501F Users Group: Controls issues, mods dominate conference program

Paul Tegen, chairman, and the steering committee (see sidebar) of the 501F User’s Group deserve high grades for the stimulating program at their annual conference March 28-31 in San Francisco. Program was jam-packed with practical presentations delivered by users and sup­pliers alike. So busy was the schedule that some speakers were allowed only their 15 proverbial min­utes of fame at the podium.

The group’s next meet­ing will held at Disney’s Contemporary Resort in Orlando, January 16-19, 2006. For more information, contact Caren Genovese, meeting coordinator, at carengeno­vese@charter.net.

TXPFocus Group leads off

The first day got off to a good start with Mike Mag­nan’s presenta­tion on the work being done by members of 501F User’s Group in collaboration with engineers at Siemens Power Generation Inc, Orlando, to resolve issues with the TXP control system.

Change is rarely easy, and cer­tainly not when it involves switching a GT product line to another control platform. A brief history: Siemens Power Corp’s V series GTs sold into the US market in the mid 1990s came equipped with the company’s XMAT distributed control system, which preceded the more widely known Teleperm XP by a few years. Siemens bought Westinghouse Electric Corp as the 1990s were winding down, but not included in the sale was the company’s Process Control Div which supported the WDPF control system used in Westinghouse GTs. The new entity, Siemens Westinghouse Power Corp (SWPC), decided that Teleperm XP—later shortened to TXP—would be the only controls platform for future V and 501 machines.

This decision was made near the beginning of the GT boom. Architect/ engineers and EPC (engineer/pro­cure/construct) firms gen­erally had little experience with the TXP, making more difficult the job of interfac­ing GT and plant controls— this at a time when schedule and cost constraints were particularly demanding.

Steering committee
Bill Wimperis, director of engi­neering and construction, Con­stellation Generation Group.
Jim Beckett, director of engineer­ing, Calpine Operations, ERCOT region.
Raymond Martens, O&M man­ager, Klamath (Ore) Cogenera­tion Plant.
Gary Giddings, O&M superinten­dent, Seminole Electric Co-op’s Payne Creek station.
Gary Olivi, engineering and main­tenance team leader, Calpine’s Calgary Energy Centre.
Matt Kaleyta, CT manager, Dyn­egy Generation Fleet Operations
Paul Tegen and the other members of the 501F family expressed their appreciation to retiring steering committee member Bill Barras, for­mer gas turbine manager, Dynegy Generation, for his dedication and hard work over the years in support of the user group.

In addition, commission­ing teams rarely had some­one who understood TXP. The fact that logic diagrams were done in the European KKS sys­tem rather than in the ISA/SAMA system used in the US just made work more difficult. Training was needed, but owners often were reluc­tant to make the investment. The bottom line: Operations staffs were not adequately prepared to operate, maintain, and optimize logic or trou­bleshoot the TXP control system.

Fast forward to the 2004 501F User’s Conference. Users had become “pretty vocal” regarding TXP issues, according to Tegen, the chief CT engi­neer for Cogentrix Energy Inc, Char­lotte, NC, and an ad hoc group was formed to present to SWPC the con­cerns of owner/operators. Magnan, plant manager, Lower Mt Bethel Energy LLC, Bangor, Pa, provided the leadership to move the initiative forward. One of the users’ concerns was their difficulty in reviewing trip logic. This is critical to deciding where runbacks can be substituted for trips to facilitate operations and reduce wear and tear on equipment.

SWPC’s response was quick and positive: It con­ducted a thorough Internet-based survey to identify areas of high customer dissatisfac­tion with TXP, reviewed the results internally, and met with the 501F’s TXP Focus Group in December 2004, promising the following:

  • Improve TXP training to enable the development of “real skills.”
  • Upgrade TXP software perfor­mance.
  • Enhance graphics and trend­ing.
  • Clarify alarms/messaging.
  • Facilitate understanding of function blocks and the retrieval of historical data.
  • Create a standard preventive maintenance program and the tools to optimize control system reliability.

Goals of the software upgrade that are particularly important include the creation of a “rollback” feature to undo unwanted logic changes and installation of a “test mode” for test­ing logic before it is activated.

Progress in the training area: A mobile simulator is now available to support on-site training. Not hav­ing to send personnel offsite makes training more cost effective and sim­plifies staff management.

SWPC set up a TXP simulator at the meeting and held a half-day TXP User’s session. Hands-on use of the Internet-based simulator by partici­pants demonstrated features identi­fied for improvement and allowed users to ask questions directly to SWPC technical experts. This session also helped both users and SWPC to more clearly define the areas of dissatisfaction and to resolve some problems.

Magnan’s progress report at the 2005 501F Conference noted posi­tively SWPC’s efforts to resolve open issues. However, it also highlighted three important unknowns:

  • Timeline for improvements supported by the company.
  • Cost of improvements.
  • Future support for TXP.

Last point was raised because Siemens recently introduced the Turbine Control System 7 (TCS7) and TXP users are concerned that effort might adversely impact parts availability and dilute support for their system long-term. The first ret­rofit of a TCS7 was profiled as part of the W251 Users report published in the COMBINED CYCLE Journal, 2Q/2005, p 63 (available at www.psi­media.info/ccjarchives.htm).

To learn more about participat­ing in the TXP Focus Group, contact Tegen at paultegen@cogentrix.com.

Mitsubishi extends its capabilities

Dave Walsh, VP service and manufacturing, pre­sented an overview on the capabilities of Mitsubishi Power Systems (MPS), Orlando, with respect to the 501F. Mitsubishi Heavy Industries’ units in this model series, have logged more than 2.5 mil­lion operating hours since 1988, providing the knowl­edge necessary to develop solutions for fleet issues.

In 2004, MPS conducted an exten­sive test program with a W501FD-2 prototype, verifying the perfor­mance characteristics of MHI parts. The company now has all the parts available for the FC, Walsh told the user group, including those for the compressor. For the FD-2, the com­pany can supply R1-R4 vanes, R1-R3 blades, and transitions, he contin­ued. R4 blades, baskets, and nozzles will be manufactured domestically by 1Q/2006.

Walsh added that MHI has invest­ed $125 million in the Orlando Ser­vice Center to support the F fleet with repairs, inventory, remote mon­itoring, and field service. The compa­ny had completed 16 W501F outages before mid year.

R3 angel wing wear

Conference organizers did a good job of alternating user and vendor presentations on the first day. Last presentation before morning coffee on Day One was made by Reliant Resources’ Joe Schneider, an active participant in several user groups. Schneider, who addressed the third-stage angel-wing wear issue, is past chairman of the Combustion Turbine & Combined Cycle Users Organization (CTC2) and current program committee chair for the Combustion Turbine Operations Task Force (CTOTF).

As Schneider explained it, cobalt-based third-stage vanes like to creep down­stream over time; Siemens calls it plastic deformation. His first expe­rience with this time/temperature phenomenon involved Frame 7B second-stage nozzles. To the naked eye, Schneider said, the parts look fine; it’s just that the inner shrouds are very slowly migrating down­stream due to a cantilever effect. Keep in mind that time/temperature-dependent phenomena are of partic­ular concern today because of initia­tives aimed at extending component replacement intervals.

The OEM’s fix to mini­mize rubbing between the inner shroud and the angel wings of third-stage blades is to cut back part of the inner-trailing-edge vane shroud. Permanent fix is to use nick­el-based vanes, which have superior creep resistance.

The same issue is in evi­dence on the fourth turbine stage as well. However, the tempera­ture is lower and rubbing is not as severe.

Compressor washing

A good salesperson is able to compel a customer to listen to his or her presen­tation. Hugh Sales (yes, that’s his real family name) must be a good salesman. All he had to tell the 501F users that one of their units in base-load service gobbles up $12 million/month in gas at today’s prices and virtually everyone realized that what he had to say about main­taining gas-turbine (GT) efficiency with water washing was worth lis­tening to.

Sales, a VP at Houston-based Gas Turbine Efficiency (www.gtefficiecy.com) began with a tutorial on water washing and finished with a case his­tory success at a plant in the refinery corridor along the Gulf Coast that is equipped with two 501FBs in cogen­eration service. Of greatest inter­est to most subscribers, perhaps, is what Sales had to say about the process variables impacting wash­ing efficiency. This back­ground may prove valuable in formulating questions for prospective suppliers should you be involved in the pur­chase of a wash system.

The optimization vari­ables for a water wash sys­tem that you should be familiar with are these:

  • Water temperature. Water that’s too cold is con­ducive to blade temperature vari­ation, loss of cleaning efficacy, and the increased need for chemi­cals. Water that’s too hot leads to increased early vaporization and loss of fluid penetration. Plus, it presents a safety risk to opera­tors. The “just right” temperature, Sales said, was about 140F.
  • Water pressure is optimum within the range of 750 to 1400 psig. The designer’s goal is to accelerate water to approximately air-flow speed, not to increase blade impact. Water pressure within the range cited provides maximum dispersion across IGVs (inlet guide vanes) and R0 stage, maximum penetration into the compressor core, minimizes the waste of water and chemicals (if used), and ensures penetration of the air-flow boundary layer on blades and vanes.
  • Droplet size. Sales said this was the most critical variable in terms of cleaning penetra­tion and erosion. If droplet size is too small—less than 40 microns according to Sales—water never reaches the blade surface, it mere­ly follows air flow and no washing is accomplished. If it is too large—over 180 microns—the droplets may damage blades and inertial forces that throw water to the outer case may occur.
  • Water volume. The opti­mized water volume ranges from four to six minutes of water flow at approximate­ly 40 gpm. Actual number depends on the size of the tur­bine.
  • Nozzle design and place­ment. Spray pattern for a nozzle should be engineered after con­sidering the other variables being optimized. If true optimization is accomplished, Sales said, you only need one set of nozzles to accomplish both online and offline washes. Also, you will require far fewer nozzles (perhaps only a third as many) than others might say you need, he continued. Typi­cal retrofit applications, Sales added, normally call for replac­ing existing OEM off-line nozzles with those capable of both online and offline washing.

Three user presentations. Next, Dennis Winn of Pacific Klam­ath Energy and a colleague presented on their plant’s experience with out­let temperature control. They were followed on the program by Calpine Corp’s Nicole Prudencio, who updat­ed the group on natural-gas quality standards. Prudencio’s presentation was of particular importance given the varying composition of gas prod­ucts finding their way into transmis­sion pipelines these days. A sum­mary of Prudencio’s comments was presented in last issue’s Regulatory Update column (COMBINED CYCLE Journal, 2Q/2005, p 2; access at www.psimedia.info/ccjarchives.htm).

R1 turbine blade failures were discussed by engineers from Semi­nole Electric Co-op. Current think­ing is that the cause is high-cycle fatigue, origin unknown. Also that this is not a generic design fault; only two machines are known to have the problem.

Renegotiating LTSAs

No user-group meeting would be complete today without discussion of long-term service agree­ments (LTSAs). Richard E Thompson II, who is based in the New York City office of the international law firm Troutman Sanders LLP (www.troutmansanders.com), is well versed on the subject. He began his presen­tation with a brief historical overview.

The majority of LTSAs were signed during the bubble of the late 1990s, he said, a sellers’ market. Such contracts were relatively new at the time and there were wrinkles to iron out. Over the last several years, many things have changed. Now we’re in a buyers’ market, Thompson con­tinued, or moving into one, depend­ing on your perspective. Owners have learned lessons from disputes that have arisen regarding the intent of the agreements, as well as from situa­tions not anticipated earlier—such as changes from hours-based to starts-based maintenance and vice versa.

Thompson addressed the follow­ing three important themes and con­siderations during his presentation:

1. More expansive OEM cov­erage. Owners are beginning to find OEMs offering more expansive risk coverage, he said, via so-called “term warranties” or “full cov­ered maintenance.” This approach can eliminate arguments as to the types of defects, as well as the need to file claims before warranty expiration, by simply stating that the OEM will repair or replace any defective part as may come to light at any time during the whole contract term. To stay competitive, Thompson foresees that non-OEMs will have to enter this territory, too. Net effect is less risk for the owner, more risk for the contractor. The key is determining whether the “price is right” to place more risk on the con­tractor.

But, Thompson cautioned, be careful: OEM contracts can set forth vague “conditions” of the coverage—like “proper” storage, operation, and maintenance. “What is proper?” he asked the users. Also, language can be unclear as to what happens if a condition is violated. Is the entire warranty voided? The solution here is precise drafting that appropriately limits the impact of any violations of such a condition.

2. Changes in maintenance profile is a well-known issue today. Plants thought to be hours-based going starts-based, and vice versa. Thompson said users should be ask­ing: “How does my LTSA cover this? How should it?” Cash-flow questions arise, too, including these:

  • Does more maintenance sooner mean more money is paid to the OEM sooner?
  • Does elimination of a combus­tor inspection (go directly to a hot-gas-path inspection) mean the owner doesn’t pay for the CI?

Possible approaches include the following:

  • Yogi Berra approach: It ain’t over ‘til it’s over. Make a chart of parts and repairs. Deal is not complete until the owner gets all that it is paying for and the con­tractor gets paid for all that it has provided.
  • Fix it now/fix it later. Do a change order each time you have a change in operating mode, to address specific issues at the time.
  • Merchant pricing approach: Pricing is based on EBH/ES ratio, so pricing tracks the actual wear and tear. Recall that EBH is equivalent base hours, ES is equivalent starts.

Each of these approaches has advantages and disadvantages, depending on the nature of your proj­ect and business goals. It is impor­tant to analyze them carefully before making a decision.

3. Negotiating leverage con­cepts. Owners may be in a position to renegotiate their contracts, now or in the near future, Thompson sug­gested. Perhaps a contractual exit ramp is around the bend. Perhaps the OEM itself has amendments it would like to make. Perhaps the OEM wants to secure future tur­bine business. Key point: Owners should assess their renegotiating leverage and analyze the opportu­nities to improve their contractual positions—including the elimination of pitfalls—through renegotiation. Think ahead strategically; the payoff can be substantial.

Suggested for further reading is “Understanding contractual ‘fine print’ can be the difference between financial success, failure,” COM­BINED CYCLE Journal, Fall 2003, 27; access via www.psimedia.info/ccjarchives.htm.

Siemens Day

Siemens Power Generation Inc (SPG), Orlando, developed the pro­gram for the second day of the three-day 501F User’s Conference. The content topics: Review of fleet-wide performance, SPG’s response to ques­tions submitted by users prior to the meeting, status of the OEM’s prod­uct improvement program developed in response to user needs, steam-turbine developments, mods and upgrades, and controls update.

Defining terms
Important to communication is an understanding of technical and business terms. Duty-cycle defini­tions used by Siemens Power Gen­eration are:

  • Base load. Greater than 75% service factor (approximately 6500 hr/yr) and operation for more than 125 hr/start. Service factor is percentage of time over a given period—typically one year—that a unit is connected electrically to the grid.
  • Intermediate duty. Greater than 30% service factor (approx­imately 2500 hr/yr) and operation for more than 12.5 hr/start—and less than base-load duty.
  • Peaking. Greater than 5% service factor (approximately 400 hr/yr) and operation for more than 4 hr/start—and less than intermediate duty.
  • Standby. Less than peaking.

Availability and reliability terms, as defined by ANSI/IEEE Standard 762 are:

  • Availability is the percentage of time that a plant is capable of providing service, whether or not it actually is in service and regardless of the capacity level that can be provided.
  • Reliability is the percentage of time a unit is not in a forced-outage condition.
  • Starting reliability is the per­centage of attempted starts that successfully synchronize with the grid.

The SPG presentations were upbeat: Customer satisfaction is improving steadily, said Ron Bauer, director of quality process. His key message was that the company is listening to its customers and ben­efits were accruing to both. Glenn Sancewich, program manager for STG6-5000F availability and reli­ability, who followed Bauer, began his presentation with an analysis of duty cycles (see sidebar for definition of terms) that showed a relatively even split fleet-wide in 2004 of units operating base-load, intermediate duty, peaking, and standby. Make note that Siemens recently changed its W501F designation to SGT6-5000F. Also, that the availability and reliability information reported by Siemens was based on data reported to it by responding users.

Reliability of the GT package con­tinued to improve, he said, finish­ing the year at nearly 99% fleet-wide (three-quarters of the fleet exceeded 99%); on-peak reliability exceeded 99%. Availability on-peak (peak-period average from June 1 through October 1 and from Janu­ary 1 through March 1) also has been improving steadily, closing out 2004 at over 97%. Sancewich reported continual improvement by custom­ers in conducting their outages as one reason for higher availability. Better planning has contributed to reduced outage duration: A double-digit percentage decrease in the time required for scheduled outages over the last three years, and a whopping 35% decrease in forced-outage hours during the same period.

Starting reliability also is improving, primarily because of upgrades that have improved the performance of the ignition system, sensors, control logic, igniters, etc. It is now close to 90%. When avail­able upgrades are installed through­out the fleet, which may be possible by the end of this year, Sancewich expects the average starting reliabil­ity to exceed 95%. Further improve­ments in development could push the number to 99% or higher when they are implemented. The time frame for this is planned at two to three years.

Customer questions answered. Phil Karwowski, director of GT opera­tions and service engineering, and Jon Kemmerling handled SPG’s technical response to customer requests, while Vijay Kapoor revealed implemen­tation plans for product upgrades in the final stag­es of develop­ment. Karwowski and Kemmerling took attendees through well over 100 slides in three hours of compre­hensive technical presentation.

They discussed advanced com­ponent designs used in the SGT6-6000G (formerly W501G) that could be adopted for the F, such as that for the inlet splitter plate modification. Also, ways to reduce outage time by extending the interval time for over­haul of key components—the thrust bearing, for example. The new main­tenance-interval recommendations for thrust bearings are described in “Optimizing restoration of thrust clearances,” COMBINED CYCLE Journal, 2Q/2005, Design/Operating Ideas section, p 87; access at www.psimedia.info/ccjarchives.htm.

A detailed report on the testing of segmented vane packs (SVPs) for rows 1 and 2 compressor dia­phragms designed to reduce wear was of considerable interest. Four units equipped with SVPs each have accumulated more than 4500 equiva­lent base hours of operation and inspection indicates a reduction in wear on the outer shroud.

Another discussion point: The benefits of installing mods to facil­itate borescope inspection. This enables early problem identification so many extraordinary items can be incorporated into the outage work­load without extending schedule.

Performance enhancements, the speakers said, should be considered during major inspections. These include installation of honeycomb seals and abradable coatings to reduce clearances and efficiency loss associated with leakage between rotating and stationary components. Result is a recovery of losses attrib­uted to normal degradation.

Icing was yet another topic cov­ered in-depth. Focus was on the damage that icing can cause and how to avoid it. An icing risk chart was presented. It plots ambient-air temperature and relative humid­ity and the position of inlet guide vanes (IGVs). Users who close IGVs to up to 45 deg to reduce part-load emissions of carbon monoxide were reminded to pay close attention to ambient conditions.

Keep in mind that as you close the IGVs, pressure drop increases and ice formation occurs at lower temper­ature and humidity than when the vanes are open more. To illustrate: At 40F and IGVs set at 45 deg, ice can form when the relative humidity is only about 65%; at 40F and 35 deg the relative humidity would have to be about 80% to get icing.

Photos shown of ice damage cre­ated during tests at a NASA lab caught the attention of virtually everyone. Protection is afforded by installation of icing probes and inlet air heaters. Speakers recommended a call to SWPC when contemplating a change in IGV position—just to be on the safe side.

An update was presented on the DLN (dry low-NOx) pilot (refer to “501D5/D5A Users focus on GT upgrades,” COMBINED CYCLE Journal, Summer 2004, p 82). All test-bed and validation tests have been completed, the speakers report­ed, and field testing is underway to verify that the design is not excited by engine operation.

Other ideas. Tuned resonators and thick thermal barrier coatings for combustor baskets were pre­sented as methods for helping to minimize potential damage from high-frequency dynamics. Chem­ical stripping was suggested for overhauling DLN transition piec­es where wall thinning is of con­cern with grit-blast stripping. Addi­tional cooling holes added to the inner-diameter shroud of R1 vanes when repairs are necessary can help reduce oxidation.

SPG’s Margie Croteau put into perspective for users many of the technology advances discussed by Karwowski, Kemmerling, and Kapoor. Her presentation focused on the specific product improvements that evolved from the rigorous pro­gram of engineering and testing—improvements customers identified as important to them and now avail­able commercially.

Croteau’s presentation divided the product enhancements into the following groups:

  • Power and efficiency improve­ments—including such things as compressor performance upgrade, hot restart option, wet compres­sion, etc.
  • GT emissions and alternative fuels—including low CO turn­down, ultra low NOx package, integrated gasification/combined-cycle syngas firing capability, etc.
  • Plant operating flexibility improvements.
  • Service interval extension.

Each enhancement was profiled with a summary of the package, potential benefits, commercial avail­ability (now or when in the future), status of first application (complete or when completion is expected), and estimated lead-time for imple­mentation.

More on controls

The TXP control system discussed earlier raises certain issues with users, the WDPF, supplied with GTs built by Westinghouse Electric Corp before the company’s sale to Germany’s Siemens AG, raises oth­ers. Of particular concern to 501F owner/operators is WDPF capacity restrictions that can hamper system O&M by making it impossible to add or modify DCS control logic, alarm­ing, or monitoring functions.

According to Mitch Cochran (mitchpcs@aol.com, 601-271-8199) of Process Control Solutions LLC, Hattiesburg, Miss, lack of available distributed processing unit (DPU) database space and data highway system IDs (SIDs) are the most com­mon capacity problems encountered with a WDPF.

Total database memory capacity is limited to 120 kB for seven- and eight-level DPUs, 80 kB for six-level DPUs. Also, each DPU is limited to 2046 total points—including period­ic and non-periodic points, received or originated (local points are not included in the limitation).

WDPFs with extended SID capa­bility are restricted to 16,000 periodicpoints (broadcast frequency of 0.1 and 1.0 seconds) and 16,000 non-periodic points. Systems without extended SID capability can accommo­date 16,000 periodic points only. For small systems—those with fewer than 10 DPUs, “SIDs maxed out” is not a problem, but for sys­tems with 15 or more DPUs this restriction often applies.

When the WDPFs went into ser­vice it’s doubtful that anyone could have anticipated capacity issues. But new logic added over the years in support of system/equipment modi­fications, new alarms, etc, have used up the readily available capacity at many plants much the same way that the empty cabinets in a new home fill up over the years.

Retrofit of dry low-NOx combus­tion systems and addition of combus­tion dynamics monitoring systems have created particularly demanding storage needs.

Cochran, who has a wealth of con­trol-system experience, offered some solutions to the space crunch and then presented a couple of case his­tories to support his ideas. Perhaps the most obvious solution is what he calls the DPU Mem-Free procedure. It’s somewhat analogous to throw­ing out items no longer used from a full cabinet, or relo­cating them to the garage or attic, to create space for new. Mem-Free identifies unused points and obsolete logic for deletion, and alarm points that can be relocated to a DPU with available space. It does not require a total plant outage; in fact, most of the work can be done offsite.

Another solution is to add a DPU processor without cabinet and Q-crates. This increases database space by 120 kB to which alarm points from other DPUs can be relocated. Taking this solution one step further, add a DPU with cabinet and Q-crate, mak­ing it possible to relocate I/O and control points from existing DPUs in addition to alarms.

Rebuilding the system point direc­tory can alleviate the “SIDs maxed out” restriction. This involves delet­ing unused points and obsolete logic and redistributing periodic and non-periodic points. However, a total plant outage of up to two days is required to load the modified DPU source code and rebuild the system point directory.

Finally, you can opt for a sys­tem upgrade from WDPF to Ova­tion (a product of Emerson Process Management, www.ovationforpower.com/power) which will permanently eliminate DPU database capacity restrictions, Cochran said. Ovation controllers have about eight times the capacity of a WDPF DPU proces­sor. This option would be most cost-effective for a plant expansion.


No user-group meeting is complete without a generator presentation. At the 501F meeting, that was done by Tony McConnon, an electrical design engineer for National Elec­tric Coil (www.national-electric-coil.com), Columbus, Ohio. In his pre­pared remarks, McConnon focused on preventive maintenance, inspec­tion, and testing. Much of what he had to say was covered by colleague Bill Moore at the Fall 2004 meeting of the Combustion Turbine Opera­tions Task Force (refer to “CTOTF tackles user issues on a wide range of aero, frame machines,” COMBINED CYCLE Journal, Fall 2004, p 52; access at www.psimedia.info/ccjar­chives.htm). Add to that the mate­rial in “Inspecting, maintaining air-cooled generators,” on p OH-52 of the 2006 OUTAGE HANDBOOK supple­ment to this issue, and you have a good overview of the subject. ccj