March 15th, 2011 – Don’t know if Japan added hardened vent but I’m sure they are well aware of it

Author: No Comments Share:

From: Beasley, Benjamin
Sent: Tuesday, March 15, 2011 5:02 PM
To: Kauffman, John
Subject: FW: SBO impact on Mark I’s
Attachments: ORNL Study Secondary Containment.pdf

I have been wondering if, after things settle down, you or I should propose a generic issue on extended station blackout.


From: Lane, John
Sent: Tuesday, March 15, 2011 2:28 PM
To: Beasley, Benjamin
Subject: SBO impact on Mark I’s

Ben, FYI–Here is a report from ORNL from the late ’80s, a time when NRC was actively studying containment/secondary containment failure issues. It provides a little bit of background information about station blackout studies undertaken then and the impact of SBO on the secondary containment.

The NRC required Mark I’s to add a hardened vent around 1990, when it was discovered (probably from NUREG 1150) that the containment was likely to fail (up to 90% likely) as a result of some core melt accidents. The fix was intended to allow for a gradual release of overpressure to maintain the containment integrity as much as possible. I don’t know if the Japanese plants added the hardened wetwell vent but with GE/Hitachi right there, I’m sure they are well aware of it.


Sheriell R. Greene
Oak Ridge National Laboratory

During the past four years, the ORNL BWRSAT Program has developed a series of increasingly sophisticated BWR secondary containment models. These models have been applied in a variety of studies to evaluate the severe accident mitigation cappbility of BW1r secondary containments. This patter describes the results of a recent ORNL study of the impact of BWR MK I primary containment failure dynamics on secondary containment integrity. A 26-cell MELCOR Browns Ferry secondary containment model is described and the predicted thermodynamic response of the secondary containment to a variety of postulated primary containment failure modes is ,7e~anted. The effects of primary containment failure location, timing, and ultimate hole size on secondary contai,.ment responise is investiga;. , and the potential impact of hydrogen deflagrations on secondary cont inment integrity is explored.

The most common boiling water reactor (BWR) plant design in the United States is the BWR-4/MK I primary containment system. These plants employ secondary corntainments (Exhibit 1) consisting of a reactor building and refueling bay that completely surround the primary containment. Detailed severe accident analyses of MK I containment designs generally indicate that the conditional probability of primary containment failure is quite high in the unlikely event that core debris escapes the reactor vessel.

Should the primary containment pressure boundary fail, the secondary containment becomes the final barrier between the plant’s fission product inventory and the environment. Traditional BWR risk studies have, however, de-emphasized the ability of the secondary containment to act as an effective fission product trap, During the past four years, the ORNL BWRSAT Program has developed a series of increasingly sophisticated BWR secondary containment models. These models have been applied
in a variety of studies to evaluate the severe accident mitigation capability of BWR secondary containments.

This paper describes the results of a recent ORNL study of the impact of BWR MK I primary containment failure dynamics on secondary containment integrity. Tle fundamental design characteristics of the Browns Ferry secondary containment are first discussed, followed by a brief description of potential MK I severe accident containment failure modes. A 26-cell MELCOR Browns Ferry secondary containment model is described and the predicted thermodynamic response of the secondary containment to a variety of postulated primary containment failure modes is  presented. The effects of primary ccntainment failure location, timing, and ultinate hole size on secondary containment response is investigated, and the potential impact of hydrogen deflagrations on secondary containment integrity is explored.

Domestic BWRs of the MK I primary containment design employ a secondary contairment which is comprised of a multi-floored reactor building and a refueling bay which completely surround and enclose the primary containment. Multi-unit plants employ separate reactor buildings for each unit but may utilize a common refueling bay to service all units. Exhibit 1 is a cross sectional view of the Browns Ferry Unit 1 reactor building and refueling bay (shared with Units 2 and 3). The Browns Ferry reactor building is a massive (1.4 million ft 3 or 40000 m3 ), five floored structure with reinforced external concrete walls. The thickness of the walls varies from 6 ft (1.8 m) in the reactor building baseme. to 2.5 ft (0.76 m) at the junction of the refueling bay siding and the reactor building wall.

Secondary containment above the reactor building is provided by a 2.75 million ft 3 (77700 m3) refueling bay which is constructed of corrugated sheet metal walls that contain large blowout panels to provide protection from the effects of tornados and steam line breaks. Not shown in Exhibit I are details such as stairways, elevator shafts, and internal blowout panels which provide communication pathways between the various floors of the reactor building and between the reactor building and the turbine building.

The Browns Ferry Final Safety Analysis Report’ indicates that the  above grade exterior walls of the reactor building are designed for pressures up to 250 lb/ft 2 (11970 Pa) without structural failure. The tornado design basis is a pressure decrease of 3 psi (20684 Pa) at a rate of 0.6 psi (4137 Pa) per second. The refueling bay siding is designed to withstand internal pressure in excess of 57.6 lb/ft 2 (2758 Pa) without structural failure. Pressures in excess of 50 lb/ft 2 (2394 Pa) will, however, be relieved by blowout panels in the siding.

The design basis accident for existing MK I primary containments is the large break loss of coolant accident in which one of the main re circulation pipes is assumed to circumferentially rupture. The purpose of the primary containment is to Limit the release of fission products from this accident to levels which will not exceed the limits of
10 CFR 100. This goal is accomplished by designing the containment to withstand the predicted tranjient pressure and temperature loads induced by the blowdown of steam and hydrogen (produced by cladding oxidation) from the reactor vessel. The design pressure and temperature of the Browns Ferry primary containment are 56 psig (487 kPa) and 281*F (411 K). The primary containment is inerted with nitrogen during reactor operation.

Recent ORNL calculations for an unmitigated short-term station blackout severe accident sequence at Browns Ferry 2 indicate that temperatures as high as 2700°F (1750 K) may be generated in the primary containment if the majority of the core was to be relocated onto the drywell floor. Maximum primary containment pressures for this case
appear to be limited primarily by the containment’s maximum pressure capability. A recent Chicago Bridge and Iron Company study3 of the ultimate pressure capability of Peach Bottom’s primary containment produced a maximum pressure capability estimate (assuming median gasket resiliency) of 140 psia (965 kPa), with failure predicted to occur via leakage past the drywell head flange assembly. Since the design of the drywell head flange assembly is plant specific, the Peach Bottom results cannot be applied a priori to other plants. It must be noted, of course, that the continued pressure increase associated with the evolution of noncondensible gases from an unmitigated core/concrete reaction would eventually result in over-pressure failure of the primary containment unless precluded by some other failure mechanism.

A second potential mechanism for MK I primary containment failure in an unmitigated severe accident is drywell liner (shell) ablation due to direct attack by molten corium. The ability of molten metals to erode steel structures is well documented. 4 While significant uncertainties surround the behavior of core/concrete reactions and corium
spreading in a MK I containment configuration, 2 preliminary analyses indicate failure of the HK I drywell liner is quite likely if core debris does contact the inner liner surfaces.

Should the liner fail near the drywell floor elevation, the most probable sites for blowdown entry into the secondary containment are the reactor building basement torus room and the second floor of the reactor building (Exhibit 2). The transport path for the blowdown is the gap between the drywell shell and the surrounding reactor building concrete, and the annular gaps surrounding the drywell vent pipes and penetrations.

These gaps provide a 145 ft 2 (13.5 m2 ) flow path into the torus room and a 135 ft1 (12.6 m2 ) flow path into the second floor of the reactor building. Since elevated drywell pressures and temperatures result in swelling of the drywell liner and a reduction in the gap between the liner and the reactor building concrete (Exhibit 3), it appears that the etfective flow path area for drywell blowdown would be limited by the actual size of the drywell shell rupture or the available space between the liner and the surrounding concrete. Significant uncertainty therefore surrounds both the ultimate hole size and the ablation time associated with opening of the hole for this drywell failure mechanism.
Given the uncertainties surrounding the dynamics of MK I primary containment failure, it appears prudent to investigate the impact of a range of failure mode assumptions on secondary containment hydrogen deflagration phenomena and building survivability. Such an investigation is possible only via detailed computer simulations of secondary containment behavior. During the past two years ORNL has developed an extremely detailed computer model of the Browns Ferry Unit 1 secondary
containment. That model is described in the following section.

Previous Article

March 15th, 2011- Benjamin Beasley – Edits to NBC FOIA GI-199 Request

Next Article

March 15th, 2011 – TEPCO fail on info-sharing / No N-crisis HQ for 4 days after tsunami disabled reactor cooling system