Document mpVEEdBp82Q2ad0rgvEkBg3pQ

later in the program These materia.1* ae^e subjected onl> to the constant emplace ment elevated temperature and long-term stress relaxation tests of Phase II. because of the timing of their release Replacing asbestos Screening tests All 33 candidate materials were subjected to a series of screening tests to determine steam path gaskets which of them qualified for more ngorous testing. The screening tests consisted of exposing precut tensile specimens and compressibility buttons to three environ Asbestos restrictions are tightening as deadlines draw near. Suitable substitutes for asbestos sheet gasket materials have been elusive, but one family of materials shows promise ments: 750 F/650 psig steam, 210 F wa ter. and 250 F mineral oil. Exposures lasted for 240 hours with specimens sup ported so that the environment contacted all surfaces. Materials were examined after they were exposed and Table I shows the By W F. Jones arid B. B. Seth, (1) inorganic fiber with a binder. (2) inor results on the better materials. Results Westinghouse Electnc Corp. ganic fiber with no binder and woven from the 750 F/650 psig steam exposures similar to cloth, (3) organic fiber (e g. proved interesting. Specimens disappeared Restrictions on the use of asbestos-con Kevlar) with a binder, and (4) flexible in the worst cases and others broke apart taining materials continue to tighten. The graphite with an insignificant amount of when touched. The better matenals speci federal Environmental Protection Agency binder. mens remained intact and they could be (EPA) has established a policy for elimi Fiber with binder and ceramic fiber handled and tested. The best matenals nating asbestos in non-cntical applications with binder materials became available appeared unaffected by exposure in 210 F (e.g., below 750 F) by August, 1993. Now, most engineers seek suitable substi Table 1. Summary of visual examination results of asbestos free samples. tute materials for applications where as bestos performed well over the years. In he case of sheet gasket materials, the larch for an asbestos-free substitute has -een more complicated than anticipated. During the mid-1980s, utilities began to replace asbestos sheet gaskets in lowpressure turbines (service temperatures below 750 F) with organic fiber/binder Material tested Organic fiber/binder Material 1 Material 2 Material 3 Inorganic fiber/binder Material 4 750 F Steam Good condition Good condition Good condition Intact/bnttle 210 F Water Good condition Good condition Good condition Good condition 250 F Mineral oil Good condition Good condition BJistered/swotlen Good condition sheet gaskets. The substitutions were Inorganic liber no binder based on gasket supplier data and recom mendations. As these asbestos-free gas kets accumulated service hours--normally about six months at temperatures above 550 F--they began to leak. The leakage was severe in most cases. This experience prompted the initiation of a research program in 1986 to qualify Material 5 Flexible graphite Material 6 Asbestos control White Black Ciotft type material Tom* Good condition Good/bnttle Good/bnttle Good condition Good condition Good condition Good condition Good condition Good condition Good condition Good condition an asbestos-free sheet gasket material. The key word here is "qual Table 2. Tensile, compressibility and recovery test data taken before and after exposure. ify," because experi ence showed that trials based on product litera ture and data were not Material tested 750 F Steam Tens* ComeressMSy V change % change Recovery % change 210 F Water Tensile Compres&Mty % eftange % change Recovery \cfiange 250 F Mineral oil Tensile CompressiMrty % change % change Recovery % change satisfactory. Organic fiber/binder Materials selection Phase t of the program sought candidate materi als. Suppliers of sheet gasket materials were given a requirements specification and asked Material t Material 2 Material 3 -94.72 Cracked -95.26 Cracked Fractured -10.0 Inorganic fiber/binder Material 4 Fractured 33.3 Inorganic liber no binder Material 5 No test No test 41 4 Cracked -6.1 444 No test *1.1 -4.4 105 3.3 No test 00 0.0 63 77 No test -1.10 -4 39 10.49 3.33 No test -5 2 23.0 -83 4 0.0 12.5 3.2 11.24 20.0 No test No test 196 26.9 NATT 32.7 No test to identify candidates that would meet the cri teria. As a result, 33 asbestos-free sheet gas ket materials entered the program and were clas Flexible graphite Material 6 11.27 Asbestos control White Black NATT NATT -25 25.7 29.0 -4.0 16.1 -75 16.10 44 24 5 NATT -43 -59.0 108 -59.0 215.1 0.0 -7.2 -6.4 -546 26.1 -54.6 -29.5 -10.9 11 sified into four groups: NATT . Not *bt tO IMt POWER ENQMCERtNOIIiARCH ltd 43 better materials. More details of these test results can be found in the paper "Evaluation of As bestos Free Gasket Materi als" by W. F. Jones and B. B. Seth, presented dur ing the ASME/1EEE Inter national Joint Power Gen eration Conference. Octo ber 21-25, 1990 (90-JPGO Pwr-58). The screening results identified the best materials for Phase II qualification tests. Primary criteria for selection, in this case, were based on the materials' ap pearance after the screening exposures. This was be cause very few of the 33 materials survived the 750 F/650 psig steam expo Figure 2. 750 F (trees relaxation test data. sure. The flexible graphite materials survived intact with no visible evidence of degradation. A few of the organic fiber/binder materi als also survived these screening tests based on appearance. Final selection of Phase II materials was based on tensile, compressibility and recovery test data. Those with the least loss in tensile strength after exposure held the most promise. Consid eration also was given to those materials that were hot affected by the mineral oil tests. Flexible graphite materi als and three organic fiber/ binder sheet materials were selected for testing. The or ganics were selected in stead of the inorganic fiber water. Only two materials appeared to be cloth materials because the cloth-type ma affected by the oil in 250 F mineral oil exposures. Tests for tensile strength, compressibil ity and recovery also were conducted after exposure and the resulting data were com pared to unexposed properties. See Ta ble 2 for the percentages of change in the terials would not seal vacuum applica tions. Further, the organic fiber/binder materials most closely duplicated asbestos sheet in handling and size availability. Phase II testing was considered the qualification portion of the program. The candidate materials were subjected to more evien-oe a,..L-service conditions Muteruis cere jected to standard compression, flamma bility and > and m factor tests, as well as specially designed tests Data from these standard tests indicated that the materials would perform adequately. It was the special tests that began to reveal the problems that account for the leaky service experience of asbestos-free materials. Long-term stress relaxation tests showed expected performance at room temperature iFigure 1). but all ex cept asbestos sheet exhibited significant load reductions at 750 F (Figure 21. The organic fiber and ceramic fiber materials showed the fastest load drops. However, the flexible graphite came closer to matching the performance of asbestos sheet. Special tests reveal leak causes Constant displacement elevated tempera ture tests also offered insight to the leaky service experiences of the organic fiber/ binder materials. In these tests, samples were loaded to typical flange conditions (3000 psi) at room temperature. Room temperature displacement was held con stant while the specimens were heated to 750 F. Load was monitored during heat ing. In all cases, gasket stress increased rapidly, but some behaved unexpectedly. The carbon fiber/binder materia smoked and oozed out of the test fixture Two of the organic fiber/binder materials pulverized and melted, a condition lead ing to leakage in service. The third organ ic fiber/binder material appeared to sur vive, but it was held together merely by its wire mesh insert. Only the asbestos sheet and the flexible graphite materials survived the test with satisfactory results. Table 3 summarizes the maximum stress es and visual conditions for materials sub jected to these tests. In the final part of Phase II. asbestos sheet and flexible graphite were subjected to simulated service tests that varied tem perature and pressure of die gasket envi ronment. The tests were run. but fixture difficulties prevented precise leak rate de terminations. However, the asbestos sheet leaked more than the flexible graphite sample. ' The results of the asbestos-free gasket qualification program through Phase II indi cated that the flexible graphite material ap peared to be the only candidate capable of substituting for asbestos sheet. However, the long-term stress relaxation behavior of the flexible graphite (Figure 2) indicated that leakage could be expected at 750 F in less than five years. As a result. Phase III of the program was initiated to develop data to predict usable service for flexible graph ite sheet. Usable service in this context is defined as the time required for the flange load to relax to the minimum flange clamping load. This is calculated in accordance with ASME Section VIII Man- ROWER ENGINEERING/MARCH 1M2 Thbfe 3. Constant displacement elevated temperature test data for materials tested. Material Maximum stress reached stress (psi)/ temp f Asbestos sheet 12.000 / 705 Organic fiber/binder 8480 / 714 Carbon fiber/binder 8475 / 729 Ceramic fiber/binder 11.100 / 746 Organic fiber/binder wire insert 13,120/745 Flexible graphite 10,200 / 750 ** Smofcdd and oond out o< ftxtur* during hMOng. Stress after S min at lemp stress (psi)/ temp F 13.040 / 750 8480 / 750 12.550/750 11.100 / 750 Visual condition of test sample alter test Intact, no distress Pulverized and melted Extremely brittle** Brittle 13.120 / 745 10,200/750 Intact, no distress Intact no distress Temperature. F Figure 4. Flexible graphite usable service prediction curves. datory Appendix 2. Stress relaxation tests In Phase III. long-term stress relaxation tests were started at 500 F. 600 F. 650 F. 700 F, 800 F and 850 F and allowed to run until minimum usable service load or 10,000 hours of exposure were reached. The 500 F, 600 F and 650 F tests contin ued to run to 10,000 hours. Figure 3 shows the long-term stress relaxation curves. The 500 F data are not plotted, because they show no relaxation. Usable service prediction curves (Figure 4) were generated from the data in Figure 3 by picking different amounts of stress relax ation. Utility survey Parallel with the long term stress relax ation tests, a utility survey was initialed to determine industry experience with asbes tos-free gasket materials. Thirty-three util ities representing 90 units responded to the survey. Of these 90 units, 46 continue to use asbestos sheet gaskets. The remain ing 44 units had tried or were currently using asbestos-free materials in some lo cations. The asbestos-free sheet gasket rowan moiwcsaiMOiiiAWCti tan materials identified in this survey included types made up of seven different organic fiber/binders: one carbon fiber/binder; four different flexible graphite types; one spiral-wound flexible graphite filled gas ket and one spiral-wound Teflon-filled gasket. Three of the materials mentioned in the survey responses were packings and not considered. In addition, seven were un known materials and their use was in locations at less than 550 F. They met with little success in applications at tem peratures above 300 F. Success in the use of organic fiber materials was mixed, according to re ports. One unit reported successful opera tion at 750 F. yet others reported failures of the same materials at 630 F and 700 F. Other materials failed at temperatures as low as 550 F. Failure times were short, generally in one to three months. Reports of operation for the carbon fiber/binder material ranged up to 1000 F. It operated successfully at 750 F and 850 F. but had one confirmed failure at 690 F. The user indicated that this materi al tended to leak in operations above 590 F, especially after a system distur bance. The use of flexible graphite sheet was limited. Although one utility repotted its successful use at 950 F, the service time was only three months. Higher tempera ture applications (to 1150 F) of flexible graphite were in the form of spiral wound gaskets. Although the results of the utility sur vey were not quantitative, they did con firm the laboratory test results. Tests of asbestos-free gasket materials confirmed that flexible graphite is the best material for replacing asbestos sheet in applica tions with temperatures above 300 F. Flexible graphite sheet equaled or outper formed asbestos sheet in all except the high temperature stress relaxation tests. This behavior of flexible graphite is adequate for use up to 650 F. However, the service life drops to less than five yean in applications above this tempera ture. As a result, the use of flexible graphite sheet gaskets in Westinghouse steam turbines is based on the usable service curves shown in Figure 4. The following guidelines are used for all appli cations eliminating asbestos Use the time vs temperature curses tor applications needing less than five-sear usable service. Use at 650 F maximum to get five years of service. Use at 600 F maximum to get 10 years of service. Use metal or flexible graphite filled spiral wound gaskets for application at temperatures higher than 750 F. Flexible graphite spiral wound gaskets also should be used for special or critical applications. Fidd tests To correlate the laboratory test data with actual service, use of flexible graphite gaskets in operating turbines is being evaluated. Flexible graphite sheet gaskets currently are in service in several loca tions and are operating without incident. As more units begin to use flexible graph ite we expea to verify its adequacy as a substitute for asbestos sheet gaskets in steam turbines. Conclusions No single material was found during the research that could substitute for asbestos sheet gaskets. However, flexible graphite is a satisfaaory replacement for applica tion at temperatures up to 650 F. As for organic fiber/binder sheet gas kets. they are not capable of performing effectively at temperatures above 350 F. Elevated temperature testing (i.e., at least operating temperature) is required on any material that is intended to be used as a replacement for asbestos. Further, long time stress relaxation testing at projected operating temperatures are required as part of qualifying any asbestos free mate rial. END AUTHORS WIIHam F. Jonas is a manager, nonde structive evaluation, at Weatinghouae Electric Corp. Ha holds a BS In metallur gical engineering from trie University of Illinois. B.B. Seth la a manager, material* and computer systems engineering, with Weatinghouae Electric Corp. He holds a BS degree In physics, chemistry and math from the University of Reteathan, a BE degree In metallurgy from the Indian Institute of Science, and MS and PhO de grees In metallurgy from the University of Toronto. 45 FIELD NOTES APPLICATIONS OF NEW TECHNOLOGY AT POWER GENERATING PLANTS MONA REYNOLDS. Special Projects Editor i- ** Central controls balance district cooling system Automated control of pan of a cogenerated district heating and cooling system in Trenton, N.J. has provided an investment pay back of less than six months because of an increase in efficiency compared to manual operation. Trigen Energy Corp. installed the Honeywell Inc. system to govern the chilled water ponion of a cogeneration project serving 18 buildings in downtown Trenton. Buildings served include the state capitol complex, offices, a hospital, a prison, and commercial structures. Control of hot water and electric power generation will be automated in the near future. At Trenton, Trigen operates two diesel electric generators total ing 12 MW, three hot water loops and a chilled water loop. A maximum of 3.73 MW is drawn by the equipment; the remainder is sold to Public Service Electric & Gas Co. "The 4.5-mile-long chilled water loop to the 18 buildings includes a 2.7-million gallon flow-through storage tank and pump house." according to Gary Fechter, Trigen director of operations. "The tank imports chilled water at night and exports it dunng the day....Loads can range to 11.000 tons." When the chilled water storage tank was installed, control was manual. Efficient use of the new tank required that automated optimization strategies for tank import and export be developed. At the same lime, it was determined that such automation could eliminate the need for operators to walk the several block distances between the generating plant and the pumphouse and chillers to start and stop and monitor equipment, adjust setpoints, and balance loads. The new control system consists of loop-and-logic controllers in three locations to provide chiller and pump interlocking, sequenc ing. multiloop analog control, and status information. Compared to previous manual operation, the new system tightened control of chilled water production and distribution. But to obtain the full benefits of automation through coordinated control --and a reduc- Coganaratlon plant supplies chiliad water for district cooling system in downtown Trenton, N.J. tion in the amount of walking required of operators -- the control lers in the field were linked to a supervisory unit located in the generating plant control room. With the supervisory unit, "the operator can balance the entire chilled water system--storage, production, and distribution--from one location." said Fechter. The operator interface to the supervi sory unit consists of two personal computers. In an emergency the PCs can access the controllers directly. According to Fechter. typical optimization calculations per formed by the supervisory unit are those employed to control the temperatures in the cooling tower that serves the absorption chill ers. By establishing setpoints that approach wet bulb limits, and also by more tightly controlling chiller operation, the sys tem reduces the amount of steam needed to produce a ton of cooling. Steam consumption has fallen from as high as 14 Ib/ton to 9.0 Ib/ton. Circle 206 on Reader Request Card Switch to helium unit improves condenser leak detection Operators of Florida Power & Light's Manatee plant decided to stop using a chlorofluorocarbon refrigerant to locate leaks because of its environmentally destructive nature. FPL also experienced problems with the product because it is heavier than air and drops to the bottom of the condenser, thus making detection of air inleakage difficult. FPL personnel were first introduced to the use of helium as a tracer gas when they called in a contractor who used a mass spectrometer to find an air inleak they could not find with their equipment. The mass spectrometer, however, is large and cumbersome, and FPL sought a portable instrument. Their search led to a helium detector manufactured by Mark Products. Inc., Sunnyvale. Calif., that can be strapped to a technician's waist or mounted on an equipment cart for testing. Testing for air inleakage with the helium unit takes two people, one to "spray" the helium, and the other to remain with the instrument and watch for indicative displays. Helium is sprayed on the suspected leak areas for about 10 seconds. The technician then waits for the helium to permeate into the leak and register at the detector. Circle 207 on Reader Request Card More field notes on page <v 46 POWER ENGINEERINGSARCH 1992