SIERRA CLUB OF CANADA
September 24, 2003
BRUCE NUCLEAR BACKGROUNDER
BRUCE POWER -- PRIVATE SECTOR LESSEE
BRUCE A RESTART
FUEL CHANNEL PROBLEMS
NUCLEAR SAFETY PROBLEMS AT BRUCE
INFORMATION BEING WITHHELD
- Bruce Power is seeking 5-year licences from the Canadian Nuclear Safety Commission (CNSC) for the Bruce A and Bruce B nuclear stations, which each have four reactors. These would be their second licences the first were 2-and-a-half years long.
- Reactor #2 at the Bruce A nuclear station was shut down in October 1995, and reactors 1, 3 and 4 were shut down in March 1998, because of technical problems, poor performance, and management deficiencies.
- On April 6, 2001, Bruce Power announced that it intends to re-start Reactors 3 and 4 at the Bruce A Station. Bruce 3 began operation in January 1978 (26 year old) and Bruce 4 began operation in January 1979 (25 years old). If re-started, they will have the oldest operating fuel channels of any reactors in Canada.
- The Bruce B reactors have operated since they began commercial operation in 1984-87.
- The Bruce B reactors are each 860 megawatts net, and the Bruce A reactors 3 and 4 are each 769 megawatts net.
BRUCE POWER PRIVATE SECTOR LESSEE
In May 2001, Ontario Power Generation closed a deal with Bruce Power (then a subsidiary of British Energy, a nuclear power plant operator in the United Kingdom) for an eighteen-year lease to operate the Bruce nuclear complex on the shore of Lake Huron in Ontario. The details of the agreement were kept secret, but it was clearly a sweetheart deal for Bruce Power. Among other things, Bruce Power has no long-term responsibility for radioactive waste management and decommissioning, responsibility for which remained with Ontario Power Generation (OPG), and ultimately with OPGs sole shareholder, the Government of Ontario. The agreement was a cash cow for British Energy, earning $120 million profit in its first year of operation.
The economic failure of nuclear power was been highlighted by the effective bankruptcy of British Energy, originally the majority shareholder of Bruce Power. Despite the advantage of an extremely generous privatization deal that saw its creation in 1996, in May 2002 British Energy posted a massive loss of 527 million pounds (about $1.2 billion CDN) for the financial year ending March 2002. Following the British Energy announcement of possible collapse on September 4, 2002, the Canadian Nuclear Safety Commission (CNSC) expressed concern about a $264 million (CDN) financial guarantee required as a licence condition for six months of operating costs in the event of an emergency shutdown. Despite the admission of Bruce Power CEO Duncan Hawthorne that the company could not deliver on the financial guarantee if required (a clear violation of an operating licence condition), the CNSC allowed the Bruce B reactors to continue operating. In December 2002, the British government agreed to accept financial responsibility for the guarantee if funds were required.
On December 24, 2002, a new Canadian consortium announced the purchase of Bruce Power from British Energy. The group consists of Cameco Corporation, TransCanada Pipelines Ltd., and BPC Generation Income Trust, part of the OMERS pension fund. The three major partners will each hold 31.6%, while the Power Workers Union will have 4% and the Society of Energy Professionals will have 1.2%.
BRUCE A RESTART
On April 6, 2001, Bruce Power announced that it intended to restart reactors 3 and 4 at the Bruce A station. At that time, Bruce Power expected that the reactors would be restarted early in the summer of 2003 at a total cost of about $340 million (CDN).
The Canadian Nuclear Safety Commission (CNSC) conducted a screening level environmental assessment on the restart of the Bruce A reactors 3 and 4. This was a low-level environmental assessment, and the CNSC ignored public requests to ask the federal Environment Minister to appoint an independent assessment panel. As the Responsible Authority for federal nuclear matters, CNSC is in charge of all lower level environmental assessments (Screenings and Comprehensive Studies) unless it refers an assessment to the Minister of Environment for a hearing by an independent panel. A panel is more independent, since its members would be specially appointed by the Minister, and it would provide funding for intervenors.
On April 12, 2002, the CNSC approved the Environmental Assessment Guidelines for the screening assessment. Bruce Power issued its Environmental Study Assessment Report in August 2002, and CNSC released its screening report in October 2002. On December 12, 2002, the CNSC held a hearing on the environmental assessment screening report for the return to service of Units 3 & 4 of the Bruce Nuclear Generating Station (NGS) A. The Board ignored the concerns of environmental groups, and decided that the Screening Report had met all of the requirements of the Canadian Environmental Assessment Act.
By the fall of 2002, the estimated refurbishment cost has escalated to $400 million, and the restart schedule was speeded up to have reactor 4 (the first to go) restarting in April 2003, and reactor 3 shortly afterwards.
That schedule has slipped significantly. It was not until August 30, 2003 that the reactor was restarted (became critical). On September 17, 2003, the CNSC allowed Bruce Power to increase the reactor power to 10%. The restart of reactor 3 is about 6 weeks behind reactor 4. In June 2003, Bruce Power said that it had spent about $486 million on Bruce A capital improvements.
FUEL CHANNEL PROBLEMS
In November 2000, Bruce Power had announced that AECL had been selected as the general contractor to lead an internal inspection and condition assessment of 70 fuel channels as well as steam generators for Bruce A reactors 3 and 4. There are 480 fuel channels in each Bruce A reactor, so AECL is making a safety judgement based on an examination of only 7% of the 960 channels in reactors 3 and 4. The assessment cost $30 million and was intended to determine if the re-commissioning of the reactors was economically justified. With the guaranteed profit of its lease agreement with OPG, there was little doubt that the restart project would proceed.
After the four Pickering A reactors, the Bruce A nuclear station has the oldest commercial reactors in Canada. The 480 fuel channels in each Bruce reactor core are prone to age-related problems due to the weight of the fuel bundles, as well as high temperatures, pressures and radiation fields in the reactor cores. Fuel channels in CANDU reactors consist of an outer calandria tube, and an inner pressure tube. The inner pressure tube holds uranium fuel bundles, and heavy water coolant is pumped through at high pressure to draw off the heat released by the fission process. Pressure tube problems include creep and sag, where the metal thins out over time and the tubes become wider and longer, bending under the strain. Various design changes were made in later stations to try to accommodate this problem, but eventual tube replacement (retubing) is anticipated on a schedule dictated by the extent of the problem in each reactor core.
When the pressure tubes sag they can come into contact with the outer calandria tube. This increases the chance of pressure tube rupture caused by embrittlement, where the metal becomes brittle due to absorption of hydrogen. This metal hydriding process happens faster where the sagging pressure tubes make contact with the cooler calandria tubes. The space or annulus between the calandria and pressure tubes is maintained by spacers or garter springs. However, at Bruce reactors 3, 4, 5, 6 and 7, the garter springs are not locked into place and have to be periodically checked and moved back into position to keep the two tubes from touching.
Bruce reactor 5 476 of 480 fuel channels (FCs) have loose fitting spacers (4 FCs have tight-fitting spacers installed in 1983) Bruce reactor 6 467 of 480 fuel channels have loose fitting spacers (11 FCs have tight-fitting spacers installed in 1983. 2 FCs have tight-fitting spacers installed in 1995 and 2002 respectively) Bruce reactor 7 478 of 480 fuel channels have loose fitting spacers (2 FCs have tight-fitting spacers installed in 1983)
The Channel Inspection and Gauging Apparatus for Reactor tool (CIGAR) and Spacer Location and Repositioning tool (SLAR) are used to inspect for contact between pressure tubes and calandria tubes. When contact is observed or predicted to occur before End of Design Life, the SLAR tool is used to move the spacers to new positions to eliminate contact. A joint report from Bruce Power and OPG on tight-fitting spacers is not expected until late 2003.
The CANDU industry has argued that pressure tubes will always leak before rupturing, allowing time to shut the reactor down before a loss of coolant accident occurs an assumption they call leak before break. However, there have been at least two cases of catastrophic pressure tube ruptures in Ontario reactors: August 1983 at Pickering 2 and March 1986 at Bruce 2. All fuel channels at the Pickering A station reactors were subsequently replaced.
Some individual tubes at Bruce reactors have been replaced in the past, but Bruce Power is taking a calculated risk, trading off safety against profit by arguing that complete replacement of fuel channels is not necessary at Bruce reactors 3 and 4. Bruce Power has taken this controversial position despite having inspected only 7% of tubes. Complete retubing of the reactors would likely double current restart costs as well as extending the outage time.
NUCLEAR SAFETY PROBLEMS AT BRUCE
Since the Bruce Power took over operation of the Bruce nuclear stations, beginning in the second quarter of 2001, there have been 218 reportable events at the Bruce A station (to the end of the second quarter in 2003), despite the fact that there were no reactors operating. From the beginning of the second quarter of 2001 to the end of the first quarter of 2003, there have been 397 reportable events at the Bruce B nuclear station. Reportable events are the more serious safety- related events at nuclear plants. Some of these events stood out dramatically as incidents of serious safety concern. There are also a number of unresolved safety concerns that merit special attention. We have identified at least 20 incidents and safety concerns that are identified below. They are not noted in any order of priority.
On June 11 2002, during a maintenance outage at Bruce B reactor 6, a Spacer Location and Relocation (SLAR) tool malfunctioned causing a high electrical current arc which burned a hole in both the pressure tube and calandria tube it was working on. This allowed heavy water coolant to mix with heavy water moderator that was in a poisoned state to guarantee shutdown. Bruce Power lied about the incident, saying only that a pressure tube had been slightly damaged and the operational impact is not expected to be significant. The incident was kept secret on the basis that public knowledge of the shutdown of the reactor would have commercial implications for British Energy. Because of concerns about nuclear safety and the publics right-to-know, secrecy about nuclear shutdowns has prompted public protest (See: Secrecy at Bruce, Toronto Star Editorial, September 27, 2002, p. A26). The damaged tubes had to be replaced, which shut down the reactor until early September.
In December 2002, after a planned outage for reactor 7 at Bruce B, the start up instrumentation (SUI) did not respond to removal of moderator poison. In other words the operators were operating blindly with respect to the neutron flux in the reactor core, which was increasing as the poison was removed. The reactor was returned to guaranteed shutdown. It was thought that the SUI had not been operating since November 30, 2002.This impaired potential operation of the emergency Shutdown System 1, one of two emergency shutdown systems.
On February 28, 2003, it was discovered that a number of check valves in the Emergency Core Cooling System (ECCS) of the Bruce A station were not operating properly. The ECCS provides cooling water to prevent a meltdown in the event of a loss of coolant accident (LOCA). The faulty valves could have impaired the ECCS in the event of a serious accident. Although Bruce A was not operating at the time, it was discovered that 3 of 6 similar check valves at the operating Bruce B station were also faulty.
Regarding the Sustained Loss of All Heat Sinks, the regulator is concerned that the heat transport (cooling) system could over-pressurize due to inadequate relief capacity. This could cause a loss of containment, resulting in a major radiation release. Bruce Power was required to respond by the end of July 2003, but did not respond until August 22, 2003. Bruce Power continues to argue that such events are of such low probability that they are not of concern. The CNSC says that staff has not yet reviewed the Bruce Power response.
The CNSC has a list of Generic Action Items that are major safety issues of concern applying to more than one reactor and/or nuclear station. For the Bruce A station the list includes:
GAI 88G02 Hydrogen Behaviour in CANDU Nuclear Generating Stations
GAI 91G01 Post-Accident Filter Effectiveness
GAI 94G01 Best Effort Analysis of ECCS Effectiveness
GAI 95G01 Molten fuel / Moderator Interaction
GAI 95G02 Pressure Tube Failure with Consequential Loss of Moderator
GAI 95G04 Positive Void Reactivity Uncertainty - Treatment in Large Loss LOCA Analysis
GAI 95G05 Moderator Temperature Predictions
GAI 98G01 PHT Pump Operation Under Two-phase Flow Conditions
GAI 99G02 Replacement of Reactor Physics Computer Codes Used in Safety Analysis of CANDU Reactors
GAI 00G01 Channel Voiding During a Large LOCA
GAI 01G01 Fuel Management and Surveillance Software Upgrade
The list is the same for Bruce B, except that for Bruce B, GAI 98G01 is closed. The CNSC indicates that Bruce Power has taken some partial action on only 6 of the 11 GAIs at the Bruce stations. However, in the absence of closure these serious safety questions remain unresolved.
Increased pick-up rate of deuterium has been recognized as a problem, particularly in Bruce B reactor 6. However, only 12 pressure tubes have been tested. Only 4 of these had previously been tested, allowing life-time uptake rates to be determined. In one tube location, the pick-up rate was 1.4 times higher than the average. Hydrogen absorption, or hydriding leads to embrittlement of the metal in the tube, which can lead to catastrophic ruptures.
In 2001, the separator loop of the Emergency Core Cooling System (ECCS) of reactor 8 in Bruce B was below a minimum level redefined in 2003. This revised minimum level is designed to prevent a water hammer from occurring during a main steam line break accident. A water hammer could disable the ECCS, which provides cooling water to prevent a meltdown in the event of a loss of coolant accident (LOCA). [response #25]
In a Large Loss of Coolant Accident (LLOCA), power in the reactor increases rapidly as a void is created in the absence of heavy water coolant (CMD 03-H27, p. 15). In 2001, new computer analysis indicated that increases in power levels had previously been underestimated. Although this affects all CANDU reactors, the Bruce A and B reactors were most affected because of their unique heat transport systems. The CNSC has described this as an unresolved issue, and has indicated that analysis by Bruce Power is still under review (CMD 03-H27.A, pp. 2-3).
Shutdown System 1 and 2 (SDS 1 & SDS2) Neutron Overpower (NOP) System are designed to prevent local power transients in the reactor core by shutting down the reactor through neutron flux detectors (CMD 03-H27, p. 16). The problem is that the SDS2 design came while the Bruce A station was already under construction, and fewer detectors were used. The CNSC has described this as an unresolved issue, and says
there are certain credible but unlikely failures of reactivity control devices wich could not be fully covered by SDS2 NOP system
(CMD 03-H27.A, p. 3).
CNSC has refused to provide the Sierra Club with the executive summary of the Bruce A Probabilistic Risk Assessment (BAPRA) (CMD 03-H27, p. 26). The complete study was only received by the CNSC on June 27, 2003, and CNSC staff has not completed its preliminary review (estimated to take 6 months ) or its detailed review (estimated to take one year).
On August 14, 2003, following loss of power in the great blackout, Bruce B reactor 6 was unable to recover, due to an adjuster rod jamming. The main function of adjuster rods is to produce a more uniform flux in the reactor, and they are positioned particularly in order to depress neutron flux in the central region of the reactor core. They are sometimes used when fuelling machines are unavailable (i.e. to compensate for lower reactivity from older fuel). However, the adjuster rods are also occasionally used to override xenon poisoning which also causes the reactors to shut down. In the event of unanticipated shutdowns (forced outages), CANDU reactor operators attempt to put the reactor in a poison prevent condition. The reactors have neutron absorbers called adjuster rods that stay in the reactor core during normal operation, and can be withdrawn to increase reactivity. Careful removal of the adjuster rods provides enough excess reactivity to allow the reactor to be restarted within thirty minutes of shutdown, and this is known as poison prevent. If the reactor operators do not achieve poison prevent within thirty minutes, the reactors take at least 40 hours to restart because of the half-life of the Xenon in the reactor fuel. Removal of the adjuster rods is a delicate operation, since it affects the neutron flux of the reactor, and a mistake lead to a transient which could result in loss of control of the reactor.
As part of its first licence in 2001, the Canadian Nuclear Safety Commission (CNSC) required that a financial guarantee in the amount of $264 million be provided to cover the cost of safely shutting the reactors if the operator should become insolvent. As it turned out, this requirement was rather prescient, given the financial collapse of Bruce Powers parent company, British Energy.
So it certainly not a remote possibility, as CNSC staff have suggested that the new owners of Bruce Power might become financially distressed (CMD 03-H27.C, p. 4). Sierra Club has several concerns about the CNSCs current direction on the guarantee. First, it is not clear to us that the amount of the guarantee ($150 million, noted in CMD 03-H27.C, p. 7) will be adequate, compared the current licence requirement for $264 million. We also believe that the Commission should specify the nature of the financial vehicle of the guarantee. There should be no question. It should be a segregated trust fund under the control of the CNSC, and this should be specified in the licence requirement.
INFORMATION BEING WITHHELD
As preparation for its submission for the Bruce A and B licence hearings, the Sierra Club of Canada requested various documents and information from the Canadian Nuclear Safety Commission (CNSC) and Bruce Power. A number of important safety-related documents noted below have been withheld. We do not believe that claims of commercial confidentiality can be justified.
Reports of the Bruce Event Review Panel were not provided. CNSC said that these reports are not in CNSC possession; they cannot be provided. Bruce Power refused to provide these reports.
Bruce Power has refused to provide its Fuel Channel Life Cycle Management Plan, and the summary of inspection results that Bruce Power gave to CNSC staff in the fall of 2002 (reference CMD 03-H28, p. 20). They have deemed it to be of a proprietary nature.
Bruce Power has refused to provide the CANDU Owner Groupss Fitness for Service Guideline (reference CMD 03-H28, p. 20). They have deemed it to be of a proprietary nature.
CNSC has refused to provide the probabilistic leak-before-break assessment (or even a summary of key findings) for Bruce Units 3 and 4 pressure tubes (reference CMD 03-H27, p. 24). They have simply suggested that CNSC staffs review of Bruce Powers submission is on-going.
Bruce Power/CNSC have refused to provide the 2003 correspondence with Bruce Power re: financial guarantee, arguing that it is proprietary in nature.
CNSC has refused to provide the executive summary of the Bruce A Probabilistic Risk Assessment. (CMD 03-H27, p. 26).