ML20127E295
| ML20127E295 | |
| Person / Time | |
|---|---|
| Site: | Fermi, Harris, Columbia, Brunswick, River Bend, Ginna, Robinson, Comanche Peak  |
| Issue date: | 09/03/1992 |
| From: | Mariotte M NUCLEAR INFORMATION & RESOURCE SERVICE |
| To: | Selin I, The Chairman NRC COMMISSION (OCM) |
| Shared Package | |
| ML17262B024 | List: |
| References | |
| CON-#193-13388 2.206, EA, NUDOCS 9210060439 | |
| Download: ML20127E295 (49) | |
Text
{{#Wiki_filter:__ _. _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ . . _ _ - _ _ _ - _ _ _ _ _ t (,' , -f C n)
/838T Ii-DOCKET NUMBER -
$ 'g{g PROD. & UTIL FAC.-s Q-Mh,2 h I w = ; ; n..=
i
~= q==M = a-v '92 tm 20 A9:35 '
Nuclear Information and Resource, Service, 142416th Street, N.W., Suite 601, Washington, D.C. 20036 (202) 328 0002 21 pQ,/ "VU Bo:rd of nitectors
*sev Drev sitoui.Mo September 3, 1992 setn occrai e
['l,'M",n lion. Ivan Selin sew sork.NY Chairman ma Hak U.S. Nuclehr Regulatory Commission
$,$dYA),'o ,,, Washington, DC 20555 '
New iork, NY R4 8o'd*" G ron. OH Dear Chairman Selint arvet hes,towenthal chase Mo Enclosed please find an appeal, under the provisions i Mar' Morna of 10 CFR 2.206, of the NRC ctaff's August 19, 1992
. )l] "'r,',[ denial of our petition for emergency enforcement a .,n.neton. oc action of July 21, 1992 and addenda of August 12, ei.n nens 1992, w aihinaion mc SuiCio""s$ "' " We look forward to your early response. -
National Advhory Board be,c stierwmi Sincerely, f ederation 01 Americ an voenints* 3 une Allen North Anna Invironmental Coahtion'
/' /
' g A ,1 / *,3 -
f/ g , Robert llat k us * -
)
Backuuhea & Moser* Michael'Mariotte fn'$$ie No'Nrn to, Executive Director Pubhc Health
- 9arbara Bonon Actrett Bruc e Cockbur n Musician David Cohen Advocacy institute * (- / ,
Harlan likson
~ Author f7yk~l
- Or, lack Geiger, M D Prof of Comm Medicine, CUNY Medical Schoot*
Marta Gbbs & Attreu , Whoopi Goldberg At trest
' Janet Hogle Blue Ridge
( Invironmental Celerwe league
- Dr. H W,1bsef Cahtornia State Unnersity*
^
'N, nvironene t ( ahtion on
' Nuricar Power * '" [ '
Charien Aomsmott ,
*'") k r
Aomanoti(nergy Anoogleg*
. Dr. Marv,n Resnektaff [ p '. e Q P l xj Radsoaclave Waste p h
, Management Anociale5* _
Mary knelair - Grert tales (nesty Allianre* t
- 7 '
{ Hasvey Wawerman Author ' .Jggj(gggdgoa gyndngn.nydggygngygypyh(y, .
*oreanuations h$ted wt.mnYmnvrnuram e anv<ias rar e'-
? ,, Ato.e} . %
D40
~
Rhg DmiAmT g igmjd_ coRocoef dhe%* . 2 Nuclear Information and Resource Seriice 142416th Street N.W. Suite 60L Washington, D.C. 20036(202) 328-0002 APPEAL TO TIIE NRC COMMISSIONERS OF NRC STAFF DENIAI, OF NIRS PETITIONS FOR IMMEDIATE ENFORCEMENT ACTION OF JUIN 21.1992 AND AUGUST '12.1992 i
- 1. INTRODUCTION On July 21,1992, the Nuclear Information and Resource Service (NIRS) submitted a petition to the Nuclear Regulatcry Commission (NRC) staff under the provisions of 10 CFR 2.206_asking.
for the immediate suspension of the operating license of Gulf States Uti!ities River Bend reactor, and for issuance of Generic Letter 92 xx (February 11, 1992). On August 12, 1992, NIRS submitted an addenda to this petition asking for immediate suspensions of the operating licenses of Carolina Power & Light's Shearon Harris and Robinson reactors, Detroit Edison's Fermi-2 reactor, Rochester Gas & Electric's Ginna reactor, Washington Public Power Supply System's WNP-2 reactor, and Texas Utilities Comanche Peak Unit 1 reactor. In addition, NIRS asked for a stop work order or suspension of construction permit for Texas Utilhics Comanche Peak Unit 2 reactor. NIRS' petitions' were, and are, based on the use of the above named reactors of a fire barrier called Thermo Lag. In the cases of the above named reactors, independent testing has proven that Hermo Lag is ineffective as a fire barrier material, and may even cause fires itself. NIRS acted because the continued use of Thermo Lag in the above-named reactors poses a clear and present danger to tite health'and safety of citizens living near these plants.. On August 19,1992, the NRC staff denied NIRS' petitions in their entirety, although, in - apparent recognition of the hazards caused by the use of nermo Lagi the staff said it would - Lissue a generic letter on this matter "in the near future." In further acknowledgment of the merits of NIRS' arguments, the staff described issuance of this generic letter as a "high priority." Also on August 19, the law firm of Winston & Strawn, acting on behalf of Gulf States Utilities, submitted a lengthy document in opposition to NIRS' petitions. NIRS now appeals to the NRC Commissioners the NRC staff's denial of our petitions, with the following qualifications: 1) NIRS removes our request that the operating license for the Ginna r 1 dedicated to a sound non nuclear eneny polics. m
( reactor be suspended. NIRS was in error when it stated the Ginna reactor uses hermo. Lag in a firewall con 0guration: in fact. Rochester Oas & Electric has informed us-to our complete satisfaction-that the Ginna reactor uses no Dermo Lag whatsoever. 2) Because of a communication error w4h a wmstleblower, NIRS incorrectly requested the suspension of the operating license of Carolina Power & Light's Robinson reactor. In fact, according to materials submitted to the NRC by Carolina Power & Light, this reactor does not use nemio Lag. Imtead, we should have requested the immediate suspension of the operating license for Carolina Power & Light's two unit Brunswick plant. We now add these two units to our petition for immediate enforcement action and remove Robinson from our petition. These two instances only highlight the difficulties involved in the Thermo Lag matter. In fact, only Thermo. Lag's manufacturer, nermal Science, Irm. (TSI) knows exactly which reactors have purchased and installed Hermo Lag, and even thL 'mpany may not know all the different condgurations with which this material has been installed nt out nation's nuclear reactors.-
- 11. IIACKGROUND ne NIRS' petitions of July 21,1992 and August 12,1992 listed numerous instances in which tests of Thermo Lag demonstrated that this material failed in its, function as a lire barrier, he petitions are incorporated by reference and attached to this appeal.
Several of these tests were conducted for the River Bend reactor, and included tests of reconfigurations. Several other tests were conducted for the Comanche Peak nuclear plants. In every instance, Thermo Lag failed, often catastrophically. On August 28,1992, the NRC issued NRC 13ulletin No. 92 01, Supplement 1, which detailed still more testing failures of Thermo Lag. Among these tests were full-scale tests of configurations used at Comanche Peak, using stock material and installed per ver. dor procedure. All but one test demonstrated failure. According the Bulletin, "the cables exhibited visible fire damage to cable jackets in all conduits," except one. In another test, " fire damage to the cables was also identdied during the post fire inspection, raising questions whether the cables would have functioned properly during a fire." Temperatures reached as high as 700 degrees Fahrenheit. The NRC noted, "Although previous 'ests conducted by TU Electric (Texas Utilities l (see Bulletin 92 01) resulted in the apparent successful performance of large diameter conduits and narrow trays, new information provided by these recent tests has led the NRC to believe the potential early failures of Thermo, Lag barriers are nc.t limited to specific sizes. He NRC considers the openings at the joints and seams of the Thermo Lag material to be of high significance " nc Bulletin also described four new NRC tests of Thermo Lag, in various installation con 0gurations. Thermo Lag failed all forr tests, in two cases, it burned completely through, with temperatures reaching as high as 1737 degrees and 1200 degrees, suggesting the possibility of
" voids" in the material.
Section 50.48(A) of Title 10 of the Code of Federal Regulations requires that each operating nuclear power plant have a fire protection plan that satisfies Appendix A to 10 CFR Part 50, 2
4 i General Design Criteria (GDC) 3, " Fire Protection." 10 CFR Part 50 Appendix R requires such fire protection plans for reactors receiving operating licenses after January 1.1979 and Appendix A to DTP 9.51 for all plants licensed prior to Januaiy 1,1979. Based on the test results listed above, all of the reactors named in this petition are in direct violation of NRC regulations, and pose an immediate threat to the health and safety of citizens . living near these plants. ! The NRC's fire protection regulations were established for good reason: the 1975 Browns Ferry fire nearly resulted in a nuclear meltdown and exposed serious weaknesses in the NRC's regulatoiy scheme. The NRC's fire protection regulations were adopted as the only responsible response to that fire. By denying NIRS* petitions, and allowing continued operation of reactors in direct violation of NRC regulations, the NRC staff has denigrated the importance of fire protection regulations and returned oui nation to the days before Drowns Ferry, when fire protection was not considered a major safety issue. The NRC Commissioners must act to re. establish the importance of fire protection at our nation's nuclear plants by approving NIRS' petitions and this appeal. Otherwise, the NRC Commissioners will be sending two, equally unacceptable messages: 1) to the American people, the NRC Commissioners would be stating that their health and safety is subordinate to nuclear industry financial considerations, and that the NRC's own critical safety regulations will not be enforced, even if this places American citizens in jeopardyt 2) to utilitic+., the NRC Commissioners would be stating that it does not intend to enforce safety regulations and that it is acceptable practice to seek to circumvent safety regulations. We trust that these are act the Commissioners' intentions, and that the Commissioners will approve NIRS' appeal. As the Commissioners well know, the NRC has denied 85 out of 85 of 10 CFR 2.206 petitions since 1985. Sooner or later, the Commissioners must act to reassure the public that safety comes first. Denial of 86 of 86 petitions would be simply too much to bear. l' and would strain the NRC's credibility beyond reason, particularly in a case as clear-cut as this one. The importance of this case can be stated in a paragraph: The NRC has no independent tests which suggest that Thermo. Lag can fulfill its vital safety function. The NRC has numerous independent tests-including some which the NRC hself has conducted-which document that : Hermo. Lag does not work in the tested configurr,tions. Fire barriers are an essential part of the NRC's fire protection regulations, nese fire barriers protect critical electrical cables necessary to control a nuclear reactor and provide emergency core cooling in the event of a major fire or accident. Without_these cables. reactors would melt down and kill necole. We note that on August 17,1992-two days before the NRC staff's denial of NIRS' petitions, a report by the NRC's Inspector General was released. This report contirms virtually every allegation made by NIRS about the inadequacy of nermo. Lag, except that it went even further than NIRS and declared that all existing TSI tests of Thermo-Lag were conducted by an unqualified laboratory using inadequate equipment and signed by personnel who had not even reviewed test results. 3 4 -. 7
n - In addition, the inspector General's report blasted the NRC staff for ten years of inaction on Hermo. Lag, despite repeated indications that the material does not meet NRC regulations. The NRC staffs denial of NIRS' petitions only underscores the IG's findings. Rathet than embark on a new course that would protect the A.c.erican public, the NRC staff has chosen to continue its policy of sweeping problems under the rug and to avoid enforcing its own regulations. The NRC Commissioners must act, now, to defend its inspector General and overturn the staffs denial of NIRS' petitions. III. RESPONSE TO AUGUST 19 1992 NRC STAFF STATEMENT IN DENIAL OF NIRS PETITIONS OF JULY 21.1992 AND AUGUST 12.1992 The NRC Staffs Response of August 19,1992 to NIRS' petitions is in error both legally and morally. The NRC staff admits that " fire resistive ratings and the ampacity derating factors" of Thermo. Lag are " indeterminate." The staff acknowledges that "some licensees have not adequately reviewed and evaluated the fire endurance test results and the ampacity derating test results used as the licensing basis for the Thermo Lag barriers: that some licensees have not adequa'ely reviewed .he Thermo Lag barriers instal lcd in their plants to ensure that they meet ' NRC requirements and guidance such as that provided in Generic Letter 8610 " Implementation of Fire Protection Requirements, " April 24,1986: and that "some licensees used inadequate or incomplete installation procedures during the construction of their nermo Lag barriers." In short, the NRC staff hz.s admitted many of NIRS' claims about the inadequacy of nermo. Lag as a fire barrier material where h has been tested. We note, however, that the NRC staff did not address NIRS' allegations in relation to the failure of the material to pass he>se stream tests (June 1992) nor the fact that the material has not undergone seismic tests. NIRS alleges that the material may not pass such tests. De NRC staff also did not address NIRS' allegation that nermo. Lag's quality assurance program is in such disarray that the company was unable to provide even the NRC with suitable material for testing. The NRC staff has acknowledged that the thickness of nermo Lag material it received for testing varied as much as one. aalf inch on a one inch specification. These issues alone are enough tc find the NRC staff's decision in error, llowever, this decision is in error in a number of other coratexts. The staff argues that " compensatory" rneasures, i.e. fire watches, are adequate to eneure the concept of " defense in depth" as it relates to fire watches, . He NRC staff essentially admits that plants using Thermo Lag are not in regulatory compliance. " Compensatory" measures, however, do not in themselves substitute for regulatory compliance. At best they are temporary actions taken to alleviate problems caused by a temporary inability of utilities to meet safety standards or regulations. In this case, the NRC staff argues that the compensatory measures ordered, i.e. fire watches. are adequate to ensure public health and safety even if Thermo-Lag cannot meet its intended 4
I 1 I
)
I function as a fire barrier. l First. as we have pointed out, fhe watches are no substitute for fire barriers. ney are an i additional means of fire Ostection, and would be entirely adequate if, for example, a video i camera used to deteci fires were temporarily out of service. Fire watches cannot. however, j protect vital electrical rables from fire: they can only alert fire brigades and offsite fire personnel that a fire exists. The NRC has provided no assurance that in every instance such fire suppression personnel can respond quickly enough to avoid cable damage absent an effective-Orc barrier. Indeed,if such assurance could be granted, there would be no need for regulations-governing fire barriers because there would be no need for fire barriers themselves. De fact ' that the NRC has insisted upon the need for such barriers, to the point of explicitly witing them in to its regulations, is itself enough proof that fire watches are an inadequete substitute for fire 'oarriers. Further, Branch Technical Position 9.5.115 (NUREO Of100) specifically states that the function of the fire protection system is "that a fire that is not promptly extinguished by the fire suppression activities will not prevent the safe shutdown of the plant." Effective fire barriers may be able to attain this goal. Iire watches cannot.
~
ne inadequacy of fire watches can easily be seen in NRC Intormation Notice 92 55 which.
> hows that Thermo Lag has failed in as little as 22 minutes. Fire watches at best only patrol on a one hour basis. In this instance, a fire could be underway for as much as 38 minutes before a fire watch could even sound the alarm, much less that fire suppression personnel could react.
Further, there is adequate documentation (at Comanche Peak. Saabrook, etc.) that utility personnel have not always taken Ore watches seriously and have falsified the logs ette ting that fire watches have been undertaken. Fire watches do no goo ( whatsoever when there is nobody watching. The use of fire watches as a substitute for fire barriers is a pervenion of the concept of , defense-in depth. Second, perhaps more importantly, the NRC staff has not identified a time. frame in which these
" compensatory" measures should be in effect. De NRC staff response gives no indication that these compensatory measures will be temporary. Thus, the NRC staff has given the nation's utilities carte blanche to ignore fire protection regulations for an indefinite time period. There can be no rationale for allowing licensees to avoid meeting NRC regulations for an indefinite period. If that were the case, there would be no reason for licensees to meet imy NRC regulation, regardless of safety significance, ne NRC has the abilityLto grant exemptions to r
regulations and amendments to licensees in cases where licensees cannot meet regulations or choose to meet them in a different manner than prescribed by regulation. In this instance, the NRC has not granted either exemptions or license amendments to any of the plants listed in NIRS' petitions. Thus, the NRC's acceptance-indeed encouragement--of compensatory measures is not only , inadequate from a public health and safety standpoint, but is legally and morally in error as well. Not only does this approach add to the public's risk from nuclear reactors but an indefinite, generic exemption from regulations is simply not acceptable and has no legal basis. If the NRC-5 4
wishes to exempt specific utilities from their requirement that fire protection systems meet regulations. it must do so on a case.by case basis by showing that in each individual Instance the exemption will not affect the public health and safety. In the case of Thermo. Lag, it will be impossible to make this determination. He NRC cannot rely, as it seeks to do in its August 19 denial of NIRS' petitions' upon the argument that 'in an actual fire situation, the fire resistance required of a barrier depends on the expected severity of the fire to which it is exposed. Typical nuclear plant fire loads are not great enough to produce a fine approaching the severity of a test fire. An actual fire at a nuclear power plant would yield a much slower temperature rise than did the test fire. Moreover, although the fire resistance ratings of certain Hermo-Lag fire barriers are considered indeterminate, tl.e NRC staff has evidence that the barriers will provide some level of fire protection. In addition, most plant areas have controlled ignition sources, which helps reduce the occurrences of fires, and are equipped with other passive and active fire protection feature which contribute to early fire detection and suppression activities." The recuirement is not that Thermo. Lag "will provide some level of fire protection" but that it will provide the specified one. hour or three. hour level of fire protection. Arguing that Thermo. Lag will provide some level of fire protection in the face of 10 years of failed tests is like arguing that an employee with a garden hose can substitute for an emergency core cooling systcm. His entire paragraph of thu NRC staff's response is nothing more than a negation of the NRC's fire protection regulations and responsibilities. If actual fires aren't great enough to produce problems which might challenge fire barriers, and if fires are so unlikely, then why have fire barrier regulations? In fact, the NRC staff's response is disingenuous and is contrary to fire protection regulations, which, we assume, were written for a reason. We remind the Coramissioners of 11 NRC 707 (1980) in which the Commissioners found that Appendix R establishes "the minimum acceptable fire protection requirements." Further, the Commissioners explicitly denied the argument that fires are unlikely to happen or that actual fires may be less severe than test fires by stating, "Our April 13 decision in no way permits reliance on probabilistic calculations to enter regulatory policy 'through the back door.' Denial of cmergency relief in this case is based upon our review of the fire protection program and the Sandia tests, and it is this review, and no probability analysis which assures us that public health and safety is not at undue risk." We further note that this decision emphasizes the importance of testing of fire protection' equipment, ". the staff's fire protection testing program is particularly important. We are concerned that the staff has still not completed plants and initiated tests which replicate typical fire protection measure being proposed for operating plants. .The Commission vien this testing-- program as a priority item ." In both MIRS original petition and addenda NIRS noted that NUREG.1150 indicates that the typical reactor will experience three to four signiticant fires over its operating lifetime, and that if there is a core meltdown, there is as much as a 50To chance that it was caused by fire. NUREG 1150 is the NRC's basic safety document. De NRC cannot have it both ways. If the NRC wishes to use this document to reduce safety risks in some areas, then it must use it in this area as well. 6
Given the NRC staff's failure over the past ten years, as documented by NIRS and the Inrpector General's report, in even addressing failed test results of Thermo Lag, these words should be taken quite literally. nird, the NRC has not adequately addressed NIRS' allegation that nermo. Lag itself is combustible, and can actually cause ilres. We elaborate on this point below (in NIRS response to Winston & Strawn brief). As for NIRS' allegation that Hermo. Lag can emit !,igh amounts of hydrogen cyanide when burned, NIRS hereby submits Proatec Report CTP 1099, July 14,1986, which references tests conducted by Southwest Research Institute and Southwest Certification Services in May 1986. This report, which tested the toxicity of Thenno. Lag and other fire barrier materials, found that levels of hydrogen cyanide and carbon monoxide released by combustion of Thenno Lag ' arc dtamatically above' lethal levels for animals According to the report, "It is a certainty that; animals, if employed during the tests reported here, would have survived only briefly in the-atmospheres produced by the thermal decomposition" of Thermo. Lag. While the report notes-some difficulty in extrapolating animal data to humans, the major difference is in the rate of-absorption, not in the ultimate effect. Levels of hydrogen cyanide released by nermo. Lag combustion were as much as 152 times higher than for another fire barrier rnaterial. Carbon dioxide releases could only be estimated, since releases exceeded the analyzer's range. A subsequent test reported on in the same document found somewhat lower, but still extremely high levels of both gasses. The Southwest tests obviously conflict with the NRC's NIST tests. Further resolution of this issue is needed. Ilowever, the NRC staff has misread our concern on this matter, We agree with the NRC staff that fire suppression personnel will typica'ly be supplied with protective clothing and usaterials. Fire watch nersonnel, on the other hand, are not typically supplied with such clothing and-materials. The NRC staff has advocated that fire watches are an appropriate compensatory measure to inadequate fire barriers. NIRS argues, however, that fire watch personnel. who do not have protective clothing and materials--could be placed in mortal danger upon finding a n:rmo. Lag
- fire which releases hydrogen cyanide gas, and may not even be able to sound the alarm before--
succumbing to the effects of the gas, thereby negating their effectiveness as a tire watch. Absent conclusive documentation which denies Promatec Report CTP 1099 or indicates that the . Dermo. Lag m'ike.up has changed in the intervening years, the NRC staff 'again digs its own grave and makes NIRS' case that fire watches are an inappropriate, inadequate substitute for: fire barriers, and do not meet NRC regulations. NIRS notes that simple mathematics shows us that even a one. hour fire watch can miss 59-minutes of observance, yet tests have shown Bermo. Lag can faii within 15 20 minutes. Further, because of errors in ampacity derating, the use of Thermo Lag can actually initiate a fire which, because it is hidden underr. cath layers of Thermo. Lag, might go unnoticed until damage to vital 7
electrical cables already has been done. In these cases, sounding the alarm may do little good. Fire suppres, ion personnel may respond adequately, but it will be too little, too late. Finally, the NRC staff argues that continued installation of Thermo Lag at Comanche Peak Unit.2 is being done at the applicant's own risk. We would argue that the risk is shared by most of north Texas. Given that the NRC knows that nermo. Lag -especially as repeatedly tested at Comanche Peak. includir.g renonfigurations, is inadequate, it is the height of irresponsibility to allow its continued installathn. nis can at best result in unnecessary costs to ratepayers, who will pay to remove this faulty material and for the delays in operation of Unit.2, and at worst i result in a previously identified but unaddressed risk of meltdown caused by fire. One would surely think that the NRC staff would not be so timid in a case where safety issues can. effectively be addressed even before a reactor begins operation. IV. NIRS RESPONSE TO WINSTON & STRAWN IIRIEF OF AUGUST 19.1992 ON BEllALF OF GUI.F STATES UTILITIES First, we must say we are quite surprised by this lengthy brief on behalf of Gulf States Utilities (GSU) in opposition to NIRS' petitions. Given the hourly rates of most Washington lawyers, it might well have been just as cost.cffective for GSU to remove and replace its Thermo Lag as it was to pay for this brief. His is particularly true for a utility which already has publicly acknowledged that it intends to remove and replace its Thermo Lag. As we understand it, there is no dispute between GSU and NIRS on the need to remove and replace its Thermo. Lag; the , only dispute is on the timing: NIRS seeks action now, GSU wants to wait a while. NIRS hopes to work with GSU to achieve a mutually beneficial schedule. That said, the GSU response reads like an industry wish list to avoid taking action on this critical safety issue, it is in error, both in its interpretation of the law and NRC regulations, and in its do nothing approach to a vital safety problem already identified by GSU and for which GSU already has promised a resolution. Frankly, we must wonder if anyone at GSU has read this brief. However, the GSU brief raises a number of issues to which we feel we must respond. First, the.brief states that " problems with Dermo-Lag material within the nuclear industry were o first identified and brought to the attention of the NRC by GSU...in February 1987,* In fact, as the NRC IG's report documents, problems with Thermo. Lag were identified as far back as 1982. Second. by this admission, it is apparent that GSU did not perform adequate Quality Control measures which would have determined the mis installation of Thermo. Lag at the pinnt (i.e. removal of stress skins). Further, the " informational" testing referred to in the bnef (page 5 6) consisted only of TSI. sponsored tests, which have been declared unacceptable by the MRC's Inspector General. While GSU may, at the time, have had reason to accept those test results, it has no reason now, and must r.ct quickly to resolve this issue. As the NRC Commissioners are aware, GSU has had
" temporary" fire watches in place since 1989-a clear indication of the dangers involved in -
accepting such indefinite " compensatory" measures. Do the NRC Commissioners truly intend to have fire watches replacing fire barriers at some 80 nuclear plants for years on 'end, just because utilities don't want to pay the expense of installing working fire barriers? 8 T
l . L Tests conducted for River Bend as early as June 1985 indicated that Thermo. Lag failed to function as an effective fire barrier in a test of fire penetration seals. GSU failed to issue an LER about this failed test (10 reporr. 9i.04N, August 12.1992, page 21). Although the NRC staff should have already recognized the failures of Thermo Lag, OSU's delay in submitting this LER and subsequent delays in issuing other LERs further hampered the NRC's ability to understand the true dimensions of the Thermo-Lag problem. We recognize, however, that in some respects. River Bend and GSU have been leaders in the - industry in identifying and coming to grips with this problem. Because GSU has acknowledged this, and has publicly stated its intent to remove and replace its Thermo Lag (answers to CNN questions of July 21,1992), there is no reason to fur *her delay this needed resolution. De NRC and the utility admit the problem, the utility states, on camera, that it will address the problem, and now they want more time? Forget it. Either there is a problem or there isn't. The problem: is acknowledged. Either it will be rectified or it won't. NIRS says it will. What do the NRC' . Commissioners say? In its section " Application of Legal Standard for Shutdown" .GSU cites numerous cases tc, buttress its argument that " mere" non-compliance of NRC regulations does not in itself. constitute an argument for "an extreme enforcement action such as a shutdown order or license suspension..." NIRS does not view a license suspension as an " extreme enforcement action," but rather as one of a number of NRC tools to assure compliance with regulations. NIRS has not sought a revocation of GSU's operating license nor a permanent shutdown of River Bend or any other reactor. Our concern is merely to ensure that nuclear reactors can meet NRC regulations and . can effectively handle fires which may arise. Temporary suspension of operating licenses is - hardly an " extreme" measure to attain this goal; rather it is an entirely reasonable method of ensuring regulatory compilance. Indeed, if this is " extreme" then we suspect that the vast - majority of the American public are " extremists," since there is ao indication that the public-favors the inability on the part of nuclear utilities to effectively combat fires. Moreover, this case law recited by GSU.only underscores the point that the NRC staff has
~
repeatedly bowed to industry pressure not to enforce safety regulations, even when it shou'd have (we also note that GSU admits that its Hermo Lag is " degraded.") If the NRC wishes to be the nation's promulgator and enforcer of nuclear safety regulations, then it must promulgate . - and enforce regulations.
- On page 13, GSU states that it "is incorrect" that River Bend is in violation of NRC requirements. We can think of no other case in which it has been more directly proven, both by the NRC and by the utility itself, which instituted " compensatory" measures three years ago, that-a reactor is in-direct violation of MRC requirements.
On pages 1415, GSU argues that Consumers Power Coa (1980) and Arizona Public Service (1990) provide precedent to avoid shutdown based on non-compliance with fire protection regulations. In both of these cases, however, even as described by GSU, the problems were not
.ncarl,y so gross as those experienced by River Bend. In the most relevant case. Arizona Public 9
I w -- ,--w- ,, , , - - - - - k
l
- l
&nirs, there was a failure to comply fully with quality assurance require'nents for fire protection systems, including fire barriers. This is a far cry from a demonstrated sepeated, end acknowledged failure to have an effecthe fire barrier system. We're not talking quality assurance here, we're describing a fire barrier which, after repeated testing, does not work and which endangen the public. j l
GSU's arguments that its Gre protection system continues to offer " defense in depth
- are i irrelevant and false. A key component of the " defense in depth" system for Gre protection is the i existence of fire barriers. One cannot remove that component and continue to as'ert ' defense in depth' indefinitely. But that is exactly what has happened at River Dend. For the NRC to )
approve the use of continued fire watches at River Bend, after seven years of failed tests and i i more than three years of fire watches would require a finding that fire betricts are no longer necessary. If the NRC wishes to remove the fire barrier requirement from its defense in depth ; philosophy, then it can do so. following the requisite public comment period and inevitable - federal court suits, ne NRC cannot, however, choose to se:cctively enforce its regulations by allowing non compliance on an indefinite, unlimited basis. On pages 26-30 of OSU's brief, the argument is again made that probably there won't be a fire, if there is one it probably won't be too bad, and the cables probably will be able to withstand it. His makes a mockery of t e h NRC's fire protection regulations. Fires can and do occur, and fire ; barriers exist to help protect us from the damages associated with fires. To srgue that fires rre unlikely, won't spread, or can easily be suppressed is to argue against NRC regulations in the f;rst place. GSU also uses probabilistic risk assessment to argue that it is unlikely that a fire willlead to core damage. To quote again from 11 NRC 707 (1980), "Our April 13 decision in no way
' permits reliance on probabilistic calculations to enter regulatory policy 'through the back door.'
He NRC has identified fire as a major contributor to core meltdown (NUREG 1150) and has adopted regulations designed to minimize this risk. GSU is again making an impermissible assault on NRC regulations. one which, in the reverse, intervenors would never be allowed to get away with. OSU makes the absurd argument (page 33) tt at fire barriers need not be non co.ibustible. In fact,10 CFR 50 Apixn, dix A and Appendix R both refer specifically to a requirement for non-combustible materials for fire barriers. The combustibility of Bermo. Lag cannot be in question. The NRC's own color photographs of nermo Lag on fire make for a stunning sisual
. confirmation of this fact. But to quote again from Promatec report CTP 1099.
- Sample 1099.3
[nermo-Lag) began to smoke almost immediately. At about 6:30 (six minutes and thirty seconds into the test), the sample spontaneously ignited and periodically evolved large amountr of flame for at least tive minutes, accatapanied by large amounts of black smoke. Post test, the sample -
- remairn were a black crusted char." ne second test referred to in CTP 1099 was described this way, "This sample began to give off a white smoke almost immediately. At approximately 1:30, the top of the sample began to bubble and started to char. Flaming combustion spontaneously !
occurred at approximately 4:30 and continued till at least 9fA" He NRC cannot possibly accept GSU's argument here. 10
k k We also note that GSU relles on " vendor information" to argue that Hermo lag has a Dame spread rate of 5. In fact, the IG's report casts severe doubt upon any vendor information ; supplied for nermo Lag. We know of no independent tests which would verify such a llame g m ad rate. ; GSU's assertion that evah.ation of a seismic event is not required by Appendix R 'in conjunction with fire" is narrowly correct. However,10 CFR 50 Part 100 does call for evaluation ! cf seismic qualiflention of plant material. In addition Branch Technical Position 9.5 3 refers to a ; required confirmation that fire protection systems components, piping, and structures are. , designed in accordance with applicable seismic design criteria. NIRS alleges that in a seismic . > event Dermo Lag could cause a shearing of cables, even without a fire; thus, hermo Lag does not meet these regulations. GSU relies only upon a discredited TSI test to address seismic ! issues, nermo Lag is not seismically quallfled by any independent tests. GSU argues (page 36 37) that NIRS mischaracterizes certain tests of Hermo Lag conducted by Texas Utilities. We do not understand the relevance of these tests to River Bend, where tests - have admittedly and conclusively demonstrated that River Bend's Hermo Lag must be removed and replacco. As we have noted above. there is no fundamental disagreement between NIRS ' ' and GSU over the inadequacy of Th:rmo Lag. Our only disagreement is over the speed in which it must be replaced. However, we note that further testing by Texas Utilities has resulted in further failures of Thermo.L.ng to act as an effective fire barrier. GSU argues (page 37 39) that failure of Thermo Lag to meet hose stream tests need not worry the NRC GSU is wrong. First. ASTM 119 was developed not to test electrical cables, but to - '! test fire barriers. Second, it is reasonable to n:sume that if any fire barrier fails hose stream tests that damage to electrical cables could ensue. Third, OSU asserts, electrical cabling at RBS ' [ River Bend Station] is dedgned 'to a!!aw wetting down with fire suppression water without-electrical faulting.' NIRS would hope that all utilities use cabling which can be wetted down ' without electrical failure. Obviously, in the case of an inadvertent sprinkler trip, cables should - not short out the reactor. But the concern is over cables which suffer degradation through early aging (due to ampacity derating errors) and early fire damage. These may well short-out with - the inttoduction of suppression water. , The purpose of the hose stream tests are to ensure that the barrier doesn't crumble, thereby offering new oxygen sources for ignition and allowing further smoldering of cables for future ignition. The literal interpretation of the hose stream test is so the product doesn't Oy off the cable trays and damage the remaining products it is protecting. In either case, the hose stream V test must certainly be of concern since it is a requirement in NRC regulations (NFPA 251 and ASnt E 119). Finally, CSU argues (page 39-41) that ampacity derating should not be a concern because it =
"has relied on the ampacity derating factors provided by TSI in its calculations for RBS installations." nis is the same TSI which has made repeated. Dagrant, errors in ampacity derating over the years, as documented in the NRC's IG report. GSU says that deterioration of ' !
cables " proceeds very slowly and is generally considered insignificant...if correct amnncity deratine . values are annlied (emphasis added), Given the IG's report: there is absolutely no reason to believe that correct ampacity derating values have been applied at River Bend. or anywhere 11 '
E cise. Indeed, one of the early selling points of Thermo Lag was its low ampacity derating valuest only recently have utilities learned that these low values were illusory and that, in fact, Hermo. Lag has comparable vnpacity derating values as other fire barrier materials. Unfortunately, utilities have learned this a[Lc their installation of Thermo. Lag. Simply because GSU 'cannot identify a single case in which ignition of cables protected in a Thermo Lag enclosure has-- _ occurred at a nuclear power plant," does not mean that such an ignition cannot occur, and does not even mean that such an occurrence is not likely. At this point, the burden is not upon NIRS to prove that Thermo Lag is inadequate as a fire barrier; the NRC already has acknowled;cd that fact. Rather, the burden is upon each-indhidual utility to prove that its particular configuration of Dermo La, is effective as a fire ' barrier. GSU has not succeeded in meeting that burden, and no other utility has even tried. Despite NRC staff protestations, the safety significance of Thermo Lag's failure is hight there is, in fact, no assurance that the reactors listed in this petition can withstand a significant fire without severe damage to their electrical cables. Such damage would virtually assure a nuclear meltdown. Given that the NRC, in NUREG.1150, already has identified fires as a significant contributor to core melt, and has stated that each reactor will experience three or four significant fires in its operating lifetime, this is somewhat more than a superfluous, casily dismissed issue. In fact, the Thermo Lag issue goes to the heart of the NRC's commitment to safety. Will the NRC enforce its regulations, or will it rule its regulations are "on the books," not to followed in real life. Will the NRC stand up for its safety studies, or will it let the utilities pick away at them whenever it's convenient? Will the NRC allow cost in this case, not even very sigcificant cost for many utilities-to override its safety first mandate?
- y. DESCRIPTION OF PINITIONERS Petitioners are as described in NIRS petition of July 21,1992 and addenda of August 12, 1992.
VI HELIEF REOUESTED NIRS asks that the NRC Commissioners overturn the NRC staff's decision of August 19,1992 and order the immediate suspension of the operating licenses of River Bend, Shearon Harris, . Fermi.2, WNP 2, Brunswick.1 and -2, and Comanche Peak 1. In-addition,' NIRS asks that a stop work order, or~if necessary under the provisions of 10 CFR 2.206, a suspension of the . construction permit, be issued for Comanche peak Unit 2. NIRS asks that these orders be in place until a tested and effective fire barrier,in accordance with 10 CFR 50 Appendix A and Appendix R, is installed. p 12 >
NIRS further asks that the Comminioners order knmediate distribution of Genetic letter 92 xx (February 11,1992). At this point. however, it would make more sense to order utilitics to remove and replace their Thermo Lag, unless utilities can prove through full: scale independent testing of specific configurations that the Hermo Lag as installed meets regulatory requirements. NIRS understands the NRC staff is considering applying testing requirements to all fine barrier materials. We agree with this approach. Respectfully subtr r.ted. Michael Mariotte September 3,1992 13 l
., r t.
s. MM
.h .
. PROGRLSSIVE MA11 RIALS AND TECHNOLOGILS. INC.
ji G fGONMBENF/A&f j a A u ,, s n u m > l H ')-
/ t. W.
FINAL REPORT CTP 1099
, y . 'j COMBUSTION TOXICITY EVALUATION ,,. OF i 'MT' BARRIER SYSTEMS JULY 14,1986 ,
l ,
~
i Referencing: . SwCS Report No. 860509
-- May1986
. and-SwRI Report No. 01 - 8818- 101 L..;
May1986 l k P.N e 6 TRRv4 772d@ e qg g@.9240 ' _2_ Ms
e6 4 , l' . FINAL REPORT CTP 1099 4 TABLE OF CCNTENTS i 4
- 1.
SUMMARY
!!. PROMATEC CTD 10 9 9 . . . . . . . . . . . . . . . . . . . . . ( 4 P a g e s )
'i !!!. SwCS REPORT.NO. 860509 -
Introduction . . . . . . . . . . . . . .-. . . . . . . Page 1 . Objective .. . . . . . . ... . . . . . . . . . . . . Page 1 . Procedure- . . . . . . . . . . . . . . . . . . . . . . Page 1. 2 i Test Results . . . . . . . . . . . . . . . . . . . . . Page'2, 5, 68 Table B . . . . . . . . . . . . . . . . . . . . . Page 3
-.i - . . Page 4 Figure 1 . , . . . . . . . . . . . . . . .
Table C . . . . . . . . . . . . . . . . . . . . . Page 4 Ot scus s ion . . . . . . . . . . . . . . . . . . . . . . Page 5 i
; Conclusions . . . . . . . . . . . . . . . . . . . . . Page 6 References . . .. , . . . . . . . . . . . . . . . ... . Page 7
*- Appendix A - Extrapolation Frca '
Animals to Humans
]
1 IV. swr! REPORT NO. 01-881S-101 E 'l - Materials (1099.1, 1099.2, 109.3) . . . . . . . . . . Page 1 1 Combustion Procedure . . . . . . . . . . . . . . . . . PageL1 e Combustion Atmosphere Analyses . . . . . . . . . . . . Page 1l
. ll
~
4
. met a
1
- l -'
tl , - , , - , - . - - .
m .. i to t1 Results - i Combustion Sumary . , . . . . . . . . . . . . . Page 2 i Analytical Sumary . . . . . . . . . . . . . .f. Page 2, 3-l Figure la - Animal Exposure Chamber and Radiant Furnsee . Il figure 1b'- Combustion Cell / Radiant i Radiant Furnace T Figure 2 - Carbon Monoxide Concentration -
) Curves ., Table 1 - Analytical Sumary ,
V. SwRI REPORT NO. 01-8818-10lb Materials (Prom 4) . . . . . . . . . . . . . . . .. . . Page 1-i i Combustion Procedure . . . . . . . . . . . . . . . . . Page l' a . j Combustion Atmosphere Analysis . . . . . . . ._ . . . . Page 1, 2 *
,, Results ,
8 Combustion Sumary . . . . . . . . . . . . . . .-Page 2 i Analytical . Sumary .:. . . . . . . . . . ., . ....Page 2 i' Figure la - Animal Exposure Chamber *
.p and Radiant Furnace I Figure ib'- Combustion Cell / Radiant Heater Assembly Fig;re 2 - Carbon Monoxide ~
Concentration Curve
. Table 1-A - Analytical Sumary 1
.l.
- l.-
m a 2
-{-
-# o.
e, . Ng ' PROGRLssfVE MATIRIALs AND TECHNOLOClLs. INC. July 14, 1986
SUMMARY
COMBUSTION ANALYTICAL TESTING OF CABLE WEAR MATERIALS CTP 1099 Refer to the attached third party consultants, SwCS Report No. 860509 entitled COMBUSTION T0XICITY EVALUATION OF M.T. BARRIER 4 and to the referenced Southwest Research Institute Final-' Report
. No . 01 -3818-101 f or specific test results,
. These series of tests were performed by a third party testing organization, Southwest Research Institute (SwRI), and results uere evaluated by an independent third party consultant, Southwest Certification Services (SwCS).under programs we believe to be reliable.
] PROMATEC
?
' "/ ,
L. Charlas Sp g Technical Services ' nager
~
P.O. 80X 4672
- HOUSTON. TEXAS 77210 * (713) 690 524n
s-
~
PROC,RL$$M. MATIA1ALS AND TECHNOLOGLS. INC. f PROCEDURE FOR: PROCEDURE NUMBER: COMBUSTION ANALYTICAL TESTlHG OF CABLE WRAP i CTP 1099 MATERIALS ~~ PROCEDURE ISSUE SUMMAR_y,
- ssut/ontt mmg cemmo canst $ .
'\. A ISSUE L . C. f 3S ( IO< *b RAIDY BROWN bo . ISSUE FOR CONSTRUCTION' ". 11/18/85 3 t 7. l t rer- r 3 0 06/06/04-
i s s. 1510t: A- , % m: CTP 1099
""*""" rc: 2 or 4 11/1B/85 .
ClP 1099 COMBUST 10N ANAL YTICAL TESTING Of CA8LE WRAP MAlERI ALS
, 1.0 PURPOSE
$ The . purpose of this sries of testing ja to determine if cable wrap
~
materials exhibit any toxic off gassing attributes that would provide a hazardous environment for the operating plant personnel and fire fighting personnel in the event of a fire. g This test chall be performed by a qualified third party testing g organization under their guidelines. This test is a preliminary test'and O drastic results may necessitate the performance nf an animal exposure 5 test to absolutely determine lethal levels. E c 2.0 SCOPE i This procedure lists the gasses to be analyzed and qualifies the J preparation of test samples.
3.0 REFERENCES
$ 3.1 swr 1 Proposal No. 01-8818-P103, dated 10/18/85 12 3 c.0 DEFINITIONS
?.
, g None
! =
5.0 RESPONSIBILITY 7 $ 5.1 The TECHNICAL SERVICES DEPARTMENT shall be responsible fort' l w determining the test objectives; preparation of the test specimens; Q
~
coordinating test performance with the testing organization; and to
- incorporate the third party test report and resultant-data into a
- final Test Report for appropriate distribution.
5.2 The QUAL.ITY DEPARTHENT shall be responsible for verifying that the
'} testing organization is knowledgeable and capable of performing and documenting the various test results.
The OUALITY DEPARTHENT is also responsible for the traceability of
.; test specimens.
5.3 The THIRD PARTY TESTING ORGANIZATION shall be responsible for determining the test methods used to satisfy the' test objectives;.
~
perforniance of appropriate tests; documentation of the results and-the issuance of an appropriate letter report.
?
O 155v0: A QM to: CTP 1099 11/18/85 ~ "~'"* " * " " * " " rW: 3 of 4 6.0 TEST REQUIREHENTS 6.1 Causes to be analyzed fort a) Carbon Monoxide b) Carbon Dioxide
, ., with a trapping system to simultaneously capture for later analyses oft
?
~~
a) fluoride ion (FI) b) Chloride ion (Cl) c) Cyanide (HCN) d) Ammonium ion (NH)) { 6.2 Samples a) 1099.1 - an approximate 31/2 x 61/2 inch camples of the HEMYC E One Hour Wrap blanket - C y NOTL: This one hour blanket is the same material as the 3
$ PROMAl[C three hour blanket when combined with the filled
. 5 powder assembly (Sample 1099.2).
[ g b) 1099.2 - an approximate 31/2 x 61/2 inch sample of the filled powder assembly. y c) 1099.3 - an approximate 3 1/2 x 6 1/2 x 1 1/2 inch specimen of
; 151 Thermolag as provided by jobsite.
=
5 6.3 Testing c h ihree separate test shall be performed as llsted belows o
-- d 6.3.1 Test 1 i W
- M C .Analyseo per SwRI proposal No. 01-8818-P103 of 1099.1 specimen of the HEMYC One Hour Blanket.
6.3.2 Test 2 Analyses per SwRI proposal No. 01-8818-P103 of 1099.2
- specimen of the filled powder assembly from the PROMATEC Three Hour Wrap system.
1 NOTE: The results of Test 1 and Test 2 shall be combined to determine the off gassing attributes of the PROMATEC Three
. Hour System.
6.3.3 Test 3 Analyses per SwRI proposal No. 01-8818-P103 of 1099.3
.- specimen of TSI Thermolag material.
5 _ _ . _ _ _ __. i__ '-
l i-1 55vt: A m CTP 1099
# ' ' " " " " " * * * " ^ ' * *
~
11/18/05 rm: a et s 6,4 flNAL REPORT 6.4.1 ' 1he testing organization shall prepare a Letter Report outlining test performance and results thereof as soon as practical. 6.4.2 The TECHNICAL SERVICES DEPAR1HENT shall incorporate the 7 *- third party information, cummation and other pertinent Jata into o final lent Report for distribution. s O E I 5 C O
.E.'
6 3 b E 5.
~
i s a 7 .8 1 w 9 M l _ S*M\.L________________m_____ _ _ _ _ _ _ _ . _ _ _ _ . _ _ _ . _ . . _ _ _ _ _ . _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . . _ _ _ _ . . _ _ _ . _ _ _ _
.i.
' Southw CerflAcatkwt Serveres, hic 6ao Asmo o .n, u,,r San Antune. Tuu 18236 '
($121 M1 $2L1 A COMBUSTION TOXICITY - EVALUATION of I , M. T. BARRIER SYSTEMS SwCS Report No. 860509 , May,1986 i . Prepared for PROMATEC, Inc.
-Houston, Texas i
~ . . . . - -; ,_. - .,;,_,,.,;_._a..:._-.....___ .. . . . _ -.
l
.y
-j
'5t t
y ABSTMCT , i 1
*4 .;y A series of combustion products toxicity tests were done utilizing; ;
V~ chemical analytical methods in lieu of animal exposure tests, In which PROMATEC, Inc.'s one and three hour M.' T.-Barrier Systems'for protecting Class'1E Electrical Cables were examined and compared toi another commonly used cable. protection matenal. The M. T. Barner-matenals_ were found to be of very low toxicity, while the other matenals-' i-
- .3 :
were shown to release combustion products of a much higher toxicity. _-. b ' .
' . i
' [; _ 1; a
1 , 73
, : . 'f4s
,y._- -
. <i
- I y ;e s' f-s f : .l 1
s R m. _ . < m
4 INTRODUCTION . With the increasing awareness of fire safety and fire protection engineering has come a growing concern for the amount of smoke released by a burning material and the 7 toxicity of that smoke. While the quantity of smoke given off during a fire is an I important parameter of a material, the toxicity of that smoke is equally important.
~
There is a movement 1 by building code agencies to require that building materials of all types be subjected to a test which will indicate or rank the toxicities of their combustion products. Of the significant test methods proposed to date, the majority utilize some method of exposing rodents (rats or mice) to the smoke, and determlning the amount of smoke it takes to kill the animalin a given period of time (LC50)- While animal exposure tests may indeed prove necessary, a great amount of rodent toxicity exposure testing has been done for which there is good concurrent gas analysis data for the major toxicants present. Gases such as carbon monoxide, i carbon dioxide, oxygen, hydrogen cyanide, hydrogen chloride, and formaldehyde have been studied in some depth in'this regard, and the data is available to the interested reviewer 2.3A.S.6.7, I OBJECTIVE in an effort to determine whether or not a toxicity problem exists with the combustion products of their M. T. Barrier Wrap system for power station cable trays and conduits. PROMATEC, Inc. embarked on a test program which would, if possible, exclude the use of animals, and still assess the relative potential for release of toxicologically significant quantities of selected combustion gases produced from the thermal. 1 decomposition of different cable wrap materials. PROCEDURE The test method chosen was a National Bureau of Standards laboratory test apparatus modified to utilize a horizontally-configured quartz radiant furnace to induce combustion of the test specimen. This test apparatus, which is currently tunder consideration for use in a proposed ASTM toxicity test procedure, was utilized to generate the smoke for this project. A complete description and drawings of the test
.- apparatus used may be found in the final reports for SwRI Projects 018818101a a b.-
A comparison of four materials was done at the Southwest Research Institute toxicity l , laboratory, based on the response of 6-1/2' x 31/2" x 1-1/2" thick test specimens i exposed to a heat flux of 8.0 Watts /cm2. The performance of the test materials was 1 measured in terms of their release of five specific toxic gases: Carbon Monoxide (CO); l
O* . SwCS Report No. 860509 PROMATEC, INC. . May,1986 Page 2 Hydrogen Cyanide (HCN); Hydrogen Chloride (HCI)1 Hydrogen Fluoride (HF)1 and. Ammonia (NH3). SPECIMEN MATERIAL
- 1 1 br M. T. Barrier System 2 3 hr M. T. Barrier System (outer wrap) 3 Prefabricated, Subliming Muterial .
4 Trowel Applied, Subilming Material Overali, four experiments were done, in which each of the materials was exposed to the radiant heat flux and the resultant combustion 3roducts were monitored for the quantity of each of the above gases, with no anima s exposed. Exact details of the analytical methods used for these analyses may be found in the pmviously referenced SwRI test reports. Specimen 1 consisted of the PROMATEC, Inc. M. T. Barrier System one hour cable protection wrap. Specimen 2 was the outer blanket assembly for the
. three hour M. T. Barrier System, which consists typically of Specimen 1 covered by .
Specimen 2. It was felt more appmpriate to evaluate the two layers of the three-hour system separately. Specimens 3 and 4 consisted of a white, trowel applied material.
, which is in common use fc.- the fire protection of cable systems, which performs by subliming at elevated tempctures and thus keeping heat away from the protected systems.
TEST RESULTS Tables and discussions of the behaviour of each of the test specimens may be found in the appropriate SwRI test reports. This analysis w!il concentrate on the products of combustion from each, and their relative significances. I The generally accepted method of determining how much smoke is evolved by a test-specimen in the field of combustion toxicity testing, is to measure the weight loss of the 7 sample and calculate the. number of milligrams of sample loss (smoke) that has been emitted into a known volume, in liters. Table B contains information on the percent of weight loss, the smoke released, and the maximum temperature reached in the smoke chamber. Under normal conditions, there would be rats in the chamber as well as gas analysis equipment. These tests simply omitted the animals. As can be seen in Table B, specimen 1 (M, T. Barrier,1 br wrap) had a very low - volatility when exposed to high heat flux levels. This is as would be expected, since the materialis mostly a mineral fiber blanket with a mineral fabric covering. The only gas evolving materials present are sizing or finishing chemicals applied at the point of manufacture to facilitate the process of weaving the mineral fabric. l l
1 SwCS Report No. 800509 PROMATEC, INC. May,1986 Page 3 Speamen 2 (M. T. Barrier, 3 hr wrap, outer bianket) consists of similar materials at specimen 1, with the addition of a layer of an inorganic hydrate. The hydrate, when heated above Hs activation point, evolves steam endothermically and thus could be expected to emit a greater amount of gases when tested than did specimen 1. 7 Specimens 3 and 4 also react to the application of heat by emitting gases in an .< endothermic reaction, although the nature of the gases released is not clear. These specimens both lost approximately the same percent of weight and resutted in the production of much greater amounts of smoke than did the M. T. Barrier. specimens. Smoke Maxirrum PERCENT Concentration Temperature SPECIMEN W T l. O S S Imau *F 1 5 17 131 2 28 271 138
?
3 78 2151 150
, 4 78 2359 162 i
TABLE B The gas analysis data presented in Table C is expressed in terms of total exposure potential. Consider the fact that a very short exposure to a high concentration of a
', toxic gas may be no more or less toxic than a long exposure to a low concentration of the same gas. Consequently,it is common to express gas exposures in terms of the integration of the concentration versus time plot. A simple example would be that a 1 ] minute exposure to a concentration of 1000 parts per million (ppm) would be considered equal to a 2 minute exposure to 500 ppm or 4 minute exposure to 250 ppm. Each would give a concentration x time (C t) product of 1000, and, for purposes of these tests, would be considered equal. The C t product, then,is very simple to calculate for a constant concentration exposure. For an exposure during which the concentration is constantly changing (increasing or decreasing) however, the C t product is not as easily derived. The integration is usually accomplished by computer, using a technique of dividing the curve into very small pieces, typically one or two seconds long (see example in Figure 1, below) and assuming that each piece is rectangular. By adding up all of the ' pieces," the computer is able to very accurately , calculate the area under the curve, which is mathematically equal to the C.! product.
t
- - .,...~.-,n , e
C *. SwCS Report Ns. 860509 PROMATEC, INC. May,1986 Page 4 l h 7 5 / ' E
/
/
5 o # z - O . c. TIME Floure 1 This entire process is performed in real time during the expe.iment, since the computer is monitoring the gas analysis instruments anyway. The C*t product is utilized, e because it is similar to_ the way a mammal's body absorbs toxicants as it breathes. During the relatively short period of a toxic exposure, the body does not have a chance io rid itself of the material, and it builds up as a function of the concentration of toxicant being breathed and the amount-of time il Was breathed. ' 4 r Realizing now, that the C't product (usually expressed in units of ppm minutes) of any gas given off and monitored during this test is really a measure of the toxic effect it may - have on mammals, one can get a better feel for the sigrtificance of the test results. co C t HCN Ct hcl C.t .- HF C4 NH3 C t ' SPECIMEN (Dom min) (com-min) (opm-min) (pom-min) (opm-min) 1 27,966 280- 610 1,770 0
- 2 32,490 280 t .390 1,770 0
+352,500 85,289- -5,848
- 3,017 0~
4 2,077,920 30,910 3,950 '1,27C' l 10,800 TABLE C Of the- gases rnonitored, carbon monlxide and hydrogen . cyanide -are.the most commonly involved in fire gas poisonings. It is strongty evident from Table lC that specimens 3 and:4 produced much higher levels of CO and HCN than the others. Even if the CO and HCN levels for specimens 1 and 2 are summed to simulate test results _which might be expected for an assembled M. T. Barrier system, the levels
.t'
SwCS Report No. 860509 PROMATEC, INC. May,1986 Page 5 recorded for specimens 3 and 4 were many times greater. (As much as 34 times as much CO and 152 times as much HCN.) Higher levels of hcl and HF were also recorded, but the differences are insignificant in view ci the extremely high CO and HCN values observed for the latter two test specimens. Tiie ammonia resutts were not shriticant for any of the specimens. [ In the anarysis of these data in terms of expected toxic (lethal) potency, a review 8,9 of published LC50 values for the gases of interest is necessary. (LC 50 is the concentrat;on of a gas sufficient to kill 50% of the animals exposed under a given set of Mst conditions.) Alarie and co workers at the University of Pittsburgh8 have , reported an extensive array of LC50 values for a wide variety of materials and gases using rodents (rats and mice) as. test subjects. Atarie has reported LC50 values of 3500 ppm for CO and 166 ppm for HCN for 30 minute exposures of mica. These values tru state into C t products of 105,000 ppm minutes for CO and 4,980 ppm-minutes for HCN. A glance at Table C reveals that the two M. T. Barrier samples (specimens 1 and 2) are well below both of the te values, while specimens 3 and 4 are
, dramatically above them.
u On the basis of those gas analysis tests a!one, lethality for eri.med rodents would not
. be exper:ted for specimens 1 and 2 (M. T. Barrist system) and is definitely indicated for specimens 3 and 4. The hcl and .HF levels recorded for all four specimens were well below the LC 50 C t products reported in the litemture for rodents 810 of from 100,000 -
300,000 ppm-minutes e r hcl and approximately 5,000 - 30,000 ppm minutes for HF. DISCUSSQ Caution must ca exorciseti in the interprotation of these data; since at least two factors enter into the extre;otation of this information to real fire conditions with people exposed. First, animals were not exposed to the combustion atmospheros, so we only have information based on literature values for the gases measured. Second, animals (rodents) were used for the development of the published LC50 values, and only approximations can be made in predicting the human response to similar exposures, it is know however, that CO and HCN are absorbed in rodents in the same manner as - in humans. The difference between rodents and humans is that these gases are. absorbad at a faster rate by rodents and therefore act more quickly in rodents than in humanc. A fairty extensive discussion of the relative responses of rodents and man may be found in Appendix A, Extrapolation From Animals to Humans. Drawing conclusior_ between rodents and humans for the toxic effccts of corrosive and irritating gases such as hcl and HF is much more complicated, since rodents breath only througn the nose and humans. ara more likely to breath through the mouth in a fire environment (much less painful). Breathing through the nose results in the: ! officiant cerubbing o' the hydroscopic gases hcl and HF, and thus dramatically
.s SwCS Report No. B60509 PROMATEC, INC.
May,1986 Page 6 } reduces the toxic effect.11,12,13 There is wide-rangi.ng opinion on the correlation cr-lack of correlation of irritant exposures of rodents to humans, and consequently the subject is beyond the range of this report.
] CONCLUSIONS
,_ lt is a certainty that animals,if employed during the tests reponed here, would have survived only briefly in the atmospheres produced by the thermal decomposition of specimens 3 and 4. The two M. T. Barrier System specimens.(1 and 2) however, show all indications of being relatively non-toxic. With the increasing' awareness of.the potential for release of toxic gases during a fire, comes a real concem for the-performance of a material during such a scenario. The current series of tests has-
,, indicated that the M. T. Barrier System does not present an unusual danger from fire c generated smokt. In fact, by comparison to other commonly utilized materials, the h;, toxicity of the products of combustion from the M. T. Barrier System appears to be
. di, extremely low and, for all practical purposes, insignificant.
1
; s Y
e
o b SwCS Report No. 860509 PROMATEC, INC. May,1986 Page 7 REFERENCES i Shatter, G. S.1984. Fire gas toxicity. Recommendation'; of the Secretary of State to the Uniform Fire Prevention and Building Code Council. Albany, NY: State of New York, Dept. State.12231. 2 Cul;is, C. F., Hirschler, M. M.1981. The Combustion of Organic Polymers. Oxford: Clarendon. CO pp.' 3 Landrock, A. H.1983.- Handbook of Plastics Flammability and Combustion Toxicology. Principles, Materials, Testing, Safety and Smoke Inhalation Effects. Park Ridge, NJ: Noyeb, 308 pp. 4 Kaplan, H. L., Grand, A. F., Hanzell, G. E.1983. Combustion Toxicology, Principles and Test Methods. Lancaster, PA: Technomic.174 pp. 5 Anderson, R. C., Daring, K. M., Long, M.1983. Study to assess the feasibility of - incorporating combustion toxicity requirements into building material and fumishing codes of New York State. FinalRep. Dep. State, Off. Fire Prevention and. Control, Albany, NY. Cambridge, MA: Arthur D, Uttle, Ref. 88712. 6 Hartzell, G. E., Priest, D. N., Switzer, W. G.1985. "Modeling of Toxicological Effects of Fire Gases: 11. Mathematical Modeling of intoxication of Rats by Carbon Monoxide and Hydrogen Cyanide," Joumalo/ Fire Sciences, Vol.3, March / April 1985. 7 Hartzell, G. E., Stacy, H. W., Switzer, W. G., Priest, D. N.1985.'Modeling of Toxicological Effects of Fire Gases:IV. intoxication of Rats by Carbon Monoxide in the Presence of an irritant," Joumalof Fire Sciences, Vol.3, July /. August 1985. 8 Alarie, Y.1985 .The Toxicity of Smoke From Polymeric Materials During Thermal Decomposition, Ann. Rev. Pharmacol. Toxicol. 25:325-47. 9 Matijak Schaper,.M., Alarie, Y.1982. Toxicity of carbon monoxide, hydrogen. cyanide and low oxygen. J. Combust. Toxicol. 9:21-61. 10 DiPasquale, L. C., Davis, H. V.,"The acute toxicity of brief expo urc: te hydrogen _ fluoride, hydrogen chloride, nitrogen dioxide, and hydrogen cyanide singly and in combination with carbon monoxide," AMRL-TR-71-120 Paper No. 20. 11 Anderson, R.C., Alarie, Y.1980. Acute lethal effects of polyvinylchloride thermal decomposition products in normal and cannulated mice.' Washington, DC: Soc. Toxicol. Meet., p. A 2 (Abstr.) 12 Morris, J. B., Smith, F.A.1980. Regional deposition and absorption of inhaled j hydrogen fluoride. Washington, DC: Soc. Toxicol. Meet., p. A-4 (Abstr.)
. 13 Monis, J.B., Smith, F.A.1982. Regional deposition and absorption of inhaled hydrogen fluoride in the rat. Toxicol. Appl. Pharmacol. 62:81'89.
( - l
v . M. 1 I
.4.
APPENDIX A' EXTRAPOLATION.
. . . from ANIMALS LTO HUMANS
~
4 0 t 4 5 l-0 4 ' 9 d 9 _\ e: 3 4 l+- 1 /
t The following section, quoted from a paper written by a recognized expert in this area (Alarie, Y.1985. The Toxicity of Smoke From Polymeric Materials During Thermal Decomposition. Ann. Rev.
~
Pharmacol. Toxicol. 25:325 47) will show that a relationship between animal and human responses to toxic gases exists. Due to the lack of accurate dose response data on humanc, this relationship is more of a qualitative one than quantitative, and must be considered in this light.
' Extrapolation from Animals to Humans Two factors must be considered in extrapolating test data to real conditions.' The Brst one is that laboratory animals, mice or tats, are used for testing. The second one is that the principal toxicants in the decomposition products of various materials are not always the same. These factors greatly complicate our task in extrapolating toxicity test results to humans.
When wood bums, carbon monoxide is the main toxicant,1M and carbon monoxide is the main toxicant in otherpolymeric materials as well. The relative toxicity ranking established for carbon monoxide in kboratory animals can Justifiably be extrapolated to humans because the mechanism of toxicity in both is the same. This gas is absorbed in mice or mts in the same manner as it is in humans; the only difference is that it is absorbed faster, and thus acts more quickly, in e:nimals due to their higher minute ventilatiarybody weight.33 The same is true for hydrogen cyanide. However, it material A, in which the principle toxicant is hydrogen cyanide, is found to be ten times more toxic in mice than mate'ial B, whose principal toxicant is carbon monoxide, can we than extrapolate that material A is also ten times more toxic than material B ll humans are exposed? From the data on the relative acuts toxicity.cf hydrogen cyanide and carbon monoxide in mice 12 and the best approximations of theirlethal levels for humans,34 such an extlapolation is justified, since the potency rat:0 between CO and HCN is similar for mice and humans.' 12 Matijak-Schaper, M., Alarie, Y.1982. Toxicity of carbon monoxico, hydrogen cyanide a_nd low oxygen. J. Combust. Toxicol. 9:21-61. 17 Levin, B. C., Fowell, A. J., Birky, M. M., Paabo, M., Stolte A., Malek, D. 1982. Further Development of a Test Method for the Assessment of the Acute Inhalation Toxicity of Combustion Products. Washington, DC: US Dept. Commerce, Natl. Bur. Stand., Rep. NBSIR 82-2532. 20 Alarie, Y., Anderson, R. C.1979. Toxicologic and acute lethal hazard evaluation of-thermal decomposition products of synthetic and natural polymers. Toxicol. Appl.
.Pharmacol. 51:341-62.
33 Haldane, J.1895. The action of carbonic oxide on man. J.. Physiol. 18:430-62. 34 Henderson, Y.. Haggard, H. W.1943. Noxious Gases and the Principles of Respiration Influencing Their Action. New York: Reinhold. 294 pp. 2nd ed.
G %- .Q . , v ,
.o ; ., ,
.- -S10 U T HEW E S TL R E S E A R C H I N:S TI T U T E POST OFF8CE OR AWim 28810.1 8220 CULT s R A MO AD -- E AN ANTONaO, TE R AS 78384: , int 363a4 6911
' O(PAstutNt os 4 .
, it(t
- 7677 -
8'ai tt gnmeotoGy 4 twa ploati g ', CopeuSTION ANALYTICAL TESTING OF 'THREE CABLE WRAP MTERIALSe. CTP 1099 f t by
- Walter G. Switzer, M.S. '
J FINAL RE' PORT; i SwRI Project No. 01-6818-101 :
= .
~'
Prepared for:
?
PROMATEC I P. O. BOX 4672 HOUSTON, TEXAS '77210 MAY 1986' ,
-i
% - . n.u < . . - : -
au. -.-,u..-.
.= u. - ., = w - u. ,w.. - w..-
-- ._. u - , a ,.u . - w %..-.
.m - - .,.ru . ,u.,u..c-w.. m -*
ru. u . Approved by: . Reviewed by:
~, :
f' M r.e + - s
, 6_.-
Dr. Harold L. Kaplan, Manager Dr. Gordon E. Hadozell,1Di/ector. Applied Environmental Toxicology Department' of- Fire Technology 7 ;
-for -
Dr. R'obert E. ' Lyle, ~ Vice:Presidenti Chemistry and Chemical Engineering' m n a se s , mi .e me no , s. e, se m.a. m e.ar., se er e- e.i a.r rue m e w;}- - - m., -,_ w . m. w ., .s me w enen me e . w ,. l t ' is
. A N Aeeescoe wet. N T O N.I O. e : Tvote E X. .A. St s' s a s ,- a s e .w a s . ..e v e . o g s ; ! ;- ,
,, ,.-, .~ .. -
a .. - PROMATEC CTP 1099, SwRI Project No. 01-8818-101 COPEUSTION AMALYTICAL TESTING OF CABLE WRAP MT1'tiAl.S-MATERIALS Three materials - were ?eceived, f rom Promatec for combustion analytical testing in the Radiant Panel System. The materials designations and descripe-tions are as follows: . A. 1099.1 "Hemyc" One Hour Wrap: Bronze colored cloth over glass blanket. B. 1099.2 Alumina Trihydrate / Fiberglass Assembly from the Procatec
*hree-Hour Wrap: Silver cloth over glass blanret.
C. 1099.3 TSI Thermolag: Of f-wnite solid -with metal grid. The samples (one for each assemolage) were made to approximately_3-1/2 by
.. 6-1/2 incnes and did not exceed 1-1/2 inches in thickness.
; COMBUSTION PRCCEDURE Prior to i ni'. i ati on of comoustion, the analyzers and recor:ers were cnecked to determine that the systems were operating ' correctly. The sample assemelage was weignec and placea in a weigneo aluminum .-foil boat... This comoination was then placed on the floor of the combustion cell anc the cell 7 moved' into position on the Radiant P'anel System under the 200-liter uchamber '
(Figure 1). The experiment was initiated with the application of heat flux t_o the sample. - The samples were subjected to a heat flux from 7.4 to 8.4 W/c-for the 30-minute duration'of the' test. COMBUSTION ATMOSPHERE ANALYSES' Analyses of the combust' ion atmospheres were made continuously for 02, CD, and CO2 with a closed-loop sampling system at a sampling rate of 500 cc/ min and the ' following instrumentation: Ber.kman OM-11 ana'yzer (0 2);.:Beckman: 865 Infrared Analyzers (C0 and CO2 ). Prior to the experime nts, each analyzer was calibrated with aporopriate calibration gases (C0 and CO2 ) or' room air (0 )'- 2 ei;
FINAL REPORT SwRI Project 01-8818-101 ., Page 2 Samples for analyses. of amonium - (NH3 ), chloride ion (Cl"), fluoride ~ ion (Fl'), and cyanide (CN) were collected by an impinger system. Amonium, C1, and F1 were analyze:, by ion chromatography and converted to ppm assuming 100-percent conversion. Cyanide analyses were done by the pyridine-barbituric acid metnod. , RESULTS , Combustion Summary: Samols 1099.1 began to generate a light smoke with a noxious odor within 30 seconds. By 1:30, the top of the sample was enarrea. Post test, the clotn at the top of tne- sample had turnea wnite, althougn the bottom side of- the' sample was only slign:ly charred. Sample 1099.2 began to generate a light smoke with a noxious odor within 30 seconds. The sample began to char on the surf ace during the test. Post , test, the silver glaze was gone on top, leaving a powdery residue. The cloth was left semi-in, tact on the ends and bottom side of the sample. Sample 1099.3 :egan to smoke almost ~ imeciately. At about 6:30, the sampli spontaneously i gnit ed ~and periodically evolved large amounts 'of flame ' for at least five minutes , accompanied by'large amounts. of black smoke. Post test, the sample remains were a black crusted char. 9
] Analytical Sumary:
The analytical data is sumarized in Table 1. The gas data are presented in terms of- the concentration time product (C*t product), which is the -ir.te-grateo value of the gas concentration. An examination of this table reveals substantial di f ferences in the amounts of CD and HCN produced, even when compared on _ a weight loss basis. Chile present, hydrogen fluoride (HF) and hydrogen chloride (HC1) were not generated in high concentrations for ^ any of - , the materials. However, a direct comparison is difficult as the two sats of materials were actually combusted under different conditions. The heat f'l ux - level was chosen on the assumption that the materials were extremely thermally o ,
g -O. FINAL REPORT SwRI Project 01-B818-101 Page 3 stable and that ignition would not occur. . Sample 1099.3 did ignite, however,
- and samples 1099.1 and 1099.2 did not. Materials can produce differ'ent k'inds and amounts of gases depending on the combustion mode and a direct comparison should be made using the same combustion moce. Another difficulty with sample 1099.3 arose when the CD evolution exceeded the . analyzer's range. Therefore, the CO value in Table 1 snould only be taken as an estimate of the range. and the actual value exceeos this. Graphical representations of the CD evolution for the three samples may be found in Figure 2. Again, the, CO level for 1099.2 levels is off 6t the maximum range of the CO analyzer. The amonium INH 3'. cata is not available at this time ano will be sent later.
e f 7 4
.k t
1' s u -_______.__.__ _________ _ _ _ _ _ _
O %'4
- t. .
j i 1- n- I -(. ry - [ C3. I /
- ~ .n .
/ \ ---
J ; O OL O 0 O a r cm
-- . n h . .
- \
4; / ..e .. N 3
. P.-:
i i i
/
..? x rigure la. Animal Exposure Chamber'and Racian: Jurnace.: -
~l ol l-
.., _ , d
- l --'" - - -
we a 6. . .e .,
/,
y
] ,-.. - . . ,
r .,... , 3 . .. . a _. _. . :
~
7 ....i 7
/
,g - -- . . .
2 avi' c'
,s . _a. /. .. , .
.. 1 J..:4 ;
o / e m...~. . .
,,.u f
i
/ s, p[_ .,,,,,, . y
\
/ _.
/-
v -
..: 4. :-
~
1
--)
.*\ /- l -
\-
. / /
/ \. e . . , _ ,
j .
. . . ... ._ s . .
- . e +=
i- i !- r 3 i Figure Ib. Comeustion. Cell /Rady ant Heater Assembly. ' e ... *
." E g g
., . ,;> ea 4, , ,
rrs.._---
.,3..:.
s -w - . . i j-. , u .,. .c
. - i.
. ,,. ' o
, . , . .r -
s 1- __
- n. . a. .s a. .>
. .z a. .. e. ; e. ,
, ., - , T
- M c'- -
t s ,. :. . . , , , - . .. i4
-.L
,. - .7 ,!
s I, A t
.f = ,
-f. ,
I - e -
. , , e - .
s.. 4 i o .
;9 \ . I-
, - e u 4
~ ' . $.
. . f. ,.
r s: +. n c . *. c e. .-
- s. .t -!-
s.. . m- . A . ! .t +
.+ . . ,. .. .. f t
J j ,~
- . -l.
v.
. .t
_f 1
./
ref ! 4 ' . - 4 . e.- ..
. c->.7 . . .. ., , , 4 .. . ., e.
Ji ; s '. . i .
. I f ,g I W .". ' , ; , , i.
Figure-2.. Carbon c M. o. noxide Concentration ' Curves. j ..
'},-
. =
i4t4 )j, *, y ,f
's -. ,
- o i...t
,+,'s ;' < f- ),
Q. t * .). .
i& : - .~;;:;, n- =
' ~ *'
2 ua. . v ~.. . . s l . , , , , ..... u
.4.w
.~
F.
?Asit 1. ANALYIICAL sumAAT imie CD tKM IICI ff ICI Man.
Iest Sample Percent Sal. . Concentratten- italght Concentration C-t C*t C*t Ct Cf T eg.- ag/L ppe-min ppe-alo pg.e-sla ppm-min pre-etn *F No. ' 39/1 _ toss 5- 16.5 21,966 283 610 1,770 0 131 1039.1 333 28' 21 32,490 280 1,390'~ 1,710 0 133 ; ' tc92.2 ~ %? 1,300 ' 267.5 60,456 560 2,000 3,540 0 -- 1099.1/.2 -- Fleming ' 18 2151- 2352,500* 85.289 5,848 3.01T 'O 350 1099.3 2.747
- CD evolutton enceed analyser's rasge.
- Actual vales enceeds this figure.
tr
'b -
B a,.h
= a-w- ,'
t-L S O U:T H W-E S T R E S-E A R C H. ~ N S T I T U T E - POST OPPICt Om Awsm 33610 8220 CULtl.8t A MO AD SA* JTO*eso, Tt R AS ?stes IS121864+6119 Ot*antut%ter Tttta 767Et
#1RE 1(C e*Not0G Y TWK 910 471 ttaa 1
Ct N USTION ANALYTICAL TESTING OF THREE CABLE WRAP MATERIALS, CTP 1099, A PGU".ED TSI THERM 0 LAG MATERIAL _ by Walter G. Switzer, M.S. FINAL REPORT SwRI Project No. 01-8818-10lb'- Prepared for: PROMATEC P. O. BOX 4672 HOUSTON, TEXAS -77210
%. n. m ., w3. - . u , o.
,. -s w .,. = w e w , m . - . .. . , .
.u u ., w. ~.u a. -a, - = , u..
-.- mo.o... -mm% w
. . eu. ., v. i m a .r a - - ~
ca. -u . s - . MAY 1986 Approved by:~ Reviewed-by
' I' f g[
L v t. im,w . / g Dr. Harold t.. Kaplan, Manager Or:. GordokE. Har sell, Difector . Applieti Environmental' Toxicology Department' of F1 e-Technology for a m Dr. Robert E. Lyle. Vice President
', Che'mistry and Chemical- Engineering I I _-' ,
- o. 11 . . . - . - - - . s .., 4 . _ .
i 1 .
..-_.,,..-~w..,- ,
}; R L 1% i +& 4 f 1^J**fidN N 5N^'k '"~ E y y y ;' .
~
y "'".'a .c . L. - 3x
- . - - . . . . - ~ . _ - .
o p *. .
.y : r
+-
g ;- 1 ., PROMATEC -
-CTP 1099,fSwRI1 Project No. 01-8818-10lb-
. COMBUSTION ANALYTICAL TESTING OF CABLE' WRAP , MATERIALS:- i MATERIALS .
; 9
h An additional material was, received from Promatec for combustion analyti-
- cal testing in the Radiant Panel System. This fLaterial?was Ldesignated asLa '
a
~
poured TSI Thermolag material . This sample measured approximately 31/2 to 6 " 1/2 in. and about 1-1/2 in. thick. The sample was slightly malleable to thej-touch. COMBUSTION PROCEDURE Prior to initiation of combustion, the - analyzers ' and recorders were? checked to . determine thac the systems were operatingf correctly. The; sample assemblage was weighed and l pl aced in a weighed Raluminum: foilo boat. ThisT q combination was then placed' on the floor of-the combustion cell.; and the cerli i moved into position on .the Radiant Panel Sysurr under thes 200-liter r.hmber ' (Figure 1). The exper.iment was initiated with the- applicationtof h'est flux: to the sample. The sample was subjected-to a heat flux. from ~ 7f4:t'oL 8 .4 W/cm 2 for' the ' 30-minute duration of the test. t . COMBUSTION ATMOSPHERE ~ ANALYSES- _, . -=
- V 4 Analyses of the combustion atmosphere was- made continuously for 02, CO,
] and CO2 with a closed lo'op = sampling system atia1 sampling ; rate of 500 cc, .in -
and the following instrumentation: Beckman.1 0M-11: analyzer (0 _2
); = Beckinan 865.:
.. o j
Infrared Analyzers - (CD and CO2 ). Prior to the experiment, each. analyzer Lwas calibrated with appropriate calibration- gases '(CD and- C02 )"or room air .(0 . 2 )*' D'O i-! . cd
/ 7,
& ~, ( g (
17 ': q . , _ >
.a m 3-3 , :.'ig f ; - 4 ,- :- ' - .
go 3 FINAL REPORT. -SwRI Project 01-8818-10lb- Page 2-Samples- for analyses of amonium (NH3 ), chloride ion (Cl"), fluoride ; ion - (Fl'), and cyanide (CN) were collected -by an impinger system. . Amonium, Cl ..
~
and F1 were analyzed by ion chromatography and converted. to ppm assuming 100-- percent conversion. Cyanide analyses were done by . the pyridine-barbituric
-- acid method, t . ,, RESULTS ,
Combustion Summary: This sample began to give off a white smoke almost immediately. At approximately 1:30, the top of the sample began to -bubble and started to char. Flaming combustion spontaneously occurred at approximately- 4: 30 : and - continued till at least 9:00. Post test, the sample remains were a biack
- crusted shape in the original form.
Analytical Summary:
. The analytical data is sumarized in Table 1-A. The gas data are pre- . - -
'sented in terms of the concentration time product (C*t product), which is the integrated value of ' the gas concentration. As in '.ae other test with tnis material -(CTP 1099, SwRI Project No. 01-8818-101), ignition did occur despite this material's purported thermal resistance., Carbon monoxide- (CO) evolution-was not as fast as in the other test for thi: material. Hydrogen cyanide
} t (HCN) produc. ion was also lower than in the other test. These- two observa-tions perhaps reflect - a- greater moisture content in this . sample. 'As ' in ~ the -
previous test, the C0 evolution exceeded the analyzer.'s range, despite an-
. increase' in the range from 12,000 to 60,000 ppm. - The - CO. value in Table 1-A:
should- only be used as an' estimate of. the .iange as the' actual value ' exceeds. this. Graphical representation of the C0 evolution for this test is L in' Fig-ure 1. The CO level in this graph' s levels off at the maximum reading Lof-the analyzer and below the actual talue. While present, hydrogen fluoride and hydrogon chloride (HClI were not generated in high concentrations. : Again', production of these . gases was lower than in the previous test. Weight loss,--
.however, is comparable with the previous test. .In an . exceptien from the
- previous -test,.1080 ppm NH3 was ' found in the last' 10-minute period of the .
test, yieldingsa 10,800 ppm min level .
,,w i
- 1) ~ M
], , . . . . . . . .
s~* f* M
. // /
.... I 3 l O boo :o: Ja -
N' j i C j r m
.l-
, . 1-i ,
w
~5
-j t ig g- j. .em , e.nes
/ . . i . .'\ ' .
s I i/ i i /= fg s..-.um / Figure 13. Animal Exposure Chamber and Radiant Furnace.
. .s m..
1 '" (~
,-u..- - -.
l
,V ._
.l . . . j . ,- u. - . ,- ... . .
~
n e
;-' } .
d C^ f,'s ~ _ /, - s : , - -., .-. . O \ **
/ s'
\ g ~~ _ "w.". m
,' m, .... . s
*? . v f :/ ,
;,.-2h.- I ' , -
' - / ~ / 35 ,
v
. - . . . . , , . . ./ ,
/
- . . . ~ . . . , ..
- ~ . . -
' Fi gure ' lb.'-- .
Combustton _ Cell / Radiant Heater Assembly. . . u.
m;
- .
- c + -
g s p c,4 e. e._ i _ { ;,
'?
u 1'
? ,
<, . 2 -
-r CROM-4
.J L .
.1
=:I c,,, ~~ ~ .
,1 g .;;. ./
. /
,, a -
.,, . .s 'n j.
c:n,
- c. - f
.t
" T5: 10 - 15 .?. O g. e,, .
- Y,3 T!11E . t1 I H >
B
.'J 7
-I
.- 7 .-
.s_
. j --
FI9ure 2. Car on Monoxide ' Concentration Curves. -
- ems W
D 0 {. e, 4
.{
d.
*auS _
t _h y
'll.
t* ='s= p .a 7,- .4
? . - - . &
ww: re o . 4 ..
-~
m.- r <.o
~
f' " _ 3 .'- ; JA. i S'i s 'm' '
;;:[, [-
EW;v.v
.l. i ;c i '* ,..
1 :,/ ~jM:d N .;rf w 2 -
'4
,' - .J
'y,,-.. - . . . . '.f"-
ar t ',-. _t-e.. <-.. , . -. .o r _.4 .pl. 3fs l~...M..;
- p; . % i >
<-C - -
-.s. -
2 -. w' . ..
;> + -- L. -- -
., .,.1. ,, w ; e.. . ,
,. 7 ,
t *
' + . .u
'%n
, y
, ...q;
# ^
.y.
,.I < r ;T.
m .. v~ i O
. b!
r [',.' _ o. v, . ;- a-a
~-1 -
IA8tf I.A. .. ANAt yIICAt - 5Uf9Wt f A
'g
- Iest- .' Sample Percent ,CG Smoke - , : HCN j leC1, 'W 188 3- fler. ;
. . . Ref; Concentration- ' Weight, Concentration C *t - C*t Ct . C't- . C-t ' - lemy. '
No. ag/L ' tossi : mg/t. -
- pre-ele pre-min pre-ala pre-min pre-min *f
~
- Pro =4 3019' ' 78 :2359 " ~2011920-- 30910 -3958 i
1210 ' 10.800: I62 4
. N a-
p
. t ig "
m
'O 1
s e 4 p. [
'4..
l g ; i
' O'"
m + 4' r ~
, b , ..-Y
e 4
..Yf
- s i m l' -.
W. 1'. , 4 3 J L '_ .l; vi j *.
-4
, +
- e'
- ~ ~'
g
' f,_
g_ gj . . . . ..
.,i s 7 . ! g.
,- e e'
-p' A g 4-g #-
.e-M .--<6 .
w-N' O '" l z_ n , ,;' . - .,'l 1 [r t. -{ u b:4 4 , , - - . 4 r J.u , % , , s. < 64 <,
- + -w w~ o-+%- - - r~--- --- - - -~~- --- -}}