insightsinsights

people power

Summer 2015

a DResseR-RanD PUBLICaTIOn

Dresser-ranD - a siemens business: the journey continues

new high-Performance gimPel

®

ValVe Picking uP steam

insights.dresser-rand.com

Visit usonline:be the change you want

to see in the worlD

CONTENTS

03

insightsSUMMER 2015

This document contains statements related to

our future business and financial performance

and future events or developments involving

Siemens that may constitute forward-

looking statements. These statements may be identified by words

such as “expect,” “look forward to,” “anticipate,”

“intend,” “plan,” “believe,” “seek,” “estimate,” “will,”

“project” or words of similar meaning. We may

also make forward-looking statements in other

reports, in presentations, in material delivered to

shareholders and in press releases. In addition,

our representatives may from time to time make

oral forward-looking statements. Such

statements are based on the current expectations and certain assumptions

of Siemens’ management, of which many are beyond

Siemens’ control. These are subject to a number

of risks, uncertainties and factors, including, but not limited to those described

in disclosures, in particular in the chapter Risks in the

Annual Report. Should one or more of these risks or uncertainties

materialize, or should underlying expectations

not occur or assumptions prove incorrect, actual

results, performance or achievements of Siemens

may (negatively or positively) vary materially

from those described explicitly or implicitly in the

relevant forward-looking statement. Siemens neither intends, nor

assumes any obligation, to update or revise

these forward-looking statements in light of

developments which differ from those anticipated.

Fellowship Program Welcomes Moll and Lucas

George M. Lucas and Randall W. Moll represent the third Fellows class inductees since this prestigious program launched in 2012.

14

engineer’s notebook Stability Testing of CO2 Compressors

This paper presents results from stability analysis and testing on high-pressure CO2 centrifugal compressors to gain a more clear understanding of compressor stability for different load conditions.

16

02

profile Service Contract Manager Not Unlike a Maestro

Danielle describes herself as an orchestra conductor, whose job is to work toward a common goal and make sure everything goes smoothly for Phillips 66.

10

Strengthening the communities in which we operate is an important aspect of our Company’s values – along with a “safety first” mindset.

12 Be the Change You Want to See in the World

Partnerships: As Strong as the Foundation You Build

A pipeline plant is in regulatory compliance for emissions and operating with increased capacity, lower operating expenses and using fewer packages than it had previously, thanks to collaborative teamwork.

08

Client Support From Every Angle

These examples aptly demonstrate how we work closely with clients to quickly resolve problems and research better ways to manage high-speed compression.

06

people power

Our new Gimpel® electro-hydraulic trip-throttle valve (EHTTV) stops the flow of steam to the turbine four to 10 times faster than latch-type valves.

New High- Performance Gimpel® Valve Picking up Steam

Chris Rossi executive vice president of global operations at Dresser-Rand - A Siemens Business, discusses how the combined Siemens and Dresser-Rand products and services will benefit clients, suppliers and other stakeholders.

candid visions Dresser-Rand - A Siemens Business: Our Journey Continues

D-R Eliminates High Vibration Levels in Gas Compressor

With just 45 days to get a charge gas compressor back up and running, Petroquimica Mexicana de Vinilo called on Dresser-Rand engineers to have a look.

04

Janet Straub Ofano, sr. communications specialist & editor, insights magazine

Willkommen

After 15 years as editor of insights magazine, and in the spirit of embracing change, I’m pleased to welcome you to this edition.

Change is such a short word, yet has profound meaning. Much research has been done around this one small word – in business, relationships, lifestyles. Sometimes we fear change; sometimes we welcome change. However, we all experience it at some point in our lives – good, bad and otherwise.

Looking back on my time with Dresser-Rand, I have indeed experienced my share of change, both personally and professionally. During this time, Halliburton sold its share of Dresser-Rand to Ingersoll-Rand, Ingersoll-Rand then sold Dresser-Rand to First Reserve Corporation, and we became a publicly traded company on the New York Stock Exchange. Thereafter, there were nearly a dozen acquisitions, which involve a great deal of change on both parts. Most recently, our family of approximately 8,000 grew to one of 350,000. We are now Dresser-Rand - A Siemens Business.

From a business standpoint, the merging of two companies is no small feat. It’s about growing, experiencing new ways of doing things, meeting new people, expanding our horizons, and connecting employees across teams and geographies. A new cultural cohesion must take place. It requires collaboration, openness, flexibility, adaptability, and strength.

You’ll find within these pages several instances of people collaborating to help our clients lower operating expenses, increase capacity, meet emissions regulations, and more. In fact, one service contract manager describes herself as an orchestra conductor who works with her client to ensure everything goes smoothly.

As part of a new, larger family, we anticipate even more opportunities to work together to solve our clients’ challenges. Rest assured, that while we make this transition into the Siemens organization, our journey remains unchanged: to earn our clients loyalty for life.

With warm regards,

“Collaborative organizations can help make the world a better place.”– Jacob Morgan

Janet Straub Ofano editor, insights magazine

1

cand

id v

isio

ns

Chris Rossi executive vice

presidentglobal operations

2 insights

BBoth companies have much in common: a long and proud engineering tradition with roots that date back to the mid-19th century; global presence; great people with know-how and understanding of clients’ needs; successful business records; and premium brand recognition within the industry.

Siemens and D-R, combined with the most extensive gas turbine driver offering with the products acquired from Rolls Royce, is expected to become the leading rotating equipment solutions provider to the oil and gas and environmental solutions markets. We can provide far more choices for our clients, particularly in relation to market coverage, local service capabilities and drivers for both compression and power generation applications.

With the new Dresser-Rand business, there are three primary advantages for our valued clients.

First, with Siemens’ well-established presence in Europe and Asia and Dresser-Rand’s stronghold in North America, we now have greater scale and reach. Our combined presence in more than 150 countries, including more than 80 manufacturing and service locations, will enable faster response through local resources.

Second, a larger product and services portfolio gives our clients more choices from a single provider. And with the extended population of installed equipment and increased services support network, we believe that we have the scale and reach to further enhance the quality and timeliness of our day-to-day service coverage.

And finally, we are bringing the best of our two companies together – unequaled products and technologies, talented people, extensive industry

know-how, and unmatched knowledge about our markets.

Our commitment to our clients is that the new Dresser-Rand business within Siemens will

be an accessible, responsive partner of choice. Beyond the expanded offerings

of products and services, we are working on a seamless integration which ensures business continuity throughout the integration period. Over the coming months, we plan a measured approach to the integration of our two businesses. The path forward includes the

following key elements:

• Continuing to be a safe and reliable supplier

• Providing technology expertise and leading in innovation

• Responding with speed and flexibility

• Acting local with global scale and reach

• Maintaining the best total cost of ownership.

If we stick to what brought us here today – ethical behavior, a culture of “safety first,” operational excellence, and high-quality products and services – we are confident that we will enjoy a successful integration with no disruption to clients, suppliers or other important stakeholders.

We remain dedicated to delivering the best-value solutions and our vision remains: earn client loyalty for life.

“Dresser-Rand was an important

acquisition in building Siemens

for the next generation.”

Dresser-Rand - A Siemens Business: Our Journey Continues

Joe Kaeser, Siemens AG CEO

commented,

It has been a remarkable year. As you know, we formed a new business within Siemens that combines the Siemens and Dresser-Rand compressor products and service businesses. To preserve the brand strengths of both companies, we are now Dresser-Rand - A Siemens Business.

Christopher Rossi Executive Vice President Global Operations

Best regards,

3

DDesigned to prevent the catastrophic failure of steam turbines, Dresser-Rand’s Gimpel EHTTV stops the flow of steam to the turbine in less than 0.3 seconds, a closing force that is four to 10 times faster than typical latch-type valves. Additionally, the back seated stem design prevents steam loss when the valve is fully open.

EHTTVs operate independently of the installed turbine oil system and only require electrical power to operate. They can be locally or remotely actuated and exercised. The EHTTV maintains a five-year overhaul cycle with no lubricating requirement and complies with API-612 standards. They can also be used to improve efficiency and, ultimately, the bottom line by saving 100-250 lbs. of steam per hour.

The valve underwent extensive testing at Dresser-Rand’s facility in Burlington, Iowa and performed as expected under designed conditions.

EHTTVs were developed for steam turbines in a wide range of industries, including power, ethylene, ammonia, pulp and paper, petrochemical, steel, medical, education, and refinery.

Dresser-Rand is the first and only known trip throttle valve manufacturer with SIL-3 certification by Lloyd’s Register Group.

“Our new Electro Hydraulic Trip Throttle Valves bring a new level of performance and efficiency to the market and are performing very well for our clients,” said Maged Mikhail, Vice President Engineered Solutions and Gimpel SBU for Dresser-Rand. “We have an installed base of more than 17,000 Gimpel valves and have an excellent track record, which is why we’re now the only trip throttle valve manufacturer with SIL-3 certification.” •

Dresser-Rand launches new Gimpel® electro-hydraulic trip-throttle valve (EHTTV).

New

Picking Up SteamGimpel® Valves

High-Performance

4 insights

D-R Eliminates High Vibration Levels in Gas CompressorA charge gas compressor train at an ethylene plant was exhibiting high vibration levels in the low-pressure casing before crossing first critical speed. These vibration levels prevented the train from reaching its design operating speed of 5,800 rpm. With only a 45-day window to identify the cause of the vibration and fix it, Petroquimica Mexicana de Vinilo (PMV) asked Dresser-Rand engineers to have a look.

5

PPMV, a flagship joint venture between PEMEX and Mexican petrochemical firm, Mexichem, represents the first joint venture between Pemex and a private company. The merger joined Mexichem’s salt, chlorine and caustic soda operations with Pemex’s ethylene and vinyl chloride monomer (VCM) operations.

PMV revamped an old Pemex plant in the Pajaritos petrochemical complex to produce an expected 120 tons of VCM, the key material used to manufacture polyvinyl chloride (PVC), commonly used for electrical insulation, films and pipes.

Francisco Moncayo, Dresser-Rand Services director for Mexico, describes the circumstances. “Initially, PMV representatives asked us to perform a vibration analysis on the compressor train, consisting of a 3 MX compressor, 3M compressor and a 4M compressor, to determine possible causes. The fact that there was no historical data on maintenance or vibration levels compounded the problem because in the past PEMEX had sourced third-party, non-Dresser-Rand parts and service.”

At the outset, the compressor was operating at 5,150 rpm; however, increasing the speed to 5,220 rpm significantly increased vibration levels whereby the protection systems would trip the compressor train and shut it down. Data analysis during the machine trip showed high vibration (6.85 mils peak-to-peak) at the 5,220 rpm level, so the machine could not operate at speeds above 5,150 rpm. Such interruptions affected plant operation and resulted in lost production and revenue.

Internal friction in the low-pressure casing, lack of rigidity in the system, train misalignment, and process piping were found to be the main causes for vibration. In addition, Dresser-Rand engineers found corrosion in the compressor casing, and in the suction and discharge flanges, a half-inch crack

on one of the shaft journals, impeller pitting, and coupling gear teeth and spacer flange pitting. In addition, several of the components (installed over the years by third-party parts manufacturers) were found unsuitable and not within Dresser-Rand OEM dimensions.

PMV accepted Dresser-Rand’s proposed solutions to repair the train which included shutting down the train to inspect and repair internal compressor components; measuring bearing clearances and compressor shaft run-out; inspecting laby seals; aligning the compressor and turbine, and aligning the suction and discharge piping to the compressor flanges; and stiffening the compressor supports and discharge lines. Upon inspecting the proximity vibrations system and bearing clearances on the 3M compressor, D-R recommended replacing it and supplied its OEM parts for the overhaul, capital spare parts and 3MX shafts.

When all was done, Dresser-Rand completed the agreed upon scope, aligned the train, commissioned it, and started it up within the 45-day deadline. The Houston, Texas service center was able to accommodate other major repairs to the 3M and 3MX rotors in a short period of time to avoid longer delays to the shutdown.

Moncayo concludes, “Meeting the project’s extraordinary time constraints was a result of many different functions working together within Dresser-Rand, including our Services team in Mexico (Sales, Proposals, Field Services, Reliability and Predictive Maintenance Team), Houston Service Center, Dresser-Rand Turbine Technology Services, Olean, N.Y. Operations upgrades and parts team, and our Technical Support Team in Venezuela.”

According to Moncayo, vibrations on the compressor train are now well below 1 mil. •

6 insights

T

Client Support – From Every Angle

Three-tier Parts Stocking ProgramThis inventory program was developed with our authorized parts resellers to support our mutual clients’ (end users’) situations on three levels, ranging from unexpected emergency shutdowns to everyday scheduled compressor maintenance.

Our Tier 1 parts – crankshafts, connecting rods, crossheads, and oil pumps – for current and non-current compressor models are regularly stocked in several Dresser-Rand parts warehouses and can be shipped at a moment’s notice, 24/7. Most of these parts support unexpected shutdowns where compressor frame damage has occurred.

Our Tier 2 parts – piston and rod assemblies, piston rods, pistons, packing cases, valves, oil pump

rebuild kits, bearings, and bushings – are stocked by our authorized parts resellers for the end user, and by Dresser-Rand. This tier supports the end user’s scheduled compressor maintenance programs and unexpected shutdowns

where cylinder damage may have occurred.

Our Tier 3 parts – packing and piston rings, valve plates and springs, oil filters, and seals – are stocked by Dresser-Rand, our authorized reseller and the end user to ensure typical

wear parts are readily available. These parts can be air freighted, pulled directly from

a local packager warehouse or taken from the end user supply located on-site. This tier supports situations ranging from planned inspections where worn parts need replacing, to minor performance issues such as broken valve elements, to a leaky gasket or o-ring. The client can have wear parts available within hours to immediately get the unit back up running.

For example, an end user in western Texas had a HOS™ unit go down unexpectedly on vibration. An investigation determined that a crosshead, piston and rod needed replacing. “The end user contacted our stocking distributor, who in turn contacted Dresser-Rand. Within 24 hours, all of the required tier 1-3 parts were on location and the unit was back up and running in less than 48 hours,” said Gary Tas, global sales manager for the HSRC business unit Aftermarket Services.

Valve Performance Critical to Peak Operation We design and manufacture our own valves from a long legacy of engineering experience – and we believe we are the only OEM in the industry that does this.

“Our engineers work through potential problems before our clients encounter them in the field,” said Steve Chaykosky, Manager of Valve Engineering at Dresser-Rand. A dynamic valve analysis (DVA) is performed for every compressor application. The comprehensive DVA software program, developed in-house, analyzes full and part-load operating

In order to help keep our clients’ equipment running at optimum performance, the high-speed reciprocating compressor (HSRC) business unit has developed several aftermarket initiatives over the years. As with the MOS™ compressor, this goes back to listening to what our clients had to say. And these benefits don’t just apply to Dresser-Rand® compressors – we offer upgrades for all brands of reciprocating compressors.

7

conditions to maximize valve reliability and minimize horsepower and capacity losses.

The need to have an engineering tool on-site so that compressor valve assemblies could be rapidly evaluated in fatigue led to the development of our valve endurance tester. This unique device provides accelerated impact and fatigue cycling of valve components at six to 10 times compressor operating speeds. “Employing this valve ‘basher’ endurance tester allows our designers to assess valve durability in the laboratory – not in the field on a client’s compressor,” added Chaykosky.

Another element of our comprehensive valve testing is the closed loop compressor test unit at our Painted Post, N.Y., U.S.A. facility that evaluates valve performance characteristics under actual conditions. The first of its kind 36 years ago, and still an industry-wide respected, robust OEM valve testing technology, this test unit can be used on a wide variety of gases: pumping process gases at speeds up to 1,200 rpm and discharge pressures up to 1,500 psig (103 barg).

Magnum® HammerHead™ ValveWhen our clients told us they wanted a product to meet the compressor efficiency gap with heavier gases such as carbon dioxide, we listened. The Magnum® HammerHead™ valve maximizes compressor efficiency for high molecular weight applications such as CO2, ethylene, propane, and natural gas. Unlike other poppet valves, the HammerHead valve can be applied at high compressor speeds. And unlike plate and ring valves, it uses one element for all valve sizes, simplifying valve inventory. “This new valve is fast becoming an industry benchmark and can be applied to all brands of reciprocating compressors,” said Chaykosky.

Computational fluid dynamics and finite element analysis were used to develop the HammerHead valve’s geometry to improve flow area by as much as 60 percent. Minimal valve loss means improved efficiency. Also, the valve’s unique element design minimizes tensile stresses, resulting in long life in demanding applications.

Chaykosky recalls (before the advent of the HammerHead valve) a client who was getting 90 days or less operating time from their (non- Dresser-Rand) high-speed reciprocating compressor valves. When we approached the client and suggested our Magnum® XF valve, he was willing to do a “try and buy” to see if the valves would last in 1,000 psi (69 bar) natural gas service at 1,350 rpm. He agreed that if the valves lasted at least 90 days he would buy them. The valves were in service for three years before they were removed for inspection. After inspection, the valves were re-installed because they were still in great shape. Pleased with the Magnum XF valves and confident in Dresser-Rand’s valve technology, the client ordered a set of Magnum HammerHead valves for one of his natural gas compressors (a D-R competitor’s unit) operating in 900 psi (62 bar) natural gas service at 1,380 rpm.

In another instance, a client’s machines were overheating during summer months disrupting production. We installed our Magnum HammerHead valves and the client’s throughput went up and cylinder gas discharge temperatures dropped by approximately 20 degrees Fahrenheit (-6 degrees Celsius).

“As is evident in the examples above, we work closely with our clients to quickly resolve problems and continue to research better ways to manage high-speed compression. We work with our packagers, distributors and end users to ensure client support from every angle,” said Tas. •

8 insights

CClients approach us with specific needs and desired outcomes for their equipment.

Dresser-Rand has a long history of working with end users, contractors and even between business units within the company itself. Our engineers recognize that the strength of a partnership can only be judged by how adeptly we are able to achieve a specific client goal. Therefore, we use a foundational approach to providing equipment, systems and controls for each client.

Our foundational approach begins with an initial discussion that focuses on the client’s situation, their specific needs and any underlying conditions that need to be addressed by the equipment or service to be provided. The first discussion with a client is not about what products we can sell them, but rather what problems they are attempting to solve. We want to understand the client in the most comprehensive way, and then develop solutions from that vantage point.

In 2012, discussions began with a North American energy provider about a project that initially involved refurbishing 11 Clark engines at one of the client’s gas compression facilities. The rework was intended to bring the Clark engines into regulatory compliance and increase station capacity. During this period, the client was involved with an acquisition which resulted in the sale of the facility to a pipeline company. Although Dresser-Rand has a long standing relationship with the North American energy provider, as a newly formed company, the pipeline company did not have a past relationship with D-R. However, due to the timing of the acquisition, confidence in the solution and focused communications between the companies, the pipeline company chose to continue with Dresser-Rand and quickly negotiated their first Master Services Agreement (MSA) with us.

Dresser-Rand engineers were asked to evaluate doing a revamp versus buying new equipment. The North American energy provider and Dresser-Rand completed a full lifecycle evaluation based on cost of ownership, including initial capital investment and ongoing operating expenses. After completing

A partnership is defined as a relationship between individuals or groups characterized by mutual cooperation and responsibility for the achievement of a specified goal.

Partnerships:As Strong as The Foundation You Build

9

the evaluation, the pipeline company decided to purchase six new engine / compressor packages as the estimate to revamp the existing engines was greater than 50 percent of the cost of new equipment. Additionally, the corporate support within the pipeline company required to execute a major revamp was very limited.

The project consisted of three phases.

Phase 1 – Design and specify the process piping, connections to the new skids, foundations, building, ventilation, automation and controls, electrical power, and other utilities to support the six new engine / MOS™ compressor packages.

Phase 2 – Procure all parts, package and deliver the Waukesha engines, MOS compressors and Enginuity® iFLEX control systems.

Phase 3 – Construct the supporting facilities designed in phase 1, and install the packages from phase 2.

Start-up and Commissioning Achieved in Record TimeDresser-Rand was able to shift engineering resources from modifying existing equipment to designing new installations within existing infrastructure, while concurrently facilitating purchasing, project management, resourcing, scheduling, and integration of the new equipment. We believed we could accomplish the task with minimal disruption to existing facility operation, which was a key element in securing the contract. The new packages would require new process piping, structures to house the equipment, as well as ancillary and gas cooling structures.

Additionally, this project was the first site for Dresser-Rand’s MOS compressors to be installed for continuous service. The MOS compressor was determined to be a good fit for the project. Dresser-Rand’s ability to design the installation and support systems, and commission these first MOS units, ensured the compressors would perform when they needed to.

Installing a new engine / compressor design can be difficult and unsettling at the best of times; doing it under a compressed schedule can greatly increase the stresses on the team. Dresser-Rand diligently and steadfastly provided service and support to ensure production would not be affected at the plant. The pipeline company’s plant personnel were

also instrumental in making the installation of these new compressors a success, a real team effort.

With engineering and procurement well under way, the focus shifted to installation. During negotiations for installation, the pipeline company determined it had achieved a staffing level sufficient to manage site construction themselves, and would therefore call upon Dresser-Rand’s Gas Engine Technology Center experts to consult on the balance of installation services, including packager inspection and automation control factory acceptance testing, to ensure that each part of the final installed package was executed to the highest quality. After construction was completed, Dresser-Rand and the pipeline company had a smooth and easy start-up and commissioning process. The result: six units fully commissioned and operational in two months.

A new, fully integrated automation and controls system was also designed and installed. The system fully communicates the data to the existing station systems, and allows the operators to visualize unit performance in real time at their stations.

To develop the knowledge and understanding needed by the pipeline company’s staff, training was on-going throughout the various phases of the project. Numerous meetings were held between D-R and the pipeline company’s staff on-site; in the final stages, a complete documentation package was furnished to the pipeline company.

Today, the plant is in regulatory compliance for emissions, and is operating with increased capacity, lower operating expense and processes more gas using fewer packages then it had previously. The project timeline of having six new compressor units commissioned and in operation in less than one year from concept to completion, allowed the pipeline company to maximize return by aggressively pursuing new volume commitments from gas producers.

Working together, Dresser-Rand, the North American energy provider (initially) and the pipeline company (ultimately) were able to create and foster a new successful partnership. Safe installation and construction were considered during the design, and the phasing allowed for the construction contractor to remain safe. No lost time or injury accidents occurred during the life of the project.

In word and in deed, those are the foundations from which we build. •

prof

ile

10 insights

I“I always wanted to be a doctor, because I wanted to help people and have a positive impact each day,” says Danielle Guccione. “But then I visited my older sister Leia, who was taking summer classes at Iowa State, and was introduced to her mechanical engineering advisor. Her advisor talked to me about engineering and everything you could do with a degree. She was so passionate and it sounded so exciting! From that point, I wanted to study engineering.”

A native of Hudson, Ohio, U.S.A., Danielle’s family moved to Doylestown, Pennsylvania, U.S.A. when she was four years old. “It was idyllic, with old stone buildings and farms with barns amidst rolling green hills, like something out of a magazine,” says Danielle. At age 14, the family moved again, this time to Stratford-upon-Avon, England, which she describes as “historic, being the birthplace of William Shakespeare. “It was an amazing experience; I went to a British school, so was completely immersed in the culture.” When she was

18, her family moved back to the U.S., to Dallas- Fort Worth, Texas. She enrolled at the University of Tulsa, and graduated four years later with a B.S. in mechanical engineering, with Honors.

Engineering Management Acceleration ProgramDanielle signed on with Dresser-Rand during her final year at Tulsa as part of the company’s Engineering Management Acceleration Program (EMAP), one of the company’s early career and leadership opportunities designed to provide participants with multiple and diverse developmental experiences. The five-year program comprises three one-year rotations in various disciplines, with the final two years comprising a major job assignment.

Danielle began at the Wellsville, New York, U.S.A. facility in the packaging group, followed by rotations in Olean, New York and Houston, Texas, where she focused on product design and new equipment.

“It was a great experience, because I was able to interact with the sales team while learning the technical side of our compressors.” Her final rotation was in Field Service in Baton Rouge, Louisiana, U.S.A. “I went full circle, from design and manufacturing, to sales, then on to services. I gained a broad overview of the organization.”

Danielle’s first job out of the program was in Applied Technology Optimization Solutions, followed by her current assignment as the Service Contract Manager for Phillips 66 – a new position created through a partnership between the two companies.

Service Contract Manager Not Unlike a Maestro

DANIELLE GUCCIONE

Danielle Guccione, service contract

manager, Phillips 66

profile

11

“I work with a number of people at Phillips 66 headquarters as well as the local level on aftermarket issues and various continuous improvement initiatives,” says Danielle. She says the job is “very dynamic and constantly evolving.” She describes herself as an orchestra conductor, whose job is to make sure everything goes smoothly. “I’m also sort of a guinea pig,” she asserts, “we’re kind of writing the book as we go along; and looking to potentially introduce this role to other clients in the future.”

It’s All About PeopleDanielle has been in Services for four years. “It’s challenging,” she acknowledges, “but it provides the most opportunities to really have an impact, knowing you helped with a major issue – such as safety or quality – something that affects the bottom line.”

She describes her job as being more than just about equipment; rather, she says, it’s about people – forming relationships with the client and working toward a common goal. Clients often call Danielle because they know she’ll do her very best to help. “Yes, there are headaches, but that’s also what makes it worth it. Having an impact and achieving something.”

Challenges “Timing is critical,’” says Danielle. “Things can always be faster in the eyes of our clients. When units are down, they’re losing money. Emotions can be high, so communication is critical. You need to get everyone on the same page.”

But, says Danielle, you also learn that you can’t control everything. “There are emergencies, unplanned calls – you have to be adaptable.” She is quick to point out that you can’t take anything too personally.

A Family of STEM Women “I come from a line of high-achieving women,” says Danielle. Her grandmother was a combat nurse in WW II and earned three battle stars. At the conclusion of hostilities, she was the chief nurse at the U.S. Military Hospital in Paris. Her mother has two technical degrees in addition to an MBA, and was an R&D chemist across multiple industry segments. She now inspires high school students by

teaching chemistry. Her older sister, Leia, double majored at Iowa State (mechanical engineering and political science), while going through Naval ROTC. Upon graduating, she attended the Navy’s Nuclear Power School, and became a nuclear engineer on the USS Ronald Reagan (CVN 76). Leia went on to earn an M.S. in sustainable development from the University of London, and is now a manager at the Rocky Mountain Institute.

Off-Hours Away from the job, Danielle loves being active, especially outdoors. She was a Division I Varsity rower at the University of Tulsa. “I’m an avid skier but also love trying new things and challenging myself. I’d like to try an Iron Man triathlon at some point,” she enthusiastically exclaims.

Danielle also considers herself a bookworm. Among her favorites is Ayn Rand’s Atlas Shrugged. “Aspects of the book will resonate with anyone who considers themselves as driven.”

To this day, Danielle is thankful for having had that interaction with her sister’s college advisor concerning a degree in engineering. But even today, not unlike a doctor, hers is a 24/7/365 job where she’s helping people and having a positive impact. “Services never sleeps... we’re always on call. Our clients have very high expectations of us, and we do our best to not only meet, but exceed those expectations.” •

peop

le p

ower

12 insights

Strengthening the communities in which we operate is an important aspect of our Company’s values. The Company and its employees recognize that strong communities are advantageous for growth and prosperity. Worldwide, our employees look for ways to give back, strengthen community programs and support worthwhile causes.

Be the Change You Want to See in the World

Attendees of the annual European Served Areas Sales and Service meeting collect money for worthy charities. They nominate charities, and the management team decides which charity will receive donations that year.

Last year, attendees opened their hearts and wallets and raised $2,135 USD (ZAR 25000) for the Laerskool General Nicolaas Smit Primary school located in the heart of the poverty-stricken area of Pretoria, South Africa. Dresser-Rand team members Shane Janse Van Vuuren, Branch Manager; Ola Adebowale, Regional Director–Sub Sahara Africa; and Ian Sloan, Regional HR manager, presented the gift to the school’s headmaster.

The school provides approximately 1,200 primary school children with schooling, uniforms and transportation for their families. It also assists with shelter where possible, hot meals for the children, community projects to increase employment, and food parcels for the families.

See the people behind our

products and services.

Laerskool General Nicolaas Smit Primary school children created a “thank you” banner to show their appreciation.

The Dresser-Rand team was greeted by the headmaster, Chris Sealie, who has run the Laerskool General Nicolaas Smit Primary school for 17 years. He provided the D-R team with a tour of the school and the local area, where many of the children’s homes are converted shipping containers.

“Witnessing some of the hardships first hand, and listening to true life stories from the headmaster was a leveling experience as we can take many basic things for granted,” shared, Sloan. “We had the privilege to present the donations to a grateful teacher, who will make the money go a long way to positively impact many youngsters.”

Despite the hardships of the surrounding area, children attending the school are very happy to have a place to learn.

ESA Sales and Service Meeting Attendees Open Hearts…and Wallets

13

D-R Congratulated on Outstanding 1st Quarter Safety PerformanceDuPont recently sent a letter to Dresser-Rand and other contractor partners, congratulating us for delivering outstanding safety performance in the first quarter 2015. Year-to-date, DuPont has seen a 60% improvement in contractor safety performance compared with the same period last year.

The letter concluded “Congratulations on your excellent contractor safety performance, and we wish you continued success. Thank you for all that you do, and for being Committed to Zero.”

We are proud to work side-by-side with clients and other contractors who consider safety and health to be a core value, just as we do.

Smile Caravan Reaches Kuito-Bie, Angola

Our D-R Brazil Environmental Solutions business unit supports ONG Aldeia Nissi

(Nissi Village), whose work positively impacts the children and teenagers

of Kuito-Bie. Last year, Dresser-Rand donated solar panels to help generate electricity for the school, materials to

construct a new classroom and household utensils for the kitchen.

Silvia Prado (see photo above, left), an employee of D-R Brazil’s Environmental

Solutions business unit, regularly participates in the “Smile Caravan.” The Smile Caravan is a volunteer group that

dedicates their vacation and holiday time to helping people in isolated communities around the world. In February, the group

traveled to Angola’s Aldeia Nissi School in the remote community of Kuito-Bie and spent 25 days educating students about

dental care, basic first aid, the importance of self-esteem, sex education, and

conflict mediation. They also coordinated enrichment activities for the students

involving music, art and leisure.

14 insights

DDresser-Rand established the Global Engineering Fellowship Program to honor employees who have attained the highest level of distinction through significant

engineering achievements. A

Dresser-Rand Engineering Fellowship is a tribute to employees who have been practicing engineers for 20 or more years and whose work has provided competitive advantages in the energy markets. Honorees are also recognized as industry experts in their engineering discipline, often serving on national, regional and local boards and organizations.

George M. Lucas and Randall W. Moll

represent the third Fellows class inductees since this

prestigious program’s launch in 2012.

“George and Randy have demonstrated exemplary

leadership and made significant contributions to establish and

maintain Dresser-Rand as a technology leader,”

says David Nye, Dresser-Rand Vice President of Technology.

“They have proven their commitment to the future success of Dresser-Rand by effectively advancing the technology of rotating equipment in the energy infrastructure industry.”

George LucasLucas began his career with Dresser-Rand in 1978 as the lead design engineer in developing new syngas turbines and standardizing the engineered turbine product line. He also served as a project engineer in developing compressed energy storage (CAES) expanders for the McIntosh, Ala., U.S.A. CAES plant.

Lucas is the inventor or co-inventor of five issued patents with six patent-pending applications. His key contributions include developing VHP CAES cycle, introducing axial exhaust lines for ETGs and updating the CAES expanders to current design standards. He also provides technical leadership for the power-generation portion of the company’s SMARTCAES® expander trains.

Lucas holds a bachelor’s and master’s degree in engineering from Cornell University and is a member of the American Society of Mechanical Engineers (ASME), the National Society of Professional Engineers (NSPE) and the New York State Society of Professional Engineers (NYSSPE). He also developed a one-hour seminar on CAES, certified by the practicing Institute of Engineering (PIE), that awards one professional development hour of continuing education credit in New York.

Hiring and retaining top-notch technical talent is essential to maintaining market leadership.

Fellowship ProgramWelcomes Moll and Lucas

15

Randy MollMoll’s career began 33 years ago as a steam turbine product engineer handling new unit and repair inquiries for thermodynamic and product engineering proposals. He also wrote or enhanced many of the steam turbine computer programs in use at Dresser-Rand today.

Through a series of assignments with increasing responsibilities, he solved test floor and field vibration issues, developed improvements in steam ends, TT-14 and syngas turbine drivers, and led the engineering group that designed Dresser-Rand’s largest steam turbines.

Currently, Moll manages a team that provides critical engineering input to new equipment and rerates of existing equipment. As a testament to his ability to think big and tackle complex issues with a fresh perspective, he is frequently consulted for assistance on turbine design, repair, overhaul, operation, and troubleshooting.

Moll earned an associate’s degree in applied science from Alfred State College and holds a bachelor’s degree in mechanical engineering from the Rochester Institute of Technology. He is a continuous contributor to the ASME PTC-6 Code Committee and also served as a liaison for the API 612 and 648 task forces.

Honoring our company’s top technical talent is important to shaping our future. •

Engineer (en-juh-neer): person trained and skilled in the design, construction and use of engines or machines, or in any of various branches of engineering. The word is derived from the Latin words ingeniare (“to contrive, devise”) and ingenium (“cleverness”).

engi

neer

’s n

oteb

ook

16 insights

ABSTRACTThis paper presents results from stability analysis and testing on high- pressure CO2 centrifugal compressors, beginning with a preliminary analysis conducted at an academic institution. It includes results of stability tests that were carried out in manufacturer facilities. The tests were done on two different types of compressor, both designed to operate with high content of CO2 in super critical condition. Due to the similarities on testing and results, only the results of one compressor type are presented. An important gain from the tests was a clearer understanding of the behavior of this compressor for the stability for different load conditions.

Stability Testing of CO2 CompressorsRoberto F. de Noronha, PhD

Petróleo Brasileiro S.A. - PETROBRASRio de Janeiro E&P Operation Unit - UO-RIO

Rio de Janeiro, RJ, BRASIL

Editor’s Note: Reproduced with permission of the Turbomachinery Laboratory (http://turbolab.tamu.edu). From Proceedings of

the Forty-Third Turbomachinery Symposium, Turbomachinery Laboratory, Texas A&M University, College Station, Texas, Copyright 2014.

Marcelo Accorsi Miranda, MSc.Petróleo Brasileiro S.A. - PETROBRAS

Rio de Janeiro, RJ, BRASIL

Katia Lucchesi Cavalca, PhDCampinas State University - UNICAMP

Campinas, SP, BRASIL

Edmund A. Memmott, PhDDresser-Rand Company

Olean, NY

Krish Ramesh, PhDDresser-Rand Company

Houston, Texas

Dr. Roberto Noronha has been with Petrobras since 2005 as a rotating equipment advisor. He was previously with Fluminense Federal University as an associate professor. He received his BSc degree (mechanical engineering) from PUC-Rio (1976), MSc. degree from UFPb (1983)

and a UK PhD degree from Cranfield University (1990).

Marcelo Accorsi Miranda is a senior advisor with Petrobras E&P Production Development Projects, Brazil. He has been in the oil and gas business for 35 years. Mr. Miranda is responsible for the conceptual design, specification, selection, and shop test acceptance of turbomachinery. He

is a member of the Turbomachinery Symposium Advisory Committee. Mr. Miranda received a B.S. degree (mechanical engineering) from Universidade Federal do Rio de Janeiro and a M.S. degree (industrial engineering) from Universidade Federal Fluminense.

Prof. Katia Lucchesi Cavalca has been with UNICAMP since 1996 as assistant professor and since 2009 as full professor, head of the Laboratory of Rotating Machinery. She received a BSc degree (mechanical engineering) from UNESP State University of São Paulo (1985), MSc degree from

UNICAMP University of Campinas (1988) and a PhD degree from Polytechnic of Milan (1993).

Dr. Edmund A. Memmott is a principal rotor dynamics engineer at Dresser-Rand. He has been with the company since 1973. He received his AB degree from Hamilton College (Phi Beta Kappa) (1962), AM degree from Brown University (1964), and PhD degree from Syracuse University (1972),

all in the field of mathematics. He was on the API Task Force that wrote the 2nd Edition of API 684 and is doing the same for the 3rd Edition. He is a member of the ASME, the CMVA, the Vibration Institute, the MAA, and the SOME subcommittee of API.

Dr. Krish Ramesh is a team leader (engineering tools) at Dresser-Rand. He has been with the company since 1996. He received his MTech degree (Mech.Engg.) from IIT, Chennai, India (1990) and PhD degree in rotor dynamics from Virginia Tech (1996). At Dresser-Rand, he is responsible for the engineering programs used in centrifugal compressor selection and rotor dynamics.

engineer’s notebook

17

TIntroductionThe so called Replicants project was developed for oil and gas production at the Brazilian offshore pre-salt fields. One of main objectives was to develop a single standard project for eight FPSOs to be located at different pre-salt sites, minimizing costs and reaching high productivity. An important design challenge was the compression system, due to the high carbon dioxide content of the associated gas, requiring its removal and reinjection. Considering furthermore that the same project should be capable to handle a high molecular weight range varying among different sites and among different modes of operation in the same site, an elaborate compression plant was devised, comprising twelve compressors in five different services: the main and export compressors, respectively upstream and downstream the CO2 removal system, the low- and high-pressure CO2 compressors, handling the side stream from the membrane separators, and the injection compressors.

Of concern of this paper are the CO2 high-pressure (HP) and the injection compressors. The first is a nine stage, back-to-back centrifugal compressor, while the second is a six-stage one, also back-to-back, but of a single section only. The CO2 compression train was specified to operate on a fixed pressure increase condition, from 4 to 250 bar, but with MW ranging from 26 to 39 (CO2 content from 33% to 83%). The operating conditions of the injection compressor are more diverse, with MW varying from 21, in the case the membranes are not in operation, to 39 (CO2 content from 3% to 83%). The suction pressure is also fixed, of 250 bar, but the discharge pressure depends on the column weight of the injection well and may vary from 300 to 550 bar. Thus, the injection compression will work on more demanding conditions, due to its higher discharge pressure and higher molecular weight range. Nevertheless, despite these differences, both are to operate with high density, super critical fluid.

Considering their highest density condition, the location in API 617 (2002) Level I screening criteria is depicted in Figure 1. Taking into account the importance of gas density on the stability of centrifugal compressors, stability tests were specified for these two compressors. Similar to API’s unbalance response test, it was specified to verify the stability calculations and, as such, the test was done in only one for each type of compressor,

i.e. one out of the 16 CO2 HP compressors for the eight FPSOs and another for the 16 injection compressors. Thus, the testing procedure had to contemplate different operating conditions and also extrapolation criteria to determine minimum stability margin applicable to all 16 compressors based on test results from only one unit.

There are two sources of instability inherent to a centrifugal compression process, the first of which is the labyrinth seal that separates the different levels of pressure inside the compressor and the second is the actual movement of the gas caused by the rotation of the impellers, generally referred to as “aerodynamic cross-coupling.” Published papers and internal records of operator companies show these fluid induced instabilities may lead to high production losses. Thus, stability analysis is part of the dynamic design of the rotor. An excellent reference on these issues is the recent book by Childs (2013) and much is in API 684 (2005).

In the current context of dynamic analysis of rotating machines, there are various procedures for stability analysis of these systems, many of them based on numerical modeling and analysis of eigenvalues and eigenvectors. It is known, however, that there are many uncertainties associated with numerical modeling, due to theoretical approximations or uncertainties related to manufacturing and assembly of the rotor and its components. Indeed, a survey on numerical modeling and stability analysis conducted by Kocur et. al. (2007) points out to variations on the results that range from 10 to 1. Therefore, techniques were developed for experimental evaluation of the stability of rotors.

Figure 1. Location of the compressors in Level I screening criteria for the highest density condition.

engi

neer

’s n

oteb

ook

18 insights

The operator has previous experience with another OEM with stability testing with a magnetic bearing exciter on a low-pressure, low-density downstream compressor, without any device to enhance damping or reduce cross-coupling (Pettinato et. al., 2010). Stability was low and testing was important in order to assure it was within the specified limits.

The operator has a freighted platform where this OEM was involved in stability testing with magnetic bearing exciters of high-pressure, high-density back-to-back CO2 compressors for injection service (Colby et. al., 2012). For those compressors there were devices used to enhance damping – squeeze-film dampers centered by a mechanical spring at the tilt pad journal bearings and hole pattern seals at the division wall, and there were devices used to reduce cross-coupling – shunt holes and swirl brakes at the division wall hole pattern seals and swirl brakes at the impeller eye toothed labyrinths. All of those compressors had dry gas casing end seals.

Overall stability is a net result between the forces caused by destabilizing mechanisms and those from any stabilizing devices. The present ones are similar to the high-pressure, high-density CO2 compressors mentioned above and have the same stability enhancing devices. They also have tilt pad bearings and dry gas casing end seals. In common with the equipment mentioned above, the present ones share the fact that the guaranteed point is a condition where compressors seldom operate. Testing is required in more than one load condition in order to obtain an experimental curve of log decrement vs. load.

Activities at UNICAMPIn order to provide the operator with a better understanding of stability testing, theoretical and experimental studies were conducted by researchers at Campinas State University (UNICAMP). The most important feature of the test bench, shown in Figure 2, is that it is basically a Jeffcott rotor supported on cylindrical bearings with an electromagnetic actuator on the non-drive end. Therefore, its dynamic behavior is well known and, hence, due to its simplicity either for practical experiments or for theoretical modeling, the test rig can be considered appropriate for preliminary testing of the proposed method.

The electromagnetic exciter

To obtain the frequency response of the rotor and perform the procedures for stability analysis, the test bench is equipped with an electromagnetic actuator as a source of external excitation (Mendes, 2011). Figure 3 provides an additional view of the support structure and of the local coordinate system.

A finite element simulation showed that the lowest natural frequency of the structure and coils set is around 576 Hz (Figure 4), which is far above the operating range of the test bench (approximately 0 to 100 Hz). To excite all the inertia involved in the test bench, the actuator is able to apply an excitation of 150 N up to 100 Hz, considering an air gap of 2.5 mm and a current of 3A.

Figure 2. UNICAMP test bench.

Figure 3. (a) Support structure of the magnetic actuator; (b) Coordinate system rotated by 45° (Mendes, 2011).

(b)(a)

Figure 4. First mode of vibration of the actuator at 576 Hz (Mendes, 2011).

engineer’s notebook

19

The actuator force control is done by controlling the magnetic field (Chiba et. al., 2005), as shown in Figure 5, where Bref is the magnetic field corresponding to the reference force, which is compared with the measurement of the magnetic field made by a Hall-type sensor. This comparison is the input to the proportional controller. The output voltage of the controller is converted into current by a PWM amplifier (operating in current control mode) and sent to the actuator coil, generating a magnetic force Fm. Thus, using the differential set with a magnetic bias field (Bb), the resulting linear force can be written as in Equation 1 (Schweitzer et. al., 2009; Maslen, 2000).

Fm = 4·Ag·Bb·Bref/µg (1)

Where Ag is the area of each pole, Bb is the bias magnetic field, Bref is the desired magnetic field, and µg is the air permeability. The force measurement model was calibrated using a short shaft assembly monitored by load cells and accelerometers in both ends (Castro et. al., 2007).

Unlike a magnetic bearing, in the case of a magnetic actuator, control is performed based on the magnetic field because the goal is that the actuator applies a force on the shaft of predetermined external excitation, regardless of the shaft displacement.

Test methodology

The methodology for the tests aims to establish a stability criterion for rotating systems based on estimation techniques of the logarithmic decrement, using an electromagnetic actuator as a source of external excitation. One of the proposed techniques, termed Blocking Testing (Kanki et. al., 1986, Pettinato et. al., 2010), is applied in a time domain and does not require the measurement of the excitation force, while in classical methods of modal analysis, it is necessary to know the external excitation as this is part of the experimental model. This procedure is compared with the classical excitation techniques in the frequency domain, called by Pettinato et. al., (2010) as Stepped Sine Testing.

Figure 5. Diagram of the magnetic actuator system (Mendes, 2011).

During the tests, the latter was applied first in order to establish, through the analysis of the frequency response, the forward and backward natural frequencies, which were then used for the Blocking Testing excitation. Although instability generally appears only through the excitement of a forward precession mode, one must also identify backward modes in order to differentiate them properly and get the log dec strictly associated with the forward mode. After all, even when they are not dominant, both are present in the dynamic response, affecting the modal damping factors and thus the logarithmic decrement (Cloud et. al., 2009). Strictly speaking, forward and backward excitation are actually not necessary. Accurate measurements of the forward mode can be obtained using appropriate ID techniques (MOBAR or PEM) using only forward or directional excitation. The best estimates for a particular mode are provided when the excitation focuses on that mode, but it’s not necessary to excite both modes in order to get accurate estimates of either. Nevertheless, in order to get

a broader insight of the results, tests were done with forward and backward frequency excitation for each operation condition.

In the test bench, the operating condition was modified changing the rotation speed. Thus, in the tests, the same procedure was applied to a single

rotor configuration, whereas with five different levels of stability represented in this case by five different rotation frequencies of the rotor, namely, 50 Hz, 55 Hz, 70 Hz, 75 Hz, and 80 Hz, i.e., above and below the critical speed which is about 62 Hz.

Because it is a validation of a testing procedure for the identification of the logarithmic decrement, it becomes essential to evaluate its sensitivity to different levels of damping. The behavior of the real part of the eigenvalues to be estimated at different rotational speeds, gives the tendency to change sign from negative to positive, thus setting the threshold of instability in the frequency domain. In the case of the logarithmic decrement, this tendency goes from positive to negative (Equation 2).

where δ is the logarithmic decrement, λ the eigenvalue and ζ the damping factor.

(2)

engi

neer

’s n

oteb

ook

20 insights

For time domain identification of the stability level of a rotating system, the MOBAR method was studied and tested. The MOBAR, Multiple Output Backward Autoregression (Hung and Ko, 2002; Cloud et. al., 2009) is an identification method of modal parameters that considers multiple degrees of freedom, which are analyzed in the time domain. An important feature of this method is that it does not require the measurement of the excitation signal.

The recommended test for generating the output response signals is the Blocking Testing, where a rotating excitation at a particular frequency is applied for a period of time and then stopped abruptly. The application of MOBAR on the results of the Blocking Testing provides a dynamic system whose eigenvalues are used to calculate the logarithmic decrement. Moreover, this rotating excitation force can be forward or backward, as shown in Equation 3 (Jang et al., 1996):

Figure 6 shows the response of the system under Blocking Testing. Note the decrease in amplitude after the rotating force has been interrupted. Thus, one can analyze the displacement decrement after excitation cessation (Cloud, 2007).

Another way to obtain the modal parameters of the system is to perform a sine sweep, the Stepped Sine Testing, where non-synchronous excitations are applied to the rotor in a predetermined range of frequency in two orthogonal directions of the vibration plane of the rotor and the frequency response function is obtained in the frequency domain. After the test, one may apply classical techniques of modal analysis in accordance with the literature, such as Ewins (1984), due to the knowledge of the external excitation force.

(3)

Figure 6. Blocking Testing – forward excitation at 4,800 rpm horizontal response on bearing 2.

Figure 7. FRFs reconstitution at 3,900 rpm excited in horizontal direction: (a) Bearing 1 – horizontal; (b) Bearing 1 – vertical; (c) Bearing 2 – horizontal; (d) Bearing 2 – vertical.

engineer’s notebook

21

In this paper, however, a model identification technique based on error minimization was applied (prediction error method or PEM). The systems identified are the type SIMO (Single Input Multiple Output). Therefore, the responses of horizontal and vertical displacement sensors were considered. Figure 7 illustrates the application of this method. After obtaining the experimental FRFs (black), the mathematical model was adjusted to the experimental points, which allowed the extraction of modal parameters such as natural frequency, damping factor and vibration modes. For each rotation, two FRFs were identified (y and z directions) in the frequency range of 40 to 80 Hz. The reconstitution at 3,900 rpm excited in horizontal direction (y) is represented in Figure 7 (green line).

Results and discussion

Figure 8 shows the logarithmic decrement and damping ratio vs. rotation for the applied

identification methods. The experimental results are compared to the responses from a software developed at UNICAMP for rotor dynamic analysis. It is noticeable that the numeric simulation values are in very good agreement when compared to forward precession experimental results. The logarithmic decrements related to backward natural frequency are more conservative concerning the instability analysis. However, it is important to point out that even for backward the tendency is in good accordance between experiments and numerical simulations.

The frequency domain tests led to more accurate results; however, they demand higher time-costs and also the previous knowledge of the excitation force which measurement can be one more uncertainty source in the entire procedure. The time domain tests are easier and faster; however, attention must be given to the measurements and identification techniques to achieve an acceptable accuracy. Moreover, in the tests carried on in the laboratory, the time domain tests led to more conservative stability analysis.

Finally, it is important to highlight the main objective of this section. The approach with the academic environment allowed a better understanding of the whole procedure theoretical basis of the stability tests and procedures. Hence, a standard procedure to be regularly adopted to qualify rotating machines (applied at the operators’ plants) could be established.

Design of Compressors for High Density Gases Against Fluid Induced InstabilityThe early history of the OEM in the usage of stability-enhancing devices to insure stable operation with high-density gases has been described in a series of papers by Coletti and Crane (1981), Shemeld (1986) and Memmott (1990 and 1992). These papers also describe the application of full-load, full-pressure tests in the shop on inert gas, starting in the 1970s, and on hydrocarbon gases starting in the 1980s, to validate stable operation.

During the 1970s, the OEM started using the stability enhancing devices described as follows and applied them to both original equipment and upgrades. Tilt pad seals were first used by the OEM in 1972 and have been applied to more than 600 compressors since then. Tilt pads were inserted into the outer rings of oil-film seals to make them

Figure 8. Simulation, SIMO (Experimental Stepped Sine Testing) and MOBAR (Experimental Blocking Testing): (a) Log Decs vs. Rotation; (b) Damping ratio vs. Rotation.

engi

neer

’s n

oteb

ook

22 insights

act like tilt pad bearings and not like instability- inducing sleeve bearings. Damper bearings (squeeze-film dampers in series with tilt pad journal bearings) were first used by the OEM in 1973 and have been applied to more than 800 compressors since then. These are used to provide optimum damping at the journal bearing locations. Shunt holes were first applied by the OEM in 1974 and have been routinely applied to most medium- and high-pressure compressors since then. Shunt holes tap the last stage diffuser to send higher pressure gas radially in to the entrance of the division wall seal of back-to-back compressors or at the balance piston seal of an in-line compressor. This eliminates the destabilizing tangential inlet swirl to those seals that otherwise would come down from the back side of the last impeller. See the papers listed in the previous paragraph for sketches of these parts and the early history of their usage. For a more recent survey of damper bearing applications see (Memmott, 2010).

Honeycomb seals in centrifugal compressors at the division wall or balance piston were first applied as damping devices in the mid-1990s. These were ideal locations to apply the large amount of direct damping that could be obtained from these seals for high density gases to stabilize the compressor. Honeycomb seals had been used years before that at the balance pistons of some syn gas compressors before rotordynamic benefits were recognized. The experience of three different OEMS showed that the honeycomb seals needed to be deswirled and shunt holes were used to do this. See the papers by Memmott (1994), Gelin et. al., (1996) and Camatti et. al., (2003). Analytically, hole pattern seals have similar dynamic characteristics as honeycomb seals and a summary of experience with the use of honeycomb and hole pattern seals is given by Memmott (2011). The OEM has applied hole pattern and honeycomb seals to more than 400 compressors, most of them of the hole pattern type.

Papers were written by Nielsen and Myllerup (1998), Nielsen et. al., (1998) and Moore and Hill (2000) analyzing and describing the application of swirl brakes, which are used to deswirl the gases entering toothed labys, such as at the impeller eyes, and at the hole pattern and honeycomb seals at the division wall or balance piston.

Since the early 2000s magnetic bearing exciters have been used by the OEM in full-load, full-

pressure tests to validate the predicted log decs and the design of high-pressure, high-density compressors for stable operation. All of those tests were conducted on compressors with hole pattern seals and shunt holes. Some of the compressors had squeeze-film dampers in series with the journal bearings and some did not. Data was taken at low, intermediate and full pressures. With the hole pattern seals, as the pressure (and thus the density) increased, the log dec increased significantly, unlike what would be expected with toothed labyrinth seals. There was good agreement between the testing, which was done with a frequency sweep and a single degree of freedom model for the data collection and the analysis of the data, which analysis is described in the next section of this paper. See the papers by Moore et. al., (2002), Moore and Soulas (2003), Gupta et. al., (2007), Soulas, et. al., (2011), Memmott (2011), Gupta (2011), and Colby et. al., (2012).

For all of the stability analyses shown in those papers (and in this paper) the log decs were calculated with the effects of the toothed labyrinths, the hole pattern seals and with the inclusion of the MPACC number for the effects of the anticipated cross-couplings from the impeller-diffuser interaction. See the discussion in the next section for a description of the MPACC number and the necessity for its inclusion along with the effects of the toothed labyrinths and the hole pattern seals.

Also see (Chochua and Soulas, 2006) where a method was proposed for computations of rotordynamic coefficients of deliberately roughened stator gas annular seals using computational fluid dynamics (CFD). Rotordynamic coefficients predicted by CFD for a hole pattern seal were in good agreement with seal test data and the ISOTSEAL 1-d code prediction.

Because of the experience described in Memmott (1994) the OEM for that compressor began to use honeycomb seals at the division wall or balance piston in all high-pressure high-density applications and also in many medium-pressure medium-density applications. Since 1999 this OEM has used hole pattern seals at those locations. There has been no magnetic bearing exciter testing by this OEM of compressors with toothed labyrinth seals at the division wall or balance piston.

A further description from the 1994 paper will illustrate more in depth why toothed labyrinths are no longer used by this OEM at the division wall

engineer’s notebook

23

or balance piston of high-pressure, high-density compressors. This is a back-to-back compressor and has a division wall. There were no problems on the low pressure mechanical test. On the FLFP HC test there were unacceptable levels of SSV (sub-synchronous vibration) with a toothed labyrinth at the division wall, even though shunt holes were used. This was not a test with a mag bearing exciter, it was done before the mag bearing exciters were used. The shunt holes had been used before on hundreds of compressors with great success. But toothed labyrinths do not provide much direct damping. When the toothed labyrinth was replaced by a honeycomb seal without shunt holes (they were just covered up) there still were unacceptable levels of SSV, and the frequency was higher. It showed that the large direct damping from the honeycomb was insufficient without deswirling. It also showed that there was a significant direct stiffness term from the honeycomb seals. The shunts were uncovered and it was stable with the honeycomb seal and this proved again that shunt holes work.

The paper by Gelin et. al., (1996) describes much the same experience. It describes two compressors which each started out with a toothed labyrinth at the balance piston and each of which ended up with a honeycomb seal with a new shunt hole system. This also was before mag bearing exciter testing.

Stability Analysis by the ManufacturerThe log decs were calculated using the OEM’s rotor dynamic software suite, which is a state-of-the-art program linking the centrifugal compressor aerodynamic selection and compressor modeling tools with the rotor dynamic programs into one cohesive engineering tool (Ramesh, 2002).

The tilt pad bearings were analyzed by the program of Nicholas (Nicholas, et. al., 1979). The stability analysis was done by the transfer matrix program of Lund (Lund, 1974, Smalley, et. al., 1974). The toothed labyrinths were modeled by the two control-volume bulk flow method program DYNPC28 by Kirk (Kirk, 1985. 1986. 1990), The hole pattern seals were modeled by the ISOTSEAL program of Texas A&M (Kleynhans and Childs, 1996, Holt and Childs, 2002, Soulas and San Andrés 2002, and Childs and Wade, 2003). Usage of the tilt pad bearing and stability programs, especially as related to the requirements of API 617, is described in (Memmott, 2003).

The excitation arising from the centrifugal impellers is estimated using a modified form of the Alford and Wachel numbers (Alford, 1965 and Wachel and von Nimitz 1981). After benchmarking the formulation on numerous test cases operating with different mole weight gases, the OEM adopted the form referred to as the Modal Predicted Aero Cross-Coupling (MPACC) (Memmott, 2000a). Instead of the factor of MW/10 that is used in the Wachel number it is replaced by 3. By taking a modal sum of these excitations at each impeller based on the first forward whirling mode shape, an effective aerodynamic cross-coupling is calculated and applied at the mid-span of the rotor for the API 617 7th Edition Level II analysis, along with the dynamic coefficients from the toothed labyrinths and the hole pattern seals. The formula at each impeller is the same as was adopted by API 617 7th edition for the arithmetic sum of the anticipated cross-couplings at each impeller used in the Level I screening criteria, without the labyrinths and without the hole patterns.

In the paper (Memmott, 2000a) it was shown that the MPACC number needed to be included in the Level II analysis in addition to the stiffness and damping coefficients for the labyrinths in order to predict instabilities seen in several different types of compressor, a 400 bara (5,800 psia) discharge back-to-back gas injection compressor, a large propane compressor with multiple side streams, a 168 bara (2,432 psia) discharge in-line CO2 compressor and a 94 bara (1,368 psia) discharge in-line syn gas compressor.

The gas injection compressor is predicted to be stable from a Level II analysis with all of the labyrinths, but without the MPACC number. It had a toothed division wall labyrinth with shunt holes. If the MPACC number was included along with all of the labyrinths it is predicted to be unstable. It was not stable until the division wall toothed labyrinth was replaced with a honeycomb seal with shunt holes. Then it was predicted to be stable with the inclusion of the MPACC number, the toothed labyrinths and the honeycomb seal with shunt holes. This experience is also discussed in (Memmott, 1994), which was written before the introduction of the MPACC number.

The propane compressor is predicted to be stable with a toothed balance piston seal and the other toothed labyrinths, but without the MPACC number. If the MPACC number is included along

engi

neer

’s n

oteb

ook

24 insights

with all of the labyrinths it is predicted to be unstable. It was not stable until application of damper type seals with small pockets and swirl brakes at all of the labyrinths. With the MPACC number and the coefficients for these damper seals then it is predicted to be stable. This compressor experience also showed that the modal sum of the anticipated cross-couplings at each impeller should be applied, and not with an overall density ratio of outlet density over inlet density, or with a section-by-section sum based on density ratios at each section. This experience is also discussed in (Memmott, 2000b).

The CO2 compressor and the syn gas compressor are predicted to be stable with a toothed balance piston seal with no shunts and no deswirling and all the other toothed labyrinths and non-damper bearings, but without the MPACC number. If the MPACC number is included along with all of the labyrinths, then both are predicted to be unstable. They were not stable until application of damper bearings and shunt holes at the balance piston. Swirl brakes were not used at those locations. With the MPACC number and damper bearings and the coefficients for the toothed labyrinths with shunts at the balance piston both are predicted to be stable. The CO2 compressor is also discussed in (Memmott, 1990), with no results on modeling of the labyrinths and again in (Memmott, 2010) with the results for modeling the labyrinths. For the syn gas compressor without the 30/MW x MW/10 = 3 modification to the Wachel formula, the instability seen with the non-damper bearings and no shunt holes is not predicted. With the modification it is predicted to be unstable. For the syn gas compressor, the MW is 9.3 in the make-up section and 11.44 in the recycle impeller. The syn gas compressor experience led to the modification of the Wachel formula. The experience with the syn gas compressor is discussed again in (Memmott, 2010), with the results for modeling the labyrinths.

For the tested CO2 HP compressor, analytical curves of log dec vs. average gas density were developed to show the striking effect of increasing log dec with increasing density at the hole pattern seals at the division wall and to study the effects of the various parts in the system. See Figure 9. The effects of journal bearing design clearances, squeeze-film damper design clearances and eccentricity, and hole pattern clearance on the stability of the compressor can be studied. Also included are the effects of minimum and maximum oil inlet temperatures.

Minimum design oil temperature was paired with minimum design bearing and damper clearance and maximum allowable eccentricity for the damper for the stiffest system. Maximum oil inlet temperature was paired with maximum design bearing and damper clearance and centered damper for the softest system. The log decs include the effect of the MPACC number. The testing was done up to a value of approximately 7 LB/FT^3 (112 kg/m3), and this was full-load and full-pressure, but not full density. The last three points on each curve had the same final discharge pressure and approximately the same load, but not the same MW. It is seen that plotting against discharge pressure or against load is not always appropriate.

Prior to the tests, the log decs of the tested compressors were calculated for the stipulated operating conditions and the measured bearing and gas annular seal running clearances throughout the compressor by the OEM. Table 1 presents the calculated log decs for the tested CO2 HP compressor based on the API 617 Level II analysis with the inclusion of the MPACC number.

Curve (b) in Figure 13 shows the first forward mode calculated values of log dec in terms of average gas density for the tested CO2 HP compressor on the three different load conditions. As the average gas density increases, so does the damping of the damper seal in the division wall which leads to the growth of the log dec, as shown in Table 1 and Figure 13.

Figure 9. Log dec vs. average gas density for the CO2 HP compressor.

Table 1: Calculated log decs for the tested CO2 HP compressor.

engineer’s notebook

25

Stability Testing at the ManufacturerTest Description

Stability tests were carried out in the first CO2 HP and injection compressors to undergo factory acceptance tests, from a total of 16 each. Hence, they are called by the operator as a type test, that is, they are not for all equipment, but only for one of a type. As discussed in the introduction, the purpose of the tests is to establish the minimum stability margin of a machine type based on test performed on one of the machines. The operator technical specifications stated that the log dec should be measured on at least two different operating conditions in order to obtain an experimental curve of its dependency to cross coupling stiffness for the tested compressor. The extrapolation of this curve to the worst design conditions, i.e., those that produce the smallest log dec is discussed next.

Figure 10 describes the procedure recommended by the operator to obtain corrected log decs from on a hypothetical compressor. It depicts, in blue, three curves of calculated log dec, of which the center one corresponds to the actual test conditions, such as bearing clearance and preload, and the lowest one is the curve of minimum calculated log dec, i.e., the calculated curve for the worst conditions of these parameters, e.g. minimum clearances. Similarly, the three curves in red represent the same curves, but corrected from the results of the test. Of these, what matters is the curve of minimum corrected log dec obtained from the minimum calculated log dec curve, translated by the value of the largest difference between the measured and calculated values of log dec for the different test conditions.

In accordance with the procedure established by the manufacturer for factory acceptance tests, the stability tests were performed in sequence with another type test, the FLFPFS (full load, full pressure and full speed) test, using the same test bench in three operating conditions, called 100%, 50% and low load. The first condition is the same as the FLFPFS test, the second with approximately 50% gas power and the last one with very small power, all three conditions were established before the test and with the log decs also previously calculated (see Table 1). The rotation speed during stability testing was the maximum continuous, the same speed of the FLFPFS. That is, while on the UNICAMP bench the stability level was varied by rotation; at the factory it was through load. The excitation applied to the rotor was accomplished by a magnetic bearing coupled to an extension of the rotor on the non-drive end, as represented by the sketch in Figure 11.

Figure 10. Hypothetical curves of log dec vs. cross-coupled stiffness – calculated curves in blue; corrected curves in red.

Test results

Stability test data was acquired and analyzed by a consulting firm specialized in modal testing. Similar to what has been done on the UNICAMP bench, a frequency domain MDOF method was used. It was based on the PolyMAX modal parameter estimation (Peeters et. al., 2004a,b and El-Kafafy et. al., 2012) and was applied to determine the natural frequencies and damping of the first forward and backward modes. As represented in Fig. 11, top left, signals that excite the forward or the backward modes were provided to the windings through the magnetic bearing controller. A sweep sine signal over the entire frequency range was used. Figure 12 shows spectra produced by successive sine wave frequency sweeps. Time domain MDOF parameter estimation was also attempted but did not produce consistent results.

Figure 11. Representation of a compressor with the magnetic bearing attached to the non-drive end.

engi

neer

’s n

oteb

ook

26 insights

Table 2 shows the measured log decs. In blue are the results that presented acceptable repeatability, that is, when at least five measurements were within the range of the mean value ± 0.03, in red when repeatability was unacceptable. For forward modes, only the frequency domain results for low and 50% loads presented acceptable repeatability. All three forward mode results, good or not, are higher than the respective calculated values presented in table 1. Hence, the stability analysis done by the manufacturer is basically conservative.

Based on the two forward log decs with acceptable repeatability, the required Figure 10 curves were constructed and presented in Figure 13, where the dashed lines are calculated and the full ones are the corrected ones based on the test results. In line with Figure 10, the corrected curves of Figure 13 were obtained as follows:

• For low load, the corrected maximum and minimum values are the predicted ones, plus, the difference between the measured and predicted values for the tested rotor.

Figure 12. Spectra generated by sine wave frequency sweeps.

Table 2. Measured log decs - CO2 HP compressor.

• For 50% load, an equivalent procedure applies.

• For 100% load, since there is no measured value for this condition, the three corrected values, maximum, tested rotor and minimum are calculated by adding to the predicted values the difference between the measured and predicted values for 50% load.

The abscissa was changed to average gas density, since, as discussed in the previous section, it is more straightforward and more appropriate for high-density compressors then cross-coupled stiffness. Once again, the figure shows that the calculated values are very conservative, e.g., while the minimum predicted log dec is around 0.5, the test result for low load shows that this value should be above 1.0. Comparing Figure 13 with Figure 1, it can be seen that the average gas density for the shop load tests did not reach that which can be seen in the field. Nevertheless, as presented next, the full load stability response already presented a flat response.

For comparison purposes, Figures 14 and 15 present frequency response functions for low load and 100% load. While for the first FRFs, the natural frequency is clearly defined by the peak amplitude and phase shift on the sensor signals, for the second set of FRFs, identifying the natural frequency is very difficult since the curves present a flat aspect. It is noteworthy that another compressor with very similar specification but from another manufacturer also presented this flat response when tested at full load.

Figure 13. Curves of log dec vs. average gas density for CO2 HP compressors – dashed lines are calculated; solid lines are corrected from test results.

engineer’s notebook

27

Results Discussion At least at the operator where the two first authors work, there is a saying that the design condition is a point where it is sure that a machine will not operate. Taking this into account and also noting that since the stability test is a factory acceptance test and not a test done on a test bench rotor, where good agreement between experimental and numerical model results are important, it should be pointed out that the results of the test, showing that the measured stability level is higher than that predicted by the OEM, is considered desirable by the operator, leading to a more robust compressor in the sense that it will also be stable in off-design conditions. This desirability is further enhanced by the survey conducted by Kocur et al. (2007), where a compressor designed and tested as unstable had its stability analyzed by many specialists, most of them with positive log dec results.

Nevertheless, a discussion on the results is an opportunity that should not be over looked since it provides an occasion to try to get a better insight on the contribution of the rotor components to its dynamic behavior. To begin with, there is the tilting pad and squeeze film damper set. The unbalance

Figure 14. FRFs for the four compressor probes – forward precession and 0% load.

Figure 15. FRFs for the four compressor probes – forward precession and 100% load.

response test, where, in addition to the rotor itself, these two components are the only ones to provide stiffness and damping, gave a good agreement with the calculated results (and also showed well damped behavior). However, on an unbalance response test there is only synchronous vibration. During the magnetic bearing actuated stability test, the tilting pad bearings are simultaneously subjected to synchronous and sub-synchronous vibration, while API 617 requires that the bearing dynamic coefficients be synchronously reduced. So, good agreement between calculated and measured behavior during an unbalance response test does not mean that the contribution of the bearing and squeeze film damper sets will be well predicted during a stability test. Then all stability contributions should be considered together: the destabilizing effects of the labyrinth seals and aerodynamic cross couplings, and the extra damping provided by the bearings, squeeze film dampers and hole pattern seal, discussed next.

In previous sections a historical retrospective of compressor design and modelling at the OEM is given, with examples on why the OEM has stopped using labyrinth seals at the division wall or balance piston of high-pressure and high-density compressors and, more important to this discussion, how the MPACC method of calculating the aerodynamic cross-coupling forces has evolved, describing four examples where only after this calculation methodology was applied that it was possible to obtain stable behavior of the compressors. In other words, only after damper seals, honeycomb or hole pattern, that could neutralize the MPACC destabilizing forces were considered. Additionally, the tendency, for the present compressor, of the calculated stability increasing with density, i.e., with pressure and MW, shows that it is the hole pattern seal that produces a dominant effect on the stability of the present compressor. It is dominant because it was oversized during design in order to neutralize the MPACC destabilizing force.

The testing procedure was specified based on a previous experience of the operator with stability testing, but on low-pressure, low-density gas compressors without any stability enhancing device (Pettinato et. al., 2010). So the main results of the present test, of higher density leading to higher stability level, full load flat FRFs and tested log decs higher than calculated ones did produce some (pleasant) surprise onto the operator staff.

engi

neer

’s n

oteb

ook

28 insights

Nevertheless, the results are consistent and in line with previous experience of the manufacturer and also consistent with witnessed tests by the operator at another manufacturer.

Conclusion Compressors for high density gas are susceptible to fluid-induced instability. Events that have produced high production losses are found in literature and also in internal records of operators. In order to avoid this problem, hole pattern or honeycomb seals with deswirling devices have become standard practice for many OEMs. For the present compressors hole pattern seals are used at the division wall and the deswirling devices are shunt holes and swirl brakes at the division wall and swirl brakes at the impeller eyes. Squeeze-film dampers are also used at the tilt pad bearings.

Numerical modeling methods have been developed in order to assure stable behavior of these machines. Although a detailed procedure for the numerical evaluation of the stability level of centrifugal compressors has been added in the API 617 standard, experimental evaluation of the stability level of centrifugal compressors are becoming more common.

Stability tests were done on two different types of compressors, both designed to operate with high content of CO2 in super critical condition. Due to the similarities on testing and results, only the results of one compressor type are presented. Test data were acquired and analyzed taking into account forward and backward modes of vibration and employing

frequency MDOF and time domain acquisition methods, obtaining results with good repeatability in two different operating conditions. The final result was a curve of experimentally corrected minimum log dec versus average gas density applicable for that type of compressor.

The results also showed that the manufacturers´ log dec calculation procedure is very conservative for these particular compressors in the sense that the experimental results were greater than the calculated ones. The operator has experienced that over designing a balance piston seal damper in a straight-through compressor might lead to excessive susceptibility to unbalance, but regards that for a back-to-back unit this is desirable since it gives higher endurance against seal degradation and robustness to off design operation conditions. This conservatism is granted by the MPACC aerodynamic cross-coupling procedure used by the manufacturer on the stability verification during the design of the compressor.

Despite the hardware limitations, the support provided by the academic institution was clearly advantageous in the aspect of allowing the preparation of the operators´ professionals across theoretical basis related to the implementation of the stability tests.

Finally, it is considered as an important gain from the tests a clearer understanding of the behavior of this compressor concerning its stability for different load and density conditions.

ReferencesAlford, J. S., 1965, “Protecting Turbomachinery from Self-Excited Whirl,” ASME Journal of Engineering for Power, Vol. 38, pp. 333-344.

API Recommended Practice 684, 2005, “API Standard Paragraphs Rotordynamic Tutorial: Lateral Critical Speeds, Unbalance Response, Stability, Train Torsionals and Rotor Balancing,” 2nd edition.

API Standard 617, 2002, “Axial and Centrifugal Compressors and Expander-compressors for Petroleum, Chemical and Gas Industry Services,” 7th edition.

Camatti, M., Vannini, G., Fulton, J. W., and Hopenwasser, F., 2003, “Instability of a High Pressure Compressor Equipped with Honeycomb Seals,” Proceedings of the Thirty Second Turbomachinery Symposium, Turbomachinery Laboratory, Department of Mechanical Engineering, Texas A&M University, College Station, Texas, pp. 39-48, September 8-11.

Castro, H.F., Furtado, R.M., Cavalca, K.L., Pederiva R., Butzek, N., Nordmann, R., “Experimental performance evaluation of magnetic actuator used in rotating machinery analysis,” Journal of the Brazilian Society of Mechanical Science and Engineering, pp. 99-108, 2007.

Chiba, A., Fukao, T., Ishicawa, O., Oshima, M., Takemoto, M. and Dorrell, D.G., 2005, Magnetic Bearings and Bearingless Drives, Elsevier.

Childs, D. W. and Wade, J., 2003, “Rotordynamic-Coefficient and Leakage Characteristics for Hole-Pattern-Stator Annual Gas Seals-Measurements Versus Predictions,” ASME Paper 2003-TRIB-211, Proc. of 2003 STLE/ASME Joint International Tribology Conference, Ponte Vedra Beach, Florida, October 26-29.

Childs, D. W., 2013, “Turbomachinery Rotordynamics with Case Studies,” Minter Spring Publishing.

Chochua, G. and Soulas, T. A., 2006, “Numerical Modeling of Rotordynamic Coefficients for Deliberately Roughened Stator Gas Annular Seals,” Proceedings of IJTC2006, 2006 International Joint Tribology Conference, San Antonio, Texas, October 23-26, IJTC 2006-12049.

Cloud, C.H., 2007, “Stability of rotors supported by tilting pad journal bearings.” PhD thesis, University of Virginia.

Cloud, C. H., Maslen, E. H., and Barrett, L. E., 2009, “Damping Ratio Estimation Techniques for Rotordynamic Stability Measurements,” ASME Journal of Engineering for Gas Turbines and Power, Vol. 131, January, pp. 012504.

Colby, G. M., Gupta, M. K., Nove, S. E., Griffin, T. R., Miller, H. F., and Sehlstedt, N. H., 2012, “High Pressure CO2 Compressor Testing for Tupi 1, Tupi 2, and Tupi 3,” ASME, Proceedings of ASME Turbo Expo 2012, Power for Land, Sea and Air, Copenhagen, Denmark, June 11-15, GT2012-70137.

Coletti, N. J. and Crane, M. E., Jr., 1981, “Centrifugal Compression on the Arun High Pressure Injection Project,” Proceedings of the IMechE Conference on Fluid Machinery for the Oil, Petrochemical and Related Industries, The Hague, Netherlands, pp. 63-70, March.

Ewins, D. J., 1984, Modal Testing: Theory and Practice, John Wiley and Sons.

Gelin, A., Pugnet, J-M., Bolusset, D, and Friez, P., 1996, “Experience in Full Load Testing Natural Gas Centrifugal Compressors for Rotordynamics Improvements,” IGT & A C&E, Birmingham, UK, 96-GT-378.

Gupta, M. K., Soulas, T. A., and Childs, D. W., 2007, “New Steps To Improve Rotordynamic Predictions Of Centrifugal Compressors,” ASME, Proceedings of GT2007, ASME Turbo Expo 2007: Power for Land, Sea and Air, Montreal, Canada, May 14-17, GT2007-27501.

Gupta, M. K., 2011, “Centrifugal Compressor Design Challenges for CO2 and Other Acid Gas Injection Services,” ASME Turbo Expo 2011, Vancouver, Canada, June 6-10, GT2011-46481.

Holt, C. G., and Childs, D. W., 2002, “Theory Versus Experiment for the Rotordynamic Impedances of Two Hole-Pattern-Stator Gas Annular Seals,” ASME Journal of Tribology, 124, pp. 137-143.

Hung, C.F. and Ko, W.J., 2002, “Identification of modal parameter from measured output data using vector backward autoregressive model,” Journal of Sound and Vibration, v. 256(2), pp. 249-270.

Jang, E., Muszynska, A., Park, Y.P., Kim, C.H., 1996, “Identification of the quadrature resonances using modal nonsynchronous perturbation testing a dynamic stiffness approach for an anisotropic rotor system with fluid interaction,” International Journal of Rotating Machinery, v. 2 n. 3, pp. 187-199.

Kanki, H., Fujii, H., Hizume, A., Ichimura, T., and Yamamoto, T., “Solving nonsynchronous vibration problems of large rotating machineries by exciting test in actual operating conditions,” Proceedings of the International Conference on Rotordynamics, IFToMM, 1986, pp. 221-225.

Kirk, R. G., 1985, “Evaluation of Aerodynamic Instability Mechanisms for Centrifugal Compressors,” ASME Paper 85-DET-147, Design Engineering Vibration Conference, Cincinnati, Ohio, Sept. 10-13.

Kirk, R. G., 1986, “Influence of Disk Leakage Path on Labyrinth Seal Inlet Swirl Ratio,” Proceedings of the Fourth Workshop on Rotor Dynamic Instability Problems in High Performance Turbomachinery, Texas A&M University, May, NASA Conference Publication 2443, pp. 225-236, June 2-4.

Kirk, R. G., 1990, “User’s Manual for the Program DYNPC28 -- A Program for the Analysis of Labyrinth Seals,” Negavib Research & Consulting Group, Virginia Tech, Blacksburg, VA, Jan.

Kleynhans, G. and Childs, D. W., 1997, “The Acoustic Influence of Cell Depth on the Rotordynamic Characteristics of Smooth-Rotor/Honeycomb Stator Annular Gas Seals,” ASME Trans. Journal of Engineering for Gas Turbines and Power, October 1997, Vol. 119, No. 4, pp. 949-957.

Kocur Jr., J.A. et. al., 2007, “Surveying Tilting Pad Journal Bearing and Gas Labyrinth Seal Coefficients and their effect on Rotor Stability,” 36th Turbomachinery Symposium.

Lund, J. W., 1974, “Stability and Damped Critical Speeds of a Flexible Rotor in Fluid-Film Bearings,” Trans. ASME, Journal of Engineering for Industry, pp. 509-517, May.

Maslen, E., 2000, Magnetic Bearings, University of Virginia.

Memmott, E. A., 1990, “Tilt Pad Seal and Damper Bearing Applications to High Speed and High Density Centrifugal Compressors,” IFToMM, Proceedings of the 3rd International Conference on Rotordynamics, Lyon, pp. 585-590, Sept. 10-12.

Memmott, E. A., 1992, “Stability of Centrifugal Compressors by Applications of Tilt Pad Seals, Damper Bearings, and Shunt Holes,” IMechE, 5th International Conference on Vibrations in Rotating Machinery, Bath, pp. 99-106, Sept. 7-10.

Memmott, E. A., 1994, “Stability of a High Pressure Centrifugal Compressor Through Application of Shunt Holes and a Honeycomb Labyrinth,” CMVA, 13th Machinery Dynamics Seminar, Toronto, pp. 211-233, Sept. 12-13.

Memmott, E. A., 2000a, “Empirical Estimation of a Load Related Cross-Coupled Stiffness and The Lateral Stability Of Centrifugal Compressors,” CMVA, Proceedings of the 18th Machinery Dynamics Seminar, Halifax, pp. 9-20, April 26-28.

Memmott, E. A., 2000b, “The Lateral Stability Analysis of a Large Centrifugal Compressor in Propane Service at an LNG Plant,” IMechE, Proceedings of the 7th International Conference on Vibrations in Rotating Machinery, Nottingham, England, pp. 187-198, September 12-14.

Memmott, E. A., 2003, “Usage of the Lund Rotordynamic Programs in the Analysis of Centrifugal Compressors,” Jorgen Lund Special Issue of the ASME Journal of Vibration and Acoustics, Vol. 125, October.

Memmott, E. A., 2010, “Application of Squeeze-film Dampers to Centrifugal Compressors,” CMVA, Proceedings of the 28th Machinery Dynamics Seminar, Universite de Laval, Ville Quebec, Canada, pp. 115-129, October 27-29, CMVA2010-15.

Memmott, E. A., 2011, “Stability of Centrifugal Compressors by Applications of Damper Seals,” ASME, Proceedings of ASME Turbo Expo 2011, Power for Land, Sea and Air, Vancouver, Canada, June 6-10, GT2011-45634.

Moore, J. J. and Hill, D. L., 2000, “Design of Swirl Brakes for High Pressure Centrifugal Compressors Using CFD Techniques,” 8th International Symposium on Transport Phenomena and Rotating Machinery, ISROMAC-8, Vol. II, Honolulu, Hawaii, pp. 1124-1132, March 26-30.

Moore, J. J., Walker, S. T., and Kuzdzal, M. J., 2002, “Rotordynamic Stability Measurement During Full-Load Full-Pressure Testing of a 6000 PSI Re-injection Centrifugal Compressor,” Proceedings of the Thirty First Turbomachinery Symposium, Turbomachinery Laboratory, Department of Mechanical Engineering, Texas A&M University, College Station, Texas, September.

Moore, J. J. and Soulas, T. A., 2003, “Damper Seal Comparison in a High Pressure Re-Injection Centrifugal Compressor During Full-Load, Full-Pressure Testing Using Direct Rotordynamic Stability Measurement,” ASME, Proceedings of the 19th Biennial Conference on Mechanical Vibration and Noise, International 2003 DETC, Chicago, Illinois, DETC2003/VIB-48458, September 2-6.

Nicholas, J. C., Gunter, E. J., and Allaire, P. E., 1979, “Stiffness and Damping Coefficients for the Five-Pad Tilting-Pad Bearing,” ASLE Transactions, Vol. 22, 2, pp. 113-124, April.

Nielsen, K. K. and Myllerup, C. M., 1998, “Parametric Study of the Flow in Swirl Brakes by Means of a 3D Navier-Stokes Solver,” 98-GT-328, IGTI, Stockholm, Sweden, June.

Nielsen, K. K., Myllerup, C. M., and Van den Braembussche, R. A., 1998, “Parametric Study of the Flow in Swirl Brakes by Means of a 3D Navier-Stokes Solver,” Transactions of the 3rd European Conference on Turbomachinery, pp. 489-498.

Mendes, R.U., 2011, Desenvolvimento de um Sistema de Atuação Magnética para Excitação de Sistemas Rotativos, Mechanical Engineering MSc thesis - Campinas State University.

Peeters, B., Van der Auweraer, H., Guillaume, P., Leuridan J., 2004, “The PolyMAX frequency-domain method: a new standard for modal parameter estimation,” Proceedings of Shock and Vibration 11, pp. 395–409, 2004.

Peeters B., Vecchio, A., Van der Auweraer, H. “PolyMAX modal parameter estimation from operational data.” Proceedings of the ISMA 2004 International Conference on Noise and Vibration Engineering, 2004.

El-Kafafy, M., Guillaume, P., Peeters, B., Marra, F., Coppotelli, G. “Advanced frequency-domain modal analysis for dealing with measurement noise and parameter uncertainty.” Proceedings of the 30th International Modal Analysis Conference, a Conference on Structural Dynamics, Volume 5, 2012.

Pettinato, B. C., Cloud, C. H., Campos, R. S., 2010, “Shop acceptance testing of compressor rotordynamic stability and theoretical correlation.” 36th Turbomachinery Symposium, pp. 31-42.

Ramesh K., 2002, “A State-of-the-art Rotor Dynamic Analysis Program,” The 9th International Symposium on Transport Phenomena and Dynamics of Rotating Machinery, Honolulu, Hawaii, February 10-14.

Schweitzer, G., Maslen, E.H., “Magnetic Bearings: Theory, Design, and application to rotating machinery,” Springer, 535p., 2009.

Shemeld, D. E., 1986, “A History of Development in Rotordynamics - A Manufacturer’s Perspective,” Rotordynamic Instability Problems in High Performance Turbomachinery, NASA Conference Publication 2443, Texas A&M University, pp. 1-18, June 2-4.

Smalley, A. J., Almstead, L. G., Lund, J. W. and Koch, E. S., 1974, “User’s Manual - MTI Cadense Program CAD-25 - Dynamic Stability of a Flexible Rotor,” Mechanical Technology Inc., February.

Soulas, T., and San Andrés, L., 2002, “A Bulk Flow Model for Off-Centered Honeycomb Gas Seals,” ASME Paper GT-2002-30286, Proc. of ASME Turbo Expo 2002, Amsterdam, The Netherlands.

Soulas, T., Desai, A. R., Colby, G., and Griffin, T., 2011, “CO2 Compression for Capture and Injection in Today’s Environmental World,” Proceedings of the First Middle East Turbomachinery Symposium, Doha, Qatar, February 13-16.

Wachel, J. C. and von Nimitz, W. W., 1981, “Ensuring the Reliability of Offshore Gas Compressor Systems,” Journal of Petroleum Technology, pp. 2252-2260, Nov.

acknowledgementsThe authors would like to thank the management of Petróleo Brasileiro S.A. and Dresser-Rand for their support and permission to co-author this paper. •

engineer’s notebook

29

insights

©Copyright 2015 Dresser-RandDResseR-RanD is a registered trademark of Dresser-Rand Company.

Editorial Statement: insights is a Dresser-Rand publication

that addresses topics of interest to its clients / readers in the oil, gas,

petrochemical, process industries, and general industrial markets (including

paper, steel, sugar, distributed power, and the U.s. navy). Comments, inquiries

and suggestions should be directed to:

Dresser-Rand Janet Ofano

insights editorial office 500 Paul Clark Drive

Olean, nY 14760 Tel: (Int’l +1) 716-375-3000

email: jofano@dresser-rand.com

www.dresser-rand.com