THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS

345 E. 47 St., New York, N.Y. 10017

91-GT-329

E s The Society shall not be responsible for statements or opinions advanced in papers or in dis-
cussion 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 in an ASME Journal. Papers are available
1^( from ASME for fifteen months after the meeting.
Printed in USA.
Copyright © 1991 by ASME

Humidity Effects on Gas Turbine Performance

Downloaded from https://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1991/78996/V002T02A025/2400497/v002t02a025-91-gt-329.pdf by guest on 23 March 2020

J. BIRD and W. GRABE
Engine Laboratory
Institute for Mechanical Engineering
National Research Council of Canada
Montreal Road, Ottawa K1A 0R6

ABSTRACT Further cooling will cause the excess water to condense.

It is assumed that this process is slow enough to permit
Moisture in the intake air of a gas turbine can affect the establishment of equilibrium conditions. In engine
its operation and performance in two different ways: by inlets, depending on the inlet Mach number, depressions
possible condensation in the inlet and by changing the in static temperature and pressure can be substantial.
gas properties throughout the cycle. Condensation can If certain conditions are met, the inlet static
be controlled by restricting engine operation with temperature may fall below the dew point, leading to the
limits on relative and absolute humidities. Two possibility of condensation. If the cooling process is
fundamental correction approaches for the effects of very rapid and is followed quickly by a temperature
humidity on major engine parameters were investigated; rise, equilibrium conditions will not be reached and an
they were found to compare very well. Both methods unstable state may exist. This state is called
correct parameters as a function of absolute humidity, supersaturation, in which relative humidity well beyond
yielding corrections of between 0.1 and 0.8%, for high 100% can be present for a short time.
humidity test conditions. Additional operational,
engine-specific humidity corrections were examined: some
notable differences were observed. Recommendations are 2.0 DEFINITION OF PROBLEMS
made for the correction of major performance data for
absolute humidity. Air humidity will have several influences on the
operation and performance of gas turbine installations.
In this paper, the discussion will be limited to the
1.0 INTRODUCTION effects on the engine. Although the consequences are
complex, they fall into two major categories:
While it has been recognized for a long time that high condensation and changes in performance through
humidity levels in the air will affect the performance different (moist) gas properties.
of gas turbine engines, no single reference appears to
exist which surveys the effects or problems associated Condensation and performance effects will be discussed
with humidity. This paper will discuss previous in the next two sections. However, the problems
experimental and analytical work, and review some associated with each category may be highlighted. The
humidity corrections used by engine operators. The consequences of condensation are:
intent is to identify the magnitude of probable
corrections, when testing engines in ambient air, and to a) decreased compressor efficiency resulting from
assess the assumptions made in the estimation methods surface wetness of compressor blading;
currently available.
b) reduced cycle efficiency, as condensation
The presence of water vapour in air is called humidity, releases latent heat, raising the inlet
measured as the mass of water vapour per unit mass of temperature; the subsequent evaporation does not
air. It is also referred to as "absolute" or "specific" compensate for this degradation, because of the
humidity, or as "humidity ratio". The amount of water irreversibility of the process (Blake 1975);
in the air depends on the temperature and meteorological
conditions and the availability of moisture. In an C) fouling of compressor blading by the deposition
equilibrium state, water can not exist as vapour beyond of particulate matter, which acted as
saturation, but must condense into liquid form. condensation nuclei (Zaba and Lombardi 1984); and
Moisture content below this saturation limit is given as
percentage saturation or, more commonly, as relative d) blocking of pressure instrumentation lines in the
humidity (RH). Relative humidity is measured in intake by water droplets.
percent, with the saturation humidity being 100%. A
good treatment of psychrometry and charts for the Even if condensation does not take place, the water
interrelationships between absolute humidity, relative vapour will affect the gas properties and therefore the
humidity, and temperature are given by Zimmerman and cycle performance. The expected performance changes
Lavine (1964). relative to that in dry air are:
As air is cooled at constant absolute humidity, its a) errors in establishing component efficiency of 2
relative humidity increases until it reaches its to 3 percent by assuming perfect gas properties
saturation temperature, also called the dew point. (Fessler 1979);

Presented at the International Gas Turbine and Aeroengine Congress and Exposition
Orlando, FL June 3-6, 1991
b) increased specific fuel consumption; turbofan engine. The manufacturer specified the
following psychrometric conditions (Blazey 1983):
c) increased fuel flow; and
relative humidity less than 75%;
d) thrust or power changes depending on control absolute humidity less than 0.01428 kg
system functions. water/kg air; and
no visible moisture in the air.
3.0 CONDENSATION The last limitation should be met by adherence to the
other two. The absolute and relative humidity specified
Theoretically, condensation may take place whenever the coincide at a temperature of 297 K.
static temperature of air, accelerated into an engine
inlet, falls below the dew point. However, this phase A plot of theoretical condensation limits for a range of

Downloaded from https://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1991/78996/V002T02A025/2400497/v002t02a025-91-gt-329.pdf by guest on 23 March 2020

change requires a finite dwell time which may not be likely ambient temperatures shows that for an inlet Mach
possible in a short intake (Section 3.1 and Spencer and number of 0.44 and a temperature of 297 K, the
Archer 1971). The dwell time is determined by the inlet saturation limit is slightly below 50% RH (Grabs and
Mach number which depends on the particular engine and Bird 1988). At 75% RH, the GE proposed limit, a
inlet. The Mach number may vary from 0.3 to 0.5, with supersaturation level of 150% exists, in the absence of
modern gas turbine engines under static conditions. condensation. This appears to be a safe condition,
considering that Spencer and Archer (1971) established
Zerkle, Colley, and Doel (1982) presented a detailed a supersaturation level of 220% in a bellmouth inlet,
analysis of the condensation process in atmospheric air. without condensation initiation. However, intakes
They pointed out the conditions required for practical should be checked for condensation if any testing is to
condensation, such as the presence of condensing nuclei, take place in the supersaturated regime.
super saturation, and duct length or residence (dwell)
time. The authors emphasized that, while their 4.0 HUMIDITY EFFECTS ON PERFORMANCE
mathematical model took into account all the crucial
elements of the condensation process, verification by With the problem of condensation controlled, the effects
experiment was imperative. of humidity on gas turbine performance parameters remain
to be taken into account. While the relative humidity
3.1 Condensation in an Airframe Duct or Bellmouth Inlet was the controlling factor for inlet condensation, it is
the absolute humidity which affects the gas properties
Spencer and Archer (1971) studied the effects of of the cycle, and the performance elements.
condensation in the long air inlet duct of the A-7
aircraft with a TF30 engine, operating in a humid Before citing individual correction methods, it is
climate. This investigation was prompted by noticeable important to present a rationale for applying calculated
shifts in engine trim, between day and night. Their humidity corrections. Humidity, as other atmospheric
report showed that increases in relative humidity from elements, such as pressure and temperature, will shift
30 to 60 percent in daytime to 90 to 95 percent at night engine performance parameters from their dry, standard
time caused increases in turbine inlet temperature of up day values. However, any shift must be considered along
to 50 K. with accompanying shifts in other parameters. To
quantify a particular shift and to derive theoretical
They compared the long, fighter-type air intake duct and corrections, certain reference condition(s) must be
a bellmouth inlet. It was found experimentally with an chosen to be fixed, ie. comparisons at constant speed,
aircraft duct, that the local relative humidity could Mach number, or power setting.
reach an instantaneous peak in excess of 500%, before
the initiation of condensation. With a bellmouth inlet, In the following discussion of correction methods, some
on the other hand, a relative humidity of only about substantial and fundamental disagreements were
220% would occur at the compressor face. This discovered. Unfortunately, fixed reference points were
relatively low state of supersaturation was possible not always documented, presenting certain difficulties
because of the extremely short residence time of the air in quantitative comparisons.
in a bellmouth and the comparatively low Mach number.
However, the authors warned that each installation would 4.1 Correction Methods
have its own aerothermodynamic characteristics, based on
mass flow and duct geometry. The work of Samuels and Gale (1950) has been widely
accepted and considered a standard in the correction of
Subsequently, Blake (1975) investigated the possibility performance parameters for humidity. Their approach
of condensation in a JT9D bellmouth inlet. From depended on the definition of the thermodynamic
psychrometric and compressible flow tables, the properties of humid air as the average of those of dry
theoretical limits for condensation were established. air and water vapour, weighted by the relative masses of
With an approximate Mach number of 0.5 for the JT9D these two components. They derived one-dimensional flow
turbofan engine, a minimum ambient RH of 40% would be expressions for performance parameters as functions of
required for possible condensation. This theoretical inlet parameters. Their work has two applications:
minimum condition would be met in many operational modelling with cycle programs and corrections of test
situations. Assuming that condensation took place, an data for inlet air humidity.
inlet temperature rise of 4 K was predicted for typical
summer conditions. This rise would increase the thrust Humidity corrections were established for engine speed,
specific fuel consumption of this engine by 0.7%. A airflow, fuel flow, jet thrust, and turbine outlet
laser system, installed at the throat of the bellmouth, temperature. Samuels and Gale defined a correction
indicated the presence of water droplets at higher Mach factor as the ratio of the performance in dry air to
numbers, when condensation was predicted by theory. that in moist air, for the moist and dry conditions at
equal flight Mach numbers.
3.2 Humidity Limits
The approximate humidity correction factors from moist
From the discussion of Section 3.1, we may conclude that air conditions (m) to dry air conditions (n) proposed by
condensation may take place in an engine intake. Samuels and Gale (1950) are:
Depending on the specific installation, testing will
need to be limited to relatively dry atmospheric
conditions. N / ✓W
Yc.n'Rc,n
Engine Speed: -
As an example, a cell correlation program between N/ / VYc.mRcm

General Electric (GE) and the National Research Council

Canada (NRCC) called for limits on tests with an F404
 Wa,nVW/a — Rc,m Yc,n and turbine inlet temperatures of 900 to 1400 K. The
Airflow: model and methods used included the following
Wa,,,f /6 Rc,n-Yc,m assumptions:

a) compressor aero-thermodynamic characteristics

(pressure-flow-speed) are invariant with moisture
/a ✓ff RcmYcn content;
Fuel flow: Wt Cpbn
Wt m /6V VRc,,Yc,, Cp,b,,, b) compressor and turbine efficiency characteristics
are unaffected by moisture; and
c) inlet temperatures were only varied from 288 to
F.18 _ Ye,n 333 K.
Jet thrust:

Downloaded from https://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1991/78996/V002T02A025/2400497/v002t02a025-91-gt-329.pdf by guest on 23 March 2020

Fm 16 Ye,m
These assumptions and inlet temperature limitations are
not supported by experimental data and the validation
process for the model is not discussed.
The engine components are identified by subscripts as:
compressor (c), combustion chamber (b) and exhaust Correction factors which were weak quadratic functions
nozzle (e). The usual definitions apply for the gas of absolute humidity (H) were derived as:
constant (R), specific heat (C ), specific heat ratio
(y), and the ratios of inlet to reference conditions for Thrust: Cx = 1 - 0.0826 H - 0.0021 H2
pressure and temperature (6,0).
Fuel Flow: CF = 1 + 0.4882 H+ 0.3828 H2
The theoretical analyses were corroborated by tests
carried out with two different engines of low overall Airflow: CA = 1 - 0.3352 H - 0.4010 H2
pressure ratio, under controlled humidity conditions.
They were tested by NACA, with the humidity corrections Compressor Speed: CN = 1 + 0.2607 H - 0.1538 H2
applied according to the detailed method given (Samuels
and Gale 1950). With all data corrected to constant Turbine Inlet : CT = 1 - 0.0845 H - 0.119 H2
engine speed, the agreement between the experimental and Temperature
the theoretical predictions was better than 0.5 percent
for specific humidities up to 0.03 kg water per kg of The humidity, H, has the units of mass ratio of water to
dry air. Samuels and Gale concluded that the factors air. The output of the model, as given in the form of
were generally applicable to any engine and convenient plots by Fishbeyn and Pervyshin (1970), for the various
to apply to correct engine performance for humidity. cycles and inlet conditions, was within 0.5 percent of
these functional expressions. Fishbeyn and Pervyshin's
Sample plots for the corrections of compressor speed, corrections of compressor speed, fuel flow, turbine
fuel flow, turbine temperature and thrust, as linear inlet temperature, and thrust, as functions of absolute
functions of absolute humidity, are given in Figures 1 humidity, have been included in Figures 1 to 4.
to 4. Corrections reach one percent for moderately high
humidity levels. Although the derivation of correction factors for humid
inlet air was aimed at turbofan engines, the report
Samuels and Gale's approach has been analyzed in detail recommended that they be used with all gas turbine
(Grabe and Bird 1988). Their methods were found to have engines. A later paper by Fishbeyn (1975) examines the
some shortcomings: errors introduced in the calculation of compressor and
turbine efficiencies, if humidity is neglected. His
a) properties of moist combustion gases must be theoretical results showed that for moderate humidities,
known; the compressor efficiency would be underestimated by 0.4
to 1.7% and the turbine efficiency would be
b) choked turbine stator flow and unchoked final overestimated by 0.7 to 1.2%, in the absence of humidity
nozzle flow are assumed; corrections.

c) component and engine temperature ratios are 4.2 Comparison of Methods and Results
assumed invariant with humidity levels; and
In general, Fishbeyn and Pervyshin derived corrections
d) reference conditions are not the usual ones used to performance parameters from engine model outputs from
for performance parameter comparisons. a variety of cycles. The gas property descriptions in
the model were modified to allow for moist inlet air.
Analytical or experimental validations are indicated for Samuels and Gale, on the other hand, formed explicit
some of the assumptions and restrictions. The validity one-dimensional expressions to define each performance
of the corrections will depend on the specific parameter. They used the same gas property descriptions
application, e.g. engine configuration, power setting, as Fishbeyn and Pervyshin, but assumed many parameters
or humidity levels. Detailed aero-thermodynamic models remained invariant with moisture content.
may be used to approximate the complex functional
relationships, offering an improvement over simple, one Three differences in their methods should be identified.
dimensional flow, closed-form expressions. However, the Firstly, Fishbeyn and Pervyshin calculated the gas
methods used are judged sufficiently valid for their properties for the mean temperature of the process.
application in assessing the relative magnitudes of Samuels and Gale did not specify what point of the
humidity corrections. process they considered for temperature calculation. If
it was the process inlet temperature, differences could
Fishbeyn and Pervyshin (1970) also investigated the result between the two methods for compressors with high
effects of atmospheric humidity on the characteristics pressure ratios.
of turbofan and other engine cycles. They generated a
series of humidity correction factors for major Secondly, for the inlet airflow correction, Fishbeyn and
performance parameters. Grabe and Bird (1988) also Pervyshin used a flow description which allowed for
studied their approach, including assumptions and isentropic flow in a duct of varying area, including
limitations. provision for choked flow. Conversely, Samuels and Gale
did not use this general expression but rather a simple
The thermodynamic properties of the moist air were one-dimensional compressible flow equation with fixed
estimated in the same way as Samuels and Gale (1950). area. Within the practical range of humidity (0 to
However, Fishbeyn and Pervyshin incorporated these 0.043 kg water per kg of dry air), the differences in
properties into a component-level, thermodynamic model. airflow would amount to 0.08%, at most.
This model was run with varying humidity and power
setting levels for cycles with bypass ratios of 0 to 4

3
The Samuels and Gale curves are linear functions, while moderately high humidity of 0.0223 kg water per kg dry
the Fishbeyn and Pervyshin curves are weak second-order air.
polynomials, which could easily be approximated by
linear equations. In general, the Fishbeyn and Some promising results have been derived from the
Pervyshin corrections are more conservative than those comparison of data published by Rolls Royce for their
by Samuels and Gale. Avon engine and predictions made by Gu and Palmer
(1985). Differences ranged from being negligible for
From Figures 1 to 4, both methods showed that all compressor speeds to about 0.8% for thrust corrections.
corrections had the same trend for a given parameter and
very good agreement existed between the two 5.0 SOME CURRENT CORRECTIONS
developments. At a reasonably high absolute humidity of
0.0223 kg water vapour/kg dry air (300 K, 100% RH), Some experimental validation of these studies was sought
differences in correction factors for the primary from available corrections from suppliers and users of

Downloaded from https://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1991/78996/V002T02A025/2400497/v002t02a025-91-gt-329.pdf by guest on 23 March 2020

parameters were at most 0.15%, and 0.2 % for the current gas turbines. These corrections were mostly
specific fuel consumption correction derived from those taken from test cell operation instructions and from
of the thrust and fuel flow. engine manuals. It is expected that most corrections
are derived from analysis, supported by test data. A
Although the Samuels and Gale method has more inherent list of these sources, together with abbreviation codes
limitations, the approach is clear and understandable used in the text are given in the references. The
and was experimentally validated. The good agreement corrections from Section 3 have been included as a
with the results of Fishbeyn and Pervyshin strengthens reference point for these examinations.
the confidence in the latter's more implicit methods.
The Fishbeyn and Pervyshin results are preferred for The "moist" over "dry" correction presentation by
general use since the corrections were generated from Fishbeyn and Pervyshin has been chosen for this
engine cycles with bypass ratios from 0 to 4, i.e. from discussion. An initial examination of the various data
turbojets and turbofans, and for fairly wide ranges of showed that corrections presented could be considered
power setting and ambient conditions. linear functions of the absolute humidity (Grabe and
Bird 1988). Therefore, corrections could be simply
If possible, the given humidity corrections should be presented (Table 1).
verified with experimental data, particularly for
ambient temperatures of less than 288 K. No data have The actual results are discussed parameter by parameter
been found which validates these methods with test data with reference to Figures 1 to 4.
for modern engines.
5.1 Compressor Speed
4.3 Discussion of Humidity Corrections
Compressor speed was the most common engine parameter to
Humidity corrections can therefore be made in two ways. be corrected for humidity. All correction factors, but
The more fundamental approach would first estimate of one, had the same trend as Fishbeyn and Pervyshin or
the changes in thermodynamic properties of the air and Samuels and Gale, i.e. the compressor speed would be
gases. The moist properties would then be used in the higher in moist air than in dry (Figure 1). For a
analysis of experimental data or in cycle calculations. representative test condition of 293 K and 80% RH, which
The computer code provided by Fessler (1979) has been yields an absolute humidity of 0.0118 kg water/kg dry
verified against tabulated values for air pressures of air, the various sources would suggest a speed
0.05 to 300 MN/m 2 and temperatures of 155 to 1473 K. correction to be 0.3 ± 0.1 percent.
Experimental data were not available for validation but
the methods were to be used for test purposes. No The sole exception was the correction given for the J79
guidance is given for the estimation of the properties (Technical Manual 33D4-6-261-1) which suggests a drop in
of moist combustion gases. compressor speed. The correction was intended for use
in collapsing test cell correlation results when plotted
Alternatively, accepting the thermodynamic property as a function of engine pressure ratio. Since this is
models for moist air (Section 4.1), the work of Samuels the same apparent application for the J79 corrections
and Gale and Fishbeyn and Pervyshin can be applied given in Technical Manual 2J-Jl-10, the reasons for the
directly to engine performance results with moist air. differences are not evident.
Returning to the example of Section 3.2, the humidity
limit of 297 K and 75 percent relative humidity (0.0143 It will be noted that for multi-spool engines, the
kg water/kg dry air or 100 grains water/lbm dry air), available data for PW100 and RB211 engines showed
yields the following corrections: decreasing speed corrections for each spool of higher
pressure.
• Compressor Speed: + 0.37%
5.2 Airflow
• Airflow: - 0.49%
The general prediction was that with moisture in the
• Fuel Flow: + 0.70% air, the airflow decreases relative to the dry-air
condition, ie. the correction has a negative slope
• Thrust: - 0.12% (Table 1). Only two agencies presented corrections for
airflow. General Electric gave different corrections
• Turbine Inlet Temperature: - 0.12% for their LM2500 stationary gas turbine, depending on
the reference power setting. For constant shaft horse
• Specific Fuel Consumption: + 0.82% power (SHP) and power turbine speed (NPT), the airflow
correction is nearly twice the correction for the
For practical purposes, this set of humidity corrections reference condition with constant power turbine inlet
may constitute limit values, since gas turbine total temperature (T54) and power turbine speed. The
performance testing would not be conducted at high choice of reference condition or independent parameter
absolute humidity to avoid condensation in the inlet. must therefore be made carefully.
Therefore, humidity corrections for most performance
parameters, under most conditions are expected to be 5.3 Fuel Flow
small but sufficiently large for consideration during
high accuracy tests. All predictions were that fuel flow would increase with
moist air (Figure 2). The Pratt & Whitney Canada
While no uncertainties are available for either the corrections for their PW100 turboprop engine were only
Samuels and Gale or Fishbeyn and Pervyshin correction 2/3 of those of Fishbeyn and Pervyshin, and Samuels and
factors, confidence can perhaps be derived from their Gale. Depending on the operating conditions, the
excellent agreement, in trend and magnitude, with each humidity corrections for the LM2500 were the largest
other. The two correction methods agreed to within 0.2 (T54 and NPT constant), as well as the smallest (SHP and
percent for primary performance parameters, at a NPT constant), among those corrections that were
available. The influence of the reference or power
setting conditions on humidity effects should be further b) Two approaches for humidity corrections of major
investigated. engine parameters have been inspected. There are
limitations in the methodologies employed which
It should be noted that for the selected ambient may be validated with experimental or field data.
condition of 293 K/80% RH, the average correction would However, the two derived sets of correction
be 0.5 percent. This is an important adjustment to fuel agreed well and yielded essentially linear
flow data for accurate performance assessment. functional relationships between engine parameter
corrections and absolute humidity.
5.4 Turbine Inlet or Outlet Temperature
C) A limited number of humidity corrections given to
Turbine temperatures, inlet or outlet, are expected to users by suppliers have been examined and
decrease with air humidity (Figure 3). The apparent compared with those provided by Samuels and Gale,
spread among correction curves can largely be attributed and Fishbeyn and Pervyshin. In most cases, good

Downloaded from https://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1991/78996/V002T02A025/2400497/v002t02a025-91-gt-329.pdf by guest on 23 March 2020

to differences in operating or reference modes (LM2500: agreement was found in trend between these
SHP, NPT fixed versus T54, NPT fixed) and gas path humidity corrections and those given by the
station (PW100 HP turbine inlet [T4] versus HP turbine theoretical/empirical methods. Some significant
outlet [T6]). However, at the selected conditions of differences in magnitude were noticed. These may
293 K/80% RH, most corrections are relatively small be the result of derences in:
(less than 0.2 percent). For example, at 1400 K, 0.2%
is only 2.8 K, which would not be detectable with the cycle parameters, e.g. pressure ratio,
uncertainty of most instrumentation. control modes, or
choice of independent comparison parameter.
5.5 Thrust or Power

The thrust or power corrections for humidity presented This limited survey has identified the need for
conflicting results (Figure 4). Two users, in three operational limits on testing in ambient air.
applications, followed the trend given by Fishbeyn & Demonstrated differences in correction factors, in trend
Pervyshin and Samuels & Gale, who predicted that and magnitude, need to be explained by careful cycle
moisture in the inlet air would lead to a reduction in studies with different control or reference parameters;
thrust. The LM2500 data (SHP, NPT fixed) and the cycle pressures and temperatures must be representative
J85/J79 (Tech. Manual 2J-Jl-10), indicated greater of the engines which have published humidity
thrust/power reductions from moisture than Fishbeyn and corrections. Only then could the apparent anomalies in
Pervyshin or Samuels and Gale. The CFM56 corrections trend and magnitude be explained.
were so small as to be virtually non-existent. However,
at 293 K/80% RH, corrections were only of the order of Facilities without air drying capability are likely to
0.15 percent. encounter problems when assessing data because of
humidity effects on engine performance, unless some
Two data sources gave corrections with an opposite accurate correction methods can be agreed upon among
trend, is. thrust/power gains from humidity. One was facilities.
the LM2500 stationary gas turbine with T54, NPT fixed,
and the other one comes from the J79 data set (Tech. It is recommended that humidity correction methods be
Manual 33D4-6-261-1) which also had the compressor speed documented as clearly as ambient atmospheric corrections
corrections with an opposite trend. These LM2500 power to standard conditions are documented now, to avoid the
humidity corrections were significantly larger than the introduction of possible errors when assessing or
J79 ones (0.6% vs. 0.3% at 293 K and 80% RH). comparing test data.
The differences in corrections for the two LM2500 7.0 REFERENCES
reference modes (SHP, NPT fixed and T54, NPT fixed) may
be explained using Table 1. Moist airflow is predicted Identification codes used in the text and figures for
to be relatively greater for the constant temperature certain references are shown in square brackets.
case, and so the power output should be relatively
greater at these higher airflow levels. Air Canada, 1988, "Air Canada Unit Overhaul Manual 901.
Engine Testing - CFM56. Testing 72-00-00 October 1985,
5.6 Specific Fuel Consumption January 1986, October 1988," Air Canada, Dorval,
Quebec. [code CFM56]
Only the LM2500 data set included corrections for
specific-fuel-consumption (SFC). While Fishbeyn and Air Canada, 1988, "Air Canada Unit Overhaul Manual 901.
Pervyshin had given SFC curves in their report (1970), Engine Testing. Effectivity: JT9D-7R4D "A" Package.
the curve for Samuels and Gale was derived from fuel 71-00-00 Testing. June 1984, June 1986, February 1988,"
flow and thrust data (Grabe and Bird, 1988). Compared Air Canada, Dorval, Quebec. [code JT9D]
to these humidity corrections, the LM2500 corrections
for either control mode/ reference condition were small, Air Canada, 1987, "Air Canada Unit Overhaul Manual 901.
about one third of the values estimated by the two RB211-524B4 Test Schedule. Testing 72-00-00. July
theoretical/empirical methods. At 293 K and 80% RH 1981, February 1982, May 1983, November 1983, January
(0.0118 absolute humidity), the humidity correction 1987," Air Canada, Dorval, Quebec. (code RB211]
would be approximately 0.2% for either mode/condition.
This correction would be in agreement with a combination Blake, J.C., 1975, "Effects of Condensation in the JT9D
of the humidity corrections predicted by Samuels and Turbofan Engine Bellmouth Inlet," AIAA/SAE 11th
Gale, and Fishbeyn and Pervyshin, for fuel flow and Propulsion Conference, Anaheim, CA, September 29 -
thrust. October 1, 1975. AIAA Paper no. 75-1325.
Blazey, D.J., 1983, "F404 Engine ESN 376044-1A Test
6.0 SUMMARY AND RECOMMENDATIONS Cell Correlation Results from Lynn Cell 108 and NRC
Ottawa, Canada, Test Cell," General Electric, Lynn
a) The presence of moisture influences gas turbine Mass., Technical Memorandum TM83AEB1159.
testing through:
Fessler, T.E., 1979, "WETAIR - A Computer Code for
condensation resulting primarily from high Calculating Thermodynamic and Transport Properties of
relative humidity which is controlled with Air-Water Mixtures," NASA Technical Paper 1466, National
limitations on testing; and Aeronautics and Space Administration, Lewis Research
Center, Cleveland, Ohio.
changes in performance parameters from the
absolute humidity, which may be corrected
depending on the chosen reference condition
with factors up to one percent for
moderately high humidity conditions.

5
Fishbeyn, B.D. and N.V. Pervyshin, 1970, "Determination Samuels, J.C. and B.M. Gale, 1950, "Effects of Humidity
of the Effect of Atmospheric Humidity on the on Performance of Turbojet Engines," National Advisory
Characteristics of a Turbofan Engine," Translation by Committee for Aeronautics, Washington, D.C., Tech. Note
Foreign Technology Division, Wright-Patterson AFB, OH, 2119. [code Samuels & Gale or S & G]
FTD-HT-23-290-68 (AD 715232). [code Fishbeyn & Pervyshin
or F & P] Spencer, J.H. and Archer, D.C., 1971, "The Effect of
Condensation within an Aircraft Inlet Duct on Installed
Fishbeyn, B.D., 1975, "Air Humidity Influence on Turbofan Engine Performance," 10th National Conference
Experimentally Determined Compressor and Turbine on Environmental Effects on Aircraft and Propulsion
Efficiency," NASA Technical Translation, Washington, Systems, United States Navy, Naval Air Propulsion Test
D.C., NASA-TT-F-17073, July 1976. Centre, Trenton, NJ, 18-20 May 1971.

General Electric Company, 1988, "Commercial Engine Test U.S. Air Force, 1963, "General Engine Test and Technical

Downloaded from https://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1991/78996/V002T02A025/2400497/v002t02a025-91-gt-329.pdf by guest on 23 March 2020

Cell Correlation Procedure," Procedure Design Manual Test Cell Instructions. Turbojet, 2J-J1-10
Engineering Department, Mail Drop G40, 1 Neumann Way, Turbofan Aircraft Engines," T.O. 2J-J1-10, Published
Box 156301, Cincinnati, OH, 45215-6301, October 1988. under Authority of the Secretary of the Air Force, 16
[code GE T/C] September 1963. [code J85/J79)

General Electric Company, 1981, "LM2500-30 Marine Gas U.S. Air Force, 1984, "Operations, Maintenance Technical
Turbine Performance Data," Marine and Industrial Engine Manual and Parts Breakdown. Test Cell Correlation 33D4-
Division, Cincinnati, OH., General Electric MID-TD-2500- 6-261-1 Set," T.O. 33D4-6-261-1. 1C3954G1 (General
8, July 1981. Electric), Published under Authority of the Secretary of
[SHP & NPT constant: code LM2500- SHP, NPT] the Air Force, 1 December 1966 (Change 1 - 30 April
[T54 & NPT constant: code LM2500- T54, NPT] 1984). [code J79]

Grabe, W.K. and Bird, J.W., 1988, "Humidity Effects on Zaba, T. and Lombardi, P., 1984, "Experience in the
Gas Turbine Performance," National Research Council of Operation of Air Filters in Gas turbine Installations,"
Canada, Institute for Mechanical Engineering, Ottawa ASME paper 84-GT-39.
Ontario, Technical Report TR-ENG-003.
Zerkle, R.D., Colley, W.C. and Doel, D.L., 1982,
Gu, Y-G and Palmer, J.R., 1985, "A Mathematical Model "Analysis of Moisture Condensation in Engine Inlet
for Computing the Effects of Air Humidity, Fuel Ducts," AIAA/ASME Joint Fluids, Plasma, Thermophysics
Composition and Gas Dissociation on Gas Turbine and Heat Transfer Conference, St. Louis, Mo., 7-11 June
Performance and Its Application," ASME paper 85-IGT- 1982, pp. 101-112.
114. Presented at the 1985 Beijing International Gas
Turbine Symposium, Beijing, China, Sept. 1-7, 1985. Zimmerman, O.T. and Lavine, I., 1964, Psychrometric
Tables and Charts, 2nd ed., Industrial Research
Pratt & Whitney Canada, 1989, "PW100 Test Bed PW100 Services, Inc., Dover, NH.
Correction Program Humidity Effects," Private
Communication with Paul R. Nutt, 31 January 1989. [code
PW100]

Table 1: Comparison of Correction Factors: Theoretical/Empirical Methods and

User/Operator Data

Data Source Change in Corrected Performance Parameter

per unit absolute humidity
(percent per kg water/kg of dry air)

Compressor Airflow Fuel Flow Thrust Turbine Temp.

Speed Inlet Outlet

Samuels 24.6 -32.3 57.1 -10.2 n/a -9.5

& Gale
Fishbeyn & 25.5 -34.7 50.0 -8.3 -9.0 n/a
Pervyshin

GE T/C 30.8 n/a n/a n/a n/a n/a

J85/J79 23.6 n/a 48.8 -14.7 n/a n/a

J79 -14.7 n/a 49.0 23.8 n/a -12.5

LM 2500
SHP,NPT 13.0 -39.2 27.0 -18.5 n/a -15.0
T54,NPT n/a -20.8 68.0 53.2 n/a -33.9

CFM56 30.8 -29.8 48.8 -1.9 n/a -9.0

JT9D n/a n/a 40.8 n/a n/a n/a

RB211 N1: 27.3 n/a 49.0 n/a n/a -9.9

N2: 24.3
N3: 20.8
PW100 NL: 32.3 n/a 32.6 n/a -23.8 -10.8
NH: 24.3

Note: Sources for data are as specified in the

references (Section 7.0) with the data source
codes above.

Sample calculation of a parameter correction using Table

1:

Let the relative humidity be 80% and the air

temperature 293 K (20°C). This yields an
absolute humidity of 0.0118 kg water/kg dry air.
For a compressor speed correction, obtain from
Table 1 the factor "24.6" (using Samuels and
Gale's correction).

Hence, the speed correction would be: 24.6

0.0118 = 0.3% or

Downloaded from https://asmedigitalcollection.asme.org/GT/proceedings-pdf/GT1991/78996/V002T02A025/2400497/v002t02a025-91-gt-329.pdf by guest on 23 March 2020

N(moist) = 1.003 * N (dry)
which agrees with the correction of Fig.1.

F&P S&G
CFM56 F&P
S&G
1.010 GE T/C
1.020 /
N / ^0 moist
W,/64 moiat
N l ^0 dry N1 / flB211 N2 W,/8Jfi dry
RB211

J]9
PW100 NL 1.015
LM2500 T54/NPT "" /
/ CFM56
1.006
J85/J79
:' HB211 N3
1.010 JT9D /''

1.004
JBSIJ79

PW100 NH

,= PW100
1.005 ,^ /
1.002 LM2500 SHP/NPT

/ ^' LM2500 SHP/NPT

(6UF )
80%RH
•^ / 20'C ' 168 - F)%AH
/ 80
1.000
0 0010 0.020 0.030 1.000
0.040 0.010 0.020 0.030 0.040
ABSOLUTE HUMIDITY, H kg Water / kg Dry Air ABSOLUTE HUMIDITY, H kg Water / kg Dry Air
FIG. 1: COMPARISON OF COMPRESSOR SPEED CORRECTION FACTORS FOR ABSOLUTE HUMIDITY
FIG. 2: COMPARISON OF FUEL FLOW CORRECTION FACTORS FOR ABSOLUTE HUMIDITY

LM2500 T54/NPT
1.0001 1.010 X/6 moist / /
r \
XIS dry /
/ 9

^-^. ... ...^.^ CFM56

1.005

10.1,

T T IO moist

0.996 T/e dry RB211

-' ABSOLUTE HUMIDITY, H

I
kg Water / kg Dry Air
PW100
0.010 0.020 0.030 0.040
1.000 -
LM2500 LM2500
0 1% NG/NPT PW100 SHP/NPT
T4
0.994
20'C
F)
80%RH
S&G
0.995 80%RH
0 0.010 0.020 0.030 0.040
LM2500 J85/J]9
ABSOLUTE HUMIDITY, H kg Water / kg Dry Air NG/NPT

FIG. 3: COMPARISON OF TURBINE INLET OR OUTLET TEMPERATURE CORRECTION FACTORS FOR FIG. 4: COMPARISON OF THRUST OR POWER CORRECTION FACTORS FOR ABSOLUTE HUMIDITY
ABSOLUTE HUMIDITY