Technical Challenges for Compressors and Steam Turbines for Efficient and Sustainable Operation in Mega Ethylene Plants

Satoshi Hata, Ph.D.

Vice President, Engineering
Mitsubishi Heavy Industries Compressor International Corporation (MCO-I)
1221 McKinney St, Suite 4250, Houston, TX 77010
Office: (713) 652-0319,
Email: satoshi_hata@mhicompressor.com

Satoshi Hata is Vice President, Engineering, Mitsubishi Heavy Industries Compressor International Corporation, Houston, USA.
He has 33 year experience in R&D for nuclear uranium centrifuges, turbomolecular pumps, heavy-duty gas turbines, steam
turbines and compressors. Mr. Hata has B.S., M.S., and Ph.D. degrees in Mechanical Engineering from Kyusyu Institute of
Technology
Akinori Tasaki ABSTRACT
Akinori Tasaki is Manager, Compressor
Engineering & Design Section, Mitsubishi Changing markets, industry demands for increased efficiency
Heavy Industries Compressor Corporation, and long term operation, and availability of new technology
Hiroshima, Japan. He has 18 years of have all contributed in development of more efficient and
experience as a compressor design engineer. reliable turbomachines. The authors present ethylene plants
Mr. Tasaki has B.S. and M.S. degrees in trends, demonstrate the challenges faced by the turbomachinery
Mechanical Engineering from Kyusyu equipment manufacturers and highlight various advancements
University. in the turbomachinery technology. History maps are
introduced for design advancements, verification tests, and
Matt Walton, P.E. application results in terms of transient fluid dynamics,
Matt Walton is Manager of Design thermodynamics, rotor dynamics, and blade vibration strength
Engineering, Mitsubishi Heavy Industries evaluation. In addition, after recognizing the need for long
Compressor International Corporation, term operation and related typical damage and deterioration
Houston, USA. He has experience in repair, modes, the authors explain various practical technologies (such
modification, and design of industrial gas as effective on-line washing, flow path surface treatment,
turbines, steam turbines, and compressors. combination of anti-corrosion and erosion prevention, stage
Mr. Walton has B.S. degree in Mechanical performance enhancement by partial component replacement,
Engineering from Texas A&M University. NDE techniques for both compressors and steam turbines, and
a unique casing replacement technique on the same footprint
Kyoichi Ikeno for increasing capacity) used to provide more efficient and
Kyoichi Ikeno is Manager, Turbine reliable machines.
Engineering & Design Section, Mitsubishi
Heavy Industries Compressor Corporation,
Hiroshima, Japan. He has 17 years of R&D
experience for mechanical drive steam
turbines. Mr. Ikeno has B.S. and M.S. degrees INTRODUCTION
in Mechanical Engineering, Miyazaki
University and Kyushu University, During this decade, the capacity of ethylene plants has tended
respectively. to increase continuously. Consequently, flow rates in the high-
and low-pressure stages of the steam turbine driver have also
Abhay Jain increased. To meet this capacity increase, it has been necessary
Abhay Jain is Engineer, Compressor to apply larger sub-components including large flow governing
Engineering & Design Section, Mitsubishi valves, longer blades, and longer bearing span rotors. For
Heavy Industries Compressor Corporation, compression equipment, larger impellers with longer bearing
Hiroshima, Japan. He has 2 year experience spans are required. Compressor and turbine trains with
as a turbine design engineer. Mr. Jain has capacities in excess of 1.5 MTPY have already been designed,
B.Tech degree in Materials and Metallurgical manufactured, and are successfully operating. Therefore, in
Engineering from Indian Institute of order to minimize power consumption and downtime losses, it
Technology (IIT), Kanpur, India.

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

is necessary to conduct machine monitoring for prevention and Figure 2 shows the relationship between MCR (max.
diagnosis of performance deterioration. continuous rotating speed) and steam turbine maximum power.
Turbomachinery equipment manufacturers have been This relation is basically exponential. The required power for a
moving forward to apply state of the art technologies for Mega ethylene plant charge gas compressor (CGC) MP/LP/HP
critical components in order to provide more efficient and driver is 95MW, and can potentially be as high as 120MW.
reliable machines corresponding to the high power An optimized MCR for steam turbine and compressor
requirements for ethylene services. The authors introduce the performance and rotor dynamics is around 4,000 rpm.
technology development history map for typical designs,
verification tests, and application results in terms of transient
fluid dynamics, thermodynamics, rotor dynamics, and blade
vibration strength evaluation. In addition, after
commissioning turbomachinery performance and reliability
tends to deteriorate during long-term operation. The typical
damage and deterioration map for machinery is introduced, and
the authors explain practical technologies that are applied, such
as flow path surface treatment, effective on-line washing,
combination of anti-corrosion, erosion prevention and stage
performance enhancement by partial component replacement,
NDE techniques for both compressors and steam turbines, and
a unique casing replacement technique on the same footprint
for increasing capacity.

Mega Ethylene Plant Trend

In the past decade the capacity of ethylene plants has tended to Figure.2 Recent Steam Turbine Range for Mega Crackers
increase continuously as shown in Figure-1. Consequently, - Simple Single Train Solution
flow rates in the high and low pressure stages of the steam
turbine driver have also increased. To meet this capacity In Figure 3, below, the application range of each service and its
increase, it is necessary to apply larger sub-components corresponding Mitsubishi Advanced Compressor (MAC) frame
including governing valves with larger flow capacities, longer size is shown on a compressor suction flow and discharge
blades, and longer bearing span rotors. For compression pressure map. Large frames (9H and larger) are applied for
equipment, larger impellers with longer bearing spans are mega ethylene plants. These same sizes are used in other
required. Compressor and turbine trains with capacities in applications, such as LNG main refrigerant compressors.
excess of 1.5 MTPY have already been designed, manufactured,
and are successfully operating.

Figure.1 Ethylene Plant Capacity Increase Supported Figure.3 Application Range for Mitsubishi Advanced
by Upgraded Equipment and Technology Compressor frame sizes

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 To produce the increase power, the required inlet steam flow
(HP section) and exhaust steam flow (LP section) become
larger and a larger component size for the flow path is
necessary. Figure.6 shows key technologies to meet the
requirement for large plant and machines which are used in
steam turbine design. A reaction optimized impulse blade is
applied for the inlet and middle stage flow paths in order to
increase efficiency and minimize the thrust load due to
differential pressure across each blade inlet and exit.

The last 3 stages in the LP section utilize a z-lock integral

shroud and high rigidity blade profiles are applied to reduce
resonance vibration stresses in the blade and to increase stage
efficiencies.

Figure.4 Recent Steam Turbine Range for Mega Crackers - A critical design concept in modern rotor design is to increase
Simple Single Train Solution rigidity as much as possible by optimizing blade stage number,
bearing span, and HP/LP section efficiency at the required
Every compressor’s performance, including rotor dynamics, is operating speed. A high rigidity rotor can reduce vibration
measured and verified as a part of the manufacturing process. response due to unbalance and increase vibration stability.
In addition, the steam turbine and compressors can be coupled
and the complete train can be tested to confirm combined Recently, operating intervals of over 8 to 10 years have been
mechanical performance, such as vibration and bearing required to meet customer demand. For this requirement,
condition, on a large test stand prior to shipment. Figure 4 several new technologies have been developed and applied to
shows a 1.5MMTA class mega ethylene CGC train in the same existing machines.
cooperation condition as that at the site.

Challenging Technologies for Critical Components

As plant capacities become larger and process flows increase

the corresponding compressor impeller and casing size must
become larger. The required power also increases. Controlled
manufacturing processes for 1 piece or 2 piece impellers are
also important to meet the design and quality requirements for
these large components. Figure.5 shows key technologies such
as 3D impellers and FFD (Flexible Frame Design) which are
utilized in compressor design. To increase the rotor dynamic
stability a swirl canceling labyrinth seal and a special damper
are applied depending on rotor rigidity.

Figure.6 Steam Turbine Key Technologies

Figure 7 shows the technical matrix for present and future

steam turbine designs. For long term operation and increased
reliability, several surface treatments are applied to protect the
flow path against environmental damage from solid particles or
wetness in the LP stages.

During operation, the flow path tends to develop deposits

caused by degraded steam quality and insufficient control.
These deposits cause performance deterioration and potential
reliability issues. On-line wash technology can clean and wash
Figure.5 Compressor Key Technologies the flow path in full power operation in less than one hour.

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

To simplify the governing valve control system, the oil free
governing valve was developed.

Figure.8 Latest Technologies for Steam Turbine and History

Figure 9 shows a centrifugal compressor research and
Figure.7 Technical Matrix for Present and Future Steam development map. Similar to the steam turbine, flow path and
Turbine Designs profile design concepts and procedures have been developing
from 2D to 3D as per the improvement of computational
technology. Compressor rotor stability is critical in high
pressure applications in large bearing span machines. To
Technological Development History Map increase rotor rigidity and maintain efficiency, the high boss
ratio impeller was developed. To increase the damping of the
Figure 8 shows the latest technologies for steam turbine rotor/bearing/seal vibration system and rotor stability, a
development history from 1980 to the present. Flow path and mechanism for gas excitation force reduction and a high
profile design concepts and procedures have been developing damping seal were developed. In terms of manufacturing
from 2D to 3D by improvement of computational technologies processes, new methods were developed such as 1 piece
and computer calculation speed/data capacity. Due to larger impellers by 5-axis machining for large high flow coefficient
machine requirements, a longer blade is necessary for the last impellers and EDM/ECM for small throat, low flow coefficient
stage. Integrally shrouded blades (ISB) for the high pressure designs.
section were developed. This ISB grouping concept for impulse
type blading is different from conventional twisted designs and
creates one robust blade grouping. TTV and governing valves
were developed for larger sizes and to meet higher inlet steam
temperatures. Direct lubricant tilting pads for journal and
thrust bearings instead of flood types reduce mechanical losses
and reduce pad metal temperatures in high velocity situations.
Rotor repair technologies have been developed for planned and
emergency outages to minimize the risk of no spare rotor.
Several types of seals have been developed to reduce leakage
and increase efficiency. For long term operation, surface
treatment technologies are applied. NDE techniques were also
developed to inspect internal damage on rotor disk grooves.
One of the latest technologies is PAUT, phased array
ultra-sonic testing. Based on measured crack length, location,
and orientation the allowable residual operational life can be
estimated for turnaround and asset planning. Figure.9 Centrifugal Compressor Research and
Development Map

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

Design, Verification Tests, and Application in terms of
Transient Fluid Dynamics, Thermodynamics, Rotor
Dynamics and Blade Vibration Strength Evaluation

Machine components are designed to meet the required

specification which is decided by process data. In some cases,
the compressor impellers and steam turbine blades need to be
newly developed. Here, several designs and performance
verification case studies are explained to show the typical
development process.

Compressor Development
Figure 10 shows the Compressor Flexible Frame Design (FFD)
Concept. The fixed frame consists of flow path suction parts,
discharge parts and bearings. These are high reliability and Figure.11 Verification Process for New Impeller
well-proven components. Flow path aerodynamics design parts
are flexible to achieve optimum performance with high
efficiency.

Figure.12 Development / Improvement of Impeller

Case Study: The High Boss Ratio Impeller

The latest impeller design concept combines a high boss ratio
and short span as shown in Figure 13. In order to increase the
rotating speed for compact designs the shaft rigidity needs to be
larger for low vibration response and high stability. In addition,
the bearing span needs to be shorter while the number of stages
Figure.10 Compressor Flexible Frame Design (FFD) Concept remains the same. In order to realize both, a new impeller
standard series was developed and is in service in operational
machines.
Figure 11 shows the verification process for a new impeller for
a compressor. As the first step, the impeller profile, disk and
cover structure dimensions are decided according to the results
of performance predictions by CFD and static/dynamic strength
evaluation by FEA. The actual impeller size is scaled to the
required diameter using the similarity rule of turbomachinery
flow paths (velocity triangles) and strength (vibration, stress,
power). Next, a standard size test impeller is manufactured
according to this similarity rule. In the third step, a single stage
impeller performance test is conducted to characterize the
performance and compare it to calculated results as shown in
Figure 12. This data is added to the impeller performance
database and is used for actual application design in FFD
process.
Figure.13 High Boss Ratio/Short Span Impeller Design
Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station
Figure 14 shows the profile modification for high boss ratio and Steam Turbine Development
short span impeller design and corresponding performance test For steam turbines, Figure16 shows the high performance of
results. Two concepts are applied: a reduced blade count for the hybrid impulse blade design used to improve the efficiency
reduction of friction losses and a modified blade shape for of the HP section of the turbine. It is possible to minimize the
optimum flow line. This design has a confirmed polytropic friction losses of the steam path by conducting blade and nozzle
efficiency gain over a conventional impeller. CFD analyses and the actual performance is verified by a test
machine and actual machine at the shop.

Figure.14 Profile Modification for High Boss Ratio and

Short Span Impeller Design and performance test results
Figure.16 High Performance of Hybrid Impulse Blade Design

The advantage of the high boss ratio/short span impeller for an Figure 17 shows the high performance of the LP section long
actual machine application is shown in Figure 15. A Mega blades. This example details improvements to efficiency by
ethylene plant PRC train is selected as a typical case. The optimizing the steam flow using CFD to prevent secondary
number of compressor casings can be reduced from two to one; flows between blades and the high Mach flow area. It was
in addition, the weight and length can be reduced by roughly possible to improve the LP section efficiency by 8% in this case
20% and 50% respectively. This represents significant cost study.
savings through reduced raw material requirements.

Figure.15 Advantage of High Boss Ratio Impeller Figure.17 High Performance of LP section Long Blade

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

Figure18 shows the application for higher inlet steam
conditions to achieve higher steam flow utilization.

Increasing steam and energy costs are resulting in the demand

for improved steam consumption. Higher pressure and
temperature steam reduces the steam consumption of the
turbine. High pressure/temperature steam generally has higher
enthalpy and a higher theoretical maximum efficiency. Figure
18 shows a relevant case study from a recent project. One
challenge is to apply advanced materials for the high
pressure/temperature steam and turbine design structure while
maintaining the same dimensions as the original. The stress
analysis is done to evaluate thermal stress under steady and
transient conditions comparing design criteria including creep
Figure.19 Reliable Operation for Rotor Dynamics
strength.
Case Study: LP Blading
In the following case study, the authors conducted an excitation
test on two LP blade rows, using an actual scale model, and an
actual load test with a scale model to verify the LP blades’
reliability. With these tests, the authors could complete the
verification program for the large steam turbine. Figure 20
shows the verification test for integrity of actual scale blades
and confirmation of the principle of similarity. The rotating
blades of steam turbines are essential elements in the
conversion of fluid force to mechanical force, and high
reliability in operation is a critical requirement. During the
design/development stage, vibratory characteristics and
strength of the actual blades were checked and verified with a
rotating excitation laboratory test. However, an actual load test
in a test turbine is the best method of verifying blade reliability.
Due to the limited quantity of steam flow in the test facility, the
actual loading test is usually carried out using a scale model
Figure.18 Application for Higher Inlet Steam Condition turbine instead of a full scale model. Furthermore, mechanical
and aerodynamic laws of similarity can be applied to the scale
model turbine. This figure shows a summary of the principles
of similarity.

Finally, Figure 19 shows the reliable operation for rotor

dynamics. API standards are widely used in the petrochemical
industry. Typically steam turbine designs must comply with
this API standard, which recommends a maximum shaft
vibration of 25.4 micron peak-to-peak during shop mechanical
running tests. Figure 19 shows historical vibration records
during mechanical running tests in the shop demonstrating the
API criterion.

Mechanical drive steam turbines are designed based on impulse

type blades. Typically an impulse turbine has a shorter bearing
span than a reaction type blade design of similar power. Shorter
bearing spans can increase the rotor rigidity and this higher
rigidity contributes to low vibration levels during stable Figure.20 Verification Test for integrity of actual scale blades
operation. Confirmation for Principle of Similarity

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

Figure 21 shows the verification test for integrity of actual scale edge. Six strain gauges were installed on each stage (L-0 to L-3
blades. Before a mechanical rotating test in a vacuum chamber stage). The authors adopted the latest available technology, an
typically used for high speed balance and the actual load test, induced electromotive type telemetry system (no
tapping test of the LP blade rows in the free-free condition was battery/self-generation type). Figure 23 shows the test rotor
carried out. The authors conducted a 3-D model study of the LP with the installed telemetry system. For reference, the cross
stage blades (actual/ scaled model) to verify the predicted section of the telemetry system is shown. The vibration
natural frequency mode shape of the scaled low pressure stage response of the blades was available during testing through this
blades. As a result, measured values for each mode showed a system. Figure 24 shows a photo taken during measurement of
±3% variation, compared with the calculated value. the vibration response.

For verification of LP blade performance, the authors measured

the Pressure and Temperature of each stage in the turbine
casing and the turbine power through a coupling torque meter.

Figure.21 Verification Test for integrity of actual scale blades

Test Turbine
Figure 22 shows the test facility set-up schematic used for the Figure.22 Verification Test for integrity of actual scale blades
scaled LP blade row verification. The major specifications of Full Load Blade Excitation Test
the test turbine and details of the manufactured test turbine and
compressor are shown in this figure.
The test turbine was planned as a 5-stage straight condensing
type, with 1 control stage and 4 LP stages. Its size was selected
considering a steam flow for blade loading at the most severe
case. For the inlet steam condition adjustment, a pressure
control valve and temperature decrease devices were installed
on the inlet steam line. Furthermore, the exhaust pressure was
adjusted, and a vent valve was installed on the condenser. A
2-stage centrifugal type impeller was selected for the test
compressor and air was used as the operating fluid in a closed
loop.

Special Measurements
During the test, the authors took some special measurements
for verification of the LP blade rows, including excited blades
frequencies, vibration stress and flow path performance, in
Figure.23 Verification Test for integrity of actual scale blades
addition to normal measurements (speed, steam flow, pressure,
Full Load Blade Excitation Test
temperature on piping, shaft vibration, etc.). In particular, to
verify the LP blade rows vibration response, the authors Test Conditions
computed blade vibration stress using strain gauges and The applicable operation range of the type 14 LP blade rows
thermocouples installed on the blades with a telemetry system. was evaluated for the case of a 1/4 scale test. Measurements
Figure 8 shows the special wiring for the verification test. were done at almost the whole applicable operating range.
Strain gauges were installed on the blade profile, tip or mean Turbine power and exhaust pressure at MCR (14,400 rpm)

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

were adjusted and a stable operating condition was confirmed. Typical Damage and Deterioration Maps
The speed was then decreased to Minimum Governor Speed Figure 25 shows the typical damage modes of mechanical drive
(MGS, 11,520 rpm), and then increased to MCR again and steam turbines. Steam flow path parts such as nozzles,
stable operating conditions were confirmed. During the stable diaphragms and blades are eroded by solid particles in the HP
condition at MCR, performance measurement was conducted, section and water droplets in the LP section. Corrosion fatigue
and as the speed changed, blade vibration response was failure can occur in the blade of the LP section at the transition
measured (twice for each measurement) zone of wet and dry conditions due to a corrosive environment.
These measurements were performed for a total of 50 hours of Excessive load due to drain carry over and poor quality of
operation over seven days. Additionally, 1.5 hours of 110% lubricant oil can induce melting damage to bearing pads.
load test at harmonic resonant speed was conducted to verify
blade stress cycles in excess of 1×107.

Figure.24 Verification Test for integrity of actual scale blades Figure.25 Steam Turbine Deterioration Map
Full Load Blade Excitation Test
Test Results Figure 26 shows the compressor deterioration map. Depending
Figure 24 shows the measured Campbell Diagrams of the L-0 on the actual operating conditions, process gases can become
last stage blades. In these diagrams, the size of the circle wet and corrosive. As a result, the impeller tends to have flow
represents the vibration stress levels. By changing conditions path erosion, cracking at stress concentration critical areas and
during the verification test the authors evaluated a variety of heavy chemical fouling which causes performance deterioration.
vibration response stress levels including harmonic resonance, At the balancing line, the differential pressure changes from
low vibration modes (below 20 harmonic), nozzle wake design condition, excessive thrust force acts on the thrust pads
resonance, high vibration modes, and random vibration levels and thrust bearing pads will see high metal temperatures. In
from a low exhaust vacuum condition (L-0 stage). worst case scenario pads can be damaged and compressor train
may need to be tripped. Dry gas seals can have damage due to
The authors then checked and obtained vibration response from oil mist invasion or lower than dew point temperature of
harmonic resonance to low vibration mode (below 20 supply.
harmonic) test results. The diagram shows a comparison
between calculated and measured results. This confirmed the
natural frequency of the low vibration mode of the scaled
blades, which agreed with the study results. (Other LP stage
results also agreed.)

The largest resonant stress measured during the verification test

for each vibration mode was tabled, and a safety margin of L-0
stage for each vibration mode was calculated based on an
allowable stress and weakest point stress. This was considered
as an actual stress measured point (tip or mean edge) in each
vibration mode. This enables the evaluation of the safety
margin for all blades. In the case of the L-0 stage, the safety
margin of the 1st mode was the lowest, but met design
requirements. Figure.26 Compressor Deterioration Map

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

Practical Countermeasure Technologies Applied diaphragms, have decades of operational experience. Machines
Flow Path Surface Treatment utilizing these countermeasures have achieved long continuous
Figure 27 shows the applicable technologies for long term operation. Without major maintenance, 24 units, including
operation. Key technologies for steam turbine reliability drivers used in LNG plants, are running for continuous periods
improvement such as special blade coatings have been in excess of 12 years. Therefore, the turbines can be operated
developed. In addition, applications of stainless steel materials with longer maintenance intervals which will lead to
to protect against drain erosion on the diaphragms, and the achievement of higher availability of the plant.
development of hybrid coatings of fluorine resin and ceramic
have been utilized. One of the essential factors for long-term
operation is coating technology. Regarding the deterioration Effective On-Line Washing
phenomenon of internal parts, the two major causes are solid To compete in today's economic climate, petrochemical plants
particle erosion and drain erosion. In the case of solid particle are strategizing on continuous long-term operation to reduce
erosion, boronizing heat treatments are applied for high maintenance costs and increase productivity. This strategy has
pressure stage nozzles, particularly the 1st stage nozzle. On the led some plants to go from eight years between turnarounds to
other hand, in case of drain erosion, Cr-TiN coatings and PTA 10 years. For rotating machinery such as mechanical drive
welding are applied. steam turbines, one factor that affects this strategy is heavy
deposition on steam turbine internals, caused by impurities in
the steam. These impurities result in fouling on the blade and
nozzle path surfaces due to contaminated materials such as
silica and sodium as shown in Figure 29. As a result, turbine
performance tends to deteriorate gradually.

As a countermeasure to performance degradation an innovative

on-line washing technique can be utilized to minimize the
impact caused by fouling of the steam path for large multi-stage
condensing steam turbines. This technique, although applied
here to extracting-condensing turbines, is also applicable to
large condensing turbines. The new technology has water
injection nozzles located at the steam chest of the extraction
control valve rack. The injection nozzles are plumbed to a
water supply source, which controls a set point temperature by
Figure.27 Applicable Technologies for Long Term Operation controlling the water injection rate. The objective is to directly
wash off deposits adhering to the blades and nozzles on the
low-pressure side with minimal power turndown, and without
impacting the turbine's long term performance. Erosion damage
and thermal stress of internal parts such as chest valves and
blades due to the injected water had to be taken into
consideration. To properly achieve this objective, and
considering the potential for damage during the online wash, a
new extraction valve box had to be designed.

The new design had to consider the effects of optimizing the

mixing zone of the steam and water injection to generate a
specific particle size, moisture propagation through the
condensing section, mechanical deflection of stationary
components, and the overall thermodynamic analysis of each
stage during the on-line wash. A prototype model was built and
several experiments were carried out based on the practical
Figure.28 Experience for Long Term Operation operating conditions of actual steam turbines. Analysis (FEA)
was used to evaluate the strength of the internal parts during
actual online washing. The final design was a compact
Figure 28 shows experience using applied technologies for long extraction box that could replace existing models without any
term operation. Key technologies, such as boronizing heat machining of the casing.
treatment, ion plating, Stellite brazing and stainless steel

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 Figure.29 Background for development of On Line Washing Figure.31 Performance Recovery by Online washing
To keep plant availability, the online washing system is
effective to wash off the deposits without a shut down. The
washing system can be furnished on steam turbine itself, not on Figure 32 shows the typical effect of steam turbine online
the inlet steam piping; it does not damage any other function, washing at field execution. The result shows high power
but improves the expected steam turbine performance. Figure improvement after online washing, roughly a 6% recovery
30 shows a drawing a map of typical online washing procedure. when compared to before washing.
The steam turbine online washing system has already been used
and applied over 30 existing turbine applications. First, hot
steam comes from the high pressure section and the steam is
mixed with injected water through spray nozzles in the turbine 10
95

valve chest. The mixed steam is controlled in order to avoid the ) 94

(%
saturated condition, and fouling material on the nozzle/blade g93
in
rows is washed by the steam in the wet condition. l 92
u
o
F91
m
ro
f 905
y About 6% recovery
r89
e
v
o88
c
e
R87
r
e
w86
o
P 2 Weeks
4 Weeks
850
Before On-Line Washing After On-Line Washing
Time

Figure.32 Typical Effect of Steam Turbine On-line

Figure.30 Online Washing for Steam Turbines

Figure 31shows typical operation data during online washing.
The lower axis is steam flow, and the left axis is after 1st stage For the compressor, Figure.33 shows a typical fouling
pressure for LP section. Before online washing, after 1st stage condition for a charge gas compressor in an ethylene
pressure is over the designed line, due to fouling condition. application. Heavy hydrocarbon deposits will accumulate at the
Though pressure gradually decreases during online washing, flow passage if the gas temperature rises above 90 °C. This
the pressure becomes stable in below the designed line after fouling causes large flow path losses and a reduction in
on-line washing. It means that performance was improved by efficiency, and requires more power from steam turbine driver
on-line washing. to maintain output.

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 Figure.35 Expected Scenario of Oil injection at Return Bend

Figure.33 Typical Fouling in Charge Gas Compressor

Figure 36 shows results of fouling survey at site successful
oil/water injection. This is the result of a fouling survey at site.
Due to effective oil injection, fouling of internal passages was
CFD analysis was conducted in order to characterize the minor. Internal hardware was very clean as shown on these
effectiveness of wash oil injection from the return bend, shown pictures.
in Figure 34. As an initial result, the injected oil does not
distribute remarkably. The effective oil wash area is only 3.6%
of the return channel even if the quantity is increased. Another
more effective injection method is necessary.

Figure.36 Results of Fouling Survey at Site Successful Oil

Wash Injection
Combination of Anti-Corrosion and Erosion Prevention

A typical example of applications for robust technologies is

shown in Figure 37. Application specific coatings are used to
provide additional improvement against steam turbine
Figure.34 Verification of Oil injection at Return Bend efficiency deterioration. ISB designs were applied for all
stages (impulse and reaction types) to increase overall
efficiency and prevention of corrosion fatigue in the dry/wet
transient zone. For the inlet 1st stage nozzles a boronizing heat
Figure 35 is the expected scenario of oil injection. First, oil is treatment was applied to prevent solid particle erosion and L-1
injected from return bend, but oil coverage is not expanded and last stage blade surfaces were treated with TiN ion plating
remarkably. However, when the oil reaches the impeller it and PTA (plasma transfer arc welding) respectively to prevent
washes the fouling deposit on the rotating impeller. After that drain erosion. In addition, wide pitch nozzles were applied for
the oil particles are scattered by the impeller. Therefore, the oil all stages to increase stage efficiency and to minimize after
coverage is expanded widely to the next stage. stage pressure due to fouling. In Figure 40, the difference in

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

after stage pressure increase due to deposits on the nozzle
profile is explained by comparing a fine pitch nozzle and wide
pitch nozzle. If deposit thickness increase rate (mm/year) is the
same, the nozzle area reduction rate of wide pitch nozzles can
be decreased by half because of the smaller number of nozzles
or smaller profile surface area than the fine pitch nozzle. The
after-stage pressure increase is proportional to steam flow and
the turbine can have a wider margin for operating time before
reaching the pressure limitation under fouling conditions, while
maintaining the required power.
Furthermore, if online washing is executed periodically,
the operating time can be further extended by performance
recovery. In this case study, the power recovery ratio after
online washing is assumed to be 80%. The power decrease rate
due to after-stage pressure increase caused by fouling is half of Figure.38 Anti-Fouling Coating for Compressor
that of a fine pitch nozzle when using a wide pitch nozzle. If Figure 39 is a typical example of an anti-erosion coating
newly developed coating and other improvements, including applied by high velocity oxygen fuel (HVOF) spraying. After
online washing and wide pitch nozzles as explained above, are four years of operation the 4th and 6th stage impellers had
applied systematically the turbine integrity and performance corrosion damage. Before changing the materials to
can be maintained as same as the original even after 8 years of anti-corrosion materials, for the spare rotor impellers of these
operation. stages an anti-corrosion coating by HOVF was applied.

Figure.39 Anti-Erosion Coating HVOF Running Rotor

Figure.40 shows the evaluation results of anti-erosion factor for
applied HVOF coating performance. The best composition of
coating was decided from the point of view of anti-corrosion
and erosion.

Figure.37 Surface Treatment Application for Steam Turbines

Figure38 is the typical fouling condition for charge gas

compressors in ethylene service. To avoid fouling, water
injection is done to keep the gas temperature low by water
vaporization. A hydrophobic, low surface friction layer is
applied as an anti-fouling coating for compressor flow paths to
avoid the accumulation of polymer on internal surfaces for
minimum efficiency losses. Oil (wash media) injection is done
to clean the compressor internals and retain high performance Figure.40 Evaluation of Anti-Erosion Factor for Applied
without overhauls. Coating HVOF Performance

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

Stage Performance Enhancement by Partial Component
Replacement

In case that existing machines need to be modified according to

a desired process change or plant capacity increase, two options
are proposed: one is a footprint replacement, casing with
internals replacement; the other is the internal replacement of
the rotor with impellers and stationary parts such as diaphragms.
Figure 41 demonstrates the case of retrofit design of a flow
field for a steam turbine LP section. Only the nozzle (stator)
profiles of the diaphragms of the last 3 stages are modified to
increase LP section efficiency. In this case study, total
efficiency could be improved by 1.5%.
Figure.42 Drain Erosion Prevention for Steam Turbine LP
Section

NDE Techniques
For long term operation, surface treatment technologies are
applied. NDE techniques are also developed to inspect the
internal damage of the rotor disk groove. One of the latest
technologies is PAUT, phased array ultra-sonic testing. In
contrast to traditional straight-beam UT, phased array sends and
receives multiple UT signals simultaneously, allowing for
greater detail and resolution. Based on measured crack length,
location, and orientation the allowable residual operational life
can be estimated for turnaround planning. Figure.43 shows the
results of a PAUT inspection for steam turbine stress corrosion
cracking (SCC) at the inside of the groove under the blade
assembly.

Figure.41 Retrofit Design of Flow Field for Steam Turbine LP Figure 44 shows a section of part of steam turbine disk and an
Section actual stress corrosion crack fracture analysis result. According
to these results, the PAUT measurement accuracy of crack
length and location is verified.

Figure 42 shows drain erosion prevention for a steam turbine

LP section. The last stage moisture is about 10% to 14% and
20% to 30% of this moisture becomes water. Small water
droplets of 50 to 300 micrometer diameter scatter to the blade
with a high velocity and attack the blade profile surface. The
number of such droplets has an adverse effect on the turbine
last stage efficiency. If the number of such droplets is reduced,
the drain damage is attenuated and efficiency can be increased.
This nozzle profile has a slit to gather the condensation on the
nozzle surface and an intake hole is provided in order to extract
the drain outside of the flow path.

The geometry of the structure such as slit and hole locations is

decided by analysis of the results of a 2-phase CFD (steam and Figure.43 PAUT Phased Array Ultra Sonic Test for Steam
water droplets). Turbine SCC

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 often nozzle positions will not fit in the existing main gas pipe
positions and/or do not set in the narrow openings of the
existing foundation. Modification of foundations significantly
increases the difficulty, time, and cost of the installation.

Figure.44 Typical SCC Crack for Steam Turbine Rotor Disk

Unique Casing Replacement Technique on Same Footprint

Figure 45 shows the basic steps for plant revamping Figure.46 Replacement for Plant Capacity Increase
considerations. As a first step, the end user should consider
whether or not process improvements alone can satisfy their In terms of the cost side of revamping versus replacement,
rerate targets. If so, this can be accomplished without Figure 47 shows a comparison of initial investment cost based
equipment modifications. If process adjustments alone do not on 300 kiloton per annum ethylene plant. It is easily imaginable
satisfy the requirements, the second step in this process is to that the equipment cost of replacement is higher than that of
consider re-rotoring the existing equipment. However, when revamping.
mechanical revamping of the existing equipment will not
achieve the plant rerate target, the end user will most likely
need to consider equipment replacement and/or parallel train However, it is less well known that the total cost difference is
applications as shown in the slide. not so large - only 18%. In the meantime, a 40% capacity
increase could be achieved by replacement, while only a 20%
capacity increase is achieved by revamping. So the initial
investment cost difference may be able to absorbed by the
profit difference from the production increase between the two
cases. For the case at hand (300 KTA plant), production
increase by replacement is 120 kiloton per annum.

Figure.45 Basic Steps for Plant Revamping

In the replacement approach, an existing compressor is replaced

with a new high efficiency design. In this case shown in Figure
46, it is important that the new large compressor is designed to
allow the casing to be put on the existing foundation and the
nozzles to set in the opening of the existing foundation. When a Figure.47 Comparison of Initial Investment Cost
standard design compressor is applied to a replacement project,

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 REFERENCES
[1] Saga.M, Hata. S. et al., Repair Technologies of Mechanical
Drive Steam Turbines for Catastrophic Damage, Proceedings of
the 34th Turbomachinery Symposium (2005), pp. 15-23.
[2] The ASME Handbook on Water Technology for Thermal
Power Systems p12
[3] J Jonas Otakar, Understanding Steam Turbine Corrosion 84
(1984) pp55 2 -55 28
[4] Ebara Ryuichiro, et al., Corrosion Fatigue Behavior of 13Cr
Stainless Steel for Turbine Moving Blade, Mitsubishi Heavy
Industry Technical Bulletin Vol.15 No.3 May 1978, pp1-10
[5] Naumann Hekmut G., Steam Turbine Blade Design
Options: How to Specify or Upgrade, Proceedings of 11th
Turbomachinery Symposium, pp29-49
Figure.48 Flexible Casing Nozzle Design for Compressor [6] Kihara Hiroshi, Fractography
[7] Wagner L.F, Griffin J.H, Forced Harmonic Response of
Figure 48 shows a flexible casing nozzle design for a Grouped Blade Systems: PartⅠ-Discrete Theory, Transaction
replacement compressor. The nozzle design is very important of the ASME, Vol.118, January 1996, pp130-136
for compressors. Compressor cases are typically formed by [8] Ebara Ryuichiro, et al., Corrosion Fatigue Strength of 13 Cr
casting. Therefore, the configuration of the nozzle is very Stainless Steel for Steam Turbine Blades, Mitsubishi Heavy
flexible if the location requirements are considered early in the Industry Technical Bulletin Vol.19 No.5 September 1982,
design process. This is especially useful for a footprint pp537-543
replacement project. If the replacement is properly done, the [9] Ebara Ryuichiro、Japan Steel Society , Nishiyama
base plate can be reused and the process piping positions Memorial Technical Session, Environment Crack of Steel,
remain fixed. Therefore, the compressor nozzle has to be October, 1982
flexible in order to replace the existing compressor to larger [10] Ebara Ryuichiro, et al., Corrosion-Fatigue Behavior of
one without significant and costly modifications. The flow of 13Cr Stainless Steel in Sodium-Chloride Aqueous Solution and
the nozzle should be analyzed by CFD analysis in order to Steam Environment, American Society for Testing and
ensure the design is successful. Materials, Corrosion-Fatigue Technology, pp155-168
[11] Steltz W.G, et al. Verification of Concentrated Impurities
in Low Pressure Steam Turbines, Transactions of the ASME,
Vol.105, January 1983, pp192-198
CONCLUSIONS [12] Bischoff James L, et al. Liquid-Vapor Relations for the
System NaCl-H2O: Summary of the P-T-x Surface from 300Ԩ
Turbomachinery manufacturers have been going forward to to 500Ԩ, American Journal of Science, Vol.289, 1989,
apply challenging technologies for critical components in order pp217-248
to provide highly efficient and more reliable machines [13] Allmon.W.E. et al., Deposition of Salts from Steam,
corresponding to the requirement of higher power for larger International Water Conference (1983), pp127-138
ethylene services. The authors introduce the technology [14] Maday M.F. et al., Stress Corrosion Cracking Behavior of
development history map in past and future for typical design, Two Ni-Cr-Mo-V Steels in Caustic Solutions and Pure
verification test and application results in terms of transient Oxygenated Water, Corrosion Vol.45, No.4, pp273-281
fluid dynamics, thermodynamics, rotor dynamics, and blade [15] Ebara Ryuichiro .et al., Corrosion Fatigue Process of 12
vibration strength evaluation. In addition, after Cr-Stainless Steel, ISIJ Inter-national, Vol.30 (1990), No.7,
commissioning, turbomachinery tends to have deterioration of pp535-539
performance and reliability for long-term operation. The typical [16] Katayama Kazuso, et al., High-Speed and Large Capacity
damage and deterioration map for machine are introduced and Compressor-Driving Turbines for Chemical Plants, Mitsubishi
the authors explain the practical technologies actually applied, Heavy Industry Technical Bulletin Vol.14 No.5 September
such as flow path surface treatment, effective on line washing, 1977, pp1-９
combinations of anticorrosion, erosion prevention and stage [17] Bhat. G.I., Hata. S. et al., New Technique for Online
performance enhancement by partial component replacement, Washing of Large Mechanical–Drive Condensing Steam
NDE techniques for both compressors and steam turbines, and Turbines, Proceedings of the 33rd Turbomachinery Symposium
unique casing replacement techniques on an existing footprint (2004), pp.57-65.
for revamps to increase capacity.

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

[18] Hata. S. et al., Recent Technologies for the Reliability and ACKNOWLEDGEMENTS
Performance of Mechanical-Drive Steam Turbines in Ethylene
Plants, Proceedings of the 34th Turbomachinery Symposium The authors gratefully wish to acknowledge the following
(2005), pp.15-23. individuals for their contribution and technical assistance in
[19] Hata, S., Sasaki, T., Ikeno, K., New Technologies of analyzing and reviewing results, and for great suggestions and
Synthesis Gas Compressor Drive Steam Turbines for Increasing leading of practical applications and testing at site: The team
Efficiency and Reliability, Proceedings of the 31st members of the steam turbine design section and compressor
Turbomachinery Symposium, Houston, (2002), pp.75-83. design section of Mitsubishi Heavy Industries Compressor
[20] Hata, S., New Technique for Online Washing of Large Corporation.
Mechanical–Drive Condensing Steam Turbines, Proceedings of
2007 AIChE Spring National Meeting Ethylene Producer
Conference, Houston, (2007).
[21] Hata, S., Case Study: Assessing a New Technique for
Online Washing of Large Mechanical–Drive Condensing
Steam Turbines, Proceedings of ROTATE 2007, UAE Abu
Dhabi, (2007)
[22] S. Hata, T. Hirano, N., T. Wakai, H. Tsukamoto, New On
Line Washing Technique for Prevention of Performance
Deterioration due to Fouling on Steam Turbine Blades (1st
Report: Fouling Phenomena, Conventional Washing Technique
and Disadvantages), Transaction of JSME Div. B, Vol.72,
No.723, November 2007, pp2589-2595.
[23] S. Hata, T. Hirano, N., T. Wakai, H. Tsukamoto, New On
Line Washing Technique for Prevention of Performance
Deterioration due to Fouling on Steam Turbine Blades, (2nd
Report: Basic Experiment and Analysis for Actual Operation of
Washing System), Transaction of JSME Div. B, Vol.72,
No.724, December 2007, pp2970-2979.
[24] S. Hata, T. Miyawaki, N. Nagai, T. Yamashiata, H.
Tsukamoto, Study on Corrosion Fatigue Phenomena of Low
Pressure Blades and Integrity Improvement for Mechanical
Drive Steam Turbines (1st Report: Relation of Corrosive
Chemicals Enrichment Zone and Corrosion Fatigue),
Turbomachinery, Vol.35, No.2, February 2007, pp8-16.
[25] S. Hata, T. Yasui, K. Yamada, H. Tsukamoto, Study on
Corrosion Fatigue Phenomena of Low Pressure Blades and
Integrity Improvement for Mechanical Drive Steam Turbines
(2nd Report: Operation Life Extension by Blade Surface
Treatment and Coating), Turbomachinery, Vol.35, No.3, March
2007, pp50-59.
[26] S. Hata, Blades Improvement for Mechanical Drive Steam
Turbines, Coating Technologies applied for Nozzle and Blade
of Mechanical Drive Steam Turbine, Turbomachinery, Vol.29,
No.5, May 2001, pp40-47.

Copyright© 2016 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station