Sharing of experiences by users a hallmark of STUG meetings

Repairing a steam leak at the horizontal joint

Repairs of leaks at casing joints in gas and steam turbines are a frequent discussion topic at user-group meetings. At STUG2016, an owner/operator shared his experience with a steam leak at the horizontal joint of a D11. The leak, on one side of the HP/IP machine in the vicinity of the first two HP stages, was noticed a couple of months after completion of a major inspection. Leaking steam tore holes in the insulation.

The speaker said the single-casing HP/IP is prone to distortion at the horizontal joint because of stresses associated with frequent starts and external piping loads. He added, “After an outage, the casing horizontal joint is cleaned of blue blush which opens path for steam to escape. Sometimes the leak will stop itself when blue blush builds up again in the joint.” Other causes of a casing flange leak include contaminants on the joint, broken casing stud, and insufficient torqueing of one or more studs.

If the leak persists, the owner/operator has two primary options: opening the machine, identifying the issue, and taking corrective action, or installing an external patch. The decision should reflect the results of a close examination of the leakage area by an experienced engineer after insulation has been cut out and the joint cleaned. Feeler gauges are among the tools used to characterize the leak path. 

Opening the machine has its risks and financial downside, of course. However, if this is the path taken, attendees were told it’s important to be sure nothing is hanging up the shell and there are no bolting issues. Then the focus should be on correcting distortion of the horizontal joint—the likely cause of the leak.

This is a significant task. First, the upper shell, rotor, and diaphragms must be removed, and the joint mapped with a laser or optical equipment. Next, a special self-leveling milling machine must be brought onsite to cut the horizontal flange. The cut complete, all diaphragms, packing glands, and the casing must be realigned. A boring bar also may be needed to establish casing fits for diaphragms and packing glands.

An external joint patch was the favored option for the case described by the speaker. The half-pipe patch for this machine was designed by an industry consultant who said D11 horizontal joint leaks have become more common because of the single shell design and the age of the fleet. He noted that the half pipe was developed two decades ago and has been used successfully on many steamers, including D11s. It allows for sealing uneven casing-joint flange sidewalls and for the injection of Furmanite later if the patch does not hold.

Steam-turbine issues since the millennium

A user with deep knowledge of steam turbines (STs) in combined-cycle service reviewed his company’s O&M experience with nearly two dozen units installed from 2000 to 2014. There are roughly equal numbers of GE, Alstom, and Toshiba machines.

The speaker began with a backgrounder on duty cycle and typical inspection intervals. The steamers, he said, originally were used as peaking units but now typically operate baseload with turndown at night. Inspection intervals: STs in combined cycles with AGP-equipped gas turbines, minor at about 32,000 hours and major at no more than 64k; no advanced gas path, minor at about 24k hours, major 48k.

The owner’s D11 steam turbines, all installed between 2000 and 2002, are managed by a long-term service agreement (LTSA) with the OEM. Historical issues mentioned: N2 packing head cracking, shell cracking/distortion, rotor bowing, stellite liberation from the seats of combined stop/control valves, and diaphragm dishing. The first three remain as current concerns.  

Regarding N2 packing head cracking, all eight of the owner/operator’s D11 steamers were affected. The worst casing crack, approximately an inch and a half deep, was found on the lower side wall of one unit. Possible causes considered were water induction, over-temperature operation, and material defects. Some casing cracks on other units were found in the heat-affected zones associated with casting weld repairs made at the foundry.

Rotor straightening was done when necessary. One rotor had a total indicated runout of 18 mils. Cracked stellite was found in valve seats of multiple units, one experienced a failure during operation. Attendees were referred to TIL 1629-R1, “Combined Stop and Control Valve Seat Stellite Liberation.” HP and reheat diaphragm dishing addressed by TIL 1589 is handled by replacing dished diaphragms during the second major (15 years).

Going forward, some plants with D11s are evaluating retrofit options—such as the OEM’s Enhanced Steam Path (single shell) and double-shell replacement.

For the company’s seven Alstom steamers commissioned between 2003 and 2008, the current concerns are air in-leakage and hydraulic turning gear. Historical issues also included IP rotor and vertical-guide-key shim migration, shrink-ring seat cracking, and HP-inlet thermal cracking.

The rotor shim-migration solution selected was to replace the first two IP stages with TurboCare’s shimless design with notch entry, redesigned blade root, and new blade material. The vertical-guide-key solution is described in Alstom’s Customer Information Bulletin 2DESER00089B01b.

For cracking at the shrink seats of HP inner casings, attendees were referred to CIB 2DESER00090B01. Cracking of shrink-ring seats and inlet-scroll blades can be mitigated, the group was told, with enhanced radii. For HP inlet cracks, the recommendation was drill-stop.

The speaker offered the following lessons learned for Alstom steam turbines:

      • Replace J-seals in the HP and IP sections at each major; replace LP seals as needed.

      • Monitor solid particle erosion at the inlet and inlet-scroll blades and replace all inlet-scroll Radex blades at the second major.

      • Suggestion to users with hydraulic turning-gear issues: Consider an electric turning-gear retrofit.

      • Sources of air in-leakage can be the LP gland piston ring and the exhaust-hood horizontal joint.

The speaker’s Toshiba fleet consists of one unit commissioned in 2003 and four in the 2011-2013 period. The latter are of a slightly different design. Current concerns were said to be valve sticking, likely corrected with refurbishment and upgrade.

Interestingly, the performance of units with abradable bucket tip seals was not as good as the unit without them. Curves shown to the group showed HP cylinder efficiency flat-lined at about 90% without abradable seals; with them it decreased over time (about five years) from perhaps 89% to 85%. IP efficiency with abradable bucket tip seals held steady about 93% over time, without them it was about 97% for the same three-year period.

HP casing repair

A 71-MW non-reheat steam turbine went into commercial operation in 1992 with inlet steam conditions of 1500 psig/950F. The first inspection/overhaul of this unit was in 2015. Two indications were found in the HP upper casing in the nozzle fit area. The crack on the right side of the casing was 18 in. long, running from the inner surface to a bolt hole, the one on the left side was 12 in. long and also ran from the inner surface to a bolt hole. Photos shown by the speaker were of high quality and can be seen by registered users by accessing the presentation on the user group’s website

First step after discovery was a root-cause analysis. Turbine operating data were evaluated for thermal excursions and excessive temperature ramp rates at the casing inner surface. Engineers found casing thermal ramp rates were excessive (as high as 424 deg F per hour) during six cold starts in 2012.

Repairs were conducted by the OEM at its manufacturing facility. Upper casing cracks were weld-repaired, with full stress relief afterward. Total schedule for repair was 25 days, including five days for hardware removal, but the work actually took 31 days. Back in operation, plant personnel were urged to follow the manufacturer’s starting and loading instructions. The steam-temperature ramp rate is limited to 100 deg F/hr during roll up to 3600 rpm, and 150 deg F/hr during loading.

D11 major

A user ran through a series of photos taken during the major inspection of a D11 that had experienced about 960 starts and operated for about 54,000 hours since going commercial in 2004. Photos were shown of valve strainer damage and foreign-object and domestic-object damage. The latter was significant, forcing the replacement of both first-, second-, and third-stage buckets.

Point of the presentation was to stimulate discussion, which it did. It also gave the speaker the opportunity to use the PowerUsers electronic polling system to get feedback from attendees. Here is some of the information shared:

      • Have you performed an ST major? Yes, 86%.

      • Did you replace first-stage buckets? Yes, 50%.

      • Did you perform major repairs to first-stage nozzles? Yes, 73%.

      • Did you have stop-valve stem damage? Yes, 55%.

      • Did you have damage to valve screens? Yes, 33%.

      • How often to you perform valve freedom tests? Daily, 50%; weekly, 43%; monthly, 7%.

      • For those performing daily valve freedom tests, what problems have you experienced? None, 64%; hydraulic issues, 18%; valve issues 18%. There were no controls-related issues and no problems left undetermined.

Posted in STUG |

Comments are closed.

Categories