User networks provide valuable guidance for outage planning

The first 10 to 15 years or so of a gas-turbine (GT) user group’s existence usually are focused on resolving outstanding issues with the “new” engine model. When a proactive group of knowledgeable owner/operators sits across the table from the OEM’s engineers, divideand- conquer and other avoidance tactics become ineffective; finding solutions becomes everyone’s goal.

Issues are identified and prioritized, and programs to resolve them are planned and implemented. Periodic web-based conferences and faceto- face meetings are conducted to keep customers apprised of progress and one-by-one the deficiencies are corrected.

If no new issues are identified during the fleet’s first major inspections, users generally can assume that the engine model is truly “fit for duty” and standard O&M practices are adequate to meet expectations. There is no compelling reason to meet regularly with the OEM except to learn about upgrades to boost output, improve performance, reduce emissions, etc.

By this time, there’s sure to be a cadre of qualified third-party vendors to provide inspections, overhauls, repairs, maintenance, replacement parts, etc, at more competitive prices than the engine manufacturer. Thus the primary reason the user group was formed no longer exists.

A user organization must evolve as the engine matures, the steering committee charged with identifying its new roles. The Frame 6 group offers a good case study in this regard. It met annually from 1986 through 2000, but interest waned as early issues with the machine were resolved. Budget cuts adversely impacted user support, as well as participation by GE Energy, Atlanta, and no meetings were held in 2001, 2002, and 2003.

But as the machines aged, new problems/information needs emerged, rekindling interest in a user network. California-based user Foster Wheeler Martinez led the effort, sponsoring a meeting in Las Vegas in 2004 with the goal of revitalizing the group as a self-funded organization. The Frame 6 Users traditionally had been supported by large industrial companies that owned and operated 6B engines—such as Amoco Chemicals, ExxonMobil, BASF, Celanese, etc.

Participants at the 2004 meeting responded positively and a steering committee was formed; Larry Flashberg and Jeff Gillis were elected co-chairmen. To ensure success as a self-funded entity, the group brought Wickey Elmo onboard as conference coordinator. It was a good decision. Annual meetings each June typically attract 80 to 100 owner/operators—about half employed by refineries and chemical plants.

The mission of the organization is to provide members an open forum for dialog and exchange of information to improve O&M practices related to the Frame 6B series of engines and to interface with the manufacturer regarding generic issues emerging beyond the first, second, and third majors.

Content of the group’s 2007 meeting at the Windham Riverwalk Jacksonville, June 11-14, proved the organization’s value to Frame 6 owners. By show of hands at the opening bell, nearly half of the attendees were first-timers. Without this organization, how would owners train new hires and introduce them to a network of like-minded plant O&M personnel always available to offer guidance, access to critical parts, etc, in an emergency?

The conference was a classic information resource for outage planning and management, as well as for dayto- day O&M activities. It included valuable user-only open discussions on each major engine component, plus several presentations by industry experts, specifically:

Three “training seminars” of particular value to users new to the engine and a good refresher for the others:

1. 1. David Brumbaugh, founder of DRB Industries LLC, Broken Arrow, Okla, delivered his popular “Introductory Short Course on Industrial Filter Design and Applicat i ons . ” Brumbaugh (dbrumbaugh@drbindustries. com, 918-286-7176) and brought along a van’s worth of filter and media samples, thereby enabling a live Show ’n Tell type of presentation that gave attendees an opportunity to “feel” what he was talking about. For details, access www.combinedcyclejournal. com/archives.html, click “2Q/2007,” click “501D5/D5A Users” on the cover, scroll to subhead “Have filter, will travel.”
2. 2. “Introduction and Operation of Hydraulic Sequencing Valves for GE Ratchet Gear Applications,” conducted by Larry Mitter, VP engineering support services, Young & Franklin Inc, Liverpool, NY, with help from Andrew Stitler. It would be difficult to believe that anyone knows more than Mitter on this subject. His resume reveals nearly four decades of relevant experience with the OEM’s GT product support group and Y&F.
3. 3. “GT Control and Protection Systems for the GE Frame 6,” conducted by Dave Lucier, Pond and Lucier LLC, Clifton Park, NY. Lucier (dave@ pondlucier.com, 518-371-1770), who managed GE’s Field Engineering Development Center in the late 1970s and early 1980s, provided a primer on fuel and inlet-guide-vane control loops for the Mark IV and V control and protection systems.

Two technology updates to help enable better decision-making at the outage planning stage:

1. 1. “Combustion Turbine Visual NDE Practices Utilizing State-of-the- Art Borescope Technology,” by Principals Rod Shidler and Rick Ginder of Florida-based Advanced Turbine Support Inc. Shidler (ATSRodShidler@ yahoo.com, 352-795-7050) and Ginder reviewed compressor issues known to exist in the fleet, plus the in-situ nondestructive evaluation (NDE) techniques used to identify problems before they result in collateral damage.
2. 2. “Gas Turbine Component Repairs: Why Value Matters,” by Doug Nagy, PE, manager of component repairs, Luburdi Turbine Services Inc, Dundas, Ont, Canada. Nagy (dnagy@ liburdi.com, 905-869-0734), discussed the importance of factoring repair quality into vendor selection criteria. He suggested that the small premium paid for top-quality repairs typically offers the benefit of lower life-cycle costs.

Two component upgrades that users might consider implementing during their next outage:

1. 1. “Starting and Jaw- Clutch Design, Installation, Operation, and Retrofit,” by Morgan Hendry, president, SSS Clutch Company Inc, New Castle, Del. Hendry (engineering@sssclutch.com, 302-322-8080) began by discussing the advantages in operating flexibility offered by the synchronous self-shifting clutch over the jaw clutch—such as its ability to enable a GT restart while the unit is rotating at 500 rpm or less. Then he focused on how to retrofit an SSS clutch, now a standard component on the Frame 6, when an existing jaw clutch no longer satisfies a user’s needs.
2. 2. “Exhaust Plenum Replacement,” a three-part presentation by David Clarida, GE Energy, Atlanta, and two users. One of the users opted for a GE replacement plenum, the other for one supplied by Braden Manufacturing LLC, Tulsa, Okla. Clarida (david.clarida@ge.com, 678-844-5031) reviewed the three generations of Frame 6B plenum assemblies offered over the years. He focused on the advantages of the most recent offering which incorporates a floating liner and several enhancements to the second-generation design. The two users reviewed issues each had with the first-generation plenum design and their replacement/retrofit experiences.

User-only discussions

Before digging into details of the formal presentations, let’s review what was on the minds of users in attendance. Open discussion sessions dedicated to the compressor, combustion system, turbine, engine auxiliaries, generator, and instrumentation and controls are particularly valuable in this regard.

Such user-only discussions often morph into self-help clinics. Delegates learn first-hand problems encountered by other owner/operators who may need help in diagnosis and in identifying a viable solution. It’s rare that someone who needs assistance does not get direction— or an outright solution—from the group. Think of it as free consulting by the industry’s top experts.

Where the issues presented have been resolved, others in attendance benefit from knowing what potential problems to look for, and where, upon returning to their respective plants.

Vital to the success of the open discussions is the participation of “old timers” who have valuable experience to pass on to the “newcomers.” Such knowledge transfer is important to avoid making past mistakes over and over again. The Frame 6 steering committee formally recognizes the contributions made by veterans with its annual John F D Peterson Award, named in honor of the first recipient.

You don’t have to sit in a discussion session for very long before you see the tremendous value of having someone like Charlie Zirkelback, the 2006 award recipient in the room. There’s little about rotating equipment that he didn’t witness first-hand or hear about in 37 years of service at Union Carbide Corp. Zirkelback is now in private practice (Z-MechTech Inc, zirkelcm@cableone. net, 361-552-5252).

Likewise, the 2007 recipient, Dr Roointon Pavri, is revered by users for his knowledge and technical leadership. Those who haven’t yet had the pleasure of working alongside Pavri know he has to be special: When did you ever hear of a user group recognizing an OEM employee with its highest honor? See Sidebar 2 for details.

Demographic data. Flashberg and Gillis polled the users frequently during the meeting to get some sense of how units represented are operated. Quick counts of raised hands certainly is not a scientific method of data gathering, but it is practical in a meeting setting. Here is some of the information collected from the nearly 70 participants:

• Majority of units are in base-load service; about a dozen peakers, half that number operate intermittently. Approximately half of the Frame 6s serve in combined cycles, another third in cogen applications.
• Fuel. Natural-gas only is most common; about a dozen gas/oil; one each for oil only, hydrogen, process off-gas.
• Emissions control. About one third of the units have dry low- NOx (DLN) combustion systems, another third inject steam, about a dozen inject water.
• Units installed. Most sites have one Frame 6, about a dozen have two units, half a dozen have three; one site has nine machines.
• Operating hours. Fleet leader has more than 185,000 hours, three units have passed 175,000, four are beyond 150,000, about a half dozen are between 100,000 and 150,000 hours, a dozen between 50,000 and 100,000, approximately a half dozen are under 50,000 hours.

The OEM’s presenters offered that there are about 1050 Frame 6s worldwide. Interestingly, the production of 7F machines just passed the 1000 mark and there are about 1100 7EAs around the world. Only about 15% of the Frame 6s are in the US and the majority is equipped with DLN-1 combustion systems.

Compressor. First topic of discussion focused on inlet guide vanes (IGVs). One user reminded the group to watch for IGV rubs on the inner barrel, this to prevent blade cracking. Two others in attendance suggested the change-out of IGV actuators at every hot-gas-path inspection or major.

The successful bellmouth recoating case history presented last year came up for discussion again. Newcomers were advised that you have to spend time at the repair facility during critical steps in the process. Surface preparation is particularly important to success. Keep in mind that the coating selected determines the specifics of the surface-prep step. Also, that you do not necessarily have to go to white metal to be successful.

Flashberg then reported that one “regular” at Frame 6 meetings was not in attendance because he was “struggling” through a second major on two engines. However, the dedicated user had just e-mailed notes on some findings he thought important to the group.

Regarding the compressor section, he found damage in one unit caused by nuts, bolts, and bits of wire left lodged between the frame and elbow body following repair/replacement of the inlet screen during a previous outage. The obvious lesson: A thorough crawl-through inspection of your machine is necessary after each overhaul. Look for tools, rags, and anything else that should not be there.

The floor discussion continued with thoughts on fogging and wet compression. One user said the fogging nozzles at his plant are located just downstream of the trash screen; system has 15 nozzles. After two years of operation, some erosion is evident on the first couple of compressor stages.

Regarding wet compression, six users indicated by show of hands that they had systems installed. Most indicated that they gained about a megawatt when their system was in operation. One said his plant, at a Southern California coastal location, picked up 2.5 MW. The wet compression system installed at that facility is located behind the evap coolers and operates 4500 hours annually; up to 25 gpm of water is delivered to the spray nozzles at 1350 psig.

A few more comments on the subject: (1) wet compression lowers NOx, increases CO; (2) a problem with one system’s plunger pump was noted with the suggestion that others pay close attention to detail on maintenance to avoid same; (3) pressure regulators require an annual overhaul to replace springs, O-rings, etc, which deteriorate in demineralized- water service; (4) buildup of moisture in the cups of the lower IGVs is another potential problem to watch.

Shim migration, a condition familiar to owners of 7EA and 7FA machines, has been identified in some Frame 6Bs. A user who experienced this at his plant said some problematic shims were removed, others clipped off to prevent a flowpath disturbance. For background on the subject, access the 1Q/2007 issue at www.combinedcyclejournal. com/archives.html, click “CTOTF” on the cover, and scroll to the “The back end” subhead on p 83.

When to change-out compressor blades is a perennial question. A couple of users said their blades had operated for more than 125,000 service hours and looked fine. They asked, “Why change them? What’s the fatigue life?”

A rotating machinery consultant in the room offered that if the blades were going to fail in fatigue, that would occur within 10-million cycles; otherwise, “they may run forever.” From this comment, one could infer that, like fine wine, some compressor blades get better with age.

No user volunteered that he had changed out stator blades that weren’t damaged. Interestingly, one reported replacing the rotor at 109,000 hours (no details given) and retaining the existing stator blades. Attendees agreed that they had not observed any age-related changes to stator blades.

There was a brief exchange on compressor-blade coatings that focused on their value for performance enhancement rather than for protection against erosion/corrosion.

Combustion system. Not much discussion on combustion. A couple of attendees offered their experience in burning refinery off-gas with methane and hydrogen. One said anyone contemplating this should check the olefin content; it tends to foul fuel nozzles.

Another reported burning natural gas “right from the field” that contained 15% CO2. This unit was equipped with a DLN-1 combustion system that was meeting expectations. Yet another was firing gas containing 70% hydrogen, but with a diffusion burner. There were no problems to report.

A user forced to change from baseload to intermediate-load service asked the group if the change would adversely impact his DLN-1 combustion system. One volunteered a “no.” But that seemed to initiate a flood of discussion on the performance impacts of various operating philosophies and equipment upgrades. One snippet: NOx emissions doubled on a unit following replacement of 12-hole first-stage buckets with 16-hole buckets. Engine had to be tuned and the firing temperature decreased from 2020F to 1985F to bring emissions back into compliance.

Spurious trips attributed to faulty flame-scanner operation came next. A user reported that he had installed a purge-air line—consisting of 1-in.- diam tubing and a needle valve— from the compressor to the scanners and that solved the problem.

The Frame 6 user working on the second major reported in after finding cracks at the bases of some secondary fuel nozzles during the inspection. Weld defects may have been the cause, he said, but a formal analysis had yet to be done.

Transition pieces. He then offered a positive report on replacement transition pieces (TPs). His units had the “old style” TPs and inner and outer floating seals; one floating seal had fallen off each unit. Plant purchased upgraded Nimonic 263 TPs for the two units and new inner and outer floating seals. “Fit up like a glove,” he said.

Group discussion on the topic ensued after Flashberg delivered the message. A user asked, “When do you change TPs?” Consensus answer: When parts get so deformed or worn that it costs more to fix than replace. TPs with a thermal barrier coating (TBC) on the inside don’t necessarily last longer, but they’re easier to fix, offered one party to the discussion. “Picture frames shrink over time, so you have to pay attention to sealing, said another.

A black-start question was raised just before the turbine discussion began. An attendee wanted to know how many in the group were so equipped. NERC (North American Electric Reliability Council) regulations require black-start capable units to be on the organization’s “black start” list.

By show of hands, four Frame 6B plants represented at the meeting are capable of black starts. Use of a diesel/generator is typical to ensure this capability. One combined-cycle unit also has a diesel-powered boilerfeed pump. Another has a GT exhaust bypass damper that it is permitted to operate in a black-start emergency.

Turbine. Flashberg opened the turbine discussion with a caution from the user-group member involved in the second major. It had to do with machining associated with the first-stage nozzle ring assembly and nozzle support ring. Work on one unit was conducted by a third-party services provider; the OEM handled the other unit. Fit-up of the independent’s work was perfect while rework was required on the OEM’s effort.

Point was not made to knock the OEM but to make others aware that no matter what a service provider’s qualifications, mistakes can occur. Users should redouble their efforts to check all work done offsite and be sure it meets specifications before shipment back to the plant—this to save both time and money.

Philosophy on first-stage bucket overhaul and replacement followed. At least one attendee representing a starts-based unit suggested 24,000 hours and replace. A couple of others were of the opinion that operating for 24k, refurbishing, operating for another 24,000 hours, and then scrapping was a good plan. Another said refurbish as necessary until you reach 96,000 hours and then scrap.

Obviously, type of service, fuel, and other factors influence such decisions. But the takeaway from this portion of the discussion was that users should keep an open mind regarding refurbish/replace decisions and let metallurgical examination be the determining factor.

Someone asked the group for opinions on repair of casing cracks; he had a couple in the lower combustionchamber casing. Group suggestion was metal stitching; welding of a ductile-iron casing was not an option. But mention was made that stitching, while effective, is a time-intensive repair process.

Generator. The generator discussion was highlighted by findings from the second major in progress. Inspection revealed oil in one of the generators. It was traced to plugged lines on bearing caps. Suggestion was that after you replace the filters, snake out the lines.

Pressure t ransmitter s wer e installed on bearing return lines during the outage; signal is transmitted to the DCS. An out-of-spec condition is identified by audible alarm. Reason this was done: It took several days for operators to notice the sheared coupling on an evacuator pump.

Lube-oil practice, intervals. A couple of users suggested a poll of the attendees regarding lube-oil replacement practice and outage intervals. Most users pump out and clean their lubeoil sumps during a major. Only one did it more frequently; about a quarter of the group less frequently, such as when the additive package is depleted. Discussion continued on varnish issues, lube-oil testing, and how best to control oil coolers. One user warned of a problem in using high-pressure water to clean finfan coolers: It blew off the fins on his cooler.

The group reported by show of hands that most users were conducting combustor inspections (CIs) on 12,000-hr intervals; one was still at 8000 hr, two at 24,000 hr, and one at 30,000. Most popular interval for a hot-gas-path (HGP) inspection was 24,000 hours; a couple of users were doing theirs at 30,000. Regarding major inspections, most were working on a 48,000-hr interval, one was at 60,000. Most interesting was the user who said “infinity.” At his plant, they lift the lid at 48,000 and look over the machine but don’t pull the rotor. If no problems in evidence, replace the upper casing half and continue operating.

OEM highlights

GE Energy brought a solutions team that included six speakers to review modifications and upgrades of value to owner/operators and to address specific concerns of attendees. The three-hour session was held the morning of Day Two; a tour of the company’s Jacksonville repair facility, conducted on the afternoon of Day Three, was the last event on the program.

The presentation on performance enhancements for the Frame 6 described how the company leveraged technology from its aircraft engine operation and from advanced frames to deliver uprates and upgrades of value to the customer. The Frame 6 has evolved dramatically since its introduction in 1978 with an 1850F turbine inlet temperature; the latest commercial model operates at 2084F.

Uprat ing to 2084F requi res upgraded first-stage nozzles with chordal hinge, the latest first-stage buckets (directionally solidified advanced alloy, increased cooling), honeycomb seals, etc. Other key points made:

• High-pressure packing brush seal cannot be retrofitted into an old inner barrel when the latter is out of round—which is typical. Machining costs to accommodate the brush seal are close to the cost of a new inner barrel with the brush seal.
• Redesigned 17th-stage stator reduces structure sensitivity and prevents blade failure. It is suggested for sites with low ambient temperatures and for units with water or steam injection or with modulating IGVs (for heat-recovery applications).
• Tilted control curve allows you to fire hotter on hot days when you need the power (increases the maintenance factor) and cooler on cooler days when you don’t (decreases the maintenance factor). Net change in maintenance factor is zero.
• Implementation of the Extendor™ “kit” permits extending combustion inspection intervals to up to 24,000 hours.
• Development of the DLN1+, which is designed for 5-ppm NOx, continues.

Compressors. One of the OEM’s compressor experts addressed two areas of particular concern to Frame 6 users in his presentation:

• R1 blade cracking. Fewer than a dozen incidences have been reported throughout the fleet— all outside the US. Crack starts below the platform and proceeds upward. Worst case: the airfoil releases. Crack may be visible before liberation occurs; visual inspections are recommended after every 25 starts.
  Root cause: Fretting emanating from cracks in a high stress area. Investigators said failed blades were made by a non-OEM supplier. A new technical information letter (TIL) was said to be in preparation at the time of the meeting.
  GE redesigned the R1 blades several years ago; the new blades have been in commercial production since 2003. They feature an undercut platform that prevents fretting. Recommendation was for users to change out blades in the first row.
• Shim migration. Attendees were referred to TIL 1562. Shim migration in the Frame 6 is similar to that experienced by many Frame 7EA and 7FA owners and covered previously in the COMBINED CYCLE Journal. GE recommended that users consider its procedure for permanently attaching shims to the bases of adjacent stator vanes.

The ensuing hot-gas-path presentation focused on bucket tip clearances and rotor life management. Attendees were advised to avoid rubs at low speed because they are conducive to seal rail loss as well as bucket shroud damage—the latter caused by bucket material deposited into shrouds. Such damage has occurred on stages 2 and 3 (Sidebar 3). Recommendation was for owner/operators, or a contractor, to take detailed clearance measurements on multiple shrouds. Correction may involve casing jacking and redoweling to make it round again.

Rotor life management is a subject of much debate in the industry today and of tremendous interest to Frame 6 users because several owners have passed the 175,000-hr mark. Some users believe that the OEMs have arbitrarily decided on rotor lifetime limits and will not be satisfied until comprehensive reports that include the manufacturers’ assumptions are made available. They say the OEMs have told them their decisions have been made for “safety’s sake.” Such words, these vocal users continue, may very well be a scare tactic and not really about safety at all.

The GE presentation made mention of a safety TIL (one requiring immediate attention) that will be released later this year; it is expected to suggest lifetime limits. A rotor-life limiter of great concern on starts-based machines is low-cyclefatigue cracking/fracture. Embrittlement over time may reduce fracture toughness. Typical lifetime limits mentioned are 200,000 factored hours—including turning-gear operation— and 5000 starts.

A “remaining life” evaluation involves rotor disassembly and inspection. Current thinking is that an hours-based rotor in good condition after 200,000 factored hours of service can be certified by the manufacturer for another 50,000 hours. The rotor is rebuilt and returned. There would be no extension for a starts-based machine—5000 factored starts is end of life.

Controls. Perhaps the most important topic to users in this portion of the GE session was support for ageing control systems. Many owners still have Mark IV and V controls and parts availability and service for those systems will never be better than they are today. Suggestion for users with Mark IVe (enhanced) was to upgrade to the VIe; ongoing support (repair only) for the IVe will be available only for five more years.

The Mark V was introduced in 1991 and production stopped at the end of 2004. Replacement parts will be available for seven years beyond that, exchange/repair/replace for two more years, and the final time to buy what’s left in inventory or to exchange/repair/replace up until a year later. Retrofit options for the V are the Ve, VI, or VIe.

A concern of users was noted regarding exhaust and wheel thermocouples. The part material was changed from stainless steel to Inconel; the connector type was changed as well. Users say the new sheaths are slightly larger, making them difficult to replace.

Training: Control, protection systems

Dave Lucier’s (Pond and Lucier LLC, PAL Engineering) one-hour seminar on GE Energy control and protection systems was a valuable introductory course for newcomers to the Frame 6 and a good refresher for others. It began by answering these two questions:

What must be controlled on GE gas turbines? Answer: turbine shaft speed and rate of change of speed (shaft acceleration), exhaust temperature and rate of temperature change, air flow through the main axial-flow compressor.

Five minutes into Lucier’s “class” you realize that you’re not going to pass this “course” unless you learn GE control terminology. For example, Lucier first hits you with TNH. Not even someone with a PhD in engineering would know what that is. The bottom line: Talking about control systems with a knowledgeable troubleshooter is impossible unless you memorize the “code” this sector of the industry uses to communicate. Lucier says the acronyms are valuable in that they abbreviate the conversation. He calls it the “GE lingo.”

So, for the controlled variables above, TNH is the speed (N) of the high-pressure (H) turbine (T) on twoshaft machines or the speed of the only shaft on single-shaft machines. Units are revolutions per minute (rpm) or percent of rated speed. Acceleration is in rpm/sec or % speed/ sec. Percent speed extends from 0% (0 rpm) to 100% (5100 rpm for the Frame 6). Extending the nomenclature to a two-shaft turbine, TNL would be the speed of the low-pressure turbine (not applicable to the Frame 6).

TTX, the average exhaust temperature, is in units of degrees Fahrenheit. The rate of change of exhaust temperature is deg F/sec. GE sometimes uses TTXM; the M stands for arithmetic mean—that is, the average of all active thermocouples in the turbine exhaust.

Air flow is in pounds of air per hour (flow is not actually measured). The variable inlet guide vanes (IGVs) control air flow during startup, operation, and shutdown without actually measuring and displaying the value.

What devices do the controlling?

Answer: In liquid-fuel operation, it is a bypass valve (BPV). In gas-fuel operation, there are two valves piped in series: a stop speed/ratio valve (SRV) and a gas control valve (GCV). Finally, the variable IGVs provide the air-flow control necessary to protect the compressor against surge and to maintain the desired exhaustgas temperature during part-load operation.

Many Frame 6Bs in service, continued Lucier, are equipped with the Mark IV (1982-1989) or Mark V (1989-2000) control systems. Recall that the GE presenter for the controls portion of the OEM session said these systems were good candidates for upgrade/replacement— particularly when applications change or the turbines are moved to new locations.

Lucier flashed a picture of the Mark IV control panel on the screen (Fig 1). The Mark IV was GE’s first venture into two-out-of-three voting enabled by triple modular redundancy (TMR) design. Note the R (top), S (middle), and T (bottom) controllers on the door. Comparing the Mark IV operator panel in Fig 2 with the Mark V screen in Fig 3 illustrates how rapidly control systems evolved in only a few years.

Control principles

About half of Lucier’s presentation focused on control principles. He began with an overview on the subject, the key point being the OEM’s minimum-value-select (MVS) philosophy. In simple terms that means the control sub-system “calling for” the least amount of fuel will be in command. Goal is to produce power with the least amount of energy possible while always being able to protect the turbine by reducing fuel low should an adverse condition arise.

Before digging into liquid-fuel, gas-fuel, and IGV control principles, Lucier explained how the MVS logic worked, discussed TMR circuitry, and reviewed temperature and pressure profile diagrams for base-load operation. The graphs of temperature and pressure were included on a diagram of the engine to give users a visual sense of machine thermodynamics. Also of particular value to the firsttimers were GT startup curves that plotted speed, exhaust temperature, IGV angle, and fuel-stroke reference (FSR, the signal representing 0 to 100% of the fuel valve stroke) from the starting of engine auxiliaries through full-speed/no-load (FSNL).

Liquid-fuel control principles. Fuel flow is proportional to bypass valve position and fuel pump speed. In mathematical terms, FFN (flow divider speed) = f (%FSR) × (%TNH). The following are important to note:

• Fuel pump (Fig 4) is driven by the turbine through the accessory gearbox.
• Bypass valve position (Fig 5) is adjusted by the Speedtronic control panel. Its job is to subtract fuel flow that you don’t want going to the combustor.
• Flow-divider (Fig 6) speed is proportional to fuel flow through the device.

Nomenclature and components identified, Lucier conducted a short clinic to show attendees how to calibrate the liquid-fuel flow rate, running through calculations of flow rate for light-off, FSNL, and base load.

Gas-fuel control principles. Lucier began this portion of the tutorial with a simplified flow diagram that identified the three gas pressures of importance: P1, the supply pressure, measured just ahead of the gas stop speed/ratio valve (SRV); P2, the pressure in the fuel line just ahead of the gas control valve (GCV) and downstream of the SRV (Fig 7); P3, the pressure in the ring manifold serving all combustors, is downstream of the GCV. Important to note is that P2 is the pressure controlled to assure the correct fuel pressure to the GCV. Lucier concluded this lesson, as he did for the liquid-fuel portion, by showing the users how to calibrate the speed/ratio valve.

IGV control principles. As mentioned earlier, the purpose of the IGVs is to control the flow of air entering the compressor. Lucier ran through a couple of control loops to show how the system works and then moved on to protection systems. IGV operation and maintenance, including a “how to” on setting IGV angle, is another presentation in PAL’s arsenal.

Charlie Pond, Lucier’s partner in the firm, recently conducted an IGV clinic at another user-group meeting. Access to that presentation is via the CTOTF’s free Internet Bulletin Board Communication Service (http://forums.ctotf.org). After logging on to the bulletin board, scroll to the COMBINED CYCLE Journal Forum for the topic entitled “PAL’s Guide to Variable Inlet Guide Vanes.” Alternatively, you can access the presentation directly at http://www.ctotf.org/forum/CCJ/3Q2007/ VIGVPAL.pdf.

Protection systems. Lucier began by identifying the four primary protections and typical settings. They are:

• Overspeed. Setting on the electronic overspeed generally is 110% TNH; mechanical, 112% TNH.
• High exhaust temperature. Nominal setting is around 1000F.
• Loss of flame.
• Vibration. Limits usually are set at about 1 in./sec and 5 mils peakto- peak.

Secondary protections include those listed below, and others as required or as requested by the owner/operator:

• Low lube-oil pressure.
• High lube-oil header temperature.
• Low hydraulic pressure.
• Generator lock-outs.

Lucier next noted that protection systems are configured as energize to run, de-energize to trip. He then discussed GE’s so-called “redundancy by association” concept. What this means is that when operating on liquid fuel and the stop valve closes, the bypass valve opens to enable full recirculation— thereby diverting fuel that would go to the combustors. For gas fuel, when the stop valve closes, the gas control valve also closes immediately. On a unit trip, IGVs immediately move toward the closed position— this to prevent compressor surge.

PAL offers comprehensive O&M seminars on a regular basis. Ondemand onsite training also is an option. Visit www.pondlucier.com for details.

Technology updates: Borescope inspection, GT component repairs

The presentation by Rod Shidler and Rick Ginder (Advanced Turbine Support Inc, ATS) was of particular value to first-timers unfamiliar with the Frame 6B and/or with the benefits of remote visual inspection using a borescope. Shidler opened by reviewing how borescoping can benefit the owner/operator. Examples included the following:

• Conducting inspections recommended by the OEM without having to remove the upper casing half.
• Trending of engine condition.
• Planning for maintenance and outages.
• Troubleshooting.
• Verifying engine cleanliness and the absence of foreign materials (tools, rags, parts, etc) after an outage.

Proper conduct of timely OEMrecommended inspections, continued Shidler, requires that technicians be familiar with both the specific engine model and with the vendor’s requirements. For GE machines, this demands working knowledge of any applicable technical information letters (so-called TILs). He stressed thorough documentation of findings to ensure correct engineering disposition.

Baseline inspections are particularly important, Shidler said, because they document the condition of the unit as received by the owner from the manufacturer. Items identified during this inspection normally are repaired under warranty or are covered by the warranty in excess of the standard one-year provided with most new equipment. Documentation developed during the baseline inspection also lets users know whether subsequent damage found during annual inspections was present since start up or if a review of operating procedures is in order.

Periodic inspections—annual is typical—allow owner/operators to chart unit condition over time and to identify any damage caused by rubs, foreign objects in the gas stream, corrosion pitting, deposits, cracks, coating loss, and component wear, movement, or loss.

General condition assessment of the 6B would include inspection of compressor, combustion system, turbine, and possibly the generator. One focus of the compressor inspection: inlet guide vanes and the first-stage rotating blades. These are the components most susceptible to foreign object damage, erosion/corrosion, etc. Inspection results are factored into the outage plan.

An experienced borescope team has a solid value proposition in troubleshooting. It often can identify the cause or causes of such things as (1) high vibration, (2) high (or low) exhaust-temperature spread, (3) NOx compliance issues, and (4) unit trips.

Regarding the last, the team generally can identify damage associated with a full-load trip if any has occurred. Some things past inspections have revealed include combustor damage attributed to a flash-back, rubs or bent blades and/or vanes resulting from compressor surge, and rubs and seal and tip damage in the turbine.

Perhaps the most instructive part of the ATS presentation was the 6B photo collection presented by Shidler. Most pictures were taken remotely so they offered attendees a primer both on the types of damage experienced by the fleet and on how to “read” borescope photos. Figs 11-16 are of compressor damage, Figs 17 and 18 reflect findings during combustor inspections, Figs 19-23 show turbine damage, and the remaining two issues identified with the generator.

GT component repairs: Why value matters

Doug Nagy’s (Liburdi Turbine Services Inc, LTS) presentation was particularly valuable for managers charged with making repair/replace decisions on hot-gas path (HGP) components and with selecting an appropriate shop when repair is the route selected. It also was an eye-opener for newcomers. The meticulous Nagy indoctrinated them on the importance of rigorous due diligence in evaluating repair shops, of proper coating selection and application, and of uncompromising quality.

Many attendees had to return home convinced that formal training in basic metallurgy, HGP repair technologies, and quality control was a prerequisite for advancement in the plant O&M hierarchy. To that end, a popular introductory course on GT component metallurgy and refurbishment is offered periodically by the ASME’s International Gas Turbine Institute, Atlanta. It typically is co-located with a major industry meeting, such as the society’s annual Turbo-Expo.

“Quality” is a word you hear from virtually every salesperson when evaluating a product or service. But what does it really mean? Nagy offered a simplistic, meaningful definition: Meets all requirements. “High quality” is a common term, he continued, but the modifier “high” is irrelevant: The product or service is either “quality,” and meets all requirements, or it is not quality because it does not meet all requirements.

Quality typically is controlled in component repair by the use of specifications for materials, repair processes and limits, dimensions, future serviceability, delivery schedule, and cost. The primary goal of your spec should be to ensure that the repair is reliable, with minimum risk of service problems, and that it will behave similar to the original new part during the next service interval.

Examples of poor quality, he continued, include the following:

• Welding in inappropriate areas—that is, repair limits are exceeded.
• Welding with an inferior alloy—for example, one that does not meet strength and/or hardness specs.
• Coating material is inappropriate for preventing oxidation/corrosion under the specified service conditions.
• Coating application process is out of spec—for example, coating is too thick or too thin.
• Critical dimensions are not fully restored.
• All defects are not identified and/or corrected.

An obvious question: Why does poor quality happen and where should users focus their attention when evaluating suppliers and work in progress? Nagy offered these comments after cautioning “caveat emptor”:

• Power generation is an unregulated industry. Hence, no industry standards for repairs to land based engines exist as they do for flight engines. Each vendor, in effect, has its own standards.
• Shop backlog and final negotiated price can adversely impact vendor decisions regarding the use of cost- and/or time-saving methods/procedures that may not benefit the GT owner. Similarly, not allowing sufficient lead time for repairs before outage dates encourages poor vendor decisions.
• Depth of knowledge and experience can vary widely among alternative vendors and over time in a given vendor’s shop. This includes both professional (metallurgists, engineers, etc) and skilled craft (welders, machinists, etc) positions. Personnel turnover is something every owner should evaluate in the due diligence process.
• Careful review of vendor documentation regarding quality assurance is particularly important. An owner also should conduct a facility walk-through to confirm that written procedures are indeed part of the shop culture.
• Audits to assure that best-available technologies are integrated into repair processes and that shop personnel are qualified to those technologies are at least equal in importance to the preceding points.

The penalty a user pays for poor quality can be significant if engine availability is adversely impacted when power demand is high. Early removal of repaired parts that did not meet expectations is bad enough, but if the parts fail in service and there is collateral damage, a machine could be out of service for weeks.

Sometimes off-spec repairs may prevent parts from being repaired again, at the end of the next service run. This means new parts will have to be purchased sooner than planned and component life-cycle cost will increase. Likewise, a vendor lacking in knowledge, experience, and the latest equipment may have a lower yield of repaired parts than a top shop. Purchase of new parts to complete a set can increase the cost of the total project beyond that expected.

Once “quality” is under control, users should consider component repair “value” instead of “cost,” said Nagy. Value, in his view, includes the following:

• A high yield of repaired parts.
• Repairs that permit future repairability—multiple service intervals—of parts.
• Operational risk reduction by the “resetting” of design safety margins.
• “Upgrade” repairs that make components “better than new” by correcting weaknesses in the original design.

Some examples offered as “value” repairs are these:

• Shrouded-blade weld repair, an upgrade that extends component life (Fig 26). To illustrate: Conventional weld repair of shroud edges for RB211 (Rolls-Royce) high pressure turbine (HPT) blades after 13,000 hours of operation are at left; upgraded extended-life repairs using advanced filler metals (right) seem almost new after 24,000 hours. Blades at the right cost only about 20% more to repair than those at the left.
• First-stage bucket rejuvenation and internal coating to promote life extension (Fig 27). The 7EA first-stage bucket pictured has 92,000 hours of service and will be repaired for yet another service cycle. After this third cycle, the bucket will have doubled its expected service life. Nagy calculated the cost saving attributed to “value” repairs on buckets for this customer’s four-unit fleet at more than $5 million over six years.
• Nozzle repairs using Liburdi’s patented high-strength powder-metallurgy (LPM) repair process results in repair joints that are stronger than the original cobalt castings, thereby promoting longer life.
• Second- and third-stage repairs, upgrades that extend component life.

Nagy then ran through some “back-of-the-envelope” calculations for Frame 6B bucket “value” rejuvenation compared to conventional repairs. Rejuvenation typically ranges from 15% to 25% of the price of new parts. Three repairs would get a user two full service intervals for—at most—75% of the cost of replacement buckets. Keep in mind that the two service intervals also would include two conventional “strip and recoat” repairs at 10% to 15% of the “new” price. Lastly, there is the cost of new bucket purchase after one or two recoat repairs.

The bottom line: The “effective” cost of achieving 100,000 service hours with “value” rejuvenations would be about half the cost of conventional repairs and associated parts replacement.

Heat treatment, like quality, means different things to different people and must be clearly defined in specifications. Nagy said that a conventional repair may be heat treated with a “partial solution”—that is, only a fraction of the alloy’s creep strength is restored. Usually the upper limit of such restoration is around 50% of the original creep strength. Thus partial-solution heat treatments are of questionable benefit and even the best processes may not recover sufficient strength on a second repair to assure problem-free operation through the next cycle.

By contrast, a “full-solution” heat treatment using a hot isostatic process (HIP) fully rejuvenates alloy creep strength. You pay a premium for this procedure because furnace time is longer than for a partial treatment and there is the added HIP cost. Metallurgically speaking, proper full solution rejuvenation would “reset” component microstructure to a virtually as-new condition (Fig 28).

A significant requirement of rejuvenation is the shop’s ability to remove the internal coatings to allow full-solution heat treatment, then to replace the coatings. Note that internal coatings are sensitive to damage if excessively heat-treated. Nagy claimed that Liburdi was one of very few companies in the industry with long-term success in removing and reapplying internal coatings.

In rare cases, he continued, an alloy may not respond to heat treatment and components must be scrapped. This typically is associated with old alloys manufactured to marginal quality standards. Users can avoid such surprises through proper qualification testing of representative alloy samples before and after heat treatment. Your repair specifications should be written around final alloy creep strength and proper precipitate microstructure as opposed to simple certification of times and temperatures.

Regarding internal coatings, Nagy had this to say: Use of internal coatings depends on the manufacturer’s capabilities and design requirements. Frame 6B first-stage buckets are coated internally, but second stage buckets are not. Originally, he continued, nozzles were uncoated to facilitate weld repair. But today many users are coating R1 nozzles and the latest nickel-alloy R2 nozzles also are coated with a simple aluminide system (Fig 29). Finally, Nagy reflected, internal coatings are expensive; if the designer specified one, respect its value.

Upgrades: Clutches, exhaust plenums

Morgan Hendry (SSS Clutch Company Inc, SSS) is no ordinary company president. He still gets his hands dirty and he remains one of the industry’s most knowledgeable people when it comes to the firm’s primary product—a synchronous self-shifting clutch. Hendry has spent more than three decades—virtually his entire professional career—designing, manufacturing, installing, and maintaining SSS clutches, which was quite evident from his presentation before the Frame 6 Users Group.

Hendry was invited to participate in the organization’s annual meeting because swapping out jaw clutches on 6Bs with SSS clutches can contribute to the bottom lines of many plants by improving their operational flexibility.

First some background: Jaw clutches were a standard issue on starting and turning gear drives for GTs produced by GE Energy until about the mid 1990s (Figs 30, 31). More than 7500 had been commissioned worldwide by that time. The SSS became standard hardware on the 6B in the latter half of the 1990s.

The jaw clutch essentially became a liability for many owners as the merchant power business ramped up. Primary reason: Following a unit trip, this type of clutch cannot be reengaged to restart the GT until the rotor coasts down to 0 rpm (something that may never happen during a windstorm). Thus if a startup were aborted for any reason—loss of flame, vibration trip, high exhaust temperature, etc—it would be about 15 minutes before a restart could be attempted. Such operational inflexibility could get expensive during periods of peak demand.

The SSS clutch, by contrast, can be re-engaged once the rotor speed slows to 500 rpm—or about five minutes after a trip at rated speed. Further, engagement is automatic (there is no servo control system) and the robust SSS requires minimal maintenance.

Other shortcomings of the jaw clutch may include the following:

• Won’t slide on the torque-converter shaft splines because of wear, dirt, rust, etc.
• Does not properly engage because of wear, or only partially engages teeth, thereby damaging the clutch on startup (Fig 32). The clutch also may fail to engage when the actuation pistons are leaking (not holding) or have sheared off because of misalignment. Obviously, if the clutch does not engage, the start is aborted.
• Limit switch is grounded (open), or has moved out of position because of high vibration or clutch hangup, and the GT will not start.

If a jaw clutch is damaged, the OEM recommends replacing the entire component, rather than just the sliding half, to ensure proper engagement and operation and to facilitate installation. The SSS should be considered as an alternative to a replacement in kind, despite its higher cost, Hendry said, because of its operational flexibility and robustness. It can contribute positively to the bottom line and save on maintenance costs as well. More than 15% of the 6B fleet currently is equipped with SSS clutches.

The SSS clutch should be purchased through GE, Hendry advised, so the shaft interfaces can be made compatible with the specific turbine.

Remainder of the presentation focused on how the SSS clutch works (access this information at www.sssclutch.com), programmable alterations required for the various generations of GE control systems from Mark I through Mark VI, and installation details.

How to replace an exhaust plenum

Five participants in the Jacksonville meeting said they had to replace exhaust plenums on about a dozen units because of metal deterioration. Two of these users shared their experiences with the group, showing dozens of before/after photos in the process.

The editors were told that there are only two suppliers of replacement plenums for the Frame 6B: GE Energy Services and Braden Manufacturing LLC, Tulsa. David Clarida (david.clarida@ge.com, 678-687-5194), GE’s CHROEM™ product line leader, presented on his company’s replacement inlet and exhaust systems; Jeff Trost (jtrost@braden.com, 918-274-2454), Braden’s general manager for global retrofit and aftermarket projects, was available at the vendor fair to answer questions.

It seems relatively easy to know when an exhaust plenum must be replaced. A couple of indicators: Peeling paint and warped exhaust compartment doors; metal temperature of the exhaust-plenum enclosure approaching the exhaust temperature; load-gear compartment too hot to go near, let alone go in.

Industry experience suggests that when the marginally designed first generation exhaust plenums have been in service for about a decade, safety and other considerations generally dictate replacement. To illustrate: The doors on the exhaust plenum compartment at one plant were removed because they warped badly and could no longer be closed (Fig 33). Plenum skin temperature was estimated at 800F. Excessive heat also had damaged the doors on the other unit at the site (Fig 34).

The second user presenting illustrated other problems resulting from leakage of hot exhaust gas. In Fig 35A, it’s difficult to see the hole in the plenum wall (circle) because of maintenance lighting inside the exhaust plenum. Turn off the lights and daylight is clearly visible (Fig 35B).

Note that the hole is adjacent to the covered expansion joint that connects the plenum to the heat-recovery steam generator (HRSG). Hot gas coming through that hole “cooked” the junction box for the exhaust thermocouple leads mounted on the outside of the plenum (Fig 35C). A steel plate was welded to the plenum to plug the hole.

Another experience: All insulation from under the plenum went missing and radiant heat caused the No. 2 turbine bearing to move. It moved just enough to jam the rotor, so operators couldn’t restart the unit immediately after a trip. They had to wait about a day until the unit cooled down enough to free up the rotor.

First-generation exhaust plenums are well known for their ability to keep welders busy. Fig 36 shows several patches in the “insulation pan” that serves as both the aft wall of the plenum and the forward wall of the load-gear compartment. To simplify explanation, think of an insulation pan as a fabricated metal container containing insulation that is held against the outer shell with through bolts. Designers intended the pans to grow thermally and seal against each other. First-generation pans typically were 3 to 4 in. deep and stuffed with mineral wool.

Normal thermal stresses “worked” the sheet metal and, over time, it cracked in places where expansion/contraction was constrained. Insulation was sucked through the cracks; in this instance it traveled into the HRSG and contributed to plugging of finned tubes.

Thermal breakdown of the binders used in mineral-wool manufacture facilitated aspiration of the insulation. The result was a pan without insulation in some areas, thereby creating hot spots that accelerated wear and tear. Hot spots in the aft wall insulation pan caused excessive temperatures in the load-gear compartment; hot spots in other pans contributed to high exhaust-plenum skin temperatures. “Normal” pan maintenance consisted of cutting out the section metal containing the crack, restuffing the pan with insulation, and then welding in a patch.

One user described the load-gear compartment as an enclosed space formed by the aft exhaust-plenum insulation pan described above on one side and the generator shield on the other. An insulated roof, back wall, and front access door complete the sound-proof room. Such a tight space is difficult to maintain at a personnel-accessible temperature under the best conditions, impossible when insulation is missing from the plenum’s aft-wall pan. At this location, compartment temperature exceeded 250F on occasion. Such high ambient temperatures contributed to varnish formation on the gearing, which reduces the 5100-rpm output of the 6B GT to the 3600 rpm required by the generator.

6B exhaust system arrangement.

Some 6B users are relatively unfamiliar with their exhaust systems because the focus of maintenance activities is in the money end of the plant (engine and generator); relatively little time is spent tending sheet metal. Same is true for users at some Frame 5 and 7 installations, which have similar exhaust systems.

Clarida discussed the OEM’s latest design for the third-generation 6B plenum assembly (Fig 37) and compared it to earlier generations. He characterized the first-generation plenums as those with insulation pans; second generation, floating liners; and the current third generation as those with both floating liners and additional enhancements.

Clarida described floating liners as individual solid stainless-steel components supported by stand-offs welded to the outer shell. Liners are designed to grow (thermally) independent of each other and create a continuous “floating seal” that protects the outer shell from the hot gases inside the plenum assembly. Fig 38 (left) shows the floating liner for the exhaust plenum floor; at right is the first-generation insulation pan in the same location.

One of the issues with the original pan design was so-called “hot flanges” (Fig 39). Explanation: Thermal growth of a pan’s exposed surfaces does not match that of the outer perimeter, thereby causing distortion along that perimeter. Gaps occur between pans, allowing hot exhaust gases to contact the outer shell; distortion and cracking result.

Floating liners have “cold flanges” (Fig 40), which are characterized by field-installed wrapped insulation pillows and liner plates. This design creates a continuously “sealed” thermal insulation barrier conducive to a cooler interface.

Typical skin-temperature reductions achieved by third-generation exhaust plenums, compared to those of the first generation, are dramatic. GE collected data shown in Fig 41.

Similar skin-temperature reductions are reported by Braden for its retrofit plenum (Fig 42) compared to the OEM’s first-generation design. Some of this information can be found at www.braden.com.

Perhaps a better source for design detail and performance data is “Retrofit Solutions for Exhaust Systems: Plenums, Wings, Cowls & Expansion Joints,” which was made available to 6B users attending the vendor fair in Jacksonville. The publication attributes a portion of the company’s success to a new insulation design. It replaced 3 in. of mineral wool with 4.5 in. of more efficient ceramic-fiber insulation compressed to 4 in. A new liner design prevents the possibility of aspiration.

Exhaust section arrangement. Vital to conducting a successful exhaust plenum replacement is knowledge of how the equipment impacted is arranged.

In Fig 43, with the old exhaust plenum removed, the exhaust diffuser is at the left, and the welder is leaning against the exhaust frame. A close-up of the lower half of the exhaust frame is in Fig 44 and the top in Fig 45. Note the “pipe-within a-pipe” design. The inner cylinder holds the No. 2 turbine bearing in place; exhaust gas passes through the annulus between the inner and outer cylinders and flows to the diffuser. Gas exits through the diffuser and is directed by the exhaust plenum to the HRSG.

The turbine shaft passes through the No. 2 bearing and bolts to the coupling on the load gear, which is under the protective tarp in Fig 46. The two large pipes in the middle of the photo provide cooling air to the load tunnel (Fig 47). In that picture, the No. 2 bearing is hidden by the shaft coupling. Aft wall of the exhaust plenum is at the back of the tunnel. Note that an insulation pan surrounds the load tunnel to protect it against the high temperature exhaust gas.

Managing replacement of the exhaust plenum can be challenging. Careful planning is particularly important if you want things to go smoothly, said one of the presenting users. Thoroughly interview the competing vendors. Have each visit the plant for a walk-through inspection to identify any issues that might impact project schedule and cost. Ask for a plan describing how the project will be conducted, special requirements, etc. Get references and call them. Don’t repeat the mistakes of the past; learn from the experiences of others.

Personnel. Inquire as to the qualifications of supervisory personnel and verify credentials with references. If the retrofit manager has never done a similar project, dig into related experience. After a background check, if you don’t have complete confidence and trust in the person proposed, do not accept him (or her). It probably will be difficult to find craft labor with specific exhaust-plenum retrofit experience, so it is imperative for the supervisor to have it. This does not mean that you can forget craft qualifications. Check them, too, particularly the welders.

Materiel. One presenter said the new plenum was received in multiple truck shipments and in many pieces—several requiring a crane to offload. This was inconvenient given all the other things going on during a major outage. Message here is to prepare appropriate laydown space, expect several deliveries, and schedule crane time.

Both users stressed several times the value of knowledgeable supervision and labor. One said the availability of suitable craft labor was the biggest challenge. The two general foremen assigned to the project were excellent and once proper labor was available the retrofit was completed within a week and there were essentially no follow-up issues.

Results. The user opting for the non-OEM exhaust-plenum retrofit said the project went smoothly and met expectations. Skin temperature of the replacement plenum now is less than 250F (Fig 48). The other user reported that there were still unresolved issues on his project.