ML19260C450
| ML19260C450 | |
| Person / Time | |
|---|---|
| Issue date: | 11/01/1979 |
| From: | Colmar R Office of Nuclear Reactor Regulation |
| To: | Adensam E Office of Nuclear Reactor Regulation |
| References | |
| REF-GTECI-A-01, REF-GTECI-PI, TASK-A-01, TASK-A-1, TASK-OR NUDOCS 8001030449 | |
| Download: ML19260C450 (4) | |
Text
,
R., td/
e p nog o,
UNITED STATES
'[$
NUCLEAR REGULATORY COMMISSION E\\
. v.
E WASHINGTON. D. C. 20555 9
E
%.',1. * #
NOV 1 1979 MEMORANDUM FOR:
E. Adensam, Section Leader B, Plant Systems Branch, D0R FROM:
R. J. Colmar, Plant Systems Branch, D0R
SUBJECT:
DRAFT REPLY TO ACRS REQUEST; THERMAL HYDRAULICS In a memo prepared by W. T. Russell to NRR Directors, dated October 9, 1979, the status of outstanding ACRS requests and recommendations made since February 1978 were presented with a request by him for responses from the appropriate individuals.
Dr. Hanauer's group on unresolved safety issues was identified as being responsible for preparing a response to the " Thermal-Hydraulics reques t/recom-mendation, under the TAP A-1 effort on Water Hammer.
The specific ACRS request / recommendation is reproduced below and the draft response which I prepared on a very short turn-around time (10/23 to 10/25) for C. Burger of RES is attached for your information.
Thermal-Hydraulics _
During the 230th ACRS Meeting (June 1979), Dr. Plesset reauested that the NRC staff provide information during the June 19-20 ECCS Subcommittee Meeting regarding the potential mechanical forces that can be developed by injecting cold water into steam and steam / water mixtures, and hot pipes.
"The ACRS also recommends that ea:h licensee and construction permit holder should examine a wide range of anomalous transients and degraded accident conditions which might lead to water hammer. Methods of controlling or preventing such conditions should be cvaluated, as should research to provide a better basis for such evaluations.
TL Committee expects it would be appropriate to have such studies done generically first, for classes of reactor designs and system types."
With regard to the above request made by Dr. Plesset coricerning the potential mechanical forces that can be developed by injecting cold water into steam and steam / water mixtures, and hot pipes, it is noted here that the physical conditions of the steam / water contact are critical in determining the magnitude of the potential forces.
If cold water is sprayed as droplets into a steam atmosphere in a vessel, such as a tank for example, condensation will occur and the forces will correspond to the change in the saturation conditions.
These are generally mild forces and the mechanical effects would be expected to be relatively mild.
L'6 1700 020 8001 030 j
E. Adensam NOV 1 1979 For the injection of cold water masses instead of droplets into a vessel containing a steam atmosphere the steam / water interface may remain relatively calm so that condensation occurs at the water boundary in a quiet way. Under these circumstances the forces would again be likely to be relatively mild.
If, on the other hand, the steam / water interface is turbulent, it is conceivable -
that some steam can be trapped by the surface water to form a discrete steam bubble within a subcooled water system.
Under this circumstance, the steam condensation can be extremely rapid and a water hammer situation arises. However, the determination of the magnitude of the forces is not straightforward for many reasons.
For example, steam entrapment causing the bubble must necessarily occur relatively near the surface for highly subcaoled water because it would not penetrate too deeply into the water without condensing.
In such a case, depending upon the bubble size, the water velocity into the void space would determine the magnitude of the pressure according to the Joukowski relationship P=/CV.
But on the other hand, the transient time, which affects the impulse load on the systems, is determined by the proximity of the surfaces to the region of collapse. That is to say, if the water surface is closer than the structural surface, then the local pressure by the water haniner can be relieved through the steam / water interface before the pulse reaches any structural surfaces.
In this case, the effects are very local and probably of little damage consequerce.
In the event that the structural surfaces are closer to the initial region of collapse than the water surface, the steam / water inter-face still controls the magnitude of the impulse but some structural loading will occur.
For each foot of submergence, for example, the duration of the impulse would be R = at 2
seconds
=.0004
=
C 4500 or about one-half of a millisecond.
This is the maximum value of the pulse width if the bubble collapse occurs at the structural surface; this value would be reduced in proportion to the distance of the structural surface from the initial bubble collapse region.
The actual magnitudes of the forces and impulse loads are difficult to calculate in general. The water collapse velocity, in particular, is an unknown quantity.
However, some insight into the possible magnitudes of these events may be inferred from the Mark I and II containment tests if the subcooling conditions are appropriate.
However, these tests are reverse in nature in that steam is injected into water (instead of water into steam), in which case the bubble penetration may be deeper than in the case under consideration here.
For this reason the Mark I and II impulse loads may be upper bounds for water injection into steam, provided the subcooling is the same in both cases.
1700 021
NOV 1 1979 E. Adensam For steam / water mixtures the situation is such that the fluid is at or near saturation.
In this case, the injection of cold water can lead to encapsulation of a steam bubble by the cold water. Here again, the bubble condensation can be quite rapid, if the water is cold enough, leading to rapid acceleration of the water into the.
void space. The magnitude of the pressure pulse is again somewhat indeterminate because of the difficult in defining the Joukowski pressure peak. Also, the geometry and configuration affect the pressure wave transit time, as noted above, so that the determination of the impulse load also requires that the configuration be specified.
For injection of cold water into steam / water mixtures it is also to be noted that the subcooling that occurs locally to the steam bubble is a critical parameter in detemining the nature of the bubble collapse and the subsequent loads; that is, whether it is inertially dominated, thermally douinated, or is in a transition region, sometimes referred to as en "anomolous" region because of multiple pressure peaks. The local subcooling is difficult to predict because of the difficulty in tracking the injected cold water.
If it mixes with the hot water first, then ancapsulates a steam pocket the results will differ, perhaps markedly, than if the cold water immediately encapsulates a steam pocket.
The latter possibility leads to an inertially dominated pulse, the former case tends towards a thermally dominated collapse, with the development of relatively mild pressure peaks.
The injection of cold water into hot pipes can cause, in addition to the thermal shock, steam entrapment and rapid bubble collapse if the pipe is initially steam filled. As indicated above the Joukowski pressure magnitudes or the impulse cannot be calculated at present with any degree of confidence or reliability. The staff is presently attempting to establish a program in which a calcula-tional tool will be developed to be used in estimating the magnitude of the water hammer forces and structural loading due to steam bubble collapse.
This will be amplified below.
Another aspect of cold water injection into hot pipe containing steam is tnat local depressurization can occur as a result of steam condensation.
This local depressurization would then induce a steam flow transient within the system that could cause transient structural loads because of slug flow. The magnitudes of such effects is an unknown at present.
1700 022
NOV 1 1979 E. Adensam At the present time the NRC staff has completed several phases of the generic effort on water hammer. This has included the issuance of reports by EG&G covering a systematic review of the LER's and incidents involving water hamer, the formulation of calculational methods, and the issuance of NUREG-0582, " Water Hammer in Nuclear Power Plants." The calculational methods are an effort to evaluate the consequences of water hammer under degraded conditions such as check valve closure associated with an upstream pipe break and the discharge into a voiJed line in order to evaluate the consequences of reactor spray initiation. An effort is also underway to establish a calculational model with which to evaluate the consequences of water hammer induced by steam bubble collapse.
Moreover, for the near term the staff is planning a meeting with utilities to discuss with them the safety significance of the many reported water hamer events in their plants and to identify the licensee's efforts to deel with this problem.
Following these short-term efforts, which incluce some calculations of consequences and discussions with licensees as well as interim licensing positions, the staff will consider a long term program which will be directed towa*ds establishing the means of preventing or controlling the water hammer events in order to fin ly esolve the water hammer probl em.
l
- h. J. Colmar Plant Systems Branch Division of Operating Reactors cc Ayco L.. ourger F. Cherny D. Fischer S. Hanauer G. Lainas C. Tan J. Zwolinski 1700 023