Trojan: all wrapped up neatly

When the decommissioning plan for the Trojan plant was approved, in April 1996, it included plans for disposal of the reactor pressure vessel and internals by segmenting the internals to fit into 45 standard shipping containers which would be transported by road to the disposal site at Hanford. After the internals were segmented, the reactor vessel itself would be segmented or shipped whole. During 1995, however, Trojan’s owner Portland General Electric (PGE) successfully completed a project to remove, package and transport several other large components – four steam generators and the pressuriser – as single pieces. The success of this project led PGE to propose, in January 1997, to change its decommissioning plan to dispose of the pressure vessel and internals as a single piece.

Under the new disposal option – known as the Reactor Vessel & Internals Removal Project (RVAIR) – the pressure vessel would be filled with low density concrete with the internals in place. It would be lifted through the containment, using the opening through which the steam generators had been removed in 1995. The concrete-filled vessel would then be transported 270 miles by barge and some 30 miles by road to the American Ecology low level waste site in Hanford, near Richland, Washington state. Once at the site, it would be placed on a special disposal saddle in a trench 850x150 ft in area and 45 ft deep, and would finally be covered with soil and rock to a depth upwards of 17 ft.

PGE’s arguments in favour of the single shipment were difficult to fault: the process would be simpler; would incur less occupational exposure; would reduce shipping and therefore both public exposure and the possibility of transport accidents; and would be significantly cheaper. Finally, it would allow the vessel and internals to be classified as low level waste, whereas portions of the internals, if segmented, would be classified as ‘greater than Class C’ (GTCC) waste and would not be eligible for disposal at the American Ecology site.

Planning to move the vessel by ship meant the timing of the project took on a new importance. Thanks to river conditions, weather limitations and other barge traffic schedules, each year there was only one short window of opportunity, stretching some six weeks from the beginning of August, when the vessel could be shipped along the Columbia River. PGE’s aim was to ship the vessel during 1999, and to meet this engineering schedule the company would have another tight schedule to achieve: approval was required from four agencies for this new method of disposal, in time to prepare the vessel. Until approval was granted from the Nuclear Regulatory Commission, the US Department of Transportation, the Oregon Office of Energy, and the Washington Department of Health, the only preparations PGE could make were those that would be appropriate for either disposal route.

Winning approvals from the various agencies within this timetable was difficult, but final approval was granted on schedule at the end of November 1998. PGE could finally move forward with its plans, and the first irreversible step was quickly taken when, in early December, the vessel was filled with low density cellular concrete.

LOW LEVEL WASTE?

One of the most important questions to be settled before the RVAIR disposal could go ahead was whether the reactor vessel and internals could be classified as low level waste. If dismantled and packaged separately some parts of the internals (around 340 ft2) would certainly have been greater than Class C waste and would have to remain on the Trojan site, since the American Ecology site would not have a licence to handle it. The low level waste issue was an important one for local people too, and some objectors in Washington questioned whether the single package could legally be classified as low level.

To settle the low level waste issue PGE had to show that the vessel would meet 10CFR61 requirements, and this was done by carrying out a pathway analysis which considered hypothetical doses to the public under a variety of scenarios up to 10 000 years in the future. The package was assumed to contain 155 Ci of inner surface activity and 2.01 MCi of activated metal (and therefore non-releasable), taking into account that the activity had been decaying for five years since the reactor was shut down in January 1993. It was also assumed that an 8-10 ft engineered closure cap would be placed above grade at site closure. The scenarios for long term exposure included:

• Agriculture – a family living close to the structure consumes potentially contaminated crops, groundwater and milk.

• Inadvertent intrusion in the form of well construction above the disposal.

• Inadvertent intrusion in the form of a residence with a cellar intruding into the closure cap.

• Waste decomposition gas, assumed to penetrate a house built above the trench.

• Natural erosion.

• Waste site flooding.

• Biotic transfer – takeup by plant root or burrowing animals.

The first agriculture scenario was shown to produce the highest dose, most importantly via groundwater transport pathways. Dose calculations produced estimates below prescribed regulatory limits: some thousands of years into the future they were estimated at 0.13 mrem per year (thyroid), 0.09 mrem per year (whole body) and 0.2 mrem per year (red marrow), at 0.05 in per year groundwater infiltration.

Several shorter term scenarios were also considered, including:

• Waste container breakage during handling operations.

• Fire affecting the vessel in an open disposal trench.

• Fire involving the vessel and transport vehicle.

• Surface contamination of the trench as a result of the vessel handling and disposal.

PGE concluded that the short term occupational and radiological advantages of handling the vessel whole far outweighed that of sectioning and removing the internals. The advantage of handling one vessel instead of 44 containers was described as substantial, in view of the fact that the external exposure on the single package was 200 mrem per hour, while for some sectioned packages exposure would be likely to reach some tens of thousands of mrem per hour. Estimated exposure to site workers at American Ecology was 0.2 person-rem for disposal of the vessel, compared to 4-5 person-rem for the sectioned internals.

IN CASE OF ACCIDENT

In gaining approval for the single vessel transport, PGE also had to consider the consequences of accidents during the transport.

For example, the vessel package, prepared for transport, weighed something over 1000 tons, and it had to be lifted several times while it was being carried from transporter to barge, to another transporter, and to its final resting place.

One possibility that had to be investigated was the likelihood and consequence of the vessel being dropped.

The contractor Bigge Crane and Rigging had experience of similar sized jobs, albeit not within the nuclear industry, and Bigge was supported by the PGE quality assurance department. A testing programme was completed for load bearing components, and suppliers of key components were audited. An analysis of the likely consequences of a drop on site at Trojan was also carried out. PGE assumed that, regardless of height, the vessel would break open in a drop and release contamination. In estimating the amount of contamination from the break PGE calculated that the vessel contained 155 Ci of surface contamination. It was assumed that 10% of this contamination would be released in a drop and that 1% of this amount would become airborne. Using NRC Reg Guide 1.25, PGE calculated that 0.155 Ci (ie 0.1% of the total 155 Ci) would be released off site. This is well below the 2.07 Ci which was calculated by PGE to meet Environmental Protection Agency protective action guidelines.

The results of PGE’s analysis bore out the view that a single transport by barge posed less risk than the 40-50 road transports that were envisaged in the original decommissioning plan. PGE was successful in winning approval from Oregon authorities for the single transport on 16 October 1998. Final approvals from all four authorities were granted by December 1998.

The reactor vessel is made of carbon steel varying in thickness from 5 3/8 in to 10 1/2 in. The internals and internal cladding are stainless steel. To prepare the vessel for transport carbon steel plates were welded over all the vessel openings. Carbon steel radiation shielding of varying thickness was attached to the outside of the vessel. The shielded package was coated with acrylic latex to fix any residual surface contamination, and finally the package was wrapped with a layer of heat shrink to provide an additional level of protection from the weather.

By June 1999 PGE had filled the vessel with cement, and had lifted the vessel from the containment and made most of the preparations for transport.

The vessel arrived at American Ecology’s site at Hanford, after a journey of 270 miles up the Columbia river and 30 miles by road from the port at Benton at a stately 5 mph. Burial began on 19 August, watched by an audience of 80 stakeholders, and is expected to take around three months.

At the Trojan site, fuel in the spent fuel pool is due for transfer to a recently-built and licensed storage facility. Remaining equipment, structures and site areas will be removed or decontaminated to a level that should allow the property to be released for unrestricted use. PGE estimates that for most areas this will be in 2002. Those structures that are uncontaminated, or have been decontaminated, and are not reused will probably be removed in 2018 or 2019.