studies conducted in other upstream and downstream areas of the river. Thermal and chemical discharges during and following the accident did not exceed the effluent limitation established to protect the aquatic environment. Although several million gallons of treated industrial waste effluents were released into the river, these discharges were not of unusual volume compared with normal operation and were a very small portion of the seasonally high spring river flows. The extent and relative location of the effluent plume were defined and the fish species known to have been under its immediate influence were identified-including rough, forage, and predator/sport fishery species. Impacts to benthic invertebrates or fishes were not detected. No unusual conditions of fish disease or mortality were noted in the river following the accident. The normal spring increases in abundance and species-composition of riverine fauna occurred, as did the onset of the fish spawning season in April with peaks of ichthyoplankton abundance in May and June.

Nevertheless, post-accident recreational fishing in the Three Mile Island vicinity underwent significant departures from historical trends. Fishing activity appeared to shift away from the Susquehanna River waters near the nuclear station to other areas, especially downstream. Anglers returned greater proportions of their catches than during any comparable period within the previous five years. This was most

notable during April when anglers fishing near the plant returned an unprecedented 100 percent of their catches. Thus, in the waters receiving station effluents during the month following the accident, the liquid radiological pathway leading to man via fin fish consumption could have been absent entirely. With the passage of time following the accident, the normal pattern of recreational fishing was approached. The investigation defined several generic aspects of the accident and lessons applicable to other facilities: the time of the accident with respect to the biological season, and to the ability to detect an impact; data availability and data needs for adequate monitoring; and the application of the non-radiological findings for radiological assessment. This investigation is described in an NRC report: "The Non-Radiological Consequences to the Aquatic Biota and Fisheries of the Susquehanna River from the 1979 Accident at Three Mile Island Nuclear Station" (NUREG-0596).

TMI RECOVERY OPERATIONS

Following the accident of March 28, a substantial effort was mounted to provide technical assistance, regulatory guidance and review of the licensee's operations procedures and system addition and modification activities. A team began to form with the arrival of the Office of Inspection and Enforcement Region I inspectors shortly after the accident and continued to expand with the arrival of the first contingent from the Office

of Nuclear Reactor Regulation (NRR) on March 29 and additional inspectors from all five regional offices. On March 30, the Director of NRR and additional NRR staff arrived on the site to assist in the recovery operation. A Public Affairs Office was also established in Middletown, PA, and staffed on a 24-hour-per-day basis to handle the flow of information to the public and the media.

NRR staff analysts in many of the major disciplines were brought to TMI to provide needed technical resources. The specific activities engaged in by the staff can be broken down into four major areas:

(1) A review was initiated of the system modifications and system additions (proposed by the licensee, the industry review group, or the NRC) as contingency measures to mitigate the consequences of the accident and to provide assurance for continued safe shutdown and long-term safe shutdown.

(2) Substantial effort was given to the review of all procedures, both emergency and normal operation and maintenance, which were necessary to post-accident activities. In many cases, because of changes in the use of normal systems and the addition of new ones, new operating procedures were necessary. Further, the facility license and technical specifications, which defined the limits for operating parameters and surveillance requirements, were no longer fully applicable to the post-accident facility, though existing facility procedures provided a mechanism for establishing specific operability limits and surveillance requirements. It was necessary, from a regulatory point of view, to have NRC's review and approval of any new procedures that might be in conflict with the pre-accident license.

(3) NRR provided close and continuous monitoring of the operations in progress to assure that system parameters stayed within expected limits and to provide prediction of future system performance and the capability of plant systems to maintain safe conditions.

(4) Lastly, substantial NRR effort was committed to providing consultation, review and analysis of the ongoing radwaste, cleanup, and health physics activities. The accident generated a significant amount of contaminated water which, in turn, contaminated substantial portions of the facility and its systems. This made it difficult to have normal access to systems important to safety and also constituted a threat of further fission product release and occupational exposure. In addition, the radiological makeup of the contamination was different from that normally encountered in operating reactors, in terms of its airborne intensity as well as its ratio

of beta and gamma activity. It was therefore an important concern-particularly in view of the intensive work activity needed to continue safe operation-that operator exposures be maintained within acceptable limits and the environment protected from undue radiological effluents.

Examples of the system review activity undertaken by the NRR on-site staff were design reviews and evaluations of the following sytems:

(a) Supplementary diesel generators (b) Supplementary filtration systems (c) Long-term cooling systems

(d) Alternative decay heat removal system (e) Pressure volume control system

(f) Tank farm for storage of radioactive liquids (g) EPICOR-II system for processing of contaminated liquids

(h) Many monitor modifications in existing systems which allowed operability in the post-accident environment.

Besides the systems reviews, approximately 250 procedures were reviewed and approved by the on-site staff. This activity was particularly important in the first two months following the accident because a serious shortage of personnel familiar with the facility existed; the NRC review constituted not only a regulatory approval of the intended operation, but also served as a quality assurance check on adequacy and operability. The review of procedures is continuing as the licensee rewrites emergency and operating procedures to reflect the changing status of the facility. It is anticipated that such procedure review will be necessary until a new set of facility technical specifications, which reflect the post-accident facility configuration, is implemented.

A substantial amount of staff effort was expended on the review and approval of the EPICOR-II operation, intended for use in decontaminating the 380,000 gallons of intermediate-level contaminated water held in the auxiliary building tanks and in the tank farm constructed following the accident. EPICOR-II was designed and constructed following the accident because it was clear that storage of water would be a significant problem and could not be accommodated with existing facility equipment. EPICOR-II is a three-stage demineralization system, constructed in an existing on-site building. EPICOR-II was provided with sufficient shielding and remote-handling capability to accommodate the processing of radioactive water up to a level of about 100-microcuries-permilliliter. When facility operation was near, court actions were initiated to prevent operation of EPICORII or disposal of the decontaminated water. In response to these actions, the Commission directed that an environmental assessment for the use of

The "EPICOR-II" system being used to decontaminate some 380,000 gallons of intermediate-level radioactive water held in the auxiliary building tanks at the TMI-2 site is shown above. It consists of three process vessels (steel liners) shielded by four-inch lead enclosures located in the chemical cleaning building. Each vessel contains ion-exchange resin. The vessel at the top of the photo at the left is the system prefilter/demineralizer, the center vessel is a cation ion-exchanger, and the third vessel is a mixed-bed polishing ion-exchanger. Each is fitted with three quick-disconnect hoses: a liquid waste influent line, a processed waste effluent line, and a vent line with attached overflow hose. Vented air from each vessel

passes through a special filter and charcoal adsorber. "Spent" ionexchange resin liners containing radioactive material removed from the water are transferred by crane to cells (shown at top right) which are housed in modular concrete storage structures (above). The cells are concrete-shielded, galvanized corrugated steel cylinders seven feet in diameter and 13 feet high. The storage module shown under construction has 4-foot thick walls and will be 57 feet wide and 91 feet long. The modules, each holding about 60 storage cells, will be built on an as-needed basis. Shipment of the radioactive liners away from the site will depend on approval of a disposal facility and availability of shipping casks.

EPICOR-II be prepared, followed by the environmental assessment for the alternatives of disposal of decontaminated water. Both of these environmental assessments would be provided to the public for comment before any actions would be initiated. Environmental assessment for the use of EPICOR-II in the decontamination of the intermediate level of contaminated water in the auxiliary building was prepared and sent out for public comment on August 14, 1979. The assessment evaluated various alternatives to the proposed cleanup and concluded that the use of the already constructed system was the best alternative, and that the processing of water would constitute a negligible environmental impact.

Based on these evaluations, the Commission, on October 16, 1979, issued a Memorandum and Order directing the use of EPICOR-II.

BULLETINS AND ORDERS
TASK FORCE

The accident at TMI-2 involved a feedwater transient coupled with a small break in the reactor system (the open relief valve). Because of the severity of the ensuing events and the potential generic implications of the accident for other operating reactors, the NRC staff initiated prompt action to: (1) assure that other reactor licensees, particularly those with plants similar in design to TMI-2, took the necessary action to substantially reduce the likelihood for TMI-2 type events, and (2) start comprehensive investigations into the potential generic implications of this accident on other operating reactors.

The Bulletins & Orders Task Force was established within the Office of Nuclear Reactor Regulation (NRR) in early May 1979. This task force was responsi

ble for reviewing and directing the TMI-2 related staff activities regarding loss-of-feedwater transients and small break loss-of-coolant accidents for all operating reactors. The task force concentrated its efforts in the areas of: assessments of auxiliary feedwater system reliability; review of the analytical predictions of plant performance for both feedwater and small LOCA-induced transients; evaluations of generic operating guidelines; the review of emergency plant operating procedures; and the review of operator training.

The task force worked with operating plant licenses, and, for the review of generic items, with owners' groups for plants of each nuclear steam supply vendor (Babcock and Wilcox, Westinghouse, Combustion Engineering, and General Electric) and with the indilvidual vendors. Initial priority was placed on plants of the Babcock and Wilcox (B&W) design, but as short-term actions on these plants were completed, priority was shifted to other pressurized water reactor (PWR) plants, i.e., those manufactured by Westinghouse and Combustion Engineering. Activities related to boiling water reactors, a significantly different light water reactor type manufactured by the General Electric Company, were pursued as a third priority.

The task force, which was composed of approximately 30 technical professionals of widely varying disciplines and areas of expertise, evaluated licensees' responses to NRC Bulletins; the issuance and subsequent lifting of Orders to the B&W operating reactors; system reliability and predicted plant performance for each of the reactor vendors, with regard to feedwater transients and small break loss-of-coolant accidents; and related follow-on activities.

Bulletins

The preliminary review of the accident chronology identified several events that occurred during the accident and contributed significantly to its severity. As a result, all holders of operating licenses were subsequently instructed to take a number of immediate actions to avoid repetition of these errors. These instructions were specified in a series of bulletins issued by the NRC's Office of Inspection and Enforcement (IE).

The initial bulletins defined actions by operating plants using the B&W reactor system, but as staff evaluations determined that additional actions were necessary, these bulletins were expanded, clarified, and issued to all operating plants for action. For example, as a result of staff evaluations, holders of operating licenses for B&W designed reactors were instructed by IE Bulletins to take further actions, including immediate changes to decrease the reactor high pressure trip point and increase the pressurizer pilot-operated relief valve settings. A chronology of bulletins issued by IE is shown below.

The task force directed the evaluations of each licensee's response to the IE Bulletins. This process involved an inter-office review group, which included representatives from IE and from the NRR Division of Operating Reactors. When it was concluded that a licensee had understood and had provided an acceptable response to the bulletins, the bulletin review was completed and the evaluation issued as a staff report.

The prompt action taken by licensees in response to the IE Bulletins was considered an important contributor to the assurance of continued safe plant operation. The bulletins and related evaluations also provided substantial input to other staff activities, such as those associated with the generic study efforts and the Lessons Learned Task Force (see below). Thus, many of the subjects addressed by the bulletins were studied in greater depth through other staff activities and studies. Further, the bulletins and the associated responses were used as a basis for IE inspection activities and for auditing of reactor operator training.

Orders on Babcock and Wilcox Plants

Soon after the TMI-2 accident, the NRC staff began a reevaluation of the design features of B&W reactors to determine whether additional safety corrections or improvements were necessary. This evaluation involved numerous meetings with the vendor and the affected licensees.

The conclusion of these preliminary staff studies was documented in an April 25, 1979 status report to the Commission. It was found that B&W designed reactors appeared to be unusually sensitive to certain transient conditions originating in the secondary system. The features of the B&W plants that contributed to this sensitivity were: (1) design of the steam generators which operate with relatively small liquid volumes in the secondary side; (2) lack of direct initiation of reactor trip upon the occurrence of off-normal conditions in the feedwater system; (3) reliance on an integrated control system (ICS) to automatically regulate feedwater flow; (4) actuation before reactor trip of a pilotoperated relief valve on the primary system pressurizer (which, if the valve sticks open, can aggravate the event); and (5) a low steam generator elevation (relative to the reactor vessel) which provides a smaller driving head for natural circulation (except for the Davis-Besse plant).

Because of these features, B&W design relies more than other PWR designs on the reliability and performance characteristics of the auxiliary feedwater system, the integrated control system, and the emergency core cooling system (ECCS) performance to recover from certain anticipated transients, such as loss of off-site power and loss of normal feedwater. This, in turn, can require greater operator knowledge and skill to safely manage the plant controls during

such anticipated transients. As a result of the work supporting the April 25, 1979 report, the NRC staff concluded that certain other short-term design and procedural changes at operating B&W facilities were necessary in order to assure adequate protection to public health and safety.

After a series of discussions between the NRC staff and licensees of operating B&W plants, the licensees agreed to shut down these plants and keep them shut down until the actions identified to the Commission in the April 25, 1979 report could be completed. This agreement was confirmed by a Commission Order to each licensee (see "Actions Directed by Orders," below). Authorization to resume operation was issued in the period late May through early July, as individual plants satisfactorily completed the short-term actions and the NRC staff completed an on-site verification of the plant's readiness to resume operation. In addition to the modifications to be implemented promptly, each licensee also proposed to carry out certain additional long-term modifications to further enhance the capability and reliability of the plant systems to respond to transient events (see "Longer Term Actions," below).

Since some of the long-term modifications involve the design, procurement, and qualification of safetygrade hardware, not all of the actions of the long-term portion of the Orders were completed in 1979. Staff involvement will continue to assure that licensees complete each long-term action of the Order "as promptly as practicable" and that the Orders are closed out by a prompt staff acceptance review.

Specific Plant and Generic Studies

For B&W operating reactors, an initial staff study has been completed and published in a staff report (NUREG-0560). This study considered the particular design features and operational history of B&W operating plants in light of the TMI-2 accident and related current licensing requirements. As a result of this study, a number of findings and recommendations resulted which are now being pursued.

Generally, the activities involving the B&W reactors are reflected in the actions specified in the Orders. Consequently, as noted earlier, a number of specific actions have been specified in the areas of transient and small break analyses, upgrading of auxiliary feedwater reliability and performance, procedures for operator action, and operator training.

Similar studies are now well underway for the Westinghouse and Combustion Engineering operating

plants. These studies focus specifically on the predicted plant performance under different accident scenarios involving small break loss-of-coolant event and feedwater transients. Based upon analytically predicted system behavior, recommended guidelines for emergency operating procedures were developed and reviewed in the study. In addition, these studies include engineering assessments of the reliability of individual plant auxiliary feedwater systems and identification of dominant failure contributors and recommendations for corrective action. A similar study of the operating boiling water reactors is also in progress, but is at an earlier stage.

As the above studies developed firm conclusions and recommendations, implementing action was initiated. For example, the results of the Westinghouse and Combustion Engineering auxiliary feedwater system reliability assessments concluded that certain improvements were necessary. Individual plant licensees were then requested by letter to initiate corrective action or to propose design solutions for NRC staff review. Additional instructions were to be issued to licensees upon completion of other aspects of these reports.

Follow-On and Interfacing Activities

It was planned that the task force would terminate its activities in late 1979, and therefore some of its activities were transferred prior to completion. Consequently, the task force concentrated on lead plants and established review guidelines and acceptance criteria that could be implemented by other NRR organizational elements.

As a result of the work performed in modeling small break and feedwater transients, longer range efforts were identified dealing with the procedures and systems available for core cooling under certain accident conditions, and with confirming analytical models through experiment or research programs. For example, plans are being implemented to conduct some small break loss-of-coolant tests at the Semiscale and LOFT facilities to obtain a better understanding of small break phenomena and to use the results to verify calculational techniques (see Chapter 11). Other recommendations in this regard are expected to result from the task force activities.

As noted previously, the task force concentrated on the immediate and near term actions necessary to assure the safe operation of operating plants. However, based on actions already completed. a number of items have been identified which warrant careful additional study. These actions have been and are continuing to be, documented for detailed assessment within the NRR organization.