for the movement of heavy loads over spent fuel to assure that the potential for a handling accident that could result in damage of spent fuel is minimized while the generic evaluation proceeds. In addition, the licensees were requested to provide information on load handling operations for use in the Task A-36 review. Responses were received from all licensees by December 1978.

The staff has completed its survey of load handling operations at operating plants, including design and procedural measures that prevent or mitigate the consequences of a heavy load handling accident and has prepared a draft report containing the NRC staff's resolution of this issue including revised criteria and other recommendations. This report is expected to be issued for public comment in January 1980. The report will provide the basis for revisions to the Standard Review Plan (SRP) and Regulatory Guides, if needed, that can be used in future reviews of new plants and will provide the basis for implementing additional requirements and procedures, in operating plants.

Although Task A-36 will result in generic criteria, implementation of these criteria will be dependent on plant design characteristics and the specific procedures in effect at each particular plant, and will consequently require a plant-by-plant review.

Seismic Design Criteria

NRC regulations require that nuclear power plant structures, systems and components important to safety be designed to withstand the effects of natural phenomena such as earthquakes. Detailed requirements and guidance regarding the seismic design of nuclear plants is provided in the NRC regulations and in Regulatory Guides. However, there are a number of plants with construction permits and operating licenses issued before the NRC's current regulations and regulatory guidance were in place. For this reason, rereviews of the seismic design of various plants are being undertaken (principally as part of the Commission's Systematic Evaluation Program) to assure that these plants do not present an undue risk to the public.

The NRC staff is conducting Generic Task A-40 as part of the NRC Program for Resolution of Generic Issues. Task A-40 is a compendium of short-term efforts to support the reevaluation of the seismic design of operating reactors, and to support licensing activity in general. The objective of the task is, in part, to investigate selected areas of the seismic design sequence to determine the conservatism for all types of sites, to investigate alternative approaches to part of the design sequence, and to estimate quantitatively the overall conservatism of the design sequence. In this manner the program will aid the NRC staff in performing its reviews of the seismic design of operating reactors.

The NRC Office of Nuclear Regulatory Research is also undertaking a related, but more comprehensive and long-term program to develop mathematical models to realistically predict the probability of radiactive releases from seismically induced events in nuclear power plants. This Seismic Safety Margin Research Program will utilize input from Task A-40 in a number of areas.

Generic Task A-40 is subdivided into two phases. Phase I includes a number of subtasks related to the response of structures, systems, and components to earthquakes. These subtasks include studies on: (1) quantifying conservatisms in seismic design, (2) electro-plastic seismic analysis methods, (3) sitespecific response spectra, (4) nonlinear structural dynamic analysis procedures, and (5) soil structure interaction. These studies were performed under NRCsponsored contracts and all were completed by October 1979. Review of the results of these studies is underway. The results will support the effort on seismic reevaluation of operating plants, particularly in the area of site-specific definition of seismic input. As of January 1, 1980, Phase I was scheduled to be completed in February 1980, with the issuance of recommendations for changes in the Standard Review Plan and Regulatory Guides in those seismic design areas related to response of structures, systems, and components to seismically induced events.

Phase II of Task A-40 includes several subtasks related to numerical modeling of earthquake motion at the source, analysis of near source ground motion, and attenuation of high-frequency ground motion. Studies under these subtasks being conducted by NRC contractors are scheduled for completion by the end of 1980. Review and implementation of the results of these studies in terms of recommended revisions to the Standard Review Plan and Regulatory Guides are scheduled for March 1981.

Pipe Cracks at Boiling Water Reactors

Pipe cracking has occurred in the heat affected zones of welds in primary system piping in BWRs since the mid-1960s. These cracks have occurred mainly in Type 304 stainless steel that is being used in most operating BWRs. The major problem is recognized to be intergranular stress corrosion cracking (IGSCC) of austenitic stainless steel components that have been made susceptible to this failure mode by being "sensitized," either by welding or by post-weld heat treatment. Although the likelihood is extremely low that IGSCC-induced cracks will propagate far enough to create a significant hazard to the public, the occurrence of such cracks is undesirable and measures to minimize IGSCC in BWR piping systems are indicated to improve overall plant reliability.

"Safe ends" (short transition pieces between vessel nozzles and the piping) that have been highly sensitized by furnace heat treatment while attached to vessels during fabrication were found to be susceptible to IGSCC in the late 1960s. Because they were susceptible to cracking, the Atomic Energy Commission took the position in 1969 that furnace-sensitized safe ends in older plants should be removed or clad with a protective material, and there are only a few BWRs that still have furnace-sensitized safe ends in use. Most of these, moreover, are in small diameter lines and are subjected to augmented inservice inspection.

Earlier reported cracks (prior to 1975) occurred primarily in 4-inch diameter recirculation loop bypass lines and in 10-inch diameter core spray lines. More recently cracks were discovered in recirculation riser piping (12- to 14-inch) in all foreign plants. All these crack locations are part of the reactor primary system. Cracking is most often detected during inservice inspection using ultrasonic testing techniques. Some piping cracks have been discovered as a result of small primary coolant leaks.

In response to these occurrences of BWR primary system cracking, a number of remedial actions were undertaken by the NRC. These actions included:

• Issuance of Regulatory Guide 1.44 on "Control of the Use of Sensitized Stainless Steel."

• Issuance of Regulatory Guide 1.45 on "Reactor Coolant Pressure Boundary Leakage Detection Systems."

• Closely following the incidence of cracking in BWRs, including foreign experience.

Encouraging replacement of furnace sensitized safe ends.

• Requiring augmented inservice inspection of lines having less corrosion resistant stainless steel, especially those that have a high potential for cracking (service sensitive lines).

• Requiring upgrading of leakage detection systems.

More recently pipe cracking and furnace-sensitized safe end cracking have been reported in larger (24-inch diameter) lines in a GE-designed BWR in Germany with over 10 years of service. Because the safe ends in that facility had been furnace-sensitized during fabrication, IGSCC was suspected. As a result of concerns regarding these furnace-sensitized safe ends, a safe end was removed and subjected to destructive examination. During laboratory examination of the removed safe end, including a small section of attached pipe, cracks were discovered at various locations in the safe end and in the weld heat affected zone of the pipe. The cracks in the pipe weld area were very shallow with the maximum depth less than 5 mm (about 1/8-inch) in a wall thickness of about 1.5 inches. Cracking in the furnace-sensitized safe end, also having a wall thickness of about 1.5 inches, was

somewhat deeper. The German experience was the first known occurrence of IGSCC in pipes as large as 24 inches in diameter.

In June 1978, a through-wall crack was discovered in an Inconel recirculation riser safe end (10-inch diameter) at the Duane Arnold facility. The crack has been attributed to IGSCC, although the material in this instance is different from the Type 304 stainless steel that has been historically found to be susceptible to IGSCC. Prior to safe end removal, ultrasonic examination showed several indications of possible cracks. Following their removal, cracking was discovered in all eight safe ends. The cracking appeared to have originated in a tight crevice between the inside wall of the safe end and the internal thermal sleeve attachment. Such crevices are know to enhance IGSCC. Differences in materials, geometry, stress levels, and crevices appear to make the problem at Duane Arnold unique to a particular type of recirculation riser safe end (Type I). As a result of this event, ultrasonic examination of the other Type I safe ends in U.S. BWRs (i.e., at the Brunswick 1 and 2 facility) was conducted. No significant indications of possible cracks were found in Unit 2 and one indication was identified at Unit 1. Although this latter indication was relatively minor and too small to be reportable pursuant to the NRC Regulations, periodic reevaluation of the Unit was deemed necessary. This ultrasonic indication at Brunswick Unit 1 was remeasured and reevaluated in the presence of NRC ultrasonic testing consultants at another plant shutdown in January 1979. It was concluded that: (1) there is no apparent change of this indication between inspections, and (2) although the existence of a very small localized area of cracking cannot positively be ruled out, the most likely cause of this indication is irregularities at the weld-tobase metal interface of the first bead weld at the thermal sleeve to safe end weld. This indication will be reexamined during the next refueling outage.

General Electric (the reactor vendor) has been asked to provide an in-depth report on the significance of recent events, including current inspection, repair, and replacement programs. They were also asked to address any new safety concerns related to the occurrence of cracking in large main recirculation piping. Based on information presented by General Electric to date and on extensive staff evaluation, the staff concluded that the recent occurrences do not constitute a basis for immediate concern about plant safety, nor require any new immediate actions by licensees.

Based on the earlier incidents of pipe cracking discussed above, the NRC formed a Pipe Crack Study Group to: (a) investigate the cause of cracks, (b) make interim recommendations for operating plants, and (c) recommend corrective actions to be taken for future

plants. The Study Group published its report (NUREG-75/067) in October 1975, containing recommendations to reduce the incidence of IGSCC in sensitized stainless steel piping. Following staff review of the Study Group's recommendations, the staff issued an implementation document (NUREG-0313) which established staff positions consistent with the recommendations of the Study Group.

As a result of the more recent incidents, the NRC reestablished a second Pipe Crack Study Group on September 14, 1978. The new Study Group specifically addressed the following issues:

• The significance of the cracks discovered in large diameter pipes relative to the conclusions and recommendations set forth in the referenced report and its implementation document, NUREG-0313.

Resolution of concerns raised over the ability of ultrasonic techniques to detect cracks in austenitic stainless steel.

• The significance of the cracks found in large diameter sensitized safe ends, and any recommendations regarding the current NRC program for dealing with this matter.

• The potential for stress corrosion cracking in PWRS.

The significance of the safe end cracking at Duane Arnold relative to similar material and design aspects at other facilities.

The new Study Group completed its evaluation in February 1979 and issued a report, "Investigation and Evaluation of Stress-Corrosion Cracking in Piping of Light Water Reactor Plants" (NUREG-0531). The new Study Group not only reaffirmed the conclusions and recommendations reached by the previous group in NUREG-75/067, but also presented some new ideas to reduce the potential for IGSCC and addressed the subject of IGSCC in safe ends. On March 13, 1979, NRC issued a Notice in the Federal Register soliciting public comments on NUREG-0531. After expiration of the public comment period and review of the Study Group's conclusions and recommendations, the staff initiated Task A-42. The work to be performed under Task A-42 was defined at that time as the development of an update to the implementation document, NUREG-0313, to incorporate the new Study Group's conclusions and recommendations and public comments received on NUREG-0531.

Revision 1 to NUREG-0313 was issued in October 1979, and public comments have been solicited on the report. Revision 1 sets forth the NRC staff's revised guidelines for reducing the IGSCC susceptibility of BWR piping. The guidelines describe a number of preventive and corrective measures acceptable to the NRC, including guidelines for: (1) corrosion resistant materials for installation in BWR piping, (2) methods

of testing, (3) processing techniques, (4) augmented inservice inspection, and (5) leak detection. The report also included recommendations for developmental work to provide future improvements in limiting the extent of IGSCC or detecting it when it occurs.

Containment Emergency Sump Reliability

Following a postulated loss-of-coolant accident, i.e., a break in the reactor coolant system piping, the water flowing from the break would be collected in the emergency sump at the low point in the containment. This water would later be recirculated through the reactor system by the emergency core cooling pumps to maintain core cooling. This water would also be circulated through the containment spray system to remove heat and fission products from the containment. Loss of the ability to draw water from the emergency sump could therefore disable the emergency core cooling and containment spray systems. The consequences of the resulting inability to cool the reactor core or the containment atmosphere could be melting of the core and/or loss of integrity in the containment.

One potential way the ability to draw water from the emergency sump can be lost is from blockage by debris. A principal source of such debris could be the thermal insulation normally installed on the reactor coolant system piping. In the event of a piping break, the subsequent violent release of the high pressure water in the reactor coolant system could rip off the insulation in the area of the break. The loose insulation material could then be swept into the sump and block it.

A Task Action Plan was under development in March 1979 when the Three Mile Island Unit 2 accident disrupted work on it. As of January 1, 1980, the Task Action Plan was nearing completion. Nonetheless, several technical studies related to sump reliability which were already underway will either be incorporated into Task A-43 or will provide input into Task A-43 efforts.

A study program investigating PWR vortex technology has been completed by the Iowa Institute of Hydraulic Research and a technical report issued. A summary report of NRC experience with containment sump testing is being prepared. This summary will be issued as a NUREG report in 1980. Based on the Iowa study program and the review of tests, NRR expects to draft interim positions on sump design guidelines and preoperational test requirements in early 1980. Criteria for the evaluation of operating containment sumps will be formulated at about the same time.

Finally, a program is being sponsored by the Department of Energy, in cooperation with NRC, to aid in resolving this issue as part of their Light Water Safety Research Program. This is an experimental pro

gram, begun in July 1979 at Alden Research Laboratory, Worcester Polytechnic Institute, to study the hydraulic aspects of containment sump operation. The program will continue through February 1981. It is anticipated that this task can be completed in 1982.

Station Blackout

Electrical power for safety systems at nuclear power plants is supplied by at least two redundant and independent divisions. The systems used to remove decay heat to cool the reactor core following a reactor shutdown are included among the safety systems that must meet these electric power supply requirements. Each electrical division for safety systems includes an offsite alternating current (a.c.) power connection, a standby emergency a.c. power supply (usually one or more diesel generators), and direct current (d.c.)

sources.

The issue of station blackout involves a study of whether or not nuclear power plants should be designed to accommodate a complete loss of all a.c. power, i.e., a loss of offsite a.c. sources and all onsite emergency diesel generator sources. Loss of all a.c. for an extended period of time in pressurized water reactors, accompanied by loss of the auxiliary feedwater pumps (usually one of two redundant pumps is a steam turbine driven pump that is not dependent on a.c. power for actuation or operation), could result in an inability to cool the reactor core, with potentially serious consequences. If the auxiliary feedwater pumps are dependent on a.c. power to function, then a loss of all a.c. power for an extended period could of itself result in an inability to cool the reactor core. Although this is a low probability event sequence, it could be a significant contributor to risk.

Current NRC safety requirements require as a minimum that diverse power drives be provided for the redundant auxiliary feedwater pumps. As noted above, this is normally accomplished by utilizing one or more a.c. power electric motor driven pumps and one or more redundant steam turbine driven pumps. One concern is the design adequacy of plants licensed prior to adoption of the current requirements.

The degree of dependence of decay heat removal systems on a.c. power supplies and their reliability with a total loss of a.c. power has recently been reviewed for a large number of plants. For some plants, modification in design and/or operating procedures were recommended in the short term. This evaluation was carried out using simplified analytical techniques.

A Task Action Plan for Task A-44 was under development in March of 1979 when the Three Mile Island Unit 2 accident disrupted work on this task. As of January 1, 1980, the Task Action Plan was still

under development. It is anticipated that the task can be completed in 1982.

Under Task A-44 a more detailed and comprehensive assessment will be performed for both PWRs and BWRs. Preliminary scoping work indicates that this should include consideration of: (1) the failure modes that can result in a station blackout, (2) the probability and frequency of occurence of a blackout including site variability and time dependence, (3) the potential consequences of a blackout, and (4) potential preventative and mitigating actions. The results of this effort will be used to determine if changes to licensing criteria are necessary and, if so, to develop criteria for use in the review of CP and OL applications and for evaluating operating plants.

OTHER TECHNICAL ISSUES

Design Errors in Control Building

In the spring of 1978, Portland General Electric Company (PGE), operator of the Trojan Nuclear Plant, reported design errors in the control building walls, i.e., conditions at variance with the design criteria set forth in the Final Safety Analysis Report for the facility and incorporated into its operating license. A detailed NRC staff review of PGE's investigation and analysis of the design revealed the following er

rors:

• The steel reinforcement in the reinforced concrete core of the walls was permitted to be generally discontinuous and, therefore, the concrete core could not be relied upon to resist shear (in case of an earthquake) to the extent assumed in the approved design.

• The shear capacity of the reinforced concrete and grouted masonry block was not correctly computed.

The steel reinforcement needed to resist shear beyond the capacity of the concrete and grouted masonry block was computed incorrectly, resulting in a lower level of conservatism than intended.

A detailed reevaluation of the control building in its existing configuration was performed by PGE to assess the capability of the structure to withstand the Operating Basis Earthquake and the Safe Shutdown Earthquake. The NRC staff determined that there had been a significant reduction in conservatism and design margins, with respect to the control building seismic capability, below the level intended and desired for the 33 years remaining in the expected plant life and that the margins should be appropriately restored by plant modification. PGE indicated its intent to make such modifications.