7F USERS GROUP

General Manager Gerald D Hill and his team of knowledgeable tour guides told the gas-turbine owners and operators that, in addition to restoration repairs, PWPS pursues design improvements with the goal of providing refurbished parts that are “better than the original” where possible. The collaborative process of component improvement involves (1) identification of issues based on field experience, (2) re-engineering and modeling of improved parts, (3) laboratory validation, and finally, (4) field validation.

The users-only sessions on Tuesday covered the 7F compressor, safety practices and lessons learned, controls, and auxiliaries. Vendor presenters included PSM, ExxonMobil Lubricants & Specialties Co, National Electric Coil, and Advanced Turbine Support LLC. Wednesday’s user-only lineup featured a combustion and turbine session, generator session, and open discussion before presentations by Environment One Corp, Praxair Surface Technologies Inc, PWPS, and Turbine Technology Services Corp, among others. Vendor fairs, including dinner, completed the Tuesday and Wednesday programs.

Thursday was GE Day. Presentations and discussion covered the entire engine from the air inlet house through the generator, capped off with a reception and OEM product fair. Friday’s program, which ended at noon, featured parallel sessions on D-11 and A-10 steam turbine maintenance and GE controls and diagnostics.

User presentations, discussion

The closed sessions at most user-group meetings serving owner/operators of frame-specific gas-turbine assets were structured similarly until recently. The standard formula was to start at the inlet air house and work back through the engine covering, in turn, the compressor, combustion section, turbine, generator, and control system. Session moderators would stimulate discussion using key words, recent experiences announced on bulletin boards, etc.

Example: “Who’s not getting expected life from their air filters and what are you doing about it?” Air filters typically were good for 30 minutes of discussion. It seemed that everyone had an opinion or experience to share on this subject. Meeting after meeting many of the same experiences were shared—certainly fine for first-timers, but not many ideas for the veterans to take away. Interspersed between discussion sessions would be one or more relevant vendor presentations. With this format, the questions raised and the ensuing discussion often were predictable.

The 7F, 501F, and 501G user groups have broken the mold, contributing to a higher level of energy at conference activities. All participants benefit from the new format, which has the following key elements:

• Steering committees deep in numbers and knowledge.
• Active participation by all steering committee members, both in making or sponsoring presentations by owner/operators and by contributing to and facilitating discussion forums.
• Eliminating vendor presentations from the user-only sessions and encouraging greater participation by equipment and services suppliers in forums held just prior to the vendor fair. At the 7F meeting, which had exhibitions on two evenings (different vendors each evening), there were six 45-min presentations by suppliers each day, conducted in two time slots following the afternoon break (three companies presenting in parallel in each session).

The 7F steering committee, 15 members strong, has a good blend of experienced “youth,” hard-nosed O&M managers, and engine experts with more than 25 years of industry service. One of the meeting’s non-technical highlights was the recognition by the current steering committee of two of the user group’s founding members in attendance, Pierre Boehler of GenOn Energy Inc and Bill Wimperis of Constellation Energy.

Wimperis, who chaired the first several 7F conferences, represented Baltimore Gas & Electric Co at the first meeting 21 years ago; Boehler was working for Potomac Electric Power Co at that time. Both engineers continue to serve the industry as members of the 501G and CTOTF steering committees, respectively. An interesting contrast: The first 7F meeting was attended by 14 managers and O&M personnel from four electric utilities, this year’s conference hosted more than 250 user attendees from perhaps a hundred generating companies—both regulated and unregulated.

In round numbers, there were a dozen user presentations at the San Antonio conference and they stimulated hours of open discussion so compelling no audible buzz from sidebar conversations was in evidence—unusual for a large audience in a cavernous ballroom. Critical to discussion management is having a team of experts on the floor to clarify questions where necessary and to answer them when others present cannot. The 7F steering committee does this to perfection.

Compressor section

The first user to present said he was concerned by the migration of compressor rotor-blade spacers discovered on two 7241s during their second HGPs. The units had about 1800 starts each but were transitioning from a starts-based regimen to base-load service. The units’ first majors probably are two years off. This owner/operator was not sure what should be done to address the migration issue, if anything, and asked colleagues in the audience to share relevant experience.

His investigation revealed no vibration problems resulting from spacer migration. In fact, calls to several users before making the trip to San Antonio did not uncover any vibration issues in the fleet that could be linked to spacer migration. The OEM’s response, he said, was “don’t worry.” However, GE expressed some concern that if the migration had occurred in rows 7 and 8, where the spacers align, a combined migration of the two rows could happen in the future.

Four attendees said spacer migration had occurred on their machines but there had been no negative effects. One of these users said he used to work for the OEM and had seen many instances of migration but no problems resulting from that movement. Yet another owner/operator had experienced migration on Frame 5s and 7EAs and thought it might be the result of poor craftsmanship—specifically, poor staking.

Someone else said that if just one or two spacers migrate it’s one thing, but if several in a small section of the rotor move, then you could throw the unit out of balance. He reported seeing this on peaking machines and attributed the migration to “thermal ratcheting.” A few others confirmed this based on their experiences with 7B-EAs. The OEM reportedly did field restaking for one user to correct for movement of one spacer.

Is your S17 solution the right one? A user reported that the back ends of some of his company’s 7EAs and 7FAs have had issues. The OEM recommended a shrouded S17 to strengthen the back end of one FA, but material distress was identified when the unit was opened up a year later. Some attendees seemed surprised to learn there are several S17 designs. The latest has staked bolts, older version welded bolts. Washers can be used, or not. In some instances where they have been used, washer rotation has caused rapid wear. Another variable: Some bolts have two tack welds, some one.

Discussion included mention of fretting on the ID shroud face near where adjacent vane segments butt against each other, as well as heavy wear in the V-seal slot. The amount of clearance between the shroud and S17 airfoil tip was brought to the floor. A couple of users measured gaps of up to about 30 mils; group consensus was that one-third that would be better and probably would mitigate the wobbling of parts experienced in some instances.

It appeared there were more questions than answers regarding S17 issues and that the discussion was likely to continue next year in Greenville (May 20-24, 2013 at the Hyatt Regency), if not sooner in the forum on the group’s new website at www.7fusers.org.

Machining might be best option for removing R0 blades to avoid damaging the wheel. One user provided a short presentation on the care required to remove compressor blades without galling. Ice damage on one of his units, initiated by cooling-tower drift, dictated the removal of R0 blades for repairs.

A 1-ton come-along and air hammer were the tools selected for the extraction which proved particularly difficult, galling 17 blades in the process. The high edges on the gall were blended to get the machine back into operation. Replacement of the first-stage wheel and R0 blades is scheduled for next year. This plant online water washes daily, offline twice annually.

Follow-on discussion revealed that a leaking wash-water valve (with the pump off) was the underlying cause of icing at another plant. In addition, a user attested to the difficulties one can face trying to remove R0 blades. He recommended machining out blades to avoid ruining the first wheel and having to replace it at significant cost. The attendee said it took two days of around-the-clock effort to machine out the R0 blades on his engine.

Discussion highlights:

• Most packages the OEM has are for the flared fleet, not unflared units. Example: There are no Package 2s for unflared units.
• Question to attendees: Any experience with OEM’s new Enhanced Compressor? One user reported having a problem-free 8000 hours on one engine.
• R0 dovetail indications building, a user reported. Be prepared for “discovery,” he said. Continuing, he suggested to colleagues that they not get complacent regarding inspection. “As the fleet ages, people who never found any indications are now finding them.”
• Bigfoot mod: Stator twisting in the case slot occurs early and appears to slow/stop. Future impact, if any, is unknown. More than 60 mods have been installed in the fleet, according to one user, and all show signs of twisting. The OEM reportedly has told users that twisting does not impact performance or reliability. It was said that no plant has yet removed stator blades from a unit with a Bigfoot mod.
• An attendee suggested that users look for cracking on the R17 wheel rim; about half a dozen users said they had found it. Recall that the 17^th stage incorporates the stub shaft. Another participant said cracking is expected by the OEM and the wheel is designed to accommodate small cracks (up to 0.125 in.), which can be blended out. However, if you blend and can’t remove the crack, that user said the crack will propagate faster than if no blending had been done.
• Another user said there also are cracks on R14, R15, and R16 wheels, but they generally are not visible to the naked eye. These cracks are said to run inward from either the front or back face of a disk. You don’t know how deep these cracks are until you grind them out. It was said that no failures have been attributed to this type of cracking, “so don’t worry about it.”
• Audience poll: One user still is operating with P-cut blades.
• Make sure R0 blades are properly staked. Over-staking, it was said, can compromise the ability of a biscuit to lock a blade in place.

Air inlet house

A fire that consumed an inlet filter house pointed to shortcomings in plant’s safety plan.

A contractor hired to install prefilters selected a halogen lamp to provide more light in the narrow passageway between the prefilters and conical/cylindrical final filters. An extension cord, draped over the access-door threshold, brought power to the lamp, located about 5 ft inside the door.

Unsafe work practice #1: Never run electrical cords, hoses for cutting torches, welding cables, etc, through open doorways. The contract crew said the door had been wired open, but photos suggested otherwise. Workers took a break and left the light on, then decided to quit for the day because job progress was ahead of schedule.

Unsafe work practice #2: This unplanned work stoppage was not reported to responsible plant personnel and the light remained on. The high heat produced by the halogen lamp ignited combustible material in the inlet house and smoke poured out from the structure. The fire was promptly reported to the control room and emergency procedures were enacted.

Volunteer firefighters arrived within about 20 minutes and took about three-quarters of an hour to extinguish the fire, even though about 80% of the “fuel” had been consumed before the firefighters arrived. The inlet filter assembly and transition duct were a total loss. The support structure suffered no fire damage; silencer panels, inlet-bleed-heat components, and downstream ductwork were salvageable after a thorough cleaning.

Unsafe work practice #3: The danger of halogen lamps and the OEM’s recommendation not to use them was covered by Technical Information Letter 1368-2 but went unheeded. Read and understand all TILs recommended for your units and act accordingly, the speaker urged. This same TIL suggested warning/caution signs on the dangers of working in the inlet air house, but they were not posted. Not surprisingly, the Job Hazard Analysis developed for this project by plant personnel never warned of fire hazards.

Unsafe work practice #4: The safety analysis conducted by company personnel after the fire revealed that the lone door in the filter house accessed the lowest catwalk for the filter assembly. The three catwalks at higher elevations were accessible only by ladder. Had there been workers at one or more of the higher elevations when the fire began, they probably would not have survived. Egress panels were installed at each of the catwalk levels. The replacement inlet air house was installed without lighting; plans are in place to add retractable LED lighting.

A key take-away from this incident was that you cannot over-communicate about safe work practices, proper tool selection, and hazardous conditions. The speaker suggested that his colleagues implement a “Friendly Eyes” program if they didn’t already have one. It involves taking experienced, safety-conscious employees from another site and have them walk-down your plant to identify unsafe work practices. They will find things the resident employees see regularly and accept, but which may not be in the best interests of health and safety.

Combustion dynamics

The user controls expert who last year offered an objective evaluation on the pros and cons of combustion dynamics monitoring systems (CDMS) installed on 7F engines in the fleet he serves returned to the podium in San Antonio. The speaker said he believes the following three CDM systems are acceptable for powerplant use:

• PSM’s AutoTune.
• WoodGroup’s (Gas Turbine Efficiency last year) Ecomax™.
• The OEM’s OpFlex™ AutoTune.

Last year most of the engineer’s experience was with AutoTune; only two engines had been equipped with Ecomax. OpFlex was not considered for deployment because of its cost. What a difference a year makes: Here are the highlights of what he had to say in San Antonio:

• PSM’s AutoTune now has more than 30,000 hours of operating experience in his fleet. Recent upgrades included firing-temperature optimization up to 2420F (going higher still this summer) and an events counter for reporting how many times the system has tuned and what’s been tuned.
• GE’s OpFlex now is offered at various levels to suit specific requirements and pocketbooks.
• Ecomax has been commissioned on six engines in the speaker’s fleet with six more being installed. Experience is now coming in from more than 16 units. He reported the following:
• The system is user configurable and offers control-curve optimization—that is, it optimizes base-load output.
• Post-outage tuning is not required.
• Fuel temperature and firing curve knobs are provided to facilitate tuning.
• Operation is flexible. Example: Ecomax will still tune for NOx when the CDMS is down.
• System’s ability to prevent lean blowout (LBO) seems apparent.
• Monthly reports are provided on low-NOx tuning events that likely have prevented LBO.
• Power optimization is noticeable.
• Installation requires minimal downtime and proceeds in a relatively smooth manner with only minor issues.

Bearing issue

A user offered a step-by-step procedure for installing thermocouples in generator bearings. It was obvious that he hadn’t intended becoming an expert on the subject, but when bearings sent out for rebabbiting came back without thermocouples, someone had to install them.

Saving time by doing the job in-house seemed like a reasonable objective. After all, how difficult could the task be? Everyone who has ever spent time on the deck plates knows there’s usually more to any job than meets the eye. So it was in this case. You can access the procedure at www.7Fusers.org.

Sensitive rotors

A scholarly, yet practical, presentation by a power-company vibration expert walked users through field data analysis, observations, and conclusions of 7F thermal transient vibration. This is another presentation you may want to review at www.7fusers.org. Perhaps the best part, and critical to easy understanding of the subject matter, is that the speaker prepared an old-school paper (no sketchy PowerPoint), complete with 19 figures.

The utility consulting engineer told the group that all rotors have a thermal transient characteristic. For some, the thermal transient can be sufficiently small as to be considered non-existent. On other rotors it can be so severe on a cold start that it can make the rotor go into vibration alarm, cause a load runback based on vibration amplitude, or even cause the unit to trip on vibration amplitude. For most rotors, it is somewhere in between. The source of this transient appears to be a combination of the basic design, manufacturing tolerances, and variations in the assembly process at the factory/shop.

When it has been determined that a rotor has a thermal transient that must be addressed, the expert said, required data should be plotted in polar format for a cold-rotor-condition start, beginning immediately after reaching the full-speed/no-load condition and ending at the steady-state base-load condition. Once a rotor’s transient is quantified, a corrective course of action can be determined.

Depending on how the unit will be operated, either tactical balancing or prewarming of the rotor before the actual start usually are the best options. Should the transient’s extreme points ever become so large that the transient cannot be controlled—even through warm starts and balancing—then the rotor must be sent to a shop for repair.