Case study #2: Excessive HP to IP leakage in a 500-MW supercritical Westinghouse combined HP-IP turbine
This case is more typical of excessive leakage that goes undetected. Pre-outage temperature split testing of this unit in 1992 indicated a very high level of IP dummy leakage (7.6% of hot reheat flow) and similarly high total HP to IP leakage of 7.4% of hot reheat flow (design is 5.4%). Since it is impossible for the IP dummy leakage to be higher than the total leakage, the absolute accuracy of the IP dummy leakage was called into question. A steam path audit calculated IP dummy leakage of 3% through a distorted ring. Distorted IP dummy rings are a generic aging problem for the BB44 design.
The distortion of the IP dummy ring was "fixed" during the 1992 outage by installing new seals and boring them out off-center. Outage inspections also revealed that a broken first-stage drain pipe and worn HP dummy seals were contributing to the high leakage. The bell seals were in good condition.
Unfortunately, post-outage testing, in 1993, produced worse numbers for IP dummy leakage (13.4%) and total HP to IP leakage (14%). The unit experienced some thrust balance problems after start-up that may have contributed to the higher leakage. Computer models indicated that the leakage raised the unit's heat rate by anywhere from 32 Btu/kWh to 109 Btu/kWh, depending on its source (first stage or cold reheat). The unit was retested in late 1993 and then again in late 1996. Both tests indicated that both the IP dummy leakage and total HP to IP leakage had decreased somewhat over time, which could not be explained.
Since the distorted IP dummy ring was a known problem, the plant purchased a "modernized" BB44M IP dummy ring from Westinghouse and installed it during a 1999 outage.
Inspections during the 1999 outage revealed some severely worn IP dummy seals (Figure 11) and some with rows missing. The HP dummy seals were also in poor condition, with numerous teeth broken off by axial movement. The LP dummy seals also were worn, and two leaking bell seals were found as well. However, a pre-outage test had not been conducted. If it had been, the total HP to IP leakage would have been much higher than reported by the last readings, in 1996. The dummy seals and worn bell seals were replaced when the IP dummy ring was upgraded.
11. Age takes its toll. A 500-MW supercritical Westinghouse BB44 turbine with a distorted and damaged IP dummy ring seal. Courtesy: Southern Company Generation
The 1999 post-outage test revealed tremendous improvements in IP dummy leakage (1%) and total HP to IP leakage (5.4%) to values near design. A test repeated one year later indicated that both leakage rates had remained near design. Without a doubt, the new IP dummy ring had worked as advertised. The apparent values of IP efficiency, which had been high due to the IP dummy leakage, returned to more normal values. Unfortunately, during the 2004 outage, the replacement IP dummy ring was found to be distorted. Recent testing indicated that the IP dummy leakage rate (4.4%) and total leakage level (7.0%), though not quite as good as the 1999 results, are overwhelmingly better than those found by the 1993 tests.
Case study #3: Excessive HP to IP leakage in a 180-MW Westinghouse combined IP1-HP-IP2 turbine
This case also is typical of excessive IP dummy leakage that goes undetected. For many years, temperature split testing of this unit indicated very high levels of IP dummy leakage (5% of hot reheat flow) and total HP to IP leakage (8.6% of hot reheat flow). Measured values of IP1 and total IP efficiency were always higher than design, indicating significant leakage into the IP1 turbine inlet. The seal clearances of the IP dummy and the horizontal joint of the IP dummy ring, and the fit of the IP dummy ring into the outer cylinder had always been found acceptable. Finally, the mystery was resolved during a 2005 outage inspection. Three 1.25-inch holes were unexpectedly found in the IP dummy ring: one on top and two in the lower quadrants (Figure 12).
12. Size matters. "Extra" IP dummy ring holes (arrows) for cooling steam on a 180-MW Westinghouse IP1-HP-IP2 turbine. Narrowing them reduced HP to IP leakage rates. Courtesy: Southern Company Generation
A review of the turbine cross section and of detailed IP dummy drawings revealed that the holes were supposed to be there. Flow calculations indicated that these holes for cooling steam were the source of the excess IP dummy leakage (a significant 40,000 lb/hr). After discussions with Westinghouse, SCG decided to reduce the size of the holes to 0.4375 inch by welding and drilling. Post-outage testing confirmed that the work lowered the measured IP turbine efficiency by lowering its HP to IP leakage rates (to 1% for the IP dummy, and to 5.4% for total leakage). On a similar, smaller unit at this plant, the holes in the IP dummy ring were completely welded up in the 2007 outage because a sister unit at another plant did not have the extra cooling holes.
Author's note: At press time, SCG was investigating a sudden rise in thrust bearing temperature in a 125-MW IP1-HP-IP2 turbine. Further review of the incident indicated sudden increases in measured IP2 efficiencies (one to a heater extraction and one to the LP crossover), along with lower reheat pressures and higher pressures in the IP2 turbine. The problem was determined to be a broken equilibrium pipe between the IP1 exhaust and the IP2 inlet. Unit full load is restricted until an outage can be taken to correct the problem.—Warren Hopson, PE (whhopson@southernco.com) is a Southern Company Generation principal engineer for generating plant performance.
Email
Comments
Print
Reuse
