Changing the way the Power Generation Industry buys parts.

Western Turbine Users (WTUI) 2006 recap

The nuts’n bolts of LM engine O&M, generator maintenance, turbine blade repairs, SCRs for peakers, duct-burner igniters headline annual meeting

The Western Turbine Users must be doing something right—many things well may be more correct—because year after year the group attracts between 600 and 700 attendees to its annual conference and exhibition. That’s about double the registration of any other user group serving the gas-turbine-based sector of the generation business.

Recall that the Western Turbine Users Inc (WTUI), headquartered in Long Beach, Calif, and led by President Jim Hinrichs of LS Power, is a membership organization of LM2500, LM5000, and LM6000 aeroderivative gas turbine (GT) users, third-party parts suppliers, and service providers.The group’s mission is to provide a forum for the exchange of technical, operations, and maintenance information and experience, with the goal of improving the reliability and profitability of generating facilities—simple cycle, cogeneration, and combined cycle—using these types of engines.

The editors of the COMBINED CYCLE Journal (CCJ) came away from the 2006 meeting (Wyndham Hotel, Palm Springs, Calif, March 12-15) with the impression that there wasn’t a question on LM engine O&M that couldn’t be answered by one or more of the following in attendance:

• Representatives of the three depots—TransCanada Turbines (TCT), Calgary; MTU Maintenance Berlin-Brandenburg GmbH, Ludwigsfelde, Germany; and Air New Zealand Engineering Services (ANZ), Auckland—licensed by GE Energy, Atlanta, to provide parts, repairs, and maintenance services on LM series GTs. These companies collectively provide the lion’s share of the prepared technical material presented at WTUI meetings in a collaborative manner without commercials.
• About 300 experienced colleagues representing owner/operators.
• Another 300 or so attendees representing about 100 exhibitors— including third-party suppliers of engine parts, manufacturers of consumables, service providers, consultants, etc.

Owner/operators of LM engines who have never attended a WTUI conference probably cannot imagine the value associated with participation. It is the rare attendee that returns to his or her plant without an idea for saving at least thousands of dollars in maintenance and/or operations.

This meeting is a challenge for an editor to cover because so much is going on all the time. For example, the core of the conference was a series of five breakout sessions totaling more than eight hours for each engine type—two were open to all conference attendees, three were user-only. The breakouts for each engine are conducted in parallel.

Likewise, many of the prepared technical presentations in the afternoon of the second day were conducted in separate rooms in the same time slot. Special presentations were made on generator partial discharge testing, generator maintenance, generator stator repair, cooling-tower O&M, duct-burner igniters, hightemperature SCR catalyst performance, and turbine blade repairs.

Thus the coverage that follows, while not complete, hits the highlights of discussion topics for each of the LM engines. And it provides an overview of several prepared presentations. Sections on the LM breakouts are based on notes taken during the open discussion sessions (not the user-only portion of the meeting) by President/CEO Sal Della Villa and his colleagues at Strategic Power Systems Inc, Charlotte, NC (www. spsinc.com). SPS, a data collection and analysis consultancy, compiles statistics on engine performance for WTUI and others, and volunteers as the “recording secretary” for the group.

Comprehensive proceedings. Users responsible for multiple engine types are challenged like the editors by wanting to be in two or three places at the same time, but the WTUI officers and board of directors have solved that problem by publishing a proceedings—certainly the most comprehensive package offered by any GT user group.

Each user attendee receives a CD with the detailed presentations on maintenance procedures and repair methods—including excellent photography— prepared by the depots for the breakout sessions. Those, together with the actual session minutes compiled by SPS, are sufficient to satisfy the information needs of virtually every attendee. Slides used in the special technical presentations on the afternoon of the second day are posted at www.wtui.com and available only to members.

For information on membership in the WTUI, contact Conference Director Gae Dow (gdow@cox.net, 619-460- 8314. Current officers—in addition to Hinrichs—and directors are:

• VP, Mike Raaker, Raaker Services LLC.
• VP, Jack Dow, Primary Energy Ventures LLC.
• Treasurer, Wayne Kawamoto, Power Services.
• Jon Kimble, Fresno Cogeneration Partners.
• Mark Breen, Calpine Corp.
• Charlie Hoock, Wood Group Power Operations.
• Chuck Casey, City of Riverside.
• Jimmie Wooten, DPS-Juniper LLC.
• Jim Amarel, Energy Services Inc.

LM2500 breakout

Session chair and discussion facilitator for the LM2500 sessions was Chris Kimmich, operations engineer for Nevada Cogeneration Associates #2, Las Vegas; recording secretaries, Della Villa and Debbie Flamino of SPS. The two open sessions were conducted on the first day. ANZ’s John P Leedom made the opening presentation for the depots and guided the discussion of damage assessment/repair procedures; he was supported by Kevin McGregor of TCT and Christian Czmok of MTU.

By way of background, the LM2500 was derived from the CF6 family of aircraft engines used on widebody passenger jets. It has a hot-end drive, two-shaft gas generator with free power turbine, and a thermal efficiency that ranges from 36% to 41%, depending on the model and enhancements.The more than 2000 landbased and marine units installed worldwide have accumulated more than 40-million operating hours.

The basic 60-Hz engine is rated 23 MW at ISO conditions and has no gearbox. It has 16 compressor stages, two high-pressure turbine (HPT) stages, and six low-pressure turbine (LPT) stages. Power enhancement by way of steam injection (so-called STIG) is an option. The LM2500+— suitable for power generation, marine, and mechanical- drive applications—can develop more than 30 MW at ISO. Power turbine comes in two- and six-stage configurations; compressor has 17 stages. NOx and CO releases each are 25 ppm when the machine is equipped with the OEM’s dry, low-emissions burner and fires natural gas.

Leedom started at the compressor air inlet and worked his way back through the machine identifying, and discussing in sufficient detail for attendees, the aggregated findings of the depots in 2005. The bullet points that follow indicate the depth of coverage in WTUI’s equipment clinics and why they should be on the “must attend” list of every LM2500 user.

• Compressor inlet air seal and No. 3 oil seal: Oil seepage caused by the worn seal liner reported by some users can be corrected by implementing the solution suggested in service bulletin (SB) LM2500-IND-085. Note that the latest IPB part number supersedes SB-IND-085.
• No. 3 bearing housing (damper configuration): Low oil pressure during startup causes metal-tometal contact. OEM has been asked to research design documentation concerning air-flow limits for both damper and nondamper configurations.
• Transfer gearbox wear: To check for wear, try to move the gearbox. Any movement is proof-positive. Wear sometimes causes high vibration.
• Duplex bearing failure: None have been reported in the last two years. Solution in SB-IND- 160 appears to have corrected the problem.
• Horizontal gear-shaft spline wear: SB-IND-160 introduced a rugged nozzle to solve this problem. Replacement must be done in a depot shop.
• Transfer gearbox clevis mount wear: Severe wear found on both bolt holes. Avoid condition with periodic inspection of bolt tightness.
• High-pressure compressor (HPC) stage 1: Engines operating at low power will cause wear of Carboloy pads on mid-span dampers. Problem not identified with land-based machines. Navy replaces wear pads every 12,000 hours (refer to SB-IND-182).
• Bolts for trim balancing: Depots have removed bolts, nuts, and shims and implemented a procedure that allows balancing without weights.
• HPC rotor air duct: Vespel® strip is missing or has come loose. Replace at next depot visit; immediate attention not necessary.
• HPC blade failures examined: Depots not sure of operating hours associated with airfoil fractures but periodic borescoping with high-resolution equipment is recommended to accurately assess the condition of blades.
• HPC Stage 15 locking blade: Solution to locking-blade platform cracking is the improved blade referenced in SB-IND-174.
• HPC rotor, excessive rub coat wear: Check rotor color when conducting a borescope examination. If you see green, coating is missing and should be replaced during the next overhaul to improve efficiency.
• Variable-stator-vane (VSV) lever arms bent, damaged, or loose: Use go/no-go gauge to measure the position of lever arms. Attendees identified one reason for the damage: parts stepped on during maintenance.
• VSV bushings and washers worn: Depots strongly recommended that users clean the package periodically to help reduce wear. One user said dry ice was used at his plant for this purpose, with the ice particles sized for maximum cleaning effectiveness. Process is used on aircraft engines.
• Compressor rear frame (CRF) No. 4 bearing failure: Minimize the probability of failure with periodic inspection (once a quarter or every other quarter) and by following the guidelines in SB-IND-98 and SB-IND-105. One user with several LM2500s said their plants have not experienced such a failure in more than a decade of base-load service.
• CRF seal support issues: Depots recommended users follow guidelines in SB-168 for all engines regardless of serial number.
• Fuel-nozzle wear: This reflects a quality-control issue identified with the OEM. One suggestion offered to reduce wear was to replace the supplied swirl cup with one of lighter material. Users said they would rather replace the fuel nozzle.
• HPT Stage 1 nozzle-guide-vane deterioration: Condition typically caused by the spalling-off of the thermal barrier coating (TBC) on the dry, low-emissions (DLE) combustor.
• HPT Stage 1 blade cracking: Cracking at the leading edge of blades near the base was reported by some users with engines operating in marine environments.
• HPT Stage 2 nozzle-guide-vane erosion: Attendees recommended a platinum aluminide coating to protect against erosion conducive to metal liberation that could damage Stage 2 blades.
• LPT tip interlock wear: Such wear leads to shingling, which is common in marine units, according to the navy.

At the end of the open sessions the depots stressed the need for “good” information from the plants to facilitate engine overhaul and repair. Specifically, the depot requires hours and starts since the last hot section, vibration levels, a summary of inspection finding, number of full/partial cycles and trips, etc.

Kaua’i Co-op’s super STIG experience

A big benefit of participating in user-group meetings is that you get to learn first-hand from the experiences of your colleagues. User presentations almost always offer a few good ideas that can be implemented at your plant. Conference attendees were treated to a presentation on the operational flexibility of the super STIG LM2500 by Brad Rockwell, production manager for Hawaii’s Kaua’i Island Utility Co-op.

Some background on the Kaua’i utility and its Kapaia Power Station (Fig 1) is helpful before getting into details on the power block. The importance to the power company of the single LM2500-PH STIG 70 GT installed at Kapaia cannot be overstated. The GT, rated 27.5 MW with steam injection, is the largest unit on an island with an all-time peak load of 77 MW. Normally, the GT supplies between one-third and two-thirds of Kaua’i’s power via a 48.9-MVA generator manufactured by Brush Electrical Machines Ltd, Houston.

The black-start-capable machine, which began commercial operation in September 2002, burns liquid fuels—naphtha is the primary energy source, distillate oil the backup. Natural gas is not available. Operation is 24/7 with limited shutdowns for maintenance. Lifetime capacity factor is 90%; availability just south of 96%.

STIG operation is enabled by the availability of quality surface water from nearby Kapaia reservoir, which previously served sugar-cane plantations. Water received by the plant is clarified and chlorinated before being pumped to the 340,000-gal service water tank, which reserves 170,000 gal for plant fire-fighting purposes. From there, an activated carbon filter removes residual chlorine before the makeup water is sent to reverse-osmosis modules and the downstream electrodeionization (EDI) unit—or so-called E-Cell. Note that the E-Cell, developed by Glegg Water Conditioning Inc, is now offered by GE Water & Process Technologies, Trevose, Pa. A mixed-bed demineralizer is installed at the plant as a backup treatment system.

Product water has a conductivity of 0.1 micromho/cm and a maximum silica content of 5 ppb, making it suitable for use in the once-through heat-recovery steam generator (OTSG) connected to the GT exhaust duct. The OTSG, provided by Innovative Steam Technologies, Cambridge, Ont, Canada, is not equipped with duct burners. Steam produced has no value other than for NOx control and power augmentation.

If you read the GE literature, it says the STIG option for the LM2500-PH allows the injection of up to 50,000 lb/hr of steam for power augmentation. The STIG 70 designed especially for Kaua’i allows injection of up to 70,000 lb/hr—the highest amount ever.

Rockwell explained the reasons for purchasing the STIG 70 to the editors of the COMBINED CYCLE Journal after his presentation. The project, originally developed as an IPP, had to combine high efficiency and low emissions at the correct size to meet the local utility’s needs.

A good balance among emissions control, efficiency, and power production was struck with the STIG 70, configured to provide about 10,000 lb/hr of steam for NOx reduction and the remainder for power augmentation. The plant’s air permit limits NOx emissions to 15 ppmc (“c” is for corrected). At full load, and with 10,000 lb/hr of nozzle steam, the engine produces about 75 ppm NOx on liquid fuel under normal operating conditions.

Further NOx reduction to levels required by the air permit is achieved by an SCR (selective catalytic reduction) using a urea-to-ammonia system to produce reagent on demand.Rockwell noted that the emissions limit could have been achieved without an SCR by using 21,000 lb/hr of nozzle steam for NOx control, which has been demonstrated onsite. However, the heat-rate penalty of this approach was too high.

When the unit is running at full load with 60,000 lb/hr of 465 psig/ 900F power-augmentation steam, the gross heat rate is 8150 Btu/kWh based on the higher heating value of naphtha, he continued. This makes the Kapaia unit the most efficient fossil-fueled generator on the island. If nozzle steam were used to meet the 15-ppmc NOx limit without benefit of the SCR (thereby reducing poweraugmentation steam to 49,000 lb/hr), the gross heat rate jumps to about 8500. However, the engine can still produce 27.5 MW.

The benefits of steam injection extend beyond emissions control, power augmentation, and higher efficiency. Rockwell said that the high amount of steam injection keeps critical hot-section parts operating at 80 to 100 deg F below the maximum firing temperature. This contributes to an extended maintenance cycle.

At the time of the WTUI presentation, the Kapaia GT had accumulated about 30,000 fired hours.Rockwell reported that the unit had completed one hot section to that point, at 20,000 actual hours. OEM guidelines for liquid fuel suggest a hot section at 12,500 hours. Considering the condition of the engine, he thought the unit could operate for 25,000 hours between hot sections.Rockwell added that the Stage 2 nozzle was the limiting factor relative to the time between hot sections, so the utility opted for a third-party solution to extend the interval time for servicing those parts.

LM5000 breakout

Session chair and discussion facilitator for the LM5000 sessions was Jimmie Wooten of DPS Juniper LLC, Bakersfield, Calif; recording secretary, Torgeir Rui of SPS. Users attending the two open sessions on the first day of the conference numbered fewer than 30—the smallest attendance of the three breakout sessions. Reason: The LM5000 fleet is dwindling because GE no longer builds the engine. ANZ’s John Callesen and MTU’s Peter Kilian teamed up for the depot presentation.

Of note is that GE offers an upgrade program for replacing the LM5000 with a new LM6000-PC engine (hot-end drive) within the existing package. Compared to the LM5000, the LM6000 is said to offer improved reliability and efficiency, with similar power output and air flow.

Callesen kicked off the program with a review of the low-pressure compressor (LPC) Stage 0 blade failure discussed last year without conclusion. Investigation revealed crack initiation at a corrosion pit. Unit was operating base-load on a North Sea platform. User reportedly had taken no steps to prevent chloride corrosion and metal attack was severe.OEM investigators considered it an unusual and isolated case.

Discussion on the LPC continued with the No. 1 bearing air/oil seal. Then new-hardware solution covered in SB-IND-179 has resolved previous problems and users in attendance had no new issues to report. Likewise, wear problems with the Stage 3 inner shroud and bushing have been fixed by following the recommendations of SB-IND-195 and -196. Wooten displayed pictures of shrouds and bushings with three years of service that showed no wear.

The front frame assembly(FFA) was covered next. Cracking identified previously is still occurring and the issue remains unresolved. Suggestion was to inspect as frequently and completely as the operating regimen allows because such cracking is not identified with any specific location. Experts also recommended that users keep a sharp eye on cracks identified and carefully monitor their propagation. Be aware that if any material is released into the HPC, expensive repairs likely will be necessary.

Cracks should not be welded onsite, the depots said, because the risk of heat distortion accompanied by new cracking is too great. Best results for corrective welding are achieved when repairs are made in the controlled environment of a qualified shop.

Experience indicates that the best way for preventing FFA cracking is to maintain LPC vibration as low as possible—lower, in fact, than the OEM’s recommended limits. Depots reported finding broken seals on units operating within the specified vibration range. An alignment check was suggested as a first step for any engine with significant vibration.

A GE representative participating in the discussion was asked if the OEM was seeing the same cracking problem on the LM6000 as it was for the LM5000. Answer was that there has been some cracking, but not nearly as much as on the LM5000. As to what the OEM was doing regarding corrective action, the reply was nothing at present.

Discussion moved on to the inlet-gearbox spline wear problem. The harder bearing and revised layout recommended by SB-IND-175 was said to have been effective where implemented with no new problems identified.

Next, HPC rotor dovetail failures were discussed and off-schedule VSV operation was identified as the cause. Risk of blade failure can be reduced dramatically, experts said, by operating the vanes near the middle of the range rather than by using the entire range. The importance of checking bushings and lever arms at each inspection was stressed.

Five final points regarding the compressor section: First, performance degradation can be reduced by frequent water washing of blades. Second, failure of a blade in Stage 11 of one HPC was reported by the OEM. Root cause is unknown at present. Third, damage to the Vespel® strip on the air duct is addressed in SB-IND-210. Good results from the new coating were reported.

Fourth, service letter 5000-06-01 issued recently recommends replacement of all oil tubes and their support brackets during a compressor mid frame (CMF) overhaul—this to avoid tube cracking in front of the CRF mid flange. Recall that these are internal pipes and not available for frequent inspection. Fifth, SB-IND-203 issued a few weeks before the WTUI meeting, suggests adding bridging brackets to the CRF mid flange. Cracks have been found on the flange and stiffeners; they represent a risk for substantial damage.

Combustors. No root cause was offered for an issue characterized by a burned combustor dome. It appears most often on oil-fired units, but sometimes is found on gas-fired GTs as well, after about 10,000 hours of operation. Fuel quality and water injection are believed to be the biggest contributors to the condition. Depots also reported that two engines with the new combustor trumpet design show no sign of burning, based information collected during a 12,000-hr borescope inspection.

HPT. Depots reported large variations in the erosion rates of Stage 1 nozzles; a correlation exists between the T44 firing temperature and erosion. Average time between hot sections is from 20,000 to 26,000 hours when firing temperature is at its limit of 1420F to 1440F. One unit with a firing temperature of about 1300F ran for 60,000 hours with the original hot section. The correlation between the thermal barrier coating and erosion also was part of the discussion.

Users reported shroud seals blowing out and creating gaps in the Stage 2 nozzle assembly. GE says there’s not much it can do about it. Air-filter screen cracking is not considered critical and trying to fix it onsite probably would cause additional problems. Filter screen cracking should not be cause for early removal of the component. Note that cracking also occurs on the trailing edge of the nozzles, and cracks up to an inch in length are allowed.

Inspection of the rotor thermal shield was emphasized in a recent service letter because cracking of the component has been detected in some units. Concern on the part of GE is that if the heat shield were to fail it could initiate a disk separation failure.

Onsite repair of cracks in the outer case of the turbine mid frame (TMF) is not recommended. The depots said there is a high risk of permanently damaging the TMF if work is done at the plant. At a minimum, repairs at the depot will be more complicated; in the extreme, the casing may have to be scrapped. Note that welding on the case without purge air from the back side can cause additional cracking.

LPT. The depot report card indicates that the low-pressure turbine is quite reliable. Occasionally, vane cracks are experienced, but these usually occur on cycling units with steam injection.Some cracks are acceptable, others require repair. Shroud wear often is reported on high-hour units. The engine tolerates this condition but there is a performance penalty associated with it. Final discussion focused on the rotor disk in one unit that exhibited cracks between adjacent rim bolt holes. Condition was caused by the OEM’s hole-drilling process, which as been changed. A service bulletin issued after the meeting suggested an inspection to see if the condition is present.

LM6000 breakout

Session chair and discussion facilitator for the LM6000 sessions was Jim Amarel of Energy Services Inc, Farmington, Conn; recording secretary, Kevin Licata of SPS. Dale Goehring of TCT led the depot presentation with support from Christian Poeppel of MTU. The LM6000 track commanded the attention of the majority of conference participants, with about 150 delegates attending the two open sessions on the first day of the meeting.

This engine is extremely popular, with more than 600 units in electric generation service worldwide. It is based on the CF6- 80C2, which powers most Boeing 747 and 767 wide-body aircraft, and has accumulated more than 100 million hours of operation on-wing. Generating units have added more than 10 million hours to that total.

The LM6000 produces between 43 and 50 MW (ISO conditions) at a design thermal efficiency exceeding 43.5%. Package is compact (direct drive, no gearbox in 60-Hz applications), enabling installation of the power block on constrained sites. Advantage for peaking service is the engine’s quick-start ability: It can reach maximum power in 10 minutes. Regarding emissions, units equipped with DLE combustion systems can meet 15 ppm NOx. GE Energy claims a demonstrated reliability of 99%, availability of 98%.

Machine details. The compressor has a total of 19 stages—five lowpressure, 14 high-pressure—and a compression ratio of 30:1. Casing is horizontally split. Turbine has two HP stages, five LP. Generator outputs electricity at 13.8 kV; power factor is 0.9.

Goehring opened the meeting asking for a show of hands regarding the particular models operated by the users in attendance—this to guide the presentation. Majority of users had the PC. PA owner/operators numbered about 15, PBs and PDs were in the single digits.

Where’s the lockwire? Discussion began with a review of the “missing- lockwire” issue associated with HPC variable bleed valves (VBVs). Goehring asked the users if they have experienced loss of lockwire and about a dozen participants raised their hands.

One delegate said that investigators assessing damage to two engines at his plant found lockwire in both machines. Thinking is that when the wire comes loose it flies up into the collector duct during shutdown; when blocker doors open the wire drops down into the engine.Depots said they have no evidence of an HPC failure caused by lockwire but that it is a possible contributor to damage.

Then came a question regarding operating regime: Are engines operating at low load more susceptible to the lockwire problem? Consensus of the group was that there is a partialload correlation to the lockwire issue and users thought it was related to an increase in air flow at low loads.

A service provider in attendance recommended that mechanics look closely at the VBV actuator bolts during every inspection. He said many users may have the problem but are not aware of it because mechanics are not specifically verifying that lockwires are intact.

One user said his experience suggested that the PC engine was more susceptible than the PA to lockwire failure. He thought that the change in air flow with the PC design increases the vibration level and the probability of breakage.

A question on how to deal with the lockwire issue elicited several suggestions. Some attendees were for keeping the lockwire, others for eliminating it. One of the pro-lockwire advocates suggested running the wire in the direction of air flow, not across it, to prevent breakage.Another recommended using RTV on the safety wire to secure it. Of those suggesting that the lockwire be eliminated, one believed use of Loctite® on the bolts was a viable alternative. Another suggested use of a tab lock bolt. Attend the 2007 meeting and contribute to the dialog.

HPC bushings. Worn bushings can lead to lever-arm stress, sometimes failure. A new multi-piece bushing, described in SB-IND- 213, is designed to rotate during operation. Greatest wear appears at the casing split-line. One user’s experience is that if you keep the bushings around the split line tight, then the remainder should be fine. The depot workbook provided to all attendees offered details on inspection limits and recommended actions.

Inlet gearbox (IGB). Goehring suggested implementation of the recommendations offered in SB-IND- 220 to prevent spline wear caused by inadequate lubrication and/or movement between the horizontal gearbox and the HPC adapter. A few users in the room reported IGB failures. One plant owner mentioned that it had inspected the IGB on one unit and found no sludge, but six weeks later the gearbox failed with only 13,000 service hours on the engine.

A point was made that corrective action presented in SB-IND-220 could be implemented in the field— provided there was no wear. Anything other than “as new” condition would require a depot visit.

HPT diffuser. Original cast part is known to crack and fail—usually during startup. SB-IND-216 recommends replacement with a forged Inconel 718 diffuser. OEM’s hours-based voucher program offers significant benefit to users whose machines don’t run very often or long. One problem for operators is that it’s virtually impossible to find a crack in the diffuser using a borescope.

G39 combustor promises to correct the following problems: splashplate oxidation, fuel nozzle/primary swirler wear, secondary swirler TBC spallation (see SB-IND-208). Field experience with this component is just beginning to be made available through the user organization. Only four users said they had installed the G39; leader had 11,000 operating hours.

Mention was made that GE was working on two improvements to its DLE combustor but no details were available.

HPT Stage 1 vanes. SB-IND- 223 introduces the N5 single-crystal vane which is said to more effectively resist oxidation and fatigue than the part supplied with the machine. In particular, users have experienced coating degradation on the platform and leading edge of Stage 1 vanes because of thermal mechanical fatigue (TMF).

HPT Stage 1 blades. Users have experienced tip rub with TBC spallation as well as TMF attributed to frequent cycling. First condition sometimes can be corrected by better matching of the TBC’s thermal expansion characteristics with those of the base metal.

Best way to mitigate the effects of TMF may be to replace the supplied DSR-142 blades during the next scheduled hot-section repair instead of repairing them. Two options here: Upgrade to single-crystal N5 blades (refer to SB-IND-191) or to a more advanced single-crystal blade with improved cooling at the tip and on the suction side (refer to SB-IND- 215).

Four users at the meeting were operating with N5 blades; one offered that they “look good.” This led to a question on the recommended frequency for borescoping to check blade condition. Answer from the depot: Semiannually for units in base-load service. A user said borescoping is done quarterly at this plant. That exchange was followed by a question regarding hot-section interval for the more advanced blades. No clear answer here; suggestion was to make a determination based on your unique duty cycle, impact of water or steam injection if provided, method of inlet cooling, etc.

HPT Stage 2 vanes. Parts produced after issuance of SB-IND-124 experienced TBC spalling and oxidation; parts after SB-IND-163 suffered TBC spalling, oxidation, and cracks in the airfoil’s trailing-edge fillet (both the TBC and base metal had cracks). While both vane configurations serve for the expected lifetime, the later parts performed better.

Following the formal depot presentation, questions and comments on a variety of subjects ensued. Here’s a sample:

Engine preservation: Operating regimes have changed and more units have significantly longer periods of shutdown than in past years. Suggestion was to review procedures to ensure proper layups. One reference is GE Work Package 3011.

Engine handling: Be sure you’re using proper lift points when shipping engines in containers.

Transportation: Hire tractor/ trailers with pneumatic air-ride suspension to prevent bearing damage.

Critical parts life management: Users must track the operating histories of their parts and remove from service any parts that have reached their cyclic life limit (refer to sidebar profiling Della Villa’s presentation on previous spread). It was clear from a show of hands that some users are slow in adapting to the new reality. Discussion plodded along after a while until Amarel ended it by stating that all users should not delay in initiating a program of parts life assessment.

Special technical presentations go beyond the engine

The second half of the second afternoon at WTUI annual conferences is devoted to technical presentations on topics of interest beyond the engine O&M subject matter covered so thoroughly in the LM breakout sessions. This year, seven presentations were conducted within a two-hour window in three different rooms. Topics generally are so timely and valuable to users, it can be difficult to narrow your selection to two presentations—the most you can possibly attend. Brief summaries of six presentations follow.

Cooling towers

Anytime you have the opportunity to listen to Bill Stroman speak on the subject of water treatment, reserve a seat in the meeting room. Stroman, now manager of water chemistry for Primary Energy Ventures LLC, headquartered in Oak Brook, Ill, has monitored and treated water in powerplants across the country, as well as internationally, for more than 30 years—so he’s a bona fide expert on the subject, one who has pretty much “seen it all.” But equally important is that Stroman is not at the podium to impress you but rather impart some practical know-how to make your primary job—generating electricity—easier.

Stroman’s topic for the 2006 Western Turbine Users conference was “Cooling Towers: Safety, Maintenance, Chemistry Control,” which he expanded to include the condenser and remainder of the circulating water system. Tough to present a comprehensive overview on the topic in the limited space available here.Stroman whipped through 118 slides in the hour he spoke, most illustrating problems to look for and what happens when you don’t pay attention water chemistry. It was an eye opener for anyone who didn’t believe in the importance of proper cooling- water chemistry when he or she walked into the room.

Stroman began with a slide that outlined his presentation— the major factors that influence coolingtower operation:

• Water quality.
• Chemistry control and program administration.
• Water and air flow rates.
• Air-to-water distribution ratio.
• Heat load.
• Maintenance program.

Physical and chemical treatment programs typically are unique to a given plant because of the many variables that influence them—particularly water quality. In certain areas of the country—primarily the arid West—strained water resources often dictate the use of grey water (typically treated municipal wastewater) and mineral-rich well water. Zero liquid discharge mandates in some states further complicate treatment.

Safety first. Before discussing treatment, Stroman stressed the need for a practical and effective safety program for handling and storing chemicals. He offered the following checklist:

• Read product literature and material safety data sheets (MSDS).
• Know what hazards (if any) exist.
• Make available protective clothing required for safe handling of chemicals and be sure it is worn.
• Know first-aid practices, etc.
• Be sure plant personnel know the locations of safety equipment, emergency eye wash/showers, safety kit, and emergency contacts.
• Clean out chemical containers prior to disposal.
• Use and maintain the chemical feed systems installed.
• Take no short-cuts.

The successful cooling-water treatment program, Stroman continued, has these four major elements:

• A technically sound mechanical and chemical program tailored to the system and its operation.
• Capable and reliable chemical feed system.
• Ongoing monitoring and control using the proper techniques, equipment, and supplies.
• Knowledgeable personnel who understand the importance of the cooling water system with regard to efficient, reliable, and safe power production. Also important is for the plant staff to commit to do what’s necessary to maintain and control the chemistry program.

Stroman spent a significant amount of time reviewing how to control what he calls “the problem trio”—corrosion, scale, and fouling (sidebar).A particularly interesting slide showed the impact on heat transfer of common foulants. Silicate is the most efficient insulator. The thinnest of coatings (perhaps a tenth of a millimeter or less) can reduce heat transfer by 50%.

By contrast, it takes about 6 mm of calcium phosphate—perhaps the most predominant scale today because phosphate-based corrosion inhibitors are used extensively—to cut the heat-transfer rate in half. In between are biofilm (1 mm), biofilm/ scale (2.5 mm), calcium sulfate (4 mm), calcium carbonate (4.5 mm), iron oxide (5 mm).

SCR for simplecycle GTs

A few years ago many simplecycle GTs—more than a dozen in California alone, the majority LM machines—experienced SCR catalyst failure. At least one LM6000 peaker known to the editors of the COMBINED CYCLE Journal suffered catalyst breakdown in less than 500 hours of service. Virtually overnight high-temperature catalyst was viewed askance by owner/operators. However, a review of plant data provided to the editors by user sources revealed that most, perhaps all, of the failures occurred with one supplier’s product.

Elizabeth Govey assured attendees that Cormetech Inc’s (Durham, NC) experience with simple-cycle SCR catalyst has been excellent. She said that her company, Cormetech, has demonstrated continuous successful operation of SCR catalyst in simple-cycle operation in the US since 1996. The presentation, prepared by Govey and colleagues Dr Christopher Bertole and Karolyn Hagan, was fairly technical and offered operating data to support the company’s claim. Several LM5000s and LM6000s, and one 7EA frame, in the database are running on replacement catalyst from Cormetech after experiencing problems with the product supplied originally.

An important point made by Govey was that while catalyst formulations are standard, the product is extruded for the specific application.Simply put, catalyst is optimized for each plant’s operating conditions. Thus, to achieve five years of reliable service—the industry standard today—owners must provide their catalyst supplier accurate data regarding turbine-exhaust-gas (TEG) flow, composition, and temperature, and expected plant duty in terms of operating hours and cycles. Other information required includes percent NOx reduction, ammonia slip, differential pressure, etc. The better the data, the tighter control you will have over O&M costs relating to emissions control.

Bertole stressed that temperature is the main driver of catalyst degradation.Cycling has relatively little impact on the lifetime performance, he added—that is, unless the temperature ramp is not maintained within recommended limits. The rate of catalyst degradation typically is not of concern until TEG temperature is above about 1000F. What designers do then is opt for an arrangement that permits staged replacement of catalyst over time. Benefit of this approach is lower backpressure and reduced operating cost.

Another option designers have when TEG temperature is high is the use of tempering air to cool the exhaust stream. However, Govey was confident that SCRs can be designed and successfully operated for simple- cycle GTs using homogeneously extruded titania-based catalysts with or without tempering air. Note that titanium dioxide is the base material in the catalyst formulation and that other active materials—such as vanadium pentoxide and tungsten trioxide—are varied according to TEG temperature.

Govey spent several minutes reviewing the cost/benefit of using/ not using tempering air. The benefits of tempering air are:

• Permits a higher ratio of vanadium to tungsten, which translates to less catalyst volume and lower pressure drop.
• Guarantees longer catalyst life.
• Allows use of larger catalyst modules, possibly ones with carbon- steel frames (lower cost than chrome/moly steel).
• Financial penalty associated with tempering air is the capital and O&M costs of the fan and related components (ductwork, controls, etc). It also requires more space, which may be an issue on small sites. A downside risk is that the catalyst can overheat if the fan fails.
• The benefits of operating without tempering air include these:
• Reduces the amount of capital equipment required, and its maintenance.
• Does not add parasitic load.
• Eliminates loss of turbine efficiency attributed to the pressure-drop contribution of dilution air.

Generators a hot topic

Most powerplant personnel excel in mechanical aptitude. In general, they’re most comfortable around gas and steam turbines and heat-recovery steam generators. When it comes to electric generators, well, they’re often taken for granted. Obviously, that’s flawed thinking.

Recognizing the importance of a continuing dialog on generators, the WTUI program committee organized an air-cooled generator track during the time allocated for technical presentations on Tuesday afternoon. If you didn’t leave the room, you surely benefited from listening to these experts:

• David Branton of Wood Group Generator Services Inc, Farmington, NM (david.branton@ woodgroup.com), who spoke on stator and rotor testing and maintenance.
• Ron Cox of Alstom Power Inc, Midlothian, Va, (ron.cox@power. alstom.com), who spoke on a new repair procedure for generator stators condemned by the OEM.
• Rajiv Sharma and Greg Stone of Iris Power Engineering, Toronto, Ont, Canada (rsharma@irispower. com), who spoke on the use of generator partial discharge testing for determining the condition of stator insulation.

Branton began with a review of three diagnostic procedures for determining stator health. Here are thumbnails of each:

1. Power-factor tip-up test examines the windings at various voltage levels. It is significant in the degree of detail provided with regard to winding serviceability. Factors such as contamination, corona suppression system degradation, partial discharge activity, insulation delamination, and voids impact test results.

2. Corona camera examination. A corona camera is a specially designed video instrument used to record the presence of partial discharge activity invisible to the naked eye under normal turbine-deck lighting conditions. Instrument enables the viewer to see what a particular winding looks like under normal operating conditions.

3. Slot discharge examination. The slot discharge probe is sensitive to the ratio- frequency signals produced by partial discharge activity within a winding. It is easy to use: A technician slowly moves the probe down each individual slot of an energized winding until the meter reading peaks. Probe also can be used to locate partial discharge activity in the end-winding regions.

When partial discharge activity is identified, Branton continued, the affected area should be cleaned and painted at a minimum. It may be necessary to fill-in the affected area with epoxy felt or RTV silicone foam. Another consideration: Removing and replacing damaged insulation with new.

He also mentioned that excessive movement of windings could be the result of the winding “basket” being in a resonant condition. This can be corrected by replacing loose blocks and ties and/or by installing an end-winding stability system.

Branton said that partial discharge activity in the stator slots can be reduced by filling voids between coils and slot walls with a conductive filler, RTV, or epoxy. Two more possibilities: (1) Replace the stator-slot wedge system, thereby tightening the windings. (2) Removing, replacing, or reinsulating the windings.

Rotors were next on Branton’s agenda. He stressed the importance of monitoring and evaluating the data gathered. For example, a winding short can be identified by a change in running vibration or an increase in excitation current to maintain a given load. The flux probe is valuable in that it will accurately pinpoint any turn-to-turn or coil-tocoil shorts. One caveat: Although turn-to-turn shorts may not be an immediate concern, careful monitoring is recommended.

Branton then reviewed rotor tests users should be aware of and concluded by quickly reviewing nearly four-dozen shop photos to show various types of damage and repair procedures. What follows are short descriptions of the rotor tests:

• Static turn insulation integrity test (STIIT™) is used to measure and compare readings for individual coils within a two-pole rotor. Readings for like coils of opposite poles are compared to each other as well as to industrial standards. Differences in readings are indicative of the severity of turn-to-turn shorts in a particular coil (Fig 2).
• A pole drop test is performed by inducing a voltage into a rotor winding, measuring the voltage drop across each individual pole, and then comparing the two agains t eac h other. If readings differ by more than 5%, turnto- turn shorts are suspected in the pole generating the lower reading (Fig 3).
• An ac impedance test is another way to test for turn-to turn shorts. It is performed with the rotor at rest or at any speed by inducing into the rotor winding a series of voltages that increase incrementally. Voltage and amperage are measured and recorded. Resistance is calculated and graphed against the applicable voltage levels.
• Nondestructive examination. Branton recommended that all critical rotor components be examined whenever possible—for example, during a GT major, when the rotor normally is pulled. The condition of retaining rings, rotor body wedges, blower hubs, couplings, and the forging itself should be examined by the most suitable method among ultrasonic, dye-penetrant, Zyglo®, eddycurrent, and magnetic-particle techniques (Fig 4).

Cox’s presentation was a big hit among the users in attendance because it illustrated how a plant was able to exact considerable savings by opting for a generator-field repair over replacement in-kind. In today’s competitive power market this can mean the difference between a profitable year and one that is not.

It also showed how rapidly repair technologies are advancing. What might not have been fixable last year may now be. With orders of new equipment down from a few years ago, OEMs, and the third-party service providers that compete against them, are investing heavily in equipment and techniques for maintenance and repair that bring value to the user.

Interestingly, for the generatorrepair case history profiled by Cox, Alstom, one of the world’s leading suppliers of electrical generators, was a third-party services provider.

What happened: A 160-MVA aircooled generator shut down on ground fault. Reason: Two balance weights had broken loose from the generator field and damaged the stator core to the extent that the OEM condemned it. The owner approached Alstom for an alternative to replacement.

Engineers conducting the investigation found core laminations badly damaged between adjacent slots (34 and 35), plus additional lamination damage in the surrounding area (Fig 5). Specifically, there was major damage to ground insulation (Fig 6), a cavity was created in the tooth sidewall where lamination melted away, and melted lamination material was found in vent ducts. A baseline ELCID (for electromagnetic core imperfection detection) test confirmed major damage in this area and found other areas of minor damage.

The inspection team’s assessment: Damage was bad but there was a good chance for a successful repair based on past experience and the use of a proprietary epoxy injection method.

The proposed solution: Machine out the visible area of tooth damage—a section about 6 in. long by 3.5 in. deep (Fig 7)—and inject a nonconductive epoxy into the remaining laminations. That done, a core loop test was performed, but the temperature rise exceeded acceptable limits. Engineers decided to mill out an additional 2.5 in. of tooth height, leaving only 1 in. of the original 7-in. tooth height remaining.

The epoxy injection process then was repeated and a second core loop test proved the repair satisfactory: only about a 4 deg C temperature rise after one hour of testing.Next, engineers designed a special non-magnetic, non-conductive G-11 (NEMA designation) filler block to help support wedges in slots 34 and 35 and to assure compression of the remaining laminations in the milled area (Fig 8).

Note that a conforming material was used between the block and the top of the milled tooth to ensure a good bond between block and tooth (Fig 9). Also, a semi-conductive paint was applied to the sides of both the filler block and the remainder of the tooth to prevent corona discharges.

The core then was partially rewound with new bars supplied by Alstom (Fig 10) and a confirming ELCID test was conducted with satisfactory results. After final electrical tests, the unit was returned to service in early February. Total repair took two months because it was partially based on a time-and-materials work scope. Cox said that, given the experience gained, a partial rewind and repair of this type could be done in less than 45 days.

Duct-burner igniters

Merrill Jones, a Forney Corp product manager located at the company’s headquarters in Carrollton, Tex (merrill.jones@ forneycorp.com), conducted a tutorial on duct-burner igniters. His presentation was paired with Stroman’s on cooling towers (see above) in one of the three afternoon meeting rooms.

Many LM plant personnel work at peaking facilities, which don’t have duct burners. But in the nomadic business of powerplant operations what is today isn’t tomorrow, and the WTUI meeting provided a learning opportunity. Of course, for those delegates already employed at cogen and combined-cycle plants, the update also was valuable. Jones began by outlining the scope of his presentation. He started with a couple of slides showing what a duct burner (Fig 11) and igniter (Fig 12) looked like so everyone in attendance could follow what he had to say. Then Jones discussed ignition requirements, the types of igniters, spark source, flame detection, and maintenance and troubleshooting.

The basic requirements of any igniter are these:

• Fuel. Natural gas at a constant supply pressure was assumed for the WTUI presentation.
• Air. Source of supply is GT exhaust or the ambient environment; a 5% to 15% fuel/air ratio is necessary.
• Ignition source—transformer or high-energy spark.

Successful ignition requires proper mixing of fuel and air and a reliable spark. Plant conditions often make achieving these design goals more challenging than it would appear. Recall that duct burners often have long periods of nonuse, the atmosphere in which they operate contains upwards of 20% moisture and is highly turbulent, turbine-exhaustgas (TEG) temperature varies from ambient to 1200F, and the oxygen content of TEG is low and may vary from about 10% to 14%.

There are two basic types of igniters: so-called “airless” because all combustion air is mixed in the TEG, and “supplemental air” where oxygen content is raised by injection of ambient air. The latter obviously eliminates many variables that could adversely impact performance. Excellent fuel/air mixing is a characteristic of the supplemental-air igniter if the pressure of both fluids is maintained within the design range. Note that mixing of fuel and air occurs before the combustion chamber—so-called “premix.” Secondary combustion is supported from TEG.

High-tension igniters (6000- 10,000 V dc) provide a continuous spark across a gap (Fig 13). Operational challenges: spark gap can vary with heat and operating profile, and the electrode can short-out when moisture and/or dirt are excessive. Advantage: Low initial cost.

High-energy igniters produce three or more sparks per second (Fig 14). Advantages: spark action selfcleans the tip, tip is sealed and will not short-out, device is cost-effective over its design lifetime.

Flame detection is extremely important. Perhaps most important is that the detector, which most often senses the ultraviolet radiation spectrum, must be able to “see” the igniter flame. Thus layout of components at the burner front has a major impact on reliable flame detection. A swivel joint helps to position the igniter accurately.

The NFPA (National Fire Protection Assn) requires only one detector for the Class III igniter typically used for heat-recovery steam generators. It must “see” both the ignition and main flames (Fig 15). End shield shown in the photo helps to protect and stabilize the igniter flame.

Igniter maintenance is an important consideration at the specification stage. For example, condensation will occur at the back end of the igniter and you can reduce its impact by specifying the use of stainless steel and a sealed high-energy spark ignition rod. It’s also a good idea to provide a steady supply of cooling air to the igniter to reduce moisture effects and to prevent particles of dirt from accumulating on it. And since igniter removal may be required with the GT in service, be sure to specify a knife gate valve or aspirator to prevent the backflow of TEG at high temperature.

Minimal maintenance is required to keep your igniters in top condition. Periodically clean and test the spark source and inspect electrical connection points for open or short circuits. Replace worn parts as necessary. Also, inspect and clean all gas orifices and ensure that the supply of supplemental air is satisfactory.

Maintenance cost optimization

As noted in the summary above of Alstom’s Ron Cox presentation on an innovative generator repair that saved a “condemned” stator, repair technologies are advancing at a rapid pace. This is nowhere more evident than in the methods used today to extend the lives of GT hot-gas-path (HGP) components.Advancements in metallurgy—particularly coatings—and software are enabling a positive change in attitude toward repairs among users. Competitive power economics demand O&M savings and repair by a capable service provider—OEM or third party— is one sure way to reduce costs.

Liburdi Turbine Services Inc of Dundas, Ont, Canada (US office in Davidson, NC), is one of the leading third-party providers of HGP repair services and a frequent participant at user-group meetings. Lloyd Cooke, Doug Nagy, and John Bottoms, seen most often, are respected by most owner/operators for their knowledge and can-do attitude. At the 2006 Western Turbine Users meeting, the Liburdi presentation was made by Martin Perrin, PE, who manages the company’s aero business unit.

The title of Perrin’s talk was wordy, but a helpful précis: “Maintenance Cost Optimization through On-Condition Maintenance Schedules and Advanced Component Repairs.” He began by reviewing the three methods typically used for scheduling maintenance:

• Fired hours, as established by the turbine OEM—for example, a hot section every 24,000 hours.
• Equivalent hours, which includes the impact of starts and stops and is calculated by using formulas supplied by the OEM.
• On-condition maintenance (OCM).

Service interval, as determined by the third method, is based on an evaluation of actual engine operation and its metallurgical effects on components.Thus the overhaul schedule is unique to the engine and its operating regime. Advantage is that the service interval is optimized and key components are known to be repairable at the time of removal.

Perrin continued with the key steps for determining the OCM service interval using a turbine blade as an example:

• Determine the blade’s life limiting factors.
• Model the blade’s thermal mechanical state as it relates to the engine’s operating performance parameters.
• Develop analytical models for the critical degradation modes.
• Combine models into a software solution.
• Track engine operating performance parameters online.
• Calculate time to overhaul.

Condition assessment is critical to doing a realistic lifetime analysis for the blade. A metallurgical examination is conducted to determine the type of damage and to quantify the extent of material degradation. Metallurgists look for oxidation/corrosion attack and microstructure degradation, check coating thickness and integrity, verify creep strength, and take precise measurements to identify dimensional changes.

Note that metallurgists consider both time- and cycle-dependent degradation when assessing remaining service life. First group includes oxidation, hot corrosion, and creep; second, thermal mechanical fatigue. Life-limiting factors vary significantly with engine design, application, and operating conditions, as well as the quality of prior repairs.

Modeling to determine OCM is a complex task that considers the thermal design of the overall engine, blade heat transfer, and mechanical stress and strain, and makes use of special life-assessment models for cyclic oxidation, creep, alloy aging, and thermal mechanical fatigue.

Liburdi’s first use of this leading- edge technology was to develop a remaining-life model for a major pipeline operator’s fleet of 43 LM1600 engines. The pilot system for the LM1600 has been installed and the concept validated. Customer’s goal to maximize service-interval hours, extend component life, and reduce maintenance costs is achievable.

In this pilot installation, the model identified that all six engines monitored could achieve more than the conventional service interval and still maintain repairability of the HPT blades. The optimized service intervals ranged from 31,000 to 42,000 hours. These results were verified and confirmed the savings that can be achieved through OCM.

Future work calls for fleet-wide deployment of the life analyzer to monitor the consumption of HPT blade life in all units. An important adjunct to fleet-wide implementation is a continuous program of model calibration. This requires periodic metallurgical analysis of blades from lead engines to validate deterioration rates and to search for unforeseen failure mechanisms.

Important to all in attendance was the company’s goal to pursue expansion of the LM1600 work to the LM2500 fleet. Engineers and metallurgists believe the process is transferable and scalable for other types of aero engines.

Perrin concluded his presentation with several slides illustrating lifelimiting factors and how Liburdi’s advanced repair technologies could be used to extend the service lives of LM2500 HPT blades (Fig 16). Benefit: Users could see first-hand what’s repairable with today’s technology and what’s not—certainly a valuable benchmark for decision-making during their outages.

LMS100 overview closes program

LMS100, the engine that GE has been touting for years, was the subject of the final presentation at the 2006 meeting of the Western Turbine Users. Program Manager Michael Michael spent most of his time at the podium reviewing the engine’s design, performance, and ease of maintainability. Perhaps of greatest interest to those in attendance were data culled from teststand results. The machine was full-load tested in Houston before the first commercial unit was installed at mid year by Basin Electric Power Coop in Groton, SD. This machine will be operated as a remote peaker.

Those in attendance who had not seen the engine previously learned that its design incorporates some very successful components from the OEM’s fleet of aeros and frames. For example, the HPC was taken from the CF6-80C2 on-wing engine, the HPT from the CF6-80E, and the LPC from the 6FA frame (Fig 17).

The machine, in standard peaker configuration with a DLE combustion system, is said to develop 96 MW (ISO) at a heat rate of 7600 Btu/kWh based on the lower heating value of natural gas and the most favorable of operating conditions. In a STIG arrangement (introduction scheduled for mid 2008), the machine should develop 112 MW at 6850 Btu/kWh. Capturing intercooler and exhaust energy can push efficiency over the 90% mark in CHP applications, according to the manufacturer .

Michael said that full-scale testing (109 hours on gas, 11 on liquid fuel; 68 starts) validated the machine’s ability to:

• Operate at maximum power (over 100 MW).
• Meet heat-rate requirements.
• Start in 10 minutes.
• Follow load.
• Meet emissions requirements.

Maintenance intervals claimed are the same as for the company’s other aeroderivative generation packages: hot section every 25,000 hours, depot maintenance every second hot section. ccj

 

Independent voice of the gas-turbine-based generation sector

© 2014 CCJ Online, Inc. All rights reserved. | Published by PSI Media, Inc | 7628 Belmondo Lane | Las Vegas, NV 89128