501 F Users Group: Proven value of participation drives record attendance

Chairman Paul Tegen called to order short­ly after lunch on Monday, January 17, the largest 501F Users Group meeting in the his­tory of the organization. It was winter and the Orlando venue is a lure for north­erners at that time of year, but the group had been to the Contemporary Resort in Walt Disney World before and never came close to attract­ing the more than 160 users it did this year. A vendor fair for attendees of the 501F and 501G annual meetings, held in separate rooms except for a couple of joint sessions, had the most exhibitors ever.

Why did so many users attend this year? Answer is simple: The meeting’s repu­tation for being content-rich, plus an opportunity to net­work with both peers and equipment and services pro­viders. Tegen and the steering committee (Sidebar 1) pack what feels like a week’s worth presentations into only three days, and that doesn’t include the evening vendor fair and a few premium social events.

You can’t come to this annual event and not leave with new knowledge that helps improve plant per­formance, O&M practices, and/or safety. Sessions are vibranta mixture of focused user presentations, selected vendor presentations, and open discussion.

1. Steering committee

Chairman: Paul Tegen, chief CT engineer, Cogentrix Energy Inc.
Vice Chairman: Russ Snyder, plant manager, Cleco Evangeline LLC’s Evangeline Power Station.
Gary Giddings, Progress Energy Florida’s Hines Energy Complex.
Ivan Kush, manager, Calpine Corp’s turbine controls group.
Mike Magnan, plant manager, PPL Lower Mt Bethel Energy LLC.
Ray Martens, plant manager, Iberdrola Renewables Inc’s Klamath Cogeneration Plant.

Controls and generators dominated the first after­noon’s program; safety, com­pressors, and the hot gas path (HGP) the second day; Wednesday was “Siemens Day,” with the OEM cover­ing the engine from A to Z; Thursday morning com­pleted unfinished business in the combustion section and in the air-inlet and exhaust sections of the engine. The “rear guard” was invited to attend a special session on steam-turbine issues and tita­nium L-0 blades.

Rotor air coolers

Virtually all mechanical engineers attending the Monday afternoon session (editors included) welcomed Scott McLellan’s presentation on rotor air coolers (RACs). He talked about corrosion, welding, materi­als selection, tube plugging, water treatment, pipe supports, expansion jointsall the topics that can give “mechanicals” an adrena­line rush.

McLellan, plant engi­neer at Arizona Public Service Co’s (APS) West Phoenix Generating Sta­tion got to the podium at around fourfollowing three hours of presenta­tions and discussion on controls and generators.

GT user group meet­ings allocate perhaps three-quarters of their “program” time to the engine proper. Remain­der is used to cover every­thing else. Thus it was interesting that rotor air coolers and torque con­verters were highlight­ed on both the 501F and 501G agendas this year.

RACs are installed in most F- and G-class com­bined cycles to reduce the temperature of compres­sor discharge air used for rotor cooling. They may be air-to-air heat exchangers (fin-fan coolers) or water can be used as the cooling medium. Siemens Energy, Orlando, often refers to its air-to-water exchangers as kettle boilers, cross-town neighbor Mitsubishi Power Systems Americas Inc (MPS) calls them TCA (for turbine cooling air) coolers.

Air-to-water exchang­ers are built to Section 8 of the ASME Boiler & Pressure Vessel Code. Their advantage over fin-fan coolers is that heat is captured by the system, not rejected to atmosphere. Result is a small improvement in plant performance.

Here’s a typical arrange­ment used to recover heat: At Portland General Electric Co’s Port Westward Generating Plant, commissioned in mid 2007, the ver­tical shell-and-tube TCA designed by MPS uses water from the HP feedwater circuit to cool compressor discharge air. Water leaving the TCA cooler is returned to the heat-recov­ery steam generator (HRSG) down­stream of the HP economizer.

McLellan began with a physical description of the kettle boilers inte­grated into the design of West Phoe­nix’s 2 × 1 Combined Cycle Unit 5. They are cylindrical heat exchangers about 60 in. in diameter at the mid section with conical inlet and outlet sections at the ends (Fig 1). Manufac­turer was Struthers Wells, Warren, Pa, which was acquired in 2005 by Thermal Engineering International, Santa Fe Springs, Calif, a unit of Babcock Power Inc.

Water is on the shell side and, when the RACs were installed, air flowed straight through 0.75-in.-diam Type-304L stainless steel tubes from inlet to outlet (single-pass arrange­ment). Tubesheet at the air-inlet end of the RAC is fixed. Expansion is accommodated by a bellows in the shell located just ahead of the outlet tubesheet. Tubes are rolled and weld­ed into both tubesheets. This design resembles, in concept, a boiler for an old steam-powered locomotive except that on trains the hot gas produced by the combustion of coal or wood flowed through the tubes.

No design is risk-free. One risk associated with the relatively small gain in efficiency offered by the RAC is that tube leaks, were they to occur, would allow oxygen to enter boiler water. Note that compressor dis­charge air is at approximately 250 psig and between 600F and 900F.

Another potential problem asso­ciated with tube leaks is that on system shutdown, if you don’t open drains quickly enough, water can enter the rotor cooling circuit and possibly back up into the GT shell and fuel systema condition that would abort firing on the next start unless action is taken prior to startup to ensure all water has been drained from the engine.

McLellan showed a photo of the tubesheet in one of the RACs (Fig 2). Most Type-304L stainless tubes in the upper half of the heat exchanger had been plugged before the plant celebrated its second birthday. He said a combination of welded and poppet-type plugs were used, the lat­ter being quite expensive. McLellan noted that once plugging started, the rate of new tube leaks accelerated and plugging was required every few weeks. Interestingly, despite the loss of significant heat-transfer surface, the RAC still was able to deliver cool­ing air to the GT rotor at the temper­ature and quantity required.

Metallurgical analysis of failed tubes revealed transgranular crack­ing characteristic of high-stress, low-cycle fatigue (Fig 3). Note that the crack identified by the arrow runs right through the metal’s grain struc­ture. Magnification of the etched sample is 200X. In Fig 4, an unetched sample viewed at 64X shows that the cracks start at the tube outside diam­eter (OD) and propagate inward.

Tube failure location, continued McLellan, almost always was at the back side of the inlet tubesheet, as shown in Fig 5. That made sense, he said, because it was where the differ­ence in temperature between the air flowing into the RAC and the 80-psig LP-drum water on the shell side was greatest.

Then McLellan told the group that Combined Cycle Unit 5, although designed for base-load operation, actually was experiencing about 130 starts annually; also that the RACs were retubed twice before the plant was four years old.

Retubing of the first kettle boil­er, in October 2005, was necessary because of metal failure attributed to a low-water condition. Cause: The water-level reference leg was inad­vertently drained, resulting in an incorrect level indication. Corrective action: Caps were installed on the reference legs of both units to prevent accidental draining.

The kettle boiler serving the other GT was retubed the following spring because of tube cracking. Both RACs were first retubed with the original material, Type-304L stainless steel (0.049 in. wall).

Next, a detailed engineering investigation was conducted to iden­tify the root cause of tube cracking. McLellan said three priority issues identified and addressed were the following:

  • Tubes were not able to expand and contract freely.
  • Pitting was found on the inside surface of many tubes.
  • Metal fatigue was induced by thermal cycling and/or vibration.

He added that the plant could not isolate a specific cause of the crack­ing, only that the issues bulleted above probably were contributing factors.

Several actions were suggested to facilitate expansion and contraction of tubes, including the following:

  • Conduct periodic inspections to ensure that all pipe supports remain active (Fig 6).
  • Add a support just before the RAC air outlet flange (Fig 7).
  • Add guides to encourage parallel tube expansion and check guides periodically to confirm free move­ment during startup (expansion) and shutdown (contraction).
  • Upgrade tube material to AL6XN; it has a lower coefficient of expan­sion than Type-304L stainless steel.

Pitting attack inside the tubes was attributed to chlorides in car­ryover from the evaporative cooler. That component was designed to use a 50/50 mixture of potable water and well water, but the quality of the lat­ter is quite poor. These three actions were taken to help the RAC by mini­mizing carryover from the evapora­tive cooler in the GT inlet:

  • Upgraded makeup water to reverse-osmosis quality.
  • Added an automatic blowdown circuit to maintain conductivity below 500 micromhos.
  • Implemented an ongoing inspec­tion program to identify pipe leaks in the evap cooler so they can be fixed before carryover becomes a problem.

The RAC suffering the low-water failure was retubed again, in October 2006, using 0.049-in.-wall AL6XN to improve its resistance to chloride stress corrosion cracking. In May 2007, the other RAC was retubed for the second timein this instance with AL6XN having a wall thickness of 0.065 in.

The heavier AL6XN was only slightly more expensive than the 0.049-wall alternative and, because it was slightly stiffer, it was easier and faster to install; the additional material cost was more than offset by the reduction in installation time. Note that Type-304L tubes for one kettle boiler cost about $30,000, sig­nificantly less than the $100,000 for AL6XN.

Water quality on the shell side of the RAC was improved by increasing the automatic blowdown rate to 2%.

To mitigate fatigue from ther­mal cycling and/or vibration, piping was modified to allow makeup water to enter the RAC at the air-inlet side immediately behind the tubesheet (Fig 8). This encourages circulation of water to reduce overheating as well as the potential for violent boil­ing and steam blanketing.

McLellan said the OEM suggested that rapid boiling promotes vibration that you want to avoid. Steam blan­keting, he continued, would cause tubes at the top of the bundle to run hot and be more susceptible to tube failures than those submerged in wateras confirmed in Fig 2.

Use of AL6XN also helps protect against tube cracking because it has greater strength at higher tempera­tures than the Type-304 family of materials.

Controls, generators

Mike Magnan, plant manager, PPL Lower Mt Bethel Energy LLC, is the user group’s controls “champion.” He was one of several members who worked together to form an I&C focus group in 2004. The team included Ivan Kush, manager of Calpine Corp’s turbine controls group, who was recently elected to the steering committee. Idea behind the initiative was to present to Siemens, in an organized manner, the concerns of owner/operators regarding the TXP control system.

Magnan continues to provide the leadership for what is known today as the TXP Focus Group. It is an especially vital group within the greater 501F user organization given Siemens’ announcement in early 2006 that it would phase out its Simatic S5 and Simadyn productscritical components of the TXPthat October.

It was a busy afternoon for Magnan, who coordi­nated the program until the coffee break at 3 pm. First order of business was an update of TXP Focus Group activities for the greater community of F users. After a pre­sentation on generator partial dis­charge testing by Mike Thawley of Iris Power LP, Houston, Magnan introduced Randy Riggs, who man­ages Powergenics, Midlothian, Va, an OEM alternative for TXP spare parts and card repairs. Riggs’ presentation was followed by a users-only TXP session and that by a joint users/Siemens TXP session. A user pre­sentation on generator rewind con­siderations and an open roundtable on generators led up to McLellan’s update on RACs.

Riggs’ message basically was that users don’t have to worry about being caught short regarding TXP parts and repairs, Powergenics is focused on helping owner/operators reduce operating costs through its:

  • Control system repairs and spares.
  • Power supply repairs and spares.
  • Electronics supply, both new and refurbished.
  • Large inventory of parts.

When repair of control-system cards is required, Riggs continued, Powergenics is equipped to provide quick turnarounds, detailed repair reports, and warranty certificates (for a year or more in most cases). The company also is capable of reverse engineering obsolete cards, thereby allowing users to postpone or avoid expensive system upgrades.

Partial discharge testing. The editors couldn’t catch up with Thaw­ley after his presentation but they did find Bill Moore, PE, National Electric Coil’s director of product line development, a frequent presenter at user-group meetings, and asked him to explain partial discharge and how to prevent it.

Partial discharge, sometimes called corona, is quite common in air-cooled generators. It is caused by a partial voltage breakdown within the generator coil insulation, in gaps between the coil and the stator core, or in the end turns when the coils are in close proximity. Because it is not a complete breakdown of the insula­tion system, it doesn’t cause a full electrical ground. Over time, howev­er, these discharges can “eat” at the insulation, causing its deterioration until a full ground does occur and the unit trips offline.

Evidence of partial-discharge (PD) activity often is visible to the eye, appearing as a white powder dusting the surface of the stator winding (Fig 9). Severe PD damage on the outside surface of a stator coil is shown in Fig 10. This was attributed to a lack of semi-conductive coating on the sur­face of the coil in the slot portion.

Special equipment is needed to detect PD activity. Stator slot cou­plers can be inserted under the stator wedge to monitor magnitude and fre­quency of the discharges. It is impor­tant to trend PD activity over time because different machines have dif­ferent baseline values.

Doubling of PD levels over a peri­od of six months is cause for concern. The machine should be opened up and inspected visually. Special coro­na suppression treatments can be applied to coil surfaces to minimize some types of PD activity. Fig 11 shows special cell- and corner-section corona suppression treatment for a stator coil. Other types of shielding arrangements can be used as well, especially for phase-to-phase dis­charges.

Mitsubishi update

The first day of the 2008 conference ended with a sponsor presentation by Mitsubishi Power Systems Americas Inc (MPS), Lake Mary, Fla. It was the ideal time slot for Dave Walsh, senior VP manufacturing and opera­tions, Bill Newsom, director of sales and marketing, and Scott Cloyd, technical director for turbine engi­neering, to update the combined F and G groups on Mitsubishi’s activi­ties in the Americas. After being hammered with technology issues and solutions for an entire day, users had an opportunity to wind down a notch and learn what one of the fast­est growing companies in the gas-turbine-based generation sector was doing.

Walsh began with per­sonnel and facilities. The company now employs more than 350 full-time staff in the Orlando area and recently spent $200-million to expand and update its Houston service facility and outage resource center.

At the time of the meet­ing, the115,000-ft2 addition to MPS’s Orlando facility for the production of R1 and R2 F- and G-class turbine blades and vanes was nearing com­pletion. When the editors visited the “space-age” manufacturing center in mid July it was in the final stages of commissioning and already making product.

The new blade and vane produc­tion lines are next door to the com­pany’s “one-stop” repair shop, which is equipped to fix virtually anything that goes into a large frame machine. Specialties include repair and recoat­ing of hot-section partsblades, vanes, transition pieces, combustor baskets, etc. Walsh showed a series of photos to illustrate the company’s repair capabilities.

The automated machining centers for making new blades and vanes are adjacent to coating and heat-treating facilities. Investment castings come in the front door and finished blades and vanes exit via the back door to a job site or warehousea continuous manufacturing process that creates 100% of the value added to the castings, according to Bob Provitola, general man­ager, plant services.

New business. Mit­subishi is actively pursuing overhaul and repair busi­ness from owners of West­inghouse and GE F-class units as well as from its growing Americas fleet (25 M501Fs and 27 M501Gs by 2011). The company helps cus­tomers monitor the health of their machines via a 24/365 remote asset intelligence center that warehouses historical data, does real-time trend­ing, performs diagnostic checks, con­ducts root-cause evaluations, and provides combustion-dynamics tun­ing support. Walsh mentioned that through 2007 the company had com­pleted more than 70 combustor, hot-gas-path, and major inspections on machinesmostly F-classmade by other OEMs.

2. Emerson, Mitsubishi form alliance

You can learn a great deal at a user-group vendor fair just by observing and listening. At the joint 501F and W501G exposition last January one found Emerson Process Manage­ment’s Power & Water Solutions divi­sion, Pittsburgh, sharing a booth with Mitsubishi Power Systems Americas Inc, Orlando.

The two companies had recently formed an alliance to provide power generators in the Americas complete turbine solutions and an OEM alter­native for long-term turbine support. The partnership leverages Emerson’s expertise in powerplant automation and Mitsubishi’s experience in turbine design, service, and upgrades.

Makes sense. Recall that before its acquisition by Siemens AG of Ger­many, Westinghouse Electric Corp had an agreement with Mitsubishi concerning design and production of some models of gas and steam tur­bines. At that time Emerson’s Power & Water Solutions was the West­inghouse Process Control Div. Thus some of Mitsubishi’s turbines were equipped with WDPF (for Westing­house Distributed Processing Family) controls.

Siemens passed on the oppor­tunity to buy the Westinghouse Pro­cess Control Div and Emerson saw an opportunity. It has continued to support WDPF control systems over the years, but developed the more advanced Ovation® along the way. As support for WDPF winds down, plants typically are migrating to Ova­tion—including the older Mitsubishi turbines.

Recall that Ovation offers fully coordinated boiler/turbine control, automated startup and shutdown sequencing, specialty turbine inter­face cards, built-in security features, and embedded advanced control applications designed to improve plant reliability and efficiency.

The Emerson/Mitsubishi alliance is targeting many old Westinghouse assets for upgrade—including these gas-turbine models: 251, 501A/AA/B/D5/D5A/F. Many have WDPF con­trols. Siemens wants to convert these engines to its T3000 control system; Emerson to Ovation. Obviously, Mit­subishi believes it can at least match Siemens on any turbine upgrades that might accompany a controls retrofit.

Emerson and Mitsubishi already have at least a couple of joint turbine/controls upgrade projects under their belts. One that comes to mind: The Ovation migration and mechani­cal upgrade at the Termocandalaria Power Plant in Cartagena, Colombia, which included a dual-fuel retrofit and the new ability to burn liquid fuel. Another: The Ovation migration and mechanical upgrades for the San Juan Repowering Project, owned by Puerto Rico’s Prepa.

He also told attendees about a compressor diaphragm upgrade for the W501F fleet that puts the hook-fit wear issue to rest without need for cylinder modification. Details on the company’s strategic alliance with Emerson Process Management’s Power & Water Solutions division, which was announced during the meeting, is in Sidebar 2.

Fleet statistics. By October 2007, Mitsubishi’s global F fleet (50 and 60 Hz) had accumulated 4-million actual operating hours. There were 107 units in service at that time and 37 on order. The Gs had accumulated more than 500,000 actual operating hours by last October; fleet had 26 units operating, 28 on order.

Starting reliability for the M501F3 fleet in the Americas stood at 98.99% last October, the equivalent avail­ability factor at 94.8%. Same metrics for the M501G Americas fleet were 99.65% and 95.96%. Lead F3 unit had 50,200 actual hours; lead G, 34,200.

Walsh closed his presentation with thumbnails of these two recent Americas projects to illustrate the broad scope of Mitsubishi’s after-market services:

  • Termocandelaria, Colombia. Original plant consisted of two W501FC simple-cycle GTs burning natural gas in a DLN (dry, low NOx) combustion system. Project requirement was to change the two units from gas-only to dual-fuel to meet new govern­ment regulations, and to complete the work in less than a year.

The Mitsubishi solution was to convert to water-injected dif­fusion-flame combustors to maxi­mize output and simplify the fuel conversion process. The water- and fuel-treatment systems installed to enable the conversion doubled the physical size of the plant. Use of a duplex oil fuel nozzle elimi­nated the need for a flow divider and atomizing air and minimizes visible emissions during startup.

Project included an Ovation® control system upgrade from WDPF (Westinghouse Distrib­uted Processing Family), which was accomplished by swapping-out processors and work stations. The original Q-line cards for the GT were reused. However, a new work station, processors, and cards were needed for Ovation to control the new water- and fuel-treatment systems.

  • San Juan Repowering Project, Puerto Rico. The two W501FCs arranged in 1 × 1 combined cycles were supposed to be placed in service about eight years ago (the WDPF system was pre Y2K), but they weren’t. Mitsubishi was hired by Prepa (Puerto Rico Elec­tric Power Authority) to take the GTs out of storage in 2006 and refurbish them. This included the removal of some new components that would have caused operation­al problems and their replacement with ones of the latest design.

Like Termocandelaria, San Juan migrated to Ovation from WDPF. Work stations, proces­sors, and control cards all were replaced. New control settings and project-specific logic was co-devel­oped by Mitsubishi and Emerson.

Look for feature articles on both projects in future issues of the COM­BINED CYCLE Journal.

Day Two, the beginning

Plant Manager Ray Martens of Kla­math Cogeneration got the attendees revved up the second day with an interactive discussion on fleet safe­ty issues. Early discussion focused on lube-oil transfer-valve issues that many new­comers were not aware of. Background on the subject is available at www.combined­cyclejournal.com/archives.html, click on 4Q/2005, click 7EA on the cover of the issue and scroll to p 12, “Find any chunks. . . .”

Scaffolding for GT outag­es was another discussion topic. The Siemens solution provides scaffold­ing for one side of the unit and users swapped ideas on how they designed and built scaffolding for the other side of the machine.

Removing ductwork and fans from the top of the package to access the unit during an outage attracted a lot of interest. Users swapped ideas on how to do this work without violating OSHA regulations.

Next, Chairman Tegen led a com­pressor roundtable. Topics of inter­est included diaphragm problems and hook-fit wear. Users discussed it on Tuesday, the OEM included it in the Siemens Day presentations on Wednesday. The editors followed up with others offering solutionssuch as Mitsubishi Power Systems Ameri­cas Inc (Sidebar 3) and Sulzer Hick­ham USA (Sidebar 4).

Exhaust-manifold cracking

You can be sure problems exist when both the user group and the OEM have panel discussions on the same subject at different times during the same meeting. On Tuesday, Vice Chairman Russ Snyder moderated a panel in a user-only session devoted to exhaust cylinder/manifold crack­ing and repair experiences. Three users made short presentations; Q&A was lively.

The following day, five Siemens experts were available to answer questions after Scott Harrell pre­sented on the same subject. Summa­ries of key Siemens presentations are compiled in the last section of this report; complete details are available to qualified 501F owner/operators on the OEM’s Customer Extranet Portal (CEP). If you do not have access to this valuable information resource, and believe you qualify, contact Dawn McCarter at dawn.mccarter@siemens.com for information on how to register.

Cracking in the exhaust cylinder and exhaust manifold sections of the GT (Fig 12) is not difficult to believe when you consider that more than 1100 lb/sec of turbine exhaust gas at about 1100F is tearing through the aft end of the machine at nearly mach speed, and that most of these engines stop and start on a daily basis. Talk about thermal cycles.

Mention was made during both the user and Siemens sessions of bro­ken struts, cracking of components, and general material distress in the exhaust cylinder, and cracking in sev­eral areas of the exhaust manifold.

Given the conditions, users enter outages prepared to make repairs. However, you get the sense while listening to them that they are frus­trated because cracks often occur in the same location during each pro­duction run and have to be repaired every outage. Long-lasting repairs often are an elusive objective.

However, some progress apparent­ly has been made in at least one area by third-party vendor KE-Burgmann USA (KEB), Hebron, Ky. A user on Snyder’s panel focused on cracking of the exhaust manifold just ahead of the aft or downstream flange that bolts to the expansion joint frame (refer again to Fig 12).

According to the speaker, most 501Fs have suffered cracks in this area, which is probably why the room was full and virtually everyone seemed to be paying attention. He added that several units also have found cracking at the opposite end of the manifold, just downstream of the upstream flange.

3. Technology transfer: One solution to the hook-fit wear issue

Compressor diaphragm problems have been a standing topic of discussion at 501D5-D5A and 501F user meetings for years. At the 2004 D5-D5A meeting, representatives of Mitsubishi Power Systems Americas Inc (MPS), Orlando, spoke about the company’s success with robust, assem­bled-type diaphragms in M501 and M701 series machines. These same components, the group was told, also could be used in 501s of Westinghouse manufacture.

Fast forward. Shortly after the 501F Users conference last January, where diaphragm issues were discussed once again, the first set of Mitsubishi assembled-type diaphragms began operating in a W501FD2. Selling someone on the idea of being “first,” no matter how good your pedigree, is particularly difficult in the electric power industry.

The editors visited MPS’s Orlando facilities in mid July to follow up on experience to date with the diaphragm retrofit and also to visit the company’s new and highly automated facilities for producing turbine blades and vanes for large frame engines.

Here’s what they learned: Mitsubishi has devel­oped replacement compressor diaphragms for the W501FC and W501FD machines that resolve the hook-fit wear issue (Fig A) without chang­ing the geometry of the hook-fit groove. Design of the replace­ment diaphragms was derived from the diaphragm used in Mitsubishi F and G machines for the last decade (Fig B). In sum, more than 70 of the large frames have aggregated nearly a million hours of service with the assembled-diaphragm design (table).

Design-validation of the assembled-type diaphragm in the W501FD2 using strain gages and accelerometers verified that the retrofit parts were meeting performance expectations; plus, data closely tracked that gathered by Mitsubishi from tests on its own F-class machines. Inspec­tion of the W501FD2 at 2200 hours showed no significant wear near the horizontal joint, confirming issue resolution.

Mitsubishi traces the success of its solution to a man­agement decision years ago to maintain an industrial-compressor design approach rather than adopting more aggressive airfoil designs used in aero engines. The evolutionary design approach versus a more significant change in technology remains a core part of Mitsubishi product development.

Repairs to address cylinder wear in the hook-fit grooves that support the compressor diaphragms some­times must be performed between 8000 and 12,000 hours of operation for the W501FD and W501G—instead of during a major inspection, which typically is performed around 48,000 hours, or less frequently. The wear results from relative movement between the diaphragm and cyl­inder, predominately near the horizontal joints. Relatively simple fixes—such as wedges, sealants, or thickened outer shrouds—generally have not met expecta­tions, at least according to some users.

The Mitsubishi solution is to increase the thick­ness of the airfoils and change from a welded fabrica­tion to a mechanically fastened diaphragm assembly. The latter provides a more rigid structure with improved mechanical damping. Increasing airfoil thickness for the first three stages reduces compressor efficiency by only about 0.1%—a small penalty for the gain in availability. MPS expects to have diaphragms for rows 4 through 6 ready for shipment in 1Q/2009.

Next, the user described the crack­ing in his unit and how KEBwell known for its large, high-tempera­ture expansion jointsdid the repair. The editors dug into the details with Paul Schubert, the company’s field service manager, who has more than four decades of applicable welding experience. He said weld repairs are a natural extension of KEB’s expan­sion-joint work on the exhaust end of the machine.

The first question that a user might ask: How will I know when cracking occurs? That’s easy: Insula­tion is burned by the hot gas escaping through the crack and it turns white as shown in Fig 13

A second question might be: How concerned should I be if a crack develops? Also easy: very. Safety is an obvious concern. The speaker said the last crack his unit experienced went almost all the way around the exhaust manifold. He showed a film clip of the escaping gas; that got everyone to sit up straight.

Another concern is that any escap­ing gas contains CO and NOx at much higher levels than would be found downstream of the emissions control systems incorporated into the heat-recovery steam generator. According to this user, cracks in the exhaust system can open wide and it may be possible to exceed emissions levels specified in your environmen­tal permit.

Thus, there are one or more reasons to repair a crack as quickly as possible. However, if you’re in the middle of a run when demand is high you might not be able to take the machine out of service for a repair that can take up to a week to do correctly. Unfortunately, no one has yet reported on a quick fix that can get you by in the short term.

Fig 14 shows just how wide some cracks can open up. The photo also illustrates a failed attempt at a quick fix from the outside of the exhaust manifold. A series of steel “tiles” were welded in place to “plug” the large open area while accommodating the expansion and contraction associated with daily starts on the unit. High-temperature refractory was worked into voids, held in place by lath-like clips easier to see in Fig 15, a photo taken on the inside of the manifold section. This “fix” lasted only a few hours.

A majority of plant personnel seem to believe cracking at the aft end of the exhaust manifold can be traced, at least in part, to the use of relative­ly thin plate for the manifold (5/16 in.) together with a massive exhaust flange (more than 2 in.thick).

This flange has 132 bolts to fasten the exhaust manifold to the expan­sion-joint frame and flow liner. The nutshidden by the expansion-joint frameare prone to backing off dur­ing GT operation. When this happens the bolts work their way out of the flange and the loose exhaust flange tries to break off from the exhaust manifold.

The “seal-weld” fix shown in Figs 15 and 16 joins the flow liner to the exhaust flange thereby preventing exhaust gas from leaking out between the flange and expansion joint. Note that the purpose of the flow liner is to direct hot gas over the flange. It is not fastened on the downstream side to accommodate expansion.

What usually happens next is that the stress created by the seal weld initiates a crack about 2 in. upstream. The crack shown in Figs 15 and 16 was from 5/8 to ¾ in. wide and extending more than 300 deg around the exhaust manifold.

The way KEB approaches this repair is to first remove the expan­sion joint (Fig 17). This requires unbolting both the exhaust flange and the flange on the round-to-square transition piece on the downstream side of the expansion joint. Next, the upstream exhaust manifold flange is unbolted from the exhaust cylinder. Then the exhaust manifold is pushed back toward the engine about 8 in. to gain access to the crack on both sides of the cylindrical section.

Preparations complete, a Type-347 stainless steel strap 2 in. wide × 0.375 in. thick is welded over the crack on the inside surface of the manifold (Fig 18). The strap is installed in seg­ments to allow for expansion. Then the crack on the outside surface is ground out and welded as shown in Fig 19. Welding done, a new expansion joint is installed and the joints remade using new fastening hard­-ware with the nuts on the outside this time and bolt heads tack-welded on the inside.

The user speaking about this repair procedure said his units were near the end of their second produc­tion run with no sign of gas leakage. Previously no repair had lasted lon­ger than one production run.

Using advanced CDM analysis to improve reliability

Ben Franklin never operated a gas turbine, but one of the many quota­tions attributed to him, “An ounce of prevention is worth a pound of cure,” certainly applies to the use of com­bustion dynamics monitoring (CDM) to prevent engine damage.

This is especially important for GTs equipped with dry, low NOx (DLN) combustion systems, because of their inherently narrow stability limits. Even at 9-ppm NOx, which is double the emissions rate of the latest combustors installed on large frames, combustion is close to the lean blow-out (LBO) limit of full pre­mix systems.

Relatively minor changes in ambi­ent conditions or in fuel quality, pres­sure, and/or flow, or wear or damage to combustion-system hardware are sufficient to cause combustion insta­bilities that can put your operation on the proverbial “slippery slope.” Harmonics created by these instabili­ties can severely damage liners, tran­sition pieces, crossfire tubes, etc, in little more than a heartbeat.

Got your attention?

Well that’s what Leonard C Angello, man­ager of combus­tion turbine technology at the Electric Power Research Institute, and Dr Tim C Lieuwen, PE, associate profes­sor at Georgia Institute of Technolo­gy’s School of Aerospace Engineering, set out to do at the annual meeting of the 501F and 501G Users Groups last January.

Angello is well known to owner/operators of W501F engines because EPRI has worked with many users in this GT community in the devel­opment of guidelines for hot-section repair, procurement of replacement parts, CDM, and combustion-system tuning, as well as to improve dam­age-tracking to optimize mainte­nance intervals.

Lieuwen is an expert in the field of combustion dynamics. He seems to be a party to virtually every serious discussion on the subject as it relates to the electric power industry.

Angello’s purpose in Orlando was to gather support for follow-on work in the fieldspecifically the develop­ment of an anomaly-detection algo­rithm. Here’s how he explained why an algorithm is necessary: Modern gas turbines cannot “fly blind” and continuous expert human monitoring is not feasible. An algorithm based on knowledge gained over the yearsand one that can be updated as new knowledge is gainedis needed to flag anomalous behavior.

The algorithm must be robust, because false alarms will make it useless, he said. It must be capable of distinguishing among sensor fail­ures, variations in ambient condi­tions, machine aging, and real fail­ures. Finally, it must be trained on what Angello specified as “healthy” dataincluding that taken up to known failures, data collected during normal variations in ambient condi­tions, during sensor failures, etc. He then demonstrated how an algorithm warned of an imminent pilot-nozzle failure.

Participation in virtually any soft­ware development project requires a leap of faith. Angello spent some time showing the users the kinds of information available that would be incorporated into algorithm so they would realistically view the goal as a step on solid ground as opposed to a “leap.” He had charts of pressure pul­sations caused by (1) a flow obstruc­tion in the premixer, (2) cracking of a combustor liner, (3) a transition-piece (TP) failure, etc.

For the TP he went into detail. You had to be impressed with how much is known about the characteristics of impending failures and, at the same time, somewhat disappointed that this knowledge is not yet “packaged” in a manner to help plant personnel make proactive operating decisions.

Consider the following: Low-fre­quency pressure oscillations grow substantially in magnitude as TP failure approaches. There is a large swing in intermediate-frequen­cy oscillation amplitude about 12 hours before failure, then three hours before failure a large increase in amplitude is clearly visible on the chart he showed. More subtle shifts in frequency, not readily discernable by eye but capable of extraction by the “smart” algorithm developed by EPRI, also are evident several days ahead. Such data support the notion that CDM systems provide added value as an indicator of machine health. Ongoing work here is aimed at ensuring these dynamic pressure signatures are unique to TP failures.

Obviously, the more data that can be gathered, the more precise the algorithm will be. One of Angello’s goals is to gather 501F data from as large an operational base as possi­blea broad spectrum of operating hours, power settings, load profiles, ambient conditions, and DLN con­figurations.

Pooling resources for R&D that will benefit all participants is a “win-win” proposition. With the prop­er support, Angello said, the fault-detection algorithm would be ready for implementation at each partici­pating company early next year.

Update. The editors spoke with Angello in early September, just before finalizing pages. He said that since January, EPRI has used oper­ating data from six 501F sites to test and refine the CDM “toolkit,” which consists of a primary algo­rithm search tool, a historical-data search tool for baselining purposes, and a false-alarm filter. The idea of these tools is to compare the dynam­ics signature of each can against neighboring cans and against data collected under similar operating conditions.

One of the things investigators learned during this phase of the proj­ect was that CDM systems at some plants were not working as intended. The tip-off was unusable data. In one or more instances the CDMS was not “on” during a part failure, the hard drive was full, or sensors had failed. Your CDMS must be maintained just like any other system. It can’t be used as a health monitoring tool if it isn’t working properly.

In June, work began on validat­ing the toolkit using the failure-data matrix derived from the six-site data­base. The first field validation dem­onstration began in August. Plan is to add several field validation sites early next year and to release the software for general use by 2010.

Looking for more host sites. Angello said he has three more 501F field-demonstration host sites lined up and is looking for as many as six more, for a total of 10. If you are a 501F owner/operator and want to participate in the CDM program, call or e-mail Leonard Angello today (lan­gello@epri.com, 650-855-7939).

User follow-up discussion. The open user discussion on the combus­tion section of the engine was part of the Thursday morning program. Vice Chairman Snyder led the discussion with help from Kush. A portion of the session concerned high-frequency combustion dynamics. Some users seemed surprised to learn that high-frequency dynamics (above about 1000 Hz) is not new and is not a GT phenomenonit’s a combustion phe­nomenon. But this is to be expected at user-group meetings where an individual’s knowledge of a complex subject such as dynamics may range from virtually nil for industry new­comers to a great deal for veterans.

In a large fleet, high-frequency dynamics may be an issue on one or more engines, but not all. Shared experience was that adjustments to PAG (power augmentation) steam, fogging spray, fuel split, etc, may help control the problem but they won’t eliminate it. Users recalled damage to crossfire tubes and flash­backs.

One participant said that resona­tor basketsyou’ll see mention of them in the Siemens Day section at the end of this reportseemed to help based on experience gained on one engine so retrofitted. However, a downside of their use was an increase in CO emissions, which went to dou­ble digits below 70% of the full-load rating. Users generally considered resonator baskets pricey; budget constraints were pushing a few par­ticipants toward process adjustments with the hope of finding a solution.

There was considerable discussion and many questions on dynamics. One recommendation: When investigating for the presence of high-frequency dynamics, check all 16 cans. Picking a couple of cans at random and running diagnostics was thought to be suffi­cient in the past, but no longer.

Another suggestion: Educate yourself on the subject before you begin trying to solve any dynamics issue in-house. Once you have a basic understanding of the problem you’re dealing with there’s a greater likeli­hood that management can be con­vinced to contract for some analytical help. One user suggested that the Jet Propulsion Laboratory and Los Ala­mos National Laboratory have done plenty of research on the subject and it is accessible via the Web.

Other subjects also were dis­cussed. For example, turndown to low loads is a priority for most users to avoid starts/stops that consume component life. However, there are concerns regarding high CO emis­sions at low loads. Another discussion point had to do with the CEP. Users had grown comfortable with the term “prod mods” and how to access the information on them needed to keep plants up to date. They say there’s not as much information available on the CEP as there was in the past and that some things no longer look like what they did when users first saw them.

Torque-converter maintenance

Torque converters often are taken for granted. Perhaps it’s because the recommended service interval is five years (sooner if there’s a change in operating characteristics). You can forget most anything in five years, certainly in seven.

Bob Wasik, who manages the after-sales market for Voith Turbo Inc, York, Pa, told the 501F users that you might be able to stretch the interval to seven years, depending on the number of operating and turning-gear hours. However, he added, you should check alignment and inspect the coupling and flexible connection annually.

When the torque converter is over­hauled, the starting motor should be as well. Inspect the booster pump for leaks and rebuild as required. Finally, check the balance of package equipment for wear and/or corrosion.

O&M case histories

Owner/operators typically make two kinds of presentations at user-group meetings. One warns colleagues of a possible problem and explains how the speaker’s plant addressed it. The other type explains an O&M philosophy or course of action and asks colleagues in attendance what they think of itessentially making the presentation a valuable self-help clinic.

Mike Voeller, plant manager for Cogentrix Energy Inc’s LSP-White­water LP 1 × 1, 501FC-powered com­bined-cycle cogeneration plant in Wisconsin, made one presentation of each type at the annual meeting.

“Basket and Nozzle Support Housing Failure” was a short, straightforward report on a flashback incident that occurred in September 2007. There was no apparent reason for the repeatable flashback temper­atures experienced in responding to load on the day of occurrence. Equal­ly baffling at the time was that the symptoms disappeared the following day, never to return.

With operation “normal” and a busy run schedule, Voeller postponed a borescope inspection for 10 days. It revealed that the flashback had damaged several swirlers (Fig 20) as well as the oil tips on the nozzle support housing (Fig 21). With the engine running well despite the dam­age, replacement of the nozzle sup­port housing and basket in the No. 8 position were scheduled for a more convenient time, about three weeks after the incident.

The culprit was a component ID plate that had liberated and lodged in one of the swirlers, thereby inter­rupting air flow and causing the flashback (Fig 22). Voeller noted that this particular ID plate had been tack-welded to the outside of the filter bleed pipe, also known as a balance tube (it provides access to the rotor for adding or removing bal­ance shot).

4. Weld repair of compressor hook-fits

Hook-fits for the first six stages of a 501F compressor had expe­rienced severe fretting of contact faces (Fig A). Repair was required to prevent stationary blades from tilting downstream and contacting the rotating blades. Sulzer Hickham USA, LaPorte, Tex, was asked by the owner to propose alternative solu­tions.

The two options suggested were (1) machining the low-pressure compressor case’s hook-fit area to accept bolted inserts that would hold the stationary blades in place, and (2) weld-repairing the hook-fits. The first is a convenient solution once the case has been modified to accept the inserts because the inserts can be replaced in the field if they wear in service. Weld repair is attractive because it does not alter the original design.

Sulzer Hickham designed the inserts and performed an engineer­ing study to determine the impact of the modification on the structure of the case. A finite-element analy­sis confirmed the feasibility of the concept and enabled selection of an insert geometry that would both minimize its impact on the case and simplify the machining opera­tion.

Though the recommendation for the insert solution was well received from a technical standpoint, the owner ultimately opted to weld-repair the case. This is not as simple as it might sound because of the A216 case’s large size (slightly longer than 8 ft, nearly 7 ft wide where its diameter is largest) and considerable weight (about 19 tons).
Restoration of the diaphragm hook-fits was achieved by depositing more than 750 lb of weld wire using the semi-automatic submerged-arc process and then machining to origi­nal dimensions. Weld parameters were carefully monitored to assure top quality work.

Fig B shows the case set-up in a large indicating table (50-ton capac­ity) with a center-of-gravity overhang of 78 inches. The welding operation is shown in Fig C. Note the suspend­ed platform that allows the operator to access the inside of the pre­heated case while it is slow-rolling. Fig D shows in detail the areas of the hook-fits that were weld-repaired.

After welding was complete, several fixtures were fabricated and installed to restrict thermal distortion during stress relief in an oven (Fig E). Next, the case was rough-machined and inspected using magnetic particle examination and ultrasonic techniques. The case was reinspected after final machining of the hook-fits.

The repair was completed in six weeks and saved a casing that would have been unusable other­wise. The repaired casing has been in service for two years; no problems have been reported.

Users questioned rhetorically the logic of tack welding (one spot in each corner of the rectangular plate) any­thing in the flow path where heat, vibration, and other phenomena exist to facilitate liberation. Other users may have experienced ID-plate lib­eration as well and not have known it had the plate fallen to the bottom of the machine instead of becoming entrained in the flow path.

“Starting Reliability Improve­ment Measures” was a short presen­tation that reviewed the reasons why LSP-Whitewater failed to start the relatively few times that it had and Voeller’s philosophy for improving its record. He is a believer in the process of “continual improvement” and was looking for the group to critique his plan, which placed more responsibil­ity on operations personnel but kept a tight rein on the cost of capital improvements.

LSP-Whitewater’s starting reli­ability traditionally has been above the fleet average. For example, in 2005, it was 96.6% (256 of 265 attempts); 2006, 97.7% (252 of 260 attempts, including two aborts dur­ing test fires); 2007, 96.8% (304 of 314 attempts). Voeller said the rea­sons for the aborts in 2007 were the following: four failures to ignite, two aborts caused by gas-valve prob­lems, two attributed to starting-motor problems, one caused by ice/snow in the purge manifold vent, and one attributed to logic issues with the OEM’s flame-detection prod mod and later corrected.

The four failures to ignite, he con­tinued, were traced to the A-stage fuel (gas) throttle valve of the vee-notch butterfly type. Voeller said its seat ring is prone to chipping and when that happens, gas can leak by the valve and cause a fuel-rich condition that prevents ignition on the next start. The two other gas-valve problems noted in the preced­ing paragraph are a separate issue. Starting-motor problems were traced to a thermocouple failure.

Voeller acknowledged an OEM prod mod that includes an upgraded valve and actuator to mitigate the issue he was dealing with, but it was too pricey in his view. Instead, Voeller chose to gather real-time data from each start and review the data when operators identified significant deviations from the normsuch as A-stage fuel flow higher than normal. Plant personnel look for changes in pilot and A-stage fuel flows on a weekly basis. Note: Ambient temperatures can impact data accuracy, but such temperature effects can be corrected with tuning.

A series of slides presented data on weekly fuel flow, real-time data collection, and acceptable blade-path temperatures, as well as the present­er’s notes on tuning moves.

Voeller says the plant has a replace­ment valve and several replacement seats in the storeroom. Changeout of valve seats typically is required twice annually for each engine. Job is a four-hour effort for two staff person­nel. Other users in the room endorsed LSP-Whitewater’s strategy because it was more cost-effective than the OEM upgrade. Those in the group who had implemented the prod mod said it was meeting expectationsspecifically, that no failed starts had been expe­rienced because of valve seat-ring wear.

Siemens Day

Ed Bancalari jump-started the Sie­mens portion of the 501F Users Group meeting with a brief explana­tion of the OEM’s regional service organization and its commitment to customers. He also explained his role as service frame owner, which is to support all regions with dedicated resources. It includes translating customer needs into product strat­egy, developing products to improve customer competitiveness, resolving technical issues, and leading fleet performance programs.

Ron Bauer, director of service programs, followed Bancalari and reviewed the recent reorganization of Siemens into three operating compa­nies: Energy, Industry, and Health­care. Siemens Ener­gy brings together a broad portfolio of busi­nesses extending from primary energy to electric distribution. Bauer concluded with comments concerning personnel resourcesspecifically that Siemens is committed to supporting upcoming outages with the appropri­ate skill sets and staffing levels.

Rich Rogalin, manager of fleet performance and risk assessment for the Americas, presented the oper­ating, outage, and availability fac­tors reported by customers. He also reviewed the four duty cycles that characterize the fleetbase load, intermediate, peaking, and stand­bynoting that a large portion of the SGT6-5000Fs are in “intermedi­ate” service. It is defined as having a service factor between 35% and 75% and averaging between 12.5 and 125 hours per start.

Rogalin said the SGT6-5000F fleet is operating at an average service factor today that is nine percentage points above the comparable figure for 2004. Fleet statistics across the board have improved steadily over the last few years, he added, with the Siemens F fleet continuing to achieve industry standard levels for advanced frames.

Bancalari returned to the podi­um to present an overview of “frame status.” The first slide identified his frame-owner targets, which included a user-satisfaction metric based on customer surveys. It showed that the majority of customers are satisfied with SGT6-5000F performance.

Bancalari said he is committed to driving product life-cycle improve­ments that benefit both customers and Siemens; also to developing and implementing technical-issue respons­es to achieve world-class levels of reliability, availability, and starting reliability. A top priority is the devel­opment of performance and flexibility enhancements to help users improve their competitive positions.

He stressed that key to achieving performance goals is better commu­nication between Siemens and its customers through bimonthly net meetings, user-group conferences, the Customer Extranet Portal (CEP), and other channels. Improving com­munications is a priority.

Major F-frame accomplishments for 2007, Bancalari said, included the following:

  • Increased durability for many hot-gas-path (HGP) parts (Fig 23).
  • Higher frequency of user net meetingsbimonthly at leastwith both morning and afternoon sessions.
  • Extensive part-load testing of an SGT6-5000F field engine as part of a development program aimed at providing users the operating flexibility to run units at less than 50% of base-load power with low emissions.
  • Launch of the SGT6-5000F4 machine.
  • Fleet leader among the SGT6-5000FD3 engines passed 8000 hours of operation.
  • Additional R&D funding to address a broad range of issues of concern to customers, which deliv­ered these benefits:

1. New R1 turbine blades with a modified seal slot geometry to increase tolerance to debris build-up.

2. Favorable inspection results of advanced combustion-system and turbine components participat­ing in a field-validation program. The components are designed for 12,500 EBH/900 ES (equiva­lent base-load hours/equivalent starts).

The compressor was the first section of the SGT6-5000F discussed by the Siemens team. Tom Gordon, engineering technical issues man­ager, reviewed the Generation 2 (Gen2) compressor-diaphragm hook-fit program and presented an evalu­ation of the heavy-strongback (HSB) 8000-EBH inspections. He discussed in detail the status of a gusset indi­cation issue and then summarized the 4000- and 8000-EBH inspections of rows 1-8 as well as the current status of production and the imple­mentation plan for Gen2 compo­nents.

The technical overview of the Gen2 diaphragm hook-fit wear solu­tion included discussion of the HSB for R1-R3, inverted seal boxes with joint keys for R4-R6, and the success­ful SGT6-5000FC diaphragms for R7 and R8.

Gen2 tests were completed on two engines in December 2007 and the final report last February with no additional findings identified. Tests were extremely thorough, requir­ing 148 strain gages on R1-R3 and accelerometers and dynamic pres­sure transducers on R2. Multiple seg­ments were instrumented for all key features on the inner shroud.

Engine test results showed that HSB endcaps effectively eliminated the high-stress region of the inner shrouds on both R1 and R2 segments. That plus weld-geometry improve­ments on R1-R3 produced test results that were well below Siemens’ stress-limit targets.

Analysis of inspection results on HSB engine hardware followed. Dimensional and visual inspections were conducted on all wear surfaces. For R1-R3, the primary loading sur­faces exhibited full contact as intend­ed. Minimal wear on one feature, wit­nessed on an earlier inspection, had not progressed. Findings on R4-R9 met Siemens’ expectationsspecifi­cally, minimal findings on key wear surfaces and no other significant findings throughout the compressor hardware.

Steve Holland, service engineer­ing frame owner, gave a status report on the R7 compressor blade developed to reduce the response from R7 and R8 diaphragms. He also discussed wear of outlet guide vanes (OGVs) on some units and improvements now in development; occasional issues with compressor-blade locking keys; and Technical Advisory (TA) 2007-007, which offers recommendations for inspecting units with high hours on turning gear.

The combustion system over­view by Khalil Abou-Jaoude, combus­tion applications manager, covered the current status of hardware, com­bustion dynamics, and auxiliary sys­tem improvements.

In the hardware portion of his presentation, Abou-Jaoude covered design enhancements to the dual-fuel (DF) pilot nozzles to improve fuel-oil atomization, reduce startup opacity, increase mechanical strength, and pro­tect against damage during handling. Plus, the pilot body was redesigned to improve its natural frequency; struc­tural design of the tip heat-shield was improved as well. The enhanced DF pilot nozzle is installed in five units; fleet release was expected around mid year when Abou-Jaoude addressed the 501F users.

Benefits offered by the advanced-design combustor basket include the following:

  • Extends repair interval to 12,500 EBH/900 ES.
  • Improves component durability.
  • Supports low NOx emissions.

The new basket and advanced combustion-system transitions are installed in three units and meeting or exceeding Siemens’ expectations. Fleet release was planned for mid year, Abou-Jaoude said.

He then conducted a “refresh­er” on high-frequency combustion dynamics (HFD). Recall that the risk of HFD increases for units with lean premixed combustion systems; reduced cooling, or higher flame tem­perature, increases the level of risk. When combustion pressure fluctua­tions correlate to component natural frequencies, HFD can be particularly destructive.

A combustion-dynamics monitor­ing system (CDMS) helps plant per­sonnel operate their facilities within the dynamics and emissions limits for various ambient, fuel, and compo­nent conditions. But to minimize the potential for damage from HFD, fre­quencies as high as 5000 Hz must be monitored. Keep in mind that accept­able pressure fluctuations vary with frequency and measurement location.

If you identify HFD, Abou-Jaoude recommends contacting your local Siemens representative. Combustion tuning, which has helped to mitigate HFD on some units, may be all that’s required. If tuning is unsuccessful, Siemens probably would recommend replacing the affected baskets with ones having a thick thermal barrier coating (TBC) or installing advanced-hardware baskets with resonated liners.

The tuned resonators provide acoustical damping and have been validated for use in low-NOx and other Siemens combustion systems. To date there have been no reported issues associated with the intermix­ing of baskets as outlined above. TA 2008-004 offers advice on HFD miti­gation alternatives.

Greg Perona, marketing man­ager for GT modernizations and upgrades, was next to the podium. He offered a portfolio of mods and upgrades to help users improve plant performance and profitability. The upgrades were based on proven and advanced-technology downloads to address such market drivers as power, efficiency, emissions, reliabili­ty, availability, operational flexibility and life-cycle cost.

Perona said that based on the most recent customer survey, the mods and upgrades available for immediate implementation were aligned with user needs. A checklist to review is provided below. Details on each are available on the CEP.

The following upgrades are intend­ed to provide significant improve­ments in efficiency, power, and operating flexibility:

1. Compressor performance upgrade.

2. Turbine efficiency.

3. Turbine power.

4. Hot restart.

5. Turbine stage-4 option.

6. Outlet temperature corrected (OTC) part-load engine control methodology.

7. Steam power augmentation.

8. Compressor icing alarm.

Note that base-load perfor­mance flexibility and inlet-heating upgrades are available but both are in the first-service validation appli­cation phase.

  • Emissions-control and alterna­tive fuels upgrades include these offerings:

1. Ultra-low NOx.

2. Low-CO startup.

3. Fuel conversion.

Siemens’ LNG upgrade is await­ing an opportunity for first-service application.

  • Interval-extension opportunities are covered in the following ser­vice bulletins:

1. The so-called SB 51009 upgrade, first made available in 2004, is designed to reduce trip factors and increase equivalent operation hours or equivalent starts between inspections.

2. The SB 55004 upgrade extends the interval between combustion inspections to 12,500 EOH or 900 ES. Experience with this upgrade at several customer sites was said to be excellent.

Implementing one or more of the upgrades listed above may not be as simple as it sounds. First, “selling” the idea to management will require a thorough financial analysis. If you make it over that hurdle without stumbling, you’ll have to identify the regulatory reviews and approvals requiredlocal, state, federal.

Questions you’ll need answers to include these: Will a major per­mitting effort be required? Will the upgrade reopen the air permit and require a public notice and/or public hearing? Will the upgrade impact air emissions, will it trigger a BACT (best available control technology) review? And the foregoing only rep­resents the tip of the iceberg.

The bottom line: Most plant per­sonnel don’t have the experience and skill sets to make a “go/no-go” decision regarding upgrades. And with staffing levels the way they are in the gas-turbine-based generation business today, they certainly don’t have the time.

The initiative begins at the deck-plates level, but a solid budget and expert personnel resources are need­ed to carry the idea forward. You may want to speak to the OEM even before approaching management, just to be sure your thinking is sound.

Its engineers can make a pre­liminary evaluation of performance upgrade potential versus the base line. For example, depending on the type of upgrade desired, rel­evant supporting data might be a comparison of existing emissions to the upgraded target given the plant’s expected operating profile, the increase in peak power produc­tion, the reduction in fuel consump­tion, etc.

Other topics targeted by Sie­mens engineers at the 501F Users Group’s annual meeting included turbine section status, exhaust cyl­inder/manifold, Aeropac generator update, the latest information on parts life, mods and upgrades under development, and much more. It’s all easily accessible after you log on to the CEP. ccj