ML20126M672
| ML20126M672 | |
| Person / Time | |
|---|---|
| Issue date: | 05/09/1978 |
| From: | Mattson R Office of Nuclear Reactor Regulation |
| To: | Kehnemuyi M NRC OFFICE OF STANDARDS DEVELOPMENT |
| Shared Package | |
| ML19250G967 | List:
|
| References | |
| FOIA-81-201 NUDOCS 8106190356 | |
| Download: ML20126M672 (3) | |
Text
%
g,m./m
[pu pf cq%
b*41ItO STAlfS O'h s't f JUCLE AR R EGU LA l oiiY Cc '.".*J ' '.lC N j QQ:'))$
w.cm mion. o. c. n s a S;nQ 7 :
'3Y 9 lfi3 8
('~$54 3
.,s MLMOPMDUM FOR:
M. Kehnemuyi, Assistant Dirx '.er for General Engineering Standards Division of Engineering Standards Office of Standards Development FROM:
R. J. Mattson, Di rector Division of Systems Safety Office of Nuclear Reactor Regulation
SUBJECT:
DRAFT SANDIA REPORT ON FIRE DETECTION SYSTEMS Per your request dated March 17, 1978, and numbered DSS-38, we have revice..ed the subject document.
Our comants are as follows:
1.
The report notes (on page 10) that the Draf t Regulatory Guide 1.120 lacks sufficient guidance with regards to specifying detector type dnd placewent throughout the plant.
What" the report overlooks is the fact that the Reg Guide 1.120 places the responsibility un the applicant by requiring a fornal fire hazards c.nalysis that must be conducted by a qualified fire protection engineer who would selact the type and location for each detector.
This should remedy the concern expressed in the Sandia report.
Consequently, recomr.2nd2-tions A.2 and A.3 lack merit.
Thus, the report should be revised to recognize that this method should result In acceptable selection and location of detectors.
,) wi n+zt 2.
On page<2P$5a% a concern is expressed that major deficiencies exist in the UL test procedures in that ionization detectors are not tested to materials such as; cable insulations, lube oils, diesel oils, or plastics coraonly found in nuclear plants.
H7.. a v e r,
UL St:ndard 167 for ionization detectors subjects the detectors to products of combustion from paper, gasoline, polystyrene, v. cod, and cotton.
Additionally, UL Standard 168 for photoelectric detectors uses gasoline, kerosene, polystyrene, wood, and punk.
Gasoline and l
kerosene are very close to diesel fuels and lube oils in fire i
behavior given ianition.
Basically, these S1s give off dense snoke with high thermal energy.
Polystyrene produces long carbon chains that is a characteristic of most plastics and cable insula-tions.
In addition, the general thrust of tSa Sandia report under Section III, Part B, is that qualification testing is beset with I
CONTACT:
I G:iarrison, ASB X27763 61061003Sg
s I
M. Hehnemuy1 gay 91?78 deficiencies and/or inadequacies.
While inprovenants can ahcays be made, the real problem is not the one cited in the Sandia repart, but, rather, manifests itself in the lack of comparison test data for each detector class.
This, then, prevents intelligent decisions from being made in the selection of detectors for specific fire risks.
The I;ational Bureau of Standards recognizes this problem and report 14BSIR 76-1172, "A 1:ew Test Method for. Autoniatic Fire Detection Devices," prepared by Mr. R. G. Bright (nationally recognized authority s
in fire / smoke detection) proposes a new test method that appears to close this gap in an acceptable manner.
The Sendia report should include discussion of the !!BS proposed test method.
3 3.
The Sandia report presents generally known shortcomings of available 3
criteria or procedures for testing detectors in situ.
liowever, their allusion to the lack of testing require; rants in Reg Guide 1.120 is l
uisleading.
The Sandia report overlooked the existence of the I RC t
Tech Specs which specify when and how frequent detectors should be
)
tested in nuclear plants.
The Sandia report should be revised to l
acknewledge that these !!RC requiremants provide for periodic detector j
testing and maintenance.
}
4.
Recom.nendations B.1, B.2, and B.3 are not considered to be worth l.
pursuing becuase of existing work ongoing (or accomplished) within the llational Bureau of Standard's comprehensive smoke detection
/
research programs. Additionally, firms involved in detection l
equip:c,ent or suole detection engineering are aware of the parameters influencing detector selection expressed in Sandia recomundation l
B.1.
j 5.
Sandia's recommendation D.1 is well taken and we concur with it.
lbwever, this aspect is also under study at the I BS smoke research program.
To date, a portable, si.wke particle generator has been developed'for in situ testing.
In sunnary, We are of the opinion that the draft Sandia report does not provide any infonnation that would serve as further guidance on fire ddection systems.
In direct response to the OSD question regarding dcvelopmnt o' a regulatory guide on fire detection based on the infor-mation presented in this report, we are of the opinion that the report l
does not provide any information that i. uld serve as further guidance
[
i O
i
,6 M. K:hncmuyi l [g 9 gj3 en fire detection systems and, therefore, does not support developnent of a regulatory guide on fire detection systems.
Although we believe that a fire detection guide would be beneficial, it is our opinion that the draf t Sandia report is not the proper vehicle to accomplish this end.
h'e unde r-stand that OSD is discussing with the !!ational Fire Protection Association the preparation of additional nuclear plant fire protection standards that will include firb detection.
This docunent should be the proper basis for preparing an !!RC guideline.
e a
f c,
f_. (
Roger ]J. Mapison, E{ Safety rector Division of%ysteias Office of t!uclear Reactor Regulation I
I e
e es k
0 e
e 0
i e