7F Users Group: Compressor issues dominate discussion at annual conference

The 7F Gas Turbine User’s Group consistently conducts a superior annual conference. Three things stand out when reviewing meeting notes:

• User involvement. The 7F is the largest frame user group, with more than 700 active members. Attendance at this year’s conference— nearly 250 user delegates, including more than 50 first-timers— was more than double that of the next largest frame meeting in 2006.
• Practical content. User presentations dominate the four-day program except on Day Three—General Electric Day—when OEM engineers review fleet operational history and update the group on new and existing issues and solutions.
• Comprehensive exhibition. The second evening of the 7F conference was reserved for a three-hour vendor fair which included reception and dinner on the expo floor. This year, 53 companies showcased products/services ranging from inlet filters to borescopes to compressor washing systems to lube-oil treatment and services.

The 2006 meeting was held May 8-11 in at the Emory Conference Center, close to GE Energy’s Atlanta headquarters, allowing maximum interaction with key OEM personnel. Nearly 50 of the manufacturer’s specialists were available on Day Three to make presentations, conduct interactive sessions, and answer questions across a wide range of subjects.

In addition to the annual conference, the group’s website, http:// GE7FA.Users-Groups.com, has been instrumental to the success of the organization, according to members of the steering committee (Sidebar 1). It allows all users, including those not able to attend the annual conference, to share and gain information and experience 24/7. Those interested in joining the organization are urged to submit profiles via the membership drop-down menu on the website. Only users can qualify as “members”; others may be accepted as “affiliates.”

Day One: Warmup

The Day One schedule allowed for arrivals, miscellaneous events, and an opportunity to meet with colleagues in an informal setting. The golfers were out on the Stone Mountain Lakemont Course by noon, before registration opened. Tournament was organized by Rick Parker, the gas-turbine (GT) compressor cleaning expert from Zokman Products Inc, Fort Wayne, Ind, and Lee C Wood, regional sales director, Wood Group Gas Turbine Services, Flower Mound, Tex (Sidebar 2).

A pre-conference event associated with the 7F meeting is a so-called Spotlight Session on F-class heatrecovery steam generators (HRSG) presented by HRST Inc, Eden Prairie, Minn. With the golfers (mostly vendors) out on the course, about 50 users signed up for an update on HRSG O&M practices and problem resolution. Presenters/discussion leaders for the afternoon program were Lester Stanley and Scott Wambeke, both registered professional engineers (Sidebar 3).

The formal portion of the 2006 meeting began in the early evening with a welcome reception for users and a keynote address by John Krenicki, president/CEO of GE Energy. Krenicki, in his current position for less than a year when he made the presentation, worked previously at the company’s silicones, lighting, superabrasives, transportation systems, plastics, and advanced materials business units.

His overview of GE Energy’s initiatives throughout the generation sector—including nuclear, coal, and renewables—helped to expand the views of power professionals focused on operation and maintenance of gasturbine- based facilities. The portion of Krenicki’s remarks on the GT sector included the company’s efforts directed at ensuring the availability of combustion systems capable of burning LNG and synfuels when they achieve commercial viability. Another positive message was his commitment to enhanced customer service.

Day Two: Compressor, Turbines, and combustion systems

Shim outages. Dominion Energy’s Paul White, a member of the 7F Gas Turbine User’s Group steering committee and the chair of the compressor session, jump-started the Day Two program with his presentation on compressor shim issues and solutions.

Keep in mind as you read through the précis of this presentation and the ones that follow that the intent of the editors and the steering committee is to bring the greater community of 7F users up to date on fleet-wide issues and on the concerns of their colleagues, not to solve problems in print or to provide details on corrective action. There is no substitute to attendance at the annual meeting for digging into issues of particular concern to you and the generating assets that you manage.

White explained that some shims, which measure either 40 or 80 mils in thickness and can weigh up to about two ounces, have begun to come loose from their locations between adjacent stator-vane ring assemblies in midcompressor- case stages S0-S4. Problem also may occur on 7EA frames in stages S1-S4. Recall that vanes for the stator stages referenced are mounted in a ring assembly saw-cut into six segments; shims are installed between segments where needed to make up for saw kerf material loss.

White encouraged users to verify by borescope inspection the number, size, and location of shims and that shims are flush with the ring segments. If protruding shims are found, you should measure the height and determine the course of action. Risk increases when shims come loose and protrude into the flow path. If outward migration continues, some industry experts contend, the resulting flow disturbance could cause highcycle fatigue damage in neighboring blades. In the extreme, if a shim or a portion of the shim is released into the flow path DOD (domestic object damage) is likely.

The migration of shims appears related to the number of starts a machine undergoes, White continued, and it occurs because ring segments grow at a different rate than the casing. Inspections have confirmed a range of observations—from several shims protruding to all shims in place. Roughly a third of the shims in one group of units inspected were found to have stress cracks in tabs or missing tabs—conditions conducive to shim migration. The tabs, sometimes called ears, hold the shims in the hook fit. In one case, a crack about 2 in. long was found below the casing ID surface in a 40-mil shim before any damage occurred, supporting the case for shim removal prior to release.

Correcting a migrating shim condition suggests removal of ring segments, which can be difficult depending on the age of the machine and contaminants in the inlet air and wash water. Shim extraction may be possible eliminating the current risk. Also, grinding a shim flush may be an interim risk solution. During an outage, soaking the lowerhalf segments and being patient is the best removal method.

If you can’t remove lower-half segments, the situation may require rotor removal and destructive extraction of the segments. White added, “The OEM worked well with us during outages to develop several creative approaches to deal with shim constraint, including welding and bolting. The mutual goal was to leave the unit with no future shim risk.”

A pinning technique developed by Rodger Anderson of DRS-Power Technology Inc, Schenectady, NY, deals with the aft individual square-base stator vanes.

7FA+e compressor repair. An international user reported on compressor issues identified after upgrading two GTs in 2004 from 7F to 7FA+e. Reason for the upgrade was to boost output by 10%. A borescope inspection in 2005, scheduled after 2500 hours or 150 starts, identified some damage but suggested no root cause. Numerous dings were found on each stage of rotating blades; 19 stator vanes were damaged on one unit, 42 on the other. Correction was by blending according to OEM guidelines.

R0 rub/blade repair. Compressor damage was identified after a highvibration event tripped the unit. User followed the OEM’s recommendations in TIL (Technical Information Letter) 1509, conducting a visual inspection of the compressor inlet. Rolled edges were identified on R0 blades. Two cracks were found using dye penetrant—one about 5?8 in. long at the trailing edge, the other about ½ in. long a short distance from the trailing edge.

User worked through the OEM’s Power Answer Center (PAC), which gives owner/operators access to a GE Energy support team capable of addressing specific technical inquiries. Recommendation was to cut out the cracks from the two blades and an OEM blade specialist was hired to do that work. Follow-up dye-pen test showed cracks had been removed.

On restart, operators noted that the vibration level had decreased and there was no obvious loss in performance. However, no formal performance test was conducted after the repair. No details were offered as to the cause of the damage that initiated the trip.

Foreign object damage (FOD). A compressor borescope inspection was initiated after user identified cracking in transition pieces. Ten R17 rotor blades and one S17 stator blade were damaged at their respective trailing edges. OEM’s investigation revealed that a wrong borescope plug had been installed in the unit and it fell into the compressor. The entire row of R17 blades was replaced and the correct plug installed.

Modified stator shims and EGV rocking. Two innovative solutions were offered by a user with a significant investment in 7FA engines to address stator shim issues and the rocking of stator and exit guide vanes. No information was provided for this report; details were made available only to users attending the annual meeting.

Clearanceometer plug issue. One of the so-called clearanceometer plugs that allow compressor entry for blade clearance checks was inserted too far and damaged a row of rotating blades. Another plug blew out because if it were screwed in further it would have contacted another row of blades. Investigation by the user revealed plugs of four different sizes were used in the machine.

Borescope inspection identified damage to R10 blades that required repair. Compressor case was lifted; R10 blades were removed (revealing additional damage) and cropped at both the tips and trailing edges.

Compressor/turbine health inspection. A user ran through a “laundry list” of issues identified during (and following) an R0/R1 compressor inspection that suggested an early outage. A companion check of the turbine revealed cracking on first-, second-, and third-stage buckets; severe cracking on the inner and outer sidewalls of first-stage nozzles. User reported that the OEM is not concerned with such cracking until it gets to the leading edge. Damaged first-stage nozzles were replaced.

Compressor casing had moved more than an eighth of an inch, locking in blades. Machine shop fabricated a jack that could be attached to the upper casing to facilitate its removal. Work scope for compressor work included the following: Mapped all blades and performed blending as required, installed P-cut R0, implemented TIL 1502-2R1, replaced a damaged S1 segment, checked/ adjusted IGV (inlet guide vane) clearances, repaired rocking stationary blades.

The compressor roundtable discussion, moderated by White, focused on IGVs and water washing of units with P-cut blades. The IGV discussion touched on one user’s attempt to change the spring washer from steel to plastic as suggested by a TIL. The participant reported that the steel thrust bearing was steel and had rusted and jammed. Consideration was given to installing grease fittings to facilitate operation.

IGV freeze-up, particularly IGVs on older machines with steel bushings, has been experienced by many users with Frame 5, 6, and 7 engines. The real challenge seems to be fixing/refurbishing/replacing lower-half IGVs with the rotor in place. For more on this technique, and on IGVs in general, refer to “Replacement of damaged lower IGVs with rotor in place saves IMPA a million,” p OH- 57, 2007 Outage Handbook supplement inserted in the middle of this issue.

Another roundtable topic: A user’s experience during a combustor inspection (CI) when the IGV ring flexed and six gears jumped a tooth. Obviously, vanes didn’t line up as they had before and the user was concerned. OEM said the machine could be operated, but the user was not comfortable and vane settings were corrected prior to restart. Interestingly, with all the advanced technology available to users today, there’s still no replacement to manual adjustment of IGVs using the ancient protractor.

Online water washing is vital to minimizing performance degradation between offline washings but there are industry concerns regarding duration of spray, droplet size, use of detergents, spray pattern, etc. Some machines are more sensitive than others to online washing. The 7F arguably is the most sensitive. Users discussed a modification to the OEMsupplied wash system that moves the erosive effect of washing R0 blades away from the particularly sensitive root area to the blade midsection.

A new online water wash system from GE Energy was mentioned as the discussion continued. The way those participating understood it, the new system would be offered as an upgrade and would allow online washing for up to 30 minutes daily on units with P-cut R0 blades, a maximum of five minutes a day on units without the P-cut blades.

Turbines, combustion systems

The afternoon of Day Two was devoted to the turbine and combustion systems. Ed Fuselier chaired the turbine session and moderated its roundtable discussion; Peter So provided the leadership on combustion systems. There were three formal user presentations on turbines, two on combustion systems.

Exhaust frame blower repairs. A user offered an in-situ solution for repairing exhaust-frame seals. Operators noticed that exhaust-frame blowers were cycling, primarily on startup; investigation revealed tripping breakers. Root cause was thought to be exhaust-frame seals split and out of position. The OEM suggested a stopgap fix: Try to slide seals back into position and tack-weld.

Plant personnel opted to fix in place and GE Energy mobilized a technical advisor and three millwrights for the job. A piece of the seal was found in the superheater of the heat-recovery steam generator, but there was no third-stage FOD. Service team cut out the support pipe/channel in one section, removed the old seals and installed new ones. Lesson learned: It was difficult to slide seals into position; others performing the same task should consider cutting out the whole pipe.

Major-outage case history. “Be prepared” was the message from this user as he told of significant unplanned issues that occurred during a major inspection of two units. Plan called for a 34-day outage on the first unit to accomplish the following tasks, as well as others: remove rotor and ship to the OEM, perform P-cut modification on R0 compressor blades, conduct eddy-current inspections (ECI) of turbine wheels, replace compressor stator row S17, upgrade pumps and Mark V control systems, change inlet filters, inspect evaporative cooler, overhaul gas control valves, service inlet bleed heat valves. Unexpected was the rotor exceeding runout tolerances and the need to replace the stub shaft.

A crack was found in a dovetail of the second unit and GE Energy supplied a used rotor with 20,000 hours of service. So-called thermal arch binding was identified on compressor stator rows S13-S16 and they were replaced. Miscommunication between the EX2000 excitation system and the Mark V control system caused a forced outage, requiring technicians to change the timing on EX2000/Mark V communications to prevent a recurrence.

Balancing a thermally sensitive rotor. Vibration at full speed/no load increased from 3 to 8 mils on startup and the unit tripped (high vibration). An axial rub was suspected but a borescope examination did not confirm that. Belly bands were installed in the generator and GE recommended changing the thrust bearing.

Turbine roundtable discussion focused on blending, polishing, and peening turbine wheels to mitigate cracking.

Combustion systems

TP failures. Plant was unable to achieve emissions compliance and brought in the OEM to retune the combustion system. Result was positive, but emissions were very close to regulatory limits. A borescope inspection identified heavy cracking on three transition pieces. All TPs were removed and cracking was found on 13 of the 14. A liberated flow sleeve damaged downstream components; first-stage buckets and nozzles were replaced, secondstage buckets recoated.

DLN tuning issues. This user presentation essentially was a primer on the tuning of DLN (dry, low NOx) combustion systems. It summarized DLN operating modes, discussed dynamics monitoring and alarms, etc. Similar subject matter is covered in “Monitoring— and mitigating—combustion dynamics,” p OH-61, 2007 Outage Handbook supplement inserted in the middle of this issue.

A review of TIL updates was scheduled just prior to adjournment of Day Two sessions and ahead of the vendor fair. Productive discussion includes the intent and impact of the OEM’s technical information letters issued since the previous conference. TILs reflect changes, modifications, and inspections suggested for mitigating issues affecting operation, reliability, maintainability, and/or safety.

TILs sent to 7FA owner/operators in the year between the 2005 and 2006 meetings were the following: 1509, 1214, 1456, 1345, 1504, 1397, 1475, 1480, 1518, 1498, 1469, 1519, 1520, 1502, 1521, 1522, 1508, 1523, 1524, 1525, 1496, 1528, 1280, 1345, 1287, 1532, 1533, 1330, 1539, and 1540.

General Electric Day

Day Three began with a review of fleet operating statistics. At the top of the agenda was an update on the OEM’s budding steam-cooled H fleet. A total of three units shipped this summer to the UK and Japan; another to California. The first H machine in the US is scheduled for commercial operation in 2008.

The 7FA fleet had an average reliability over the last year of 99.1%, and a 97.0% availability—this based on ORAP® data compiled by Strategic Power Systems Inc, Charlotte. For background on the Operational Reliability Analysis Program, refer to “Proactive management of GT parts life key to controlling maintenance cost,” 2006 Outage Handbook supplement to the 3Q/2005 issue of the COMBINED CYCLE Journal, available at www.psimedia.info/ccjarchives.htm.

The nearly 800 engines in the fleet had accumulated just under 250,000 fired starts and 8.5-million fired hours at the time of the user-group meeting. Breakdown by operating regime: peaking, 19% of operating units; cycling, 43%; and base load, 37%.

The Power Answer Center received this report card from the OEM: PAC cases continue to increase (36,000 in 2004, 42,000 in 2005, 45,000 projected for 2006); nearly 90% of the responses are provided on-time; nearly 6500 customer surveys reveal 97% customer satisfaction, with more than 40% of the respondents citing the service as “world class” or “exceeded expectations.” Users in attendance asked for faster response and fewer generic answers. Some noted that PAC answers sometimes differed from those provided by the onsite technical advisor.

TILs were a major topic of discussion. Mention was made that alerts and safety were separated among the compliance categories in 2005. If you receive a TIL with Compliance Category “A” it means that failure to comply with the recommendations within the specified operating time—for example, “at first opportunity” (next shutdown)—could result in equipment or facility damage. Category “S” (for safety) means that failure to comply could result in personal injury.

Some of the TILs reviewed during the discussion:

• 1397, addresses cracking that forms along the annulus of the forward combustion casing. Recommendation is to check during routine/annual inspections and submit a PAC case if cracks are found. Weld repair is suggested until casing can be replaced.
• 1532-2, addresses creep of fabricated compressor discharge casings. Recommendation is that when you remove the CDC take clearances/ measurements and submit a PAC case. Support team will provide instructions for reinstalling the CDC and an estimate on creep movement.
• 1533-1, addresses general casing inspection for turbine and CDC. Recommendation is to look for cracks propagating from welds. Options include continued operation, weld repair, or replace casing.
• 1517-2, addresses reliability improvements for the Renk turning gear. Turning gears have experienced a high failure rate after about 150 starts because of fast ramp-ups, over-torquing, and other causes. A soft-start device and other enhancements are described to boost reliability.
• 1528-3, addresses lube-oil varnishing. It discusses the positives associated with electrostatic-type filtration and balanced-charge agglomeration for preventing varnish accumulation. Recent studies support the contention that these technologies are only one part of the varnish-elimination solution. For the latest thinking, refer to “Gas-turbine valve sticking, the plot thickens,” p OH-4, 2007 Outage Handbook supplement inserted in the middle of this issue.

Reliability topics were next on the program. TP wear on extendedinterval pieces, sparkplug reliability, decreasing hot-air leakage from outer crossfire tubes with better seals were included in this part of the discussion.

Compressor update, just before lunch, included the following:

• R0. Fleet experienced five liberations in 2005. Each event was unique regarding cause—such as dovetail fretting. Attendees were referred to TIL 1509. OEM reported that 520 units were inspected and no R0 indications were found; no units were removed from service. P-cut solution was said to provide 10 times the damage tolerance of non P-cut blades; the dovetail undercut eliminates fretting potential by dramatically reducing stress. The fleet leader was said to have about 700 starts and 5000 hours on P-cut blades.
• Clearances were checked on 520 machines. Nine revealed heavy rubs in row 0, four in R1; five had cracks/corner loss in R0, eight in R1. A slight increase in clearance— 0.027 in. for R0, 0.021 in. for R1—to prevent such wear and tear reduced output by about 0.9 MW, heat rate by 7 Btu/kWh.
• Stator shim migration was reported in 42 units. Root-cause analysis: high-cycle fatigue caused by COMBINED CYCLE JOURNAL, Second Quarter 2006 rotating stall during startup (forward stages) and thermal ratcheting (aft stages). There is a risk of downstream damage if shims or parts of shims are released into the flow stream. Currently available OEM solution is to mechanically attach shims to vane segments (first five rows) or vane platforms by bolting. A pinning solution is being developed.
• S15 rocking. Four units reported severe rock in S15 vanes, two units reported liberated S15 vanes. Engineers report the cause is not high-cycle fatigue, but rather a wear-out/looseness issue. Recommendation for severe rock is to replace vane or vanes, install back-side shim, and/or possibly repair the casing. The design solution, to strap four to six stator vanes to increase their wheelbase, already has been incorporated into the FA+e and FB machines.

Turbine update focused on wheel cracking with locations of greatest interest being the dovetail coolingair slot and lockwire tab. Last is more prevalent in the Frame 6. Cooling- slot cracking can be found only by eddy current inspection (ECI); cracks are too tight for identification by dye penetrant. If a crack is found, the wheel must be retired. Base-load units with unpeened wheels and square cooling slots are at greatest risk. In the extreme, cracks can lead to bucket liberation.

Recall that shot-peening overcomes machining-induced residual tensile stress. To illustrate: Cracks have been identified in only four of more than 700 7F machines with shot-peened wheels. By contrast, 15 of the 50 engines with unpeened wheels had cracks. The original square slot design had sharp or adverse edges that lacked surface to shot-peen. New contoured slot significantly reduces stress concentration.

TIL 1539 and 1540 recommend the following: ECI inspection, backcut dovetails in buckets, and blending, polishing, and peening of the cooling slot edge on the wheel. Note that back-cutting is required only on stage-1 and -2 bucket dovetails only; stage 3 is uncooled.

Specifically, ECI unpeened units at every exposure to verify wheel integrity. For turbines with peened wheels and square slots, ECI at next hot-gas-path inspection and install back-cut buckets. For turbines with peened wheels and contoured slots, ECI at next major inspection and install back-cut buckets.

General Electric Day concluded with breakout sessions for controls, accessories, generators, and fuel programs. Controls upgrades—Mark Ve and Mark Vie—highlighted that breakout session. Accessories discussed included online water wash (medications available), exhaust thermocouples (pay attention to connectors when replacing), No. 1 bearing lift hose (borescope inspection recommended to detect damage), and inlet air system (inspection requirements). Flexibility of the 7F for burning LNG, and the backup liquid fuel system, were top topics in the fuel breakout.

Day Four: Auxiliaries, liquid fuel system

Mark V improvements were featured in the auxiliaries session which opened the final day of the meeting. A new overview screen was developed for Cimplicity to facilitate operator review of data. Other enhancements included a (1) flow diagram to assist in training that includes a steamturbine stress-holds page and added vibration information; (2) part-load starts counter to assist in the calculation of factored starts; (3) hydrogen page for the steam turbine to cover safety issues associated with handling this generator coolant; (4) water wash screen to be sure permissives have been satisfied before washing.

The session roundtable discussion touched on a variety of subjects— including fire protection, oil filtering, valves, and inlet bleed heat.

Dual-fuel reliability issues headlined the liquid fuel session. The user group worked with the OEM to increase the reliability of operation on liquid fuel during winter operation. Numerous modifications were made to the standard design to improve reliability by 360% based on starts/ trips. Mods include:

• Re-engineered and heat-traced the problematic liquid fuel/waterinjection purge-header drain system to prevent freeze-up.
• Re-engineered the water injection skid to prevent drain-line freezeup and eliminate leaks.
• Added Swagelok mud daubers to the false-start drain system to prevent freeze-up.
• Replaced flow dividers.
• Redesigned the spool piece for the fuel forwarding skid.
• Installed RTDs to control heaters in the turbine compartment thereby ensuring tighter control of temperature.

Roundtable discussion focused on ways to stop coking in liquid fuel lines (more in “GT fuel-system advancements,” p OH-105, 2007 Outage Handbook supplement inserted in the middle of this issue). ccj