insights - Dresser-Rand

insights
Editorial Statement:
“insights” is a periodical publication of the
Dresser-Rand Company. Its editorial
mission is to inform our readership in the
areas of energy industries, as well as
business, and world affairs that have an
impact upon our mutual concerns.
Comments, inquiries and suggestions
should be directed to:
Janet Ofano
Communications Coordinator
DRESSER-RAND insights Editorial Office
P.O. Box 560
Olean, New York 14760 USA
Phone: (716) 375-3000
FAX: (716) 375-3178
10077 Grogan’s Mill Road, Suite 500
The Woodlands, TX 77380 USA
Phone: (281) 363-7650 FAX: (281) 363-7654
www.dresser-rand.com
©Copyright 2000 Dresser-Rand Company
A PUBLICATION OF THE DRESSER-RAND COMPANY
insights
VOLUME 3, NO. 2
Featured in this issue of insights:
New Magnum Valve: Reliability With A
Common Element
First Field Test Installation On Gulf Of Mexico Deepwater
Platform For Multiphase Turbine Technology
Compressed Air Energy Storage: A Technology
Whose Time Has Arrived

insights
VOLUME 3, NO. 2
CONTENTS
1 Candid Visions:
Dresser-Rand's president, Vince Volpe, details advancements in technology to better serve
expanding global marketplace.
2 New Magnum Valve: Reliability With A Common Element
Dresser-Rand’s Magnum valve establishes benchmarks for reliability, long life, and affordability.
4 Configurator Cuts Cycle Time, Shortens Long-Range Planning
Dresser-Rand’s on-line Product Configurator has cut equipment delivery time in half.
6 First Field Test Installation On Gulf Of Mexico Deepwater Platform For Multiphase
Turbine Technology
MPPT installs Two-Phase Rotary Separator Turbine onto an offshore oil platform in the Gulf of Mexico.
9 Expanded Scope Solutions Offer Single Source For Dresser-Rand Control Systems Clients
Total project expertise meets client needs for single supplier.
10 Compressed Air Energy Storage: A Technology Whose Time Has Arrived
Innovative power generation method stores excess electricity during low peak periods,
dramatically reducing the cost.
13 Distant Equipment Monitoring At Your Fingertips; Models, Active Simulation Available On-Line
Dresser-Rand’s clients are now making use of the system that allows them to monitor equipment and
controls performance from anywhere in the world.
14 Engineer’s Notebook: Performance Evaluation And Fluid Flow Analysis In Low Flow Stages Of
Industrial Centrifugal Compressor
Dresser-Rand engineers describe flow field and performance of centrifugal compressor stages.
19 Dresser-Rand Expands Service Capabilities For Reciprocating Compressors And
Integral Gas Engines
New Applied Technology Department provides service and analysis for all process
reciprocating compressors and integral gas engines.
20 Global Visions:
Dresser-Rand Wins Contract For Combined Cycle Power Generation
Alliance With Liburdi Engineering To Extend Life Of Power Turbine Components
Dresser-Rand Forms Alliance To Improve Air Quality In China
Engeturb Offers Full Service For Latin American Clients
Dresser-Rand Publishes Guide To Naval Steam Turbines, Compressors And Blowers
Dresser-Rand Field Support Services Brochure Available
www.dresser-rand.com
For more information on Dresser-Rand, as well as technical articles and past issues of
insights magazine, visit our website at www.dresser-rand.com.
Editor's Note:
In this issue of insights,
the Candid Visions column is
written by Vince Volpe, who
was recently named president
of Dresser-Rand Company.
Volpe succeeds Dave Norton
who announced his intention to
leave Dresser-Rand after
more than 30 years of service.
To the average person, the
new technology era consists of
the Internet, mobile communications
and buying stocks from
an on-line broker, or having the
capability to order just about
anything from the vast world of
merchandising dot coms.
Indeed, those are nice things
to be able to do, but little do
most people realize that their
new lifestyles are the result of
industries who have taken
great technological leaps that
keep pushing the technology
edge even further. The real
advances, often overlooked by
the public, are those made by
the major industries of the
world.
In this issue of insights, we
examine two major technology
breakthroughs that Dresser-
Rand has developed that will
greatly aid our clients. The
results include reduced costs
for clients, reduced cycle time,
and reduced downtime. Both
innovations were introduced to
the marketplace in the last
12 months.
At Dresser-Rand Control
Systems in Houston we
launched distant equipment
monitoring through our Global
Access MC system. The other
is the Product Configurator,
which enables our product
engineers to make a firm bid
within two to five days right in
the client’s office – or from just
about anywhere in the world.
Internally, we have introduced
our Enterprise Execution System
(EES), which is a proprietary
computer technology to
customize project planning
and execution. This allows us
to continue our ongoing efforts
to reduce cycle time and
improve delivery time. (From
1997 to 1999, Dresser-Rand
has reduced cycle time by
45 percent, and improved ontime
delivery from 55 percent to
more than 90 percent.)
The Configurator is a software
package that allows our sales
team to meet with clients to
determine their needs and
requirements. With a laptop
computer our engineers, with
crucial input from the client,
can provide the client with not
only a bid, but also a set of
drawings and essential data,
and be ready to manufacture
as soon as the contract is
signed. Imagine how that
helps to speed up cycle time.
Our engineers can access our
base computer in Olean, New
York, from anywhere in the
world – from a platform in the
North Sea to a liquid natural
gas plant off the South China
Sea. Further details on the
Configurator can be found
on page 4 of this issue.
While the Configurator has
far-reaching capabilities, so too
does our Remote Access
Monitoring. From any location
anywhere – an oil refinery in
China, a pipeline bringing
natural gas from Siberia to
western Europe – Dresser-
Rand can now monitor
equipment and controls
performance, providing instant
situation diagnosis and longterm
equipment health
analysis. Imagine how that
vigilance and quickness can
reduce or forestall downtime
(see details on page 13).
In the world of major equipment
engineered to order,
Dresser-Rand has taken the
world of high technology and
mobile communication to help
our clients bring their essential
products to market, to fuel an
expanding economy. ■
Vince Volpe
President
Dresser-Rand Company
11

Dresser-Rand engineers have
developed a new valve for all
models of reciprocating
compressors, incorporating
an innovative design that
establishes advanced benchmarks
for reliability, long life,
and affordability.
The Magnum valve can be
configured for a wide range of
operating conditions and
virtually any gas process.
With a precision guided valve
element and springs, the
Magnum valve increases the
reliability of moving parts for
long-term operation.
Backed by more than 100
years of valve design and
manufacturing leadership,
Dresser-Rand developed
the Magnum valve to
provide clients with a highperformance
valve that would
also give the company a
competitive advantage in the
compressor valve market.
Key to the success of the
Magnum valve is the use of
superior materials. “A highstrength
thermoplastic known
as PEEK ® (PolyEtherEther-
Ketone) is used for the sealing
2
New Magnum Valve:
Reliability With A
Common Element
element in the Magnum
valve,” said Joel Sanford,
supervisor of valve engineering
for Dresser-Rand’s Painted
Post, New York, operation.
“This high-strength valve
element can withstand high
compressor speeds and
pressure differentials while the
streamlined flow path optimizes
the effective flow area.”
In addition to the PEEK
sealing element, the Magnum
valve can be manufactured
using elements made of
alternate materials for exotic
services such as chlorine.
After selecting PEEK,
Dresser-Rand engineers
began to concentrate on
designing a reliable element
geometry. Realizing the
limitations of traditional
element designs, they used
finite element analysis (FEA)
to develop the unique design
of the Magnum valve’s sealing
element.
“Traditional poppet and plate
designs have geometries that
are subjected to bending under
normal operating conditions,”
Sanford said. “This bending
induces tensile stresses in the
element that can lead to
fatigue failures in certain
circumstances.” The Magnum
valve element has been
designed so that the stresses
are compressive rather than
tensile to improve reliability by
reducing fatigue failures.
After successfully testing the
element in a high-pressure
non-lube air compressor that
delivered 3,200 psi of
differential pressure at 300° F,
it was proven that this unique
design could survive in very
harsh operating conditions.
The Magnum valve also
employs a unique grid pattern
that balances seat and guard
areas with the lift area. The grid
pattern concept was enhanced
using Dresser-Rand’s valve flow
tester to optimize effective flow
area (EFA).
To ensure optimal performance
and reliability, the
design of the Magnum valve
springs was given much
consideration. “With the
rugged design of the other
components, it was critical for
us to put a lot of thought and
effort into the design of the
springs, as they could have
become the weak link,” said
Derek Woollatt, senior staff
engineer at Dresser-Rand’s
Painted Post operation.
To verify the spring design,
Dresser-Rand used a valve
endurance tester. “Conditions
were set to an extreme value
to produce rapid failures,”
Woollatt said. “The severity of
the test produced an initial
failure rate of 34 percent.
After the first modifications
were made, a three percent
failure rate was achieved, and
with the final production
design we successfully
achieved a zero percent
failure rate.”
An added benefit of the
Magnum valve is the proprietary
Dynamic Valve Analysis
(DVA) software used to
custom design every Magnum
valve. The DVA software
enables Dresser-Rand
engineers to design the valve
to meet specific application
requirements.
“By taking the basic pattern of
the valve and machining it to
any size and configuration,
cost and cycle time are
reduced because it takes less
time to design and manufacture,”
Sanford explained.
“Additionally, machine
tools are dedicated
to manufacturing
Magnum
valves.
Special tooling
was developed, and
processes were optimized.”
Using a common element
for all valves is cost effective
because it increases interchangeability
and availability of
replacement parts, minimizes
replacement parts, and
reduces parts inventory for
both Dresser-Rand and the
client.
A prototype design was made
for testing in the Dresser-Rand
development laboratory.
Three compressors and other
special test rigs were used to
prove the design’s worthiness.
After more than a year of
testing, Dresser-Rand
manufactured “development
valves” that were reviewed
and tested by clients under
actual field conditions. Based
on feedback from these tests,
Dresser-Rand engineers and
designers further refined the
design of the valve.
When the valve was released
to the market, it had been
installed in more than 100
cylinders around the world in
both lubricated and nonlubricated
compressors, with
gases ranging in molecular
weight from 2 to 71, speeds
from 327 to 1,000 RPM, and
pressures ranging from 35 to
4,000 psig.
Originally, Dresser-Rand
engineers began their
research as an effort to
improve performance and
reliability of existing valve
designs, and to extend the
application range to high
speeds and high pressures
at a lower cost. But one
innovation led to another, and
in the end it became the
Magnum valve – a creative
engineering design that has set
new benchmarks for reliability,
long life, and affordability for
Dresser-Rand and its clients. ■
Magnum Valve
Literature Available
A comprehensive brochure
detailing information on
Dresser-Rand’s new
Magnum valve product line
for process reciprocating
compressors is now available.
The field-tested valves have
proven to be superior in
reliability and performance.
Dresser-Rand applied more
than a century of design and
manufacturing experience to
create the Magnum valve and
make it compatible with all
models of reciprocating
compressors and an extensive
range of operating conditions
and gas processes.
The four-page brochure from
Dresser-Rand describes the
Magnum valve’s reliability and
high-performance characteristics.
A copy of the brochure may
be obtained online at
www.dresser-rand.com by
using the “Contact D-R” form,
or by emailing your request to
info@dresser-rand.com.
To request a brochure by
telephone, call (U.S.)
713-973-5353. ■
3

Configurator Cuts Cycle
Time, Shortens Long-
Range Planning
As part of an effort to cut
equipment delivery times,
Dresser-Rand has developed
a technologically complex
system that simplifies the
proposal process, allowing
the company to produce a
proposal in two to five days in
the client’s office.
Called the “Product
Configurator,” Dresser-Rand
product engineers can
compile client needs to devise
an engineered solution with
drawings ready for manufacture.
“We call it ‘building what
we bid,’” said Allan Kidd,
project manager, product
configuration.
“The Configurator is the
starting block in a sprint to
faster client solutions,” he
explained. “It allows us to
discuss with clients their
exact needs and convert
them into manufacturable
information on the spot.”
The Configurator is a software
package written by Dresser-
Rand engineers. It allows the
regional business team (the
product engineers, application
engineers, etc.) to select and
configure machinery and
services, establish a price,
and guarantee that it will build
what it has bid. And it can
actually be done in the client’s
office within two to five days,
depending on the project
scope.
“We can start building our
equipment as soon as a
contract is signed,” said Gary
Jacobik, quality assurance
manager on the Configurator
development team.
“There are no coordination
meetings required, as the
project application is developed
with the client in the
client's office. Normally it
would take 10 weeks to
complete the first issue of
engineering documents, then
meet with the client for
coordination meetings.
The time for preliminary
engineering is totally eliminated
with the Configurator.
With the Configurator, we’re
able to cut off at least six
weeks, reducing lead time
before manufacture by as
much as 16 to 20 weeks.”
Typical industry cycle time
from quotation to delivery is
12 to 14 months.
With the Configurator and
other timesaving solutions
that Dresser-Rand has
developed in recent years,
delivery time is reduced to six
months.
Highly Engineered
Products
Because Dresser-Rand’s
equipment is highly engineered
– individual for not
only each client, but also
each client application –
solutions for clients had been
complex and time
consuming. The Configurator,
however, allows Dresser-Rand
to speed up the process.
Informational fields are
populated across a series of
templates and modules. This
allows the product team to
quickly define the performance
and configuration of
Dresser-Rand’s entire line of
turbo equipment and package
services. The Configurator
can delineate all of the turbo
machinery features required.
“We automate the creation of
an entire project bill of
materials and services,”
Jacobik explained. “They're
developed for each part of the
compressor. A bill of materials
is developed for each component.”
The end result is that
at the close of the initial
meeting with a client, the
Dresser-Rand product
procurement team has a set
of drawings, data, and details,
coupled with a pricing
structure that has all of the
information that the client
needs to make a decision. “In
a matter of a few days in the
client’s office, he knows
everything he needs to know
about a Dresser-Rand
solution,” Jacobik said.
Although the Configurator is
unrelated to the EDS
Unigraphics computer aided
drafting system Dresser-Rand
developed, “we’ve designed
them so they shake hands,”
Kidd said. “You can type
information into the
Configurator and it shows up
on CAD.”
Launched in 2000
The Configurator was initially
launched in January 2000,
for pipeline booster (PDI)
centrifugal compressors.
DATUM centrifugal compressors,
packaging, and gas
turbine drive modules will be
added this summer and will
include lube oil systems, dry
gas seal systems, and gas
turbine modules. European
modules will be released
throughout 2000.
The completion date for the
entire Product Configurator
System, including mechanical
drive steam turbines and all
reciprocating products for
process and separable
applications, is June 2001.
Kidd estimates that the
Configurator is 65 percent
complete at this time. “We’re
way ahead of schedule,”
Kidd added. The deadline
next year does not include
expanders, axial compressors,
or power turbines.
With the Configurator, a
regional business procurement
team, which includes engineers
with expertise in all
areas of the equipment,
meets with a client.
“Typically, there will be four
or five people on a specific
project,” Jacobik explained.
The servers for the Windowsbased
Configurator are
located in Dresser-Rand’s
Olean, New York, facility. A
business procurement team
can access the server by
laptop computer from
anywhere in the world.
“Providing remote access
was the challenging part,”
according to David Prince,
president of Databranch,
Dresser-Rand’s consultant.
“Dresser-Rand is one of the
first companies ever to do
this,” Prince said. “This is truly
global in scale and that’s
what’s impressive about it.
Dresser-Rand is a true leader
in developing this technology
and deploying it worldwide.”
Client Reaction Positive
Initial reactions from the half
dozen clients who have
witnessed the Configurator
during the first four months of
2000 have been very positive.
“I spent some time with one
client recently and I couldn’t
tell who was more excited,
him or us,” Kidd recalled.
There are additional benefits
to clients other than the cycle
time cut from Dresser-Rand’s
end. “Those clients that
understand the principle of
cycle time reduction are also
interested in reducing their
own cycle time,” Kidd
explained. “We also shorten
the client’s cycle time – from
project concept to completion
– by giving them more
information sooner that's
accurate and reliable upon
receipt. We can sit with them
while they’re going through
their concept stage and, as
a result, shorten their concept
cycle time.”
This has a dramatic impact on
client satisfaction. “The clock
starts ticking when they first
start thinking about a project,
long before they call us to help
them with their energy
solutions,” Kidd said. “This is
significant when you consider
some of the time saved from
initial concept to plant start-up
– say, an LNG plant in
Malaysia, 300 to 400 days off.
The Configurator allows our
product engineer teams to
move with alacrity and
accuracy to fulfill client needs.”
Kidd concedes that the
Configurator and other efforts
by Dresser-Rand to reduce
cycle time give a whole new
meaning to long-range
planning. In fact, the long
range just got shorter. ■
4 5

6
First Field Test Installation On Gulf
Of Mexico Deepwater Platform
For Multiphase Turbine Technology
Link Dresser-Rand’s durable
and reliable machinery design
and manufacturing capabilities
with Kvaerner Process
Systems’ unparalleled
expertise in separation
technology and what do you
get? A unique multiphase
turbine that separates gases
and liquids and generates
power from energy previously
wasted in oil and gas
production.
Multiphase Power &
Processing Technologies LLC
(MPPT), formed in 1998, is a
joint venture between
Kvaerner Process Systems
and Dresser-Rand Company.
MPPT was created to
develop and market proprietary
multiphase turbines that
separate gas and liquids,
while generating power from
pressure let-down operations.
The biggest market is the
offshore oil and gas industry.
Offshore platforms possess
the unique combination of
high operating pressures,
multiphase (gas, oil, and
water) flow, and minimal
operating space.
MPPT’s equipment is smaller
than conventional systems,
and is based on biphase
turbine technology, acquired
from Douglas Energy
Company of Placentia,
California. Douglas Energy
works with Dresser-Rand and
Kvaerner as a consultant in
the development and
demonstration of the multiphase
turbine technology.
This equipment is designed to
enable oil and gas producers
to reduce the size of their
production platforms and
subsea modules, and to
generate power needed for
production, with no greenhouse
emissions. This
creates significant footprint,
weight, and total installation
cost savings.
Turbine separators use high
centrifugal forces (up to 5,000
times the force of gravity)
created by pressure letdown
and a rotating drum to rapidly
expel gas from liquid.
Pressure letdown occurs in
nozzles specially designed for
two-phase flow expansion.
Expansion energy of the gas
flashing out of the solution is
effectively transferred to a high
velocity gas-liquid jet mixture
that is used to spin a drum
connected to an output shaft.
The high centrifugal forces
from the rotating drum and
fluid result in effective separation
of gas and liquid within
two to three seconds – as
opposed to 20 minutes in
conventional gravity vessel
separators. This enables
more efficient compact
separation, with lower fluid
inventory (volume).
Originally, the turbine
separator was developed to
separate water and steam
and increase the total energy
produced at a steam turbine
facility in a geothermal
application. The transfer
of this technology from
geothermal to offshore oil and
gas was undertaken to meet
the demanding needs for
deepwater hydrocarbon
production.
An important milestone will
be realized for MPPT this
summer, with the installation
of a full scale Two-Phase
Rotary Separator Turbine
(RST) on an offshore oil
production platform. MPPT
will install this Two-Phase RST
skid package onto the Shelloperated
Ram/Powell tension
leg platform. This platform,
submersed in more than
3,200 feet of water in Viosca
Knoll Block 956, is 125 miles
southeast of New Orleans in
the Gulf of Mexico.
The Ram/Powell field test
marks the turbine’s first
installation into a live
producing facility. “The future
path of MPPT relies on the
test at the Ram Powell
platform,” said Hank Rawlins,
project engineer for MPPT.
The package will be field
tested for three months.
Design conditions are a liquid
flow rate of 20,000 barrels of
oil per day (BOPD) and a gas
flow rate of 30 million standard
ft 3 of gas per day (MMSCFD)
at an inlet pressure of 1,250
psig. Estimated power output
will be 200 kilowatts at full
flow conditions.
During actual well flow
conditions, the turbine will be
required to handle fluctuating
oil, gas, and water flow rates.
Other constituents that may
be present in the well fluids
are produced solids (sand),
scale, foam, wax,
asphaltenes, and hydrates.
Each of these conditions have
been factored into the design
of the turbine.
The skid was built in the
Kvaerner Process Systems
shop in Perth, Australia.
It is a self-contained, selfoperating
skid complete with
control, operating, monitoring,
and testing equipment all
wrapped into one package.
The skid package is 20 ft.
long, eight ft. wide and 10 ft.
tall, and weighs 19 tons. It is
designed to require minimal
operator intervention, and has
the capability of remote
monitoring of turbine performance
from Dresser-Rand’s
Wellsville, New York, facility.
The turbine that was
developed for this package
was originally tested at the
Texaco-Humble flow test
facility in 1997, and is suitable
for operation with live oil and
gas wells. Extensive factory
testing was further performed
at Dresser-Rand’s Wellsville
facility, on the hydraulic, safety,
operating, and control
systems to ensure full
mechanical reliability upon
package commissioning in an
offshore environment.
Dresser-Rand was in charge
of package, integration,
assembly, and factory testing
of the RST. The skid and the
turbine were shipped to the
Dresser-Rand facility in
Wellsville in early spring for
completion. Package
integration involved installing
the turbine into the package
and finishing the piping,
instrumentation, and control
systems.
“The purpose of the threemonth
test is to show that the
RST can efficiently separate oil
and gas, and that we meet
the predicted power output
with full mechanical reliability,”
stated Rawlins. ■
7

Could your bid evaluations use
a little wait reduction?
WE GIVE YOU ANSWERS IN HOURS, NOT WEEKS.
Now you can get custom energy conversion solutions in record
time, without even doing an RFQ. Our new Product Configurator
Software makes it possible. Its flexible design lets us work
directly with your engineers to quickly and easily analyze
various equipment scenarios. Then it generates complete
specs, drawings and a bill of materials, so you’ll know exactly what you’re getting
and what it will cost. We’ve streamlined the rest of our production process as
well, cutting cycle times by months. To learn how you can save time and money
while getting your new equipment on-line—and generating revenue—faster than
ever before, give us a call. We’ll show you how we can lighten your load.
North and
Latin America:
Tel: (713) 467-2221
Fax: (713) 935-3490
Europe, Middle East,
Eurasia and North Africa:
Tel: 33-235-25-5225
Fax: 33-235-25-5367
Asia-Pacific:
Tel: 60-3-253-6633
Fax: 60-3-253-2622
E-mail: ad@dresser-rand.com
©2000 Dresser-Rand Company
Expanded Scope Solutions Offer
Single Source For Dresser-Rand
Control Systems Clients
In order to maintain a competitive
edge, companies are
continuously changing their
business practices to remain
up-to-date with current marketplace
requirements, new
technologies, and shareholder
demands.
In the past, clients would
coordinate with several different
suppliers to bring together the
products and services that
made up an entire project.
Today this practice is obsolete.
Companies are now focusing
on single suppliers that can
effectively provide the same
products and services as
several vendors had in the
past in order to reduce
procurement, management,
installation, and start-up costs.
With control system technology
changing at such a rapid pace,
what was state-of-the-art 10
years ago may not be suitable
for today’s monitoring and
control requirements. In
addition to producing the most
advanced machinery control
systems, Dresser-Rand has
provided expanded scope
solutions for more than
40 years. These capabilities
include site studies to determine
the condition of equipment,
retrofit replacement planning,
engineering, training, project
management, and equipment
removal and installation.
Past projects include supplying
complete onshore and offshore
control systems and auxiliaries
for the compression of gas for
process plants, oil recovery
applications, and many other
services. Dresser-Rand has
also done complete retrofits of
plant control and auxiliary
systems.
Control Building Project
Expanded scope solutions vary,
but one typical example
demonstrates a retrofit project
that Dresser-Rand completed
on control and auxiliary
systems for eight Dresser-Rand
25,000 hp offshore compression
modules on two offshore
platforms in Latin America.
The control and electrical
equipment had been in
operation for more than
18 years. Many of the
replacement components were
obsolete and no longer
available, so the client was
experiencing problems with the
availability and reliability of the
compression modules. The
client needed a solution to this
problem and with the help of
Dresser-Rand, chose to retrofit
the entire system.
The solution involved the
replacement of the control
buildings, including HVAC and
building pressurization equipment.
Dresser-Rand replaced
the motor control centers,
generator control panels,
automatic transfer switches,
battery systems and chargers,
inverters, gas turbine driven
compression train control
panels, gas turbine fuel valve
and controls, fire detection and
suppression systems, and gas
detection systems. To reduce
downtime for the client, the
buildings were constructed
and completely tested on shore
before being placed on the
platform as a complete
assembly.
Dresser-Rand completed the
installation and interconnection
of the new control buildings and
finished up the project by
putting the client’s new
equipment on line.
Turnkey Project
On another platform – this one
off the coast of Nigeria –
Dresser-Rand performed a
complete retrofit of controls for
a compression system with
three gas turbine drivers.
This turnkey project required
engineering expertise to
determine how to integrate the
new control panels into the
client’s existing facility and
operation buildings, remove the
old control systems and install
the new ones, and start-up the
new system.
The control system’s availability
and reliability were increased by
using redundant (hot standby)
control system processors,
which allow for maintenance
and repair while the equipment
is running. Removal of the old
systems, installation of the new
systems, and commissioning
were executed in just 22 days.
As more businesses seek
total solutions from partnervendors
for complete project
management, Dresser-Rand’s
expanded scope capabilities
become increasingly valuable
to clients, and integral to their
long-term facilities management
plans. ■
8
www.dresser-rand.com/go/innovation Compressors Concept Engineering
Contract/Rental Compression Control Systems Expanders Field Repairs
Gas & Steam Turbines Operation & Maintenance Process Audits Upgrades
By offering expanded scope
solutions and upgrades for
compression and associated
ancillary equipment, Dresser-
Rand has uniquely positioned
itself to provide total solutions
to the oil, chemical, gas,
and petrochemical industries
through its innovative products
and services.
9

For nine years a prototype
compressed air energy storage
(CAES) plant in McIntosh,
Alabama, has been producing
up to 110 megawatts of
electrical power within 14
minutes of start-up during
periods of high-peak demand.
For nine years the concept has
been proving itself over and
over again, just waiting for the
right market alignment.
And that time is now.
“It’s what our nationwide grid
needs,” says Lee Davis, who
has been manager at the
McIntosh plant since the latter
part of the CAES plant’s
construction. “The electric
industry ought to be excited,
especially as the industry
deregulates.”
The McIntosh plant went on-line
in May, 1991 for the Alabama
10
Compressed Air Energy Storage:
A Technology Whose Time Has Arrived
1.
LEGEND
1. High-pressure compressor
2. Intermediate-pressure
compressor
3. Speed-increasing gear
4. Turning gear
5. Low-pressure compressor
6. Clutch
7. Motor/generator
8. Clutch
9. Low-pressure expander
10. Low-pressure combustors
11. High-pressure expander
12. High-pressure combustors
13. Turning gear
14. Air throttle valve
15. Air trip valve
Electric Cooperative (AEC).
AEC is a generation and
transmission cooperative that
serves member distribution
cooperatives and municipalities
in south central Alabama and
most of the Florida Panhandle.
Since 1991, construction of
new electric generating plants
has come to a standstill
throughout most of the United
States. “That’s why the grid is
in such danger right now,”
Davis says. He cited outages in
Chicago last summer as a
prime example of not having
enough power at the right time
in the right place. Davis thinks
CAES plants are the solution.
It works this way: during low
electric usage periods in the
night (off-peak periods) the AEC
CAES plant compresses and
stores air in an underground salt
2.
3.
4.
cavern. When electric power
demand peaks during the day,
the process is reversed. The
compressed air is returned to
the surface, heated by natural
gas in combustors, and run
through high-pressure and lowpressure
expanders to power
the motor/generator to produce
electricity. It takes less than 15
minutes to bring electric
generation up to capacity.
Using electrical energy to
compress air during off-peak
hours when the cost of
electricity is at its lowest allows
the utility to generate electricity
from the CAES-stored air and
sell it during peak periods. It
uses its CAES plant to boost its
power capabilities during peak
daytime periods when demand
for electric energy skyrockets.
“We buy cheap and displace
high-priced power to our
cooperatives,” Davis explains.
“Basically, I’m very much for the
CAES concept,” Davis says.
“Our load is primarily residential
and CAES fits well with our load
shape. We have a coal-fired
plant 20 miles up the road.
Excess electricity generated by
the coal-fired plant during offpeak
hours is used in McIntosh
to compress air for storage.
5.
“Normal startup for us is 14
minutes to 110 megawatts,” he
says. “I can run down to 10
megawatts. It’s a better
regulating tool.” Control of the
plant for power generation and
compression requirements is
accomplished via microwave
signal by the dispatcher 120
miles away from AEC’s
corporate complex in
Andalusia, Alabama.
In addition, the equipment has
“black start” capability. “Let’s
say you have a blackout. You
can start this equipment without
any outside power,” says Ivan
R. Lehman, product manager
for Dresser-Rand Company in
Wellsville, New York.
Integral to the McIntosh CAES
plant is Dresser-Rand, which built
most of the machinery for the
system. The company’s Olean,
New York, operations manufactured
the turbocompressors;
Wellsville operations produced
the gas turbine expanders and
developed the controls; Painted
Post, New York, operations
manufactured the fuel gas
compressors.
“It’s one of the longest trains of
equipment in the world,”
Lehman says. The 140-foot
train has a centrally located
motor/generator with clutches
at each end. On the motor
side, a low-pressure
compressor, intermediate
compressor, and high-pressure
compressor work to compress
the air into the salt dome up to
1,200 psig. Compressor
intercoolers are used at each
6.
of the three compression
stages to maintain equipment
efficiency.
And the equipment has come
to meet expectations in
McIntosh. “After undergoing a
few design modifications, the
Dresser-Rand equipment is
now performing excellently for
us,” Davis says of the CAES
plant he manages.
The equipment is from Dresser-
Rand’s various product lines
that have been time- and fieldtested
for decades in various
other applications. The product
7.
140'
line includes single stage
turbines, standard multistage
turbines, packaged geared
turbine generators and engineered
turbine generators,
centrifugal and axial compressors,
gas turbines, and
reciprocating compressors.
Single stage turbines operate up
to 2,500 kilowatts, and the
standard multistage turbines
operate up to 5,000 kilowatts.
Engineered power generation
units are up to 110 megawatts
plus, with an inlet pressure up to
2,000 psig and an inlet temperature
up to 1,000° F (540° C).
8.
9.
Turbine speeds are 2,500 to
20,000 rpm and generate
speeds for 50 cycle or 60 cycle
applications. Packaged turbine
generators come with three
condensing packages and two
back-pressure packages.
AEC used an existing underground
salt dome for
compressed air storage. “We
solution-mined it for 629 days,”
Davis recalls. That created the
space for 19 million cubic feet of
compressed air. For the solution
mining, fresh water was
pumped in to dissolve the salt.
10.
11.
The solution was pumped out to
a nearby chemical manufacturer
that uses brine in its processing.
Dresser-Rand’s Broken Arrow
facility supplied the portable
compressors that provided the
initial charge for the solutionmined
cavern.
The McIntosh plant uses
existing salt domes, but hard
rock caverns and aquifers can
also be employed. “Dresser-
Rand is negotiating with a
prospective client with an aquifer
that has a capacity in excess of
230 million cubic feet, compared
to McIntosh's salt cavern of
19 million cubic feet,” says
David Hargreaves, Dresser-
Rand’s director of business
development.
While the CAES plant has
performed reliably and met
AEC’s needs and requirements
for the last nine years, the
concept otherwise has sat idle
throughout the world since its
beginning. The time is now ripe
for CAES units to become key
components of utilities and
independent power producers.
“We’re involved in several
projects, in the development
stage of issuing proposals,”
Lehman says. And the
company is fielding inquiries
12.
Continued on page 12
13.
14.
11
15.

SPEED RPM
Compressed Air Energy
Storage: A Technology
Whose Time Has Arrived
Continued from page 11
from all parts of the world,
including projects in the United
States, Middle East, Far East,
and Asia.
“There has been a growing
power need around the world,”
Lehman explains. “Utilities have
traditionally met their requirements
for large blocks of power
during peak periods by adding
gas turbine/generators. But the
climbing cost of natural gas and
new emission regulations,
coupled with the long leadtimes
for gas turbines are now
forcing them to rethink their load
management tactics.”
Or, as Davis, the manager of
the McIntosh plant put it, “The
gas turbines had been so
cheap.” But times have
changed over the last nine
years. “Gas turbines are now
very expensive, and right now
because of high demand for
them, you can’t even buy a gas
turbine.”
Gas/turbine generators also
face another problem,
according to Hargreaves.
Output decreases as altitude
START-UP FOR POWER GENERATION
GENERATION
360
3000
240
1800
1200
600
12
0
1
SPE
POWER
EMERGENCY
START
POWER
NORMAL
START
2 3 4 5 6 7 8 9 10 11 12 13
TIME IN MINUTES
and ambient temperatures rise.
“A simple-cycle gas turbine will
put out 85 megawatts at ISO
conditions,” Hargreaves
explains. “The heat rate is
10,500 BTUs per kilowatt hour,
which comes to a cost of
three-plus cents a kilowatt hour.
When the temperature goes up
outside, the turbine loses
capacity. Air that goes into the
front of the turbine becomes
less dense. At 95 degrees the
gas turbine is providing 74
megawatts, and it begins to
take more energy to compress
air. The CAES unit will produce
its rated power in all ambients.”
A CAES unit, on the other
hand, uses less energy. “The
compressor system we use is
intercooled,” Hargreaves says.
“It takes 10 to 15 percent less
power and uses nighttime offpeak
electricity to do
compression. In essence,
CAES is a battery.” He
estimates that CAES-generated
power is produced at a cost of
two cents per kilowatt hour.
CAES equipment is engineered
in modular trains capable of
producing 100 to 135
megawatts of power. As few
as two modules upwards to
more than 15 modules are
being considered per location
to meet future expected local
110
100
90
80
70
60
50
40
30
20
10
POWER MW
demands. CAES is uniquely
qualified to provide reliable
power on a flexible basis to
meet all power requirements
from “intermediate and peak”
demand, including spinning
reserve and standby. Dresser-
Rand is currently discussing
four-module CAES plants with
START-UP FOR COMPRESSION MODE
360
3000
240
1800
1200
600
SPEE
This Alabama CAES facility, built by Dresser-Rand, has been
performing reliably since 1991.
0 1 2 3 4 5 6 7 8 9 10
TIME IN MINUTES
110
100
90
80
70
60
50
40
30
20
10
POWER MW
SPEED RPM START-UP FOR COMPRESSION MODE
EXPANDER
POWER
MOTOR
POWER
independent power producers
as merchant plants. “That’s
enough to produce over 500
megawatts of power at one
site,” Hargreaves says.
“Merchant plants merchandise
electricity for anyone who needs
it,” Hargreaves explains. “It’s
like the spot gas market.” It’s
an offshoot of the continuing
deregulation of the utility
industry.
Dresser-Rand is sitting in the
driver’s seat for future CAES
plants. It is the only company
with an enviable track record,
meeting the peak load periods
of the AEC on a daily basis.
Clearly, storing excess electricity
generated during low peak
periods is the ideal answer.
And CAES has proven for the
past nine years to perform with
unmatched reliability. ■
Distant Equipment Monitoring At Your
Fingertips; Models, Active Simulation
Available On-Line
Instant situation diagnosis and
long-term equipment health
analysis are now at the
fingertips of Dresser-Rand
clients anywhere in the world
who have signed on for remote
monitoring and control for the
operation of rotating equipment
systems.
With remote monitoring through
its Global Access MC
system, available since January
2000, Dresser-Rand clients are
now making use of this system
that allows them to monitor
equipment and controls
performance from anywhere,
said Tim Walker, Marketing
Engineer for Dresser-Rand
Control Systems in Houston.
Remote monitoring is available
through Dresser-Rand’s Global
Access MC system via Internet,
direct phone line, or by satellite
link. Other than a personal
computer, no special equipment
is necessary. “If they have a
computer and a net browser,
they don’t need any other
equipment,” Walker said.
Clients interested in the system
can log onto a static demonstration
model on the
company's web site at dresserrand.com
and the Global
Access MC demonstration link
on the Controls page and
navigate through the fixed
model.
The model contains areas for
comments and for requests
for additional information.
A Dresser-Rand representative
will contact the user and
provide a password for access
to the company's active
dynamic system on simulation
at the Control Systems’ product
development laboratory in
Houston, Walker said.
Remote monitoring holds vast
potential for both clients and
Dresser-Rand, Walker said.
“It facilitates the monitoring of
clients’ equipment. We’re
looking at reducing or eliminating
on-site service personnel,
which will very likely have a
major, favorable impact on cost
and the feasibility of long-term
service agreements.”
Remote monitoring provides
immediate turn-around for
solving problems. “Historically,
the client and Dresser-Rand
would go to quite a bit of
expense and time to even
evaluate a problem. With
Global Access MC, that difficulty
is eliminated. We can evaluate
the situation before a technician
gets on a plane. Equipment
and control engineers can
examine the data in counsel
with the client’s staff experts.
“We can see some things, and
log some events. We’re all
looking at the same data and
making a determination whether
maintenance or replacement is
required. We can determine
whether we need a mechanic
or a control system technician.”
The remote monitoring system
allows Dresser-Rand to expand
its Asset Management business
by offering better, faster service
for clients’ heavy capital
equipment. It also gives
Dresser-Rand the timely
informational wherewithal
to provide more accurate
projections of equipment health
and maintenance schedules.
Benefits to the user are
immediate. The continuous
acquisition of data through
Dresser-Rand’s Global Access
MC system provides equipment
operators with access to
temperature and vibration
displays, comprehensive analog
summaries and status lists,
surge control displays, current
and historical alarms, and
current and historical trends of
all analog signals.
The Global Access MC
server on the control
system includes
Windows NT ® with remote
access enabled, web server
software, and a 56k modem.
The remote PC requires
Windows 95/98 or NT ® ,
Netscape Navigator ® 4.07 or
higher, or Microsoft Explorer ®
4.01 or higher, and a 56k
modem. Connection between
the server and the remote PC
can be by phone line, Internet,
or satellite.
“There are ways to provide
security through dedicated
phone line or local area
network,” Walker said.
“Offshore platforms and remote
sites have satellite links and
microwave repeaters.”
From offshore platforms, to
pipelines, to processing plants
— all are candidates for remote
monitoring and control through
Global Access MC. ■
13

Performance Evaluation
And Fluid Flow Analysis In
Low Flow Stages Of
Industrial Centrifugal
Compressor
by Yuri I. Biba, David A. Nye
and Zheji Liu Dresser-Rand
Company, Olean, New York
Editors Note: This paper
was presented at the 8th
International Symposium of
Transport Phenomena and
Dynamics of Rotating
Machinary (ISROMAC-8),
6-30 March 2000.
Abstract
A comparative study of the flow
field and performance of
centrifugal compressor stages
is presented for low volume
flows and high-pressure
applications. Two different
impeller designs and stage
configurations are considered
and modeled using commercial
Computational Fluid Dynamics
(CFD) codes. Internal stage
designs are evaluated by
qualitative and quantitative flow
analysis with the goal being to
obtain more efficient stages.
The resulting improved
configurations are implemented
into one of Dresser-Rand’s
compressors. Computational
and experimental results are
discussed and conclusions are
made regarding the existing
model, as well as future
improvements, both in
modeling and design concepts.
Nomenclature
D 2 – overall impeller diameter
D 5 – diffuser outlet diameter
Q 0 – volumetric flow at stage
inlet
u 2 – impeller speed at overall
diameter
� – low coefficient,
�=4Q 0 /(�u 2 D 2 2 )
� - polytropic efficiency
� - polytropic head coefficient
Introduction
Centrifugal compressor
impellers for low flow coefficients
� are subjects of specific
interest due to their application
in process industry and gas
Figure 1. Cross Section of Last Three Stages of a High-Pressure
Centrifugal Compressor.
14
injection markets. In these
applications such impellers are
often used in the last stages at
high-pressure levels.
A systematic study of low flow
compressor stages has been
presented by Casey et al.,1990.
It is noticed that a simple
dissipation loss model was
unable to predict, with any
certainty, which of the impellers
of different design would have
the best performance.
Extensive experimental studies
for all impellers are necessary to
identify the most efficient one.
Impeller maximum efficiency
drops rapidly with the decrease
of design volume flow (Balje,
1981) making it very difficult to
improve stage efficiency for
lower flow values. That general
trend has not changed with
years of extensive development
efforts by major turbomachinery
manufacturers (Dalbert et al.,
1999). Dresser-Rand has been
using CFD to design new, more
efficient, families of low flow
impellers to apply in the last
stages of high-pressure
compressors.
Figure 2. Stage Performance at Various � and D 5 /D 2
Calculated by STGPERF.
Recent progress in CFD code
development and decreased
computing cost make more
detailed computational analysis
possible. At the same time,
experimental testing remains
very expensive and will likely
need to be reduced. CFD
modeling or a Virtual Test Rig
(VTR) approach is routinely
used, while experimental testing
remains a final verification tool.
Dresser-Rand has implemented
the VTR concept since the
company began its use of
CFD software. CFD has been
routinely used to compare
analytical predictions with test
results to augment test data and
validate predictions (Sorokes
and Koch,1996).
Many issues need to be
addressed before a sufficient
level of confidence is gained to
assess performance change via
CFD studies. Part of this
responsibility is certainly part of
the VTR study, where the effects
of grid density, windage, leakage,
surface roughness, frame size,
etc. will need to be addressed.
However, full confidence can
really only be gained through
calibration and comparison with
test data. This test data can
originate from extended scope
instrumentation on production
rigs or improved Single Stage
Test Rig (SSTR) testing. While
the VTR is being developed, the
new stages can be optimized
Figure 3. Typical CFD Grid Arrangement.
using CFD to assess additional
design improvements.
Background
It was decided to implement
a new design for the last three
stages of a multistage highpressure
barrel compressor.
Flow coefficients � for these
stages are 0.0146, 0.0117 and
0.0097, respectively. These
three stages (Figure 1) were
selected to replace original
impellers that are characterized
by relatively small backsweep
and narrow flow passages.
The existing designs are replaced
by the new impellers with wider
passages and a significantly
larger blade backsweep angle.
The intent of the design was to
reduce flow velocity and loading
throughout the stage, make the
impeller exit flow more uniform to
reduce stage losses, and
produce slightly less head than
the original design.
Preparatory study was
performed to determine the
optimum diffuser size. This
study used Dresser-Rand’s
proprietary semi-empirical, onedimensional
analytical code
STGPERF. A number of cases
were run on STGPERF to
investigate the influence of the
diffuser radius ratio. This ratio
was varied between 1.25 and
1.49 for stages with flow
coefficients from 0.0032 to
0.0275. Efficiency plots are
shown in Figure 2. The results
showed that for a stage
F=0.0275, the peak efficiency
was achieved at the larger
diffuser radius ratio. For the
lowest flow stage, peak
efficiency was achieved at the
lower value of diffuser radius
ratio, with indications that even
lower values of radius ratio may
be beneficial. For a stage � =
0.0069 STGPERF predicts a
neutral influence of diffuser radius
ratio on peak efficiency.
The conclusion drawn from the
investigation is that higher flow
stages require larger radius ratio
vaneless diffusers than smaller
flow stages. The positive
influence of vaned diffuser
systems in lower specific flow
stages does not change that
conclusion; in fact, it may expand
this conclusion. A vaned diffuser
tends to move the stage with
neutral influence of diffuser radius
ratio to a larger � value and also
tends to move the minimum
diffuser ratio to a lower value.
As in the existing design, an
impeller with a flow coefficient
0.0275 is chosen as the parent
impeller of the whole family.
Each impeller with a flow
coefficient less than 0.0275 (but
greater that 0.0077) can be
obtained from the parent impeller
by a proper meridional contour
cut. The parent impeller is
included in the analysis because
it is the only low flow wheel of
existing design that had been
tested in the SSTR until this time.
The main focus of the current
work revolves around past and
present stage configurations in
a CFD study comparing
performance of the original and
new stages. The operating
conditions chosen are for a single
wheel speed under typical SSTR
conditions: inlet pressure 30 psia,
inlet temperature
100° F, and nitrogen as the
working gas. A description of the
SSTR can be found in the work
of Sorokes and Welch,1992.
The stages compared have
significantly different stationary
component configurations, which
are not necessarily optimum.
The two fundamental differences
that should be noted in this
discussion are the shorter radius
ratio used in the new design
versus a larger radius ratio
used as the current standard.
In addition, due to the high
backsweep used in the new
impeller designs Low Solidity
Diffusers (LSD) are required,
whereas the old design is better
served using a vaneless diffuser.
Therefore, the new and old low
flow configurations should be
compared as a stage for any
proposed adjustments to overall
compressor performance.
Continued on page 16
Figure 4. Velocity Field Colored by Mach Number at Design Flow,
� = 0.0275, New Design.
15

Performance Evaluation
And Fluid Flow Analysis In
Low Flow Stages Of.....
Continued from page 15
It must be noted that the
CFD results should only be
compared directly with other
CFD results, since the modeling
does not completely capture all
physical reality (disk friction,
leakage, etc.).
Figure 5. Velocity Field Colored
by Mach Number at Design
Flow, � = 0.0275, Old Design.
CFD Analysis
Obtained flow fields are
analyzed to identify possible
flow separation or other specific
areas of losses (like flow nonuniformity)
that may cause
decreased efficiency or
insufficient head rise.
The stage grid consists of four
parts: vaneless radial inlet region
(not shown), impeller, LSD
including one-half of the return
bend, and return channel with the
remaining portion of the return
bend, as shown in Figure 3.
A commercial CFD code CFX-
Tascflow ® ‚ a product of AEA
Technology Engineering
Software, was used for the
stage analysis. Three “frozen
rotor” interfaces were applied to
attach grid regions; the total
number of nodes is approximately
150,000. One blade
passage was considered for
each component, with
16
circumferentially periodic
boundary conditions. Mass
flow boundary was utilized at
the stage outlet, while averaged
flow angle and stagnation
pressure and temperature were
imposed at the inlet.
For the parent impeller stage
(� = 0.0275) at design flow,
separation is not apparent in the
stationary components or
impeller (Figure 4). Flow velocity
Figure 6. Velocity Field Colored
by Mach Number at 62% of
Design Flow, � = 0.0275,
New Design.
remains relatively high in the
inlet region due to leading edge
incidence effects and blade
thickness blockage. However,
the relative Mach number
increase is not as high as that
observed in the old impeller
(Figure 5). Impeller loading
decreases around trailing edge,
thus minimizing the wake
originating from the suction side
of the blade, and making the
flow field behind the impeller
more uniform. Most of the
head is generated by approximately
60 percent of the blade
in the mid-passage region.
At 78 percent mass flow, no
recirculation zones were found
in the impeller. A low-energy
region, mainly near the shroud
surface, occurred at the suction
side midway through the blade
passage, but the high exit
backsweep does not allow
the wake to spread.
A small local recirculation zone
occurs near the hub at the
trailing edge of the suction side
of the LSD vane, but the wake
is relatively small due to the
higher velocity flow coming from
the pressure side.
At even lower mass flow
(62 percent of design value), a
small separation zone with a
recirculation was found at the
hub-suction corner of the
Figure 7. Velocity Field Colored
by Mach Number at Design
Flow, � = 0.0117, New Design.
impeller blade passage (Figure 6).
High velocities were observed at
the impeller inlet on the suction
side of the blade, resulting from
incidence and flow turning
around the leading edge. In the
mid and exit part of the blade
passage a low-energy zone can
be seen at the suction side, but
without separation.
A local separation zone
occurred at the hub-suction side
of the LSD vane near the trailing
edge that results in a wake flow
region in the return bend
entrance.
At overload (120 percent design
mass flow), impeller flow was
without separation, but flow with
a higher relative Mach number is
found at the inlet, both on the
suction and pressure sides of
the blade. The latter is due to
incidence effects and the
corresponding turning required
around the leading edge. For
this flow case the LSD region
has a strong recirculation zone
on the pressure side of the
blade, and the resulting wake
washes out only in the second
half of the return bend.
For smaller stages (� = 0.0146
and lower), the flow field in the
impeller and diffuser looks similar
to that in the parent stage, both
at design and off-design points.
The separation zone is found in
Figure 8. Static Pressure Field
in the Impeller and LSD At
Design Flow, � = 0.0275,
New Design, Mid-Span.
the LSD region at the low flow
condition on the suction side of
the vane, close to the trailing
edge and shroud surface.
Figure 7 illustrates the flow field
in the newly designed impeller
with a flow coefficient of 0.0117.
Static pressure plot in the
impeller and the LSD mid-span
region is shown in Figure 8.
At low flow, the recirculation
eddy is washed out in the LSD
quickly enough to smooth the
velocity profile in the return
bend. A typical low velocity
zone in the impeller was located
at the outlet on the suction side
of the blade, becoming more
severe toward the shroud
surface. In addition, the wake
entering the diffuser from the
impeller is not very strong.
A CFD model was built to
analyze the flow in the
discharge volute. As shown in
Figure 9, about 400,000 nodes
were used to mesh the return
Figure 9. Volute Grid for CFD Model.
Figure 10. Volute Static Pressure Contour Plot.
Figure 11. Comparison of Normalized Optimum Efficiency and
Head Coefficient at Best Efficiency Point for New and Old Design
Impellers and Stages.
bend, the volute, and the
discharge nozzle. A commercial
code Star-CD ® from
Computational Dynamics Ltd.
was used to run this model.
A mass flow rate boundary
condition was applied to the
inlet, at the entrance to the
return bend. At the exit of the
discharge nozzle, an outflow
boundary condition was applied.
Figure 10 shows the static
pressure distribution computed
by the CFD model. The flow field
in the volute was uniform with no
major loss regions.
Performance Comparison
Figure 11 shows peak efficiency
and polytropic head coefficient
at peak efficiency based on the
CFD results and SSTR. The first
item that should be noted is that
CFD overpredicts the test data
for low flow coefficients, while
underpredicting data at higher
values of flow coefficients. The
underlying causes for these
differences are most likely due to
influences external to the main
flow path which were not
modeled (e.g. friction in the
impeller cavity and leakage),
as well as numerical sources
(e.g. grid dependency), and
turbulence modeling. There are
several other items of interest
that should be noted from
Figure 11. First, the impeller
efficiency curves predicted by
CFD track relatively closely to
one another, although the new
impeller appears to exhibit
slightly lower efficiency in the
lower flow capacity stages.
On a stage basis the new
stages appear to exhibit a near
constant peak efficiency across
the entire flow range. The older
stage peak efficiencies follow
more closely to the empirical
curve.
An alternative presentation of
the peak efficiency and head
coefficient at peak efficiency is
shown in Figure 12. The CFD
data shown on this plot are
normalized by maximum values
of peak efficiency and head
coefficient, while the empirical
data were normalized to agree
with the CFD results for the old
stage with � = 0.0275. These
results demonstrate more clearly
the differences in curve shape
and relative performance of the
stage designs. Clearly, the new
stage shows significant
improvement in performance
when compared to the old
stage. The actual level of
improvement cannot be
determined without further
calibration of the predictions
with empirical data.
Head coefficient � curves
(shown in Figures 11 and 12)
correspond to the peak
efficiency operating points.
The results indicate higher head
levels at peak efficiency for both
the current (“old”) impeller and
stage designs. The results also
imply better pressure recovery in
the stationary components for
the new design, based on the
comparison of � for the impeller
and stage curves.
The general shape of the CFD
predicted curves for � are
relatively flat with a peak in the
0.010 – 0.015 flow coefficient
range, while the empirical curve
and SSTR data suggest an
increasing trend for � with flow
coefficient. The reason for this
discrepancy most likely lies in
the lack of leakage and external
secondary flow path windage
losses; internal flow path
roughness may also be a factor.
In addition, previous comparison
with SSTR data indicates that
the CFD results appear to
underestimate the impeller
slip factor.
Figure13 illustrates individual
stage performance along a
speed line. The new stage with
� = 0.0146 exhibits the typical
curve shape expected with
LSD’s higher peak efficiency
with slightly shorter range.
Continued on page 18
17

Performance Evaluation
And Fluid Flow Analysis In
Low Flow Stages Of.....
Continued from page 17
Relative peak efficiency
adjustments for STGPERF were
suggested based on the CFD
results obtained, assuming that
the difference in head and
efficiency between the new
stage design and the original
design may be identical to the
actual change in performance
of the physical machine.
Comparison With
Production Test
The one-dimensional analysis
code STGPERF was modified
based on the aforementioned
CFD results to ensure performance
prediction for stages
with newly designed low flow
impellers. Based on stageby-stage
calculations, the overall
compressor performance was
predicted at test conditions and
compared with test results.
Intermediate test data were
gathered between stages
where total pressure and
temperature probes were
installed in the return bend area.
Comparison of newly predicted
and tested data for the entire
compressor is shown in Figure
14. The results demonstrate
very good agreement of head
coefficient and relatively good
agreement of efficiency values.
The discrepancy does not
exceed 2.5 percent and is
greater at low flow capacity
conditions.
Conclusions
The proposed new stage
design demonstrated improved
efficiency compared to the
previous design that made new
geometric configuration
successful in a high-pressure
compressor and for similar
applications. It should be noted
that when applying the newly
designed low flow stages,
operating range is reduced due
to utilization of LSD’s. Additional
18
design improvements are
possible, e.g. adjusting impeller
leading edge position and/or
changing the number of blades,
optimizing return channel/bend
geometry, etc.
The efficiency improvement has
been predicted by a series of
CFD calculations at various flow
capacity values and verified
experimentally by testing the
compressor with the newly
designed changes. Qualitative
agreement between computational
and test data has been
obtained. Quantitatively, the
CFD calculations slightly
over-predict the performance.
This discrepancy is caused by
numerous factors not being
taken into account in the
computational model and CFD
tools such as secondary losses
in labyrinth seal passages,
surface roughness, inlet and
intermediate flow non-uniformity,
transient effects, turbulence
modeling, near-wall flow
resolution, etc. Some of these
effects represent a challenge for
CFD methods to predict them
accurately. At its current level of
development CFD cannot be
used for direct performance
predictions. However, CFD is
an excellent tool for providing
data for relative dimensionless
comparison and analysis of
the impact of design and flow
parameters on performance of
centrifugal compressor
components.
For practical reasons, it is
possible to define a set of key
parameters that will allow a
designer to perform a parametric
study of the impact on
stage performance using CFD.
Results of that study, together
with available test data, can be
used to tune the existing
models implemented in a
semi-empirical 1D code (like
Dresser-Rand’s STGPERF)
to enable it to predict stage
and overall compressor
performance with accuracy.
Ultimately, this parametric
approach will also make the
1D code more generic in nature
(i.e., able to analyze the full
range of stage geometry) as
well as open the doors to its
application as an inverse design
tool that could be used to tailor
stage design to operating
specific conditions.
In current conditions of rising
research and production
test costs, the problem of
calibration of CFD codes for
turbomachinery becomes a
key factor in performance
prediction. Consequently,
Dresser-Rand continuously
strives to develop its test and
computational programs,
including both SSTR rigs and
expanding CFD capabilities. ■
Figure 12. Comparison of Normalized Optimum Efficiency and
Head Coefficient at Best Efficiency Point for New and Old Design
Stages. Alternative Normalization.
Figure 13. Comparison of CFD-based Performance Curves for
New and Old Stages at � = 0.0146.
Figure 14. Comparison Between Tested and Predicted Overall
Compressor Performance Characteristics.
Dresser-Rand Expands Service
Capabilities For Reciprocating
Compressors And Integral Gas Engines
Dresser-Rand has announced
the formation of an expanded
engineered solutions operation
to provide expert analysis and
service for all makes of process
reciprocating compressors and
integral gas engines, including
non-Dresser-Rand equipment.
The new operation, the Applied
Technology Department,
located in Grove City,
Pennsylvania, is an extension of
Dresser-Rand’s Engineered
Solutions group based in
Painted Post, New York.
Applied Technology is headed
by George Elefteriou, project
manager, who joined Dresser-
Rand in March, after more than
25 years experience with
Cooper Energy Services.
Elefteriou brings to Dresser-
Rand a team of engineers and
technicians with a combined
150 years of experience in field
service and engineering.
The group reports to Fred
Tanneberger, vice president of
Engineered Solutions for
Dresser-Rand Product Services
in Painted Post.
“Our clients will be able to
upgrade or revamp their
existing equipment – regardless
of make or model – with the
most advanced engineered
components and assemblies,”
Tanneberger said. “We’re not
replicating parts, but instead,
providing the latest technology
and designs for improved
efficiency and reliability of
operation.”
Formation of Applied
Technology is a direct response
to clients’ desire to have one
supplier work on all of their
equipment. It also expands
Dresser-Rand’s expertise as a
single source supplier to the
industry. During the past 100
years, Dresser-Rand has
shipped more than 8,000
process reciprocating compressors
and 9,000 integral gas
engines. This vast experience
as an equipment manufacturer,
combined with extensive
experience in field service,
benefits Dresser-Rand’s clients
by providing advanced
capabilities for technology and
total solutions capabilities,
including:
- Site Audits
- Field Analysis and Problem
Solving
- Equipment Upgrades
- Complete Revamps
- Remanufactured Units
- Field Installation and Service
- Total Solutions
“There are no restrictions on the
makes or models of equipment
that we can service,” says
Elefteriou. “As an OEM, we
understand the challenges of
the equipment. And no other
company has this range of
manufacturing experience
combined with Dresser-Rand’s
technical support and extensive
field service organization.”
For information on Dresser-
Rand’s Applied Technology
services for reciprocating
compressors and integral gas
engines, contact a D-R sales
representative; call Dresser-
Rand at (724) 450-2081;
or send an email to:
george_k_elefteriou@dresser
-rand.com. ■
19

Dresser-Rand Wins
Contract For Combined
Cycle Power Generation
Arizona Public Service/
Pinnacle West Energy has
awarded Dresser-Rand a
contract for a 45 megawatt
steam turbine generator set to
be installed in Phoenix, at the
Pinnacle West Energy
combined cycle plant.
The steam turbine generator
set will be delivered in October
2000, with plant start-up
scheduled for June 2001.
Dresser-Rand will provide onsite
training for APS/Pinnacle
West Energy associates,
installation, and start-up for
the project.
“This project reflects a great
opportunity for Dresser-Rand
to further demonstrate our
ability in the steam turbine
power generation market,”
said Doug Gewand, product
manager for Large Steam
Turbines at Dresser-Rand.
Alliance With Liburdi
Engineering To Extend Life
Of Power Turbine
Components
Dresser-Rand and Liburdi
Engineering Limited have
entered into an alliance
agreement to provide
advanced component repair
and rejuvenation services for
Dresser-Rand’s complete line
of power turbines.
According to company
officials, Liburdi Engineering’s
experience in advanced repair
processes for turbine blades
and vanes, coupled with
Dresser-Rand’s turbine design
expertise and Power Turbine
Life Extension program, will
permit clients to safely and
economically extend the
operating life of their power
turbines beyond the original
“design” life.
Joe Liburdi, president of Liburdi
Engineering, stated, “This
alliance will allow us to quickly
introduce proven repair
technologies and provide
Dresser-Rand’s [turbine] users
with the combined benefit of
component life extension and
OEM support.”
Walt Nye, executive vice
president of Dresser-Rand’s
Product Services organization,
commented, “Liburdi
Engineering’s expertise in
component repair, combined
with our design knowledge,
allows us to economically and
safely develop repair and
rejuvenation processes that
extend the operating life of
power turbine components.
We have the opportunity to offer
our clients a choice between
new and refurbished parts.
“We’re committed to
supporting our products, and
we’ll continue to develop
product upgrades, repair
techniques, and support
services that benefit our
clients.” ■
Dresser-Rand Forms
Alliance To Improve Air
Quality In China
Engeturb Offers Full
Service For Latin
American Clients
Dresser-Rand recently
acquired controlling interest in
Engeturb to provide a full line
of products and services for its
Latin American client base.
Engeturb’s steam product line
has remained viable
throughout its ten-year
association with Dresser-
Rand. The company will
maintain its current steam
turbine product line to serve
the Latin American steam
turbine generator market
under 20 megawatts.
Engeturb’s Campinas facility,
65 miles north of Sao Paulo,
Brazil, brings Dresser-Rand’s
worldwide total of service
centers to 23, which includes
another Latin American-based
service center in Maracaibo,
Venezuela.
The facility is making a
transition to a fully integrated
Dresser-Rand service center.
When completed, it will
provide total energy conversion
solutions and services
throughout Latin America
for all reciprocating, centrifugal,
and steam turbine product
lines.
Dresser-Rand Publishes
Guide To Naval Steam
Turbines, Compressors
And Blowers
To better serve its marine
industry clients, Dresser-Rand
has published a comprehensive
guide to the steam
turbines, compressors and
blowers it supplies to the U.S.
Navy. The eight-page
brochure describes the
innovative engineering, reliable
equipment, and total solution
capabilities used to meet the
Navy’s shipboard compression
and turbine drive requirements.
The nuclear-powered CVN 77,
the Navy’s next aircraft carrier,
will have eight Dresser-Rand
turbines. The four low-pressure
and four high-pressure turbines
will be driving the carrier’s four
propellers.
A copy of the Naval Steam
Turbines, Compressors and
Blowers brochure may be
obtained online at
www.dresser-rand.com by
using the “Contact D-R” form,
or by emailing your request to
info@dresser-rand.com.
To request a brochure by
telephone, call (U.S.)
713-973-5353. ■
Dresser-Rand Field
Support Services Brochure
Available
Dresser-Rand has published a
comprehensive guide to the
field support services it offers
around the world. The fourpage
brochure describes how
the field support service
organization provides solutions
and services for every phase
of operation, from design,
installation, and financing to
training and service.
As one of the many solutions
provided by Dresser-Rand, field
support includes ‘round-theclock’
reliability. With 160
service representatives in more
than 20 strategic locations
around the world, field service
personnel can be dispatched
to a site to provide routine
maintenance or emergency
service needs.
A copy of the Field Support
Services brochure may
be obtained online at
www.dresser-rand.com by
using the “Contact D-R” form,
or by emailing your request
to info@dresser-rand.com.
To request a brochure by
telephone, call (U.S.)
713-973-5353. ■
Since 1957, Dresser-Rand
has designed, manufactured,
tested, and shipped more than
2,000 megawatts of steam
turbine power, specifically for
electric power generation.
Although the primary focus of
the agreement is the repair
and rejuvenation of power
turbine components, Liburdi
Engineering’s expertise also
will allow Dresser-Rand to
offer new repair procedures
for other product lines.
Liburdi Engineering is a world
Dresser-Rand and Jinan Diesel
Engine Co., Ltd. (JDEC) of
China recently entered an
alliance to provide fueling
station equipment to the
compressed natural gas (CNG)
industry in China. The two
companies will cooperate
exclusively to produce CNG
compressor packages and
market them throughout China.
Claudio Jose N. Oliveira is
the general manager of the
Campinas facility, located at
Rua Altino Arantes 1010,
13051-110, Campinas – SP,
Brazil. Please email inquiries to
engeturb@aleph.com.br, or
phone 011-55-19-227-2060. ■
APS is Pinnacle West’s major
subsidiary. It generates, sells,
and delivers electricity and
energy-related products and
services to wholesale and retail
customers in the western
United States. Pinnacle West
was recently added to the
S&P 500 index. ■
leader in gas turbine component
repair, and a pioneer in
gas turbine blade and vane
rejuvenation techniques,
possessing such technologies
as LAWSTM automated
welding systems, LPMTM
powder metallurgy, and
advanced coating and
rejuvenation processes.
Dresser-Rand will provide its
compressor packaging
technology to JDEC, a
subsidiary of China National
Petroleum Corp., and allow
JDEC to package the
compressors.
The alliance will provide
equipment and services to
the CNG industry that will help
meet the national goals of
20
improving air quality and living
standards in China. ■
21