Document RJYNqgE1DBN3EGMxwNV5paj5V

BFGoodrich The BFGoodrich company Technical Center P.O.BOX 122 Avon UKe, Ohio 44012 2164334100 March 12, 1993 TO: Ron McCreedy Chairman Vinyl Institute Technical Committee Technical Committee Bob Burnett Roy Gottesman FROM; Alan Olson DISCUSSION OF PROPOSED CORROSION HAZARD ASSESSMENT WORK 1 have had extensive conversation* with Fred Clarice, Bill Abbott, Greg Smith and Caren Quisenbcrry regarding the subject project. Fred Clarke (Benjamin (Clarke Associates) pointed out that combustion gas flows through the corrosion chamber (a lab desiccator) currently used in the NBS Cone Corrosuneter apparatus at the rate of 4.5 liters per minute, which produces one changeover every 25 minutes. (Space Velocity of 24/hr.). The gas that is trapped in the chamber for the 24 hour post-burn period actually is generated during the final 2Jj minutes of the 20 minute bum (For reference, the total output of the cone burner is about 24 liters per second). If the test material bums non-uniformally, the combustion gas at end may not be representative of the gas at the beginning. To rectify this, B31 Abbott (Battelle) proposed using existing Baticllc corrosion chambers of about 200 liter volumes. A partial vacuum could by pulled on a chamber before a bum started, and combustion gases could be metered and pulled Into the chamber throughout the run. Muffin fans in the chamber would mix the combustion gas. Under the current scheme, some combustion products are likely swept through the corrosion chamber. The new procedure alleviates this, and allows Clarke to make his corrosion hazard assessment. By generating data in the form corrosion/unit mass burned, Clarke can scale the results to fire models. An example could be a wastebasket fire igniting Noryl or FVC articles and then calculating relative amounts of resulting corrosion. Separately, Rohrbach probes could be used to compare results from the new chamber to the old chamber. Combustion conditions could be altered to study those effects on corrosion. Per Jim Hoover, DuPont is inclined to participate. DuPont has a spare NBS burner, which could be loaned to Battelle for this study. DuPont should pay for all extra polymers that it wants to proride and have tested. Already we have pointed out some deficiencies in the NBS Cone Corrosimeter test. The Battelle Phase I and Clarke Corrosion Hazard Assessment work are new and necessary technical pieces. Together they will form another VI benchmark scientific work. iOO oxtv HDiaaoos da C0SO CC6 0TZS> c;tt ce/sx/co Regarding B actelle Phase II, speciation of the firegases, there have been recent advances in FTIR technology which could make this relatively straightforward. Nicolet contacted Greg Smith and Caren Quisenberry at Geoa/EFG recently. Nicolct has adapted an f l'IR to an NBS Cone Burner, with a prototype in operation at NIST and a second one planned soon for RAPRA in England. Abbott reported that the Battclle FTIR people can also do this work in Columbus, OH. However, some GC/MS work is still required upfront to calibrate the system. Once calibrated, the FTIR can fallow firegas concentrations throughout the bum generating new data. According to Abbott and Clarke, the shortcomings with speciation are: some important, low level species will cot be identified; and the corrosive effects of mixtures of gases are additive, if not multiplicative. (Compare to combustion toxicity where the effects of individual gates are not considered even additive with one another.] Thus, missing a small active component may leave an important gap in explaining the total corrosion picture. Nevertheless, speriatmg the combustion products from a material such as PE may show formation of formic and acetic acids, plus other components which can be quite corrosive, and thus dispel the notion that non-halogen containing materials should not cause corrosion from burning. - AJ. Olson 3/12/93 R&S 144306 800 onv HDinaooo da C9S0 CCS 9T2ffi iC^TT C6/ST/C0 REVISED CORROSIVITY BUDGET 1. BatteUe Phase I base cost PVC, PE, Nylon 200 liccr corrosion chamber $ 66,000 2. Add Acrylic and Nylon S 10,000 3. Cost of NBS Burner $ 15.3^000 4. Add Fluoropolyrners (S 5,000 ea.) -0- 5. Clarke's risk assessment (2 5 days at $l,QQ0/day) $ 6,000 max. Minimum cost to VI Maximum cost to VI $76,000 $113,000 Notes? 1. Abbott agreed methodology stays the 2. Abbott's estimate: $ 5,000/incremental polymer 3. DuPont may loan a spare NBS Burner 4. DuPont to pay for Fluoropolymers 5. Produces a hazard assessment comparing relative corrosivity to unit mass burned. Scaleable in fire m0u(ls ajo/3-93 R&S 144307 AisMommnoc 600 DITV HOIHdOOS da COSO CC6 9Uffi 8C-`TT C6/ST/C0