Discussion sessions at Frame 6 meeting stimulate interaction among users

Discussion sessions, chaired by members of the steering committee, run about an hour each. At previous meetings, discussion leaders began at the compressor end of the machine and moved aft. The order was changed for the 2016 conference because all the discussion time often had been consumed before the “small stuff”—instrumentation, for example—got air time. This year, the lineup was: auxiliaries, generators, and excitation; instrumentation and controls; compressor section; combustion section; turbine section. What follows are some of the highlights captured by the editors:

Controls. Chevron Corp’s Mike Wenschlag, the steering committee’s go-to person on control systems and related equipment, put up on the screen a long list of possible topics and said something like “Pick one and we’ll talk about it.”

Discussion began with chatter on the types of control systems, from Mark IV through VIeS. Based on a show of hands, a couple of plants were equipped with each of those “bookend” systems, but the most popular product for this group was the Mark V. However, there also were several Mark VI and Mark VIe owner/operators in the audience.

The VIeS designation was not recognized by a few in the room, so one attendee offered that the “S” was for “safety.” This system, he continued, is a safety management system engineered for enhanced cybersecurity in critical process applications—such as plant emergency shutdown, burner management, critical process control, fire and gas detection, and turbomachinery safety.

Another attendee said safety critical protection devices—such as an overspeed trip—are designed with a SIL rating (Safety Integrity Level) to ensure a level of protection against “tampering” that might disable a safety feature or change settings for when the safety device initiates “this or that.”

Technical Information Letters (TILs). A particularly helpful portion of Wenschlag’s overview was a checklist of TILs pertinent to Frame 6 operations issued by the OEM since the 2015 meeting. Anybody can miss something of importance, but this allowed everyone to catch up to the same level. Here’s the checklist the discussion leader presented:

      • 1420-2R1, “Lube-Oil Logic Enhancement,” Oct 30, 2007.

Purpose: To change the Speedtronic control panel to protect turbine bearings following a plant AC power loss by maintaining continued operation of the DC lube-oil pump.

Reason for revision: Added logic block with 30-min timer. Specifically, the new control software

Extends the operation of the DC lube-oil pump by 30 minutes following a turbine trip caused by loss of AC power or loss of lube-oil pressure.

      • 1963, “Mark VIe ControlST Firmware Update Notifications,” May 6, 2015.

Purpose: Notify users of impactful recent ControlST releases.

      • 1968, “Inlet Guide Vane Trip Solenoid Compliance,” Aug 11, 2015.

Purpose: To notify select users of a non-compliant electric trip solenoid installed on certain Woodward IGV actuator assemblies.

      • 1939-R1, “Improved Mark VIe Network Switch Availability,” Oct 16, 2015.

Purpose: To inform users of the availability of an alternative IONET network switch.

Reason for revision: GE-IP has stopped working on failure analysis of N-TRON switches for tracking. No need to ship the failed N-TRON.

      • 1881-R2, “Network Security TIL for Mark V, VI, and VIe Controller Platforms, Jan 21, 2016.

Purpose: To advise sites of new recommendations to help improve control-system robustness to potential cyber attack.

Reason for the revision: To update the reference documentation to reflect migration of GHT-200042 to GEH-6808.

      • 1988, “Z420 HP Hard-Drive Failures, BIOS, and RAID Updates,” Feb 16, 2016.

Purpose: To notify customers of a potential issue with Z420 HP engineering workstations running BIOS versions before 3.87 and INTEL® RAID drivers before TIL describes the problems reported by some customers and the recommended actions.

Compressor section. J C Rawls, an engineer at BASF Corp’s Geismar facility, put up a list of possible discussion topics from drift eliminators on the air inlet house through the compressor—including how to deal with practical operational issues such as icing and hoar frost.

One of the discussion points concerned getting filter-replacement intervals to match the hot-gas-path schedule—now at 32,000 hours (four years for many gas turbines in this fleet). One question in the users’ minds: Should HEPA filters be installed to achieve this goal? An attendee championed the idea that each plant should put in place a testing program and do the engineering to ensure the optimal choice of filters for its location.

Interestingly, the Frame 6 group has had several presentations on HEPA filters in the last few years. Previous discussion sessions reflected strong support for hydrophobic HEPA filters from several users. Not so many acolades this year, perhaps because attendance was down by about 20%: The oil-price collapse froze travel budgets at many of the refineries and process plants which rely on the 6B engine.

There was a brief discussion on compressor efficiency and the importance of maintaining a clean compressor. Attendees were reminded that a 1% loss in compressor efficiency actually reduces generator output by 1 MW. Everyone attending a 6B meeting should know how to calculate compressor efficiency by following the straightforward methodology presented by discussion leader Rawls at the 2013 meeting.

Combustion section. John Vermillion of Atlantic Power Corp led this discussion session. Remedies for overheating of the combustion compartment and exhaust plenum was one topic. A user reported that the problem at his plant was caused by a “huge” crack in the plenum area. A recommendation for avoiding overheating headaches: Replace aft flex seals and associated insulation at each major.

Other discussion topics included experience with 32,000-hr maintenance intervals, DLN tuning, and primary re-ignition in DLN combustion systems. Regarding the first point, an attendee said there was a tradeoff between maintenance interval and power. More specifically, if your unit is running at 2042F you likely will make the 32K; if running at 2084F to maximize output you will not make it to 32K based on his experience.

Turbine section. Attendees had to wait until Thursday morning for the turbine roundtable discussion, led by Huntsman Corp’s Zahi Youwakim. Fit-up of replacement transition pieces (TP) to first-stage nozzles was the first item discussed. It is wise to verify fit-up four to six weeks in advance of the outage, participants were told, so if there’s work to do it won’t impact the outage schedule. Full engagement of TPs to first-stage nozzles is recommended to avoid the adverse performance impacts of leakage. Be sure to check the floating seals.

First-stage shroud blocks come without cooling holes drilled, attendees were reminded. You have to drill holes after marking their locations with the blocks in place. The axial and radial holes should be drilled in a qualified shop.

A caution on first- and second-stage buckets: Flow test after repairs to identify cooling holes plugged as a result of grit blasting or other reason. You don’t want to fail buckets because of insufficient cooling.

There was a significant discussion around the lifecycle of first-stage buckets. It began with the observation that the OEM’s change from a 24K to 32K hot-gas-path (HGP) cycle is not in synch with today’s typical maintenance cycle for strip (stripping of the coating), HIP (hot isostatic pressing, a heat treatment for restoring material properties), and replace (coating).

Do you do this every HGP? If so, you’ll likely need an extra set of buckets on the shelf. A user said the GT-33 coating was much improved over the GT-29 coating and, in his opinion, you can make the 32K interval with it. The alternative is to just run the parts for 48K or 64K and toss. The optimal lifetime for your engine is an economic decision involving risk analysis. The user cautioned that every repair cycle removes some base metal and this has to be factored into your analysis. If the trailing edge gets too thin, he noted, you could be getting into problems.

Nozzles were another discussion topic. It began with a user saying the OEM expects 72K on first-stage nozzles; 96K on second- and third-stage nozzles. The current alloy for first-stage nozzles was said to be GE’s FSX-414, a cobalt–base alloy cited for its superior strength at very high temperatures. GTD-222, a nickel-based alloy, was developed by the OEM to provide improved creep strength in second- and third-stage nozzles; plus, it’s weld-repairable.

TIL 1523, “Exhaust Temperature Modifications,” spawned another discussion topic. Attendees were in general agreement that if an exhaust thermocouple (t/c) fails it should be jumpered out. At this point, a user interjected that third-party thermocouples installed at his plant are performing better and lasting longer than the OEM’s. That certainly was of interest.

Back to the main subject. If you jumper out a failed t/c, be sure not to make the connection to the same processor as the surviving t/c for that can. Recommendation to the attendees: Have a procedure for jumpering to avoid creating a trip condition and when you come down you can quickly change out the t/cs bypassed. Also, move the junction box to an area unaffected by heat.

One attendee said he wouldn’t jumper out more than two t/cs; he’d either live with the condition or shut down the engine to replace the defective instruments. A couple of users said they change out t/cs every HGP and major, acknowledging that there is an infant-mortality risk in doing this.

Posted in Frame 6 |

Comments are closed.