report 050908

Oldbury reactor core integrity

A report based on correspondence between Oldbury and the nuclear regulator, followed by two TV documentaries

Introduction:

Reactor I at Oldbury nuclear power station shut down in May 2004 and restarted in June this year, an unprecedented 13 month 'outage' or inspection and overhaul.

The twin-reactor 'Magnox' power station is 37 years old and was designed for a 25 year working span. It is acknowledged as the 'lead' station in terms of graphite depletion, ie the material composing the two 2,000 tonne reactor moderator cores has dissipated through heat and radiation more quickly than at any other UK nuclear power station (20)

The Shut Oldbury campaign (under Stop Hinkley) together with Wales Anti-Nuclear Alliance ran a campaign to shut the reactors in 2000-01 (also the subject of a BBC West documentary in 2001), which has been revived recently due to the long delay in Oldbury gaining permission from the regulators to restart this reactor. Reactor 2 began its outage in June and is still closed.

Correspondence received through the Freedom of Information Act:

In January Stop Hinkley requested from the operators, BNFL, all information relating to the current outage to include correspondence with the Nuclear Installations Inspectorate. Disappointingly BNFL's website database crashed for a week when the request was sent, delaying the starting date for processing the data by about a month. About 200 pages were sent altogether, which were copied to John Large, independent nuclear consultant and BBC West. The BBC broadcast a documentary highlighting Oldbury's safety risks on Monday 5 th September. ITV West also broadcast a more detailed documentary on 6th September.

It emerged through these sources that the graphite depletion in the worst sampled graphite bricks in reactor 1 was 33.8%. The NII had licensed the reactor to run till mid next year when the level of depletion is predicted to reach 35%, at which time Oldbury must make a new more robust safety case to continue operating that reactor. Reactor 2 was said to have already reached that point and so a stronger case is already required. Oldbury station manager, Joe Lamonby, declared BNFL's plans to continue operating till 2008, as long as safety cases are valid.

Points raised in the correspondence

1. NII suggested throughout their correspondence that depletion in the graphite could render the material unable to load-bear. A NII commissioned paper by Manchester University reported that at 35% depletion, the graphite could only bear 15% of its original capacity. In this circumstance or earlier, the reactor core or part of it could collapse (John Large BBC & ITV).

2. NII suggested that cracks in the graphite bricks could allow coolant gas to by-pass the hot magnesium alloy fuel cans (containing uranium changing into plutonium*), causing overheating and clad-melt and fuel-melt (7,14).

3. This in turn could lead to these* and other fission products being vented from the reactor, potentially contaminating thousands of people (John Large BBC & ITV).

4. Even small distortions in moderator channels caused by cracks or movements in the graphite bricks could hinder or trap and damage fuel cans and (less likely) control rods. The control rods are inserted to shut down the reactor and reduce heat (John Large BBC & ITV).

5. Cracks were suggested by NII to exist as observed on video footage although there is some dispute about their nature, BNFL saying they were just scratches (1, 16,17), but ultimately not ruling out their possible existence (14).

6. Out of the hundreds of channels existing in reactor 1 only thirty were examined through sample-taking and remote control video. This modest increase was grudgingly conceded by Oldbury above the normal four channels examined in their outages (2). BNFL had cited commercial reasons for objecting to expanding their remote visual graphite inspections in the 2001 outage (2).

7. In July 2002 (after the 2001 outage) NII wrote about their concern of "Cliff-Edge" failures , repeating this concern voiced at a previous meeting with Oldbury staff (5).

8. Despite dozens of letters exchanged and argued with scientific papers, NII reiterated their "cliff-edge" concern in October 2004, over two years later and a full six months into the outage (6).

Just a month before consent was granted to operate reactor 1, the NII wrote saying they were not convinced of the safety case in respect of the consequences of potential brick cracking leading to fuel melt (14).

9. In 2001, examination of 'trepanned' (drilled out) graphite samples was said to be delayed by almost a year across all BNFL reactors due to broken equipment. Oldbury reactor 2's samples taken in its 1999 outage had still not been analysed! (3)

NII complained about the backlog of testing due to the equipment failure and the planned removal of testing facilities from Berkeley to Windscale. They now insisted on Oldbury graphite samples being analysed on a fast-track basis . BNFL offered a revised work programme which NII then modified to allow other even more severely deteriorated reactors at Calder Hall and Chapel Cross to be examined first. (4) These reactors were both shut down the following year.

It seems awful that BNFL couldn't keep to its programme of crucial graphite inspection and there is a very strong sense of urgency to get Oldbury properly assessed.

10. Although there is an acknowledgement that cracks may exist with up to 3 mm gaps (14), BNFL continue to argue they are confident about restarting the reactor. Their argument is that two parallel, vertical cracks the full height of the brick would be needed to allow by-passage of coolant gasses or to constitute a risk. BNG said there was no evidence of this occurrence (in the 30 inspected channels) and crack development would be a slow occurrence (13).

It is complacent to say that 37 year-old carbon bricks could not develop cracks catastrophically. Pencil-lead is made from graphite. (John Large BBC & ITV)

11. If a fuel-fire were to occur, Oldbury states that the Burst Clad Detection (BCD) system would pick up the fault in eleven minutes. The controller is trained to respond to the yellow then red lights and an alarm that the BCD system triggers and he should 'trip' the reactor (23).

A fire could develop very quickly so that it would be impractical to protect the public . (John Large ITV)

12. Assuming the vigilance of the controller, Oldbury suggest the risk of vented fission products after such a fuel fire is only very slight (13) and "the extreme event was tolerable on the basis of its radiological impact" (26).

See point 11.

13. NII requested for the first vulnerable week or two after start-up there should be measures taken to enhance the vigilance of staff on the control-desk. Oldbury only agreed to put on an extra controller during start-up This could be a few hours or a shift at the most. They rebutted that no other enhancements were practicable or necessary. There is a back-up supervisor available when the controller has to leave the desk for short periods. (23)

14. The predicted strength of the graphite bricks could vary depending on the original purity of the 1970's fabricated material, which was unknown (22).

15. NII wrote their concern at restarting the reactor when the graphite depletion was approaching 40%. Oldbury argued back that the bricks would load-bear at 60% depletion, ie only 40% of the original mass of the bricks was still in existence! (9)

The operators are working in unknown areas. No other reactor has operated at this level of graphite damage. There was nothing in the documents viewed to support this assumption. The NII are saying you don't really know what you're doing. (John Large BBC & ITV)

16. Oldbury found their tested bricks had up to 33.8% weight loss. Fifty percent of the samples were depleted by 30% or more. All 44 samples were depleted by at least 21% (25). This compared to Hinkley 'A' which in 2002 had 11% weight loss (11), having shut down in 2000 through steel corrosion.

17. NII wrote indicating they would give consent to start up with some conditions. Five lines of the conditions were blanked out. One condition was a ceiling imposed on the amount of irradiation to the reactor core, before which the reactor must be switched off. (12) On ITV the regulators claimed that graphite accidents could only occur if the radiation was allowed to reach high extremes. But John Large has responded saying high pressure in this particular reactor also causes graphite deterioration . (See Shut Oldbury press release 6 th September)

Other letters suggested a one year time-limit imposed on running reactor 1, confirmed in the documentaries, before another inspection outage (24). Info gathered from reactor 2 would be made available to provide more data on reactor 1. Reactor 2 has however already reached the 35% degradation point requiring a stronger safety case. (BBC & ITV)

18. The injection of methane into the coolant gas-stream as a method of slowing down the graphite weight loss in 1994 was withdrawn at an early stage as it increased steel corrosion in the reactor structures (10).

19. BNG argued that there was extensive research in the UK and worldwide into the issue of radiolytic corrosion of moderator graphite (21). But the NII stated that there is little research in this field especially after 34% depletion and asked Oldbury to enhance their safety case accordingly (19). BNG showed no evidence from outside the UK to support Oldbury's safety case.

Other issues:

1. How much is money driving a decision to push ahead with a reactor with the finest of safety margins? BNFL has been technically bankrupt for some years and will have lost millions through this outage. The BBC were told the reactor earns £1 million per week in electricity charges. Why should BNFL even consider allowing commercial reasons prevent crucial extra remote visual inspections in 2001?

2. One German nuclear expert, Helmut Hirsch, has claimed the UK Magnoxes are equivalent to Eastern European reactors in their design and potential safety risk. Like Chernobyl, Oldbury has no 'secondary containment' considered essential for modern designs. Secondary containment is a large, often sphere-shaped vessel which holds escaping fission gasses in case of fuel-fires or other escapes of radiation from a reactor. Oldbury would not be allowed to be built if it were designed today.

Summary:

The tone of the letters between the Inspectorate and Oldbury seemed often to be 'intellectually combative'. BNFL seemed to prevaricate on some issues and dictate what happens on others. The NII began from the position that this reactor was probably unsafe to operate and tested BNFL's responses, demanding more and more information when insufficient evidence was offered.

That they twice questioned whether the reactor was at a 'cliff-edge' in its safety, taken with other repeatedly aired doubts over the validity of the safety case, suggests at the very least they felt the safety margins were too narrow. On the other hand BNFL's asserted confidence seemed often unrealistic but was applied to the NII in a 'broken record' manner with no doubt expressed on their part.

BNFL's repeated assertions won over NII's caution but it is difficult from these documents to see at which point or why that caution was dropped. On the other hand John Large's doubts and concerns seem justified if only because of the severe consequences of an error of judgement in the safety case or indeed the actual operation of the reactor.

It seems on balance and taking a common-sense view, that the commercial advantages to BNFL are greatly outweighed by the safety risks involved in continuing to operate this, the most damaged graphite reactor in the UK.

References:

(1) BNFL letter 9th July 2001

(2) NII letter 6th Feb 2001

(3) BNFL letter 31st Aug 2001

(4) NII letter 21st Sep 2001

(5) Acknowledged in BNFL letter 3rd July 2002 p.2

(6) NII letter 13th Oct 2004 p.2

(7) NII letter 28th Nov 2003

(8) BNG letter 13th Jan 2005

(9) BNG letter 26th Nov 2004

(10) BNFL letter 10th Feb 2003

(11) BNG attachment "NII Identifyer 001" to letter 6th Aug 2004