GE Standards
GE Standards
GE Standards
GE Power Systems
ii
Detroit Edison
The Detroit Edison Company air-blow experience centered on the Belle River Power Plant, which consists of two 660-MW, subbituminous coal-fired, single reheat, steam drum units. Due to the high sodium content of the specified fuel, the Belle River Power Plant furnaces were designed to be extra large in comparison with similar Eastern Coal units. Since this plant was being constructed and placed in service during a time of economic challenge, the overall plant reliability and cost effectiveness were of major importance to Detroit Edison. Starting with the question of why steam-blow at all?" Detroit Edisons Belle River Start-up Team researched the steam-line cleaning techniques and criteria of the major electric power companies of the United States. Most companies that were contacted utilized steam- blows as their steam-line cleaning standard. Several major companies, however, were found to be utilizing air as their cleaning standard. Those that used only steam-blows had no scientifically based rationale for these standards and actually indicated that the standard was based primarily on tradition. Since it was a known factor, they saw no reason to change. The utilities using air as their cleaning medium, did so for a variety of reasons: 1. They did it traditionally. 2. They had found air-blowing to be bet-
Figure 1. Belle River air-blow compressor layout. Note areas barricaded during blows. GE Power Systems GER-3636A (08/03)
I I
Basin Electric
Background. The Basin Electric Antelope Valley Station (AVS) consists of two 440-MW units. The lignite-fired Combustion Engineering superheat and reheat boilers supply steam to Westinghouse turbines. Main steam piping on AVS Unit 1 was steam-blown, while compressed air was used on Unit 2. The general feeling is that the air-blow on Unit 2 was more effective than the steam-blow on Unit 1. The initial concept of using compressed air to clean the steam piping of AVS Unit 2 was suggested by a representative of Ingersoll-Rand (IR). The main advantage appeared to be the ability to remove the initial steam-line cleaning from the critical path, although in the end, Basin Electric used the air-blow for other reasons, mainly economic. At first, we were skeptical, citing reasons typical of those who have performed the traditional steam-blows. A list of utilities using air-blows as a replacement for steam-blows was supplied by I-R for interview purposes. The utilities contacted were Allegheny Power System and Detroit Edison. No negative comments were encountered, and we received valuable advice on improving the effectiveness of the air-blows. Fortunately, Detroit Edison was using air-blows at the Belle River plant at the time we were considering air-blows. Clarence Brookins invited Basin Electric to observe the air-blow operation. Based on those observations, the utility interviews, and projected cost savings, Basin decided to proceed with the air-blow rather than steamblows. Cost Savings. There was no need for an earlier start-up date on AVS Unit 2 since Basin did not need the additional capacity sooner than originally planned. Thus, there was no advantage in starting the unit early. There were direct savings in the performance of the air-blow as compared with a steam-blow. Since a steam-blow was used 7
taken to reduce the steam-line swing was decreasing the blow valves opening rate, which may have led to another problem. In a short time, the valve, an 18-in. ball valve, started to operate slower and slower. Finally, higher pressure air from a portable air compressor was used to actuate the valve. A possible cause of the increasingly difficult actuation may have been that the slower rate exposed the ball sealing surface to the debris-laden flow. Damage to the ball surface would have increased the drag on the seats. While the ball of the valve had a blasted appearance, the damage did not appear to be the absolute cause of the valve actuation problems. The trunions were not removed and examined. Consideration for a high number of cycles and seal surface roughness should be built into the ball valves in the future, if possible. There are two indications of the air-blows effectiveness. After a 2-wk operating interval, Unit 2 was shut down to examine the screens ahead of the Main Steam stop valves. Observers indicated the screens looked good with no sign of debris or pluggage. The screens were removed, and the unit returned to service. Sometime later, Unit 2 was again shut down for a routine outage, and the high-pressure shell was removed as scheduled. The turbine blading was found to be in good condition. The general feeling is that the air-blow of Unit 2 had been more effective than the steam-blow of Unit 1. The preoperational cleaning sequence was: 1. Air-blow of superheater, reheater, and steam lines. 2. Hydro. 3. Boil out of drum and waterwalls with caustic and trisodium phosphate. 4. Operate at full load for 5 weeks. 5. Chemical clean drum and waterwalls with Dowells vertan. 8
AVS Unit 2 Steam-Blow Max. Drum Pressure Circuit Main Steam and CRH Hot Reheat BFPT HP Steam BFPT LP Steam (psig) 800 800 800 150 Specific Volume v (ft3/lb) .23 .23 .23 1.23 Max. Exit Nozzle Pressure (psig) 40 20 60 20 Max. Mass Flow Q (lb/hr) 3,140,000 3,140,000 124,000 233,000 Momentum Term (Q2v) 2.27*1012 2.27*1012 3.54*1012 6.68*1012
Circuit Main Steam and CRH Hot Reheat Steam BFPT HP Steam BFPT LP Steam
Blowing Time 17.4 sec 29.3 sec 12.4 sec 18.6 sec
A momentum ratio greater than one ensures that the mass velocity during cleaning is greater than that developed at the maximum flow condition. Therefore, it will be unlikely that any debris not removed during cleaning will be blown into the turbine during operation. R where: R Qv Qc
2
ments of the pipe to be cleaned is greater than one. In order to obtain a minimum ratio of 1.2 at the critical point in the steam line, the maximum momentum ratio may range from 2 to 5 or higher, depending on the configuration. Experience with Steam Blows. Several variations of steam-line cleaning have been effectively used by the GE Power Plant Engineering group. The common criterion has been the requirement to select conditions so that the minimum momentum force ratio is greater than one by some margin. The standard steam-blow was done with saturated steam. Boiler pressure was increased to the selected initial pressure, firing was terminated, and the blowdown valve was opened as quickly as possible. The steam initially would be slightly superheated, but would rapidly become saturated since firing was terminated. Water in the boiler flashed to steam as pressure decayed. The blowdown valve would be closed at a pressure corresponding to the maximum change in saturation temperature allowed by the boiler vendor (usually 75F). The process would be repeated until targets showed the lines were clean. In this approach, the required cleaning conditions exist only during the initial part of the transient, while the largest part of the transient would be a flushing operation.
= Q2c vc/Q2max vmax = momentum or cleaning force ratio = momentum term = calculated flow during cleaning (lb/hr) = specific volume at pipe segment inlet during cleaning (ft3/lb)
vmax = specific volume at pipe segment inlet at max steam flow (ft3/lb) The momentum term in the denominator is defined by the design conditions. The variables in the momentum term in the numerator are selected so that the ratio is greater than one. Table 1 shows that similar cleaning forces were obtained for steam and air-blows on twin units. A lower specific volume requires more flow to obtain the same value in the numerator. The same cleaning force can be obtained with air, steam, or other cleaning medium. Initial conditions are selected so that the minimum cleaning ratio at the inlet to various seg-
10
TEPCO, CHUBU, and TALKHA are combinedcycle projects where the steam-blowing time ranged from 3 to 20 min. Cleaning time is limited by the amount of demineralized water available. One extended blow at the desired cleaning condition is equivalent to many standard blows where the cleaning condition exists for several seconds, and therefore, will significantly reduce the total cleaning time. Experience with Compressed Air. The motivation to use air-blows was a compressed schedule and the need to minimize the risk associated with bonus/penalty construction clauses by removing the steam-blow from the critical path in the start-up cycle. Bob Kelety of Tidewater Compression, the supplier of compressors for the following jobs, circulated a brochure and references advertising compressed-air cleaning of main steam-lines. A reference check indicated that compressed air-blows should be seriously considered because: 1. A theoretical comparison of the potential cleaning ability of steam versus air was made by GE's Steam Turbine Thermal Engineering group in 1984. Using an air temperature of about 120F, in accordance with information obtained from electric utilities, the groups analysis showed that, for the same initial boiler pressure, the cleaning force (momentum ratio) is about the same. 2. None of the electric utilities contacted observed any indication of abnormal turbine erosion problems over a period of time. 3. Field engineer comments quoted stated that the condition of the temporary strainers was as good as the best observed for steam-blows and better than some. (A good indication of effec-
11
ratio of 1.2. After 6 mo of operation, which included 20 to 30 brief excursions to 120% power, 72 hours at 105% power, and many startups and shutdowns, the turbine screens were clean. Boroscopic examination of the first-stage buckets by the GE service department (and witnessed by the customer) showed no indication of wear. This result is consistent with industry experience in that the Start-up Manager, Andy White, stated: The screens were as clean, if not cleaner, than screens observed in 10 steamblows." Thus it appears that effective compressed air-blow cleaning can be done at minimum momentum ratios near 1.0. ANR is a gas/oil-fired STAG plant which consists of a Frame 6 GE gas turbine, a supplementary fired ZURN heat recovery steam generator (HRSG), and a 28-MW GE steam turbine. The main steam line and gas turbine injection steamline cleaning at ANR was done with nitrogen instead of air with the same result, an exceptionally clean screen. A Union Carbide Industrial Services Company (UCISCO) liquidnitrogen truck and pumper were rented instead of air compressors. The economic trade-off compares the high transportation charges of the air compressors with the cost per hundred cubic feet for nitrogen. Nitrogen blowing has an economic advantage for small volume systems. The capacity of the pumper allowed a continuous blow on the gas turbine steam injection line. Brian Palmer, the Start-up Manager at ANR, said, The nitrogen-blow of the main steam line and the gas turbine injection steam line resulted in substantial cost savings due to reducing the impact of steam line cleaning on the ANR commissioning schedule. Steam turbine inlet screens were inspected after a 30-day operational test at plant full load with no evidence of particle impingement or accumulated debris. COGEN TECH is a gas/oil-fired STAG plant, 12
3. Effectiveness: Inspection of the fine mesh screen and stop valve revealed one minor impact on the screen. The air-blow was very effective in both cost of operation and cleaning at REDDING and PERC. BURNEY FOREST is a duplicate of Redding. The steam line cleaning procedure was identical, but more blows were required. Temporary screens were removed January 1990. Warren Behrens, the Project Manager, reports that the compressed air-blow was quicker and less expensive, with results equal to or better than steam blows, with the added benefit of being completed earlier in the construction schedule. TBG COGEN is a gas/oil STAG plant with two GE LM2500 gas turbines, two supplementary fired HCG boilers, and a 13-MW GE steam turbine. After approximately 200 hr of operation, with slightly more than 100 hr at full load, Andy White, the Site Manager, stated the screens were beautiful and consistent with experience at PERC. Figure 4 shows the turbine strainer. Figure 5 focuses on the only trapped particle. Figure 6 displays a larger segment of the clean strainer.
Figure 6. Larger segment of clean TBG COEN strainer. FAYETTEVILLE is a gas/oil STAG plant with eight Frame 5 GE gas turbines, three VOGT HRSGs, and a 55-MW GE steam turbine. Six Frame 5 gas turbines feed into three HRSGs (2 GTs per HRSG) and the steam generated is delivered to the steam turbine. Two gas turbines operate simple cycle. Main steam lines were blown individually and flushed simultaneously. Individual blows were 18 on HRSG 2, 35 on HRSG 1, and 94 on HRSG 3. There were 32 simultaneous blows of HRSGs 1, 2, and 3. It appears likely that debris from HRSG 3 was pushed back into the other legs. The target following the simultaneous blows was accepted by the GE turbine erector and the Black and Veatch representative. Dave Buchyn, the mechanical engineer, observed the strainer during an early shutdown for valve repair said, "The
Conclusion
Experience over the last 20 yr on power plants ranging from 35 to 700 MW, with supercritical boilers, drum-type boilers, and heat recovery steam generators has shown that compressed airblows may be the most effective and economical approach to cleaning main steam lines. Benefits of compressed air-blows are: 1. Flexible scheduling because air-blows can be worked into the start-up schedule without affecting other major critical path activities, except for work on the boiler and steam-line pressure parts. Balance of plant equipment
Reference
GEI-69688E, Cleaning of Main Steam Piping and Provisions for Hydrostatic Testing of Reheater, 1989, General Electric Company.
15
List of Figures
Figure 1. Belle River air-blow compressor layout. Note areas barricaded during blows. Figure 2. Belle River typical debris collected from silencer. Figure 3. Belle River silencer during mainstream air-blow. Figure 4. TBG COGEN strainer. Figure 5. Particle trapped in TBG COGEN strainer. Figure 6. Larger segment of clean TBG COGEN strainer. Figure 7. Fayetteville strainer closeup. Figure 8. Fayetteville strainer. Figure 9. Fayetteville main steam stop/control valve.
List of Tables
Table 1. Performance data. Table 2. AVS Unit 2 effective air blowing times.
16