View metadata, citation and similar papers at core.ac.

uk brought to you by CORE

provided by Texas A&M Repository

Proceedings of the Forty-Second Turbomachinery Symposium

October 1- 3, 2013, Houston, Texas

GAS TURBINE PERFORMANCE AND MAINTENANCE

Rainer Kurz Klaus Brun

Manager, Systems Analysis Director, Machinery Program
Solar Turbines Incorporated Southwest Research Institute®
San Diego, CA, USA San Antonio, TX, USA

Cyrus Meher-Homji Jeff Moore

Bechtel Fellow Manager, Rotating Machinery Dynamics
Bechtel Corporation Southwest Research Institute®
Houston, TX, USA San Antonio, TX, USA

Francisco Gonzalez
Principal Rotating Equipment Engineer
Enterprise Products
Houston, Tx, USA

Rainer Kurz is the Manager, Journal. He is the chair of the ASME-IGTI Board of
Systems Analysis, at Solar Directors and the past Chairman of the ASME Oil &
Turbines Incorporated in San Gas Applications Committee. He is also a member of
Diego, California. His the API 616 Task Forces, the Fan Conference
organization is responsible for Advisory Committee, and the Latin American
analyzing compression require- Turbomachinery Conference Advisory Committee. Dr.
ments, predicting compressor and Brun is an editor of Global Gas Turbine News,
gas turbine performance, for Executive Correspondent of Turbomachinery
conducting application studies, International Magazine, and an Associate Editor of
and for field performance testing. Dr. Kurz attended the ASME Journal of Gas Turbines for Power.
the Universitaet der Bundeswehr in Hamburg,
Cyrus Meher-Homji is an
Germany, where he received the degree of a Dr.-Ing.
Engineering Fellow and Senior
in 1991. He has authored numerous publications
Principal Engineer at Bechtel
about turbomachinery related topics, is an ASME
Corporation assigned to the
fellow, and a member of the Turbomachinery
LNG Technology Center of
Symposium Advisory Committee.
Excellence as a turbomachinery
Klaus Brun is the Director of the advisor to ongoing LNG projects
Machinery Program at Southwest on the aeromechanical design,
Research Institute. His experience selection, and testing of large
includes positions in engineering, compressors and gas turbines. His 32 years of
project management, and industrial experience cover gas turbine and
management at Solar Turbines, compressor application, design, and troubleshooting.
General Electric, and Alstom. He Cyrus is a registered Professional Engineer in the
holds four patents, authored over State of Texas, a Fellow of ASME, and he is active in
100 papers, and published a several committees of ASME's International Gas
textbook on gas turbines. Dr. Brun won an R&D 100 Turbine Institute. He has a Master's Degree in
award in 2007 for his Semi-Active Valve invention and Engineering from Texas A&M University and an MBA
ASME Oil Gas Committee Best Paper awards in 1998, from the University of Houston. He is a member of the
2000, 2005, 2009, 2010, and 2012. He was chosen to Texas A&M University Turbomachinery Symposium
the "40 under 40" by the San Antonio Business Advisory Committee.

1 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 Jeffrey Moore is the Manager of Mr. Gonzalez has coauthored several technical papers
the Rotating Machinery for Turbo Machinery Symposium and ASME Power
Dynamics Section at Southwest Gen as well as articles on Improving Reliability in
Research Institute in San various publications. Mr. Gonzalez graduated from
Antonio, TX. He holds a B.S., the University of Houston with a Bachelors degree in
M.S., and Ph.D. in Mechanical Mechanical Engineering in 1990
Engineering from Texas A&M
University. His professional ABSTRACT
experience over the last 20 years
Proper maintenance and operating practices can
includes engineering and
significantly affect the level of performance
management responsibilities related to centrifugal
degradation and thus, time between repairs or
compressors and gas turbines at Solar Turbines Inc. in
overhauls of a gas turbine. Understanding of
San Diego, CA, Dresser-Rand in Olean, NY, and
performance characteristics of gas turbines helps
Southwest Research Institute in San Antonio, TX. His
proper applications, as well as driven and process
interests include advanced compression methods,
equipment sizing. Proactive condition monitoring will
rotordynamics, seals and bearings, computational
allow the gas turbine operator to make intelligent
fluid dynamics, finite element analysis, machine
service decisions based on the actual condition of the
design, controls, and aerodynamics. He has authored
gas turbine rather than on fixed and calendar based
over 30 technical papers related to turbomachinery
maintenance intervals. Maintaining inlet air, fuel, and
and has three patents pending. Dr. Moore has held the
lube oil quality will further reduce gas turbine
position of Oil and Gas Committee Chair for IGTI
degradation and deterioration. This tutorial provides a
Turbo Expo and is the Associate Editor for the Journal
discussion on performance characteristics and how
of Tribology. He is also a member of the
performance degradation can be minimized.
Turbomachinery Symposium Advisory Committee, the
Recommendations are provided on how the operator
IFToMM International Rotordynamics Conference
can limit degradation and deterioration of the gas
Committee, and the API 616 and 684 Task Forces.
turbines through proper maintenance practices. The
effects of water-washing and best washing practices
are discussed. Emphasis is on the monitoring of gas
Francisco Gonzalez is a Principal
turbine performance parameters to establish condition
Rotating Equipment Engineer for
based maintenance practices.
Enterprise Products. Enterprise
Products currently operates INTRODUCTION
machinery over 3 MM installed
How Does a Gas Turbine Work?
horsepower in Natural Gas and
NGL Processing Plants and over Explanations of the working principles of a gas
50,000 miles of pipelines turbine have to start with the thermodynamic
transportation various NGLs, and principles of the Brayton cycle, which essentially
Natural Gas, Crude and Refine Products in the USA. defines the requirements for the gas turbine
components. Since the major components of a gas
Mr. Gonzalez has over 22 years of experience in turbine perform based on aerodynamic principles, we
Operations and Maintenance of Rotating Equipment. will explain these, too.
The Reliability Department at Enterprise is primarily
responsible for providing technical support to
operations and maintenance on all rotating
equipment. The Reliability Department provides
services to operation and maintenance which include
rotating equipment vibration monitoring, aerothermal
performance monitoring, troubleshooting, overhaul,
performance testing and gas turbine and compressor
optimization as well as evaluating short term and long
term reliability improvements for all rotating
equipment. Enterprise currently operates over 385 gas
Figure 1: Brayton Cycle
turbine compressor and pump packages, and over 450
reciprocating compressor packages. The Brayton or gas turbine cycle (Figure 1)
involves compression of air (or another working gas),
the subsequent heating of this gas (either by injecting

2 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

and burning a fuel or by indirectly heating the gas) derived from aircraft engines sometimes have two
without a change in pressure, followed by the compressor sections (the HP and the LP
expansion of the hot, pressurized gas. The compressor), each driven by a separate turbine
compression process consumes power, while the section (the LP compressor is driven by an LP
expansion process extracts power from the gas. Some turbine by a shaft that rotates concentric within
of the power from the expansion process can be used the shaft that is used for the HP turbine to drive
to drive the compression process. If the compression the HP compressor), and running at different
and expansion process are performed efficiently speeds. The energy left in the gas after this
enough, the process will produce useable power process is used to drive a power turbine (on a
output. This principle is used for any gas turbine, from third, separate shaft), or the LP shaft is used as
early concepts by F. J. Stolze (in 1899), C.G. Curtiss output shaft.
(in 1895), S. Moss (in 1900), Lemale and Armengaud
The energy conversion from mechanical work
(in 1901), to today’s jet engines and industrial gas
into the gas (in the compressor) and from energy in the
turbines (Meher-Homji, 2000). The process is thus
gas back to mechanical energy (in the turbine) is
substantially different from a steam turbine (Rankine)
performed by the means of aerodynamics, by
cycle that does not require the compression process,
appropriately manipulating gas flows. Leonard Euler
but derives the pressure increase from external
(in 1754) equated the torque produced by a turbine
heating. The process is similar to processes used in
wheel to the change of circumferential momentum of a
Diesel or Otto reciprocating engines that also involve
working fluid passing through the wheel. Somewhat
compression, combustion, and expansion. However, in
earlier (in 1738), Daniel Bernoulli stated the principle
a reciprocating engine, compression, combustion, and
that (in inviscid, subsonic flow) an increase in flow
expansion occur at the same place (the cylinder), but
velocity is always accompanied by a reduction in static
sequentially, in a gas turbine, they occur in dedicated
pressure and vice versa, as long as no external energy
components, but all at the same time.
is introduced.
The major components of a gas turbine include
the compressor, the combustor, and the turbine.
The compressor (usually an axial flow
compressor, but some smaller gas turbines also use
centrifugal compressors) compresses the air to several
times atmospheric pressure. In the combustor, fuel is
injected into the pressurized air from the compressor
and burned, thus increasing the temperature. In the
turbine section, energy is extracted from the hot
pressurized gas, thus reducing pressure and
temperature. A significant part of the turbine’s energy
(from 50 – 60 percent) is used to power the
compressor, and the remaining power can be used to Figure 2: Typical Industrial Gas Turbine
drive generators or mechanical equipment (gas
compressors and pumps). Industrial gas turbines are
built with a number of different arrangements for the
major components:
• Single-shaft gas turbines have all compressor and
turbine stages running on the same shaft
• Two-shaft gas turbines consist of two sections: the
gas producer (or gas generator) with the gas
turbine compressor, the combustor, and the high
pressure portion of the turbine on one shaft and a
power turbine on a second shaft (Figure 2). In this
configuration, the high pressure or gas producer Figure 3: Velocities in a Typical Compressor Stage.
turbine only drives the compressor, while the low Mechanical Work h Transferred to the Air is
pressure or power turbine, working on a separate Determined by the Change in Circumferential
shaft at speeds independent of the gas producer, Momentum of the Air.
can drive mechanical equipment.
• Multiple spool engines: Industrial gas turbines

3 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 While Euler’s equation applies Newton’s between full load and about 40 or 50% load. In order
principles of action and reaction, Bernoulli’s law is an to accomplish this, the air flow into the combustion
application of the conservation of energy. These two zone has to be manipulated over the load range.
principles explain the energy transfer in a
Gas Turbine Performance Characteristics
turbomachinery stage (Figure 3).
The gas turbine power output is a function of the
The compressed air from the compressor enters speed, the firing temperature, as well as the position of
the gas turbine combustor. Here, the fuel (natural gas, certain secondary control elements, like adjustable
natural gas mixtures, hydrogen mixtures, diesel, compressor vanes, bleed valves, and in rare cases,
kerosene and many others) is injected into the adjustable power turbine vanes. The output is
pressurized air and burns in a continuous flame. The primarily controlled by the amount of fuel injected
flame temperature is usually so high that any direct into the combustor. Most single-shaft gas turbines run
contact between the combustor material and the flame at constant speed when they drive generators. In this
has to be avoided, and the combustor has to be cooled, case, the control system modifies fuel flow (and
using air from the engine compressor. Additional air secondary controls) to keep the speed constant,
from the engine compressor is mixed into the independent of generator load. Higher load will, in
combustion products for further cooling. general, lead to higher firing temperatures. Two-shaft
Another important topic is the combustion process machines are preferably used to drive mechanical
and emissions control: equipment, because being able to vary the power
turbine speed allows for a very elegant way to adjust
Unlike in reciprocating engines, the gas turbine the driven equipment to process conditions. Again, the
combustion is continuous. This has the advantage that power output is controlled by fuel flow (and secondary
the combustion process can be made very efficient, controls), and higher load will lead to higher gas
with very low levels of products of incomplete producer speeds and higher firing temperatures.
combustion like carbon monoxide (CO) or unburned
hydrocarbons (UHC). The other major emissions
component, oxides of nitrogen (NOx), is not related to
combustion efficiency, but strictly to the temperature
levels in the flame (and the amount of nitrogen in the
fuel). The solution to NOx emissions, therefore, lies in
the lowering the flame temperature. Initially, this was
accomplished by injecting massive amounts of steam
or water in the flame zone, thus ‘cooling’ the flame.
This approach has significant drawbacks, not the least
the requirement to provide large amounts ( fuel to
water ratios are approximately around 1) of extremely
clean water. Since the 1990’s, combustion technology
has focused on systems often referred to as dry low
Figure 4: Conventional and Lean-Premix Combustion
NOx combustion, or lean-premix combustion (Figure
Systems
4). The idea behind these systems is to make sure that
the mixture in the flame zone has a surplus of air, Figure 5 shows the influence of ambient pressure
rather than allowing the flame to burn under and ambient temperature on gas turbine power and
stoichiometric conditions. This lean mixture, assuming heat rate. The influence of ambient temperature on gas
the mixing has been done thoroughly, will burn at a turbine performance is very distinct. Any industrial
lower flame temperature and thus produce less NOx. gas turbine in production will produce more power
One of the key requirements is the thorough mixing of when the inlet temperature is lower, and less power
fuel and air before the mixture enters the flame zone. when the ambient temperature gets higher. The rate of
Incomplete mixing will create zones where the change cannot be generalized, and is different for
mixture is stoichiometric (or at least less lean than different gas turbine models. Full-load gas turbine
intended), thus locally creating more NOx. The flame power output is typically limited by the constraints of
temperature has to be carefully managed in a maximum firing temperature and maximum gas
temperature window that minimizes both NOx and CO. producer speed (or, in twin spool engines, by one of
Lean-premix combustion systems allow to keep the the gas producer speeds). Gas turbine efficiency is less
NOx, as well as CO and UHC emissions within impacted by the ambient temperature than the power.
prescribed limits for a wide range of loads, usually

4 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 Figure 5: Performance Characteristics
The air humidity does impact power output, but to 1. For a given TIT, gas turbine specific work
a small degree, (generally, not more than 1 to 3%, increases with pressure ratio, reaching a
even on hot days). The impact of humidity tends to maximum and then decreasing with further
increase at higher ambient conditions. pressure ratio increase.
2. The specific work increases with increasing
Lower ambient pressure (for example, due to
turbine inlet temperature.
higher site elevation) will lead to lower power output,
3. The maximum specific work as the TIT is
but has practically no impact on efficiency. It must be
increased occurs at increasing pressure ratios.
noted that the pressure drop due to the inlet and
exhaust systems impact power and efficiency
negatively with the inlet pressure drop having a more
severe impact.
Gas turbines operated in part load will generally
loose some efficiency. Again, the reduction in
efficiency with part load is very model specific. Most
gas turbines show a very small drop in efficiency for at
least the first 10% of drop in load. In two-shaft
engines, the power turbine speed impacts available
power and efficiency. For any load and ambient
temperature, there is an optimum power turbine speed.
Usually, lowering the load (or increasing the ambient Figure 6: Key Cycle Parameters
temperature) will lower the optimum power turbine
speed. Small deviations from the optimum (by say +/- Pressure ratios and TIT parameters of a wide
10%) have very little impact on power and efficiency. range of industrial gas turbines are shown in Figure 7.

Key cycle parameters for any gas turbine are its

specific work and thermal efficiency which are related
to the cycle pressure ratio and turbine inlet
temperature. The general qualitative relationships
between PR and TIT are indicated in Figure 6 (Meher-
Homji et al., 2009).
The salient relationships of this figure may be
conveniently summarized as follows:

5 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 occasional use of pressurized air to pulse clean the
filter material.
Applications also may have to deal with issues
like heavy rain, cold and freezing temperatures, snow
or sand storms, swarming insects, very high or very
low relative humidity and combinations thereof.
Filtration Mechanisms
Filters in the filtration system use many different
mechanisms to remove particles from the air. The
filter media, fiber size, packing density of the media,
Figure 7: Pressure Ratios and Firing Temperature particle size, and electrostatic charge influence how
Parameters of a Wide Range of Industrial Gas the filter removes particles. Each filter typically has
Turbines various different mechanisms working together to
remove the particles. Four filtration mechanisms are
AIR AND AIR INLET SYSTEMS shown in Figure 8.
Gas turbines ingest a large amount of ambient air The first filtration mechanism is inertial
during operation. Because of this, the quality of the air impaction. This type of filtration is applicable to
entering the turbine is a significant factor in the particles larger than 1 micron in diameter. The inertia
performance and life of the gas turbine. A filtration of the large heavy particles in the flow stream causes
system is used to control the quality of the air by the particles to continue on a straight path as the flow
removing harmful contaminants that are present. The stream moves around a filter fiber. The particulate
system should be selected based on the operational then impacts and is attached to the filter media and
philosophy and goals for the turbine, the contaminants held in place as shown in the top picture of Figure 8.
present in the ambient air, and expected changes in the This type of filtration mechanism is effective in high
contaminants in the future due to temporary emission velocity filtration systems.
sources or seasonal changes.
The next filtration mechanism, diffusion, is
Filters used in Gas turbine applications include effective for very small particles typically less than 0.
• Barrier Filters 5 micron in size. Effectiveness increases with lower
flow velocities (see Figure 8 h, second picture). Small
• Self Cleaning Filters
particles interact with nearby particles and gas
• Vanes/Separators
molecules. Especially in turbulent flow, the path of
• Screens small particles fluctuates randomly about the main
Gas turbine applications typically require stream flow. As these particles diffuse in the flow
filtration systems that use combinations of different stream, they collide with the fiber and are captured.
filter types. Often, the face velocity is used to The smaller a particle and the lower the flow rate
distinguish high velocity and medium/low velocity through the filter media, the higher probability that the
systems (Wilcox, et al., 2011). particle will be captured.
Filter systems have to deal with a number of The next two filtration mechanisms are the most
contaminants, such as large and small particles, as well well known; interception and sieving. Interception
as liquids. Particles can be removed from the air occurs with medium sized particles that are not large
stream by various types of filtration materials. Water enough to leave the flow path due to inertia or not
(and salts dissolved in the water) can be removed small enough to diffuse. The particles will follow the
either by inertial separation (i. e., vane separators), or flow stream where they will touch a fiber in the filter
by filtration materials that repel water. media and be trapped and held. Sieving is the situation
where the space between the filter fibers is smaller
For off-shore applications, where salt water than the particle itself, which causes the particle to be
ingestion is a major concern, two filtration concepts captured and contained.
have emerged: High Velocity Systems and
Medium/Low Velocity System. These systems are also A mechanism not shown in Figure 8 is
often used in coastal areas. For land-based electrostatic charge. This type of filtration is effective
applications, most applications use either barrier filters for particles in the 0.01 to 2 micron size range (Figure
or self cleaning filter systems. Self-cleaning filter 9). The filter works through the attraction of particles
systems are essentially barrier filters that allow the to a charged filter. In gas turbine applications, this

6 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

charge is applied to the filter before installation as a The figure shows the difference between the
result of the manufacturing process. Filters always lose filter’s efficiency curve before and after the charge is
their electrostatic charge over time because the lost. The performance of the filter should be based on
particles captured on their surface occupy charged the discharged condition [Felix, P. C., 1977, Felix, P.
sites, therefore neutralizing their electrostatic charge. C., 1979].
As the charge is lost, the filter efficiency for small
particles will decrease. On the other hand, as the filter
is loaded, the filtration efficiency increases, thus
counteracting the effect of lost charge to some extent.
This will offset some of the loss of filtration efficiency
due to the lost charge. Figure 9 shows a comparison of
a filter’s total efficiency based on the various filtration
mechanisms that are applied.

Figure 10: Comparison of Fractional Efficiency for

Filter Elements from Different Suppliers and Different
Face Velocities in New and Dirty Conditions (Brekke
and Bakken, 2010)
Different filter systems show different
effectiveness for different dry particle sizes (Figure
10). The paragraph above outlines methods to remove
solid particles from the airstream. However, as
outlined earlier, ingestion of liquids is also an issue.
Liquid water can contain dissolved salts and can
penetrate conventional filter materials regardless of
their capability filter our solid particles. At high
humidity, water can actually form in the filter system.
Water can dissolve dirt already captured in the
filtration material and seep through the filter.
Figure 8: Common Filtration Mechanisms (Courtesy Therefore, water removal is a crucial feature for air
Camfil Farr) filtration systems especially in off-shore, marine and
coastal environments. There are two fundamental
methods to remove water: either with vanes upstream
and downstream of the filter system or with special
filter material that cannot be penetrated by water,
typically in the last, high-efficiency, filter stage. Water
removal with vanes favors high air flow velocities, and
thus, high velocity filtration systems. Water removal
with special high efficiency filter material favors
low/medium velocity systems. Both methods, if
applied properly, are capable of preventing liquid
water to enter the gas turbine. However, as can be
determined from Figure 10, high-velocity systems tend
to have a lower filtration efficiency for solid particles
than can be achieved with low/medium velocity
systems.
FUEL AND FUEL TREATMENT
Figure 9: Combination of Filtration Mechanisms to
Industrial gas turbines allow operation with a
Obtain Filter Efficiency at Various Particle Sizes
wide variety of gaseous and liquid fuels, while
maintaining very low emissions. Recent discussions

7 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

include the use of LNG as fuel gas and the operation other energy resources, in reducing pollutant
of gas turbines with fuel gas from a number of emissions, as well as concern about energy security,
different sources. Lean premix combustion systems, have motivated interest in utilizing associated gas and
widely used for emissions control, require particular raw natural gas, coal-derived syngas or fuels from
considerations regarding the fuel quality and fuel other sources, such as biomass, landfill gas, or process
composition. To determine the suitability for operation gas. For many applications, dry, low-emissions
with a gas fuel system, various physical parameters of systems also have been designed to burn liquid fuels
the proposed fuel need to be considered: heating value, like Diesel.
dew point, Wobbe index and others. Attention is given
Gas fuels for gas turbines are combustible gases,
to the impact of fuel constituents on combustion
or mixtures of combustible and inert gases with a
characteristics and the problem of determining the dew
variety of compositions covering a wide range of
point of the potential fuel gas at various pressure
heating values and densities. The combustible
levels. The impact of fuel properties on emissions, and
components can consist of methane and other low
the proper operation of the combustion system are
molecular weight hydrocarbons, hydrogen and carbon
discussed. Suggestions about how to approach fuel
monoxide. The major inert components are nitrogen,
suitability questions during the project development
carbon dioxide and water vapor. It is generally
and construction phase, as well as in operation are
accepted that this type of fuel has to be completely
made (Kurz et al., 2006).
gaseous at the entry to the fuel gas system, and at all
Gas Fuels points downstream to the fuel nozzle.
The quality and composition of fuel burned in a The amount of energy that will be released during
gas turbine impacts the life of the turbine, particularly the combustion process for a specific gaseous fuel
its combustion system and turbine section. The fuel composition is determined by the heating value. For
specified for a given application is usually based on simple cycle gas turbines, the lower heating value is
availability and price. Natural gas is a typical fuel of generally used since the latent energy of the steam is
choice for gas turbines due to its low (but increasing) not recovered. Gaseous fuels can be categorized by the
cost, widespread availability, and low resulting lower heating value, but another parameter is used
emissions. However, the composition of fuel gas can more commonly in the gas turbine industry. That
widely vary, from gas with significant amounts of parameter is the Wobbe index. From the lower heating
heavier hydrocarbons1 (butane and heavier), to value (LHV) in Btu/scf [kJ/Nm3] and the specific
pipeline quality gas consisting mostly of methane, to gravity (SG), the Wobbe index (WI) of the gas can be
fuel gas with significant amounts of noncombustible calculated:
gases (such as nitrogen or carbon dioxide).
LHV
The fuel composition impacts the entire gas WI =
turbine system. For example, fuels with a large SG
amount of noncombustible components can move the
operating point of the gas turbine compressor closer to Because the fuel supply temperature Tf has an
its stall limit. They also have an impact on the power impact on the actual volumetric fuel flow, a
and efficiency of the gas turbine. Fuel choices impact temperature corrected Wobbe index is often used,
the pollutant emissions levels. Durability issues, where the reference temperature Tref is usually 520°R
especially of the hot section, due to corrosive or 288°K:
compounds or the formation of liquids in gaseous
fuels, have to be addressed. Combustor operability LHV Tref
WI = ⋅
(and durability) issues, especially for lean premix SG Tf
combustion systems depend on the choice of fuel.
Certain fuel types pose safety challenges for the gas If two different fuel gas compositions have the
turbine package due to their flammability, containment same Wobbe index, the pressure drop in a given fuel
(hydrogen is an example), and/or toxicity (carbon system will be the same for both gases and in general,
monoxide, hydrogen sulfide) issues. direct substitution is possible and no change to the fuel
Current dry low emissions technology primarily system is required. The Wobbe index is, thus, an
focuses on burning natural gas, a fuel that is mainly indication of energy flow in the system at the same gas
composed of methane. However, interest in utilizing pressure and pressure drop.
The Wobbe index is used as a parameter to
indicate the ability of the overall fuel handling and
1 Hydrocarbons in fuel gas are usually alkanes, with the
summary chemical formula CnH2n+2
injection system to accommodate the fuel

8 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

composition. If the Wobbe index varies too far from Gaseous fuels can be classified based on changes
the design value, changes to the fuel system need to be required to the injector, fuel or control systems as
made. A high Wobbe index often indicates the specified in Table 1. The table includes special start-
presence of heavier hydrocarbons in the fuel, while a up requirements and limitations in loading based on
low Wobbe index is often caused by the presence of the class of gas fuel. Both Lean Premix and
significant amounts of noncombustible fuel com- conventional combustion gas turbines can be designed
ponents or by the presence of significant amounts of to use the standard fuel with Wobbe index range of
(highly combustible) hydrogen or carbon monoxide. 1220 ±10%. This range would be typical of pipeline
quality natural gas. High heating value fuels are
Because a lower Wobbe index means, that the fuel
accommodated with minor changes to the control
mass flow, relative to the engine air mass flow
systems and can be used in conventional combustion
increases, the power output of an engine increases
and, to a lesser degree, in gas turbines with lean
slightly (Figure 11) when the Wobbe Index is lowered.
premix combustors. Very high heating value fuels
However, the Wobbe index does not capture the usually require that the fuel temperature be monitored
effects of other fuel properties, such as dew point, closely to prevent two-phase (gas and liquid) flow. For
flame speed and combustion chemistry. Therefore, the medium heating value fuels, conventional combustion
entire fuel composition must also be considered and, if may require a change in injector flow area. Low
more reactive species such as hydrogen, alkenes and heating value fuels require even more extensive
carbon monoxide are present in significant quantities, changes to the injector and fuel system. The last
additional changes to the fuel system may be required. category of very low heating values has limited
A good design criterion is that gases having a Wobbe application for conventional combustion only. DLN
index within ±10% can be substituted without making combustion systems in industrial gas turbines are
adjustments to the fuel control system or injector flow usually designed for standard natural gas. The use with
area. If the fuel injector flow area has to be changed, other gases (either high BTU or medium BTU) usually
the gas Wobbe index should be inversely proportional requires careful evaluation and testing.
to the effective controlling area of the injector.
Gaseous fuels can be divided into four distinct
categories as indicated in Figure 12. These categories
are broadly based on the fuel source. A description of
each of these fuels is included in Table 2.
Fuel Gases in Oil and Gas Upstream and Midstream
Applications
Gas fuels produced from oil and natural gas
reserves include: raw natural gas, associated gas,
pipeline quality natural gas, liquefied natural gas
(LNG), natural gas liquids (NGL) and liquefied
petroleum gas (LPG).
Table 1: Classification of Gaseous Fuels
Figure 11: Impact of Wobbe Index on Gas Turbine
Power Output (Typical) Fuel Energy Wobbe Index
For a typical landfill gas, the Wobbe index is one-
Density BTU/scf (MJ/Nm3)
third the value of standard pipeline quality natural gas. Very High >1600
The controlling orifices on the injectors must be (NGL & LPG) (>59. 6)
enlarged to three times their previous area. This allows 1342 – 1600
the fuel flow rate of the landfill gas to have an High
(50. 0 – 59. 6)
equivalent pressure drop across the injector at full-load
condition. This will provide stable, high efficiency Standard 1098 – 1342
combustion, with the desired turbine inlet temperature (Pipeline) (40. 9 – 50. 0)
distribution, for long combustor and blade life. 700 – 1098
Medium
As the fuel heating value decreases below (26. 1 – 40. 9)
standard levels, the torch igniter and the combustion 350 – 700
Low
system may require standard natural gas or liquid fuel (13. 0 – 26. 1)
for startup or shutdown, as well as possible restrictions <350
on turbine transient load operation. Very Low
(<13. 0)

9 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 of natural gas and generally referred to
Natural gas is extracted from the ground either interchangeably with pipeline-quality gas (>90%)
with crude oil or from natural gas reservoirs. It is because it is the major constituent. Mercaptans are
primarily a mixture of naturally occurring paraffin typically added to pipeline gas to give a distinctive
hydrocarbons including methane (CH4), ethane odor that facilitates gas leakage detection.
(C2H6), propane (C3H8,) butane (C4H10), pentane
(C5H12), hexane (C6H14), and higher molecular weight Liquified Natural Gas (LNG) is pipeline quality
compounds. Hydrocarbons heavier than propane form natural gas extracted remotely, chilled to a liquid
isomers, such as iso-butane (iC4H10) and iso-pentane (-260°F/-162°C), transported via tanker, heated,
(iC5H12) that have different properties than their revaporized, processed as required and injected into an
‘normal’ counterparts. Additionally, other gases such existing natural gas transmission and distribution
as carbon dioxide, hydrogen sulfide, mercaptanes, network. LNG has a similar composition to pipeline
water vapor, and nitrogen can be present. natural gas.
Natural Gas Liquid (NGL) is a general
Gaseous Fuel Sources
classification for those paraffin hydrocarbons heavier
than methane, which can be transported and
distributed in liquid form. At atmospheric pressure,
Solid Fuel
Oil and Gas Industry Biomass Coal Derived
Gasification NGL is in the gaseous state. As a gas, NGL is
Operations
separated from methane by liquefying it through the
• Raw Natural Gas • Coke Oven Gas • Landfill Gas • Coal Mine Methane
• Associated Gas • Refinery Gas • Digester Gas • Coal Bed Methane
process of pressure, absorption, or a combination of
• Pipeline Gas • Gasified Solids • Coal Gas - AirBlown both. Because of the relatively high dew point of this
• LNG • Coal Gas - Oxy Blown
• Natural Gas Liquids gas compared to natural gas, NGL is usually handled
• Liquefied
Petroleum Gas
as a liquid fuel for most industrial gas turbine
applications. A determination must be made for each
Figure 12: Gas Turbine Gaseous Fuel Categories source of NGL based on the fuel analysis as to
whether the fuel should be handled in the gaseous or
Associated Gas is extracted with crude oil and
liquid phase.
found either dissolved in the oil or as a cap gas above
the oil in the reservoir. Historically, associated gas has Liquified Petroleum Gas (LPG) is a sub-
been treated as a by-product of oil extraction and has classification of NGL referring specifically to propane
either been flared or re-injected to extract more oil. and butane mixtures offered commercially. LPG is
Burning associated gas in the gas turbines used to always transported in liquid form and as with NGL.
extract the crude oil has been recognized in the past LPG is usually handled as a liquid fuel in gas turbine
two decades as a more economical and applications.
environmentally sensitive approach. The majority of
Fuel Gases Produced in Industrial Applications
associated gas is produced offshore, complicating
further efforts to recover and utilize these gases. The Several industrial and chemical processes
composition of associated gas depends on the type of generate significant quantities of gas by-products and
reservoir from which it originated. waste streams with acceptable fuel composition and
adequate calorific value for use in gas turbines without
Raw Natural Gas or wellhead gas is extracted extensive equipment modifications. Primary industries
from gas wells. The gas reserves are generally producing such fuels include steel mills and chemical
classified as either dry or wet. Dry reserves do not and refining plants.
contain appreciable condensable heavier hydrocarbons
(pentane+), while wet reserves do. Natural gas Refinery Gas is a by-product of gasoline
containing hydrogen sulfide (H2S) is referred to as production. This gas contains high hydrogen (H2)
sour gas. Excessive H2S creates extensive corrosion content, along with several other reactive species
problems on fuel-wetted parts; thus, stainless steel including ethylene and propylene, mixed with
and/or coatings are often prescribed for equipment methane, ethane and propane. Refinery gas is
handling sour gas. corrosive and hydrogen has a great tendency to leak
and is explosive, therefore, safety is the main concern.
Pipeline Quality Natural Gas that is transported With the proper safety considerations incorporated
and distributed for industrial and residential use is a into design modifications and minor fuel control
subset of raw natural gas. Pipeline gas has been adjustments, refinery waste gas has been successfully
processed to remove the heavier hydrocarbons and used to fuel industrial gas turbines. Hydrogen has a
adjust the heating value of the fuel within prescribed relatively low molecular weight and high flame speed
limits. Methane is the lightest combustible component compared to natural gas.

10 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 Coke Oven Gas is generated during the production fuel gas, potential explosion hazards need to be
of coal and petroleum coke for iron and steel addressed. The methane content in the gas/air mixture
production. It contains methane, hydrogen and some should always be kept with a sufficient margin above
carbon monoxide and has a lower heating value than the upper flammability limit, at the highest anticipated
pipeline gas. Coke oven gas composition is quite pressure and temperature.
variable and requires significant treatment to remove
Fuel Gases from Biomass
contaminants for use in a gas turbine combustion that
produces gases with high carbon monoxide and Biomass can be converted into potential gas fuels
hydrogen content. for industrial gas turbines through thermal gasification
and anaerobic digestion.
Anaerobic Digestion Gas is the gas recovered
from the decomposition of organic matter by bacteria Thermal Gasification is a process of heating wood
in the absence of oxygen. Anaerobic digestion occurs and other biological substances without combustion
in: that produces gases with high carbon monoxide and
hydrogen content.
• Liquid sewage
• Residues from fruit and vegetable canneries Contaminants
• Animal and crop wastes In many systems, the gas composition and quality
• Solid wastes disposed in sanitary landfills may be subject to variations (Newbound et al., 2003).
• Marine plants, including macro-algae, water Typically, the major contaminants within these fuels
hyacinth and sea kelp are:

This process produces gas consisting of methane, • Solids

nitrogen, carbon dioxide and air with a low heating • Water
value. Large variations in gas composition are • Heavy gases present as liquids
common from different landfill and digester sites. • Oils typical of compressor oils
• Hydrogen sulfide (H2S)
Fuel Gases Derived from Coal
• Hydrogen (H2)
Coal-derived gases include coal gas produced • Carbon monoxide (CO)
through several different gasification processes and
• Carbon dioxide (CO2)
gases that are collected during coal mining.
• Siloxanes
Coal Gas produced by gasification contains
Other factors that will affect turbine or
hydrogen, methane, and carbon monoxide as their
combustion system life and performance include lower
combustible components, with large proportions of
heating value (LHV), specific gravity (SG), fuel
carbon dioxide and nitrogen as inerts. Depending on
temperature, and ambient temperature.
the type of gasification process, air-blown processes
generate an LHV of from 3. 0 to 7. 9 MJ/Nm3 (75 to Some of these constituents and contaminants may
200 Btu/scf), while oxygen-blown processes yield co-exist and be interrelated. For instance, water, heavy
medium BTU gases with a LHV of from 7. 9 to 21. 2 gases present as liquids, and leakage of machinery
MJ/ Nm3 (200 to 570 Btu/scf). lubricating oils, may be a problem for turbine
operators at the end of a distribution or branch line, or
Coal Mine Gas is generally extracted in two
at a low point in a fuel supply line.
forms: diluted with air as Coal Mine Methane (CMM)
or as a high quality natural gas as Coal Bed Methane Water in the gas may combine with other small
(CBM). In a gaseous coal seam, methane concentration molecules to produce a hydrate – a solid with an ice-
can be up to 94-98% with 1-5% carbon dioxide. For like appearance. Hydrate production is influenced, in
safety reasons, this gas has to be drained from turn, by gas composition, gas temperature, gas
underground coal seams, before mining, in the form of pressure and pressure drops in the gas fuel system.
CBM. To prevent explosions in coal mines, extensive Liquid water in the presence of H2S or CO2 will form
ventilation is required to extract gaseous methane acids that can attack fuel supply lines and components.
trapped between coal seams and the surrounding rock. Free water can also cause turbine flameouts or
In practice, the gas/air mixture from the coal mine is operating instability if ingested in the combustor or
further diluted with air during extraction and the fuel control components.
resultant CMM gas contains 30 to 70% methane by
Heavy hydrocarbon gases present as liquids
volume. This gas/air mixture can be compressed and
provide many times the heating value per unit volume
directly used as a fuel in a gas turbine. Since the
than they would as a gas. Since turbine fuel systems
methane fuel and its air oxidant are premixed in the
meter the fuel based on the fuel being a gas, this

11 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

creates a safety problem, especially during the engine One of the issues is that heavy hydrocarbons, even if
start-up sequence when the supply line to the turbine they are only present in traces, have a large impact on
still may be cold. Hydrocarbon liquids can cause: the dew point. In many instances, a gas analysis lumps
all the heavier hydrocarbons into one number (e. g.,
• Turbine overfueling, which can cause an C6+). Since the dew point depends on the distribution
explosion or severe turbine damage. of the heavy hydrocarbons, estimates can be made
• Fuel control stability problems, because the based on characterization methods (Campbell, 2000).
system gain will vary as liquid slugs or However, these estimates are often not very accurate,
droplets move through the control system. and direct dew point measurements may be the
• Combustor hot streaks and subsequent engine preferred method.
hot section damage.
• Overfueling the bottom section of the In particular, there is no reproducible relationship
combustor when liquids gravitate towards the between heating value and dew point. Because most
bottom of the manifold gases will see a reduction in temperature during
• Internal injector blockage over time, when isenthalpic expansion (this is the Joule-Thompson
trapped liquids pyrolyze in the hot gas Effect), it is possible that even a dry gas can develop
passages. liquids if it is subject to the pressure drop in a typical
fuel supply system. It is, therefore, necessary to
Liquid carryover is a known cause for rapid provide fuel gas sufficiently superheated. Values of
degradation of the hot gas path components in a 28°C/50°F of superheat at turbine skid edge are
turbine (Anderson, 1980; Meher-Homji, et al., 1998; frequently used as a requirement, but the appropriate
Newbound et al., 2003 and 2004). amount (which can be higher or lower) of superheat
The condition of the combustor components also can be determined by a detailed dew point analysis.
has a strong influence and fuel nozzles that have The amount of superheat needs to include allowances
accumulated pipeline contaminants that block internal for uncertainty in the fuel gas composition at present
passageways will probably be more likely to miss and in the future, as well as the potential for heat loss
desired performance or emission targets. Thus, it of the fuel system.
follows that more maintenance attention may be Blowout refers to a situation where the flame
necessary to assure that combustion components are in becomes detached from the location where it is
premium condition. This may require that fuel nozzles anchored and is physically “blown out” of the
be inspected and cleaned at more regular intervals or combustor. Blowout is often referred to as the “static
that improved fuel filtration components be installed. stability” limit of the combustor. Blowout occurs
With a known gas composition, it is possible to when the time required for chemical reaction becomes
predict dew point temperatures for water and longer than the combustion zone residence time. It can
hydrocarbons. However, the prediction methods for be an issue because the chemical kinetic rates and
dew points may not always be accurate. In fact, it is flame propagation speeds vary widely with fuel
known that different equations of state will yield composition. For example, many candidate fuels have
different calculated dew points under otherwise similar heating values but also have chemical kinetic
identical conditions. Furthermore, the temperature in times that vary by an order of magnitude (Lieuwen et
an unheated fuel line will drop because the pressure al., 2006).
drop due to valves and orifices in the fuel line causes a The opposite problem is flashback, where the
temperature drop in the gas. This effect is known as flame physically propagates upstream of the region
the Joule-Thompson effect. Most fuel gases (except where it is supposed to anchor and into premixing
hydrogen) will exhibit a reduction in temperature passages that are not designed for high temperatures.
during an adiabatic throttling. Hydrogen, on the other Similar to blowout, flashback is an issue because of
hand, actually shows an increased temperature when the widely varying flame speeds of candidate fuels.
the pressure drops, which creates a potential explosion Again, fuels with similar heating values but with flame
hazard. speeds that vary by factors of five or more are
Important Fuel Properties common (Lieuwen et al., 2006). Flashback occurs
when the turbulent flame speed exceeds the flow
Dew point of a gas is a function of gas
velocity along some streamline, allowing the flame to
composition, pressure and temperature. It describes the
propagate upstream into the premixing section.
boundary between the single phase gas and the two
Flashback often occurs in the flow boundary layer,
phase (gas and liquid) state of a fluid (Elliott et al,
since this is the point of lowest flow velocity. As such,
2004). Usually, the water dew point of a fuel gas is
the effect of fuel composition variations upon
reported separately from the hydrocarbon dew point.

12 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

flashback depend upon the corresponding change in flame speed is one of the most important parameters
turbulent flame speed. Flame speed is a propagation of governing flashback. High flame speeds occur, for
the flame front moving in the combustion zone. example, in associated gases containing high
Changes in the fuel composition, fuel to air ratio and percentages of propane or butane (Figure 13).
inlet temperature affect the flame speed (Glassman,
1996).
The laminar flame speed, also called flame
velocity, or burning velocity, is defined as the velocity
at which unburned gases move through the
combustion wave in the direction normal to the wave
surface (Glassman, 1996). A key point here is that the
flame speed does not vary linearly between the
respective pure values of the mixture constituents. For
example, the addition of H2 to CH4 does not have a
significant impact upon the flame speed until H2 is the
dominant constituent of the mixture. The impact of
adding diluents, like CO2, is, that the flame speed is
still lower than the non-diluted mixtures, even if the
temperature is maintained by increasing the
equivalence ratio (Lieuwen et al, 2006). Figure 13: Fuel Type Effects on Flame Speed

However, most issues are related to the turbulent Autoignition is a process where a combustible
flame speed, which depends, besides the laminar flame mixture spontaneously reacts and releases heat in
speed, also on the turbulence levels of the gases in absence of any concentrated source of ignition such as
question. In particular, data show that, as the a spark or a flame (Lefebvre, 1998). Rather than the
turbulence intensity increases, the turbulent flame flame propagating upstream into the premixing
speed initially increases, then asymptotes to a constant section, autoignition involves spontaneous ignition of
value, and then, at very high turbulence intensities, the mixture in the premixing section. Similar to
begins to decrease. The most obvious effect of fuel flashback, it results in chemical reactions and hot
properties on the turbulent flame speed is through the gases in premixing sections, but its physical origins
laminar flame speed. For example, for a given are quite different from those of flashback. In lean
turbulence intensity and a given burner, fuels with premix combustors, or in general, in any combustor
higher laminar flame speeds should have higher where fuel and air are premixed prior to combustion,
turbulent flame speeds. However, turbulence intensity this spontaneous ignition inside the injector barrel has
and laminar flame speed alone do not capture many to be avoided, because it can damage combustor
important characteristics of the turbulent flame speed. components, and yields high pollutant emissions. The
Two different fuel mixtures having the same laminar autoignition delay time of a fuel is the time required
flame speed, turbulence intensity and burner can have for a mixture to spontaneously ignite at some given
appreciably different turbulent flame speeds condition. This parameter is a function of the fuel
depending on the diffusion characteristics of the composition, the fuel to air ratio, the pressure, and the
species involved (Lieuwen et al., 2006). mixture temperature. Ignition delay time is of
importance to the combustion specialist because it is a
Flame propagation velocity is also strongly direct indication of potential autoignition in the mixing
influenced by the fuel/air mixture ratio; the leaner the barrel, and it is a useful parameter that defines the
mixture the lower the velocity. If the flow velocity chemical kinetic time scale at any temperature, such as
exceeds the flame propagation velocity, then flameout in the main burner, thus, playing an important role in
could occur. If the flame propagation velocity exceeds the position of the flame relative to the injector tip.
the flow velocity, then flashback within the premixing
injectors could occur that can cause damage by Leaner mixtures tend to have a longer delay time,
overheating the injector tips and walls. To maintain while higher mixture temperatures and higher
flame stability at a point, the velocity of the fuel-air pressures tend to shorten the delay time. In a lean
mixture must be within the flame-propagation speed to premix injector, the flow velocities thus have to be
prevent flashback (Lefebvre, 1998). high enough to avoid autoignition inside the injector at
the prevailing temperatures. Increasing the content of
Flame flashback from the combustion chamber heavier hydrocarbons in an associated gas leads to a
into the premixing zone is one of the inherent decrease of delay time. This is mainly caused by the
reliability problems of lean premixed combustion. The non-symmetry of all higher hydrocarbons: Heavy

13 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

hydrocarbons can be attacked much easier than maximum fuel percentage (volumetric) mixed with air
methane molecules, resulting in reduced ignition delay that will still light and burn when exposed to a spark
times (Figure 14). or other ignition source. The lower is the minimum
fuel percentage to sustain combustion. Different gases
have different ranges of flammability. Hydrogen, for
example, will burn with as little as 4 percent fuel and
96 percent air (lower limit) and as much as 75 percent
fuel and 25 percent air (upper limit). Outside of this
range (less than 4 percent or more than 75 percent
fuel) the hydrogen-air mixture will not burn.
Therefore, hydrogen has a ratio of flammability limits
of 75/4 equal to 18. On the other hand a typical coal
gas has a ratio of 13. 5/5. 3 equal to 2. 5. Coal gas
typically contains methane, CO2, and CO. CO2 is not
combustible. Therefore, if the coal gas contains too
much CO2, the flammability range will decrease and
this ratio of flammability limits will decrease as well.
The stoichiometric flame temperature impacts the
Figure 14: Autoignition Delay Time Depending on
amount of NOX emissions. It is also a parameter to
Fuel Constituents
help verify that a given fuel composition can be
In general, the characteristic kinetic times burned at all gas turbine operating loads and idle.
decrease with the addition of hydrogen, with the Across the flammability range, the mixtures of fuel
lowest times (and hence faster chemical kinetics) and air will burn at different temperatures. As the fuel-
corresponding to mixtures of CO and H2. For the to-air ratio is increased from the lower flammability
ignition times (given in milliseconds) for 1000 K limit, the flame temperature will increase. Upon
temperatures, the longest times (and hence slower further increase in the ratio, a point will be reached
chemical kinetics) are attributed to the mixtures where the amount of fuel and air will be perfectly
containing mostly methane. Similar conclusions can matched so that all the oxygen in the air is reacted
be drawn for the higher temperature (1400 K) with all the fuel - this is called the stoichiometric fuel-
conditions. For example, CH4- only mixtures ignite at to-air ratio. It also corresponds to the maximum flame
approximately 800 µs at 1400 K, while most H2- and temperature. As the fuel-to-air ratio is increased
CO-based mixtures ignite in less than 10 µs (Lieuwen further still, the flame temperature starts to decrease
et al.). In general, few data are available for specific and continues to decrease until the upper flammability
mixtures, and engine specific tests are often necessary limit is reached. In standard combustion systems, with
to avoid problems. a very heterogeneous mixture, the flame temperature is
close to the stoichiometric flame temperature. The
Another important parameter is the ratio of
flame temperature has a significant impact on the NOX
flammability limits. In the combustor, the fuel and air
production rate (Figure 15). Fuel gas with a high
must be continually burned to keep the engine
amount of non-combustibles will usually cause low
running. When the flame in the combustor is
flame temperatures, while fuel gas containing amounts
extinguished it is called a flameout or blowout. The
of hydrogen, carbon monoxide, or heavier
fuel to air ratio changes with the engine load, as
hydrocarbons, will exhibit higher flame temperatures.
described earlier. In order to prevent flameout the
Therefore, a fuel gas with a high amount of diluents
combustor must support combustion over a range of
will yield lower NOX emissions even in standard
fuel to air ratios. Each fuel composition has its own
combustion systems.
flammability range (Ratio of Flammability Limits). If
the engine required fuel-to-air ratio range is equal to or Combustion instability refers to damaging
larger than the fuel flammability range, then, at some pressure oscillations associated with oscillations in the
point, the engine will experience flameout and will not combustion heat release rate. These oscillations cause
be able to operate at that point. Knowing the ratio of wear and damage to combustor components and, in
flammability limits allows a decision whether the fuel extreme cases, can cause liberation of pieces into the
composition has a broad enough flammability range to hot gas path, damaging downstream turbine
support combustion for all operating points of the components. Combustion instability issues are
engine. The ratio of flammability limits is defined as highlighted here because the conditions under which
the upper flammability limit divided by the lower they occur can be influenced by the fuel composition
flammability limit. The upper flammability limit is the (Lieuwen et al., 2006). A necessary, but not sufficient,

14 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

condition for instability to occur is that the unsteady surface. Hot corrosion comprises a complex series of
pressure and heat release oscillations must be in phase chemical reactions, making corrosion rates very
(or, more precisely, that their phase difference is less difficult to predict. It is the accelerated oxidation of
than ninety degrees). Fuel composition variations alloys caused by the deposit of salts (e. g., Na2SO4).
affect combustion instabilities by altering this phase Type I or high-temperature hot corrosion, occurs at a
angle. temperature range of 730 to 950°C. Type II or low
temperature hot corrosion occurs at a temperature
range of 550 to 730°C. Inside gas turbines, hot
corrosion is probably the most potent cause for
premature metal weakening and failure. For example,
hot stress corrosion cracking of turbine blades, which
is effectively the continuous weakening of the blade
metal surface from hot corrosion and the subsequent
growth of subsurface cracks under mechanical
stresses, has been the recognized root cause for many
aircraft and ground based gas turbine engine failures.
Hot Corrosion requires the interaction of the metal
surface with sodium sulfate or potassium sulfate, salts
that can form in gas turbines from the reaction of
sulfur oxides, water, and sodium chloride (salt) or
potassium chloride, respectively. Hot corrosion is
Figure 15: Impact of Fuel Gas on NOx Emissions caused by the diffusion of sulfur from the molten
In order to understand how variations in fuel sodium sulfate into the metal substrate which prevents
composition affect the phase difference between the formation of the protective oxidation film and
pressure and heat release fluctuations, it is necessary results in rapid removal of surface metal. For hot
to consider the specific mechanism responsible for the corrosion to occur, both sulfur and salt (e. g., sodium
instability. Two mechanisms known to be particularly chloride or potassium chloride) have to be present in
significant in premixed systems are fuel/air ratio the very hot gas stream in and downstream of the
oscillations and vortex shedding. In the former combustor. Sulfur and salt can come from the inlet air,
mechanism, acoustic oscillations in the premixer from the fuel, or water (if water is injected). Liquid
section cause fluctuations in the fuel and/or air supply fuel contaminants, such as vanadium or lead, also
rates, thus, producing a reactive mixture whose assist hot corrosion, and will be covered in the section
equivalence ratio varies periodically in time. The about liquid fuels.
resulting mixture fluctuation is convected to the flame Sulfidation, is the reaction between a metal and a
where it produces heat release oscillations that drive sulfur and oxygen-containing atmosphere to form
the instability. The coupling of the premixer acoustics sulfides and/or oxides. This means, in particular, that
with the fuel system is affected by the pressure drop sulfidation only requires the presence of sulfur in the
across the fuel injector. The vortex shedding combustion air or the fuel, but it does not require the
mechanism, as its name suggests, is due to large scale, presence of sodium or potassium. In essence,
coherent vortical structures. These structures are the sulfidation attack is a form of accelerated oxidation
result of flow separation from flameholders and rapid resulting in rapid degradation of the substrate material
expansions, as well as vortex breakdown in swirling due to loss of corrosion protection. During oxidation,
flows. They are convected by the flow to the flame protective oxide scales can form; the metallic sulfides
where they distort the flame front and thereby cause formed are not protective. This accounts for the rapid
the rate of heat release to change. Fuel/air ratio rate of degradation produced by sulfidation attack.
oscillations and vortex shedding become important
when the resulting heat release perturbation is in phase Liquid Fuels
with the pressure fluctuation. In certain regions of the world natural gas is either
Corrosion in the Gas Turbine Hot Section not available or in short supply, and in these situations
the local power producers must turn to alternative
Also called high temperature corrosion, hot liquid fuels with qualities ranging from light distillates
corrosion requires the interaction of the metal surface to heavy residual-grade oils or un-refined crude oils.
with another chemical substance at elevated Liquid distillates such as diesel oil or kerosene are also
temperatures. Hot corrosion is a form of accelerated widely used as secondary back-up fuels in dual-fuel
oxidation that is produced by the chemical reaction applications, which is often a contractual obligation in
between a component and molten salts deposited on its case of disruptions to natural gas supplies. Dual-fuel

15 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

capability may also enable the primary fuel (i. e., U. S. Probably the largest use of naphtha is in India,
natural gas) to be purchased at a more favorable price where it is burned as the primary fuel in almost all gas
(Meher-Homji et al., 2010). turbines in that country.
Fuels should be appropriate for the intended It also should be noted that the fuels listed under
application and, especially in the case of low-grade No. 2 distillate may contain different amounts of
fuels, must also be suitable for the particular gas sulfur. US Highway Diesel No. 2 contains 15ppm
turbine model. Residual-grade oils and crude oils will sulfur, while many other fuels in this category contain
require additional balance of plant and on-site significantly higher levels of sulfur.
treatment programs, and these items also need to be
Distillates normally do not require additional
carefully considered during the initial planning and
treatment at the gas turbine power plant, but certain
design stages. The use of residual oil and crude oil in
fuel properties can influence reliable operation. For
gas turbine applications and the underlying
example, some diesel oils and kerosenes have
requirements for fuel treatment has been reviewed by
relatively high wax content and (more importantly)
Bromley (1986, 2006). BOP design considerations and
high wax melting temperatures, so fuel heating
the importance of realistic fuel specifications for
systems must be designed to prevent wax
heavy fuel projects are described by Moliere et al,
crystallization and filter plugging. Also, naphtha and
2002. A detailed treatment of fuels and combustion is
certain kerosenes are known to have poor lubricating
provided in Meher-Homji et al., ( 2010).
properties, and lubricity improvement fuel additives
Distillates and residual-grade fuels originate in the are often needed to protect critical components such as
oil refinery, and are produced from the fractionation of fuel pumps and flow dividers.
crude oil. The lower boiling point distillates are
Distillate Contamination
relatively free of sulfur and trace metals, but
contaminant concentrations increase in the higher It should always be remembered that “clean liquid
boiling point fractions toward the bottom of the fuels” can also become accidentally contaminated
distillation column. In fact, these components become during delivery, such as by seawater contact from
concentrated in the “residuum”, so that residual-grade marine transportation. Unfortunately this is an all too
oils contain higher contaminant levels than the crude common occurrence, and numerous gas turbine blade
oil from which they were produced. This is an failures (caused by sulfidation corrosion) have been
important consideration for gas turbine applications, reported that relate directly to sodium-contamination
because trace metal impurities create ash deposits of distillates. Ironically, the potential for hot corrosion
during combustion and, unless treated, are responsible with distillate fuels is greater than with residual grade
for various high-temperature corrosion mechanisms. oils or crude oils. This is because ash-forming trace
Other properties such as density and viscosity also metals are anticipated in heavy fuels, so treatment
increase on moving down the distillation column. programs are always applied at the power plant.
However, distillates are often assumed to be
Distillates are classified as “clean” fuels because contaminant-free, so fuel quality monitoring may not
of the absence of ash-forming trace metal impurities. be as diligent and fuel treatment systems may not be
Combustion characteristics are similar to natural gas installed. Furthermore, gas turbine firing temperature
and, without considering price, they are the most is normally higher for distillate fuels and protective
logical fuel alternative. blade coatings may not be applied, so the risk and
Several different grades of distillate are available consequence of hot corrosion is higher (Figure 16).
and those most commonly used as gas turbine fuels are The possibility of sodium contamination during
No. 2 distillate (also known as No. 2 diesel oil, No. 2 the marine transportation of fuels results mainly from
fuel oil or heating oil) and kerosene – which in Britain the fact that sea water is used as “ballast”; i. e., to
is often called paraffin. Naphtha is also gaining adjust the buoyancy of an empty tanker after
popularity as a gas turbine fuel, and is actually the discharging its cargo. Most of the new, large,
generic name for a group of volatile and low boiling oceangoing super tankers are now equipped with
point distillates normally used as a feedstock for segregated (dedicated) ballast tanks, but many older
gasoline production. However, depending on the vessels and coastal barges still use their product tanks
desired product mix at a particular refinery, portions of for sea water ballast. Note that OEM specifications for
the naphtha fraction can be sold directly as fuel. Due sodium plus potassium (actually for total alkali metals)
to the high volatility of naphtha, explosion-proof fuel in gas turbine fuels range from as little as 0. 1 ppm up
handling systems are needed and have been developed to 1. 0 ppm maximum, depending on the gas turbine
for this purpose, and many successful gas turbine model, and it does not take much sea water to exceed
applications now exist around the world including the

16 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

these limits. In fact, 8 gallons of sea water in 1 million refinery. Generally, residual oil quality around the
gallons of distillate will contribute 0. 11 ppm Na + K. world is deteriorating. This is simply because
refineries try to maximize their yield of light distillates
Note, also, that in coastal or marine locations, the
and high-value products such as gasoline. The net
ingestion of airborne sea salt can also initiate hot
effect is that residual fractions become heavier, more
corrosion regardless of fuel type, and this risk applies
viscous, and more contaminated with trace metals.
equally to natural gas or liquid fuel applications.
As a result, so-called “straight-run” residuals
Fuel contamination with traces of lead (Pb) and
(typically termed “No. 6” grade) are normally not sold
zinc (Zn) has also been reported, and which also
without some degree of additional blending with a
causes high temperature corrosion of gas turbine
light- or middle-distillate, in order to meet minimum
blades. These metals are not natural components of
sales specifications for the major markets – such as
petroleum and contamination can usually be explained
maximum sulfur or maximum viscosity. However,
by contact with leaded-gasoline or (respectively) with
these basic sales specifications may not adequately
galvanized materials.
address all important items needed for gas turbine
applications - such as trace metal content, which is
usually limited to a maximum of 100 ppm by the
OEMs. In many cases, therefore, it may be necessary
to establish a project-specific fuel specification with
the supply refinery, which may involve additional
distillate blending to further reduce trace metal
content; but of course this will also increase the price
of the residual fuel oil.
Residual oils coming from different (or unknown)
source refineries should also be carefully monitored
for stability and compatibility before mixing different
batches. This issue relates to asphaltenes (a heavy
Figure 16: Hot Corrosion on Turbine Blades after bitumen-type component of petroleum), and the fact
about 1,000 hrs of Operation on Distillate Fuel that asphaltenes can be destabilized if oils with
Containing 2 ppm Sodium from Sea Water incompatible chemistries are inadvertently blended. If
Contamination (Courtesy Turbotect Ltd. ) destabilization occurs, asphaltenes will agglomerate
(flocculate) and eventually precipitate out of the oil as
sludge. This can result in very serious fuel system
Residual-Grade Oils and Crude Oils problems such as the plugging of fuel filters, flow-
dividers and burner nozzles, etc.
These fuels are classified as “contaminated”
because they contain relatively high concentrations of Various names and abbreviations are used around
various trace metals (sodium, vanadium, etc.) and the world to describe residual-grade fuels including
other impurities such as solids and sediments. Trace “bunker fuel”, “furnace oil”, “mazut”, “LSHS” (low
metals create ash deposits during combustion, which sulfur heavy stock), “LSWR” (low sulfur waxy
impacts gas turbine performance and availability but residual), etc. The most common term within the gas
can also lead to serious high temperature corrosion turbine industry is “HFO” (heavy fuel oil).
issues unless fuel treatment programs are applied at
Crude oil properties vary throughout the oil-
the power plant. Such treatment programs use a
producing countries of the world, but generally they
centrifugal separation processes to remove those
remain relatively constant and specific to a particular
contaminants that are removable, such as water-
oil field or group of oil fields within a region. Crude
soluble sodium salts and water-extractable solids.
oils can be “heavy” or “light” (referring to their
Then, special chemical additives are injected to inhibit
density), “sweet” or “sour” (relating to their sulfur
the corrosive effect of oil-soluble trace metals
content), and can also have different chemical natures
(primarily vanadium) remaining in the fuel.
that influence their chemical properties and the type of
Residual oils are usually blends of several products derived during refinery processing.
different refinery residues (i. e., from atmospheric
Gas Turbine Considerations with Liquid Fuels
distillation, vacuum distillation, and various cracking
processes, etc. ). Their properties vary tremendously The type and model of gas turbine must be
depending on the crude oil feed-stock and the degree suitable for the particular grade of oil, and the
of fractionation and processing within a particular application must be supported by the OEM. Heavy-

17 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

duty engines with lower firing temperatures are used Classification and Origin of Liquid Fuel
for burning residual oil and crude oil, and also for Contaminants
distillates if fuel contamination is expected The most common fuel contaminants are
De-rating is often recommended for ash-forming classified as being either “water-soluble”, “water-
fuels and turbine inlet temperature (TIT) at base load extractable” or “oil-soluble”. This classification is
may be limited to a maximum of about 1050oC useful because it gives an indication as to whether
(1922oF) or lower. This is only a guideline, however, contaminants can be removed from the fuel by a
and actual OEM recommendations will vary according purification processes such as fuel washing, or
to the contractual fuel specification, blade cooling, whether they are chemically bound to the oil and will
metallurgy, hot gas path coatings, and the particular enter the hot gas path.
GT model. There are also slight differences within the Sodium (Na), potassium (K) and calcium (Ca) are
industry as to the specific definition of “turbine inlet water-soluble contaminants, and result primarily from
temperature” and/or “firing temperature”, and the the entry of underground formation water or seawater
actual position in the hot gas path where this during crude oil production. Much of this water is
temperature is defined. For example, variances of separated in the oilfield via dehydration processes, and
about 30 to 40oC can be accounted for if this salt content is further reduced at the refinery prior to
temperature is defined as being just upstream of the crude oil distillation. However, some quantity of Na
1st-stage turbine inlet nozzle, immediately after the and K always remains in the residual fraction and
1st-stage inlet nozzle, or at a point between the first- additional contamination often results from the
stage nozzle and first-stage rotating blade. In all cases, disposal (into residual oils) of refinery waste streams –
the specific OEM recommendations should be which, though not good practice, still commonly
followed. The main reason for de-rating and limiting happens. Contact with seawater during marine
TIT is to minimize the formation of hard ash deposits transportation to the power plant will also introduce
on turbine blades, which is a function of ash chemistry additional sodium, and can occur with any grade of
and the prevailing temperature at which the ash is fuel including distillates. Water-soluble trace metals
exposed. Higher firing temperatures create harder and can be removed (or at least significantly reduced) by a
more compact deposits that are difficult to remove. physical separation process known as fuel washing.
Advance technology models and aero-derivative Heavy oils also contain various solids and
turbines are normally restricted to firing only natural sediments that can lead to filter plugging and erosion
gas or high-quality clean distillates. This said, the problems. These contaminants are classified as “water-
possibility of distillate contamination and/or the extractable” because they are normally associated with
ingestion of airborne sea salt become very critical the water phase and can be removed during the fuel
considerations for these higher firing temperature washing process. Typical materials include sand, silt,
engines. rust, scale, or even asphaltic sludges that may have
High Temperature Corrosion and Liquid Fuel separated from the petroleum. Refinery processing can
Treatment also introduce catalyst fines (cat fines), which are
small and very abrasive particles containing alumina
Residual-grade fuel oils and crude oils are
(Al2O3) and silica (SiO2). This material is used as an
classified as “contaminated” or “ash-forming” fuels
inert catalyst support in refinery cracking units, but if
because they contain relatively high concentrations of
it contaminates residual oils it can be very abrasive to
various trace metals that create ash deposits during
gas turbine fuel system components such as pumps
combustion. Ash fouling impacts gas turbine
and flow-dividers.
performance and availability, but can also lead to
serious high temperature corrosion problems of hot Vanadium (V) and nickel (Ni) are natural oil-
gas path (HGP) components unless appropriate fuel soluble components of most crude oils. They are
treatment programs are applied at the power plant. present as organo-metallic complexes known as
Such treatment programs use a physical separation porphyrins, and are chemically bound to the heavy
processes to remove those contaminants that are asphaltic fractions of the petroleum. Concentrations
removable, and special chemical additives to inhibit can range from a few parts per million (ppm) to
the corrosive effect of oil-soluble trace metals that several hundred ppm according to the type, age and
remain in the fuel. geographic location of the crude oil Refinery
distillation has the effect of concentrating these trace
metals in the high-boiling point residue, so that levels
found in residual-grade oils are considerably higher
than in the crude oil feed-stock.

18 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

Effect of Trace Metal Contaminants and balancing issues. Similar surface oxidation
All trace metals in fuels are undesirable because mechanisms are also initiated by traces of lead (Pb)
they create ash compounds during combustion and and zinc (Zn) in gas turbine fuels.
these mixtures of oxides and sulfates will deposit It is important to note that fuels containing both
within the hot gas path. Ash fouling reduces turbine vanadium and sodium can generate a variety of mixed
output, and in severe cases, it can also cause plugging oxides known as sodium vanadates, and these ash
of critical air-cooling channels on blades and vanes. mixtures can have extremely low melting points
Various high-temperature corrosion mechanisms can around 532oC (990oF). This significantly increases the
also be initiated if ash deposits melt (and remain in the risk of molten-ash corrosion (even with de-rated firing
liquid phase) on component surfaces. However, if ash temperatures), and is the main reason why water-
melting point is sufficiently high that deposits remain soluble sodium salts are reduced to extremely low
in the solid state, these type corrosion processes levels during the fuel treatment process.
cannot proceed. Thus, ash melting point and blade
surface temperature are important parameters that Fuel Treatment to Control High Temperature
jointly determine whether corrosion will occur, and Corrosion
also influence how it can be controlled. Treatment programs for crude oils and residual-
grade oils involve physical separation processes to
Several different types of high temperature
reduce water-soluble sodium and potassium salts to
corrosion are possible, and more detailed descriptions
acceptable low levels. Chemical additives containing
can be found in Stringer (1977), and in Stalder and
magnesium are then applied to inhibit the corrosive
Huber (2000). For gas turbines firing residual-grade
effect of vanadium which remains in the fuel. Overall
oil, crude oil or contaminated distillates the two most
design of the treatment program must be based on the
important corrosion processes are known as
contractual specification of the raw (untreated) fuel
sulfidation corrosion and vanadic oxidation.
that will be received, as well as the treated fuel
specification for the particular gas turbine model.
The effect of fuel contaminants and their impact
on high temperature corrosion is described by Moliere
et al (1995) and Johnson (1996). Valuable early
studies in this field that are still valid today are
reported by Felix (1977, 1979). A useful overview of
gas turbine fuel treatment and the use of magnesium-
based additives can be found in Zubler and Stalder
(1998).
Removal of Sodium and Potassium
A process known as “fuel washing” is used and
Figure 17: Accelerated Surface Oxidation Caused by involves the injection of a quantity of fresh wash water
Vanadium in Heavy Fuel Oil (Courtesy, Turbotect into the raw fuel to dilute and extract water-soluble
Ltd.) salts and water-wettable solids. After mixing, the
emulsified oil and water phases are then re-separated
to produce outlet streams of treated (i. e., washed) fuel
Vanadic Oxidation and effluent water, which now contains the extracted
salts and solids. The most widely used separation
This type of corrosion can occur whenever fuels
technology for gas turbine fuel treatment systems
containing vanadium (V) are burned, such as residual-
involves centrifuges, which increase oil-water
grade oils and crude oils. It is caused by molten ash
separation rate by increasing gravitational force.
deposits of vanadium pentoxide (V2O5), melting point
675oC (1,247oF), which condense on blade surfaces Maximum allowed limits for sodium and
and rapidly destroy protective oxide films. This leads potassium (Na+K) depend on the type of gas turbine
to further oxidation of the alloy surface and further involved, and there is some variance between the
destruction, etc., so that normal oxidation rates are different OEMs. However, typical guidelines are as
accelerated and the service life of components and follows:
coatings is significantly reduced. An example of
vanadium-induced surface oxidation is shown in • Lower-firing temperature heavy-duty models:
Figure 17. The accumulation of ash deposits and hard 0. 5 to 1. 0 ppm max.
corrosion products can also result in serious vibration • Higher-firing temperature heavy-duty models:

19 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 0. 2 to 0. 5 ppm max. ash composition and the impact of blade temperature
• Most aero-derivative models: 0. 1 to 0. 5 ppm are reported by Lay (1974).
max.
Key Elements in the Successful Use of Liquid Fuels
By definition, these limits apply to “total Na+K” In the absence of natural gas, low-grade and
entering the hot gas path from all sources; i. e. from lower-cost residual oils or crude oils can be burned
fuel, air and any water or steam that may be injected successfully in heavy-duty gas turbines, and numerous
for NOx control or power augmentation. As can be examples exist around the world. In many cases, the
seen, these limits are very strict. local population would have been without electricity if
Additive Treatment to Control Vanadic Oxidation these projects had not been undertaken. Although
these type projects are often considered to be risky, it
Vanadium is not removed in the fuel washing is a fact that more high-temperature corrosion failures
process, so chemical additives based on magnesium have occurred with contaminated distillate fuels than
(Mg) are used to inhibit the vanadic oxidation process. with heavy fuels – simply because distillates are
They function by modifying ash composition to “assumed” to be clean and often they are not. Heavy
increase ash melting point, thus ensuring that deposits fuels are known to be contaminated and are treated
on turbine blades do not melt and remain safely in the accordingly. Key success factors for heavy fuel
solid state. The main ash component, without Mg, is applications include:
vanadium pentoxide (V2O5), with a low melting point
of 675oC (1247oF), and this ash would be molten at • The selected fuel must be appropriate for the
typical blade temperatures. However, with the addition particular gas turbine model, and vice-versa.
of sufficient magnesium a different ash component is • The consequences of de-rated performance
created, called magnesium orthovanadate (Mg3V2O8). and lower availability must be clearly
This mixed-oxide has a significantly higher melting understood at the planning stage, and installed
point of about 1243oC (2270oF) and remains solid at capacity should be designed accordingly.
typical blade temperatures. • Realistic and detailed raw fuel specifications
Assuming Na+K has been reduced to OEM must be established early in the project, so that
specification limits as outlined above, a minimum suitable fuel treatment systems can be
treatment ratio of Mg/V=3 is normally recommended; designed and included in balance of plant
i. e., the addition of 3 ppm magnesium for each 1 ppm (BOP) engineering.
vanadium in the fuel on a wt/wt basis. This treatment • High-quality and field-proven fuel additives
formula may need to be adjusted if other oil-soluble should be selected and dosed at the
trace metals such as nickel, lead or zinc are present, recommended treatment ratio.
and specific guidelines set by the particular OEM • A well-equipped site laboratory and
should be followed. Nickel, for example, is not experienced laboratory staff is essential for
regarded as a corrosion-promoting contaminant heavy fuel operations.
because it forms relatively-high melting point ash • Untreated and treated fuel quality should be
deposits. However, nickel interferes with the monitored regularly and routinely to ensure
vanadium inhibition mechanism by combining with continued compliance with fuel purchase
some of the magnesium, so that higher additive dosage contracts and OEM specifications.
may be required. DEGRADATION AND WATER WASHING
Competing reactions with fuel sulfur also occur The fouling of axial flow compressors is a serious
and create magnesium sulfate (MgSO4) as an operating problem and its control is of supreme
additional ash component, and this is another reason importance to gas turbine operators especially in the
why sufficient Mg must be injected. Actually, deregulated and highly competitive power market. It is
formation of MgSO4 is very important, because it is also significant in the mechanical drive market where
the only water-soluble component of the ash mixture a loss in gas turbine output directly affects plant
and enables deposits to be removed by periodic water- throughput. Foulants in the ppm range can cause
washing of the hot gas path. However, if blade deposits on blading, resulting in severe performance
temperature is too high, MgSO4 decomposes to deterioration. The effect of compressor fouling is a
magnesium oxide (MgO), and ash deposits become drop in airflow and compressor isentropic efficiency,
harder and much more difficult to remove. Note that which results in a “rematching” of the gas turbine and
this is the main reason why turbine firing temperature compressor causing a drop in power output and
is limited (as discussed above) when burning ash- thermal efficiency. In extreme cases, fouling can also
forming fuels. Important thermodynamic studies on result in surge problems, as it tends to move the

20 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

compressor surge line to the right, i. e., towards the fouling deterioration, and an example is the damaging
operating line. Details on compressor fouling, and its effects of fouling on blading integrity as discussed in
causes detection and control may be found in Meher- the following sections. This is another important
Homji and Bromley ( 2004), as well as in Kurz and reason to keep the compressor clean. Some typical
Brun (2012). photos of fouled compressors are shown in Figure 19
through Figure 22.
Estimates have placed fouling as being
responsible for 70 to 85 percent of all gas turbine
performance losses accumulated during operation.
Output losses between 2 percent (under favorable
conditions) and 15 to 20 percent (under adverse
conditions) have been experienced.
Gas turbines ingest extremely large quantities of
air, with larger gas turbines having airflow rates of as
high as 1,500 lbs/sec (680 kg/sec). A scatter plot of the
airflow rate versus power for 67 gas turbines is
presented in Figure 18.

800 45
Thermal Eff
700 40
Mass Flow, kg/sec

600
35
Thermal Eff, %

500
Figure 19: Examples of Fouled Compressor Blades:
30
400
(a) Heavily-Fouled Air Inlet Bell-Mouth and Blading
300
25 on a 35 MW Gas Turbine; (b) Oily Deposits on
200
Mass Flow
20 Blading; (c) Compressor Blades Fouled with a
15
Mixture of Salts and Oil (Courtesy, Turbotect Ltd. )
100
0 10
0 100,000 200,000 300,000 400,000
Power, kW

Figure 18: Airflow Versus Power for 67 Different

Industrial Gas Turbines
In a gas turbine, approximately 50-60 percent of
the total work produced in the turbine is consumed by
its axial compressor. Consequently, maintaining a high
compressor efficiency is important for the plant’s
revenue stream. Parasitic losses and mechanical
efficiency have also been considered in the model,
which accounts for the output being somewhat less
than the difference between the turbine and
compressor work. It is useful to present this type of
data to operators to convince them of the importance
of maintaining clean axial compressors by both high Figure 20: Oily Deposits on Axial Compressor Blades
quality air filtration and appropriate washing from No. 1 Bearing Oil Leakage on a Large Heavy
techniques. Duty Gas Turbine

Solids or condensing particles in the air and in the

combustion gases can precipitate on the rotating and
stationary blades causing changes in aerodynamic
profile, reducing the compressor mass flow rate and
affecting the flow coefficient and efficiency; thus,
reducing the unit's overall performance. Further,
contaminated air can cause a host of problems that
include erosion, fouling, corrosion and, in some cases,
plugging of hot section cooling passages. There is also
a close co-relation between mechanical reliability and

21 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 Tank Vapors.
• Mineral Deposits such as limestone, coal dust
and cement dust.
• Airborne Materials - soil, dust, sand, chemical
fertilizers, insecticides, and plant matter.
• Insects – This can be a serious problem in
tropical environments.
• Internal Gas Turbine Oil Leaks – Leakage
from the front bearing of the axial compressor
is a common cause. Oil leaks combined with
dirt ingestion cause heavy fouling problems.
• Impure Water from Evaporative Coolers
(carryover) or inlet fogging.
• Spray Paint that is ingested.
• Vapor plumes from adjacent cooling towers.
Figure 21: Salt Deposits on Compressor Often, the inlet struts and IGVs get severely
Blades(Courtesy, Turbotect Ltd. ) fouled. Hand cleaning the IGVs and first stage will
restore a considerable amount of performance.
However, some cleaning methods simply transport dirt
from the front stages to rear stages.
Ambient air can be contaminated by solids,
liquids, and gases. Air loadings can be defined in
mg/m3, grains/1000 ft3 or ppm (mass of contaminant
per unit mass of air). In general, particles up to 10
microns cause fouling, but not erosion. Particles above
10 to 20 microns cause blading erosion. Some typical
air loadings are as follows:
• Country 0. 01 - 0. 1 ppm by weight
• Coastal 0. 01 - 0. 1 ppm by weight
• Industrial 0. 1 - 10 ppm by weight
Figure 22: Low Solubility Deposits on Compressor • Desert 0. 1 - 700 ppm by weight
Blades(Courtesy, Turbotect Ltd. ) Felix and Strittmatter (1979) have detailed the
type of analysis that should be done at a gas turbine
plant site. In most industrial areas, the air quality can
Underlying Causes of Fouling create quite acidic conditions in the axial compressor.
Experience has shown that axial compressors will
The importance of climatic conditions, such as
foul in most operating environments; be they
rain showers or relative humidity cannot be
industrial, rural or marine. There are a wide range of
overemphasized. Several operators have reported
industrial pollutants and a range of environmental
dramatic drops in gas turbine output coincident with
conditions (fog, rain, humidity) that play a part in the
rain showers. Often, air filters will exhibit a sudden
fouling process.
growth in differential pressure as the filters get
Compressor fouling is typically caused by: saturated with water due to high humidity. Under
certain conditions, the filter may suddenly unload into
• Airborne Salt the airflow causing rapid compressor fouling.
• Industrial Pollution – hydrocarbons, fly ash,
smog, exhaust emissions from traffic etc. This It is instructive to investigate the sensitivity of
causes a grimy coating on the early compressor fouling deterioration on simple cycle gas
compressor stages and can get "baked on" in turbine performance. To this end, simulation runs
the latter stages because of the high using GTPRO software have been made for a 40 MW-
compressor discharge temperatures (This is Class heavy duty gas turbine in simple cycle
especially true of high pressure ratio configuration. Runs were made with natural gas fuel
compressors). with an LHV of 50046 KJ/kg (21,518 Btu/lb) and with
• Ingestion of Gas Turbine Exhaust or Lube Oil typical inlet and outlet pressure drops 100 & 125 mm

22 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 350 700
WG (4 and 5 inch WG) respectively. The machine has
695
an ISO pressure ratio of 11. 8:1 and a mass flow rate Mass Flow

Mass Flow( Lbs/sec) & CDP (Psia)

300 690
of 138 kg/sec (304lbs/sec) and a firing temperature of

Compressor Disch Temp, F

1104°C (2020°F). 685

250 680
The simulation was run at an ambient temperature CDT 675
of 15°C (59°F), and imposing deterioration steps in the
200 670
following sequence: CDP 665

• Step 1: New and Clean, Mass flow drop = 0%, 150 660

Comp. efficiency drop = 0% 655

• Step 2: Mass flow drop = 1%, Comp. 100 650

efficiency drop = 0. 833% 1 2 3 4 5 6 7

Degradation Steps (1 = New and Clean)
• Step 3: Mass flow drop = 2%, Comp.
efficiency drop = 1. 67% Figure 24: Change in Mass Flow Rate, Compressor
• Step 4: Mass flow drop = 3%, Comp. Discharge Temperature and Pressure with
efficiency drop = 2. 5% Compressor Degradation Steps Imposed on a 40 MW
• Step 5: Mass flow drop = 4%, Comp. Class Gas Turbine
efficiency drop = 3. 33%
• Step 6: Mass flow drop = 5%, Comp. The drop in efficiency causes the discharge
efficiency drop = 4. 167% temperature to increase by approximately 10. 6°C
(19°F) and the compressor discharge pressure to drop
• Step 7: Mass flow drop = 6%, Comp.
by about 0. 7 bar (10 psia). The variation in axial
efficiency drop = 5%
compressor work, turbine section work, and the output
Output and heat rate variation with the after losses is shown in Figure 25.
deterioration steps is shown in Figure 23. The output
at the end of the 7th deterioration step has dropped 5.
Compressor Power, Turbine Power, Output, kW

100,000
5 MW while the heat rate has increased by 850 Total Turbine Work
90,000
Btu/kW hr. On a relative basis, the power deterioration
is twice as big as the deterioration in heat rate. 80,000

39,000 11,800 70,000

Output
38,000 11,600 60,000
37,000
11,400 Compressor Work
Heat Rate, BTU/kWhr

50,000
36,000
Output, kW

11,200
35,000 40,000
11,000
34,000 Output Work
30,000
10,800
33,000
Heat Rate 20,000
10,600
32,000 1 2 3 4 5 6 7
31,000 10,400
Degradation Steps (1 = New and Clean)
30,000 10,200
1 2 3 4 5 6 7
Degradation Steps ( 1 = New and Clean) Figure 25: Change in Compressor Work, Turbine
Work and Overall Output with Compressor
Figure 23: Output and Heat Rate Change with Degradation Steps Imposed on a 40 MW Gas Turbine
Compressor Degradation Steps Imposed on a 40 MW
Class Gas Turbine Whereas the axial compressor work is seen to
drop slightly due to the reduction in mass flow (middle
line of the figure), there is a steep drop in the turbine
The change in mass flow rate, compressor work (upper line in the figure) resulting in a drop in
discharge pressure and compressor discharge overall gas turbine output of 5. 5 MW. The steep drop
temperature corresponding to the simulated in turbine section work is due to the reduced mass
deterioration steps is shown in Figure 24. flow and the smaller expansion ratio available as a
result of the loss in compressor discharge pressure.
NON-RECOVERABLE DEGRADATION
Unrecoverable performance deterioration is the
residual deterioration that will exist and detract from

23 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

gas turbine performance even after the gas turbine has Hot Section Erosion: Erosive particles entrained
been washed and cleaned. Damage that occurs to the in the air or fuel can cause turbine section erosion. The
flow path due to erosion, corrosion and increased tip damage is particularly severe if cooling hole blockage
and seal clearances or due to distortion of the casing occurs which can lead to excessively hot blades and
itself will contribute to this. ultimate creep rupture. A reduction in the blade
section size further compounds the stress problem.
During a major overhaul, deteriorated gas turbine
Erosive particles in the fuel can result in nozzle wear
components are cleaned or replaced, and airfoils are
resulting in a distorted temperature profile and severe
recoated (compressor and turbine section as needed)
hot spots at the turbine inlet.
and tip and seal clearances restored as close as
possible to the initial condition. Even after the Hot Gas Erosion: Apart from particulate erosion,
overhaul, depending on overhaul procedures, however, there is also the phenomenon of hot gas erosion which
there may be some performance loss due to: results from localized overheating and thermal cycling
due to intermittent loss of cooling or a breakdown in
• Casing distortion the coating. After several cycles, damage takes place
• Increased flow path surface roughness and the increased roughness (erosion) worsens the
• Airfoil untwist problem. This problem can occur in the first stage
• Increased leakage areas nozzle segments at the platforms. Typically the most
Dikunchak (1993) estimates the extent of this type severely affected parts are those in the hottest gas path
of unrecoverable deterioration to be usually less than 1 (e. g., central to the transition piece).
percent. The extent of the unrecoverable loss will be Hot Corrosion: Also called high temperature
greater if heavy fuels are being utilized. corrosion, hot corrosion requires the interaction of the
Some manufacturers offer overhauls to the extent metal surface with another chemical substance at
that the overhauled engine meets the same elevated temperatures. Hot corrosion is a form of
performance criteria as a new engine. These overhauls accelerated oxidation that is produced by the chemical
in general require the removal of the gas turbine, and reaction between a component and molten salts
the overhaul has to be performed in a dedicated deposited on its surface. Details are covered in the
overhaul facility. Engine exchange programs help to section about fuels. The rate of formation of the
minimize downtime if necessary. sodium and potassium salts necessary for hot
corrosion, and thus the rate of hot corrosion, can be
Erosion controlled by avoiding the presence of sodium or
Due to the presence of air filtration systems, potassium and sulfur in the engine hot section. Sodium
erosion is not a major problem with industrial gas or potassium often enter the engine via the combustion
turbines, though it has been a problem in areas where air. Therefore, the air filtration system plays an
the filtration system could not cope with very large important role. Sulfur, sodium, and potassium can
dust and sand loading such as exists during a enter the engine via the fuel.
sandstorm.
CONTROL OF COMPRESSOR FOULING BY
Erosion is the abrasive removal of blade material ON LINE AND OFF LINE COMPRESSOR
by hard particles such as sand and fly ash, usually WASHING
greater than 5-10 microns in diameter. Erosion impairs
Fouling is best controlled by a combination of two
blade aerodynamic performance and mechanical
methods. The first line of defense is to employ a high
strength. The initial effect of erosion is an increase in
quality air filtration system. However, as fouling will
surface roughness and a lowering of compressor
inevitably occur, compressor washing should be used
efficiency. As it progresses, airfoil contour changes
to control its impact.
occur at the leading and trailing edges and at the blade
tip. Thinning of the trailing edge is detrimental to the This is an area in which strong and divergent
fatigue strength and can result in blade failure. A opinions exist. Washing efficacy is so site specific that
significant loss in tip solidity can promote compressor approaches which work for one site may not be
surge. The typical area of metal loss for a rotor blade appropriate for another. Controversy is often caused
is at the tip while for a stator it is near the root. by polarized opinions relating to wash procedures,
Typically, the erosive particles are centrifuged to the wash media, and techniques. Some of the highlights
outer diameter of the compressor. As a rule of thumb, are presented below in an attempt to present the
blade replacement should be considered when loss of overall picture and the practical field experiences of
cross sectional area exceeds 10-15%. Stalder (1998). Operators must determine the best
approach for their gas turbines by trial and error in

24 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

terms of wash technique, use of on-line washing, visually for color and clarity (which should improve as
which cleaners should be used, and the frequency of dirt is removed), or by a simple conductivity meter to
washing. This is a complex technical-economical monitor the removal of salts.
scenario, and also depends on the service that the gas
On-line washing is now very popular as a means
turbines are in. For example, by not having the ability
to control fouling by avoiding the problem from
to shut down for crank washes, IPP operators, LNG
developing. The primary objective of on-line washing
Operators and merchant power plants may need to be
is to extend the operating period between off-line
more aggressive in controlling fouling. A useful set of
washes by minimizing the build-up of deposits in the
papers relating to compressor washing have been
compressor, and thereby reducing the on-going
provided in Stalder (1998), Stalder and van Oosten
incremental power losses. On-line washing is
(1994) and Bagshaw (1974).
performed with the unit in full operation, and
Several different methods of gas turbine techniques and wash systems have now evolved to a
compressor cleaning have been applied over the years, point where this can be done effectively and safely.
but “wet cleaning” has been found to be by far the Outages or shutdown periods are not required.
most effective and economic technique. However, Depending on the nature of the fouling material, on-
today's sophisticated large industrial engines and blade line washing is sometimes performed with water
coatings require appropriately designed cleaning alone. In most cases, however, the use of an approved
systems to ensure operational safety, reliability and cleaner (detergent) will improve the effectiveness of
optimum efficiency. Two different wet cleaning the washing operation. This is particularly true if the
techniques are generally applied, known as off-line fouling material contains any quantity of oil or grease.
(crank wash) and on-line cleaning. Under extreme Demineralized water quality is almost always
fouling conditions, hand washing of the IGVs may specified by the OEMs for on-line washing, to avoid
have to be conducted if time permits. During the possibility of introducing harmful trace metal
overhauls, hand cleaning of the full axial compressor contaminants (such as Na+K) into the combustion
is most effective. There has also been some recent turbine.
interest in foam cleaning of gas turbines.
Optimal compressor cleaning can normally be
Off-line washing is almost always carried out with achieved by adopting a combined program of regular
the aid of a detergent, and extremely effective power and routine on-line washing (for example, every few
recovery can usually be achieved. However, it is days or weekly), plus periodic off-line washing during
important that the manufacturer's recommendations planned outages.
are followed with respect to water quality,
detergent/water ratio and other operating procedures.
Typically, wheel space temperatures must be below
200°F to avoid thermal shock, and the off-line water
wash is done with the machine on crank. The
downtime for a crank wash depends mainly on the
time it takes for cooling the engine. Larger, heavy-
duty engines can take 8-10 hours to cool, whereas, on
light aeroderivative engines, only 1. 5-3 hours may be
needed because of the low metal mass. Off-line
cleaning is most effective when carried out in several
steps that involve the application of a soap and water
solution, followed by several rinse cycles using water
alone. Each rinse cycle involves the acceleration of the
machine to approximately 50% of the starting speed, Figure 26: Off Line and On Line Cleaning (Courtesy,
after which the machine is allowed to coast to a stop. Turbotect Ltd. )
A soaking period between each wash and rinse cycle is
Two main types of cleaning agents (detergents)
also very important, and allows the soapy cleaning
are available for compressor washing and are normally
fluid to penetrate into the fouling deposits, thus
classified as “water-based” or “solvent-based”
dissolving salts and emulsifying oil and grease
products. Most of the new-generation products contain
components. A useful method of determining the
surfactants, wetting agents and emulsifiers, and
effectiveness of the off-line wash, and perhaps the
involve either an aqueous or petroleum-based solvent
need for additional wash or rinse cycles, is to collect
system. Both products are normally supplied as
samples of the effluent water from all available drain
concentrates, and are diluted on-site with water
ports (Figure 27). The samples can be checked

25 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

(typically one part cleaner with four parts water) to Typical water quality requirements for an off-line
produce the cleaning fluid. Solvent-based cleaners wash are:
have traditionally been recognized as being more
effective in removing oil and grease deposits, but • Total Solids (dissolved and undissolved) ≤ 100
certain new-generation water-based cleaners are ppm
formulated to be equally effective. Most water-based • Total alkali metals (Na, K) ≤ 25 ppm
products also have the advantage of being • Other metals which may promote hot
biodegradable, which is an increasingly important corrosion (V, Pb) ≤ 1 ppm
requirement within this industry. • pH 6. 5-7. 5

Off Line (On-Crank) Washing This water would be used for cleaner dilution and
also for rinsing.
The basic objectives of off-line cleaning are to
clean a dirty compressor and to restore power and Off-line crank washing systems should be
efficiency to virtually “new & clean” values. When designed to achieve the highest washing efficiency
performed correctly, and provided the operating period with the smallest injection mass flow. This is
between off-line washing is not too long (site important for the following reasons:
specific), this type of cleaning will typically restore
virtually 100% of the lost power and efficiency • Gas Turbine users are interested in minimizing
attributed to compressor fouling. However, the quantity of effluent water to be disposed
irrespective of the compressor performance of.
degradation actually encountered, experience has • Some users claim that off-line water effluent is
indicated that users of both base-load and peaking gas transported up to the exhaust during the wash
turbines should incorporate a minimum of three or procedure, and may wet and soak into the
four off-line compressor cleanings per year in order to expansion joint fabric - resulting in damage of
remove the salt laden deposits on the downstream the expansion joint by lowering its insulation
stages. properties.
• A lower off-line injection mass flow will also
Off-line wet cleaning (also known as crank reduce the potential risk of trace metal
washing) is a typical "soak and rinse" procedure for contamination in exhaust systems, where
which the gas turbine must be shut down and cooled. selective catalytic reactors (SCR) for NOx
The compressor is rotated at crank-speed while a reduction or CO catalysts are installed.
cleaning fluid is injected via nozzles or jet lances. • A smaller off-line injection mass flow will
Hand-held jet lances were widely used in the past and significantly reduce the required size, volume
are still fairly popular with some operators. However, and cost of washing skids and the overall
permanently mounted off-line nozzle systems installed water and cleaner consumption.
in the air intake plenum are now preferred, and are
generally offered as standard by most of the major Important Considerations During Off Line Washing
turbine manufacturers. Nozzle design, system Wetting of IGVs: Effective wetting of the IGV’s
operating pressure, and total mass flow parameters suction area can be achieved by using full cone jet
vary widely, however, between the different spray nozzles. The number of nozzles will be defined
manufacturers. by the area to be wetted, which is usually the area
between two intake struts. The necessary off-line
The injected cleaning fluid is normally a mixture
injection mass flow characteristic will, therefore, be
of chemical detergent and water. Both solvent-based
determined by the area to be wetted and impacted by
detergents and water-based products are used,
the jet spray, and the distance between nozzles and the
depending mainly on the type of fouling material
IGVs. The injection pressure is generally between 5. 5
found in the compressor and local plant experience.
to 6 bar (80 to 87 psi). As spray jet trajectory is subject
After a soaking period the compressor is rinsed to gravity, the nozzle should be designed to provide an
with a quantity of fresh water. The amount of rinse adjustment of up to five degrees to compensate for the
water required and the number of rinse cycles vary gravity effect. Crank washing (soaking) and rinsing
from site to site, according to the gas turbine model can be considered as a mechanical “erosion” of the
and the amount of dirt removed during the off-line deposit layer, and soaking time will allow the cleaner
wash. Note that demineralized water is usually not to penetrate and soften the deposit layers. Systems
specified for off-line cleaning and fresh water quality with high atomization pressure will have no impact
is normally acceptable. Effluent water drained from pressure on the IGVs, because the spray pressure will
the compressor has to be disposed of according to have decayed approx. 20 cm from the nozzle outlet,
local regulations.

26 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

and the atomized droplets will need to be carried by prior to a crank wash procedure. Some typical items of
the relatively low air flow rate produced at crank importance include, although they are very engine
speed. Most high-pressure systems do not show the specific:
same effectiveness in removing salts and insoluble
compounds on downstream stages. Typical location of • Ensure that OEM’s Wheel Space Temperature
wash nozzles for both on-line and off-line systems is criteria are met prior to the crank wash.
shown in Figure 26. • Seal and atomizing air pipes should be closed-
off to avoid water entry.
Drainage: The effluent water collection system to • IGVs (if applicable) should be in the open
drain and remove dirty water out of the engine is of position (maximum airflow) prior to the crank
prime importance. Equally important is the isolation wash.
scheme to prevent run-off water from penetrating • All drains should be opened.
sensitive areas such as sealing and cooling air systems, • Flame detector valves should be closed.
and instrumentation air systems. The physical location • Auxiliary air compressor should be
of bleed air taps on compressor casing is also disconnected.
important. If they are located on the bottom, they are • Special precautions may have to be taken for
likely to drain run-off water, and therefore, should
regenerators, if these exist.
preferably be located in the upper part of the casing.
• If possible, the plenum should be hand washed
Drains in air inlet and compressor casing, combustion
to avoid dirt being washed into the
chamber, and exhaust should be located at the lowest
compressor.
point. The drain diameters should be sized for an easy
• Follow the OEMs or Wash system supplier’s
run-off. Care must also be taken that they do not
recommendations regarding the duration and
become plugged with dirt.
amount of fluid injection. These guidelines
may need to be modified by evaluating the
results.
• Approved anti-icing agents will be required if
off-line cleaning is performed at ambient
temperatures below or near to freezing. Even
at crank speed, some degree of temperature
depression will occur at the bellmouth.
The rinse cycles need to be repeated until the
effluent is clean (Figure 27).
On Line Washing
Figure 27: Effluent from Off Lne Water Washing: The basic objectives of on-line cleaning are to
Visual Assessment and Conductivity Evaluation. maintain the cleanliness of a compressor after off-line
(Courtesy Turbotect Ltd., Bromley & Stalder). washing, to maintain power and efficiency by
Engine Speed Variation: Off-line crank cleaning minimizing ongoing losses, and to extend the
is very efficient for removing all deposits on all the operating period between shut-downs required for off-
compressor stages. To enable penetration of the wash line (crank) washing. On-line washing for fouling
and rinse fluid through the entire axial flow control has become increasingly important with base
compressor, off-line washing should be conducted at load combined cycle plants and CHP plants. It is also
variable speeds; for example, by injecting the important for gas turbines in mechanical drive service,
solutions during coast down of the shaft, after an where little or no redundancy is installed and where
acceleration of up to 500 to 600 rpm. By doing this, the downtime associated with crank washing must be
the pattern of the centrifugal forces on the injected minimized.
solution through the compressor will decrease and On-line wet cleaning is performed while the gas
allow better wetting and distribution on the blade and turbine is in operation and at load. The procedure
vane surfaces of all stages. By contrast, off-line involves the injection of a mixture of water and
washing at high and constant cranking speed will chemical detergent via atomizing spray nozzles
result in a lower cleaning efficiency. Conductivity positioned around the compressor air intake plenum.
measurements and checks on the clarity/turbidity of This is followed by a flushing period using pure water.
the drain water will help assess cleaning efficiency. With on-line cleaning it is mandatory to use
Procedures and Precautions: The OEM’s demineralized water for preparing the cleaning fluid
recommendations and checklists should be followed and for flushing. This is because the turbine is in

27 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

operation, and high temperature corrosion damage and this can only be considered as “good” for the gas
may occur if sodium or other contaminant metals enter turbine. To summarize:
the combustion path.
• Frequent on-line cleaning with a low injection
The water specification of the particular OEM mass flow system minimizes the risk of
should be followed, but typical values are presented deposits being suddenly washed from the front
below: stages onto downstream compressor stages.
This also addresses concerns that removed dirt
• Total Solids (dissolved and undissolved) ≤ 5
may enter airfoil film cooling systems of
ppm
turbine blades.
• Total alkali and other metals which may
• Low injection mass flow reduces blade
promote
loading, and creates less stress on downstream
• Hot corrosion (Na, K, Pb, V) ≤ 0. 5 ppm blading.
• pH 6. 5-7. 5 • If the gas turbine uses flame detectors, high
An on-line washing program should always be water mass flow may interfere with flame
started on a clean engine, after an overhaul or crank detector intensities, and may also create a trip
wash. It is not recommended to perform on-line by causing fogging of flame detector lenses,
washing on a heavily fouled engine because large etc.
quantities of dirt removed from the front stages would • Higher CO emission levels may be observed
instantaneously pass through the compressor. during on-line washing, which becomes more
Therefore, after starting an on-line wash program, the significant with higher water injection rates.
time intervals between subsequent washings should be • Typically, high mass flow nozzles create
kept short; approximately every three days to weekly, larger sized droplets. This means they are
depending on the local conditions. Also, depending on more influenced by gravity, and have a greater
the type of deposits (i. e., portion of water-insoluble tendency to fall and be deflected onto inlet
compounds), detergent cleaners may be used for every plenum surfaces. This also increases the risk of
on-line wash, or for every second or third on-line blade erosion.
wash, but not less frequently than once per week. Note • Effective on-line cleaning requires that the
that the longer detergent washing is not done, the IGVs are thoroughly wetted with appropriately
greater the risk of downstream contamination due to sized water droplets (typically 50-250 microns
large portions of insoluble compounds suddenly being range). Any excess water is likely to “stream”
removed when the next detergent wash is performed. over inlet plenum surfaces and struts, etc., and
Thus, frequent on-line washing using detergents is serves little or no purpose in the washing
advisable to minimize the accumulation of insoluble process. As mentioned above, excess water
foulants. impacting the root area of IGVs may also
initiate erosion of the rotating blades.
The duration of each on-line wash can also be
varied according to the degree of fouling, engine size, • Lower injection mass flow rates reduce
and plant experiences, etc. Typical on-line cleaning demineralized water and cleaner consumption,
cycles are in the order of 10-20 minutes, and a and also reduce the size, volume and cost of
flushing or rinsing cycle (using only demineralized the wash skids.
water) of about the same duration should be applied • Experience has shown that a low water wash
after each cleaning cycle with detergent; for example, mass flow does not impair on-line cleaning
10-20 minute cleaning cycle with detergent, followed efficiency.
by 10-20 minute flushing cycle without detergent. On-Line Compressor Cleaning Efficiency
This type of regular on-line wash regime will extend
Fouling of the first stage guide vanes is the
the operating period between outages required for off-
primarily cause of reduced air mass flow through the
line cleaning, which is particularly important for base
compressor. On-line cleaning is most effective in
load plants.
removing such deposits, and therefore, restoring
The design of an on-line washing system should design air mass flow and lost power. Regular, on-line
attempt to obtain the highest possible cleaning cleaning will keep inlet guide vanes clean and free
efficiency with the lowest injection mass flow rate, from deposit build-up. Droplets of cleaning solution or
and this can be achieved in combination with an water may survive up to the 6th stage in some gas
optimum washing regime as discussed above. turbines (depending on the stage loading of the
Frequent on-line washing (to keep a clean compressor compressor), but most will have vaporized by then.
clean) enables the use of low injection mass flow rates, On-line cleaning has no effect on downstream stages,

28 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

but there is some evidence that dirt from the front condition monitoring technique(s) are based on the
stages may deposit on rear stages. following criteria (Meher-Homji, 1996):
Blade Wetting Considerations with On-line 1. What are the specific plant objectives in installing
Washing: Effective wetting of the IGVs is obtained by the condition monitoring system? What
a uniform and finely distributed atomized cleaning specifically do users hope to get from the system?
solution. Droplets must be stable in size and small 2. Where is the machine located? For example,
enough that they do not cause blade erosion (due to the offshore locations may have special requirements.
high blade speeds during operation). Droplets are Unmanned stations would also have special
subject to gravity, so they must also be light enough condition monitoring requirements.
that they do not drop out of the air stream before they 3. What particular group of machines are under
reach the compressor blade surface. Non-uniform consideration? What are the key failure modes?
wetting of the IGVs will result in spot cleaning and 4. What is the criticality of the machinery being
heavier droplets will most likely fall to the bottom, monitored?
wasting some of the injected cleaning solution. 5. What are the mission objectives (i. e., is it
performance, reliability, readiness)?
Optimum Washing Regime for Axial Flow
6. What is the relevant plant operation maintenance
Compressors An optimum-washing regime would
philosophy and can the condition monitoring
combine on- and off-line cleaning to minimize power
technique(s) be successfully implemented?
deterioration and extend the operating period between
Implementation is a key issue.
outages required for crank washing. Depending on
local site conditions, the number of off-line washes An appropriate set of condition monitoring
may be reduced to two or three per year, compared to technologies should be chosen based on failure modes,
four or more per year without on-line cleaning. detailed engineering studies and an assessment of how
Information on some important field studies involving valuable information is when difficult value judgments
combined on-line and off-line wash programs is have to be made relating to machinery operation. For
presented below. large critical gas turbines, it is usually a combination
of technologies that have to be implemented to obtain
Recently, Brun et al (2013) reported on an
good results.
experimental evalution of on-line water washing
procedures. The objective of this project was to Condition Monitoring Approaches
provide a thorough experimental evaluation on the The concept of a slow degradation of component
effectiveness of online turbine cleaning with various strength or performance is the basis for condition
cleaning agents, including high-purity deionized (DI) monitoring systems. The condition monitoring system
water. The results show that most of the cleaning relies on a variety of sensors that help in the
occurs shortly after the cleaning fluid is introduced observation of component degradation.
into the flow stream.
Aerothermal Performance Analysis
The authors of the study found that while online The aerothermal performance of turbomachinery
cleaning works, but that there is no particular provides invaluable insight into its operating health.
difference between the results with various cleaning Performance analysis can vary from relatively simple
liquids. calculations to advanced gas path analysis techniques
They also found that a significant fraction of the dirt used to pinpoint faults. It involves computation and
removed during online cleaning will redeposit correlation of all performance variables in the gas
on downstream blades. The actual quantity of the path. An aerothermal performance analysis system
redeposit depends strongly on the local flow field and should be capable of modeling and detecting degraded
the type of particles that are being carried in the performance.
freestream. The cleaning detergents used did not affect
It is often valuable to integrate such a system with
the level of redeposition.
vibration analysis as several vibration problems are
manifestations of underlying aerothermal problems.
Further, this technology provides insight into how
efficiently fuel is being utilized and thus facilitates
CONCEPTS OF CONDITION MONITORING
significant fuel savings if degradation is controlled.
Choice of Condition Monitoring Approaches Included within “performance analysis” are items such
The choice of the condition monitoring as power turbine inlet temperature or exhaust gas
philosophy (on-line vs. off-line), and the mix of the temperature spread monitoring which is a critical and
valuable indicator of hot section health. Actions such

29 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

as this can significantly extend hot section life. • Bearing Wear and damage can be detected by
Excessive spreads can occur due to a variety of evaluating the transient data for changes in
reasons including excessive air leakages, blockage of amplification factor (Q factor).
nozzles and cracks in the combustor liner/ transitions. • Thermal Bows occurring in the rotor. Very often,
Vibration Analysis this is the result of a hot restart when the rotor has
taken on a bow and the casing has cooled non
With the correct choice of sensors and analysis uniformly.
techniques, vibration analysis is an excellent condition • Rotor Rubs can be detected by analysis of
monitoring tool when used in conjunction with other vibration signatures
condition monitoring techniques.
Dynamic Pressure Analysis
Transient Analysis
The use of dynamic pressure transducers has
Condition monitoring based on transient data worked well to detect certain blading instabilities and
relating to both performance and vibration is an compressor instability. This is an important facet of
important facet of any analysis system. An overview condition monitoring that has not received much
of transient analysis techniques is made by Meher- attention.
Homji and Bhargava (1992). In a gas turbine, valuable
condition monitoring information is available by When applied to the combustors, the dynamic
examining the profile of startup acceleration, coast pressure signatures can provide information regarding
down times, EGT response during light off and other combustion instabilities, and subsequent combustor
transient behavior. liner wear. With the use of dry low emission
combustors, dynamic pressure measurement is a very
Mechanical Transient Analysis valuable condition monitoring tool to ensure
Compared to aerothermal transient analysis, operability.
mechanical transient analysis has been an area that has
been used extensively in condition monitoring systems Lube Oil Debris Analysis
over the years. Transient techniques include: Debris is generated by rolling and sliding surfaces
that are subject to normal, accelerated and incipient
• Vibration Cascade Analysis: These permit a failure. This debris can be generated at differing
visual representation of the frequency content of production rates, and be of different materials size and
vibration signal during startup or shutdown as distribution. Depending on failure mode, debris
shown in Figure 11. Techniques are available to production can increase in certain size ranges. In
check rotor related and fixed frequency rolling element bearings (such as used in aero
amplitudes during transient conditions. Vibration derivative gas turbines) under full elastohydrodynamic
cascades can be formed either on a Delta time or lubrication (EHL) conditions, where film thickness is
Delta RPM basis. large compared to average bearing surface roughness,
• The Bode and Nyquist plots are classic rotor the predominant failure mode is spalling or macro
dynamic displays of particular use for displaying pitting induced by surface fatigue. Particles produced
the location of critical speeds and their are typically in the 100 - 1000µ range. In the boundary
amplification factor. The separation margin from lubricated and mixed mode (partial EHL) regimes,
the critical speed to the operating speed range can particles are typically smaller (< 100µ). Under these
be reviewed for compliance with manufacturers or lubrication conditions, abrasive and adhesive type
industry recommendations. accelerated wear models occur. Bearing skidding can
• Another useful representation for transient occur when bearing loads are light and can produce
analysis may be based on a vectorial summation small debris (< 25µ).
technique of proximity probe gap voltages. This
can be used in a display to depict the movement Borescope Inspection
of the shaft within the bearing during One off-line Condition Monitoring technique that
startup/shutdown. is successful and very necessary is borescope
• For startup analysis of other mechanical inspection. This is usually carried out at fixed intervals
parameters (lube oil temperatures, and pressure), dependent on the machine and a video camera is used
it is possible to prepare maps showing these to record results. Borescope inspections are usually
parameters as a function of rotor RPM & load as very quick and result in a minimum loss of turbine
appropriate. availability. For borescope inspections, it is important
to have well trained personnel and clear cut procedures
Problems which can be detected by Mechanical defined in concurrence with the machinery
Transient Analysis include: manufacturer, to ensure full coverage of the critical

30 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

components2. By using a video camera to record the Bromley, A. F., 1986, “Gas Turbine Fuel Treatment -
inspection, one can enlist expert outside help to The Reasons Why,” Diesel and Gas Turbine
interpret the data. Eddy current checking is also being Worldwide, June 1986.
done to detect cracks.
Bromley, A. F., 2006, “Gas Turbine Fuels and Fuel
Condition Monitoring as a System Treatment,” Combined Cycle Journal, Outage
The previous sections outline specific methods, Handbook, 3Q 2006, pp. OH33-OH39.
technologies, and processes that can be done manually Brun, K., Foiles,W.C., Grimley,T.A., Kurz, R., 2013,
and ad hoc. However, there is a trend towards ”Experimental Evaluation of the Effectiveness of
automating condition monitoring and remote condition Online Water Washing in Gas turbine
monitoring, especially in the context of fleet Compressors”, 42nd Turbomacinery Symposium,
management. Remote monitoring and diagnostics, in Houston, Tx.
conjunction with centralized data storage creates the
ability to perform comparative analysis to similar Elliott, F. G., Kurz, R., Etheridge, C., and O’Connell,
engines (Fleet Statistics). Monitoring ancillary system J. P., 2004, “Fuel System Suitability
performance identifies risk factors to engine durability Considerations for Industrial Gas Turbines’,
(filter fouling rates, lube oil temperature and pressure TransASME JEGTP, Vol. 126, No.1, pp 119 -126.
control, fuel temperature, machine event diagnostics, Felix, P. C., 1977, “Corrosion and its Prevention in
etc. ). Rigorous tracking of availability and reliability Modern Stationary Gas Turbines,” Brown Boveri
as key performance indicators and downtime analysis Review, 1-77, pp. 4—46.
enables understanding and analysis of what systems
and situations are having the most negative impact on Felix, P. C., 1979, “Practical Experience with Crude
these metrics. and Heavy Oil in Stationary Gas Turbines,”
Brown Boveri Review, 2-79, pp. 89-96.
Finally, it must be emphasized that a key
requirement for successful condition monitoring is that Felix, P. C., and Strittmatter, W., 1979, “Analysis of
all measures have to be performed regularly and there Air Pollution on the Erection Site of a Brown
is sufficient expertise available and accountable to Boveri Gas Turbine,” Brown Boveri Review, Vol.
analyze and act on the input as it is received. 66, February 1979, pp 97-103.
CONCLUSIONS Glassman, I., 1996, Combustion, Academic Press,
New York, 3rd Edition.
The operation and maintenance of gas turbines
requires an understanding of the working principles of Johnson, K. W., 1996, “Operation of Combustion
the gas turbine and its components. Additionally, fuel, Turbines on Alternate Fuels,” Proceedings of
as well as air filtration,has an important impact on the ASME IGTI Turbo Expo, Birmingham, UK, June
operating characteristics, and the life of the gas 10-13, 1996, ASME Paper No. 96-GT-7.
turbine. The results of all measures can be evaluated Kurz, R., Wen, C., Cowell, L. H., Lee, J. C. Y., 2006,
by a thorough condition monitoring regimen. This ‘Gas Fuel Flexibility Considerations for Low
tutorial is intended to further the understanding in Emissions Industrial Gas turbines’, The Future of
these areas. Gas Turbine Technology, 3rd International
REFERENCES Conference 11-12 October 2006, Brussels,
Belgium.
Bagshaw, K. W., 1974, "Maintaining Cleanliness in
Axial Flow Compressors", 1st Gas Turbine Kurz,. R., Brun, K., 2012, Fouling Mechanisms in
Operations and Maintenance Symposium, Axial Compressors, TransASME JEGTP, Vol.
National Research Council - Canada, October, 134, GTP-11-1153.
1974. Lay, K. W., 1974, “Ash in Gas Turbines Burning
Brekke, O., Bakken, L. E., 2010, Performance Magnesium-Treated Residual Fuel,” Transactions
Deterioration of Intake Air Filters for Gas of ASME Journal of Engineering for Power, April
Turbines in Offshore Installations, ASME 1974, pp. 134-137.
GT2010-22454. Lefebvre,A. H., 1998, Gas Turbine Combustion, 2nd
Ed., Taylor and Francis, Philadelphia.
Lieuwen,T., McDonell,V.,Petersen, E., Santavicca,D.,
2006,” Fuel Flexibility Influences on Premixed
2 Borescope inspection can show up component cracks,
erosion, corrosion and buckling.

31 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 Combustor Blowout, Flashback, Autoignition, and Stalder, J. P., and van Oosten, P., 1994, “Compressor
Stability”, ASME paper GT2006-90770. Washing Maintains Plant Performance and
Reduces Cost of Energy Production,” ASME
Meher-Homji, C. B. and Bhargava, R. K., 1992,
International Gas Turbine and Aeroengine
“Condition Monitoring & Diagnostic Aspects of
Congress, The Hague, Netherlands, June 13-16,
Gas Turbine Transient Response,” ASME Paper
2001, ASME Paper No: 1994-GT-436.
No. : 92-GT-100, 1992 ASME Gas Turbine
Congress, Cologne, June 1-4, 1992. Also in Stalder, J. P., 1998, “Gas Turbine Compressor
International Journal of Turbo & Jet Engines, Vol. Washing State-of-the-Art Field Experiences,”
11, No 1, 1994. ASME International Gas Turbine and Aeroengine
Congress, Stockholm Sweden, June 2-5, 2001,
Meher-Homji, C. B., 1996, “Condition Monitoring of
ASME Paper No: 1998-GT-420.
Power Plants,” Handbook of Condition
Monitoring and Maintenance Management in Stalder, J. P., and Huber, P. A., 2000, “Use of
Industry, Elsevier Publishing, 1996. Chromium Containing Fuel Additives to Reduce
High Temperature Corrosion of Hot Section
Meher-Homji, C. B., and Gabriles, G., 1998, “Gas
Parts”, ASME International Gas Turbine &
Turbine Blade Failures- Causes, Avoidance, and
Aeroengine Congress, Munich, Germany, May 8-
Troubleshooting,” Proceedings of the 27th
11, 2000, ASME Paper No. 2000-GT-138.
Turbomachinery Symposium, Turbomachinery
Laboratories, Texas A&M University, Houston, Stringer, J., 1977, “Hot Corrosion of High
Texas, September 22-24, 1998. Temperature Alloys”, Annual Revue Material
Sciences, vol. 7, 1977, pp. 477-509.
Meher-Homji, C. B., 2000, “The Historical Evolution
of Turbomachinery,” Proceedings of the 29th Zubler, R. and Stalder, J. P., 1998, “Review of Gas
Turbomachinery Symposium, Houston, Texas, Turbine Fuel Treatment,” Chinese Association for
September 2000. Gas Turbine Power Generation Users Conference,
Shenzhen, P. R. China, June 1998.
Meher-Homji, C. B., and Bromley A. F., 2004 “Gas
Turbine Axial Compressor Fouling and
Washing,” Proceedings of the 33rd
Turbomachinery Symposium, September 2003,
Houston, Texas.
Meher-Homji, C. B., Chaker, M., Bromley, A. F.,
2009, “The Fouling of Axial Flow Compressors-
Causes, Effects, Susceptibility and Sensitivity,”
Proceedings of the ASME International Gas
Turbine Congress, Turboexpo 2009, Orlando
Florida, June 8-012, 2009. ASME Paper No:
GT2009-59239.
Meher-Homji, C. B., and Bromley A. F., “Gas
Turbine Fuels- System Design, Combustion and
Operability,” Proceedings of the 39th
Turbomachinery Symposium, George R. Brown
Convention Center, Houston, Texas 4-7 October,
2010.
Moliere, M., Gazonnet, J. P., and Perin, S., 1995,
“Updated EGT Experience with Ash-Forming
Fuels: A Step Ahead,” National Conference of
Gas Turbine Generator Users, Hainan, P. R.
China, Nov 27 to Dec 3, 1995.
Moliere, M., Huber, P., Stalder, J. P.,and Bromley, A.
F., 2002, “Key Considerations When Embarking
on a Heavy Fuel Project,” Proceedings of Power-
Gen Asia, Singapore, October 2-4, 2002.

32 Copyright© 2013 Turbomachinery Laboratory, Texas A&M Engineering Experiment Station