Users may struggle with some issues for years before viable solutions are confirmed through fleet-wide operating experience of components re-engineered by the OEM. Coking in the D/E sump drain and D-sump scavenge line, located in the turbine rear frame (TRF) at the 6 o’clock position, is one of those. A review of editorial notes indicates coking has been covered in presentations and discussions in the LM6000 breakout sessions the last three years, at least. To summarize:

At the 2014 meeting (Palm Springs), one of the 2013 depot findings presented was “TRF D&E Sump Drain Coking”—specifically, moderate to severe coking in the No. 6-bearing D-sump oil-scavenge-tube outlet. SB LM6000-IND-244, “TRF Strut End Adapter Gasket Replacement” (Jan 1, 2014), explained that the drain hole could become blocked by the old square seal gasket.

A new gasket, with a larger center-hole diameter was introduced to reduce the risk of blockage. However, early adopters still were reporting moderate to severe coking after implementation of SB-244. The depots acknowledged more work was required to completely fix the issue.

They recommended that users follow, at a minimum, the inspection and cleaning intervals presented in Table 12-1a, “Recommended Preventive Maintenance and Servicing Checks,” of service letter SL6000-05-03 R3. The depot speaker suggested that the prescribed intervals might require adjustment on a site-to-site basis.

At the 2015 meeting (Long Beach), one of the 2014 depot findings presented was Round Two of “TRF D&E Sump Drain Coking.” The first slide in both the 2014 and 2015 presentations was the same, except for the photos. The Table 12-1a reference was repeated as well. But research into the root cause of the issue continued, with the following updates presented in 2015:

• Drain-line clogging, together with minor oil leakage into the TRF frame vent cavity from the sump, creates a layered buildup of oil deposits. The presenter attributed leakage of oil in the sump area to (1) a loss of sump pressurization during low power operation, and (2) hardening and embrittlement of O-rings in the joints of the sump components (Fig 1).

Later in the meeting, during an OEM presentation on the same subject, a GE engineer reported that coking in the units affect typically was observed after two to three years of service. He pinpointed the location of the O-rings of concern and showed some photos of O-ring damage (Fig 2).    

The speaker also mentioned another possible root cause of the coking issue: Insufficient sump drain-line slope such that some oil remained in the line and coked because of the high heat. Attendees were urged to follow the recommendations in product bulletin PB-303, “Turbine Lube-Oil Drain Tubing Inspection.” It ensures the slope for the D&E sump drain tubing from the TRF to the unit drain is greater than the 0.25 in./ft cited in the specification.

• Regarding the inspection and cleaning intervals, some users reported scheduling drain-line cleaning monthly.

At the 2016 meeting (Palm Springs), presentation of 2015 depot findings again included moderate to severe coking in the No. 6-bearing D-sump oil-scavenge-tube outlet. Information provided during the previous two meetings on this issue was reviewed. Progress since the 2015 meeting included release of SB LM6000-323, “TRF D- and E-Sump Preformed Packing Material Change for Improved Durability,” which changed the O-ring material for the speed probes to AFLAS® to mitigate the hardening attributed to the high-temperature environment.

Regarding the coking issue, a depot representative echoed the advice offered a year earlier regarding more frequent drain-line cleaning, saying “The only thing you can do is clean it, clean it, clean it!” Comments from attendees included the following:

• “We’ve had this problem. We had a blown gasket. We replaced the gasket and still had the issue.”

• “We also had this issue a number of times. We increased the diameter of the line and made the line shorter. Before it would be 8-10 months to see leakage again; now it’s been over two years with no leaks.” An attendee looking to clarify the comment asked if the machine was a peaker or baseload. It was the latter, which helps to mitigate the problem.

• “We see minor oil leakage in the TRF vent area. We ended up with a completely coked housing. O-ring hardening can be the cause of this.”

One user’s experience

A user case history closed the presentation/discussion of TRF coking—for this meeting, at least. Two engineers providing maintenance support for a European LM6000 fleet of more than two-dozen engines shared their experiences, beginning with typical findings:

• Oil coking at the 6 o’clock position.

• Oil droplets coming out of the sump drains at part load (not all engines impacted to the same extent). They called this a “bleeding sump,” suspecting one or more of the following seals for underperforming: (1) labyrinth seals, (2) No. 6-sump O-ring, (3) TRF vent-tube O-rings, (4) No. 7-sump O-ring, and (5) open-end dynamic seals.

The depot contracted to correct the problem found coking at all but the labyrinth seals. Further, that the seals were hard and brittle with sections missing.

Here’s what this fleet learned from its experiences:

• Coking is most severe in engines having the most part-load operating hours.

• At baseload, the VBVs are closed, sump air is at nominal pressure (so-called P25), and the air pressure in the frame exceeds that in the sump.

• At part load, P25 is lower and sump pressure exceeds the frame air pressure.

• Suspected leaking seals are the cause of the bleeding sump, the positive differential pressure across the oil seals having been compromised.

The road to a solution. The original O-rings were made of Viton®, which was limited to 400F service. First replacement material was AFLAS, as noted earlier. However, it was not recommended for applications over 450F. The users presenting said AFLAS was installed in some of their engines and the O-rings degraded after about 25,000 operating hours.  

This owner/operator’s engineers pursued an independent solution. The choice was Kalrez® 7075, good to 620F, which was developed for chemical and hydrocarbon sealing applications. The jury is still out on this material for LM6000 use, but the speakers said they would update the group at the 2017 meeting.

In addition, this owner/operator found open-end seals installed with the open-end out, rather than open-end in as one might expect—to keep oil from leaking out. The OEM said it didn’t matter how the seal was installed because the differential pressure was so low, and the open-end out orientation facilitated installation. The fleet owner decided on changing to open-end in, towards the sump, as the IRM suggests.

Another leak-prevention alternative: The OEM’s sump evacuation system (SES)—a/k/a sump sucker—described at the 2015 meeting. It is a good solution, said the users presenting, based on 4000 hours of operating experience. Only minor package mods were required to accommodate installation. These were described by the presenters, who suggested the SES might become obsolete if the leakage issue were resolved another way.

For those unfamiliar with the sump sucker, here are some of the system’s highlights:

• Consists of a motor-driven blower installed in the air/oil separator discharge line.

• Operates when the gas generator is running at low speeds and during load transients.

• Maintains the necessary differential pressure across the oil seals to avoid leaks during low-power operation.