Chemical decontamination helps reduce dose during Borssele upgrade31 March 1999
In 1997 a major modification programme was carried out at the Borssele nuclear power plant. This upgrade programme is considered one of the most extensive of its kind in the world. The attendant collective dose was reduced considerably through several ALARA measures, one being the chemical decontamination of part of the emergency and post-cooling system. Using the Siemens CORD UV process an average decontamination factor of 17 was reached after only two of the three planned decontamination cycles. The decontamination programme took three days.
In the period from 1994 to 1997 an extensive Modification Project (MOD) was conducted at the nuclear power plant at Borssele, which has been operational since 1973. The project chiefly concerned the nuclear section of the plant and was part of a ten-yearly integral safety evaluation. The greater part of the project was carried out during the overhaul of 1997, which took approximately four months.
The MOD project resulted from the licence requirement to test the design basis of the power plant every ten years according to the latest nuclear regulations and, if necessary, to make modifications. This led to a total of 16 subprojects being carried out. As a result, the risk of a fuel melting accident has been reduced tenfold and the effect of such an accident on the surroundings can now be controlled more effectively. The total cost of the project was approximately NLG 470 million (about $250 million).
The first estimates of the collective dose of the MOD project were 10 person-Sv.
Contractual arrangements were made with the turnkey supplier (Siemens), taking account of ALARA as it is applied in practice by the plant operator, EPZ. This meant that Siemens was committed to take dose-reducing measures, applying a monetary value of NLG 1000 (over $500) for each mSv reduced, and that the maximum individual dose was to be 15 mSv/year (internal limit of EPZ in 1993).
One of the modifications of the installation was the introduction of a full separation of all subsystems (strands) in the entire cooling chain in order to exclude the risk of total system failure due to failure of individual components. One of the sub-projects concerned the so-called line separation of the residual heat removal system (TJ). With respect to dose this appeared to be a very extensive job, especially because of the large number of man hours estimated.
In order to control this dose a large shielding could be applied which, unfortunately, would hamper the work, as the workspace was limited already. The alternative of a chemical decontamination of those parts of the TJ system where the construction would require most man-hours was also considered, and after comparing decontamination with the shielding concept, it was decided to decontaminate.
The decontamination was effected using the CORD UV process, patented by Siemens. This is a chemical oxidation-reduction decontamination by means of UV light to bring about the photocatalytic breakdown of the decontamination chemicals.
THE CORD UV PROCESS
The CORD UV process is a multicyclical soft-decontamination process, in which the concentration of chemicals does not exceed 2500 mg/kg. The operating temperature is approximately 95°C.
Ion exchange resins retain the corrosion products dissolved during the process. The following four phases are completed per cycle:
In this phase a permanganic acid solution (HMnO4) is used to oxidise the chromium (III) present in the oxide layer to a well-soluble chromate (VI).
The chromate created and the remaining quantity of HMnO4 is reduced with an overmeasure of oxalic acid. The chromium ends up in the liquid as a complex.
The oxides present on the wall surfaces, are dissolved and the remaining metal ions, radioactive or otherwise, also end up in the liquid as complexes, after which they are led across the ion exchange resins together with the Cr, where they are retained.
The oxalic acid used for decontamination is fully converted into CO2 and H2O by means of UV light, so that a practically clean liquid results. After this a new cycle may start.
During the installation preparations for the TJ subproject an ALARA inspection was conducted to determine whether the cost of decontamination would be justified.
In the ALARA inspection a dose reduction of 700 mSv was calculated on the basis of a decontamination factor of 8. The dose without decontamination was estimated at 1029 mSv, and with decontamination at 335 mSv. The subsystem to be decontaminated (inclusive of post-cooling pumps and dewatering and vent lines) was divided into three process loops, between which switches had to be made during a decontamination cycle. For this purpose checklists were drawn up. Blind flanges replaced safety valves and check valves were removed in view of the need for circulation in two directions. As a precaution the interior parts of the post-cooling pumps were removed.
The permissibility of the decontamination process was also assessed by checking the list of materials and substances.
On the basis of experiences with the decontamination of warm systems (TJ is a cold system) it was expected that a total of three decontamination cycles would have to be completed. On the basis of the volume (25 m3) and the area (1050 m2) to be decontaminated and by estimating the thickness and composition of the oxide layer (2.5 mm), it was possible to determine the target quantities of chemicals to be dosed.
Thirty-two measuring points were determined by the Radiation Protection Department of the Borssele nuclear power plant in order to monitor the result of the decontamination, ie to determine the decontamination factor.
The parameters to be monitored by the Chemistry Group were the concentrations of Fe, Cr, Ni, Mn, Co and Sb, the activity concentrations of Co-58, Co-60, Cr-51, Mn-54, Sb-122 and Sb-124, as well as the total gamma activity.
When all decontamination installations were connected, the system was filled with demineralised water and a leak test was conducted. After this the temperature was increased to 95?C and the first decontamination cycle was started.
During the decontamination the Chemistry Group analysed two samples per 30 minutes during the pre-oxidation. In the other phases, two samples per hour were analysed. The results, reported directly to Siemens, are shown in the following diagrams.
After one cycle an average decontamination factor of 9 was already reached at the 32 measuring points. After two cycles the factor was 17. The quantity of activity removed during the second cycle was only 10% of that of the first cycle. The third cycle would have yielded even less. In view of the low dose reduction to be expected, the cost of achieving a further increase of DF was not thought justified, and the planned third cycle was cancelled.
Hence the total decontamination operation was concluded in three days.
In all 13.7 kgs of metal with a total activity of 1.8 X 1011 Bq were removed. The spatial dose rate decreased from an average 477 mSv/h to 32 mSv/h. On the basis of the decontamination factor realised the dose estimate was adjusted from 335 to 270 mSv. The dose actually realised was 118 mSv.
EVALUATION AND POST DECONTAMINATION
As indicated above, a third cycle was not conducted for the following reasons:
• The effect would probably be negligible given the results of the second cycle (activity removal of only 10% of that of the first cycle).
• At most measuring points the difference between the contact measurement and the spatial measurement was already small.
• The cost of the third cycle was too high for the limited dose profit expected.
After the decontamination the dose estimate for the work on the TJ system was decreased in conformity with the DF. In practice, the dose for this subproject appeared to be considerably lower. The cause of this is still unknown, but it is probably a result of the number of man-hours being lower than planned.
The quantity of ion exchanger used was higher than expected, as more metal mass was removed than planned. The quantity of activity removed, however, was lower than expected. In the following table, the performance realised and the performance expected are shown.
Seven months after the installation was put into operation, measurements of the lines were conducted again. This showed that recontamination was occurring. The radiation levels were still lower, but the DF was only 3. The long-term effects must, therefore, be deemed very limited.