ML20206G928
| ML20206G928 | |
| Person / Time | |
|---|---|
| Site: | Surry, 05000000 |
| Issue date: | 02/06/1987 |
| From: | Michael Corradini WISCONSIN, UNIV. OF, MADISON, WI |
| To: | Eltawila F NRC |
| Shared Package | |
| ML20204G644 | List:
|
| References | |
| RTR-NUREG-1150 NUDOCS 8704150200 | |
| Download: ML20206G928 (4) | |
Text
_
i.
s j,
ff I
To: Dr..Farouk El-Tawilo USNRC Regulatory Branch Washington DC 20555 g'
o From: M.L. Corradini University of Wisconsin Madison WI 53706 4
Subject:
SCOPING ANALYSIS FOR DIRECT HEATING IN SURRY The physical process known as ' direct heating' may i
occur during a severe accident in a LWR when the core melts down and slumps into the RPV tower plenum at high system pressures (>1-2 MPa). If the primary system remains pressurized as the vessel is breached and a large mass of molten corium (>25%) is ejected, the containment con be significantly pressurized. This pressure and accompanying temperature rise is due to release of the melt thermal energy to the atmosphere as well cs the exothermic metal / oxidant chemical reactions 4
that may occur. Another concern from the direct heating process is that the metal / oxidant reaction may produce hydrogen if the oxidant is water vapor. In this case the metal is oxidized and the hydrogen produced may be transported to other regions of the containment for possible combustion. This physical process has been recognized before in steam explosions with metallic fuel (FITSC, D and G series at Sandia and the CWTI tests at ANL). Although this is a hydrogen source term problem it should not be forgotten.
In order to assess the likelihood of such an event hand calculations were performed to determine the quantitative range of various parameters.
The approach taken was to focus on one specific PWR containment (SURRY) for which there were detailed drawings j
available. Quantities such as the average velocities and Weber numbers thrc9gh the cavity and lower compartments were computed. A summary of the results is presented in i
this memo. The detailed hand calculations con be provided if one is interested in the details.
The Surry PWR is housed in a large dry subatmospheric containment (50000m3). In the event of a high pressure meltdown and dispersal of the molten corium the melt would be ejected into the reactor cavity and may be dispersed into the lower containment compartment and perhaps the upper compartment, depending 1
i 8704150200 870408 PDR NUREG t
1150 C PDR 99
I l
pags - z l
l upon the relative velocity between the gas and the l
droplets. In this scoping analysis these regions are treated as control volumes.
For determination of the significance of direct i
heating for a particular containment one must consider a few particular physical processes:
- 1) The potential for entrainment of liquid corium (or ovallable water) in terms of the onset of entrainment and the rate of entrainment;
- 2) Deentralment of the droplets once the gas velocity has fallen below this threshold;
- 3) The chorocteristic time needed for release of its thermal and possibly chemical energy;.
- 4) The potential for water / fuel droplet mixing in the during the gas dispersal;
- 5) Effect of freezing on structure, i e. concrete and steel.
The possibility for entrainment by the RPV vapor / gas blowdown has been analyzed in detail by Pilch.
1 The onset of entrainment con be determined by calculation of the Weber number based on the length scale of the Laplace constant (i.e. Kutateladze number).
1 When one uses overage flow velocities through the reactor cavity for Surry one finds that the Kutateladze number exceeds the critical value of 10 by one or two orders of magnitude for RPV pressures of 2.5 to 16 MPa.
This result is even stronger if one considers the i
heating of the gas in the reactor cavity by the corium droplets. In addition the rate of entrainment is also large ranging from 0.25 to 1.75 m/s. This entrainment rate implies that the fuel would be quickly (<<10 sec) dispersed in the cavity gas flow during blowdown. In addition one should note that because water is less dense it would be even more easily dispersed in the gas phase during blowdown in the cavity. This will be discussed subsequently. The final' point to make about liquid corium entrainment in the reactor cavity is that i
the entrainment rate would actually be enhanced by two processes not even considered in this scoping analysis.
First, Powers has pointed out that hydrogen may become dissolved in the corium melt at high pressures and then may come out of solution when the corium is ejected from the RPV. This gas release jet breakup if correctly estimated by the theoretical solubility calculations of Powers would add to the breakup of the corium and its eventual entrainment.
Second, if water is present in the cavity the steam production from the corium water thermal interaction i
would also increase the local vapor / gas velocity and enhance this entrainment.
For the Surry containment the lower compartment above the reactor cavity is quite large in volume (10000m3) g
-v
--c-y
,vi-yrs y-,,,
__a
-,-e,,w-,
s
_-,_m
__-.--__--y---_.-wwww+=N a
i pags - o and torlodal in overall shape. The floor of the lower compartment is at the some elevation as that of the cavity. Therefore, when the corium is dispersed into the lower compartment, it would be swept into the large room above ground. Once exiting the cavity the gas velocity would quickly decay to a low value and the corium debris would take a ballistic path. If unimpeded the debris l
would reach a peak height and settle by gravity. In all the situations considered in this scoping analysis the vapor / gas velocity in this compartment was too low to cause.entrainment or to sustain its path with the gas flow streamlines; i.e. hand-calculations suggest a debris diameter of 0.3-3.0mm. In this situation the time i
for heat transfer in the atmosphere would be no shorter than the time to. travel to the top of the compartment (e.g. 0.1 sec). In this time the corium fuel can lose a significant fraction of its heat (50% of thermal energy). The effect of impact with miscellaneous structure in the compartment would be important in reducing the path length and allowing for structural interactions. However, this complex yet realistic effect could not be accounted for easily here.
Transport of the corium debris into the upper compartment con only occur two ways. First, the debris may be directly blown into the upoer dome along the vessel wall. Based on drawings this accounts for about 15% of the total exit area, therefore would account for some contribution unless obstructed. Second, the debris may ballistically enter the upper dome after cavity exit into the lower compartment. Both of these pathways must be considered although for the Surry geometry this seems to be a small fraction of the total debris.
The effect of corium-water interactions would be important when water is present and hcs been investigated by the ANL researchers in the CWTI tests.
One should realize that some water will always be present in one of three locations:
-l within the vessel during and after RPV breach; 11 within the reactor cavity; 111 on the containment floor.
The extent of corium-water droplet mixing is quite dependent on the location and the accident sequence; e.g. containment sprays may be operating which supply water droplets premixed in the oatmosphere. The details of each situation has only been considered in general.
Scoping analysis is still underway on this topic, because of the large mitigation effect it may have on the direct heating phenomena.
The final topic to consider is the amount of energy that can be removed when the corium debris impacts solid structure. If one considers this energy transfer to be analogous to elastic impact of spheres with the wall the i
n
-.,,n
pags - o amount of heat transfer per droplet is small, although the overall integrated effect may be large. When one considers the effect of fuel freezing on the solid structure one finds that only a thin crust will be formed during the time of the fuel / gas blowdown (<1cm).
This thin crust along with the overall droplet impact time and frequency controls the amount of heat transfer.
For the case of concrete structure this heat transfer is also controlled by the low thermal conductivity wall as well as the ability to remove energy without outgassing the concrete and assisting the reentrainment process.
Based on the case of basaltic concrete one finds that the maximum heat flux is small (<0.35 MW/m2 in the cavity). This limits the amount of core that could freeze on the structure, yet does not cause signficant gas production. To consider this effect one must remember the large amount of miscellaneous structure available in a containment.
We are continuing our analysis of the Surry containment. Our aim is to develop a methodology that con be used in categorizing other containments.
i l
.-