Performance Det in GTs
Performance Det in GTs
Performance Det in GTs
]^L
91-GT-228
The Society shall not be responsible for statements or opinions advanced in papers or in 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 in an ASME Journal. Papers are available
from ASME for fifteen months after the meeting.
Printed in USA.
ABSTRACT
This paper describes the most important factors affecting the industrial gas
turbine engine performance deterioration with service time and provides
some approximate data on the prediction of the rate of deterioration.
Recommendations are made on how to detect and monitor the performance
performance deterioration will remain and perhaps get worse with time.
Permanent Performance Deterioration
During an engine overhaul the flow path components are thoroughly
cleaned, damaged parts are replaced or damaged areas "blended out", tip
and seal clearances are restored to "as new" condition, any obvious leakage
paths are sealed up, the pertinent airfoils recoated, etc. These actions
ensure that the engine is restored as closely as possible to the "as new and
clean" condition. After the completion of a major overhaul the engine
performance would theoretically be expected to be as per the initial
Presented at the International Gas Turbine and Aeroengine Congress and Exposition
Orlando, FL June 3-6, 1991
This paper has been accepted for publication in the Transactions of the ASME
Discussion of it will be accepted at ASME Headquarters until September 30, 1991
performance acceptance test. However, it is usually the case that the engine
performance is not restored to the "as new" condition because of cylinder
distortion (and hence eccentricity in clearances and increased leakage
paths), increased surface roughness of flow path components (due to
erosion or rust scale deposits on compressor discs and annulus surfaces),
distortion in platforms causing loss of aerodynamic performance and
increased leakage, airfoil untwist, etc. Fortunately, under normal
circumstances the unrecoverable performance deterioration is small.
FACTORS AFFECTING PERFORMANCE DETERIORATION
Contaminants
Air borne contaminants are listed below:
1. Hard Particles (causing erosion and fouling): dust, dirt, sand, rust, ash,
carbon particles.
2. Soft Particles (causing fouling): oil, unburnt hydrocarbons, soot, air
borne industrial chemicals, fertilizers, herbicides, insecticides, pollen,
plant spores, air borne insects, air borne salts.
About 80% of particulate contaminants are below 2m in diameter. If an
evaporative cooler is used, then water borne contaminants, such as dirt and
salts, may be introduced. These may result in fouling and corrosion.
Contaminants affecting the hot end include some of those listed above, plus
fuel borne contaminants, those produced as a result of the combustion
process, contaminants in water or steam injected for NO, control, and any
particles, such as rust scale, etc., introduced through the cooling air system
piping.
Fouling
Fouling is defined as degradation of flow capacity and efficiency caused by
adherence of particular contaminants to the gas turbine engine airfoil and
annulus surfaces. The result of fouling is build up of material which changes
the shape of the airfoil, changes the airfoil inlet angle, increases surface
roughness, and reduces the airfoil throat opening. The end result is reduced
component performance.
Fouling, which can normally be eliminated by cleaning, can occur in both the
compressor and the turbine. Compressor fouling is more common and has
the more serious effect on engine performance. Typically about 70 to 85%
of all gas turbine engine performance loss accumulated during operation is
attributable to compressorfouling. The contaminants that cause compressor
fouling enter the compressor with the inlet air. Those that cause turbine
fouling enterthe turbine with the inlet air, cooling air, fuel, fuel additives, and
water/steam used for NO, control.
All compressors are susceptible to fouling and the degree of fouling, the rate
of fouling, andthe effect on performance depend on the following: compressor
design, compressor airfoil loading, airfoil incidences, airfoil surface
smoothness/coating, type and condition of the air borne contaminants, the
site environment, and the climatic conditions (high humidity increases the
rate of fouling) (Seddigh, Saravanamuttoo, 1990). Compressor fouling not
only reduces the inlet flow and compressor efficiency, but also reduces the
compressor surge margin (Mezheritsky and Sudarev, 1990) and may result
in compressor surge. Turbine fouling will depend on the ability of the
contaminants to reach the turbine and to adhere to the gas path surfaces.
Operation of a gas turbine engine on residual fuel oil, heavy oil, or crude oil
will have a corrosive effect on the hot end gas path components. The
corrosion plus deposits on the flow path surfaces will result in engine
performance deterioration.
Types of Filters
The function of gas turbine inlet filters is to remove particles, such as dirt,
dust, soot, etc., from the compressor inlet flow and prevent compressor
fouling, erosion and corrosion. The three main types of filters are listed
below:
1. Inertial Filter - This is the simplest and least expensive type of filter and
it functions by centrifuging out particles from the inlet flow. The
accumulated dirt is then removed (continuous 5 to 10% bleed). The
velocity in the filter has to be fairly high to make it work. This filter is
effective in removing particles above 20m diameter. Thus, this type of
filter is good in preventing erosion due to the presence of large dust or
dirt particles in the inlet air. But it is not very effective in removing
particles less than 20m in diameter, and hence it is not very efficient in
preventing compressorfouling. Compressor fouling is generally caused
by particles with diameter of 5m or less. The pressure loss inthe inertial
filter is about 1 in. (25 mm) H 2 O, and it does not increase with service.
airfoil surface coatings and blockage of turbine airfoil cooling holes. Dry
abrasive cleaning is usually not recommended for coated compressor
airfoils. Figure 1 shows the effect of progressive fouling on the mass flow
of an industrial gas turbine, as well as the results of on line dry cleaning
and hand cleaning of the front components of the compressor.
Washing - Water or water mixed with detergents is injected into the
compressor inlet when the engine is on line. The water used in this
operation must be clean and free from such impurities as sulphates,
carbonates and chlorides, which will be deposited as scale in the rear
compressor stages where the airtemperature is above the boiling point.
The contaminants adhering to the compressor surfaces are 'washed"
away or dissolved and removed. Although sometimes the contaminants
removed from the front stages may be deposited on the downstream
stages, this is a very effective cleaning method. Even more effective
method is to soak wash when the engine is shut down.
The selection of the most effective compressor cleaning method (whether
dry cleaning or water/detergent washing or both) depends on the type and
condition of the contamination. Dry cleaning is usually most efficient if the
compressor is fouled by dry hard deposits. Contamination by oil or salt
deposits is best taken care of by washing. Experimentation and good record
keeping may be required to arrive at the optimum cleaning procedure.
start-up, the transient differential expansion rates between the large mass
stationary components and the lighter mass rotor assembly may result in
blade tip and seal rubs. The increased clearances will result in increased
leakage and hence in performance deterioration. Operation at peak rating
will have similar results, as described above, with the exception of rubs due
to running through criticals, and will cause performance deterioration.
Frequent emergency trips from base load will also have a negative effect on
engine performance. Therefore, an engine which is subjected to many
starting and emergency trip cycles and/or is operated for considerable
periods of time at peak rating will experience a, more severe performance
degradation than an engine which is operated at or below base load.
Maintenance practices of a gas turbine engine have a considerable impact
on the rate of engine performance deterioration. The control system must be
properly maintaned to ensure correct fuel scheduling during starting and
normal operation, in order to prevent excessive temperature gradients. The
fuel system and the water/steam injection systems must be maintained
properly to prevent contaminants from getting into the engine. Proper
maintenance of the combustion system is required to ensure low
circumferential and radial temperature gradients. It is very important that the
engine be operated within the specified safe operating envelope. Regular
inspection of the inlet filter system and regular compressor cleaning should
be instituted. Following the correct engine maintenance practices will result
in reduced rate of performance deterioration.
PERFORMANCE DETERIORATION
Inlet
Fouling of high efficiency filters occurs progressively with time. The blockage
of the filter pads results in increased inlet loss and hence in engine
performance deterioration. This performance loss is recovered completely
when new filter pads are installed. The performance deterioration for an
industrial gas turbine engine with increasing inlet loss is shown in Figure 2.
For a 6 in. (150 mm) H 2 O inlet loss increase the engine output power
decreases by about 2.6% and the heat rate increases by about 1%.
Compressor
Some information available from open literature indicates that fouling causing
a 5% reduction in inlet flow will also reduce the compressor efficiency by
about 2.5% (Saravanamuttoo and Lakshminarasimha, 1985). The resulting
decrease in engine power will be about 10%. Usually the fouling trend is
assumed to have a linear characteristic with time. Experimental tests on the
compressor of a small turboprop gas turbine engine indicated that the prime
effect of fouling is on the inlet air flow rather than on compressor efficiency,
and that the reduction in flow varies with operating speed (being higher at
design speed).
Site test data obtained on a large industrial gas turbine indicated that
compressor fouling resulted in a 5% reduction in inlet mass flow and 1.8%
reduction in compressor efficiency. This amount of fouling would reduce the
engine output power by about 7% and increase the heat rate by about 2.5%.
Bleed valve leakage, flange and horizontal joint leakage, and the improper
positioning of the inlet guide vanes will have a deleterious effect on the
compressor performance. The compressor control bleed valves have a
tendency to bind in the open position orto leak during full load operation. The
latter is the more likely occurrence, which will get progressively worse with
time and result in a performance penalty. The improper stagger of the inlet
guide vanes will also result in a decrease in compressor performance.
The performance deterioration due to "nicks", tip and seal rubs, erosion,
damage, etc., may be significant in some cases, but usually it is considerably
less than that due to compressor fouling. It should be pointed out that
compressorfouling and other compressor performance deterioration causes
have no effect on engine firing temperature.
Hot End
Under "hot end" are included: combustion system, turbine, and exhaust
diffuser. The combustion system is not likely to be the direct cause of
performance deterioration with time. Irrespective of the fuel used (whether
natural gas, distillate oil, or even crude oil) and even if the fuel nozzles are
coated with carbon deposits, the combustion efficiency will not decrease.
However, carbon deposits breaking off from the nozzles and soot produced
as a result of incomplete fuel burning will result in performance deterioration.
Also, changes in the combustor outlet Pattern Factor may result in a
temporary or a permanent deformation of downstream components, which
in turn will result in performance deterioration. Hot end cylinder warping,
which will increase with time, may result in flange and horizontal joint
leakage, and hence in performance deterioration.
Fouling of the turbine airfoils and annulus, surface erosion, "nicks" and
deformation, blade tip and seal land rubs, and increased leakage and
cooling flows will result in performance deterioration. Experiments carried
out on a multi-stage axial turbine showed that both reduction in airfoil profile
thickness (as caused by erosion) and increase in airfoil profile thickness
(resulting from surface deposits) result in a significant performance reduction
(Bammert and Stobbe, 1970). Any changes in the exhaust diffuser
performance can be included in the overall turbine efficiency changes. A
decrease of 1% in the overall turbine efficiency in an engine such as the
CW251 B12 will result in about 2.5% decrease in output power and a similar
increase in heat rate. In addition, because of a decrease in the turbine
temperature drop, the engine will be underfired by about 12F (6.7C) (note
that the engine is controlled to a specified exhaust temperature). This will
result in a further 1.2% decrease in output power and about .2`/o increase in
heat rate. Thus the cumulative result of a 1% decrease in turbine efficiency
with time is about 3.7% decrease in output power and about 2.8% increase
in heat rate.
The components included under this heading will also suffer performance
deterioration with time and as a result will affect the overall engine
performance. It is more difficult to detect the performance losses in these
components or to predict their rate of increase with time. The exception is
the exhaust system loss. The increase in exhaust loss can be measured and
steps can be taken to remedy the situation if the loss becomes excessive.
Figure 2 shows the effect of exhaust loss increase on the CW251 B12 engine
performance. It should be noted that an increase in exhaust loss will result
in a reduction in the firing temperature and hence a further reduction in
engine performance. Thus a 6 in. (150 mm) H 2 O increase in exhaust loss will
result in about 4.2F (2.3C) decrease in the firing temperature. Hence the
total engine performance deterioration will be a 1.4% decrease in power and
1.1% increase in heat rate.
Overall Plant
conditions (Scheperet al, 1978). For instance, a 2 in. (50 mm) H 2 O increase
in inlet loss, a 5% decrease in inlet mass flow, a 1.8% decrease in
compressor efficiency, and a .5% decrease in turbine efficiency, will result
in a 10% decrease in output power and a 4.2% increase in heat rate.
The control system can also affect the perceived performance deterioration.
If the control system is not properly adjusted, the engine may be operated
belowthe desired base load setting. Thiswill lead to decreased output power
and higher heat rate.
Compressor Fouling
Compressor fouling, and hence engine performance deterioration, can be
detected using the following methods:
1. Change in Combustion Shell Pressure - The combustion shell pressure
is one of the parameters used by the control system to maintain the
engine at base load. Therefore, it is a readily available and accurate
when the compressor is "new and clean". In order to get a large enough
variation of inlet speed, the data may have to be recorded over a length
of time, in which there is a significant ambient temperature variation. The
inlet cylinder static pressure ratio obtained at any time in the future is
then plotted on this graph. If the new points fall below the "clean" curve,
then the compressor is fouled. The reduction in compressor inlet flow is
proportional to the square root of the reduction in the inlet cylinder static
pressure ratio.
3. Inlet Scroll Calibration Curve - The compressor inlet scroll calibration
curve is obtained during a shop test of the engine when accurate
measurements are made of the inlet flow and the static pressures at the
compressor inlet flange and just upstream of the inlet guide vanes.
Figure 5 shows a typical inlet scroll calibration curve, which is the nondimensional inlet mass flow plotted versus the inlet cylinder static
pressure ratio. Occasionally there is an inconsistency between shop
and site inlet flow measurements, even when the compressor is "new
and clean", I. e. there is a difference in the inlet flow, for the same N /T,
between shop and site. This may be explained by the differences in the
inlet ducting, pressure tap, and pressure line installations between shop
and site. To be on the safe side, a new "clean" curve should be obtained
from the site data as follows: As soon as possible after engine
commissioning, when the compressor is clean, data points should be
obtained over a wide ambient temperature range. With the compressor
pressure ratio and non-dimensional speed determined from site data
obtain the expected non-dimensional inlet massflowfrom the compressor
map (see Figure 6, derived from shop tests) at the different ambient
temperature base load operating conditions. Then plot the nondimensional inlet mass flow versus the site measured inlet cylinder
pressure ratio. A smooth curve should be drawn through these points
parallel to the original shop inlet scroll calibration curve. Then, as the
engine operating hours increase, with the site measured inlet cylinder
pressure ratio the inlet mass flow is read from the calibration curve
(preferably the one established at site). This mass flow is then compared
to the "clean" compressor mass flow from Figure 6 to determine the
amount of compressor fouling that has occurred.
With the aid of the shop test derived compressor map it is also possible to
determine the amount of compressor efficiency decrease that has taken
place with operating time. A graph showing the base load compressor
efficiency variation versus non-dimensional speed can be used for this
purpose (see Figure 7). The compressor efficiency may be estimated
approximately from the overall pressure ratio (combustion shell pressure
divided by the ambient pressure, or preferably the compressor inlet pressure)
and the compressor temperature rise from inlet to the combustor shell. The
compressor efficiency estimated in this manner will be lower than the actual
one because of the probable inclusion of inlet duct loss, inaccurate compressor
exit pressure, and temperature pick-up across the combustoroutlet transition
ducts. However, as with the compressor inlet mass flow, efficiency data can
be collected when the compressor is "new and clean" and can subsequently
be used for comparison when compressor fouling has occurred.
Engine Performance
The extent of compressorfouling and the otherdeleterioug effects that occur
in an engine with increasing operating time can be determined by periodically
monitoring the engine output power and comparing it to the expected value,
orto the value when the enginewas "new and clean". Thistype of monitoring
is more complicated than just detecting compressor fouling. The simplicity
or complication may range from the measurement of just a few key
performance parameters to a full blown on line health monitoring system. To
detect the power shortfall some or most of the following parameters must be
measured: output power, auxiliary power, power factor, speed, inlet
temperature, relative humidity, barometric pressure, exhaust temperature,
combustor shell pressure, inlet loss, exhaust loss, water/steam injection
rate, and instrumentation bleed air, if any. With the above information and
with the aid of the correction curves for the deviation of site conditions from
the standard or guarantee power the site output power is corrected to
standard conditions and compared to the "clean"engine power. The amount
of the power shortfall from the expected value will indicate the extent of
compressor fouling or other problems causing performance deterioration.
A further complication could be introduced by measuring fuel flow, fuel
temperature, fuel Lower Heating Value, etc., so as to be able to determine
the engine heat rate and compare it to the "clean" value. For this exercise
similar correction curves, as described above, will have to be used.
4.
The resulting yearly power shortfall = .03 x 46500 x 8000 = 11.2 x10
At one installation site test data indicated that up to 8% power loss occurred
over a two week period. Most of this power shortfall could be recovered by
cleaning the compressor with new catalyst. At another installation, after
incorporation of coated compressor airfoils, the fouling rate was reduced
considerably and on line washing of the compressor recovered all of the
power shortfall. At a third installation, after six months of continuous
operation without compressor cleaning, the engine experienced a 4.3% loss
in power. A series of on line, at reduced load, water washes recovered all
but .5% of the power shortfall. It should be pointed out that this engine had
a good inlet filtration system and coated compressor stator airfoils.
6.
5. Good filtration system used, clean operating environment (no oil leaks,
no corrosive chemicals, no sand), no major foreign object damage,
coated compressor airfoils.
3.
KW.HR .
crude or residual oil fuel. The residual performance decrement is higher for
crude/residual oil fuel because of the more severe deleterious effect of
these fuels on the hot end, compared to natural gas or distillate oil. Figure
9 shows the typical performance deterioration (non recoverable with
frequent cleaning) with operating time for the three cases. After the
overhaul the engine performance deterioration will experience a similar
rate as when it was new, that is the shape of the curve will be the same as
in the first 24,000 operating hours, but its starting point will be the residual
performance decrement.
RECOMMENDATIONS
Based on the considerations described in this report the following
recommendations are made to obtain minimum engine performance
deterioration rate and to achieve optimum engine performance over its
entire service life:
1. To retard the rate of compressor fouling the engine site installation
should incorporate the optimum inlet filtration system for the particular
site environmental conditions.
2. Smooth and/or coated compressor, and if possible turbine, airfoils
should be used.
3. Clean fuels and injected steam/water should be used.
4. The official site performance acceptance test should be carried out
immediately after engine commissioning and initial synchronization in
order to obtain reliable "bench mark" engine performance and health
monitoring type of information when the engine is in the "new and clean"
condition.
5. Engine performance deterioration and compressor fouling should be
monitored regularly and recorded.
6. Appropriate and regular compressor, and if necessary turbine, cleaning
procedure should be used.
7. Good operating and maintenance procedures should be employed.
8. The control system should be carefully monitored and adjusted as
requiredto eliminate possibilityof overtemperaturingon starts, operation
above base load firing temperature for prolonged periods, and spurious
engine trips.
9. Any required remedial actions as regards engine performance or
mechanical integrity should be executed promptly.
10. During a major overhaul the engine should be restored as closely as
possible to its original "new and clean" condition.
CONCLUSIONS
1. The most important factors affecting industrial gas turbine engine
performance deterioration with time were discussed.
2.
3.
4.
.28
1.05
26
S
S
.95
AVERAGE
FOULING
.24
90
UNFOULDED
COMPRESSOR
CLEANED
ON
LINE
SHUTDOWN
HAND
CLEANED
CLEANED
ON
LINE
.85
.22
0
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
DAYS
.20
.18
N = ROTATIONAL SPEED, RPM
1.05
INLET LOSSES
0
H
U
LL
ULU
8 = Tw R x GAS CONSTANT
540xGAS CONSTANT
.16
HEAT RATE
5000
1.0
5200
5400
N
EXHAUST FLOW
cr
5800
5600
6000
RPM
POWER
0
U
/J,
0.95
1.05
0
F
U
LIw
LL
Q to
H
0
D=
0
U
0.95
15
10
0
C
20
.10
.12
.14
.16
.18
.20
.22
.24
240
220
S
a
200
10
m
w
a
w
180
I
N
0=
I-
160
0
O
U
140
-20
20
40
60
80
100
120
AMBIENT TEMPERATURE. F
26
90
88
>: 86
0
z
W
U
UU-
Ui
84
0
LU
LU
w
Q
a
p 82
0
AFTER OVERHAUL
80
78
5,000
5,200
5,600
5,400
5,800
6,000
RPM
REFERRED SPEED
AFTER OVERHAUL
00
10
15
20
a
25
I^
w
w
C)
w
0
w
0
Uw
w
B-
II II
I
I i
FOULDED
/ I I
COMPRESSOR
EFFECT OF
CLEANING/
WASHING
BEFORE
OVERHAUL
AFTER
OVERHAUL
NON-RECOVERABLE WITH
CLEANING/WASHING
MAJOR
OVERHAUL
SERVICE HOURS