Document ZBReqVNVvwj6gn7mvr6mmqyj8

4 R.D. GAMBLIN IWISrOiVvCV CommuvilcolkMt From: Date: Subject; J.P. Warner September 13, 1965 VALVE POSITION INDICATORS It la my under8tending that Aberdeen hat requested funds ae part of their 1986 Capital Budget to Improve the valve position Indicators on some of their reactor and VCM charge system valves. I evaluated a similar project for the Oklahoma City PVC Plant and determined that there was not sufficient justification for the project at Oklahoma City. The following la a summary of my understanding of the issues facing the two plants; Aberdasn 1. Aberdssn has had approximately 14 Incidents where the valve position indicator showed that a valve was closed when it was at least partially opened. These incidents resulted In the uncontrolled release of material from the reactor module. These incidents were caused by either the twisting of the valve stem or by the failure of the coupling connect ing the valve stem to the pneumatic actuator. In most cases, Aberdeen's records are not sufficient to Identify which failure mode was responsible for the release, but it Is known that both failure modes have occurred, 2. Aberdeen has developed a valve position indicator capable of detecting the valve*s true position even if the coupling falls. The new position Indicator will not detect the valve*s true position If the valve stem twists* 3. Aberdeen believes that $trengthlng the coupling between the actuator and valve stem will result In an Increase in the number of twisted valve stems. They believe that the coupling should be the "weak link" in this assembly. 4. Aberdeen believes that a failed coupling is preferable to a twisted valve stem because a failed coupling can be detected before a release occurs if you are using the new valve stem indicators. 5. Aberdeen hae 6 actuated valves per reactor that are single block valves, Oklahoma City hae two single block valves on sach reactor. 2 *d qR/RI/PH VAB.0001098784 -)MQ...A3 OH HJPTA Lind -I Page 2 9/13/85 Oklahoma City 1. To the best of our knowledge, the Oklahoma City PVC Plant has not had a coupling failure on any actuated valves. In addition, we are not aware of any plug valves where the valve stem has twisted. We have found twisted valve stems on Jameebury Ball Valves on the main dump valve on the reactor and on the water drain valve on the water stripping tanks * Normally, we detect the twisted valve stem on the reactor main dump valve through our monthly valve leak checka and from higher than normal VCM content In the stripped slurry. The main dump valve la double blocked reducing the number of VCM releases. A significant VCM release has not occurred ae a result of these twisted valve stems. A significant VCM release from the waste water strippers did occur in December, 1979, as the result of a twisted valve stem. A copy of the near miss investigation Is attachad. A rwiAi-eH vAiy* stem hoc not been found in Lilly verviue since proper winterization of the valve was completad. 2. While we have not tested the system, we believe that the system proposed by Aberdeen Is a well designed system capable of detecting coupling failures. 3. We believe that we have not seen any coupling, failures because we either purchase couplings that are of sufficient strength or we reinforce the coupling by installing a metal sleeve around the coupling. This has not resulted In an increase in twisted valve steams. All twisted valve stems have occurred on Jamesbury Ball Valves equipped with stan dard factory supplied couplings. 4. We believe that the approach to the problem that we should follow is to design a system that does not fall. We believe that our couplings meet this requirement, and we are investigating modifications that we believe will prevent the twlatlng of valve stems on the main reactor dump valves such as eliminating ball valves is these services and upgrading the valve stem metallurgy. We believe chat there should not be a "weak link" in the actuator-coupling-valve system and that potential for failure should be eliminated or min imized--'not engineered as the desirable mode of operation. We also believe that it would be extremely difficult to design a system where the coupling will always predictably be the "weak link". P 'd "IdlOi Page 3 9/13/85 During the last two-three years considerable capital has been expended on upgrading the valve position limit switches (AFE No, 5660-Improved EPA Compliance System Reliability, AFE No, 5621-Massive VCM Release Prevention and Miscellaneous Under $15 M funda)* Tliuae n*W limit switches are similar to Aberdeen's proposal (also made by GO Systems) but are mounted on the actuator to allow access to the coupling and packing nuts. They have performed extremely well in our critical services and have provided reli able indications. Based upon our experiences, we do not feel that there is justi fication for installing the new valve position Indicators at Oklahoma City and request that this project be eliminated from our 1986 Capital Budget. 4 J.P. WARNER SENIOR PROCESS ENGINEER JPW/lo P `d 6: I t 8/0 I/60 VAB.0001098786 D>10 AlOd dlSin NOdd To/ D. W. Hollis Interoffice Cofflmunication From: Date: Subject: F. 6. Jeanson August 27, 1985 LIMIT SWITCH REPLACEMENT - BUDGET DESIGN Attached is the budget design to replace the limit switches on valves in VCM and chemical wash service. These new limit switches will operate on the position of the valve stem, instead of the position of the actuator, and thus provide each valve with accurate and reliable indication in the control room. These new type of limit switches are proposed to be installed on valves on the reactors and the VCM charge systems as part of the compliance plan capital items. This design will provide these same type of limit switches on all the other valves with existing limit switches in VCM, and chemical wash service. A trial limit switch, that is the same model as the limit switches specified in this design, was installed in the plant in the first quarter of 1984. This limit switch has operated trouble free for approximately one and a half years. By copy of this design to REP a budget estimate is requested, If there are any questions or comments please refer them to e. F. G. Jeanson Process Engineer b- VAB.0001098787