Document Njjjoz9OL255xqe4wMVjZk2b

.I B.F.floodrlch Cbsmical Company i v i111 o k o* ini i r toiDiitN cHm too oak TIICE Oo'uLtWAIIO CLEVELtttO.OHIO 44191 FHOKE: 2t0-4X-020 July 20, 1976 TO: VCM/PVC PRODUCERS GROUP and PPl/UNI-BELL DEFACTO COMMITTEE ON RVCM Gentlemen: A proposal to establish a RVCM third party certification program at the resin producer's plant has been presented to the DeFacto Committee. The attached report represents our understanding of the problem that brought about this proposal and offers an alternate solution that we believe will better meet the needs of the PVC pipe and fittings manufacturers. We think this alternate solution will also better serve the public and the plastics pipe industry. The solution offered for your consideration is based on the assumption that a system providing for frequent sampling and RVCM measure ment at the resin manufacturer's plant is the first requirement for a program that will have as its objective the furnishing of accurate lot-by-lot RVCM test results having an extremely high level of confidence. R. W. TANNEH1LL Senior Product Manager Rigid Plastic Products Director of Industry Affairs ucc 049271 i KVCM 3rd PARTY CERTIFICATION AND AN ALTERNATIVE SYSTEM A program of resin or conpound sampling and RVOI measurement by an indepen dent laboratory. Samples to be obtained at the material manufacturer's plant periodically but with un-announced visits. NSF is being considered as the out side laboratory but it has been suggested that some other laboratory would be better especially if NSF tests the pipe and fittings for RVOI. Why? It is reported that some of the pipe manufacturers who have been requiring their resin suppliers to furnish certification of RVOI level, have not found the results to be satisfactory. Therefore, they believe that some other system is required to assure them that the RVOI level is below the established limit. They see the 3r4 party certification program as the answer to their problem. The pipe manufacturers have also expressed concern about the very costly penal ties that will befall any pipe manufacturer who ends up with pipe having a RVOI , above the allowable. They believe that the penalty might even include recovery of pipe from suppliers's stock and possibly removal of installed pipe. WhaOfould^S^^^aTtj^Cer^fication^ccongUsl^^^jrha^^Jeeded? Perhaps it would bring about some minor improvement in the conditions, but our experiences with 3r4 party certification programs indicate that it would accomplish far less than is needed. In our opinion what is needed is a system for dot-by-lot RVOI certification which is based on extensive and statistically sound ^sampling and testing of resin (or compound) on a lot-by-lot basis. Ho^Woul^^h^^lteniativ^^stemWorjj? The system would have two parts - the first providing for in-house RVOI measurements and certification and the second providing for periodic verification of RVOI measurements by an outside laboratory. The first part does work and is already being used by some resin producers and the second part remains to be de veloped and implemented. Both parts are further describe below. PART 1- Each resin (or compound) producer must do this at every plant. 1. Set up and staff a laboratory equipped with a gas chromatograph to measure RVOI using the established Head Space (HS) or Direct Injection (DI) method. 2. Determine their test variability (St) for RVOI measurements at the 1, 10, and 100 ppm levels. 3. Establish a sample gathering procedure. Continued UCC 049272 I RVCM 3r<* PARTY CERTIFICATION AND AN ALTERNATIVE SYSTEM fCon'tl 4. Conduct a statistically controlled experiment to establish the RVCM sample-to-sanple variation (S$) within a shipping con tainer or lot by taking a larger number of samples from a variety of containers over an extended period of time, 5. Using Ss and St determine the confidence level and variability for each shipment and the maximum RVCM level. 6. Using the data obtained above, each lot should be accompanied by a certification statement in this form. PART 2 "The maximum RVCM level in this lot is less than --_ Bpji with 99% certainty". Each resin (or canpound) producer shall arrange for a periodic verification of his RVOl measurements by any one of several pTe-qualified independent labor atories. Verification shall be accomplished by sending to the independent labor atory resin samples accompanied by the lot number and the measured RVOl level for that lot. Each producer shall have two resin samples tested per 100 lots of resin shipped except that no producer shall have less than two samples per week, per plant tested. The results of these tests shall be sent to the Tesin producer by the laboratory and shall be made available to a Task Group of the PVC producers. Wha^Ar^The^dvantages^^Thi^S^stem? This system will provide the test data required for lot-by-lot certificates that will have a very high level of confidence which cannot be attained by 3r^ party certifcation. It provides the necessary data to the res-in manufacturer at the lowest possible cost. It provided this data to the resin manufacturer before the material is shipped so that he can monitor variations and take corrective action. And it provides independent laboratory verification of RVCM measurement capability. RCW/cpc cc: R.K. Tannehill July 14. 1976 B.F. GOODRICH CHEMICAL COMPANY Robert C. Wilging, P.E. Manager Piping Standards and Testing Plastic Materials Sales ucc 049272.01