Case history: O&M issues associated with one model of air-cooled generator

Stator phase-connection conductor fatigue was one of several significant O&M issues mentioned by a central engineering staff engineer from a large owner/operator with seven copies of the affected air-cooled generator in its fleet. All of the units are coupled to steam turbines and are rated 21 kV, 313 MVA. The OEM did not participate in the STUG meeting.

Plant became aware of the problem in August 2010 following a stator-winding in-service failure. A root-cause investigation pointed to phase-bar resonance. The generator was repaired and returned to service, and an aggressive plan was implemented to inspect and repair all at-risk units in the fleet. Solution was to repair damaged strands, where necessary, and improve the support-blocking arrangement for phase bars. The blocking mod has “evolved,” the speaker said, and periodic maintenance is expected. Periodic assessments of generator condition now include natural-frequency testing. Plus, the company’s M&D center has a program in place to warn of conditions that might lead to another failure of this type.

A review of operating data indicated a correlation between conductor fatigue and temperature. The OEM installed an RTD in six phase-lead slots (both top and bottom phase bars) on this model of generator. The RTDs installed in the slots closest to the first and last of the top phase bars revealed a temperature rise higher than that indicated by the other RTDs prior to cracked-strand repair. Data recorded following repairs showed the temperature rise reported by each RTD returned to values consistent with the average of all 12 stator RTDs. The bottom line: There was a definitive correlation between loss of conductor cross-section and temperature rise under load.

Slides posted on the Power Users website describing the findings and work above contain well-labeled high-quality images of value to anyone considering a similar project. Access them at your convenience.

Another generator case history: This one deals with locating, correcting a stator ground

Another user reported on his 3 × 1 F-class plant’s experience troubleshooting a stator ground on the steamer’s generator and the corrective action taken. The steam turbine, with nearly 45,000 operating hours and 950 starts, tripped when the generator field flashed at 94% speed. Gas turbines were shut down and troubleshooting began about an hour after the incident. Generator lock-out/tag-out and visual inspection of principal components—exciter, power potential transformer (PPT), isophase bus (IPB)—were among the first actions.

Next, megger readings were taken on the generator, IPB, and step-up transformer (GSU). The PPT was disconnected from the IPB and assessed as “fine.” The unsealed IPB was checked for water; it was dry and there was no sign of a ground. The generator-to-IPB flex coupling was disconnected, but the generator voltage couldn’t be raised to check for a ground.

GE was called in and its technician verified a ground in the generator. The neutral bus bar was disconnected and all three phases meggered individually. Personnel found the A phase grounded; B and C were fine. The top and bottom bushings then were disconnected and they were fine; the ground was in the bars. The three sets of A phase bars were disconnected; two revealed grounds. Upper end shields then were removed.

Next, the field was removed and an El Cid test performed on the stator core. An acrid smell from an undetermined location was revealed. GE recommended a full rewind after finding grounded-bar insulation cut at the end windings; this was attributed to inward migration of the outside space block (OSSB).

The speaker continued by outlining the steps to replace the OSSBs, which included loosening the belly bands, removing the compression ring, replacing the OSSBs one at a time, reassembly, etc. To mitigate a recurrence of the failure, the following OSSB actions were implemented:

• Increase torque on compression nuts from 2000 to 2500 ft/lb.

• Specify a finer finish on the OSSBs.

• Add dovetail retaining devices in previously unused dovetails.

• Add a punching with master bond adhesive.

• Put a rounded edge on the compression ring.

Final segment of the presentation was a detailed account of the generator rewind using a series of about 20 photos; view them on the Power Users website. One suggestion for projects of this type is to be sure there’s sufficient deck space before you begin. At this plant, every bit of space on a generously sized turbine deck (built out after commissioning), plus additional scaffolding, was required to do the job. Another recommendation: Take care of your field by storing it out of the weather; keep it warm and dry using heaters and dehumidifier and a burlap cover. Megger weekly. Access presentation.

Switching from Fyrquel to EcoSafe® 46 reduces maintenance, benefits personnel safety

A plant manager responsible for multiple 2 × 1 combined cycles in a desert location subject to periodic heavy rains and winds presented on several steam-turbine (ST) improvements. Most of his podium time focused on the installation of doghouses on the collector ends of the ST generators, and conversion from Fyrquel hydraulic fluid to EcoSafe® 46. New platforms for lube-oil/hydraulic skids to facilitate equipment access, plus sample-panel improvements, also were discussed.

Water intrusion at the collector end of one unit initiated a field ground which triggered the field ground detector causing the field excitation breaker to open. Low-cost remedies were tried first, but tape and other barriers installed to prevent water intrusion were not successful and a doghouse was installed over the collector end to achieve the desired outcome.

Experience with the original Fyrquel phosphate ester electrohydraulic control (EHC) fluid was not satisfactory. It has poor hydrolytic stability and reacts with water in the oil to reduce the fluid’s effective lifetime. The hydrolysis reaction, the speaker said, is not rapid at “normal” ambient conditions but accelerates as temperature increases (recall that this is a desert location) and is catalyzed by the presence of strong acids. Given a hydrolysis byproduct is phosphoric acid, the reaction is said to be autocatalytic (self-sustaining) and the reason for the neutralization filter loop installed in EHC systems.

EcoSafe EHC was said to be relatively stable and not subject to the degredation problems affecting phosphate ester fluid. In addition, the reaction products of EcoSafe EHC are weak organic acids, which are less aggressive than the acids formed during hydrolysis of phosphate ester. EcoSafe also contains beneficial corrosion inhibitors.

In addition, the speaker reminded that Fyrquel is a known carcinogen and removal from the site has a safety benefit. Also, he cited EcoSafe as being less expensive than Fryquel. Access presentation.

How to reduce steam-turbine outage time

The competitive nature of the generation business today demands relentless cost-cutting. One area power producers focus on is reducing outage time for inspection and overhaul. A user offered attendees several ideas on how to shorten a D11 outage based on his company’s experience on three comparable outages involving a base scope of work with the following elements:

• Replace N2 packing box.

• Fix dished diaphragms.

• Inspect and repair steam-turbine valves.

• Conduct a steam-path audit.

• Re-align the steam path.

• Inspect and replace steam-path seals.

• Inspect and repair bearings and lube-oil pumps and motors.

The first of the three outages was conducted at a regulated plant in fall 2011 and lasted 60 days; the second at a deregulated plant in fall 2014 took 40 days (and this also included replacing one row of bucket covers); the third at a deregulated plant in spring 2015 took 30 days. Such progress piqued the interest of attendees.

Experience and better and more rigorous planning played a major role in the company’s achievement. To begin, the same, highly experienced steam-turbine (ST) outage manager oversaw all three outages. Planning was conducted in great detail and involved experienced participants from both the owner and the OEM. This gave the owner a leg up on getting a commitment for outage support by one of GE’s most highly regarded site technical teams. Details, details, details: In one planning meeting involving the owner and GE, the speaker said the outage schedule was reviewed on an hour-by-hour basis.

An aggressive, confident plan in hand for the second outage, the owner purchased diaphragms in advance of the plant shutdown for ST stages having a high risk of dishing. Diaphragms removed were sent to a third-party vendor for restoration and stored pending re-use in the turbine serviced during the third outage.

A third-party services provider, equipped with a portable machine shop, was selected to supply and install gland and inter-stage packing and spill strips as needed. An alternative provider also was selected for tops-off laser alignment of steam-path components. Yet another third party provided a vertical boring mill to cut the radial root seals to the correct diameter and the steam-seal face to the proper thickness based on as-found condition.

The ST audit focused on performance losses attributed to both the thermal and structural condition of steam-path components—such as losses caused by damage to rotating and stationary blades, tip seals, packing seals, erosion, deposits, excessive clearances, etc. Recoverable losses include excessive seal wear, trailing-edge erosion of blades, and foreign-object damage. For each recoverable loss identified an estimated repair cost was calculated and an economic evaluation identified which repairs were cost-effective.

The speaker noted the value of diaphragm storage racks which the company had custom-built to facilitate NDE work and grit blast cleaning, optimize floor space, and minimize the need for cranes to move around diaphragms during inspection, cleaning, and repair tasks.

Installation of a temporary freight elevator from outside the ST building to the turbine deck reduced labor man-hours and provided a safety benefit over the use of stairs. Plus, spares were provided for critical valve assemblies to eliminate the need for conducting valve inspections and rebuilds during the outage.

Not following the normal practice of flipping over the HP shell also saved time. All diaphragm removal, stud extraction, NDE, and casing repairs were performed with the casing top side up. GE used a robotic tool to remove the diaphragms in the upper half; it offered both time and safety benefits.

Bonus/penalty arrangements are beneficial in most cases. The speaker suggested bonus payments be contingent on having no OSHA recordable violations, and no startup issues—including steam and oil leaks—as a result of poor workmanship. Access presentation.

Work at getting quality intel before planning your next outage

An engineer representing a non-utility generator operating eleven D11s with CODs extending from 2000 to 2013 told the group that two of the units have completed second majors; four more STs are expected to have their second majors in the next few years. General fleet-wide experience was presented first: Water induction reported at a few sites has been attributed to failed attemperators, drains, and instrumentation. Plus, piping-system stress resulting from balance-of-plant ageing issues has caused D11 vibration problems, particularly during hot restarts.

Specifics regarding the two units with the most operating hours: Site 1 has no record of water induction ever occurring, but piping hang-ups during expansion/contraction were cited as the cause of high ST vibration. Greasing of slides reduced vibration from 8 to 3.5 mils.

At Site 2, plant personnel reported an increase in vibration in the high-/intermediate-pressure (HIP) section of their D11 during a startup in 2012, then reported they had experienced issues in the previous month or so which translated to two step changes in vibration level. A data dive was initiated and assumptions made based on unit history and experience to guide inspection planning. Some assumptions were correct, the speaker said; some were “educational.”

Decisions made during the planning process included the following:

• Purchase buckets for HP Stages 1-3 and have them on-hand before opening the unit.

• Have on-hand an N2 packing box of the latest design, plus IP stages 12 and 13.

• Make arrangements for bucket replacement, possible straightening, and balance.

• Expect major diaphragm repairs will be required for five stages.

• Prepare for a 60-day outage.

Inspection findings displayed in several photos reveal severe wear and tear of internal components.

Lessons learned and experience gained during the second major at Site 2 included the following:

• HP stages 1-5 had to be replaced, not just 1-3 (look at the photos and you’ll see why).

• Major diaphragm repairs were required on Stages 1-13. Hindsight: It would have been less expensive to purchase new components before the outage, indicating the value of planning based on quality intel.

• All piping to the HIP case was disconnected, realigned, and rewelded.

• The actual outage ran 62 days.

• Vibrations started to increase when the 2 × 1 unit was put back into operation and cycled in a 1 × 1 configuration. The pipe hanger system is being evaluated for additional corrections. Access presentation.

Finding the source of valve issues can challenge most experienced turbine engineers

An engineer responsible for the owner/operator’s steamer fleet discussed experience regarding the failure of a main-stop-valve (MSV) pressure-seal-head (PSH) gasket. The background: Unit was commissioned in 2004 and an issue-free valve minor was conducted five years later. Several months after that, the right PSH was found leaking. A major in spring 2012 revealed PSH gasket damage and the left and right gaskets were replaced. A month later, the unit experienced a stop-valve nut failure and gaskets were replaced.

Less than a year later, the right PSH was leaking again and the drain pipe below the valve seat had ruptured. Mechanics found that both the left and right gaskets had failed and replaced them. Only weeks later, the right PSH was leaking yet again. In February 2015 an outage was taken to replace valve internals and modify stem leak-off piping. There have been no further issues to date.

A root-cause investigation revealed the pressure setpoint for the HRSG IP section was too high, causing the MSV to throttle to less than 20% open. The result: Severe turbulence within the valve.

In addition, the stem leak-off piping was routed to the drain tank rather than the steam seal header as the OEM’s design handbook specified. Failure mechanism: During startup, the drain tank would overfill and condensate would back up to the valve through the leak-off piping causing the soft-iron PSH gasket to fracture.Access presentation.