Document EdkpvZ02yz8Obw6xe6joqyGjV

/ ' f-* j' , <, Project Title: Swedish/Norwegian PVC Incineration Project Objective of Project: To carry out research demonstrating that incineration of PVC in municipal solid waste streams does not increase dioxin and furan emissions. Across Europe there continues to be a wide spread perception that the presence of PVC in MSW is responsible for dioxin and furan emissions. Through our consultant, Dr. Richard S. Magee of NJIT, we are involved in the planning of Norwegian and Swedish laboratory studies which are intended to provide greater scientific rigour than previous investigations on the role of PVC in dioxin/furan emissions and consequently to gain greater acceptance for the incineration of PVC in the Euro pean community. These studies are being guided by a steering committee consisting of incineration experts from Belgium, Germa ny, and the USA (Dr. Magee). Status Report: The initial meeting of the steering committee was in Umea Sweden on October 30, 1990. The committee has met four more times (Oslo, Gottenberg, Umea, Karlsruhe), approximately once every six months. Progress has been painfully slow. A major difficulty for both groups has been the design, fabrication, and construction of the combustion reactors. The most recent progress reports (Jan. 93) indicated that Umea spent zero time on the project, "this month we he had an other big project to be concern about. However, next month we are back in the ECVM business again." The SI report indicated that, as we suggested, they are unable to synthesize the radioactive labeled PVC and have sug gested proceeding without labeled PVC. While I have agreed, any value of their research contribution has been significantly re duced. Further, recent test runs seem to indicate that the combustor performance is poorer than reported in December. The next steering committee meeting is scheduled for April 29-30, 1993 in Brussels. Recommended Actions & Future Programs: Despite the extremely slow progress on this project it is very important that we continue our participation to ensure that the "good results" from the Pittsfield, MA study are not lost. R&S 144312 Meeting Agenda: 1) Are any decisions needed by the Technical Committee (Y/N)? 2) Is a presentation and/or discussion needed (Y/N)?...................... 3) If yes to either #1 or #2, how much total meeting time should be scheduled for presentation, discussion and/or decision (Minutes)?................................................................................................................ Prepared by: Richard s. Maaee Date: 2-22-93 1113923.RM /f si "fd, J8, " s* #/*/? PBQnEgT fff>R PROJECT TO RVAIiPATB COMPARATIVE ECONOMICS OF MUNTPTPAT. gQLin WASTE AMP HQSPTTaT. WASTE INCIMERATJON PROCESSES T - BACTtGRQTTTm Polyvinyl chloride (PVC) is the second largest thermoplastic material produced in the world today. The U.S. industry produced over 9.2 billion pounds in 1992. PVC, or vinyl, as it is sometimes called, is used to make a wide variety of industrial and household consumer products including water and sewer pipe, electrical conduit, wire and cable coatings, appliance housings and electronic products. Other major uses include wall coverings, home siding and window frames. Most of these are durable products which have long service lives and hence do not appear in the waste stream for many years. However, over 500 million pounds per year of PVC are used in disposables such as product and food packaging, water bottles and food wrap and the majority of these items contribute to the waste stream. IT. TSSTTES Because the management of municipal solid waste (MSW) is a major societal problem, attention has been focused on minimizing the amounts of all disposable materials, including plastics, that are sent to incinerators or landfills. Among the strategies that are receiving the highest attention are source reduction, reuse and recycling. In the case of polyvinyl chloride, programs have been undertaken by The Vinyl Institute and its individual member companies to develop technology to facilitate the recovery of vinyl containers and to encourage their recycling into second-generation products. While these recycling programs are technically feasible, several problems have been experienced, which limits their success. These include: Difficulty in having vinyl containers included in muni cipal recycling programs because t: comprise less than 0.5% of the municipal waste stream. High capital costs for automatic sortation equipment needed to facilitate the separation of vinyl from other plastics. Unfavorable economics for recovery/recycling programs. While plastics recycling programs will undoubtedly continue to receive emphasis and attention, incineration of the combustible portion of municipal solid waste with energy recovery has begun to receive greater attention. R&S 144314 2- - In the past, questions have been raised relating to the generation of toxic chlorinated dioxins and furans in the incineration of municipal solid waste (MSW) containing polyvinyl chloride. Studies carried out in the United States and abroad have demonstrated that incineration of MSW can be carried out safely, that the formation of doxins/furans occurs as a result of any chlorinated materials being present in the MSW being incinerated and that dioxins/furans can be minimized by the use of proper incinerator operating conditions and state-of-the-art pollution control systems. Incineration is also used to dispose of waste generated at hospitals including "red bag" waste containing pathogens as well as tubing, blood bags and bags used for administering parenteral solutions, some of which are PVC. Older hospital incinerators are generally not capable of operating at the higher temperatures ( above 1500 0 F) necessary to minimize dioxin/furan formation. However, incineration of either MSW or hospital waste does result in the generation of hydrogen chloride (HC1) . This has been alleged to be a contributor to or cause of "acid rain." III. CQITFRQHTING THE ISSUE Since polyvinyl chloride contains 56.7% chlorine, its incineration results in the generation of 0.584 lbs. of hydrogen chloride per pound of PVC incinerated. State and federal environmental protection regulations require that this HCl be removed before incinerator off-gases can be discharged to the atmosphere. Generally the removal of HCl can be accomplished by alkaline scrubbing using either caustic soda (aqueous) or lime (dry system). While alkaline scrubbing ( or some other economically viable method) would be required on any state-of-the-art MSW or hospital incinerator, little information is available on the economic impact of this air pollution control step. Particularly needed is an economic analysis of the impact of the presence ( or absence) or polyvinyl chloride in MSW or hospital waste on the overall incineration cost. - XV. REQPEST FOR PBQPOHAT. R&S 144315 The Vinyl Institute has decided to seek proposals from a number of organizations to accomplish the following: 1. Development of current information on the PVC content of typical MSW and hospital waste. -3- 2. Development of capital costs for incinerators for MSW and hospital waste that are representative of the sizes of incinerators that can be expected to be built in the future. Costs for three sizes, large, medium and small, should be prepared. The sizing of the scrubbers should be such that they are of the appropriate size dependent on whether or not PVC is contained in the waste being fed to the incinerator. Thus, a total of twelve capital costs are required, i.e. six (6) for MSW incinerators and six (6) for hospital incinerators (three sizes for each type of incinerator, with and without PVC in the incinerator waste feed). 3. Development of twelve (12) comparative economic analysis and costs (both capital and operating) of MSW and hospital incinerators under conditions where PVC is present in the incinerator waste feed and where PVC has been removed prior to incineration. V. PROPOSAI, COMMENT The proposal should cover the following points: 1. Background of the submitting organization: Size Capabilities Relevant experience and/or special expertise Curriculum vitae of principal and supporting investigators. 2. General work plan to address elements in the Request For Proposal ( See Section IV, pages 2 and 3. 3. Timetable: The Vinyl Institute has established a target date of six months for completion of this project. The proposal should therefore include mileposts to be met at agreed-upon dates in order to achieve completion by this target date. 4. Deliverables: The proposal should include the proposed contents of the final report that will be submitted at the conclusion of this proj ect. 5. Cost: The proposal should include the cost as well as the payment schedule. R&S 144316 -4- 6. Date for Receipt of Proposals: Ten (10) copies of the proposals should be submitted by to: Mr. Robert H. Burnett Executive Director The Vinyl Institute Wayne Interchange Plaza 155 Route 46 West Wayne, New Jersey 07470 II 7. Other Any additional information or ing organization believes will be objectives of the Vinyl Institute proposal. suggestions that the respond useful in accomplishing the in evaluating the submitted VI. ORGANIZATIONS BEING INVITED TO SUBMIT PROPOSALS The following organizations are receiving this Request for Proposal (RFP) and are being invited to submit proposals: Organization Location -- Battelle -- Ogden Martin Systems -- /'7'V ,5- 6/a y / Columbus, Ohio 43201-2693 Fairfield, New Jersey R&S 144317 FROM:UISTPi R8,D Tu; TO: 407 735 0567 Ron McCreedy, Dow USA mR A, 1993 4:20PM 8858 P.02 )c I) IntOroffica Communication From: H. J. Hall Dili: March 4, 1993 Sulijici: Fire Science Objective of Project: Monitor and promote scientific reasoning in the development of fire codes, standards and regulations that ensure a fair treatment for PVC products. Status Report: A. Corrosion The Polyolefins Fire Performance Council continues to attempt to resurrect tiie ASTM E -5.21.70 radiant test which most closely resembles an acid gas test. At the February 23 meeting of the Electrical Materials and Products Subcommittee of the Coordinating Committee on Fire Safety (CCFS), a group representing Polyolefins, PVC and Fluropolyracrs was formed to resolve the negatives on this protocol. It appears doubtful that the negatives can be resolved. All indications are that the present negatives on ASTM D9.21.04 should be answered and this protocol will be the U.S. test represented at IEC89. This uses the cone calorimeter we favor, We are continuing to monitor the formation of CSA and UL study groups on corrosion. Further information is needed before type and degree of VI participation can be determined. B. Toxicity 1. Toxicity NYC has apparently adopted an "interim protocol" which is the U-Pitt test (NY State modified) with a cut off number of 19.7. NYC has not been generous on information on the why, how and when they did this. Rich Gottwald of the SPI was asked at the CCFS Executive February 24 Meeting to investigate the status of tills action further. At this time, the only products affected are interior and floor finishes. 2. ASTM No fbrther information on ASTM E5.21 (NIST based method) from last meeting. Supposedly, a rewrite is underway to resolve eleven negatives. R&S 144318 FRQMIUISTR R2,D TO: 407 735 0567 MAR 4, 1993 4:21PM {1856 P.03 C. Defense Electronic Supply Center (DESC) Marcelo Hischler obtained copies of two reports from Support Systems Associates, Inc. of San Antonio which as reported January caused PVC products to be placed on the Hazardous Materials Minimization (HMMIN) list. Marcelo has established contact with both the audiors and die Air Force office diat audiorjzed the reports. Both B.F. Goodrich and Vista have copies of die report for review. D. New Business 1. VSI Fire Test Program The VSI is planning to conduct fire test on Type 3 and Type 4 non combustible construction. Plans are to establish a relationship between large and small scale tests. Work will be done at SWRI. 2. NFPA 90A Research Project Recent statements by the Southern Building Code Congress (SBCC) on die total amount of combustibles in air handling plenums have prompted Panel 90A of the National Fire Protection Association (NFPA) to consider a research project on the amount and type of combustibles in plenums. The National Fire Protection Research Foundation (NFPRF) would conduct die research, Funding is being solicited now. R&S 144319 FROM:UISTfl R&D TO: 407 735 0S67 MAR 4. 1993 4:22PM H858 P.04 RECOMMENDED ACTIONS AND FUTURE PROGRAMS: A. Corrosion Continue to monitor ASTM and IEC89 actions. Promote the standardization of ASTM D9.21.04 as the most scientific useful corrosion protocol. B. Toxicity Support NIST activities on toxicity protocol development in ASTM. Monitor NYC through SPI and determine with other SPI Divisions a proper course of action. C. Defense Electronic Supply Center (DESC) Attempt a stealth approach with Marcclo alerting his contacts to the bad science involved. Both B.F. Goodrich and Vista will provide input for presentation. D. Monitor developments on VSI and NFPA 90 A research. 3J 0) 44Wroo** S' 407 705 0567 ROY T GOTTESMAN A ppe^i* E P02 PROJECT REPORT FOR VI TECHNICAL COMMITTEE MEETING r Date of Mooting: March 23, 1993 Objective of Project: Develop a compendium of critical papers dealing with the combustion toxicity on fire hazard of PVC. Statue Report This project is on hold until the SWRI primate and rodent studies with HC1 and PVC are published. Response to reviewers' comments are expected to be completed by the end of March, 1993. heoommended Aotto n A Future Program*: Wait until April to restart activities. Compendium is expected to be completed and ready for publication by the end of 1993. UMtlng Agenda: 1)Ars any decisions needed by the Technical Committee (Y/N) ?...... Z) Is a presentation and/or dtecussslon needed (Y7N)7................... 3) If Yes to either #1 or #2, how much total meeting time should be scheduled for presentation, discussion and/or decision (Mlnutee)?, N N Prepared Robert K. Hinderer, Ph.D. n, Uoiui, ,,,,,, Pete: 03/22/93 R&S 144321 FEB- 23"a'j--l 1 1 4 "FROM ID: PAGE 2 f*' PROJECT REPORT FOR VI TECHNICAL COMMITTEE MEETING Date of Meeting; March 23, 1993 Objective of Project: To Prevent Solvent Cements from being regulated adversely by the Clean Air Act. Statue Report: The present status quo Is maintained. With no new regulations from SCAQMD, the next deadline Is 1/1/98 at which time new cements would have to be developed to meet the 260 g/l VOC requirement. Recommended Action & Future Programs: The various air management districts In CA have different test methods, VOC level targets, product size exemptions, and timetables from South Coast, so that there Is a need for harmonization between the districts. Because South Coast 1$ the dominant player, we hope to use our results as the model for the other districts. We are also considering legislation In CA to direct all districts to accept laws based on the South Coast mode). Meeting Agenda: 1) Are any decisions needed by the Technical Committee? 2) Is a presentation and/or discussion needed? 3) How much meeting time should be scheduled for presentation? N V 10 min. R&S 144322 Prepared by: Donald Goodman DG:djf 2/It/a3 Date: February 18,1993 VI TECHNICAL COMMITTEE PROJECT PROPOSAL Project Title: Nature of Program: Continuation of Existing Project 0 New Project Objective of Program: Description of Planned Activity: Deliverables: Relevance/Importance of Project to Vinyl Industry: Timing: Project Starting Date:_________________________ Anticipated Completion Date:___________________ Project Cost: Finances needed in FY 193-`94 $_________ _ Other Resources Needed__________________ Will funding be needed in future years ? (Y/N) If yes, how much? Priority: Urgent - Highest priority Important - Include in program if finances permit. Can be delayed until next fiscal year. Project Proposed by: Date: Technical Committee Action: Approved and Included in Technical Committee Program/Budget " Postpone until next fiscal year 5 Not approved. Date of Technical Committee Action: R&S 144323 VI TECHNICAL COMMITTEE PROJECT PROPOSAL Project Title: Seminar Program for Vinyl Fabricators Nature of Program: [ ] Continuation of Existing Project [X] New Project Objective of Program: To increase awareness of fabricators of vinyl products of environmental issues facing the PVC industry so they can become involved in grass-roots activity in their communities to refute anti-vinyl allegations by industry's critics . Description of Planned Activity: a series of regional seminars (3 or 4 in different geographical areas) will be conducted with in dustry speaker covering the topics given in the attached letter to Ron Me Creedy dated December 14, 1993. Appropriate takehome materials similar to those developed for the outreach "On the Firing Line" seminars would be provided to attendees. Deliverablea^Three or Four regional seminar programs and distri bution of information for vinyl fabricators to use in defending vinyl products from attacks by industry's opponents. Relevance/Importance of Project to Vinyl Industry: This program will help educate fabricators so they can become spokesmen in defense of the industry. Timing: Project Starting Date: After June i. 1993 Anticipated Completion Date: About 3-4 months after start Project Cost: Finances needed in FY *93-`94 i owf -ms 16to * si Other Resources Needed : VI Company tersonne) as speakers Will funding be needed in future years ? (Y/N) Yes, for follow-up If yes, how much? $ 2,500 to S 3,oqo Priority: Urgent - Highest priority H Important - Include in program if finances permit. Can be delayed until next fiscal year. Project Proposed by: Date: Technical Committee Action: Approved and Included in Technical Committee Program/Budget Postpone until next fiscal year Not approved. Date of Technical Committee Action: R&S 144324 0 Chemical Management Resources, Inc. FO Bex 6 <3 Glen Rock. NJ 07452/Tel. (201) 447-9573 December 14, 1992 Mr. Ronald L. McCreedy Dow Chemical Co. 2020 Willard H. Dow Center Midland, Michigan 48674 Dear Ron: You will recall that at the last Technical Committee Meeting on September 17, it was decided that you and X would work together to arrange and set up a seminar/meeting with vinyl fabricators of consumer products on environmental issues of interest to the industry. In anticipation of our next Technical Committee meeting on January 20, listed below are the general subject matters that I believe would be desirable and appropriate for such a seminar: 1. Overview of environmental issues affecting the industry. 2. Health and Safety Issues a) Angiosarcoma/Unit Risk Factors b) Fire/Combustion toxicity 3. Solid Waste Management Issues a)Recycling including VI programs for sortation and reclaiming b)incineration including Dioxin generation 4. Life Cycle Analysis (LCAs) -- Review of current "state of the art" on PVC LCAs and proper use and application of LCAs. 5. Global Issues including the Greenpeace anti-chlorine activity, linkage with PVC and what industry is doing to counter this attack. 6. The role of the vinyl processor - - What canthey do to assure the continued, health and growth of the industry. Let me have your thoughts on this and we canplan to discuss this at the next Technical Correnittee meeting. I will be leaving Florida on December IS to spend the next few weeks in Rhode Island where our daughter is expecting her fourth child. If you have any project summary reports, please send them to me at the address below. I will put together a proposed agenda shortly thereafter and send it to you for your review and approval. if you have any new items you can alsosend them to me at the following address: Dr. Roy T. Gottesman c/o Dr. Cheryl S. Gottesman 293 Doyle Avenue Providence, Rhode Island 02906 or call me at (401) 272-4683. Hope you and yours have a Merry Christmas and a Happy New Year. cc: Robert H. Burnett Sincerely yours, R&S 144325 PROJECT REPORT FOR VI TECHNICAL COMMITTEE MEETING Date of Meeting: 3-23-93 Objective of Project: Determine if membrane separation is technically feasible for recovering HC1 from incineration and pyrolysis processes. Status Report: Contacted John Oir (Air Products 216-447-8877) and Pat O'Brian (Air Products/Permea 313-995-3499). Perinea's Prism membranes are a variety of polyimidcs, and can tolerate less than 5 ppm HCL A Japanese and a U.S. company make polyamide membranes which have similar tolerance. Per Tom Salpesio at W,R. Grace (410-531-4000), Grace's cellulose acetate membranes are prone to attack by wet HQ. Dry COj, SOj and Nj will permeate with dry HQ in all of the above membranes, as well as others like polysulfone. Pumping the gas stream to 100-200 PSIG has very poor process economics, plus poor separation (only 4 to 10 X concentration rise). Recommended Action Be Future Programs; Follow "perstraction" in the literature. This involves a caustic solution on one ride of an HQ tolerant membrane, with the gas stream or aqueous stream on the other ride. See work by R. Klaasett of TNOIMET in Apeldoom, Netherlands, specific to HQ separation in flue gases with hybrid polysulfone membranes. Meeting Agenda: 1. Are any decisions needed by the Technical Committee (Y/N)7 No 2. Is a presentation and/or discussion needed (Y/N)7 3. If Yes to either #1 or #2, how much total meeting time should be scheduled for presentation, discussion and/or decision (Minutes)? R&S 1443 AMIOUI2ft.DOC M> soo IV} O) 0I1V HDIHQ009 dS C9S0 CC6 9T2S SC=11 C6/ST/C0 PROJECT REPORT FOR VI TECHNICAL COMMITTEE MEETING Date of Mgefincr Objective of Project: Develop methodology with Fred Clarke and Bill Abbott (Battellc) for a "Corrosion Hazard Assessment.' States Report: Clarke reviewed the BatteUe Proposal and SPI Polyoefins Divisional Report Clarke can do a hazard assessment comparing relative corrosivity per unit mass burned from the phase one BatteUe work. Abbott agreed to substitute a larger corrosion chamber so combustion gases are sampled and coUected from the endre burn. Greg Smith from Geon believes this approach would give credible results. DuPont (Jim Hoover) was informed of our plan, and DuPont is inclined to participate. Recommended Action & Future Programs: 1. Inquire if DuPont will participate and loan an NBS Burner. 2. Proceed with Battellc Phase 1 and Clarke's Hazard Assessment. 3. Review attached revised budget and discussion. Meeting Agenda: 1. Are any decisions needed by the Technical Committee (Y/N)? Yes 2. Is a presentation and/or discussion needed (Y/N)7 Y5 3. If Yes to either #1 or #2, how much totai meeting time should be scheduled for presentation, discussion and/or sion (Minutes)? | Pri-parcd V Pate: 3/10/93 J30 Ai\AJO\JIMJ.CX>C 000 DITV HD IMOOD dS COSO CC6 9TS& 9C:TT C6/ST/C0 R&S 144327 BFGoodrich The BFGoodricU Company Technical Center P.0. Box 122 Avon Lake, Ohio 44012 216-933-0100 S^pp-OisiAlX C- February 4, 1993 Dr. Fred Clarke III President Benjamin/Clarke Associates 10605 Concord St., Suite 501 Kensington, MD 20893 re: Vinyl Institute Smoke Corrosivity Project Fred, The Vinyl Institute is seeking to fund a study comparing relative corrosion risks from several type of polymers exposed to a fixed fire scenario. This risk assessment would entail modeling a fire scenario (such as a wastebasket fire in an office) and estimating the corrosion potential as a function of the material's fire properties and corrosivity of its combustion products. As a starting point, we would like to compare PE to nylon, PVC, acrylic, PPO/PS and a fluoropolymer. As a follow-up to our telephone conversation, I am enclosing two background documents. The first is work published by the SPI Polyolefins Division regarding corrosivity of smokes from various plastics burned in three different combustion tests at UL. The second is a proposal from Battelle to verify that corrosion does occur from plastic materials burned in the NBS calorimeter, and to speciate major components of those firegases. Please judge whether the SPI Polyolefins Div. data from UL would suffice for the risk assessment, or whether additional data would be required from Battelle. If you wish, please contact DuPont (one of your other clients) to determine if they might participate in this study. If you have any questions or comments, please direct them to me at the above address. I can be reached by phone at 216-933-1958 and by Fax at 216-933-1655. Bud Hall from Vista Chemical (phone 512-331-2425) will also be working with me on this project. Best Regards, R&S 144328 A.J. Olson cc: Lou Maresca (BFG) Chuck Daniels (BFG) Bud Hall (Vista) Ron McCreedy (Dow) jTob Burnett (VI) PF PC The Society of the Plastics Industry, Inc. POLYOLEFINS FIRE PERFORMANCE COUNCIL 1275 K Street N.W., #400 Washington, DC 20005 Carl F. Tripp Managing Director TEL: 203-438-1754 FAX: 203-431-4375 November 23, 1992 Mr. Allen Olsen B.F. Goodrich 6100 Oak Tree Blvd. Cleveland, OH 44131 Dear Allen, We're happy to supply, as you requested, Jim Bennett's paper presented at the recent FRCA Conference. As I mentioned on the phone, if you have any questions after reading it, please give me a call. Sincerely, CT:s Enclosure carl F. Tripp R&S 144329 CORROSIVITY TEST METHODS FOR POLYMERIC MATERIALS PART 4-CONE CORROSIMETER TEST METHOD JAMES G. BENNETT, JR GE Plastics STEPHEN L. KESSEL Quantum Chemical Corporation, USI Division CHARLES E. ROGERS Union Carbide Corporation ABSTRACT This is the fourth in a series of papers to investigate corrosivity test methods published by the Polyolefins Fire Performance Council, an operating unit of The Society of the Plastics Industry, Inc. In the first paper, 24 polymeric materials were evaluated for smoke corrosivity following the test method proposed by ASTME0S.21.70 which uses a radiant combustion/exposure apparatus. The second paper discussed the evaluation ofthe same materials using the CNET corrosion test method under consideration by ISO TC61/SC4AVG2 and IEC TC89/WG3 and compared the CNET results with the ASTM E05.21.70 results. In the third paper, the 24 polymeric materials were evaluated using a modified DIN acid gas test method and the results were compared to both the previous ASTM E05.21.70 and CNET results. These commercially available polymeric materials cover a broad range of compositions used for wire and cable insulation and jacketing. In this paper, the same polymeric materials were evaluated following the "Fire Response Standard For Determining the Corrosive Effect of Combustion Products Using a Cone Corrosimeter" proposed by ASTM D09.21.04. In this test method, a specimen is subjected to radiant heat at the recommended heat flux using a spark igniter to ignite combustible vapors. A portion ofthe products of decomposition or combustion are channeled in a dynamic mode through an exposure chamber in which corrosion targets are placed until the specimen has lost 70% of its total available mass loss. The mass loss is determined from previous experiments at the recommended heat flux. When the specimen has lost 70% of its mass loss, the exposure chamber is sealed and isolated. The corrosion of the target is determined by exposing the target to the now static combustion products 20 & 3* CO co o for one hour measured from the start of the test. The target is then placed in an environmental chamber at 75% relative humidity at 23C for 24 hours. The test method measures the increase in electrical resistance of a metallic circuit. This increase is related to the decrease in conductive cross-sectional area resulting from metal loss due to corrosion. The increase in electrical resistance of each target is determined throughout the test and correlated to its metal loss. The 24 hour corrosion value is reported as metal loss in Angstroms. In this study, heat fluxes of 25 and 50 kW/m2 were used to simulate two different fire scenarios. All of the materials were run at SOkW/m2 and 12 materials were run at 25 kW/m2. Two targets, one with a span of 2,500 A and the second with a span of 45,000 A, were used during each test at each heat flux. The results of this study indicate that the measured corrosivity of materials: fil does not correlate consistent with the expectations based upon the known chemistry of their compositions ('iil varies numerically with the heat flux under which the tests are run andon the target used to obtain the corrosion data and fiiil although numerically different, loosely ranks the corrosive potentials of the materials in a consistent manner at both heat fluxes and with both targets. The test protocol does not specify either the heat flux or the targets to be used recommending both in the Appendix. As corrosion values are numerically dependent on the conditions and target used to obtain the data, it is questionable how this test method can be used as a standard for determining and comparing the corrosion potentials of materials without requiring that both the specific heat flux and the target be specified in the test protocol as well as be reported with the results. To complete the review of corrosion test methods, a comparison of the corrosive potentials of the 24 materials using the four test methods will be made and one test method recommended for use as a global standard. INTRODUCTION The effects of smoke and corrosive gases released in a fire are gaining increased attention due to the damage they cause to electronic equipment. While tests to determine smoke obscuration have been standardized and tests to determine toxicity are available, only recently have tests to measure smoke corrosivity been developed. Recent fire events have underscored the need for standardized tests that measure corrosivity, as corrosion is particularly important with fires in and around computer and telecommunications equipment To better define test methods that measure smoke corrosivity of polymeric materials and products used in the wire and cable industry, the Polyolefins Fire Performance Council (PFPC) initiated a program to compare corrosivity test methods under discussion around the world today. The specific goal of the program is to determine the suitability of proposed test methods for measuring corrosive potentials of polymeric materials. The Council has been an operating unit of The Society of the Plastics Industry, Inc. since 1989. It consists of the eight member companies listed in Table I. 2 R&S 144331 Table I. Member Companies of the Polyolefins Fire Performance Council Carl F. Tripp, Managing Director BP Chemicals Dow Chemical Company GE Plastics Himont Inc. J.M.Huber Corporation, Solem Division Lonza Inc. Quantum Chemical Corporation, USI Division Union Carbide Corporation To accomplish this program goal, the PFPC decided to evaluate the four methods currently under consideration for global standardization: (i) the radiant combustion/exposure smoke corrosivity test method proposed by ASTM EOS.21.70, (ii) the CNET combustion test method under review by ISO TC61/SC4AVG2 and IEC TC89AVG3, (iii) the DIN 57 472 acid gas standard test method which measures pH and aqueous conductivity changes and (iv) the cone corrosimeter test method under review by ASTM D09.21.04 which uses a conical radiant heater system for generation of combustion gases. The results using the ASTM EOS.21.70 proposed "Standard Test Method For Measuring the Corrosive Effect of Smoke From the Burning or Decomposition of Materials and Products" with a radiant combustion/exposure apparatus were presented at the November 1991 International Wire and Cable Symposium (1). The data demonstrated that this test method did differentiate the corrosion potential of polymeric materials consistent with the expectations based on the known chemistry of their compositions. The results of the CNET test method "Plastics - Smoke Generation - Determination of the Corrosivity of Fire Effluents (Static Method)" were presented at the Seventeenth International Conference on Fire Safety held in January 1992 (2). The CNET test method did not appear to measure an expected corrosion response for some materials, such as chlorinated polyolefins, fluoropolvmers and polvethvlenes with chlorinated and brominated additives. The information generated in the CNET test may be useful for some type of corrosive index where it is desired to take into account the presence of a flame to maintain combustion, but based on the incomplete sample weight loss, not for determining the corrosive potential of materials. The results of the modified DIN test method "Testing of Cables, Wires and Flexible Cords; Corrosivity of Combustion Gases DIN 57 472 Part 813 Standard" were presented at the Fire Retardant Chemicals Association March 1992 meeting (3). The corrosive potentials of the 24 materials were differentiated as determined bv solution conductivity and pH and correlated well with the results obtained from the ASTM E05.2I.70 test method, The modified DIN results did not correlate with the % Corrosivity Factor "COR" obtained from the CNET test method The polymeric materials evaluated include a wide and balanced selection of wire and cable coating materials. The list contains non-halogenated filled polyolefins, polyolefins R&S 144332 3 containing halogen flame retardants, halogenated polymers, high temperature thermoplastics, polyolefin and nylon base resins plus a Douglas fir reference material. All of the cone corrosimeter experiments were performed at the Underwriters Laboratory at Northbrook, Illinois under the direction of Dr. Pravinray Gandhi. In a future publication, the corrosive potentials of these 24 polymeric materials determined using the four different corrosivity test methods will be compared. This will be the first time a direct comparison of the test methods using identical polymeric materials has been made. It is particularly important as corrosivity test methods are being standardized to determine and understand whether each test gives accurate results as well as relative results in terms of ranking materials by their corrosive potentials. MATERIALS EVALUATED Table II lists the 24 polymeric materials evaluated for smoke corrosivity. These resins and compounds represent a broad sampling of polymeric materials that are used in wire and cable jacketing and insulation applications. They are all commercially available. The table covers polymeric materials that retard flame spread by several means. Some materials contain hydrated mineral fillers compounded with polyolefin resins or resin blends (Samples 1,4,10,11,12,13,14 and 15). They retard flame spread by endothermic release of water-of-hydration. One material is an intumescent polypropylene compound containing a phosphorus flame retardant additive (Sample 8). This material retards flame spread by formation of a char barrier and its flaming characteristics are different from most of the others listed in the table. There are two types of halogenated flame retardant materials included. Some contain halogenated additives (Samples 16 and 25) and others have halogen chemically bonded to the polymer backbones (Samples 2,3,17,18,19 and 20). The brominated and chlorinated materials deter flame spread by halogen release and subsequent formation of flame poisoning radicals. They act in the vapor phase to extinguish flame, often through the synergistic action of the halogenated species with antimony oxide. The fluoropolymers are halogenated materials that have a very high limiting oxygen index and retard flame spread by being difficult to ignite. Included arc three high temperature thermoplastic materials. Sample 5 is a polymer alloy containing a phosphorus flame retardant additive. Sample 6 is an unfilled PEI resin and Sample 7 is a PEI/siloxane copolymer that resists flame spread by the formation of a char barrier. Samples 6 and 7 are also difficult to ignite so they also retard flame spread by resisting ignition. Also included are three unfilled base resins (Samples 21,23, and 24) from different suppliers so that a comparison of filled versus unfilled resins could be made. Finally, Douglas fir (Sample 22) was included as a standard as it is widely used as a 4 R&S 144333 control in the flame testing of materials for the building industry. EXPERIMENTAL SCOPE OF THE A STM DQ9.21.04 TEST METHOD This fire test response standard measures the corrosive effect, defined as the loss of metal, from exposure to combustion products of materials, components or products using a cone corrosimeter. A cone corrosimeter is defined as the equipment used to measure corrosion by metal loss using a conical radiant heater, a load cell, an exhaust system and a gas sampling system. SUMMARY OF TEST METHOD In this test method, a specimen is subjected to radiant heat at the recommended heat flux using a spark igniter to ignite combustible vapors. A portion of the products of decomposition or combustion are channeled through an exposure chamber in a dynamic mode in which the corrosion targets are placed until the specimen has lost 70% of its total mass loss, determined from previous experiments at the recommended heat flux. When the specimen has lost 70% mass loss, the exposure chamber is sealed and isolated. The corrosion on the target is determined by exposing the target to the now static combustion products for one hour measured from the start of the test, followed by 24 hour exposure of the target to 75% relative humidity at 23C in an environmental chamber. The test method measures the increase in electrical resistance of a metallic circuit using a corrosometer, a modified Kelvin bridge type instrument that measures the change in resistance of targets. This increase in resistance is related to the decrease in conductive cross-sectional area resulting from metal loss due to corrosion. The increase in electrical resistance of each target is monitored throughout the test and correlated to its metal loss. The 24 hour corrosion value is reported as metal loss in Angstroms. The test protocol does not specify either the heat flux or the targets to be used. Both are recommended in the Appendix. The Appendix however, is not part of the test method. In this study, heat fluxes of 25 and 50 kW/m2 were used to simulate two different fire scenarios. Two targets were used during each test at each heat flux, Model 610 with a span of 2500 A and Model 030788-S0.35-8061 with a span of 45,000 A. The targets are available from Rohrback Cosasco Systems. CONE CORROSIMETER APPARATUS The test apparatus, shown in Figure 1, uses a Cone Corrosimeter or the Cone Calorimeter test equipment as described in ASTM E 1354-90 with the addition of a gas sampling system. The system consists of a conical heater, a temperature controller, an exhaust system, a load cell, a specimen holder, an ignition circuit, a gas sampling system and a corrosion target. 5 R&S 144334 The gas sampling system consists of a conical funnel, stainless steel gas sampling tip, stainless steel tubing, electrical heating tape, silicone rubber tubing, flowmeter, exposure chamber, target support stand and pump. The exposure chamber is shown in Figure 2. Table II. Polymeric Materials Evaluated PFPC Company Product Number Designation Material Description 1 BP 2 DOW EXP 839 5435-30-11 XL olefin elastomer with metal hydrate filler Blend of HDPE and chlorinated PE elastomer 3 DOW 5348-40-1 4 EXXON EX-FR-100 Chlorinated PE with fillers EVA polyolefin with ATH filler 5 GEP NORYLPX1766 Polyphenylene oxide/polystyrene blend 6 GEP ULTEM 1000 Polyetherimide 7 GEP SILTEM STM 1500Polyetherimide/siloxane copolymer 8 HIMONT EXP 127-32-6 Intumescent polypropylene 9 LONZA 10 UCC UNIGARDTMRE Nylon with mineral filler Polyolefin copolymer with mineral filler DFDA-1736NT 11 UCC UNIGARDTMRE XL polyolefin copolymer with mineral filler HFDA-1393BK 12 QUANTUM PETROTHENE XL polyolefin copolymer with ATH filler XL 7403 13 QUANTUM PETROTHENE XL polyolefin copolymer with ATH filler YR 19535 14 QUANTUM PETROTHENE EVA polyolefin with mineral filler YR 19543 15 UCC UCARSIL Polyolefin with mineral filler FR-7920NT 16 BP POLYCURETM 798 XL polyethylene copolymer with chlorinated additive 17 COMMERCIAL SAMPLE A polyvinylidene fluoride material 18 COMMERCIAL SAMPLE A polytetrafluoroethylene material 19 COMMERCIAL SAMPLE A PVC material 20 COMMERCIAL SAMPLE A PVC building wire compound 21 UCC DGDK-3364NT Polyethylene homopolymer 22 Douglas fir 23 QUANTUM ULTRATHENE EVA polyolefin copolymer UE 631 24 NYLET P50 Nylon 6,6 25 UCC UNIGARDTMHP XL polyethylene copolymer with HFDA-6522NT brominated additive R & S 144335 (> Figure 1 Equipment for Corrosion Testing W/t'/wnov 7 R&S 144336 SAMPLES The test protocol states that the types of specimens permitted are materials in the form of a flat plaque, constituents of a product, or an end-use product When the specimen is a material or a constituent of a product, the specimen size is 100 x 100 x 6.3 mm (4 x 4 x 0.25 inches). The specimens used for this study were molded plaques. TEST PROCEDURE After setting the exhaust flow rate to 0.024 m3/s, both the heater flux and the load cell are calibrated. The initial mass of the specimen is recorded. The specimen is wrapped in a single layer of aluminum foil and the foil cut along the open edges of the specimen so that the top of the specimen is exposed. The grid, used to prevent fouling of the spark igniter with samples that intumesce, is placed on top of the specimen and the assembly is placed in the specimen holder. The sampling system is checked for leaks by pinching off the flexible tubing with clamps and noting that the flow from the flowmeter drops to zero. Prior to conducting corrosion tests, two mass loss determination tests were conducted, without corrosion gas sampling, on each material to establish the total mass loss at the designated heat flux. The data is reviewed to determine 70% mass loss using an average of the data from the two tests. The test is started by placing the specimen, held in the sample holder, on the load cell and simultaneously starting the ignition timer and the data collection system. The spark plug is placed over the test specimen and the power turned on. Termination of the sampling flow occurs when the specimen has lost 70% of the total mass loss determined previously or 60 minutes have elapsed after the start of the test The inlet and outlet lines of the exposure chamber are sealed with clamps and the combustion products are allowed to react with the corrosion targets for a total of 1 hour measured from the start of the test. The targets are then placed in an environmental chamber with a dry bulb temperature of 23C and 75% relative humidity for 24 hours. After 24 hours in the environmental chamber the resistance value of the corrosion target is recorded and the metal loss determined. RECORDED DATA The following information is recorded during the test: sample identification and composition, date, room temperature and relative humidity, specimen thickness and mass, radiant heat flux, corrosion target span and identification, volumetric sampling rate, time to sustained flaming, mass lost data, exposure chamber temperature, time of test start, sampling time and corrosometer values of the targets at the start of the test, at the end of the 1 hour combustion products exposure and at the end of the 24 hours of post-exposure in the environmental chamber. 8 R&S 144337 FLAMMABILITY PARAMETERS During the experiments to determine total available mass loss the following flammability data was obtained for the 24 materials and will be reported in a future publication: time to ignition, maximum heat release, total heat release, average heat of combustion and average specific extinction coefficient. This is an extremely valuable portion of the cone corrosimeter test method as it provides a variety of flammability parameters on materials in addition to corrosion data. SUMMARY.DF CORROSION DATA The corrosion results obtained at a heat flux of 25 kW/m2 using both targets are shown in Table III with the results at 50 kW/m2 shown in Table IV. For ease of comparison, the 25 kW/m2 data is plotted in Figure 3 with Figure 3A illustrating the data from the 2,500 A span target and the data from the 45,000 A span target shown in Figure 3B. The 50 kW/m2 data is shown similarly in Figure 4. A comparison of the two heat fluxes is plotted in Figures 5A and 5B. DISCUSSION OF. RESULTS DRIREINQ During the testing, a problem occurred with the PTFE Sample 18, the PE homopolymer Sample 21 and the EVA homopolymer Sample 23. At both heat fluxes all of the specimens of these materials melted over the edges of the sample pan dripping off the load cell onto the bottom of the combustion chamber. When this occurred the determination of mass loss became invalid as did the corrosion results. This problem could be corrected by the use of a sample pan with higher walls or by placing the smaller sample pan in a second larger pan, as long as the irradiance level remains constant, so the sample materials do not fall off the load cell when this phenomenon occurs. COMPARISON OF TARGETS Several observations were made after reviewing the data. The first is the apparent severity of the test method. At 50 kW/m2 with the 2,500 A span target, most of the materials appear to be corrosive. This data is difficult to correlate with the compositions of the materials tested. It is not clear why Samples 1,4 and 11 containing metal hydrate, ATH and mineral fillers respectively and unfilled nylon Sample 24 showed corrosive potentials greater than Sample 25 containing a brominated flame retardant additive. This raises serious questions regarding the accuracy and usefulness of this test method. The second observation is that the corrosion values varied numerically with the target used to obtain the data. The differentiation of the corrosive potentials using the 2,500 A span target is limited by the large number of materials where the corrosion exceeded the capacity of the target driving them off-scale (Samples 3,17 and 19 at 25 kW/m2 and Samples 2,3,16,17,19 and 24 at 50 kW/m2). 9 R&S 144338 The 45,000 A span targets are less sensitive especially with materials having low corrosion. A distinct advantage of the higher span targets is that they provide a broader overall differentiated picture of corrosive potentials. The samples that went off-scale in Figure 4A are clearly differentiated in Figure 4B. This data indicates that there is no one target available to determine, with equal sensitivity, the corrosivity of materials with both high and low corrosion potentials. The sputtered 2,500 A span targets have thickness and surface morphology limitations inherent to the sputtering process and the 45,000 A span targets, constructed from the thinnest copper film currently available, appear to exhibit a slower rate of corrosion man the sputtered target. Few field products use either technique raising the question of which target design is representative of end-use applications. The test protocol does not specify the target to be used. The Appendix, a nonmandatory part of the test method, recommends the use of the 2,500 A span target for materials with low corrosivity and the 45,000 A span target for materials with high corrosivity. The protocol states that the corrosion values obtained using either target at either heat flux are numerically interchangeable and suitable for use in fire risk assessment The data presented clearly shows this is not the case. The 2,500 A span target yields corrosion values consistently higher than those obtained using the 45,000 A span target except when the capacity of the lower span target is exceeded. As it is not known which results are accurate, the authors recommend that no attempt should be made to interchange the numerical corrosion values from the two model probes. The third observation is that the corrosion results, although numerically different, loosely ranked the materials in a consistent manner with both targets at both heat fluxes. Additional testing would be required to determine if a correlation truely exists between the corrosion values obtained from the two targets. COMPARISON OF HEAT FLUXES A comparison of the two heat fluxes used in this study, 25 kW/m2 and 50 kW/m2, is shown in Figure 5. Four of the samples did not ignite at the lower flux. All of the samples ignited at the higher flux level. Figure 5A compares the corrosivity using the 2,500 A span target At the higher heat flux, the corrosivity of some materials increased (Samples 1, 6 and 12), some stayed the same (Sample 25) other than those that went off-scale and some decreased (Samples 8 and 11). In Figure 5B with the 45,000 A span target, no corrosion levels increased, some remained the same (Samples 1,6, 8,11 and 25) and some decreased (Samples 3,17 and 19) at the higher flux. However, the same materials did not respond similarly at both heat fluxes and with both targets. Lacking an understanding as to the reasons for these materials to behave differently, the conditions under which the corrosivity is determined must be specified in the test protocol as well as be reported with the results. 10 R&S 144339 REPRODUCIBILITY OF THE TEST METHOD To better define the reproducibility of the test method, 6 specimens of Sample 6 were run at 50 kW/m2 so that a reasonable statistical treatment of the data could be performed. Previous papers used Sample 21 for reproducibility data but in this study Sample 21 dripped off the load cell invalidating the corrosion data so data from Sample 6 was used. Table V. Standard Deviation Calculations For Sample 6 FPC# 6A 6B 6C 6D 6E 6F AVERAGE STANDARD DEVIATION Time to Ignition (sec.) 108 124 138 135 134 114 126 12 Time to 70% Mass Loss (minutes) 13 15 13 15 14 15 14 1 24 hr Metal Loss 2,500 A Target (Angstroms) 1546 697 126 2305 2004 (2730) 1568 24 hr Metal Loss 45,000 A Target (Angstroms) 360 248 0 473 608 653 390 992 244 The standard deviations of both corrosion values are considered high as they are 63% of the average, indicating a wide measure of dispersion in.the data. This points out a needed improvement with this test method, especially if it is to be considered for use as a global standard. It is interesting to note that Samples 6A, B and C were run consecutively at one time and Samples 6D, E and F were run consecutively at a later date. A review of' the data on Tables III and IV reveals many instances where one data point appears to be out of line, skewing the average. Additional runs are needed to provide sufficient data to determine whether such points can be statistically discarded. In this paper all data is presented with none discarded. R&S 144340 ll CONCLUSIONS AND RECOMMENDATIONS 1. THE CORROSIVITY OF SOME OF THE POLYMERIC MATERIALS EVALUATED USING THIS TEST METHOD DID NOT CORRELATE IN A MANNER CONSISTENT WITH EXPECTATIONS BASED UPON THE KNOWN CHEMISTRY OF THEIR COMPOSITIONS. THE TEST METHOD DID NOT APPEAR TO MEASURE AN EXPECTED CORROSION RESPONSE TO SOME MATERIALS, SUCH AS UNFILLED MATERIALS, MINERAL FILLED MATERIALS AND A MATERIAL CONTAINING A BROMINATED ADDITIVE. 2. THE NUMERICAL CORROSION VALUES OF THE MATERIALS VARIED INCONSISTENTLY WITH BOTH THE HEAT FLUX AND THE TARGET USED. IT IS RECOMMENDED (A) THAT THE NUMERICAL CORROSION VALUES NOT BE USED INTERCHANGEABLY AND (B) TO AVOID CONFUSION, BOTH THE HEAT FLUX AND TARGET SHOULD BE SPECIFIED IN THE TEST PROTOCOL AS WELL AS REPORTED WITH THE RESULTS. 3. THE 2 300 A SPAN TARGETS YIELD CORROSION VALUES CONSISTENTLY HIGHER THAN RESULTS OBTAINED USING THE 45,000 A SPAN TARGETS, GOING OFF-SCALE WITH MATERIALS WITH HIGH CORROSION. 4. THE TARGETS WITH 45,000 A SPAN ARE LESS SENSITIVE RESULTING IN LOWER CORROSION VALUES THAN THOSE OBTAINED USING THE 2300 A SPAN TARGETS. THEY PROVIDE DIFFERENTIATION OF MATERIALS WITH BROADER CORROSIVE POTENTIALS, BUT WITH LESS RESOLUTION. 5. CORROSION OCCURRED EVEN WHEN THE SAMPLES DID NOT IGNITE. 6. SEVERAL MATERIALS BUBBLED OUT OF THE SAMPLE PAN DRIPPING OFF THE LOAD CELL. A CHANGE IN PAN DESIGN TO KEEP THIS MATERIAL ON THE LOAD CELL WOULD PREVENT LOSS OF DATA WHEN THIS PHENOMENON OCCURS. 7. THE REPRODUCIBILITY OF THE DATA USING THIS TEST METHOD IS POOR, POINTING OUT A FUTURE DEVELOPMENTAL NEED. AT THIS TIME IT IS NOT UNDERSTOOD WHETHER THE LACK OF REPRODUCIBILITY RESULTS FROM THE COMBUSTION METHOD, THE GAS SAMPLING TECHNIQUE OR THE TARGET MEASUREMENT SYSTEM. 12 R&S 144341 ACKNOWLEDGMENTS The authors would like to thank Dr. Pravin Gandhi, Bob Backstrom, Gary Gardell, John Donovan and Dan Plyman of Underwriters Laboratory Northbrook, Illinois for their contributions in obtaining these results. REFERENCES 1. Kessel, S.L., J.G.Bennett, Jr. and C.E.Rogers, "Corrosivity Test Methods For Polymeric Materials", International Wire & Cable Symposium Proceedings, pg. 348, November 1991. 2. Rogers, C.E., S.L. Kessel and J.G. Bennett Jr., "Corrosivity Test Methods for Polymeric Materials Part 2-CNET Test Method", International Conference on Fire Safety Proceedings, pg. 392, January 1992. 3. Bennett, J.G. Jr., S.L, Kessel, C.E. Rogers, Corrosivity Test Methods For Polymeric Materials Part 3-Modificd DIN Test Method", Fire Retardant Chemicals Association Proceedings, pg. 79, April 1992. 13 R&S 144342 PFPC # Table HI. Summary of Test Results at 25kW/m2 Time to Ignition (sec.) Time to 70% Mass. Loss (minutes) 24 hr Metal Loss 2,500 A Target (Angstroms) 24 hr Metal Loss 45,000 A Target (Angstroms) 1A IB 1C 1 AVERAGE 3A 3B 3C 3 AVERAGE 6A 6B 6C 6 AVERAGE 8A 8b 8c 8 AVERAGE 418 - "0 a 410 160 163 175 166 DNI DNI DNI DNI 114 111 122 116 24 25 26 25 17 16 17 17 15 19 14 16 29 31 25 28 592 2151 1338 1694 (2675)* (2675)* (2675)* (2675)* 284 120 266 233 1965 1966 2026 1986 11A 1 IB 11C 11 AVERAGE 389 400 392 394 23 21 21 22 (2675)* 1934 2342 (2317) 12A 12B 12C 12 AVERAGE 673 690 690 684 22 22 21 22 206 96 156 153 17A 17B 17C 17 AVERAGE DNI DNI DNI DNI 23 23 23 23 (2675)* (2675)* (2675)* (2675)* 18A 18B 18C 18 AVERAGE DNI DNI DNI DNI +* ** ** ** ** ** ** ** * READING IN EXCESS OF TARGET SPAN ** SAMPLE DRIPPED; MASS LOSS / CORROSION DATA NOT VALID ** SAMPLE NOT AVAILABLE AT TIME OF TESTING () ASSUMES MAXIMUM TARGET SPAN DNI = DID NOT IGNITE NR - NOT RECORDED 0 1980 450 810 7380 6345 2790 5505 135 113 158 135 810 540 585 645 495 608 810 638 0 0 180 60 2790 2340 4455 3195 ** ** ** ** R&S 144343 14 PFPC ii Table in. Summary of Test Results at 2kW/mJ Continued Time to Ignition (sec.) Time to 70% Mass Loss (minutes) 24 hr Metal Loss 2,500 A Target (Angstroms) 24 hr Metal Loss 45,000 A Target (Angstroms) 19A 19B 19C 19 AVERAGE 21A 21B 2IC 21 AVERAGE 25A 25B 250 25 AVERAGE DNI DNI DNI DNI 227 228 234 236 516 514 492 507 14 14 14 14 ** ** ** ** 12 12 11 12 (2675)* (2675)* (2675)* (2675)* **' ** ** ** 1882 1990 1098 1657 17505 20745 23220 20490 ** ** ** ** 698 630 113 480 PFPC# Table IV. Summary of Test Resultsat 50 kW/m1 Time to Ignition (sec.) Time to 70% Mass Loss (minutes) 24 hr Metal Loss 2,500 A Target (Angstroms) 24 hr Metal Loss 45.000 A Target (Angstroms) 1A IB IC 1 AVERAGE 2A 2B 2C 2 AVERAGE 3A 3B 3C 3 AVERAGE 4A 4B 4C 4 AVERAGE 91 86 90 89 54 49 50 51 38 38 39 38 86 88 86 87 16 18 20 18 11 11 11 11 9 NR 10 10 14 NR 15 15 (2675)* 1568 (2675)* (2305) (2675)* (2675)* (2675)* (2675)* (2675)* (2675)* (2675)* (2675)* 2030 1949 1398 1792 606 450 743 600 4095 4320 5738 4718 4905 1350 189 2148 428 450 315 398 15 r &S 144344 PFPC # Table IV. Summary of Test Results at 50 kW/m1 Continued Time to Ignition (sec.) Tir'e to 70% Mass Loss (minutes) 24 hr Metal Loss 2,500 A Target (Angstroms) 24 hr Metal Loss 45.000 A Target (Angstroms) 5A 5B 5C 5 AVERAGE 6A 6B 6C 6 AVERAGE 7A 7B 7C 7 AVERAGE 8A 8B 8C 8 AVERAGE g *** 10A 10B 10C 10 AVERAGE 11A 1 IB 11C 11 AVERAGE 12A 12B 12C 12 AVERAGE 13A 13B 13C 13 AVERAGE 55 55 55 55 108 124 138 123 66 80 74 73 27 24 24 25 131 127 126 128 123 122 115 120 139 138 160 146 70 74 67 70 5 5 5 5 13 15 15 14 8 10 9 9 7 9 10 9 17 16 17 17 15 15 15 15 12 12 12 12 11 11 12 11 (2675)* 797 1445 (1639) 1546 697 126 790 1755 1511 1394 1553 1715 1883 1418 1672 1926 1440 1899 1755 2009 1935 2036 1993 263 380 350 331 1386 1621 2386 1798 405 405 630 480 360 248 0 203 585 653 383 540 653 585 383 540 608 563 675 615 583 473 810 622 90 0 135 75 518 653 765 645 R&S 144345 16 PFPC # Table IV. Summary of Test Results at 50 kW/m1 Continued Time to Ignition (sec.) Time to 70% Mass Loss (minutes) 24 hr Metal Loss 2,500 A Target (Angstroms) 24 hr Metal Loss 45,000 A Target (Angstroms) 14A 14B 14C 14 AVERAGE 15A 15B 15C 15 AVERAGE 16A 16B 16C 16 AVERAGE 17A 17B 17C 17 AVERAGE 18A 18B 18C 18 AVERAGE 19A 19B 19C 19 AVERAGE 20*** 21A 21B 21C 21 AVERAGE 22A 22B 22C 22 AVERAGE 65 45 63 58 86 91 88 88 83 61 75 73 766 891 803 820 204 212 NR 208 33 41 39 38 67 61 NR 64 29 27 29 28 7 7 9 8 14 13 13 13 7 6 6 6 12 12 13 12 ** ** ** ** 9 10 10 10 t* ** ** ** 13 NR 13 13 1519 (2675)* 1962 (2052) 1901 1736 (2675)* (2104) (2675)* (2675)* (2675)* (2675)* (2675)* (2675)* (2675)* (2675)* ** ** ** ** (2675)* (2675)* (2675)* (2675)* ' ** ** ** ** 1645 (2675)* 1420 1913 473 495 518 495 630 450 608 563 1845 1485 2565 1965 2925 2385 2048 2453 *+ ** ** ** 2903 5760 6953 5205 ** ** *+ ** 585 698 383 555 R&S 144346 17 PFPC # Table IV. Summary of Test Results at 50 kW/m1 Continued Time to Ignition (sec.) Time to 70% Mass Loss (minutes) 24 hr Metal Loss 2.500 A Target (Angstroms) 24 hr Metal Loss 45.000 A Target (Angstroms) 23A 23B 23C 23 AVERAGE 24A 24B 24C 24 AVERAGE 25A 25B 25C 25 AVERAGE 47 49 56 51 114 70 96 93 89 93 114 99 ** ** ** ** 6 5 5 5 6 6 5 6 ** ** ** ** (2675)* (2675)* (2675)* (2675)* 1893 1654 1522 1700 ** ** ** ** 878 878 3375 1710 450 315 540 435 R& s I44347 18 FIGURE 3A CORROSION USING THE ASTM D09.21.04 TEST METHOD 25kW/m2 2500 SPAN <r OZD 1 368tt12fT18l92125 PFPC SAMPLE NUMBER FIGURE 3B CORROSION USING THE ASTM D09.21.04 TEST METHOD 25kW/m2 <o 45000 SPAN PFPC SAMPLE NUMBER Figure 3 Comparison of Corrosivity Using Different Targets at 25 kW/m1 19 R&S 144348 FIGURE 4A CORROSION USING THE ASTM D09.21.04 TEST METHOD 50kW/m2 2500 SPAN 2000 1000 1 2 3 4 5 6 7 8 10 11 12 13 14 15 16 17 18 19 21 22 23 24 25 PFPC SAMPLE NUMBER FIGURE 4B CORROSION USING THE ASTM D09.21.04 TEST METHOD 50kW/ra2 45000 SPAN 1c/r5 3 PFPC SAMPLE NUMBER Figure 4 Comparison of Corrosivity Using Different Targets at 50 kW/mJ 20 R&S 144349 FIGURE 5A CORROSION USING THE ASTM D09.21.04 TEST METHOD 25kW/m2 50 kWn\2 PFPC SAMPLE NUMBER 1cZ03r 6000 CC FIGURE 5B CORROSION USING THE ASTM D09.21.04 TEST METHOD 2SkW/m2 45000 SPAN 50kW/m2 45000 Span 00,490) 5 4000 oc I < z W5 2000 <O/> [idLU 2 JO. s s * ii i2 tr PFPC SAMPLE NUMBER 5 L 21 25 Figure 5 Comparison of Corrosivity at 25 kW/m2 and 50 kW/m2 Using Targets With Different Spans 21 R&S 144350