ML20004C308
| ML20004C308 | |
| Person / Time | |
|---|---|
| Site: | McGuire, Mcguire  |
| Issue date: | 05/13/1981 |
| From: | Tinkler C Office of Nuclear Reactor Regulation |
| To: | Butler W Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20004C307 | List: |
| References | |
| NUDOCS 8106020576 | |
| Download: ML20004C308 (2) | |
Text
.
e 4
p reg ENCLOSURE 1
,O/
h UNITED STATES Q **
,7, j
NUCLEAR REGULATORY COMMISSION
,e t
WASHINGTON, O. C. 20555
> la j
N [,, #
MM 13 E81 MEMORANDUM FOR:
W. Butler, Chief, Containment Systems Branch, DSI THRU:
. Shapaker, Section Leader, Centainment Systems Branch, OSI FROM:
C. Ti.,kler. centainment Systems Branch. OSI
SUBJECT:
EVALUATION OF R&D ASSOCIATES REPORT RE:
IGNITER RELIABILITY As part of the staff efforts to evaluate the acceptability of glow plug igniter systems, NRR initiated a short tem testing program to be carried out at Law-rence Livemore National Laboratory (LLNL). The staff also arranged for R&D Associates to assist LLNL, as a subcontractor, in evaluation of the test data.
At the conclusion of their participation,'R&D Associates issued a report en-titled, "Some Hydrogen Control Considerations For Ice-Condenser Nuclear Plants,"
February, 1981.
In this report the authors discuss a possible reliability prob-lem for igniters.
The authors have conjectured that anomalous test results measured in tests 34 and 43 were produced as a result of the femation of a dense steam-cendensatien fog. The effect of a fog is to either suppress the temperature and pressure increase from combustion or inert the mixture depending upon the droplet sizes and the fog density.
Assuming that such an inerting fog is present in the ice condenser containment, the authors have attempted, in a quick evaluation, to detemine the effect1"e-ness of the mechanisms which would tend to remove or reduce the density of the fog. Most of their efforts were focused en the fog scavenging ability of the spray system. As a result of their calculations, the authors concluded further work was necessary to achieve confidence in the fog removal ability of the sprays.
The authors go on to state that other mechanisms might assist in prompt fog re-moval, but in a qualitative assessment they judge that these mechanisn6 are either inefficient or difficult to evaluate.
R&D Associates then concludes en the bases.of these arguments that the reliability of the igniters is in doubt.
At the request of the staff, LLNL has perfomed a brief review of the R&D Associ-ates report. A sucinary of the LLNL positien is as follows:
- 1) The experience of tests 34 and 43 is insufficient to support the conclusion that igniters are unreliable. Tests involving steam addition, other than tests 34 and 53, were perfomed using the same procedure and successful ig-M tien was achieved. So if fog existed in the unsuccessful tests at 50".
steam it most orobably existed in the successful 40" steam tests. There-fore, this evidence is insufficient to support the conclusion that the ig-niters failed de to the presence of fog.
P00R BRIGlh 8106 02 0 $76
MAY 131981 W. Butler J
- 2) The approach used in the calculation of the fog scavenging ability of the spray is satisfactory for its intended purpose. There is, however, no justification provided for certain key parameter asstsnotions, or the ex-ample calculation assumption that the fog concentration need! to be re-duced by one exponential factor.
- 3) The conclusion, that there is an apparent lack of nethods (other than the spray) for reduction of a steam condensation fog, has not considered ther-modynamic processes which would tend to eliminate fog.
It may not be neces-sary to rely on active methods for fog mduction.
The staff has concidered the views of 1.LNL and those of R&D Associates and has presently concluded that the igniters would mliably function in a TMI-type degraded core accident environment. That is they would provide with reasonable assurance protection against breach of containment. We do not believe the ar-guments put forth by R&D Associates are sufficiently persuasive to alter our conclusion that the igniter system is an appropriate interim solution to the problem of hydrogen control for degraded core accidents. In any event, the R&D report simply again poses the same basic question previously raised by the San-dia report, that is that there may be some instances of containment inerting.
That matter was discussed at length in the hearing, a es G. Tinkler Containment Systems Branch Division of Systems Integration cc:
L. Rubenstein l
l I
l l
P00R ORIGINAL
A.
ENCLOSURE 2 o
11 LAWRENCE LIVERMORE LABORATORY k
NUCLEAR SYSTEMS SAFETY PROGRAM
<t ! m g i
e j.
n TF81-055 March 12, 1981 2
t t3 s
h Mr. Charles Tinkler J/
Containment Systems Branch i
J. /
Division of Systems Integration 9
Office of Nuclear Reactor Regulation C~
U.S. Nuclear Regulatory Commission Washington, D.C.
20555
Subject:
Assessment of RDA report "Scme Hydrogen Control Considerations for Ice-Condenser Nuclear Plants"
Dear Charley,
We have evaluated the subject R and 0 Associates recort that was written as a result of their contractural obligation to assist in the analysis and interpretation of the LLNL Hydrogen Ingiter Test Program. That report consists primarily of two sections, one dealing with analysis of the LLNL program and applications of its results, and the other suggesting alternatives to the use of thermal igniters as a hyorogen control method. This second section is of a conjectural nature and will not be addressed in this assessment, although several uf the suggested alternatives do merit consideration.
The assessment of the LLNL test results essentially confirms our statepoint determinations in the steam test series.
Small differences exist due to inconsistencies in available data (gas analysis of test #43, for example) but these are generally minor and inconsequential. The two anomalous tests (134 and #43) are examined in some detail, and these results are the basis for the authors' stated positice regarding a reliability problem with the igniter system.
The subsection entitled " Discussion of a Possible Reliability Problem for Igniters" addresses specifically the influence of condensation-generated fog on the cerformance of thermal igniters.
In the previous section the authors stated "it is our conjecture that a dense condensation fog existed in Tests 34 and 43".
It is further implied that this fog inhibited the caDability of the glow plug to ignite the mixture, and that the ice condenser containments are conducive to the production of fog by bulk condensation of moisture in saturated air as it flows through the ice compartment. One postulated mechanism for fog removal is the containment spray system. A model is developed to determine the length of time required for a given spray system / containment combination to clear the atmosphere of fog to some undefined level. A " relaxation time" is defined as a function of several geometric f actors and the " fog particle collection efficiency".
In the Q
)(d 8103240485 S( 0
- .nwsty of Casfer se scacn Sc8 :.wmore Caserria 94550 C *eaeonone'45447 f100 C c
9 aEC U.1 ' vuF
1 Mr. Charles Tinkler Page 2 March 12,1981 TFS1-055 specifig example provided, a lcwer limit for this collection efficiency of 3 x 10-4 is suggested which would give a relaxation time of 55 minutes.
It is implied that 55 minutes is the longest allowable relaxation time.
The report cites values " reported in the literature on aerosol coagulation" for the collection efficiency as small as 10-2 to 10-3, which would give relaxation times in the example problem of 17 and 167 minutes, respectively.
The concluding remark of the section is as follows:
"At present we consider the reliability of igniters to be in dcubt because of:
(1) the experience of tests 34 and 43 (2) tne above arguments on fog coagulation by containment sprays, and (3) an apparent lack of other methods for om'* raduction of condensation fog."
Our position on the RDA report is directed toward these conclusions, and is provided after they are restated below.
1)
"the experience of Tests 34 and 43" Tne authors state it is their conjedture that fog existed in these two tests, and imply this led to the f ailure of the glcw plug to ignite the mixture. However, there is a high probability that fog also existed in the 40% steam fraction tests (which were successful).
In the subsequent observations of the vessel during similar steam injections it was noted that the fog visually disappeared between 40% and 30% steam as the concentration dropped.
In the two anomalous tests the steam concentration fell to as low as 23% while the glow plug was activated. We do not feel this evidence supports the conclusion that the igniters f ailed due to the presence of fog.
2)
"the above arguments on fog coagulation by containment sprays" The model used to identify the relaxation time is basically sound, for a rough engineering calculation. The geometric parsneters are str ai gh t-forward. However, the fog particle collection efficiency is ill-defined and does not appear to be an easy parameter to assign a value to. That the " literature" (no references) provides values as small as 10-2 or 10-3 leaves us with no real feel for the magnitude of the problem. Utilizing a " relaxation time" to determine how long it would take the containment sprays to diminish the fog presupposes that we know to what level the fog must be diminished.
That we do not knew.
In general, we feel the model is satisf actory as a rough tool, but the problem is too complex to reacn significant conclusions without a better understanding of the input carameters. We do not know what level of fog is unacceptable.
P00R ORIGINAL.
Mr. Charles Tinkler Page 3 March 12, 1981 TF81-055 3)
"an apparent lack of other methods for prompt reduction of condensation fog" It is not clear to us that enough effort has been directed t3 ward studying the LOCA phenomena in terms of fog generation / elimination in such an environment.
It is kn wn that thermodynamic processes, for example temperature and pressure increases, tend to eliminate fog.
Nor is it clear that the fog will remain suspend 6d in the containment environment for long periods with the presence of containment air circulation and the expected temperature gradients.
It may not be necessary to rely on active methods for reduction of fog.
These statements summarize the LLNL position on the RDA report.
If you need any further information or clarification, please let us know.
Sincerely, Yb nD William Lowry Thermo Fluid Mechafn s Group 1 f.E h.'it A'. J Barry R, 85iman', Ph.D.
Group Leader Therm (Fluid Mechanics Group cc: Harmon Hubbard (RDA)
Samuel Zivi (RDA) t P00R BRIGINAL l
.