501F User’s Group covers much more than the engine itself

Selecting appropriate dates and venues for usergr oup me e t ings is one of the most difficult challenges facing the industry’s steering committees.Program topics generally are obvious to veteran plant personnel, but when and where to conduct the annual conference and vendor fair to assure maximum participation for the most productive interchange of ideas is not so clear-cut.

User-group meetings for the gasturbine- based sector of the electric generation industry typically are held in the spring and in the fall.Problem with scheduling more than two dozen conferences—including those sponsored by independent users groups as well by the gasturbine (GT), heat-recovery steam generator (HRSG), and steam-turbine (ST) manufacturers (OEMs)—in just two seasons is that overlaps are bound to occur. The editors know of three such overlaps this year, and there may be more.

The 501F User’s Group, chaired by Cogentrix Energy’s Paul Tegen, deserves a salute for recognizing this and breaking the mold by conducting its 2 006 annual meeting in mid January in Orlando. The timing must have been “perfect”: More than 100 users attended. And the location certainly rated a “10” because it allowed maximum participation by technical experts from OEM Siemens Power Generation Inc.

The programs for model-specific user groups focus on engine issues, as you would expect. Siemens experts spent one day of the three allocated for the 501F meeting to review fleet performance, update the group on technology enhancements, and provide details and experience on product modifications to improve performance and reduce emissions. Users also conducted closed-door sessions moderated by steering-committee members to address their specific experiences with controls issues, combustors, compressors, and the turbine section.

In addition, this year’s 501F meeting featured several detailed presentations of interest to virtually all GT owners. They were:

  • New EPA rules governing the calibration of flowmeters in the fuel system.
  • Operation and maintenance of station batteries.
  • Disaster preparedness and recovery, based on lessons learned by Gulf Coast facilities during and after last year’s devastating hurricanes.
  • Life evaluation of air inlet filters.
  • Impact on engine performance of operating on liquefied natural gas (LNG).

Results of tests conducted at Dynegy Generation’s Calcasieu Power Project, presented by Tony Denton, plant supervisor, was the subject of a feature article in the 1Q/2006 issue of the COMBINED CYCLE Journal (CCJ), p 58.

“Burning LNG may require mods or upgrades to turbines designed for pipeline gas” can be accessed at www.psimedia.info/ccjarchives.htm.

Know all you should about fuel-gas flowmeter certification, calibration?

Perhaps you did.

Just when you were feeling comfortable and thought you knew all you had to know about recalibrating and recertifying fuel-gas flowmeters, EPA’s rules and reporting requirements for continuous emission monitoring systems (CEMS) changed— effective 1Q/2006.

1. Steering committee

Chairman: Paul Tegen, chief CT engineer, Cogentrix Energy Inc.
Bill Wimperis, director of engineering and construction, Constellation Generation Group.
Raymond Martens, plant manager, Klamath (Ore) Cogeneration Plant.
Gary Giddings, O&M superintendent, Seminole Electric Co-op’s Payne Creek Generating Station.
Matt Kaleyta, CT manager, Dynegy Generation Fleet Operations.
Mike Magnan, plant manager, PPL Lower Mt Bethel Energy LLC.
Russ Snyder, plant manager, Cleco Evangeline LLC’s Evangeline Power Station.

Historical information on flowmeter certification now is required as part of CEMS record-keeping and this requires your attention to avoid possible operating restrictions and/ or fines. Here are five questions to ask yourself—and answer—before reading further:

  • What type of flowmeter is installed for my GT and/or duct burners?
  • When were my flowmeters first calibrated?
  • What method was used for the initial certification of my flowmeters?
  • Is my plant a peaking unit?
  • Does my plant perform fuel flowtoload check testing?

One of the first speakers at the annual meeting, which got underway shortly after a welcoming luncheon January 16, was Mike Magnan, plant manager, PPL Lower Mt Bethel Energy LLC, and a member of the 501F steering committee (Sidebar 1).

Magnan made an insightful presentation on changes to the way EPA audits the calibration and certification of natural-gas flowmeters serving GTs (Fig 1) and duct burners on heat-recovery steam generators (Fig 2 , Sidebar 2 ). The content of his talk is of importance to virtually every plant and asset manager responsible for GT-based generating facilities, no matter what type of engine is installed.

Most plant supervisory personnel are knowledgeable about CEMS and EPA’s electronic auditing procedures, and may even be familiar with flowmeter accuracy testing and the documentation requirements specified in Appendix D, Section 2.1.5.1 to 4 0CFR75—Title 4 0 (“Protection of the Environment”) of the Code of Federal Regulations Part 75 (“Continuous Emission Monitoring”). If not, consider it required reading.

However, relatively few are aware that the EPA began electronic auditing of flowmeter test data—including the initial certification date—at the beginning of 2 006. The carefully researched message disseminated by Magnan in mid January pointed out the risks to plants identified by the new audit procedure as being out of compliance—specifically the potential for both a fine and a forced shutdown until flowmeter calibration deficiencies can be corrected.

Some of the first experiences with the new audit procedures shared with the editors of the COMBINED CYCLE Journal by knowledgeable sources in the field, weeks after the 501F meeting, indicate that Magnan’s concerns were well founded.Penalties of $1 million or more reportedly have been threatened for significant infractions. Another threat: “On-the-spot” fines of technicians responsible for flowmeter calibration.

Magnan said one of the rules governing flowmeter testing that demands the attention of owner/operators is the official initial certification date. Plant managers and engineers know that most flowmeters require recertification after 2 0 quarters (five years) of service. They typically use the “first fire” date as the initial certification date programmed into the electronic data reporting (EDR) system, which tracks flowmeters by their serial numbers and includes all initial certification information. Under the “new” rules, he continued, the initial certification date is the actual date the flowmeter was certified by the laboratory conducting the test. This may be months—and possibly years—before the “in service” date.

2. The basics of fuel-flow measurement and reporting

Fuel-gas flowmeters for gas turbines, and for duct burners supplied with heat-recovery steam generators, usually are comprised of (1) an orifice plate or flow nozzle that produces a differential pressure proportional to the square root of fuel flow, (2) a differential-pressure transmitter, (3) a line-pressure transmitter, and (4) a temperature transmitter. With the information provided by these instruments, mass flow rate can be calculated at the density of the gas as supplied. A gas chromatograph also is incorporated into the instrumentation package when a plant’s fuel-gas resources have a wide range in composition.

Knowing the calorific value of the gas allows the user to input into the continuous emissions monitoring system (CEMS) the total number of Btus released during combustion in the GT and/or the duct burners. The Btus released are added continuously by the CEMS data acquisition system (DAS). Later, this information is input into EPA’s electronic data reporting (EDR) system for audit by the agency’s monitor data checking (MDC) software.

The individuals in your organization responsible for the CEMS EDR, and air permit compliance, normally are those who provide specific guidance on how to answer flowmeter questions. The answers to some questions may vary from state to state and among EPA regions because many of the regulations are being questioned for the first time.

Wait, there’s more. Knowing and properly recording the initial certification date is not your only concern, said Magnan. The certification standard is very important also. The predominant type of fuelgas flowmeter serving GT-based generating plants is the honed-orifice metering run. Section 2 .1.5.1 of Appendix D to 4 0CFR75 specifies the acceptable ASME (American Society of Mechanical Engineers) and AGA (American Gas Association) standards for orifice meters.

Note that foreign certifications— such as those based on the ISO 5167 (International Standards Organization) method—are not currently acceptable to the EPA.

Planning for recertification

If your flowmeters require recertification, incorporate this activity into the program for your next outage as soon as possible. Keep in mind that outage duration may be impacted by the negotiated turnaround time at your certification facility, as well as by your ability to remove the flowmeter and get it to the lab in timely fashion.

The key to a successful outage is job planning. Work package for the flowmeter recertification effort should include (1) procedures for lock-out/tag-out, (2) identification of consumable parts (gaskets, nuts, bolts, etc) and special tools (torque wrenches, for example), (3) development of guidelines that ensure safe rigging and lifting of the flowmeter section (can be up to 18 ft long), (4) supply of nitrogen for purging fuel gas, (5) supply of a calibrated methane gas detector to determine when purging is complete, (6) a plan to clean and lubricate nuts and bolts, (7) provision of blanking flanges or flange covers to prevent foreign material from entering the pipe, (8) a digital camera to visually document as-found conditions, (9) verification of straightening vanes, (10) construction of a durable shipping crate for the flowmeter assembly, and (11) hiring of a reputable carrier to deliver the flowmeter assembly to your certification facility.

Knowing whether your fuel-gas flowmeter is flanged or welded in place is something else to investigate at the planning stage. Most meters on GE frames are welded in and most serving Siemens engines are flanged If you’re at a loss identifying a suitable lab, consider one of these as a starting point:

  • Alden Research Laboratory Inc, Holden, Mass.
  • Southwest Research Institute, San Antonio, Tex.
  • Colorado Experiment Engineering Station Inc, Nunn, Colo.

Schedule. Recertification can take anywhere from about one to four weeks, depending on lab workload. However, Magnan did say that if you had the luxury of scheduling recertification when your lab of choice is not busy, turnaround could be as little as four days plus transit time.

An option for decreasing outage time is enlisting a flowmeter OEM that can juggle its production’s need of calibrations back in the queue to accommodate a timely calibration. Alternatively, it could install a qualified spare flowmeter or spool piece while the plant’s flowmeter is out for recertification.But substituting a temporary spool piece for the meter could require reporting CEMS data at maximum fuel input under all operating conditions, which would consume additional NOx emissions allowances.

The outsourcing option

Use of a third-party service provider for the entire process of flowmeter removal, crating, shipping coordination, calibration scheduling, and reinstallation increases your out-ofpocket cost, but may be warranted depending on the time available and your experience in completing the required EPA and calibration paperwork and in handling the other required tasks.

The course to steer obviously depends on the realities of budget, outage duration, recertification schedule, and available staff. Normally, the flowmeter recertification would be performed during a regularly scheduled outage—such as a GT combustion inspection or HGP inspection.

If plant personnel are doing all the work—including gas-line purge, flowmeter removal, crating, scheduling of shipping, installation of a spool piece or a spare or leased certified flowmeter, etc—Magnan says the total cost of recertification for a 2 × 1 combined-cycle plant with duct burners could run into tens of thousands of dollars. Flowmeter recertification alone (lab fee) is at least $3500.

Important to remember is that labs require you to communicate the specific meter dimensions, number of test points, number of taps to be tested, and Reynolds number range in advance of the calibration. Also, have a copy of the initial certification ready to send the lab if that’s requested as well.

Virtually all flow tests determining discharge coefficient are performed using water because the dimensionless Reynolds number is independent of the flowing medium. Air or natural-gas flow calibrations are costly—and, therefore, rare.

AGA, ASME standards

Russell Robinson, director of process flow measurement for Control Center LLC, Orlando, Fla, who was contacted by the editors after the 501F meeting, warned that orifice metering design standards have changed in recent years.Case in point: The AGA recognized in its April 2000 revision to Report No. 3 meter specification and installation standard that flow conditioners and thicker orifice plates were required to meet the low uncertainty in flow measurement.This is particularly important to owner/operators of GTs manufactured by GE, which specifies AGA design standards for its fuel-gas flowmeters.

AGA, Robinson continued, revised its orifice metering standard to improve the accuracy of flow measurement that previously had been impaired by excessive bending of thinner plates and non-uniform or underdeveloped flow profiles caused by elbows, valves, and other in-line fittings upstream of the flowmeter.Most GT plants do not have the real estate to accommodate up to 145 pipe diameters of unobstructed straight length upstream of the orifice, so their only option is to retrofit flow conditioners. Recertification and calibration are necessary after such changes are made.

Magnan pointed out that the EPA recognizes both AGA and ASME orifice metering runs to accurately measure fuel flow rates based upon differential-pressure measurement theory. These codes are the two dominant design standards governing the geometry, construction, and installation of orifice meters where low uncertainty is critical. The two codes have subtle differences in their design and application but offer the user with the same quantifiable method to measure mass flow to within about 0.6% uncertainty.

EPA also recognizes certification by flow calibration. An independent NIST (National Institute of Standards and Technology)-traceable flow lab is commonly employed by knowledgeable parties to verify the calibration of either style meter.The cost of this method for initial certification is warranted since heat rates commonly are contested on new GT installations. Proper documentation of meeting AGA or ASME geometry is one important cornerstone to backing up these claims. A fresh lab calibration also provides all concerned parties with accuracy validation. Degradation or damage to the meter’s internal dimensions, surface finishes, and geometry can invalidate a lab calibration or OEM’s certification to ASME or AGA.

Recalibration required

Annual testing is required to ensure proper flowmeter calibration. The relative accuracy test audit (RATA) correlates fuel flow (Btu input) to GT load (MW). A minimum of 168 operating hours is required to complete this test, which must be conducted to extend the period between recertifications from four to the 2 0 quarters specified earlier.

What this means is that peaking plants generally will not be able to meet the RATA 168-hr test period and may be forced to recertify flowmeters annually or to conduct yearly visual inspections. Sometimes a visual inspection every 2 0 quarters will suffice. According to the Code of Federal Regulations, a combustion unit is a peaker if it has an average annual capacity factor of 10% or less over the previous three years and has not exceeded 20% (capacity factor) in any of those years.

Annual recertifications for peakers are more an inconvenience and expense than a detriment to operational flexibility. Plant personnel should consult their corporate environmental specialists responsible for EPA reporting before finalizing a recertification plan.

Visual inspection

A portion of Magnan’s presentation concerned visual inspection of the orifice plate and flow straightener (if installed). He suggested using a borescope and following the guidelines presented in AGA-3. This requires taking and retaining digital images of the upstream and downstream edges to verify their condition. Inspections should be reported in the quarterly EDR and filed with other CEMS QA/QC records. Magnan warned that breaking the flange to check orifice condition can impact certification.

Robinson explained to CCJ editors that for a typical 8-in.-diam gas line serving an F-class engine there’s about 0.125 in. of “slop” between the inside diameter (ID) of the orifice- flange bolt circle and the OD of the orifice plate. Failure to exactly recenter the orifice can produce readings of from 1% to 5% off the actual flow rate.

Orifice flanges machined with alignment pins are recommended to ensure accuracy after pulling the orifice plate for inspection. Plant staffs also should install new gaskets specified by the flowmeter OEM, Robinson continued, to ensure that the proper tap-to-tap spacing is maintained. He also said to be sure to use the star bolt-tightening sequence to prevent misalignment of the orifice plate, and to put the bevel facing downstream.

Robinson recognized the value of using a borescope to identify damage to the orifice plate but said it is not effective for quantifying the extent of any damage. Every third year, he said, orifice plates should be removed for a close-up bench inspection that verifies flatness, surface condition with a profilimeter, orifice diameter and concentricity with a three-point bore gage, orifice sharpness, etc.

Periodic borescope examinations may be especially valuable when a fuel other than conventional pipeline natural gas is burned. Liquefied natural gas (LNG) is a case in point.LNG is getting considerable press today as an alternative to indigenous supplies typically burned in the US.

Use of LNG in GTs designed for natural gas can impact emissions, combustion dynamics, reliability of hot-gas-path parts, etc, said Robinson. Specific to this discussion, he added that some LNGs are rich in byproducts that can adversely affect flowmeter accuracy. Figs 3 and 4 show rings of LNG byproducts plating out on a flowmeter orifice plate.

In this case, there was noticeable pitting and surface-finish deterioration on the upstream face of the orifice plate associated with a black dusty coating. Pitting and buildup, continued Robinson, were well in excess of the 50 micro-inch surface roughness limit for orifice plates built to ASME and AGA standards. He estimated that negative bias errors of over 2 % may result, causing the flowmeter to read lower than expected.Buildup of LNG byproducts in the orifice pressure taps, instrument tubing, or delta-p transmitter may result in even more significant bias errors.

User presentations, roundtables

Magnan isn’t the only steering committee member very involved in content development and presentation at 501F meetings—they all are.Tegen, Cogentrix’ chief CT engineer, led an active discussion on disk-cavity temperature issues and rotor-aircooler orifice changes early on the second day.

Some users have experienced high disk-cavity temperatures after outages. Excessive seal clearances can cause this condition; changes in settings for vane flow coefficients during the refurbishment processes can be another reason, as can stuck Stein seals (spring-loaded axial seals between the blade rings). Guidance on where to look first for a solution was what users came away with.

Turbine section, generator.

Tegen also headed up the turbinesection roundtable on the last day and worked with Gary Giddings, O&M superintendent at Seminole Electric’s Payne Creek Generating Station, to coordinate a generator roundtable. Generators were not part of the Siemens Day program this year (see discussion of the OEM’s sessions below), so the user roundtable was the only generator session at the meeting.

It focused on 18-kV air-cooled machines. One topic of interest was winding failures and their precursor, high ozone generation. Another was the need for permanent partial-discharge monitoring to ensure reliability and long life of the unit.

Ray Martens, recently promoted to plant manager of the Klamath (Ore) Cogeneration Plant, was a busy man on the second day, chairing the compressor section roundtable, combustion section roundtable, and leading a discussion on ways to improve starting reliability.

Compressor section. Icing was the primary topic covered during the compressor session and Martens made a presentation to get attendees primed for productive “give and take.” At Klamath, he said, he has cameras installed on both sides of the plant’s GTs for visual confirmation of icing and has done some programming to sound an audible alarm when conditions are ideal for ice formation. When warned, he said, operators increase load immediately. IGVs (inlet guide vanes) open and the pressure drop across the vanes decreases to prevent icing. Load and other operational adjustments are fine-tuned later.

3. MPS expands service capabilities

Mitsubishi Power Systems, which used to build the 501F engine under a Westinghouse license, remains active in the service business. Michael Robeson, service sales manager, updated the user group on its capabilities. The company has service operations based in Houston and Orlando and has increased this segment of its business dramatically over the last couple of years. Mark Bissonnette, manager of steam and gas turbine field services, told COMBINED CYCLE Journal editors in May that MPS’ repair and overhaul man-hours in 1Q/2006 alone nearly equaled that for all 2005.

He said the company now has the personnel to conduct up to eight inspections simultaneously and expects to expand staff by about 15% this year. Engines currently supported include the 501 Fs and Gs that it manufactured, plus 501Ds and F/FD2s built by Siemens. Most recently, MPS has done combustor inspections (CI) on GE Energy’s 7FA. Typical outage durations are five or six days for a CI, depending on the scope of work; 11 days for a HGP inspection, and 24-28 days for a major.

Three thoughts on an alternative idea discussed at the session were these:

  • Divert some hot compressed air to locations ahead of the air filter to warm the inlet. A disadvantage of this approach is that a relatively large amount of expensive hot air is required to produce a positive benefit.
  • Inject hot compressed air downstream of the filters and evaporative cooler, and upstream of the silencers.
  • Design/install a manifold to inject hot air as close to the bellmouth as possible. Advantage: Less air is required the closer to the inlet that you inject it.

The need to protect the compressor against icing was foremost in the minds of the OEM’s engineers in the development of a product modification (prod mod) to sound an audible alarm when an icing condition exists and/or initiate an automatic runback. The upgrade was announced as commercially available during the Siemens Day presentations. Klamath Cogen is considering the upgrade.

Other topics addressed at the compressor roundtable included hook-fit wear experienced predominantly on the W501FD2 diaphragms for rows 1-6 and experience in water washing online and offline.

Combustion section. Refurbishment of transition pieces (TPs) took the spotlight in the combustion section roundtable. One reason is the considerable expense associated with this task. Owner/operators discussed the pros and cons of overhaul practices by the two manufacturers of 501F engines—Siemens and its Orlando neighbor, Mitsubishi Power Systems (MPS, see Sidebar 3).

Users said that what Siemens typically does is cut off the exhaust section of the TP, weld on a new piece, machine, and coat. MPS appears parsimonious by contrast. The company does not believe radical surgery is necessary in most cases. What it does, according to users, is to clean up fretting wear, fix cracks, straighten the exit area, and recoat.

Martens said he did a trial comparison of both methods. Only problem he found with the MPS work: There were some correctable small cracks in the picture-frame area of the flange—the part Siemens normally would cut off.

One user said he was working with Siemens to reduce the cost of refurbishment by restoring the profile of exit flanges via cold working of the metal—this as an alternative to surgery. The OEM will have to modify its fixtures and procedures, he continued, before the option could be considered as a commercial offering.

There also was considerable discussion on pilot nozzles. Experienced users recalled nozzle failures on early 501F machines; they often found parts of broken nozzles in combustor baskets. First solution was to run dual-fuel nozzles with oil ports plugged. However, with this “fix,” engines had an abnormally high number of trips on temperature spread because of flow problems.

Siemens responded with a new ultra-low-NOx gas-only pilot nozzle.It successfully passed test-bed and validation tests—and its first commercial success was reported during the Siemens Day presentations.

Starting reliability of the 501F has improved dramatically over the last couple of years—a tribute to the efforts of both users and the OEM. The open discussion on this subject brought owner/operators up to speed on the various ways performance can be improved—this to ensure that every plant is capable of at least achieving the current fleet average. The session illustrated well the value of attending user group meetings. In a competitive world, especially, you want to be sure the unit will start after the button is pushed.

By way of illustration, one of the discussion points involved failure to restart reliably after a trip. Users said that following a trip, gas is hot and if your fuel valves are tuned for the cold gas normally expected during startup, you may have insufficient energy for restart. This can be corrected by modification of gas piping to admit cold pilot gas to both the pilot and main gas lines on restart.When the engine achieves synchronous speed hot gas is admitted and the cold-gas supply shut off. Starting reliability on restarts went from 60% to 98% at one plant.

Kettle boilers are installed in most 501F combined-cycle installations for the purpose of reducing the temperature of rotor cooling air.The advantage these air-to-water heat exchangers, built to Section 8 of the ASME Boiler & Pressure Vessel Code, offer over alternative air-to-air exchangers is that heat is captured by the system, improving performance, rather than being rejected to atmosphere.

The cylindrical heat exchanger is about 60 in. in diameter at its mid section and has conical inlet and outlet sections at either end. Water is on the shell side, air flows through 0.75- in.-diameter Type-304 stainless steel tubes arranged for a single pass.Tubesheets at either end are fixed with tubes rolled and welded into them; expansion is accommodated by a bellows in the shell.

Everything comes at a price. And the price for a small gain in efficiency is that tube leaks allow oxygen to enter boiler water (rotor air being cooled is at 2 50 psig), which is connected into the HRSG (heat-recovery steam generator) circuit. Another problem with 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 fuel system—a condition that would abort firing on the next cycle.

Scott McClellan of Arizona Public Service Co made the presentation based on experience at one of the company’s plants. APSC identified tube cracking consistent with low-cycle fatigue. The condition was caused by improper sizing of the shell expansion joint for the thermal growth experienced. The heat exchanger was retubed. In case you haven’t already guessed, the boiler manufacturer is no longer in business.

Web forum facilitates communication between users, OEM

A controls focus group was formed by the 501F User’s after the organization’s 2 004 meeting to present to Siemens in an organized manner the concerns of owner/operators regarding TXP. A Siemens customer survey completed in 2 004 had identified TXP as an area that needed improvement. Mike Magnan provided the leadership to move the initiative forward.

Magnan compiled a historical perspective of control systems for engines of both Siemens and Westinghouse design, and outlined the goals of the group’s collaborative effort with the OEM, for the CCJ’s report on the user group’s 2 005 annual meeting (3Q/2005, p 117; available at www. psimedia.info/ccjarchives.htm).

In January, he updated the 501F group on the progress in resolving controls-related issues. Magnan summed up the first year’s effort this way: “Siemens has responded to TXP issues to benefit users and improve customer satisfaction.”

Siemens hosted six web-based I&C forums in 2 005 between June and December. Discussion focused on engineering-related fleet-wide I&C matters; site-specific issues were not addressed. Sessions were limited to two hours. A published agenda kept the interchange focused.However, each session has a “fleetoriented controls question period” for participants to introduce items not on the agenda.

Anyone who has been on a conference call at sometime has questioned the efficiency of such communication. Siemens’ Web-forum organizers were careful to establish specific groundrules for participation so no one person would speak too long and interruptions would be at a minimum. To illustrate: Participants must keep their phones on mute when not directly addressing the forum. Questions that arise while under “listen only” are introduced via a telephone prompt that the moderator will recognize and introduce the person with the question; alternatively, questions can be submitted via the website.

Each web session begins with a review of the (1) goals and objectives of the forum, (2) protocol, and (3) review/status of action items from past forums. Sessions end with the fleet-oriented controls questions noted earlier and development of the agenda for the next forum. The primary discussion topics are sandwiched between the opening/closing formats. One of the most valuable components of the open discussion is a short session entitled “Tips and Trick of the Day.” Best practices and practical ideas are keys to successful plant operation.

Here’s a sample of discussion topics conducted last year:

  • Single-point failure concerns.
  • Motor-operated valves, including field setup and control system overview and setup.
  • Training—what’s been successful, what’s beneficial?
  • Starting sequences from ignition to full-speed no load.
  • Tie-wrapping quarter-turn valves in the open/closed position to protect against inadvertent operation caused by operator error or fieldinduced vibration.

Magnan noted that from five to 10 users participated in the web forums on average. Those not able to participate can obtain a summary of the discussion on the OEM’s Customer Extranet Portal.

Remainder of Magnan’s presentation concerned the future of TXP support, parts obsolescence, and parts/ service risks. It suggested strategies for maintaining high plant availability/ reliability if rumors circulating in the industry proved correct that Siemens was going to migrate to the SPPA-T3000 and curtail its support of TXP and other controls platforms.

Rumor no longer. Less than a month after the 501F annual meeting, Siemens Energy & Automation announced that it would phase out its Simatic S5 and Simadyn D products— critical components of the TXP control system—effective Oct 1, 2006. In brief, they will no longer be manufactured or available for sale by Siemens. In letters sent to users impacted by the OEM’s decision, Siemens suggested that its customers “evaluate their individual situation and take appropriate steps to meet their needs.” Suggested options included these:

  • Purchase onsite spare-parts inventory of affected products.
  • Purchase the spare parts replacement/repair service as a part of the company’s long-term TXP service plan. This guarantees that Siemens will have the obsolete part in stock ready for overnight shipment to the plant. The company will accept the failed part in exchange for the new one. Siemens has stated that it will maintain an adequate stock of parts to service all customers with a “parts agreement” on all Simatic S5 and Simadyn D systems until Oct 1, 2015.
  • Plan a migration strategy to the T3000 control system (see article elsewhere in this issue describing the new control system and its implementation at two facilities).

Wrapping up, Magnan reminded attendees that plant and asset managers are responsible for managing risk and that user-group meetings are extremely productive forums for identifying and managing risk because they allow you to learn from other users’ experiences.

Preparing for, and recovering from, natural disasters

The frenetic work to get industrial and commercial facilities back into operation following the disastrous Gulf Coast hurricanes of 2 005— Katrina (late August) and Rita (mid September)—was pretty much over by mid January when the 2 006 meeting of the 501F User’s Group convened in Orlando. Some people impacted by the storms had time to reflect on their experiences by then and to share lessons learned.

Storms and floods don’t just happen along the Gulf Coast, they occur regionally, at different times. One reason so many facilities are adversely impacted is that most don’t have up-to-date disaster plans. Perhaps it is human nature not to pay much attention to a late-summer hurricane in the Gulf when your plant is in New England where businesses typically are disrupted by spring flooding. It’s easy to convince yourself that you will get to updating your plant’s disaster plan before rivers overflow next.Often that doesn’t happen.

Several user groups serving the GT-based powerplant community have followed the 501F’s lead in addressing the importance of having a formal document that enables your staff to prepare for, and recover from, a natural disaster. With repeated attention given the subject, perhaps the idea will gather momentum.

The impacts of Rita certainly were fresh in Maintenance Manager Paul Terry’s mind when he stepped to the podium. Muffled “wows” were heard with each slide as he showed the damage that occurred at his plant, RS Cogen LLC, which sat right in the path of the Category 5 storm.

By way of background, RS Cogen is a 2 × 1 combined-cycle cogeneration plant with 501FD2 GTs that began operation in fall 2 002. It is jointly owned by Entergy Asset Management and PPG Industries Inc, and is located within the latter’s Lake Charles (La) chemical complex.

What Terry had to say about dealing with Mother Nature sounded simple enough—essentially common sense. But as he told the CCJ editors after the presentation, you can only imagine what it must be like—not really, however—until you’ve been through it.

Terry divided lessons learned into three categories—equipment, people, other. They provide valuable guidance for updating your disaster plan.

Equipment

1. Shut down equipment as completely as possible; de-energize circuits, conserve batteries by securing all dc loads. Terry said they systematically shut down the combinedcycle plant and its substation over about a day and a half. An error in judgment, he continued, was that they didn’t open the breakers serving the DCS, lube-oil pumps, and turning gear. Thus, the batteries drained when the plant went black. You can’t start a 501F without good batteries, Terry warned.

A big challenge the RS Cogen staff faced after danger had passed and it was safe to restart the plant was finding new batteries. The plant went black on a Friday night and it was the following Wednesday before power was restored to portions of the facility. At the cogen plant, some available power was routed to the DCS so the staff could begin checking circuits—one at a time. It took another four or five days after that work had begun to get new batteries into the plant and initiate startup activity.

Terry told the editors that one of the problems in coming up with a realistic disaster plan is that you really can’t visualize things taking the time they do because of the devastation. New batteries might be obtained in a day or so under normal circumstances. But when many facilities need batteries and you outstrip local supply and transport, how long would you expect this to take?

2. Pump out lube-oil reservoirs to storage tanks. Drain lubeoil tanks if you believe a tidal surge is likely. You don’t want to risk an environmental cleanup.

3. A cooling-tower contingency plan is top priority. Towers are susceptible to wind damage and you don’t want cooling-tower repair as the critical-path item. How will you cool lube oil? Where will makeup water come from?

4. Electric motors will have to be dried out and refurbished. If they were submerged in saltwater, as at RS Cogen, a desalination cycle is a necessary part of the overhaul. Shop time, if you can locate a shop with capacity to spare, immediately increases to about a week—on an expedited basis.

5. Get rental equipment onsite before the storm hits—virtually impossible to get it after. This includes generators, air compressors, pumps, etc. Don’t forget to stock sufficient fuel at the plant and store it out of harm’s way or the rental equipment won’t do you much good.

Personnel considerations

1. Evacuate everyone if possible. Timing of the evacuation must be part of your plan. Release staff earlier rather than later; get nonessential personnel offsite quickly. Remember that highways will be jammed and dangerous.

2. Provisions, sanitary facilities, etc. In certain situations, it may be safer for some people to remain onsite. Others may not have family concerns and can stay as part of the restart crew if their safety can be assured. Still others may be able to return shortly after the storm passes but not be able to commute home daily. Expect to accommodate these employees for two weeks. At a minimum, here’s what you will need:

  • Toilets, including sewer; showers; clothes washer and drier.
  • Decent food and cooking facilities; water and ice.
  • Safe area for sleeping.

Improvise, adapt, overcome—as the saying goes. At RS Cogen, improvising included building a makeshift shower that used water from the condensate storage tank.

Other things to think about

Everything changes when Mother Nature decides to wreak havoc.

Here’s a list of some of the things you may take for granted but will not be available:

1. No communications, including office phones/fax, mobile phones, plant radios, e-mail.

2. No deliveries, including USPS mail, UPS, FedEx, etc.

3. No hotels and restaurants.

4. No fuel. This impacts employees as well as vendors and contractors.

5. No utilities onsite.

6. And don’t forget the real possibility of curfews enforced by police and National Guard.

Batteries are critical Terry stressed the need for a fully charged battery bank to implement a successful restart. The 501F steering committee anticipated this and invited Fred Fasting and Jerry Estes of Interstate Battery Powercare, Houston, to participate with a presentation on how to keep station batteries in tip-top condition and to highlight pitfalls to avoid in operation and maintenance.

Fasting was at the podium and began with an overview of the reasons batteries fail catastrophically— essentially an outline of the talk he was about to give. Then he went back to “go” to review the types of batteries used in powerplant service. Flooded liquid-electrolyte batteries (so-called floodedcell lead/acid) are the standard for this industry, Fasting said. Their attributes: low cost, good performance, and long life (1000 full discharges or more for some models) when properly maintained. Batteries are arranged in series and typically have 60 cells. Such a system would provide current down to the 1.75 V/cell required for powerplant application without damage to the battery.

Each flooded cell is encased in a flame-retardant, shock-absorbent, transparent “jar” that allows operators to visually inspect electrolyte level and the condition of cell plates. Electrolyte is maintained at the proper level by removing the vent cap and adding water to the cell. Note that hydrogen is produced during recharge, so all battery rooms must have adequate ventilation.

So-called VRLA ( for valve-regulated lead acid) batteries championed initially by the telecommunications industry sometimes are considered as an alternative because they are hyped as “maintenancefree.” However, they are seldom used in the power industry because their internal condition is not visible. Estes stressed during the discussion period that the term “maintenance free” is a misnomer; VRLA batteries require periodic inspections and actually demand more care than their flooded cousins.

Interstate Battery did a considerable amount of business after the two hurricanes. Reason, as Terry mentioned earlier, was that as plants were shuttered, emergency lights and other loads remained connected to the batteries and they drained completely. Perhaps this experience suggests the need for disconnect switch between the battery bank and the plant that can be opened when a plant is evacuated to ensure the availability of power on return.

Healthy batteries, Fasting continued, have a specific gravity of between 1.215 and 1.250 and should recharge to full capability, even after a severe discharge to a per-cell voltage of 1.67. If discharge continues, to say 1.5 V/cell, termed “very deep discharge,” quick action is required to prevent permanent damage. This is almost at the point of so-called “cell reversal” which means the affected cell will reverse polarity. Avoid this by monitoring the discharge until the voltage drops to a specified point and then disconnecting the battery bank and recharging.

The specific gravity associated with a deep discharge is very close to unity—1.01, for example. This means you have water (the condition is known as hydration), not acid, and lead is soluble in water. If the battery bank is not recharged within 24 hours of a deep discharge, Estes said “they’re toast.” Even if you are able to resuscitate your batteries, it’s unlikely they’ll ever attain more than 70%-80% of their design capacity.

Permanent damage is caused by the formation dendrites—think of metallic snowflakes—when lead and water coexist. Essentially, the lead becomes porous and “grows,” puncturing the insulator between positive and negative plates, thereby shorting the cell (Figs 5, 6).

If your battery bank becomes scrap, there are at least two immediate concerns: (1) Paying about $30,000 of unbudgeted funds for a new set of batteries, and (2) disposing of the damaged batteries. Latter is expensive, too, and obviously unbudgeted. If you have the responsibility battery disposal, be sure to hire a reputable environmental firm with appropriate experience—and one that will take title to the scrap. It is particularly important for the disposal firm to assume “ownership” of the spent batteries from the plant before they leave the property to break the chain of custody and EPA liability.

Periodic testing is suggested to ensure that batteries are in top condition. It limits the number of “unknown factors” that always come to light during an emergency. Testing can be as simple as a recordable rundown test or a formal IEEE load test. The test selected should fit the severity and critical nature of the plant that the battery bank supports.

The full-load test is the MRI of battery testing. For flooded batteries, IEEE standards recommend that it be done before the plant has completed its first year of service, then again after the third year of service. Following that, every five years is sufficient. In years past, the fullload test consisted of confirming a specific amp discharge for one hour; a salt bath served as the load. Today a resistive load bank is used in place of the salt bath. Instruments monitor every connection point every 10 seconds and the performance claims on the purchaser’s spec sheet are confirmed.

The projected end of life for station batteries typically is eight or 16 years; project performance at EOL is 80% of the new rating. This assumes proper maintenance and a batteryroom temperature at a relatively constant 77F. Higher temperature increases capacity, shortens life; lower temperature decreases capacity, increases life.

Siemens Day: Robust program includes new topics, faces

Siemens developed the program for the third day of the meeting, expanding its traditional content and introducing users to members of the senior customer-care team that they might not have had contact with previously. For example, Sallie Bachman, director of the company’s GT repair network, was on hand to address user concerns regarding shop reports and repair lead times; Bob Allen of Siemens’ new boiler technology services group presented on the OEM’s expanding capabilities for optimization, maintenance, and upgrades of heat-recovery steam generators (HRSGs).

Ron Bauer, director of quality management, opened the meeting as he has for the last several years. Customer satisfaction in the 501F (engine has been renamed SGT6- 6000F by the OEM) fleet continues to improve, he said. Expectations are that this improvement will continue because Siemens is listening to its customers and benefits are accruing to both, Bauer added.

Harvey Grassian, director of market and customer analysis, is the person responsible for quantifying customer satisfaction and for reporting to management where the company is not adequately responding to market needs. The information collected is used to improve products, services, processes, etc, thereby creating a stronger bond between the OEM and end users. In simple terms, Siemen’s six sigma approach to customer satisfaction follows this flow path: (1) Assess the voices of customers; (2) analyze what was said; (3) improve business processes, commercial offerings, technologies, services, etc; and (4) validate with customers the improvements achieved.

The amiable Grassian extracted more responses to his 2 006 survey of SGT6-5000F users (conducted late fall 2 005) than ever before—nearly 100. He said that while the company’s overall satisfaction rating by SGT6-5000F users improved only slightly over 2 005, there has been a statistically significant improvement in the group’s rating of Siemens products and services since the current program of quality record-keeping began in 2 003. This also is true of its response to GT technical issues, dissemination of technical information, and in the timeliness of outage-report preparation.

Users recognized improvements in starting reliability, operational reliability, and availability between 2005 and 2 006, and significant improvements in the lives of combustor baskets and transition pieces. Looking ahead, areas of greatest importance to users include extending inspection intervals and the lives of GT components, and reducing parts costs.

Bachman expended a significant amount of her podium time discussing improvements in communications processes and quality of content to facilitate a productive interchange between the OEM and end users regarding component repairs. The saying, “simple things mean a lot,” has been proved once again.

For example, improved content of shop reports is having a positive impact. Their value has been enhanced by providing greater detail on inspections, removing jargon, presenting details on product mods implemented in repair, compiling a meaningful executive summary, ensuring faster delivery, and giving greater attention to accuracy.

Repair reports are now posted on the OEM’s Customer Extranet Portal to increase their accessibility; email notification of each new posting ensures that users can be as current as they choose.

Repair quality improvement is a continual goal, as are improvements in on-time and lead-time performance. Bachman talked about renewed focus on repeat issues to get at the root cause of problems, greater visibility of problems through monthly “quality incident” reviews by managers who can drive change and improvement, and more effective communication of best practices and lessons learned.

Mark Kamphaus, representing the Siemens group responsible for service engineering in the Americas, put real-world numbers to the improvements in availability, reliability, and starting reliability noted by Grassian. Terms important to this discussion are presented in Sidebar 4, p 23.

Kamphaus reported that the SGT6-5000F fleet’s service factor has remained stable over the last three years at nearly 4 5%. This is interesting given the dramatic increase in the price of natural gas over that period of time. Recall that many energy industry experts predicted a dramatic falloff in GT-based kilowatt-hours because of high fuel price. But statistics indicate that while fuel cost and efficiency are important, high availability, reliability, and starting reliability are the keys to maximum use of GT-based assets.

Kamphaus said that operating data show the reliability of the Econopac—the gas-turbine package— is near world-class level, but he mentioned a downward trend in combined-cycle reliability from 2 004 to 2 005. Note that the average reliability of the GT fleet-wide is more than 99%.

The 1% drop in plant reliability was attributed to a fall-off in Rankine-cycle performance—in particular I&C (sensors, cards, wiring), the steam-turbine (ST) proper, and the steam system (valves, HRSG, etc). The ST reliability number was skewed by two events at one site, including a lube-oil leak and fire.

The good news: Two-thirds of the plants reporting data reported reliabilities above the fleet average.

Fleet-wide Econopac availability was 95% in 2005, an improvement of more than half a percentage point over 2 004, with planned outages having the most significant impact (approximately 2 %) on the availability number. Overall plant availability was better in 2005 than a year earlier despite the reliability impacts noted above.

Kamphaus showed users charts of starting reliability that indicated a substantial increase in the number of starts during the 2 005 summer months compared to 2 004; also, curves that revealed more than a 4 % increase in fleet starting reliability year over year. He congratulat ed the owner/operators for their efforts in improving operating and maintenance practices that contributed directly to the significant achievement.

Another statistic that illustrated the progress made throughout the fleet with respect to startup performance is that the 12-month rolling average of equivalent starts at the end of 2 005 had decreased by nearly 50% within a three-year period. A major benefit of this effort is increased time between inspections and overhauls.

Jon Kemmerling, manager of GT service operations for the SGT6- 5000F, opened his presentation with a snapshot of the fleet that showed the lead unit in operating hours, with just under the 100,000 plateau in January, was an F machine. Leader in the FD series, which comprises the majority of SGT6-5000F engines in operation, had passed 36,000 hours at that time. He mentioned that one-third of the fleet had performed its first hot-gas-path (HGP) inspection.

Interestingly, of the engines in service, which were designed for base-load service, only slightly more than one quarter is actually operating in that mode. Roughly one-quarter each are in intermediate, peaking, and standby service (Sidebar 5, p 24 , defines terms).

Kemmerling addressed primary user concerns. One had to do with the potential for highcycle fatigue (HCF) failure of row 1 (R1) turbine blades. Siemens’ investigation revealed the potential for the buildup of very fine particles (lime-like) behind the platform seal pin that decreases clearances and changes the blade frequency with respect to machine frequency. Source of the debris in the cooling air and how long it takes to get through the system was not fully known at time of the meeting.

In extreme cases, HCF could initiate cracking at the leading edge of the blade near the base of the airfoil. Alternatively, the potential for blade failure also exists if the debris blinds the RAC filter and the flow of rotor cooling air is reduced below the amount required.

The value in attending user-group meetings such as the 501F, of course, is to learn of issues that require your attention. In this particular case, owner/operators were advised to implement routine checks for debris buildup during combustor inspections until further notice. If found and blade “lockup” is in evidence, R1 blades should be removed and cleaned.

Through the end of last year, more than a third of the fleet had been inspected and only 14% of that sample was found with debris buildup. No outages had been extended to address that build-up and no cracking was found.

Kemmerling also reported on several other analyses conducted by the OEM that received positive response from users. Examples:

  • Engineering investigation of a problem identified with a modification made to early service-run R2 blades during refurbishment resulted in a recommendation of close monitoring and removal of the row if cracking is identified. Only a handful of units is affected and no indications of cracking had been identified before the meeting beyond the one that triggered the study.
  • A new vane design and repair method for R3 turbine blades was developed to prevent the potential for blades to rub against vanes.
  • The potential of cracking of the outer exhaust cylinder near struts essentially has been eliminated by (1) the development of weld techniques that maximize joint strength and minimize defects, (2) better welder training and qualification, (3) progressive inspection, and (4) replacement of degraded material. Mention also was made of a new strut shield (in validation at the time of the meeting) that is more effective in accommodating thermal growth of both that component and the outer diffuser. Retrofit is expected to take about a week and a half; cover removal is not required.

The foregoing was only the “tip of the iceberg” for Kemmerling’s portion of the program. He spoke for more than an hour and a half, also covering experience with the new dual-fuel pilot nozzle that Siemens designed for the SGT6-5000F, inlet struts, inlet splitter enhancements, solutions to R1 vane inner shroud erosion, coating improvements, R2 vane enhancements, inspection guidelines for R2 vanes, mitigation of icing damage (including cutback of R2 compressor blades to increase resistance to foreign object damage), etc.

Given the quality and depth of material covered at user-group meetings, such as the 501F, it’s difficult to believe that owners and/or operators, or their representatives, from every plant in the fleet would not be in attendance. Technology development by Siemens, as well as other OEMs, is an on-going process and experience with new mods and repair methods should be monitored at least annually by plant personnel to help maintain a competitive edge. The cost of attendance at user-group meetings is “chump change” when you consider the revenue lost by not being dispatched on a hot summer day.

Kamphaus returned to the podium in the afternoon with a presentation that covered in overview fashion the nearly 300 product modifications that Siemens has developed for the F fleet, including number of applications and experience. This was a good review for attendees planning for an outage to help identify mods make most sense given outage duration, current operating regimen, and payback analysis.

Greg Perona, marketing manager for GT modernization, followed Kamphaus with a complementary presentation. He presented details on the OEM’s product improvement portfolio for the SGT6-5000F. Discussion was divided into three major sections:

  • Interval extension.
  • Performance and operational flexibility.
  • Emissions and fuel alternatives.

Interval extension was covered by review of Service Bulletin 51009. The upgrade components discussed, all now available, are intended to reduce trip factors by 4 0% to 60% depending on unit load range. They permit a reduction of fast-start factors and elimination of instant load-change factors with a resultant double-digit increase in equivalent starts and a meaningful increase in equivalent base hours (EBH) between inspections.

Upgraded components that are part of the package include combustor baskets with extended swirlers, R1 turbine blades with improved tip and platform cooling, R2 turbine blades manufactured with improved undercut, improved material for R3 turbine blades, etc.

A special interval-extension upgrade for installation during a combustion inspection (CI) outage was announced with commercial availability expected by this summer. One of its benefits: An hoursbased CI interval of 12,500 EBH, up from 8000. To learn more about interval extension, access “Operating experience, analytical procedure help OEM extend intervals between GT inspections,” CCJ, 4 Q/2005, p 2 1, at www.psimedia.info/ccjarchives. htm.

Performance and operational flexibility upgrades are designed to help plants increase their dispatch rate. For example, the first plant to install the compressor upgrade package described by Perona had just completed the work before the 501F meeting convened. This package can be optimized to allow an improvement in combined-cycle heat rate of up to 1%, or to increase combined-cycle power by up to 3% (improvements expected are for a 2 × 1 facility with model 501FD2 GTs operating base load at ISO conditions).

The upgrade, which requires a commitment lead time of about 4 5 weeks, and from three to four weeks to implement on average, is designed for installation during a major inspection or compressor outage.

Enhancements include improved compressor sealing (upgrade from labyrinth to honeycomb seals), upgraded R16 blades, optimized rotor cooling air temperature, and IGV (inlet guide vane) optimization. All these enhancements are incorporated into production units beyond the FD2.

Likewise, turbine upgrade packages are available to improve combined- cycle heat rate by up to 1% and to improve both heat rate (by up to 1%) and combined-cycle power output (by up to 3%). Replacing R4 turbine blades and vanes can boost the heat-rate improvement by up to 1.5% and power by up to 5%. All these options are designed for installation during HGP or major inspection outages.

Perona was bullish on the OTC (outlet-temperature corrected) upgrade, an idea borrowed from the V fleet that is now installed on nearly two dozen F machines. In brief, OTC is a turbine control process that manages operation in a closed loop to a corrected exhaust temperature based on variations in ambient temperature and engine speed. It is a beneficial alternative to setting IGV angle according to load. Two benefits are particularly impressive: improved part-load heat rate of up to 2% and reduced emissions drift.

A new offering awaiting its first sale at the beginning of the year is an icing conditions alarm. It is designed to mitigate compressor icing potential via a warning and alarm protocol that identifies conditions related to compressor icing. Operator intervention is required to raise load to a recommended level, shut down the unit, or risk ignoring the alarm.

An alternative to purchasing the alarm system is to install an inlet-air heating system designed to maximize GT power during potential icing conditions. It minimizes the amount of heated bleed air from the compressor to that required to prevent ice formation.

Enhancements to reduce emissions and permit the use of alternative fuels rounded out Perona’s presentation. He reviewed the company’s new ultra-low-NOx combustion system that promises singledigit NOx and less than 5 ppm CO when operating on natural gas at 70% or more of the GT’s full load rating. For liquid-fuel applications NOx emissions are less than 42 ppm. Low- CO startup and turndown upgrades are two more alternatives.

Fuel conversions—for example, converting oil-only sites to dual fuel or natural gas only, or gas-only sites to oil or dual fuel—were covered in brief. Looking ahead, Perona discussed mods that might be required to burn liquefied natural gas (LNG) as that becomes more readily available, and the company’s progress in developing a syngas compatible GT to meet marketplace demands as they evolve.

Last presentation on Siemens Day was one that outlined the capabilities of the company’s new boiler technology services group. Mike Davidson, who manages the unit, was supposed to address HRSG optimization, maintenance, and upgrades but he was called out of town at the last minute. Bob Allen filled in. No matter. Both are long-term pros in the boiler business. Each has worked for a couple of the OEMs—designing, building, and operating both fired boilers and HRSGs.

4. Defining terms

Availability and reliability terms, as defined by ANSI/IEEE Standard 762 are:

  • Availability is the percentage of time that a plant is capable of providing service, whether or not it actually is in service and regardless of the capacity level that can be provided.
  • Forced-outage factor is the proportion of time a unit is in a forced-outage condition over a given time period. Reliability (not defined by ANSI/IEEE 762) is the percentage of time a unit is not in a forced-outage condition over a given time period.
  • Starting reliability is the percentage of attempted starts that successfully synchronize with the grid.
  • Service factor is the percentage of time that a unit is electrically connected to the system over a given time period.

And Siemens is no newcomer to the boiler business, having licensed the Benson technology to other OEMs and building Benson-type HRSGs. The service group was formed about a year ago to offer owner/operators of cogeneration and combined-cycle plants a one-stop alternative capable of addressing all of their powerplant needs. The company offers condition assessment, tube sample analysis, and other services to help maintain plants at maximum efficiency and reliability.

Siemens offers physical-condition and thermal-performance assessments for a wide variety of generating facilities worldwide. Beginning with a detailed review of original design data, operating history, and thermalperformance analysis, a plan is developed to address the most important areas for improvement. For HRSGs, these may be related to tubing wastage, flow-accelerated corrosion, plant systems integration, etc.

Allen said that regardless of boiler type or manufacturer, the new service group has the personnel, tools, and experience to maintain, repair, or upgrade HRSGs. A fact that may have been surprising to some users in attendance is that Siemens holds all the necessary ASME Code stamps to design, certify, and fabricate any pressure parts required.

Upgrade GT inlet-air systems to accommodate the local environment

There are good reasons that almost every user-group meeting features a presentation on GT air-inlet system design and filter options, including these:

  • Any foreign material that enters the compressor can erode, corrode, and/or foul the unit, often contributing to poor performance and expensive maintenance.
  • The air inlet house generally does not get much attention by plant personnel until there’s a problem. Continual reminders—almost to the point of nagging—are necessary to stress how important a clean compressor is to plant profitability.
  • There are about a dozen companies vying for your replacement filter business and the resulting competition drives the introduction of new products that may better suit your situation and warrant consideration.
  • Filters supplied with the GT often are part of a bulk purchase by the OEM and not the best alternative for your specific site conditions.

5. Duty cycles

Important to communication is an understanding of technical and business terms. Duty-cycle definitions used by Siemens are:

  • Base load. Greater than 75% service factor (approximately 6500 hr/yr) and operation for more than 125 hr/start. Service factor is percentage of time over a given period—typically one year—that a unit is connected electrically to the grid.
  • Intermediate duty. Greater than 30% service factor (approximately 2500 hr/yr) and operation for more than 12.5 hr/start—and less than base-load duty.
  • Peaking. Greater than 5% service factor (approximately 400 hr/yr) and operation for more than 4 hr/start—and less than intermediate duty.
  • Standby. Less than peaking.

Richard Smith, a VP at Pneumafil Corp’s gas turbine division, Charlotte, NC, presented on GT inlet air handling at the 501F meeting (rsmith@pneumafil.com, 704-399- 7441). He began with an overview on the need for efficient filtration and the impact of plant location on inlet air characteristics critical to filter selection. Variations in regional air quality presented in the accompanying table illustrate why a thorough evaluation of site environmental conditions is necessary before a buying decision is made regarding replacement filters.

Design, operational objectives

To ensure proper performance of inlet air systems, he continued, designers and plant operations personnel must ensure that the following five goals are met—at a minimum:

1. Prevent ingress of birds, trash, and insects. Simple #4 mesh screens at a face velocity of 500 ft/ min and above will stop large foreign materials; extended-area 14 × 14 mesh screens at a face velocity of 2 50 ft/min will handle the insects.

2. Remove free moisture from the air stream —including rain, snow, mist, and fog. Intake hoods reject precipitation by drawing air vertically upwards at a velocity that can discriminate against rain or snow falling downward at some higher terminal velocity. The terminal velocities of different forms of precipitation vary widely and a given hood design will be more effective at preventing the entrance of fast-falling large raindrops than it will slowfalling snowflakes.

Rain and snow. Rain hoods should provide an entrance velocity between about 600 and 900 ft/min; snow hoods 400 ft/min or less. Hood elevation is important to prevent wind-driven ground snow from being driven upward into the filter house. A rule of thumb: The flux of blowing snow drops by a factor of about five when hood elevation is raised from 5 to 2 5 ft. Locating a snow hood at a still higher elevation has little benefit.

Intake louvers for GT air intakes are inertial impingement separators designed for horizontal gas flow. They comprise a series of sinecurve- shaped flow contours (profiles) assembled in a frame with a sump to form a self-contained droplet separation module. A wide range of operating conditions can be met by changing the spacing of the profiles, materials of construction, etc.

A shortcoming of louvers is their susceptibility to icing and plugging by snow under severe winter conditions. By contrast, intake hoods do not have this problem and have demonstrated their ability to work well in arctic climates. Keep in mind that neither intake hoods nor louvers can prevent the ingress of fog and mist into the filtration system.

Mist, or water droplets in the size range of 60 to 150 microns can be removed effectively by means of a commercially available mist eliminator, such as Munters Corp’s (Ft Myers, Fla) DRIFdek-IL. When this product is installed as recommended, mist coalesces into droplets that drain from the inlet side without becoming re-entrained. Removal efficiency of droplets 60 microns and larger is said to be 99%.

Fog. GT inlet-air filtration systems can be challenged in their ability to maintain low pressure drop during periods of heavy fog. It is not unusual to experience restrictions that exceed one or more alarm setpoints. Operation at reduced power output may be necessary.

The challenge presented by fog is that very small water droplets are suspended in air at or close to 100% relative humidity. Researchers say that droplet diameters in natural fog generally range from 2 to 4 7 microns and that the mean diameter is 10; also that a disproportionate number of droplets have diameters of 2 to 3 microns.

Vane profiles like those discussed above under “rain hoods” are one option, but to be successful they require an approach velocity of 1100 ft/min. However, this results in a pressure drop of over half an inch of water. Heating the air stream is an alternative, as is a dedicated fog coalescer.

Smith noted that his company’s Fogstop is once such device. It consists of a coalescing medium of stainless within a steel frame and mounts vertically or horizontally downstream of the intake hood and moisture eliminator. He said that at an approach velocity of 700 ft/ min, Fogstop can remove 90% of the droplets 15 microns and larger with a pressure-drop penalty of only slightly more than three tenths of an inch of water.

3. Temper inlet air to maintain optimum temperature and humidity. Inlet icing is one of the hazards associated with operating GTs in cold climes. Ice can block filter-house inlets and the filters themselves, leading to performance loss, the possibility of the compressor ingesting unfiltered air, buckling of ductwork, etc. In extreme cases, ice can build up on the bell mouth and later be sucked into the compressor where it can do considerable damage.

Inlet heating systems mix hot gas—such as hot compressor bleed air—with air entering the filter house. Or they use hot water or steam as the heating medium. One design objective is to prevent ice formation at the bell mouth by keeping air entering the compressor at 4 0F or higher; this generally is sufficient to compensate for isentropic cooling as the air accelerates in the inlet area.

In warm weather, inlet air must be cooled to maximize engine output. There are several ways to do this; each comes with a price. Details are presented in “GT inlet-air cooling boost output on warm days to increase revenue,” CCJ, Winter 2004, p 4 5 (available at www.psimedia. info/ccjarchives.htm).

4. Prevent solid particles from reaching the compressor. Prefilters usually are the preferred method for removing coarse dust from ambient air. Most often, one of the following designs is selected: disposable flat pad, pleated-media panel, disposable pocket type. They are specified and selected based on their gravimetric (weight) arrestance ability in accordance with ASHRAE Standard 52.1 or equal.

For more on filtration economics, standards, and test methods, consult “Life-cycle cost analysis key to identifying optimum replacement filters,” CCJ, Summer 2 004, p 9, and “Selecting gas-turbine inlet air systems for new, retrofit applications,” CCJ, Spring 2 004, p 37. Both articles can be accessed at website referenced above.

For fine airborne dust, disposable high-efficiency filter elements are widely used. These include extended- media-area rigid mini-pleat and rigid deep-pleat panel filter elements with extend-area pocket filter elements. Static fine-dust filter elements almost always should be preceded by prefilters.

Periodic dust and sandstorms common to most dry, arid, desert regions create airborne particulate concentrations thousands of times higher than those found in typical US cities. Concentrations of this magnitude can quickly overwhelm a conventional multistage static barrier filtration system.

One solution is to specify a selfcleaning barrier filter system, which uses sharp bursts of compressed air to periodically remove entrained dust. These systems also offer excellent resistance to ice and hoar frost formation and rarely need to incorporate dedicated anti-icing capability.

Usually there is no other filtration stage before or after a self-cleaning filter. However, a final-stage nonleaching static barrier filter may be used where unusually high salt concentrations are present or a very high level of cleanliness efficiency is required.

5. Remove metal salts. GTs that are installed in coastal environments, or on offshore platforms, present unique problems regarding inlet-air contamination. The proper selection of filter media and stage efficiencies ensures that the metal salts ingested by the GT are maintained below specific threshold levels (usually 0.01 ppm by weight for sodium chloride) to protect against both low- and high-temperature corrosion.

Degradation mechanisms

Filters and moisture separation systems are installed to remove particulates, moisture, and other foreign matter entrained in the inlet air, as Smith detailed above, thereby protecting critical GT components from erosion, fouling, corrosion, and particle fusion. Here are some comments he had with respect to these degradation mechanisms:

Erosion can cause permanent damage of rotating and stationary components, making their replacement inevitable. An important point: Particles smaller than about 10 microns do not cause erosion, larger ones do when present in sufficient quantity.

Fouling of compressor blades has a negative impact on performance, increasing fuel consumption. In extreme cases, fouling can reduce engine output by as much as 15%. Experience indicates that particles smaller than 2 microns contribute most to fouling and that sticky substances— such as oil vapors, smoke, and sea salt—are the main culprits. They change the aerodynamic shape of the blade. A regular program of online and offline compressor washing minimize the impact of fouling.

Corrosion initiated by wet deposits of sea salt, acids, and other deleterious materials can initiate blade failure in as little as 2 000 hours of operation. It probably is the most serious consequence of inadequate filtration.

Particle fusion occurs when dry particles, nominally in the range of 2 to 10 microns, stick to hot metal surfaces because the fusion temperature of the contaminants is lower than the turbine operating temperature. Result can be permanent blocking of cooling air passages and thermal fatigue. ccj