INSTALL_2004 - Laserfiche WebLink

Home Browse Search

Ameron Dualoy° Pipe Monitoring System in/deg. This is significantly larger effect than for the no-vapor tests. If this result is extrapolated to a 300 ft line with a temperature change of 30 deg F, the total level change in the brine reservoir due to temperature is of the order of 6.35 inches. This change in fluid level is the full range of the change. By adding a "service factor" multiplier of 1.5 to this number, the false alarm probability is reduced dramatically, well below the required maximum. As noted in earlier discussion, the direct extrapolation of the trapped vapor results from 21 ft to 300 ft will lead to exaggerated effects. It is estimated that the results would be about 30% to 40% higher than for an actual 300 ft installation due to having proportionally fewer fittings, the primary location of trapped air. Gravity filled Interstice The interstice was originally filled with brine using gravity feed. These tests were more exploratory in nature and the results are not as well defined. They are provided in Table 2 for ease of comparison with the other test conditions for the evaluation. The average result of these tests was 0.0129 in/deg. These results show the gravity filled interstice performs between the evacuated (best case) and vapor injected (worst case) results. Designing to the worst-case conditions described in the previous section is conservative. Time to Detect a 0.1 Gal/h Leak Liquid volume transferred into the interstice from a leak in the primary line or a loss of liquid from a leak in the secondary pipe is immediately and directly seen as a level change in the reservoir of this system. This was demonstrated by removing brine from the interstice at a rate of 0.10 gal/h. The results of this test are shown in Figure D- 6 in Appendix D. The rate of change in the brine reservoir level for a 0.1 gal/h leak was observed to be 0.05 in/min. This rate of level change in the reservoir is independent of line size. If the level must change by 5 inches to produce an alarm, the alarm would occur in approximately 100 minutes. A leak in the secondary pipe would release interstitial liquid to the environment and would not necessarily be immediately critical. A leak from the primary pipe would however, present a potential loss to the environment and should result in an immediate shut down of the turbine. Catastrophic Failure Test The effect of a catastrophic failure of the primary pipe was of some concern. To simulate these conditions a pressurized tank of brine was attached to each of the bleed ports. Brine at a pressure of 40 psi was then introduced into the interstice at the maximum rate the interstice would accept the liquid. For the port located 21 ft from the reservoir the flow rate into the interstice was approximately 0.2 gal/min or slightly less than 0.7 liters/min. This would produce a high level alarm is less than two minutes. At a distance of four feet from the reservoir, the flow rate into the reservoir was observed to be approximately one gal/min, a substantially large flow rate than at 21 ft. A failure of the primary pipe at this distance would result in an alarm is approximately 15 seconds. Longer distances would undoubtedly produce lower flow rates. Page 11 of 13