Study finds no reactor designs "inherently safe"

In the first major critical study of the so-called "advanced" reactor designs, the US-based Union of Concerned Scientists (UCS) has concluded that no such design is "inherently safe", a claim many nuclear industry representatives are making. The study's findings are important not just to the US, but to Third World countries that nuclear vendors are also targeting as potential buyers.

(342.3421) WISE Amsterdam - An attempt was made by the US Nuclear Regulatory Commission (NRC) to rule that licensing standardized, advanced reactor designs would involve early site approval, certification of the plant design by rule (avoiding public hearings on plant design for each plant built), and simultaneous issuance of a construction permit and operating license (the so-called "one-step licensing" or "combined licenses"). Although a major portion of that rule has just this week been over-turned (more on that next issue), the nuclear industry (both in the US and abroad), in anticipation of this sort of licensing procedure, has developed a number of design concepts for advanced reactors. This study, which was conducted for UCS by MHB Associates, reviews the issues surrounding the licensing and deployment of these advanced reactors in the US. In the course of the study, three specific designs were reviewed:

• the Westinghouse Advanced Passive AP-600 pressurized water reactor
• the General Atomics Modular High Temperature Gas-Cooled Reactor (MHTGR); and
• the General Electric Power Reactor Inherently Safe Module (PRISM) liquid metal cooled reactor.

Although the study found that "advanced reactor designs appear to hold some promise for resulting in improved operations safety compared with existing plants," serious problems were found in all three designs.

For example, the AP 600 is described by Westinghouse as "passively safe." Yet UCS found that the plant does not provide truly passive emergency core cooling; while reliance on some pumps is avoided, instrumentation must send the proper signals and valves must open for the cooling to work, just as in current plants. In addition, UCS said the AP 600 "appears vulnerable to a loss-of-coolant accident caused by misoperation of its automatic depressurization system."

The MHTGR was found to have several significant safety issues, including a potential vulnerability to graphite fires; possible rupture of steam generator tube ruptures on the reactor's core, which could result in a reactivity insertion and damage to the graphite core; and vulnerability to large reactivity insertions from control rod ejection.

The PRISM design received even more criticism, with UCS question-ing whether it can even achieve a level of safety equivalent to current reactors. UCS found the plant vulnerable to sodium fires, and more importantly, to "large reactivity excursions due to sodium boiling in scram failure events." The Chernobyl reactor also had a positive void coefficient, which resulted in a rapid and uncontrollable rise in power culminating in the explosion.

Neither the MHTGR nor the PRISM designs have containments, a fact which the study also criticized.

UCS said there would be "little basis" for licensing the current PRISM design without a containment, and noted that the NRC staff is leaning toward requiring a containment as well. However, the design could well be uneconomic if one were required.

All of the designs received criticism for not classifying all safety-related systems properly. Safety-related systems must meet a variety of regulatory requirements (environmental qualification, quality assurance programs, etc.) that other systems need not meet. Consequently, safety-related systems tend to be much more expensive than those classified as non-safety grade. UCS is concerned that in their efforts to reduce costs, the designers may be cutting corners.

In other areas UCS found that little data exists to determine whether the plans meet all fire protection issues and recommended that more attention be paid to possible seismic hazards.

A key safety issue for the PRISM and MHTGR was identified as operator staffing. Both designs project using far fewer operators than are now employed in reactors, and suggest that control rooms need not be classified as safety-related. UCS said there are "common sense reasons for rejecting the limited staffing levels." Of particular concern to UCS was that in some cases manufacturers suggest that one operator could operate more than one reactor at the same time.

GE SELLS UNTESTED REACTOR DESIGNS. Activists have new ammunition against General Electric, a leading defense contractor and the target of an international boycott. Washington Public Power Supply System (WPPSS) took GE to court for selling them untested reactor designs. Presiding Judge Alan McDonald determined that GE knew that certain problems with its containment design "should be examined but decided to adopt only an analytical approach, because the full-scale tests required to prove that the phenomena would not occur in its then present containment design would be, I am quoting from the documents 'very expensive." Although the decision is not legally binding (because the jury split, causing a mistrial), WPPSS has re-quested a new trial, and other utilities, including Pennsylvania Power & Light and Cleveland Electric Illuminating sent attorneys to the trial in preparation for possible lawsuits against GE. Long Island Lighting Co. has already filed a similar suit. Contact: Infact International, P0 Box 80013, Minneapolis MN 55408, US. Nuclear Monitor (incorp. Radwaste Report) (US), 8 Oct. 1990, pp.1-2