibility to design combustors
envelope. to handle a wide range of low heating
value fuels as well as providing ample
The General Electric Heavy Duty Gas opportunity for emission abatement
Turbine utilizes a multi-chamber, reverse designs.
flow combustion system to provide the
maximum capability for fuels flexibility The combustion system must perform
while maintaining the shortest possible satisfactorily over a wide range of fuel
turbine shaft length. The gas turbine has and air flows during its normal operation.
six to fourteen combustion chambers, It must operate stably from the very low
depending upon the output and size of the air flow rates at ignition conditions,
machine series. Generally, two of the through acceleration to full speed, and
chambers have retractable spark plugs and then across the load range. For example,
two of the chambers have flame detectors fuel flow ranges from 30 to 1 and air mass
to provide a signal to the control system flow ranges from nearly 15 to 1.
to indicate fired operation. All of the
chambers are interconnected by crossfire Full Scale Combustor Tests
tubes to provide a path for flame

4
propagation during the starting cycle. A A standard production MS5001
typical reverse flow combustion system is combustion liner and dual gas nozzle
illustrated in Figure 8. (Figure 9) was chosen for the combustion
RETRACTABLE test program. Because in terms of
SPARK PLUG COMBUSTION combustion and cooling air flow
AIR distribution, the combustion system is
COOLING DILUTING TRANSITION identical to that currently being used in
AIR AIR PIECE an operating MS3002 machine burning a 319
TURBINE
NOZZLE BTU/SCF N 2 /CH 4 fuel, it is possible to
make a comparison of the combustion
performance when burning a fuel having a
large percentage of CO 2 . For purposes of
comparison, laboratory data was used;
FUEL however, the MS3002 data was taken from
NOZZLE // tests run with a direct-fired preheater
_ \ prior to the installation of an indirect
CROSSFIRE TUBE ^' U -
fired heater in the laboratory. As a
COMPRESSOR result, the CO and UHC data may be as much
DISCHARGE AIR
19305C as twice that if non- vitiated air had
been used.
FIGURE 8 REVERSE FLOW COMBUSTION SYSTEM

6
 The major objectives of this
particular series of tests were:

1) Determine the performance of the

combustor on the 425 BTU fuel.
2) Determine the effect of air
temperature on performance.
3) Determine the effect of heating value
PASSAGES on performance.

In this way, predictions could be made of

the combustor performance at regenerative
SAGES cycle conditions and of the combustion
response to upsets in fuel quality.

Figures 11, 12 and 13 compare CO and

UHC emissions and combustion efficiencies
VBTU of the MS5001 combustor burning low
heating value gases to the MS3002 N /CH 4
data at simulated firing temperatu res
corresponding to full speed no load up to
base load. Data from both tests 4114 and
4117 are included. The MS5001P, -20 F °
ambient cycle conditions were again chosen
while heating value variations were
investigated primarily at the lowest inlet
temperature (440 F) for this cycle, which °
would be a worst case situation. The fuel
temperature varied from a 225-235 F range °
(moderate heating) to a 75-120 F range (no °
heating).

FIGURE 9 MS5001 DUAL GAS FUEL NOZZLE 1600 *^

Two tests (test number 4114 and 4114) B/O TEST 4114
were conducted using a standard MS5001 265 BTU/SCF
combustion system. It should be noted 1400 LHV (BTU/SCF)
that a post Test inspection after Test ♦ 442 t 10
number 4114 was completed revealed that AA 393 3 7
one of the "dummy" crossfire tubes was not • 357 t 12
fully engaged as shown in Figure 10. ❑n 307 3 10
1200
OPEN SYMBOLS: TEST 4114
CLOSED SYMB.: TEST 4117
T 3 = 438-513 F
1000 W. = 24.9-12.7 Ibm/SEC
E CESSIVE \ O 2
HOT GAS a TFUEL = 234-71 F
FLOW 1
D O

1 °
0 800
3002 S/C T 3 = 555-507 F
O
N 2/CH 4 T FUEL = 85 F
I_ D !' X-FIRE TUBE W a = 18 Ibm/SEC (VITIATED)
2I
f COLLARE 600 319 BTU/SCF
QFLOW
MMY X-FIRE TUBE
IMPROPER (PROPERLY ENGAGED)
ENGAGEMENT
aoo ■ ^m ` ■

o \\
FIGURE 10 IMPROPERLY INSTALLED CROSS FIRE TUBE
200
The tubes are located approximately
5 inches from the fuel nozzle and any air
leakage here results in a leaner reaction 0
zone and, in turn, a lower combustor 800 1000 1200 1400 1600 1800 2000
pressure drop. The lower pressure drop TF — F
results in poor mixing. It is believed
FIG. 11 STANDARD MS5001 COMBUSTOR EFFECT OF HEATING VALUE ON CO
this is the reason for the higher CO and EMISSIONS AT LOW INLET TEMPERATURES
UHC emissions for these test points than
corresponding test points run later when
the problem was corrected.

7
 14 100

*'.._B/O TEST 4114 r7 COMB

12 ^
265 BTU/SCF T 395 = 1595± 5F
700
> 1 CO LHV = 396 BTU/SCF 10 ] 90
f
W. = 20 Ibm/SEC >
LHV (BTU/SCF) a 8 a

V442±10 O
U UHC 4 =
SA 393 ± 7
• 357 ± 12
❑n 307 ± 10
2
OPEN SYMBOLS: TEST 4114
a CLOSED SYMBOLS: TEST 4117
2 400 500 600 700
V T 3 = 438-513 F T3 — F
I
W a = 24.9-12.7 Ibm/SEC
FIG. 14 EFFECT OF INLET AIR TEMPERATURE, T 3 , ON COMBUSTOR
T FUEL = 234-71 F PERFORMANCE
300
The data clearly shows that the
MS5001 simple cycle machine would perform
3002 S/C T 3 = 555-507 F
well on the 425 BTU fuel. At -20 F °
200 N2/CH4, TFUEL = 85 F ambient conditions, the emissions data of
W a = 18 Ibm/SEC (VITIATED) Figures 12 and 13 and the inlet
319 BTU/SCF temperature effect, Figure 14, indicate
that the CO and UHC emissions would be 100
100 ppmv and 20 ppmv or less, respectively, at
MS5001P rated load conditions (=1750 F °
exit temperature, 510 F inlet) for heating °
I
values above approximately 390 BTU/SCF.
800 1000 1200 1400 1600 1800 2000 As Figure 14 demonstrates, the inlet
T F — F temperature has a pronounced effect on
emissions and therefore the CO and UHC
FIG. 12 STANDARD MS5001 COMBUSTOR EFFECT OF HEATING VALUE ON emissions from regenerative cycle machines
UHC EMISSIONS AT LOW INLET TEMPERATURES
are expected to be well below the above
levels over a wide load and ambient
temperature range.
393 442 BTU/SCF
100 f It is also interesting to note that,
again, the increase in CO (and to a lesser
357 / / ^^ 307 degree UHC) with increased exit
3002 S/C T 3 = 555-507
temperatures as the heating value is
90 lowered (Figures 11 and 12). Here, too,
N7/CH4 , TFUEL = 85 F it is believed that this is due to the
m ^ Wa = 18 Ibm/SEC (VITIATED) combination of high fuel nozzle pressure
B/O TEST 4114 319 BTU/SCF ratio producing high fuel injection
80 265 BTU/SCF velocities and the slower reaction rates
NOTE: SYMBOLS SAME AS FIG'S.
associated with the lower flame
temperature of these fuels. The data
illustrate one reason why heating value
800 1000 1200 1400 1600 1800 2000 variations on a given design are
TF — F restricted in order to maintain design
FIG. 13 STANDARD MS5001 COMBUSTOR EFFECT OF HEATING VALUE ON
pressure ratio limits. The dual gas
COMBUSTION EFFICIENCY AT LOW INLET TEMPERATURES design extends this range considerably by
providing two sets of metering orifices.
Figure 14 shows the effect of
combustion inlet air temperature on CO, Figure 15, 16 and 17 also show the
UHC and efficiency. effect of increased inlet air temperature
on combustor performance. As the data
indicate, CO and UHC emissions are
diminished considerably for a given
heating value and the operable heating
value range has been extended. Here, the
CO does not appear to increase as fuel
flow (or exit temperature) increases until
the heating value falls to about 270
BTU/SCF. Of course, this is partly due to
the fact that the fuel requirements are
less at higher inlet temperatures for a
given exit temperature, thus the fuel
nozzle AP is also lower.

8
 •

2000

1500
LHV (BTU/SCF)
i 0464±5
LHV (BTU/SCF) 1000 •356t36

O 464 - 5 ❑ 303 ± 10

• 356 ± 36 1 4 269±6

❑ 303±10 a 500 IF V 202±8

a
4 269 ± 6
2 ^
♦ 202 ± 8 40

30

4VV
20

200 ❑ 4
❑ a 10

0
800 1000 1200 1400 1600 1800 2000
01 I I I I I I
TF - F
800 1000 1200 1400 1600 1800 2000
FIG. 15 STANDARD MS5001 COMBUSTOR EFFECT OF HEATING VALUE ON CO T F-F
AT T 3 = 641 F
FIG. 16 STANDARD MS5001 COMBUSTOR EFFECT OF HEATING VALUE ON
UHC EMISSIONS AT T 3 = 641 F

of

m
f
0
U

800 1000 1200 1400 1600 1800 2000

TF - F

FIG. 17 STANDARD MS5001 COMBUSTOR COMPOSITION EFFICIENCIES AT T 3

= 641 F

9
 Another way of representing the data necessary when full load capability is
is shown in Figure 18, which illustrates required on two fuels which differ greatly
the CO emissions as a function of LHV at in heating value.
constant combustor AT.
Overall combustor pressure drop,
although low during test 4114, was, in
general, within the expected range. The
2000 pressure drop decreases with decreased LHV
for a given air flow and combustor
temperature rise. This is due to the
increased fuel/air (f) requirement, since
for a given air flow (Wa) and inlet
temperature (T ) and pressure (P), at
the first stage 3nozzle 3
150C
Wa (l+f) = P (1 - AP/P)p
RT3 95
g
where p = function of specific heat ratio
> 50( and OP/P = liner pressure drop.
a-a
a
Thus, it can be shown that as f
1 40( increases, AP/P decreases. Although this
is beneficial from a machine performance
standpoint, it also results in poorer
30C
mixing and can be detrimental to the
combustion process. Thus, this too could
20(
contribute to the increasing CO levels
with increasing f at low LHV.

1OC Liner metal temperatures were lower

than those measured on the conventional
combustor burning high heating value
fuels, and would be beneficial to
0 200 300 400 500
combustion hardware life. Pattern factors
LHV - BTU/SCF were comparable to those measured when
using high heating value fuels in this
FIG. 18 CO EMISSIONS VS LHV AT CONSTANT COMBUSTOR TEMPERATURE
RISE
combustor, therefore are not reported
here.
Although flame could be maintained
down to approximately 200 BTU/SCF (90% Part 2. LM2500 Test Proeram
CO 2 ), the combustion efficiency here is
<50% (Figure 14). A practical lower limit A. Description of Combustion System
for high CO 2 fuels would appear to be 250
BTU/SCF. TFiis represents a 41% decrease The General Electric Aircraft Engine
in heating value from the nominal 425 BTU Business Group produces a variety of gas
fuel and certainly illustrates the turbines for aircraft and marine/
flexibility of the conventional combustion industrial applications. Current
system. production marine and industrial engines
are derivatives of the General Electric
Pressure fluctuations in the reaction CF6-6 and CF6-50 high bypass ratio
zone were measured during the test series. aircraft turbofan engine family. The
The levels measured during the test aircraft derivative marine and industrial
program were comparable to those seen with engines operate on simple cycle principle.
high heating value fuels in standard Basically they are two-shaft engines
combustor systems. consisting of a gas generator and a power
turbine with the shaft power extracted by
One important influence on pressure means of a flexible coupling through the
fluctuations is the fuel nozzle pressure exhaust duct. The combustion systems of
ratio. Whereas high pressure ratios may marine/industrial engines such as LM2500
result in inefficient combustion or cause and LM5000 have many common design
fuel to impinge on the combustor wall and features. These engines have been used in
cause overheating, low fuel nozzle applications with a range of liquid fuels
pressure ratios can result in combustion and with natural gas. The following
instabilities amplified by feedback of description of the combustion system
pressure fluctuations between the gas applies to typical aircraft-derivative
supply and combustor. combustors.

This did not appear to be a problem The compressed air leaving the
over the range tested here. However, to compressor is directed into the annular
avoid such problems, a dual gas nozzle is combustor section where it mixes with the
fuel from the 30 fuel nozzles. Combustion
of the fuel-air mixture is initiated by
two igniters. Once started, the

10
 combustion is self-sustaining, and the Test Facility
igniters are turned off. The hot gases
are directed into the high pressure A brief description of The Aircraft Engine
turbine which drives the compressor rotor Business Group's Combustor Component Test
and subsequently directed into the power Facility is in the Appendix.
turbine which drives the output shaft.

The cross-section of an aircraft-

derivative LM2500 combustor and flowpath
is shown in Figure 19. The annular
Holes, 4 Places
combustor consists of four major
components riveted together: the cowl
(diffuser) assembly, dome, inner liner,
and outer liner. The cowl assembly, in
conjunction with the compressor rear
Fuel Fuel Nozzle
frame, serves as a diffuser and Heating Orifice Pressure
distributor for he compressor discharge Nozzle Value Effective Ratio
air. It furnishes uniform airflow to the Configuration (BTU/SCF) Area, (In) 2 P1/P2
combustor throughout the operating range,
A, 1000 0.0259 1.22
thereby providing uniform combustion and
even temperature distribution at the A 3 450 0.0518 1.53
turbine. Thirty vortex-inducing axial (2xA,)
flow swirl cups in the dome (one at each
A 4 300 0.1036 1.33
fuel-nozzle tip) provide flame
(4xA 1 )
stabilization and mixing of the fuel and
air. The interior surface of the dome is
protected from the high temperature of
FIGURE 20 DESIGN FEATURES OF STANDARD AND LOW BTU FUEL NOZZLES
combustion by a cooling air film.
Accumulation of carbon on the fuel nozzle
tips is prevented by venturi-shaped spools Laboratory Tests
attached to the swirler.
The combustion test program with the
LM2500 consisted of (a) full scale annular
combustor rig tests conducted at the
General Electric Aircraft Engine Business
Group, Evendale, Ohio complex; and (b) an
LM2500 three cup sector combustor test
conducted at the General Electric Gas
Turbine Division Development Laboratory in
Schenectady, New York. The Evendale full
annular combustor tests were conducted at
one atmosphere pressure. In these tests,
the combustor was evaluated at simulated
steady state engine operating conditions
FIG. 19 CROSS SECTION OF LM2500 COMBUSTOR FLOW PATH
at cycle T3's (combustor inlet
temperature) and fuel-air ratios.
A standard production LM2500 Combustor flameout limits were determined
combustion system was chosen for the at each of the test points. The starting
medium BTU fuel combustion test program. characteristics of three fuel nozzle
The combustion system length is 27.5 configurations were also determined during
inches from the compressor discharge to separate tests at a combustor inlet
the turbine inlet. The LM2500 is temperature of 80 0 F (ambient) and one
nominally rated at 27,500 HP using natural atmosphere pressure at representative
gas fuel. combustor air flows simulating engine
start-up. Test fuels consisted of
The only LM2500 combustion system two-component mixtures i.e., one
hardware modification incorporated in this combustible and the other a diluent to
test program was the use of enlarged simulate low BTU gas mixtures. These were
orifice fuel nozzles sized to accommodate obtained by blending (a) natural gas with
higher gas fuel flows required with low CO 2 , N , and steam diluents separately;
BTU fuels. Figure 20 illustrates the and (b) propane with CO 2 and N 2
design features of three test fuel respectively.
nozzles. The "A " and "A 4 t ' nozzles are
identical to the standard natur
al gas fuel
nozzles (referred to as "A L " nozzles)
except for the orifice size. or example,
disregarding changes in the ratio of
specific heats (gamma), and assuming equal
temperatures and pressure ratio, the A 4
fuel nozzles are designed to flow four
times as much gas (by volume) as the
standard LM2500 natural gas fuel nozzles.

11
 The objective of the three cup sector A comparison of the starting
test at Schenectady was to supplement characteristics of these three gas nozzles
Evendale tests at higher combustor indicate the significant improvement in
operating pressures. Besides blowout the light-off capability of A nozzles
limits, combustion efficiency was (sized for operation on 450 BTU?SCF fuel)
determined at a number of simulated LM2500 over that of the standard A nozzles which
steady state cycle test points from CO and are designed primarily for operation on
HC measurements. These steady state cycle natural gas fuel. Further, it is also
points were representative of simulated evident from this Figure that A nozzles ,

LM2500 gas turbine combustor operation at (sized for operation on 300 BTU/ CF fuel) '

ambient temperatures down to -55 F. have even better starting ability compared
Primary test fuel for the sector test was to A 3 nozzles. Similar data trend was
the 425 BTU/SCF gas mixture (Fuel VI, also observed at other representative
Table 1) consisting of saturated combustor airflows. A summary of these
hydrocarbons heavily diluted with CO 2 . results is given in Table 2.
TABLE 2 SUMMARY OF STARTING CHARACTERISTICS OF STANDARD
Full Annular Combustor Tests AND LOW BTU FUEL NOZZLES

Typical sea level ignition and Test Fuel: Natural Gas/CO 2 Mixture
starting characteristics of the three test
T 3 = 80°F P 3 = 1 Atmosphere
fuel nozzles at one combustor airflow are
shown in Figure 21. A standard spark
ignitor ignition system was used to ignite COMBUSTOR FULL PROPAGATION LIMIT (BTU/SCF)
the fuel-air mixture. Data were obtained AIR FLOW NOZZLE CONFIGURATION
by maintaining a constant diluent (CO ) ? (PPS) Al A3 A4
flow rate and slowly increasing the
combustible flow rate while trying to 6.0 606 550 505
ignite the mixture. The individual
component gas flow rates were then 8.0 665 605 560
recorded at ignition/full propagation.
10.5 680 650 630
This procedure was repeated for a range of
diluent flow rates to obtain a series of
full propagation points. The Efects of Diluents
combustible-diluent flow rates obtained at
full propagation are plotted in Figure 21. Effects of diluents such as CO 2 , N 2 ,
The locus of these points is referred to and H2O (steam) on combustor flameout are
as the starting characteristic. Lines of summarized in Table 3 for natural-gas-
constant heating value (BTU) are cross based, simulated-low-BTU mixtures and
plotted in this Figure for reference. plotted in Figure 22 for comparison. Note
A^ NOZZLE that any given inlet temperature (T 3 ), the
800 test data indicate a range of BT U 3 levels
700
at which the combustor would flameout.
700 ERTU/scF / / 450 The data scatter from the average was
largest at the lowest air temperature
I 600 (600 ° F) tested. This Figure shows that
nitrogen has a more favorable effect on
flammability limit than CO as indicated
500
C by classical flammability data. For
3
0 exampled at a combustor inlet temperature
q^
U)
of 700 F, the combustor flameout limit is
A^ NOZZLE about 26% lower with N ` than with CO .
O

300 \ Although classical data indicates that tie

CC A, NOZZLE
flammability limits for steam should be
LM2500 FULL ANNULAR COMBUSTOR
TEST FUEL. NG/CO 2 MIXTURE
higher than for N 2 , the full annular
200
T AME = 80 F, P, = 1 ATM PRESSURE
combustor flameout 2 data show flameout
COMBUSTOR AIR FLOW = 8 PPS limits of the steam diluted fuel mixture
100 to be equal to that of N diluted fuel.
This is probably due to the higher

temperature of the (NG/steam) fuel mixture
400 800 1200 1600 2000 (=320 ° F) compared to 160 0 F for the NG/N 2
CARBON DIOXIDE (CO 2 ) FLOW RATE, PPH mixture.
FIG. 21 STARTING CHARACTERISTICS OF STANDARD AND LOW BTU FUEL NOZZLES

It can be seen that the full

propagation line (starting characteristic)
becomes somewhat asymptotic as the diluent
flow rate increases thus indicating that
combustor full propagation cannot be
achieved below heating values
corresponding to this asymptote. This
implies a physical limit (in terms of
heating value) for any given gas nozzle
configuration.

12


TABLE 3 EFFECT OF DILUENTS ON

0.35
r .. LM2500 FULL ANNULAR COMBUSTOR

COMBUSTOR FLAMEOUT LIMIT* (BTU/SCF)

NATURAL GAS/CO, MIXTURE T3 = 1000°F

Combustor Inlet Test Fuel Composition w PROPANE/CO2 MIXTURE

Temperature (°F) NG/CO 2 NG/N 2 NG/Steam y "0
STRAIGHT PROPANE

600 340 245 235 0

700 290 215 210 0.25

\-STRAIGHT NATURAL GAS

800 260 200 200

"Average" Values 0 400 800 1200 1600 2000 2400

FUEL LOWER HEATING VALUE, BTU/SCF

FULL ANNULAR COMBUSTOR AT 1 ATM PRESSURE

FIG. 23 COMPARISON OF COMBUSTOR PATTERN FACTOR FOR DIFFERENT
A a NOZZLE CONFIGURATION BTU LEVELS
600
NATURAL GAS INS) BASED FUEL MIXTURES
Typical pattern factor for current
U production LM2500 engines running on
500
S
H natural gas is 0.24 - 0.26. Figure 23
m
shows that the PF increases as the BTU
400 NO/CO2 MIXTURE content of the fuel is decreased. This is
\ \ _ / NG/N z MIXTURE not surprising in standard combustors
300 Q^ designed to operate on "rich" fuels
probably because of the difficulty of
mixing the increased volume of reduced
U 200
ti BTU fuels with the combustor airflow.
O NO/STEAM
w MIXTURE
100 It is interesting to note that even
though the PF is higher with low BTU
fuels, at constant engine horsepower
400 500 600 700 800 900 output, the peak temperature is actually
T 3 - COMBUSTOR INLET TEMPERATURE, °F reduced, or lower, with low BTU fuel
FIG. 22 EFFECT OF DILUENTS ON COMBUSTOR FLAMEOUT because the overall combustor temperature
rise required (T 9 -T 3 ), is lower due to?

Combustor Exit Temperature Pattern the higher mass }low through the turbine.
This is shown graphically on Figure 24.
Figure 23 illustrates the results of Combustor temperature uniformity appears
full annular combustor tests conducted to to be compromised by low BTU fuel
determine the pattern factor (PF) on combustion, but overall high pressure
reduced BTU fuels. These tests were turbine life will be enhanced for not only
conducted at combustor inlet temperatures are peak temperatures reduced at the
of 1000 0 F and fuel (combustible)-air stator, but also the overall average
ratios simulating max power conditions. temperature imposed on the turbine rotor
Pattern factor is a parametric indicator blading is reduced because of lower T 4 for
of the quality of the uniformity of equivalent horsepower levels.
combustor exit temperature distribution, 2400

i.e., a low pattern factor is indicative LM2500 COMBUSTOR PEAK TEMPERATURE CHARACTERISTICS

of a more uniform temperature distribution

free from local hot spots (high temp
peaks). STANOARO NATURAL GAS /
IFF"=0.26)

By definition, PF = T Max - T 3_9 2200

J.7 — S

Where TPF = Pattern Factor S

2100
Max = Combustor Exit
Maximum Temperature
T 3.9 = Combustor Exit
2000
Average Temperature
/ PF — PATTERN FACTOR
T 3 = Combustor Inlet
' CO Z DILUENT
Average Temperature
1900 L
12000 14000
I
16000 18000
I I
20000
I I I
22000 24000 26000
SHAFT HORSEPOWER, SHP (PWSOI

FIG. 24 PEAK TEMPERATURE CHARACTERISTICS OF STANDARD AND LOW

BTU GAS

13
 Combustor Liner Temperature
1200 -
Combustor liner wall temperatures are LL
also reduced with low BTU gas fuels for
similar reasons, i.e., the addition of w 1100 -
high volumetric quantities of diluents
(CO 2 , N reduces combustion a
a

LM25000OMPNNENTPR.SECTORTEST
temperatures. This in turn results in
reduced heat flux to the combustor walls, W T 815°F
and hence lower metal temperatures. The J l000
P3 234 PSIA
temperature of the combustor liners a 3
F/A 0.022 (COMBUSTIBLES)
depends on the local flame temperature and
on the radiation and convection heat
m
NOTE: LOW BTU FUEL SIMULATED BY
900
WATER INJECTION INTO THE
transfer coefficients. In the combustor
primary zone, the predominant mode of heat O PRIMARY COMBUSTION ZONE

transfer to the liners is radiation. 0

Radiant heat transfer will tend to be 900 1000
800400 500 600 700 800
reduced with the low BTU fuels because of
HEATING VALUE — BTU/SCF
reduced primary zone flame temperatures.
The convective heat transfer coefficient FIG. 25 COMBUSTOR LINER TEMPERATURE REDUCTION WITH LOW BTU
will tend to increase due to higher FUELS
primary zone velocities with the low BTU
gases, but this effect will be Three Cup Sector Test Results
counteracted to some extent by the reduced
flame temperatures. Overall, primary zone As was mentioned earlier, a three cup
metal temperatures would decrease with low sector combustor was evaluated at
BTU fuels if liner cooling flow is simulated steady state engine operating
maintained at the same levels. Downstream conditions at cycle T 's and fuel-air
of the primary zone, convective heat ratios. The test press
ure level was
transfer predominates; liner temperatures restricted to approximately 140 PSIA
in this region should be almost because of test facility limitations
independent of fuel type, again assuming imposed by the gas subsystem. Fuel
that cooling levels are maintained. (mixture) temperatures at the fuel
manifold were maintained at 120 ± 10 ° F
During full annular combustor throughout these tests. Combustion
component testing at one atmosphere efficiency was calculated at each of the
pressure, the trend of lower liner test points from CO and HC emissions
temperature as the diluent gas measurements. The CO and HC emissions
concentration increased was measured for were averaged from the two exit emissions
both methane-steam and propane-carbon rakes for the purpose of computing
dioxide mixtures. The quantitative combustor efficiency. (The combustor
values, however, were not well defined efficiency is defined as the ratio of
because the overall heat transfer actual heat released to the theoretical
characteristics in one atmosphere test are maximum heat released from the fuel,
so low that relative differences tend to assuming complete combustion to CO 2 and
be obscured by other extraneous effects. H20).

A series of tests was performed Figure 26 is a plot of combustor

earlier at high pressure on a three nozzle efficiency versus combustor inlet air
36 degrees LM2500 sector test vehicle temperature, T 3 , for three (3) different
using methane fuel and water injection, fuel heating val ues. The lower heating
These tests were conducted to investigate value gas mixtures (300 and 280 BTU/SCF)
NO abatement; however, the results were obtained by additional dilution of
obtained parallel a low BTU fuel composed the gas mixture (Fuel VI, Table 1) with
of methane and steam. The fuel and water carbon dioxide (CO 2 ). This Figure shows
injection rates were converted to that the combustion efficiency of 425
equivalent low BTU gas values, and results BTU/SCF gas at T 's above 600°F
presented on Figure 25 which shows that (corresponding to gas turbine power
low BTU gas operation results in settings above 10,000 SHP on a 59 o F
significant reductions in liner ambient day operation) is over 99% and
temperatures, that at T 's below 600 ° F the combustion
efficiency 3 drops below 99%. The measured
combustion efficiency is 96.7% at a T of
400 ° F. This T corresponds to less 4han
1000 shaft horsepower at 59 ° F ambient
operation. With the 300 BTU/SCF mixture,
the decrease in combustion efficiency is
more pronounced compared to the 425 BTU
gas mixture at T 's below 6000F.
Combustor flameout occurred at
approximately 275 BTU/SCF during this
testing. Results of this test indicated
stable and efficient combustor operation

14
 over a wide range of engine operating 1. Higher combustor operating pressures
conditions with the 425 BTU/SCF gas fuel, exist in the actual gas turbine
During these tests, internal conditions engine. Facility pressure
equivalent to 10,000 HP at temperatures of limitations are usually encountered
-55 0 F were demonstrated. Facility during sector testing which result in
limitations prevented demonstrations of higher CO emissions. Since combustor
lower horsepower levels under these low efficiency is calculated from
temperature conditions; however, flameout emissions, lower efficiencies are
tests performed on propane-CO fuels shown obtained.
in Figure 27 indicate that the gas turbine
is capable of operation at levels of 5,000 2. Quenching of the exhaust gases is
HP at -55 0 F with 425 BTU/SCF fuel. Stable encountered in sector tests because
and efficient operation was also of the side wall cooling requirement.
demonstrated with fuel heating values as This, again, leads to higher
low as 300 BTU/SCF at conditions emissions and lower combustion
corresponding to gas turbine power efficiency.
settings above 5,000 SHP at ambient
temperatures of 59 0 F. In the past, CO emissions from sector
LM2500 THREE CUP SECTOR TEST combustor tests have been measured at
A 4 NOZZLE CONFIGURATION approximately six (6) times those levels
100 measured on full scale engine tests.
-' -0 Based on the above, it is expected that
the combustion efficiency in actual gas
turbine combustors (operating at higher
cycle pressures) burning low BTU fuels
>- would be higher than that indicated in
w
0

U
BTU/SCF Figure 26. Figure 28 shows the combustion
efficiency of a typical LM2500 natural gas
U-
0 425
w 95 ❑ 300 fired engine compared to the efficiency
0 280 obtained from a three-cup sector test.
0
I- TYPICAL LM2500 ENGINE ON NATURAL GAS (A 1 NOZZLES)
C

0
O 100-------
0
p //
LL

N Y Y /
Y Y 11 ^n Y 0 /

V
'n ° ^^v o //
90 = I LM2500 THREE CUP SECTOR DATA
N /1 O / ON 425 BTU/SCF FUEL
w ^^ A4 NOZZLE CONFIGURATION
200 300 400 500 600 700 800 900
T 3 -COMBUSTOR INLET TEMPERATURE, °F ,'
w 95
FIG. 26 COMBUSTION EFFICIENCY VERSUS COMBUSTOR INLET Z
0
TEMPERATURE C
D

LM2500 COMBUSTOR (A 4 NOZZLE CONFIGURATION)

0
U
n FULL ANNULAR COMBUSTOR, C 3 H 4 CO, MIXTURE. 1 ATM PRESSURE
❑ FULL ANNULAR'COMBUSTOR, NG/STEAM MIXTURE, 1 ATM PRESSURE
O ENGINE TEST, NG/STEAM MIXTURE
600 -
• THREE CUP SECTOR, 425 BTU DILUTED FUEL, P 3 = 50 PSIA - A 4 NOZZLES 90

C
U Q THREE CUP SECTOR, 425 BTU DILUTED FUEL, P 3 -50 PSIA - A, NOZZLES
0 1425 BTU/SCF)
a 200 300 400 500 600 700 800

T 3 - COMBUSTOR INLET TEMPERATURE, °F

w 400

U FIG. 28 COMPARISON OF ENGINE AND SECTOR COMBUSTOR EFFICIENCY

U
A comparison of the flameout
0
° 0 characteristics of the LM2500 combustor is
0 200 shown in Figure 27 and is a summary plot
C of data obtained from various test
sources. The one atmosphere combustor
component test data obtained on
propane-CO mixture parallels the natural
0
0 200 400 600 800 ;000 gas-steam data. The three cup sector data
T 3 -- COMBUSTOR INLET TEMPERATURE on low BTU fuel nozzles (A 4 and A ) do
indicate a significant improvement (Power)
FIG. 27 LM2500 COMBUSTOR FLAMEOUT CHARACTERISTICS
in flameout limit as was anticipated
Past experience has shown that combustion because of higher combustor operating
efficiency from actual engine operation pressures. Based on these test data, it
will be higher compared to sector tests is concluded that the LM2500 gas turbine
for the following reasons: would have adequate flameout margin over a
wide range of combustor operating
temperatures with the 425 BTU/SCF fuel.

15
 Engine Starting and Low BTU Fuel A review of the three cup sector
Transit inn combustor test data on low BTU fuel during
simulated operating line investigation as
It was concluded from Figure 27 that shown in Figure 26 indicate combustion
the basic unmodified LM2500 combustor efficiencies of 99% plus at T 's above
would have adequate flameout margin when ° °
600 F. Below 600 F, T , the com
? bustion
operating on the 425 BTU/SCF heating value efficiency drops below 9% and decreases
fuel at ambient temperatures of -55 F. ° with decrease in combustor inlet
However, blowout margin, in general, temperature. The measured combustion
decreases at lower engine power settings °
efficiency at a T of 400 F is 96.7%.
because of lower combustor inlet Therefore, there 3 appears to be a
temperatures (T ). Therefore, the performance penalty for operation below
question of LM2500 engine starting and °
600 F, T 3 , on 425 BTU/SCF fuel.
transition to low/medium BTU fuels has
been one of considerable interest. Full Blowout Margin
annular combustor tests at one atmospheric
°
pressure and 80 F, T 3 , on simulated An examination of atmospheric test
s have indicated
methane-based low BTU fuel data plotted in Figure 27 for the C 3 H 8 /CO 2
that (Figure 21) the present unmodified mixture at combustor inlet temperature
LM2500 combustor will need a higher BTU °
below 600 F (simulated operating line
fuel (600-700 BTU/SCF) for lightoff. investigation at T b -55 F) indicates a
Therefore, LM2500 engine starting and combustor flameout Timit of 340 BTU/SCF at
acceleration to idle on low BTU fuel, per °
400 F, T . The blowout margin is
se, is not contemplated at this time. It therefore 5 BTU/SCF for the low BTU fuel
is expected that the normal loading °
(425 BTU/SCF) at 400 F, T . Since these
sequence will be to light-off on natural figures are based on tesc data at one
gas and then transition to the low BTU atmosphere pressure, it is expected that
operation at some intermediate power the blow out margin would be even higher
setting. in actual engine combustors operating at
higher cycle pressures. Indeed, flameout
Figure 29 illustrates the transition data obtained from the three cup sector
curve showing the ambient tests at higher combustor pressures on an
temperature-engine power setting alternate fuel nozzle configuration (A
relationship for a constant compressor fuel nozzles) do indicate a significant
discharge temperature (T 3 ). For example, improvement (lower) in flameout limit as
at an ambient temperatur e of 59 F, the ° was anticipated. However, since three cup
transition to the 425 BTU/SCF heating sector tests to identify the effect of
value fuel could be effected at engine various parameters such as fuel nozzle
power levels of 1,000 horsepower or above. orifice size and combustor pressures on
AT -55°F ambient temperature, the blowout limit have not yet been completed
transition would be at power settings of at the time of writing this paper, no
5,000 horsepower or above. Some control further discussion of it follows. Based
system modifications will be needed to on the above discussions, it is concluded
satisfactorily transition from the that the blowout margin of the LM2500
starting fuel to the operating low BTU combustor operating on 425 BTU/SCF fuel at
fuel. °
simulated ambient conditions of -55 F would
be higher than 85 BTU/SCF. This
7000 corresponds to safe combustor operation
even with fuel heating value changes by as
much as ± 20%.
6000
—CONSTANT T 3 = 350°F VII. Summary
(Compressor Discharge Temp.)
5000
Heavy Duty

4000 The results of the full scale single

OUTPUT burner tests showed that the nominal 425
IGHP
NORMAL BTU/SCF fuel can be burned in a conventional
3000
425 BTU/SCF combustor at all full speed conditions.
OPERATION Emission requirements, in particular CO and
2000 UHC, can be met over a wide load range and
ENRICHED ^-- ambient conditions when operating on the 425
STARTING BTU/SCF gas. A variation in heating value of
1000 FUEL OPERATION more than ± 20% can be tolerated in the
standard combustor while still maintaining
IDLE adequate performance. High fuel nozzle
(5000 RPM) pressure ratios, which will occur on a fixed
—60 —40 —20 0 +20 +40 +60 +80 geometry fuel nozzle may contribute to
AMBIENT TEMPERATURE, °F increased CO emissions as the heating value
decreases. However, the program has clearly
demonstrated the ability of the standard
FIG. 29 LM2500 TRANSITION CURVE combustion system to operate over a wide range
of heating values.

2.6
LM2500 APPENDIX

The basic objectives of the work described The Aircraft Engine Business Group's Combustor
were to determine the minimum practical gas Component Test Facility is designed to conduct a
heating value which can be used in an variety of combustion related tests. The test
unmodified LM2500 gas turbine combustor and to facility contains inlet ducting, exhaust ducting,
quantify the effects of fuel chemistry fuel and air supplies, controls and instrumentation
variations on the minimum heating value. required for the conduct of full scale combustor
component tests. The test cell piping is arranged
Test results based on LM2500 full annular to accommodate two test vehicles simultaneously, and
combustor component tests and three cup sector even greater utilization is effected by mounting
pressure tests on a variety of representative test vehicles on portable dollies with quick-change
low BTU gas mixtures show that the LM2500 gas connections so that build-up operations are
turbine has satisfactory combustor performance accomplished in another area and a vehicle occupies
when burning a 425 BTU/SCF gas over a wide the cell only for the duration of its actual
range of operating conditions. In fact, the testing. Control consoles and data recording
LM2500 combustor has stable operation with 300 equipment are located in the adjacent control room.
BTU/SCF heating value fuel down to 5,000 SHP at
ambient temperatures of 59 F. When the ambient The test facility has a process air supply
temperature drops to -55 F, the combustor will capable of delivering up to 100 lbs. of air per
have adequate flameout margin at engine power second. Test fuel supply capabilities include
levels of 5,000 horsepower or above when conventional aircraft fuels, marine diesel, natural
operating on a 425 BTU/SCF gas. gas, and propane. In addition, it is possible to
blend combustible gases such as natural gas and
Although combustor temperature uniformity propane with carbon dioxide, steam and nitrogen
appears to be compromised by low BTU fuel separately to simulate low and medium BTU operation.
combustion, overall high pressure turbine life The fuel blending system capability includes
will be enhanced, for not only are peak provisions for accurate flow rate measurements of
temperatures reduced at the stator, but also each of the gas fuel mixture components, total fuel
overall average temperature imposed on the flow rate, and digital displays for on-line
turbine rotor blading is reduced because of monitoring.
lower T 4 for equivalent horsepower levels.
Significant reductions in combustor liner
temperatures and gas stream temperatures can be
expected on low BTU gases for a given shaft
horsepower, as compared to natural gas
operation. During cold ambient temperature
operation, the LM2500 engine starting can be
accomplished on standard natural gas, and
transition to the 425 BTU/SCF fuel could be
accomplished at approximately 5,000 horsepower
level or above with ample blowout margin.

Overall results have shown that the

standard, unmodified LM2500 gas turbine
combustor can operate on the CO 2 -diluted fuels
with heating values as low as 300 BTU/SCF.
Specific investigations aimed at 425 BTU/SCF
gas capability have shown that the standard,
unmodified LM2500 gas turbine will produce the
power range performance required for a specific
application without compromising heat rate or
component life over the ambient temperature
spectrum for this application.

REFERENCES

Gordon, Sanford, "Computer Program for

Calculation of Complex Chemical
Equilibrium Compositions, Rocket Performance,
Incident on Reflected Shocks, and Chapman -
Jouquet Detonations," N71-37775, 1971 National
Aeronautics and Space Administration,
Washington, D.C.

Coward, H.F. and Jones, G.W., "Limits of

Flammability of Gases and Vapors,"
Bulletin 503, Bureau of Mines, Washington,
D.C., 1952.

Zabetakis, Michael G., "Flammability

Characteristics of Combustible Gases
and Vapors," Bulletin 627, Bureau of Mines, WP744/C307/DS/150
Washington, D.C. 1965.

17