GE Roundtable highlights fall meeting with meaningful discussion of issues/solutions for Frames 7B-FA

South Shore Harbour Resort & Conference Center, League Ci ty, Tex, whi ch ho s t e d CTOTF’s 32nd annual Fall Turbine Forum and trade show, September 16-20, is a superior meeting location: easy access to the airport, close to many gas-turbine-based powerplants, a stone’s throw from Houston’s many repair shops, firstclass accommodations, and good food.

Forget the property’s many amenities, however; not enough time to enjoy them. Conferences planned by the Leadership Committee of the Combustion Turbine Operations Task Force run four days from breakfast at seven until five in the afternoon. Then there are important business events in the evenings. This year they were a welcome reception on Sunday, trade show Monday, and shop tour on Tuesday.

Classroom sessions started on Monday with Chairman John Lovelace, a consulting engineer at Arizona Public Service Co, presiding at the morning Plenary. The Generic Roundtable, with a program addressing non-engine-specific concerns, was attended by all users in the afternoon.

The Siemens and FT8 Users Roundtables ran concurrently from dawn to dusk on Day 2; four half-day roundtables— Legacy, Alstom, LM6000, and ZLD & Environmental—were held on Wednesday; Thursday was reserved for the GE Roundtable. Nearly three dozen prepared presentations were delivered over the four days, in addition to the traditional user-only discussions at each roundtable.

Once the program is planned and speakers selected, the Leadership Committee turns over the job of “making the meeting and vendor fair happen” to Wickey Elmo (wickelmo@ ctotf.org, 704-753-5377), Goose Creek Systems Inc, Indian Trail, NC. The gold stars she gets for her work in the energy industry recently got Elmo recognition from ConventionSouth. She was one of several meeting professionals honored by the magazine in late October for their outstanding professionalism, creativity, and dedication.

Focusing on the Frame 7

The marquee session at the fall meeting was the GE Roundtable, which featured as its theme alternative outage solutions for Frame 7 engines. Roundtable Chairman John Gamble of TVA and Vice Chairman Larry Rose of Dominion Energy Inc, both very knowledgeable on Frame 7 issues, put together a bell-to-bell session that included 10 prepared presentations and a user-only discussion period.

The CTOTF report in this issue is devoted to expanded coverage of the GE Roundtable because of the importance of the 7FA to the gasturbine- based sector of the electric power industry in North America.The majority of large combined cycles installed since the millennium are powered by this frame. Reports on other sessions at the Fall Turbine Forum will appear next issue.

Gamble and Rose had the GE Roundtable program organized just like an outage. The first two speakers scoped out the project. Hans van Esch of Turbine End-user Services Inc, Houston, covered major technical and commercial considerations that required considerable calendar time: component condition assessment and work-scope definition based on those findings, qualification and selection of repair vendors, etc. Next, TVA’s Zach Cowart reviewed milestones in outage planning, beginning with the “kick-off” meeting 26 weeks out.

Following these two “scene-setting” presentations, an insurer offered an industry-wide view of 7FA compressor issues, two users talked about specific challenges identified with the 7FA compressor during inspections and outages, and two non-OEM shops offered proven cost-effective repair solutions for compressors and combustion hardware. GE Energy, Atlanta, then had its opportunity to update the group on fleet experience with the 7FA compressor.

Insuring quality repairs of HGP components

Van Esch (hvanesch@ teservices. us, 281-291-0447) got the day off to a fast start after Gamble’ s br ief o p e n i n g comments. Van Esch, an expert on shop practices and component repairs, summarized in less than an hour his popular three-day training course, “Metallurgical Aspects of Industrial Gas Turbine Component Refurbishment.” He has more than two decades of experience in the development of repair techniques and coatings for industry leaders Sulzer Elbar, Sulzer Hickham, and Chromalloy both here and in Europe.

Van Esch started “at the beginning,” suggesting how users might assess the condition of their GT parts onsite and use this information to prepare meaningful component repair specs and then select the appropriate repair vendor for their particular situation.

Justice could not be done to van Esch’s thorough handbook-style presentation here. A better approach is to read his “Six steps to successful repair of GT components,” a series of four articles written for this periodical and available at www.combinedcyclejournal.com/archives.html. Here are the details:

  • Steps 1 and 2, click 2Q/2005 on web page referenced above, click “Six steps. . .” on the issue cover. Step 1, onsite condition assessment, covers dimensional checks and how and where to take them, visual inspection, and the importance of operating history.
  • Step 2, repair-specification preparation, recommends dividing the repair process (and spec) into logical stages. Reason is that you want the repair facility to report its findings and recommendations after each stage and not proceed with the next tasks before you or your representative approves.
    Van Esch offers a detailed flow chart that suggests dividing the repair process into these four stages: (1) receive parts to be repaired and conduct the initial inspection; (2) disassemble, clean and strip, heat treat, and inspect; (3) repair, heat treat, and inspect; (4) coat, assemble, and inspect.
  • Step 3, click 3Q/2006, click “Selecting the appropriate vendor to refurbish parts for your turbines” in the Outage Handbook. Sample scorecards included in this article offer a methodology for evaluating alternative vendors in terms of experience and reputation, technology and subcontractor experience, and personnel capabilities (engineering and skilled labor) and management systems. Also presented is a handy chart of acronyms used by repair contractors and their meanings.
  • Step 4, click 4Q/2005, click “Shops verify as-received condition of your components with an ‘incoming’ inspection.” First action in Step 4 is initial inspection, which involves taking dimensions in the as-received and assembled condition at the repair facility, as well as conducting a visual inspection and a metallurgical evaluation.
    After the scope of work is confirmed, or modified to reflect the findings of the initial inspection, the component is disassembled from its retaining ring, blocks, etc, and all hardware—such as seals, pins, core plugs, etc—is removed. Key actions required during this stage to assure accurate verification by the repair vendor of component condition include these: visual inspection, metallurgical evaluation, stripping/cleaning and surface inspection, heat treatment, nondestructive examination (NDE), and engineering review. Last determines if the existing scope of work for the refurbishment effort is conducive to achieving the desired results.
  • Steps 5 and 6, click 1Q/2006, click “Verifying repairs, final inspection results.” Step 5, verification during the refurbishment process, starts with preparation for welding, brazing, and coating. Note that before any of these processes can be performed, surface must be prepared to ensure that it is free of cracks, coating, oxidation and corrosion products, and other contaminants. The article offers general guidance on welding, brazing, and coating processes—such as the importance of process quality and repeatability.
  • Step 6, verifying final inspection, focuses on the review of all quality records for conformance to specifications. A checklist of specific items to check is included. NDE and dimensional verifications are necessary to ensure both proper and timely reassembly of the engine as well as its reliable operation during the next duty cycle. A final visual inspection is suggested before packing for shipment to ensure that the parts are clean and not susceptible to handling damage.

Outage planning

Cowart began with a pertinent quote from Benjamin Franklin, “By failing to prepare, you are preparing to fail.” TVA has more than 80 frames in its 5000+ MW arsenal of GT assets so the company has considerable expertise in outage planning that other users may benefit from. Twenty-six weeks out, its milestones are these:

  • Identify the need and type of outage required, including a preliminary scope of work.
  • Develop an outage-team responsibility matrix and assign key personnel. Also, specify tasks that will be assigned to the OEM, third-party field service firms, and in-house personnel.
  • Obtain management approval for both budget and schedule. TVA’s outage experience allows engineers to develop relatively detailed budgets for labor, parts, services, and equipment at this point in the planning process. The sample budget for a 7B compressor overhaul that Cowart presented showed line items down to $1000.
  • Schedule includes start date, duration, and work-shift plans (24 × 7s, 10 × 5s, and planned overtime, if any). Considerable detail is provided, as it is for the budget. A sample 7B HGP inspection schedule displayed on the screen went as deep as the one hour estimated for removing spark plugs.
  • Identify long-lead-time items, including parts and tooling, and issue RFQs (request for quotation). Eighteen weeks out, order longlead- time materials, “lock in” about three-quarters of the work scope at this point, identify spare components that may be required during the outage, and update the schedule. Six weeks later, conduct a planning meeting for supervisors with specific outage assignments and begin workpackage development.

Cowart offered as sample workpackage description the following:

“Inspect the aft end of the compressor and the exit guide vanes. Check for deposits, FOD [foreign object damage], warped casings and cracking. Record findings on GE form ISE/ GT-FF6077-MS7001 Series Compressor. Note: This procedure provides the steps needed for inspection of the No. 2 bearing. If the No. 2 bearing is not to be inspected in the hot-gas inspection being performed. . . .”

Ten weeks from the start of the outage, planning and preparations shift into high gear. Labor and technical direction are scheduled, contracts for capital parts (purchases and repairs) and hardware are firmed up, and orders are placed for tools, trailers, and specialty equipment—generally on a rental basis.

Over the ensuing month engine history is researched. By six weeks out, key participants have reviewed findings/ results from previous outages, known liabilities, operational issues, and lessons learned. Plus, everyone has been updated on correspondence from the OEM and others that recommends specific items to inspect, repair, modify, and/or replace.

About a week before the outage, site/ unit preparation begins. Site mobilization includes delivery and set-up of office and storage trailers; expansion of parking facilities; installation of temporary hookups for plumbing, electrical, potable water, etc, required by work crews; deployment of portable toilets; mark-off of laydown areas for both deliveries and for components/ parts from dismantled equipment; tagging of key electrical, gas, water, steam, wastewater, and other connections to assist contract personnel unfamiliar with the plant, etc.

An insurer’s perspective

FM Global, Johnston, RI, one of the world’s largest commercial and industrial property insurers, has 43 clients that own/operate 170 7FAs on four continents. The company probably is in second or third position behind the OEM in terms of most first-hand knowledge of 7FA compressor issues that have resulted in equipment damage, because of its focus on property loss prevention.

Although the insurer’s engineers have seen a great deal, client confidentiality prevents them from discussing publicly the specifics of their findings. Nonetheless, FM Global’s Brett Dupre, VP staff underwriting, and Terry Cooper, assistant VP and principal engineer (power generation) were valuable participants in the GE Roundtable because of what they said in general terms.

Cooper took the podium first and began with a review of the company’s values, people, processes, etc, thereby enabling attendees to better understand its culture and behaviors. At the top of his list was FM Global’s engineering resources—including scientific research, product testing, and loss experience—which Cooper said is critical to success.

Additionally, more than 10% of its 1500 engineers work with clients in the generation sector of the electric power industry to help them understand what their property threats are and how to better protect their business operations from equipment breakdown, fire, natural disasters and other risks.

Monitoring of fleet events, validation of information received (engineering bulletins, operating standards, web casts, etc) and evaluation of known facts and RCA (root cause analysis) findings, when available, are all part of this process.

Issues identified with the 7FA compressor that FM Global is following, Cooper said, include the following:

  • P-cut R0 cracking.
  • Standard R0 erosion.
  • Third-stage stator cracking.
  • Corrosion in the forward section of the compressor.
  • Stator shim migration.
  • Aft-stator base wear.

Important to the group was the insurer’s recommendations. Simply put, FM Global accepts as technically sound and appropriate the OEM’s corrective actions and risk-mitigation measures for dealing with the F-class forward-compressor distress issues. The company suggested that clients follow GE’s current recommendation of more- frequent inspections, which Cooper said would have minimal impact on the O&M of the units involved. However, he added that any client’s position contrary to the OEM’s recommendations would be evaluated as a possible alternative on a case-by-case basis.

Dupre said all turbines in a fleet are not created equal: Some have higher risk profiles than others. One reason could be that the maintenance and personnel training practices at a given plant are not as rigorous as those of the fleet leaders; another could be a more challenging operating paradigm.

Users profile outages

Presentations by owner/operators add tremendous value to user-group meetings because attendees hear directly from colleagues who can identify specific problems others in the fleet may experience, where and under what circumstances these problems might occur and why, how to identify and quantify damage found, and alternative solutions that warrant consideration.

At the GE Roundtable, two users profiled outages that involved 7FA compressor inspections and repairs: Dave Brunson, plant manager, Redhawk Generating Station, Arizona Public Service Co (APS), located about 50 miles west of Phoenix; and Bill O’Brien, plant manager, Union Power Station, Entegra Power Group LLC, El Dorado, Ark.

Brunson covered in detail the progressive deterioration of rotating and stationary blades in the compressors for both gas turbines (GTs) serving one of Redhawk’s 2 × 1 combined-cycle units. Note that both units are covered by long-term service agreements (LTSAs).

Damage was first identified in January 2004 during a routine borescope inspection, about a year and a half after COD (commercial operating date). Subsequent annual inspections revealed increased damage. Corrective action was deferred from the time of discovery until the first hot-gas-path (HGP) inspection in April 2007.

Brunson showed nearly 50 slides, most with multiple photos, so colleagues would thoroughly understand the extent of damage and exactly what it looked like. Rolled metal and transfer of metal to case was observed for R0 blades in one compressor in 2004; same was identified in the second compressor except one blade in that machine also experienced tip liberation (Fig 1).

Rubbing and/or impact damage also was found in the first compressor in 2004 among rotating blades in rows 4, 8, 10, and 15, plus stator row 6. In the second compressor, such damage was identified in rotating blades for rows 1, 3, 4, 7, 11, 14, and 16, plus stator row 15. An R11 blade with heavy impact damage is shown in Fig 2; heavy tip rub and metal transfer to the case is easy to see for the R1 blade in Fig 3.

Inspection of both compressors the following year showed more pronounced damage in most of these rows, plus new damage in other rows.

Once the upper compressor cases were removed for the spring 2007 HGP outage and the extent of the damage confirmed visually, an emergency PAC case was submitted to the OEM. At a minimum, the OEM recommended replacement of all stator vanes and blending of all rotating blades; exit guide vanes already were part of the outage scope. APS identified the source of the domestic object damage as metal and oxide liberation from blade tips (stator vanes on wheel and rotating blades on case) from excessive friction caused by inadequate clearance. GE rebladed both compressors. Work beyond that originally planned extended the outage of each GT by nominally one month.

O’Brien’s presentation on the HGP inspection experience for Power Block 1 at the 2200-MW Union plant was different from Brunson’s report in several respects: (1) It covered more than just the compressor; (2) the OEM had a relatively small role in the outage; (3) multiple thirdparty contractors were selected to perform specific segments of the workscope based on demonstrated expertise.

The 2 × 1 Power Block 1 is powered by naturalgas- fired 7FA GTs (Mo d e l 7 2 4 1 ) , which began operat ing in 2003.

Controls are Mark V. Original plan was to conduct the first HGP at about 900 factored fired starts (FFS), but the schedule was moved up when a routine borescope inspection identified a damaged first-stage bucket in the Unit A turbine at 739 FFS and 9864 fired hours (Fig 4). The Unit B GT underwent its inspection at 862 FFS and 11,003 fired hours.

Scope of work for the outage, said O’Brien, included the following:

  • Inspections. HGP, compressor with the upper casing removed, GT and steam-turbine generators, heat-recovery steam generator, balance-of-plant equipment.
  • Testing and calibration. Highvoltage electrical system.
  • Replacement. Remove P-cut R0 compressor blades and replace with standard blades.
  • After the upper casing was removed from the machine with the damaged turbine blade, plant personnel found the compressor had started to rust because of the extended shutdown period (Fig 5); however, no permanent damage was in evidence.

DRS-Power Technology Inc , Schenectady, NY, was contracted to pin various stator vanes from the fifth stage back through the exit guide vanes (EGVs)—this to capture the shims and to dampen vane movement. The number of pins (one, two, or three) used per vane depended on the row. For details on this procedure, access www.combinedcyclejournal. com/archives.html, click CTOTF on the issue cover, and scroll to “GE Roundtable. . . .”

The inner barrel was exposed on Unit A’s compressor to install counterbore plugs recommended in the OEM’s TIL (Technical Information Letter) 1315-2R1 (Fig 6). According to that document, the root cause of excessive stator groove wear at row 17 or the EGV row and/ or S17 distress with EGV cracking is air flow disturbance across the counterbore holes in the inner barrel. The phenomenon is said to dissipate as load and/or ambient temperature increases.

O’Brien stressed several important benefits of early compressor inspections with the upper casing removed, including these:

  • Discover installation problems.
  • Tighten loose stator vanes.
  • Repair damage from rubs.
  • Use NDE (nondestructive examination) techniques to identify cracked vanes and blades.

Regarding the first point, O’Brien referred to a condition found in one of the company’s Gila Bend units where a S15 stator vane was installed backwards and had loosened to near liberation (Fig 7); weld repair of the case was required. Fretting wear also was identified in Union’s Block 1 units at S15 (Fig 8), requiring replacement of several vanes and pinning of most others. Damage incurred early in commissioning was found (Fig 9, left) and blended (right).

In his conclusions, O’Brien highlighted the benefits of using qualified alternative vendors:

  • Get quality work at affordable prices.
  • Increases your choice of options.
  • Services are performed with the “customer in mind.”
  • Potential exists for partnering to establish fleet-wide savings.

In addition to work done by DRS, the Union outage benefited from the participation of these alternative suppliers:

  • Pratt & Whitney Power Systems, Windsor, Conn, first- and secondstage buckets and repair of HGP parts.
  • PIC Turbine Services, Marietta, Ga, outage support labor.
  • Wood Group Generator Services, Farmington, NM, generator inspections.
  • Siemens Power Generation, Orlando, first- and second-stage turbine disc inspections.
  • Dashiell Ltd, Houston, highvoltage electrical system maintenance.
  • Allied Power Group, Houston, repair of combustion system components.
  • Wood Group Gas Turbine Services, East Windsor, Conn, fuelnozzle repairs.
  • Power Systems Mfg (PSM) LLC, Jupiter, Fla, fuel-nozzle repairs.

Third-party repair projects

The GE Roundtable at the Fall Turbine Forum focused on the Frame 7. Most of the session, including GE Energy’s presentation, was directed at the 7FA—the compressor, in particular— but many in attendance had earlier models. Chairman Gamble, for example, is responsible for the maintenance of about four dozen 7Bs and 7EAs at TVA.

With that in mind, one non-OEM services provider, ReGENco LLC, Milwaukee, presented on refurbishment of 7B compressor vane segments; National Electric Coil (NEC), Columbus, Ohio, addressed 7EA field unbalance; and Allied Power Group, Houston, o f f e r e d r e p a i r alternatives for 7FA combustion hardware.

Frank Oudkirk, a regional service manager for ReGENc o wi t h over 40 years of experience in the electric power industry, began with an overview of a recent 7B vane refurbishment project, which was initiated by a compressor failure. The owner’s O&M personnel were surprised to discover extensive vane FOD during a scheduled outage. No replacement parts were on hand.

Upper-half casings and lowerhalf ring segments were shipped to ReGENco; many of the lowerhalf segments required destructive removal. Refurbishment included a combination of (1) dressing existing vanes where possible, (2) replacement with purchased used blades as available, and (3) reverse engineering and manufacturing of new vanes to complete the sets. Last were made from bar stock by ReGENco in its highly automated new-parts manufacturing center (Fig 10).

Step 1: Incoming inspection and triage. All vane segments were inventoried and sorted. Visual assessment determined which vanes and rings were salvageable; those were blastcleaned and rechecked using nondestructive examination (NDE) techniques (Fig 11).

Step 2: Sources of used parts were identified. Inspection teams consisting of ReGENco and customer engineers visited sellers to inspect and select vanes and rings (Fig 12). Parts were shipped to ReGENco, cleaned, qualified by NDE inspection, and blended where necessary.

Step 3: Ring-segment and vane accounting. Each row (stage) of vanes in the 7B has six segments for a total of 102. On this project, 50 segments could be reused, 25 were purchased, and 27 were manufactured by ReGENco (Fig 13). The total number of vanes in the machine is 1504; 575 were repaired and reused, 642 were purchased, 287 were made by ReGENco. Details of individual vanes and rings were compiled into a database which was included as part of ReGENco’s report for the customer.

Step 4: Vane segment assembly. Vanes were inserted in the segment slots, shimmed at the back to assure a tight fit, and staked. Next, the arc of each ring segment, and the height of vane tips, were checked against specifications and adjusted if necessary. After the first seven stages were coated to retard corrosion, the upper-half segments were installed in the casing; lower-half segments were shipped to the customer (Fig 14).

NEC’s Bill Moore, PE, director of product line development, is a frequent presenter at user-group meetings. At the Fall Turbine Forum, he focused on 7EA generator rotor vibration issues caused by shorted turns, shorted coils, and tight wedges. Last is of particular concern on this frame because certain series groups of this generator frame size have a single one-piece wedge per rotor slot and its removal and installation can be challenging. The long length of a 7EA generator rotor compounds the difficulty.

One reason CTOTF invited NEC to speak had to do with the success of the company’s Specialized Engineering Solution ™ (SES) in solving the shorted-turn problem. NEC’s experience includes the rewinding of eight 7EA rotors from a single customer over a nominal five-year period (1998-2003). The rewinds were prompted by the OEM’s TIL 1308, which addresses shorted turns caused by the migration of turn insulation (Fig 15).

Moore said that insulation migrates because the adhesive holding it to the copper can break down over time. The SES, he continued, relies on a superior adhesive to hold the insulation in place and employs a specially designed positive locking system to prevent the insulation from migrating out of the end turns. Solution’s success has been demonstrated in more than 10 rotor rewinds over the past several years.

Next, Moore talked about end-winding distortion (Fig 16) and blocking design. He said NEC has encountered severe end-turn distortion on some units. Contributing factors include (1) lack of, or insufficient, slip planes; (2) blocking that was installed incorrectly or moved during operation; (3) thermal instability; and/or (4) a high degree of unit cycling.

Moore described a problem associated with the OEM’s blocking design and what NEC did to eliminate it. The manufacturer’s single-sided Nomex tabs were found torn away on some rotors that NEC worked on. Exposed pop rivets can create galling points that, in turn, can tear retaining- ring insulation. Nomex also can tear out around the pop rivets.

NEC’s blocking design differs in that holes are not pre-drilled in the Nomex. Rather, the Nomex is folded, inserted in a slot in the wedge, drilled, then glued and permanently secured with G-11 dowel pins. In addition, two Nomex tabs are used for each block instead of the OEM’s one.

When an end-winding distortion problem is identified, engineers must pinpoint the cause and customize a solution to prevent a recurrence. Replacing windings in kind is not the answer if, when the unit returns to service, it will operate under the same conditions that contributed to the distortion.

The repair solution, Moore said, depends on the extent of distortion and conductor characteristics, including hardness. Replacement coils, when required, generally are made from CDA (Copper Development Assn) 107, a silver-bearing copper material with a hardness of about 85 Rockwell F. The quality of all brazing is verified by 100% ultrasonic testing and appropriate quality assurance/quality control procedures are in place to assure an appropriate level of dimensional accuracy and shape consistency.

Purpose of rotor slot wedges is to hold slot contents—copper, turn and ground insulation, top fillers, and creepage blocks—during rotation. They are secured radially by dovetail teeth cut into the rotor forging. Moore noted that wedge design is influenced by ventilation design and typically is characterized by a tight fit in the slot. One or both retaining rings must be removed to slide out one or more wedges. Many 7EAs have one-piece, single-dovetail type wedges (Fig 17).

Important to note is that installation (or removal) of any wedge may cause galling between the wedge and the rotor dovetail. This can “lock” the wedge into place, restricting thermal differential expansion between the wedge and forging. Where the wedge/ slot fit was particularly snug during original manufacture, air hammers were used to drive in the wedges, making their removal particularly difficult.

Close attention must be paid to the surface condition of both rotor teeth and wedges. If a small burr or foreign particle is trapped between a rotor tooth and wedge the resultant damage should be dressed out to eliminate stress concentrations conducive to fatigue cracking.

Moore helped users develop a greater appreciation for the idiosyncrasies associated with generator rotor repair when he discussed the relationship between wedge tightness and rotor vibration. Specifically, single-piece wedges can cause rotor vibration issues if they have unequal amounts of tightness after installation. Rotors can become thermally sensitive, he said, if wedges are not properly installed.

For example, if a group of wedges in one area is significantly tighter than the other wedges, there will be unequal amounts of expansion among the wedges and the rotor may bow. If so, vibration will result. One way to reduce the probability of thermal- expansion issues is to use short wedges with small gaps between them.

Moore then presented a case history of a 7EA rotor rewound by NEC that developed a thermal sensitivity/vibration issue after a period of service. Although the unit was no longer under warranty, he said, the company took it back for analysis and correction. First step was to conduct a thermal sensitivity test in NEC’s balance pit.

This was done by running the rotor in both a hot and cold condition and determining the thermal balance vector between them. Vector length, Moore continued, which ideally should be less than 3 mils in length, can be affected by vibration magnitude or by change in phase angle. Tests run hot to cold and vice versa produced the same result: a vector of about 7 mils in length.

NEC engineers suspected that the thermal sensitivity might be caused by a galled wedge that was locked in a slot, or by unequal tightness among the wedges that initiated a rotor bow. Challenge was to find out what the underlying cause of the thermal sensitivity was, and where it was, without tearing apart the rotor.

The company’s service personnel were familiar with Adwel International Ltd’s (Mississauga, Ont) wedge tightness detector, having used the instrument in stator overhauls. Note that Adwel recently was purchased by Iris Power LP, Toronto. It eliminates the human error associated with hand-held tap testing of stator wedges with a hammer. Here’s how it works: A calibrated tapper is placed over a wedge and the tapper activated with the push of a button. The natural frequency of each wedge is calculated and the results are recorded automatically in an accompanying laptop.

The thought NEC engineers had was that this device might also be used to check the tightness of one-piece rotor wedges with the location of too-tight wedges identified for correction. Field tests proved the method worked well in this service (Fig 18). The few wedges identified as being excessively tight were removed, sanded, reinstalled, and retested. Severe galling was found on one of these wedges and in its mating slot.

The location matched that of thermal imbalance point identified during the thermal sensitivity test conducted after the rotor was returned to the shop. Upon completion of work, another thermal sensitivity test was run with good results. The thermal vector between hot and cold conditions was reduced to 2 mils and the phase-angle change was low.

Allied Power Group’s John Yelincic focused on DLN-2.6 combustion hardware refurbishment for the 7FA, 7FA+, and 7FA+e. Attendees knew well the importance of maintaining combustion systems in tip-top condition to assure emissions compliance, and since most had DLN-2.6 systems in their GTs, Yelincic’s presentation was particularly significant.

He identified modes of degradation and repair strategies and techniques for the following HGP components:

  • Combustion caps.
  • Liners (8000 and 12,000 hr).
  • Transition pieces (TPs, 8000 and 12,000 hr).
  • First-stage shroud-block assemblies for the 7FA+e.
  • Compressor stator vanes for the 7FA.

Degradation of combustion caps most often is caused by (1) cracking (Fig 19) and implosion/distortion (Fig 20) and/or (2) fretting of spring seals. Yelincic began by showing photos of damage that users typically encounter. Next, he discussed the challenges associated with developing both a thermal barrier coating (TBC) to reduce the degradation of the effusion plate during a service interval for combustion caps and an application process that would meet stringent specifications for air flow through the approximately 4000 holes in each effusion plate (Fig 21).

This latter effort involved extensive comparison flow testing and sampling of the OEM’s effusion plates and of Allied’s refurbished parts to ensure repeatable results. Yelincic said that two sets of combustion caps refurbished by Allied are in service and meeting expectations.

Evidence of liner degradation includes (1) body cracking, with or without coating-system damage; (2) liner significantly out of round; (3) fretting of spring seal; (4) breaking of “too-cool” welds; (5) burning/oxidizing of aft end; (6) coating loss downstream of PM3 nozzles.

Recall that the DLN-2.6 combustor has six premix burners: five identical nozzles surrounding a smaller center nozzle. The three premix manifolds (PM) are configured such that any number of burners (from one to six) can be operated at any given time. PM1 fuels the center nozzle, PM2 the two outer nozzles located at the crossfire tubes, and PM3 the remaining three outer nozzles.

Yelincic showed a couple of dozen photos to illustrate the points he made—for example, coating loss from service in Fig 22, liner cracking in Fig 23, new double-curve spring-seal assembly (Fig 24), and final customer inspection (Fig 25). Photos such as these are “educational” for illustrating the types of deterioration found and the look and feel of quality refurbished parts.

However, they do not provide the user an appreciation for the complexity of HGP parts refurbishment. That perspective was offered by a couple of other slides Yelincic showed. One was an example of an interactive Excel data sheet used for quality assurance/quality control (QA/QC) purposes to ensure that components don’t leave the shop until they meet customer expectations. The other presented a very detailed scope of work for component refurbishment— including seemingly myriad dimensional checks, welding procedures, QA steps, etc.

TPs. Most common degradation issues associated with transition pieces for the 7FA series of engines, continued Yelincic, are the following: (1) cracking of the aft bracket on the impingement sleeve; (2) bottom side cracking on the impingement sleeve; (3) picture-frame wear; (4) distortion and severe wear and tear of the inner and outer cloth seals between the TP and first-stage nozzle assembly; (5) ring wear on the forward end clamp for the impingement sleeve; (6) improper assembly of the impingement sleeve to the body of the TP.

Then he showed some of the techniques that Allied uses to assure high-quality refurbishment and improve parts life, including:

  • An aft bracket for the impingement sleeve that reduces stresses at the predominant failure location. Also, a forward clamp-ring modification offers a tight fit between the TP and impingement sleeve thereby improving service life (Figs 26, 27).
  • Dimensional qualification fixtures— such as the one that simulates the compressor discharge casing on one end and the firststage nozzle row on the other so TPs can be checked for form, fit, and function prior before returning them to the plant (Fig 28).
  • Special fixturing and weld repair, in some cases, permits Allied to refurbish the cloth seals between the TP and first-stage nozzles.
  • Condition of multiple sets of DLN-2.6 hardware refurbished by Allied in 2004 and recently returned for service after demanding combustion inspection intervals revealed component degradation that met and, in most cases, exceeded customer expectations (Fig 29).

First-stage shroud-block assemblies for the 7FA+e often experience (1) coating loss; (2) heavy rubs from the first-stage turbine buckets; (3) impingement/metering plate detachment; (4) oxidation and burning of substrate material; (5) block-to-block cloth seal loss.

Yelincic said Allied has shipped 20 sets of fully refurbished firststage shroud-block assemblies back to their owners. Engineering of the coating system and dimensional requirements assure a refurbishment quality that meets or exceeds customer expectations, he added.

The company grinds inner shroud blocks to increase the clearances and alleviate the hard tip rubbing and associated coating loss being experienced in the fleet from turbine casing distortion. A fixture that simulates the casing allows fit-up checks of shroud blocks to ensure form, fit, and function.

Yelincic closed by explaining how Allied addresses and refurbishes severe T-fit and butt-to-butt wear on 7FA compressor stator vanes.

OEM presentation

Ross Youmans, manager of product service, and Andy Baxter, GE Energy’s 7FA group leader, participated in the CTOTF’s Fall Turbine Forum with a four-part presentation that included a technical update on 7FA compressor issues, results of rootcause analyses conducted by the company, a general overview of fleet experience, and the OEM’s communications program for keeping customers current on work in progress.

Roundtable Chairman Gamble said the users generally were pleased with GE’s candor in discussing the issues and the work being done to solve problems. Questions put to Youmans and Baxter after their prepared remarks were answered to the degree possible. If a question could not be answered, a reason was given.

Compressor technical issues addressed included R0 cracking, stator-base wear at the aft end of the machine, and corrosion and stator distress at the forward end. Mid-span tip rubs were covered as well; solution here probably is tip grinding. The turbine was mentioned to the extent that disk issues are being addressed and that wholesale replacement of disks is not being considered as a solution based on findings thus far.

The prepared remarks concluded with a review of fleet recommendations for operation of compressor washing systems (online and offline), a checklist of items to run through during compressor maintenance outages, and another checklist for annual inspections and minor outages. ccj