Decontamination at Biblis A&B24 October 2018
A new process was used to achieve decontamination factors of up to 85, with the added benefit of utilising only existing plant systems. Anna Prüllage, Markus Thoma, Laura Schneider, Hartmut Runge and Markus Becker present the results from full system decontamination of Germany’s Biblis A&B.
IN DECOMMISSIONING A NUCLEAR POWER plant chemical decontamination of the primary circuit is usually required to allow dismantling activities to take place at low radiation level. The main decontamination target is therefore minimising activity level in the primary circuit (the reactor coolant system and its primary auxiliary systems) to reduce the personnel dose received during dismantling work.
The radioactivity is predominantly in the system’s oxide layers. Chemical decontamination is therefore focused on dissolving these oxide layers and the subsequent removal of liberated radioactivity from the decontamination solution.
The ASDOC_D-MOD process is based on several short decontamination cycles with low concentrations of chemicals per cycle. Small quantities of methane sulfonic acid are added to control the pH-value and to avoid precipitation of iron complexes from the decontamination solution.
In contrast to other decontamination processes, ASDOC_D-MOD does not require a standalone decontamination facility. The existing plant systems (the chemical dosing system, ion exchange clean-up systems, main coolant pumps and the control instrumentation) are used to operate and control the decontamination process.
The ASDOC_D-MOD process was originally developed to avoid damage to materials sensitive to chemical corrosion and to suppress hydrogen production during the decontamination process. The main aim of development, therefore, was to find a process smooth enough to avoid corrosion of weak materials while to guaranteeing effective dissolution of the oxide layers.
The ASDOC_D-MOD process is therefore a smooth chemical decontamination process which can be applied in nuclear plants containing not only austenitic steels but also ferritic or martensitic chromium steels or other materials which might be sensitive to organic or inorganic decontamination acids. The decontamination efficiency of ASDOC_D-MOD is comparable to other decontamination processes and usually delivers decontamination factors better than 85.
Extensive material researches were done as part of the development process. These results are taken together to a comprehensive material report which can be used as basic reference for FSD planning with ASDOC_D-MOD. The researches were mainly made in the NIS pilot plant, which reproduces the main components of a nuclear primary system.
In each full system decontamination (FSD) project special attention should be paid to the material situation, to avoid any chemical corrosion damage to sensitive components.
How it works
The aim of an FSD is to minimise the radioactivity within the primary circuit. During the operational phase of a nuclear power plant the activity (mainly Co-60) resides in the oxide layers, which mainly build up on interior austenitic surfaces. To remove the activity after plant shutdown the oxide layers on the internal surfaces have to be removed. In any FSD process the oxide layers and the included activity are dissolved by chemical agents. Metallic ions from the oxide structure, and the activity included, are dissolved in the decontamination solution and subsequently extracted by means of ion-exchange resins.
The ASDOC_D-MOD process uses methane sulfonic acid (MSA) to adjust the pH-value of the decontamination solution, to avoid precipitation of iron ions from the solution. Permanganic acid (HMnO4) is used as a strong oxidant and oxalic acid (C2H2O4) as a chelating agent.
The number of single decontamination cycles required is 15 to 20. This allows low chemical concentrations of about 100ppm or less per cycle. With this low chemical concentration the process can be safely controlled and avoids the hazards of hydrogen generation and unintended base material attack.
The ASDOC process characteristics requires significantly less chemical and results in less secondary than other decontamination processes.
In the process dissolution of the oxide layer can be controlled and recorded very precisely. The final oxide-free status of the system surfaces will be detected precisely so that the process can be terminated precisely, without the risk of unnecessary waste production.
It is an outstanding advantage of the ASDOC_D-MOD process that it does not require extensive external decontamination equipment, using instead the plant’s own operational systems: chemical dosing systems, volume control system, ion exchange clean-up systems and residual heat removal system.
Figure 1 illustrates the procedure of a single FSD cycle with the ASDOC_D-MOD process.
The cycle starts by adding MSA to the decontamination loop at an equilibrium concentration of 50-100ppm.
In the following step permanganic acid is added until its concentration in the primary circuit reaches 40-70ppm. The permanganic acid oxidizes the Cr3+ ions of the chromium oxide layers to more soluble Cr6+. During this step manganese dioxide builds up. The process phase takes 6-8 hours (depending on system volume). After the oxidation reactions have come to an end the dissolved chromium is removed from the decontamination solution by anion- exchange resins. This cleaning process requires 12-16 hours.
In the next step, oxalic acid is added to the primary circuit and will react with manganese dioxide produced in the preceding step, as well as with the pretreated oxide layer. First manganese dioxide will be dissolved, consuming about 30ppm oxalic acid. Chemical dissolution of the pretreated oxide layer will use up another 100ppm oxalic acid.
By adding oxalic acid to the primary circuit the metal ions from the oxide layer are transferred into the liquid phase. Thereby iron, nickel, manganese and zinc are dissolved. This dissolution reaction requires 5-7 hours.
Dissolved ions are finally removed from the decontamination solution in a bypass cleaning process by cation-exchange resins. Together with the metal ions the mobilized activity (mostly Co-60) is deposited on the cation-exchange resins. This step takes about 20-24 hours. During the ion exchange cleaning phase the oxalic acid concentration decreases steadily, due to chemical reactions and thermal decomposition. The remaining oxalic acid is finally decomposed by adding small amounts of permanganic acid.
The four steps are repeated in up to 20 cycles until the decontamination target is finally reached.
Results from Biblis
The ASDOC_D-MOD process was conducted at Biblis A (in 2016) and B (in 2017/2018). The FSD in unit A was completed within 39 days, corresponding to 13 separate cycles with a mean duration of three days per cycle. The overall duration of the FSD in unit B was 51 days. This period included 19 decontamination cycles and the final system clean up (purging). The mean duration of one cycle was 2.2 days.
At Biblis A the DF-average taken over 72 measuring points was 90 and in unit B the DF-average taken over 46 measuring points was 85 (see Figures 2&3). The selected measuring points had been distributed across the decontamination circuit in a representative way.
In general, the decontamination efficiency will vary, and will depend on a unit’s operation history, materials, flow rates and the operating temperature. All will influence the component activity and thereby the dose rate and decontaminability.
The dose rate development during the FSD in Biblis B is shown in Figure 4 for the recuperative heat exchanger and Figure 5 for one of the steam generators.
For both components, a measuring position on the component’s wall was chosen. At the recuperative heat exchanger, the initial dose rate of 2825Sv/hr fell during decontamination to 54μSv/hr, corresponding to a decontamination factor of 52. The initial dose rate at the steam generator was 178μSv/hr and its final value of 0.4μSv/hr corresponded to a decontamination factor of 445.
Fluctuations of the dose rates were caused by main coolant pump switching operations.
Despite high decontamination factors the total amount of waste (ion exchange resins) and the corrosive ablation on material samples are low. The total amount of waste for Biblis A was only 6.65m3 and in Biblis B was 12.75m3. Both include the amount of ion exchange resins spent for the final removal of chemicals from the system cleaning. Overall, 143kg (Biblis A) and 220kg (Biblis B) of metal ions were extracted from the circuit by the exchange resins.
Beside the control of chemical parameters and of the dose rate during the FSD the corrosive ablation on materials was controlled by material samples which were installed in a separated bypass loop to the main circuit. In this bypass loop some materials with low corrosion resistance were inserted. The corrosive ablation was at each material very small (Figure 6).
During the licensing procedure, a threshold and a limit value for some sensitive materials were defined. The threshold for the chromium steel 1.4923 was set at 80μm and the limit at 150μm. At the end of the decontamination the corrosive ablation on the 1.4923 specimen was only 1.08μm. This value is well away from the approved limit and from the threshold value and underlines the extremely low corrosion risk of the ASDOC_D-MOD process.
In addition to the corrosive control of sensitive materials during a chemical decontamination process, special attention should be paid to production of hydrogen gas.
During the ASDOC_D-MOD process the concentration of hydrogen is continuously measured in the liquid and gas phases. In both projects, the hydrogen concentration never reached detection level (1ppm). Large scale corrosion of structural metals (e.g. RPV internals) can therefore definitely be excluded from the ASDOC_D-MOD process.
Full system chemical decontamination projects at Biblis A and B had convincing results.
Mean decontamination factors of 90 (Biblis A) and 85 (Biblis B) were achieved. Chemical corrosion of chromium steels and even of some isolated ferritic base material was very low.
The formation of hydrogen was below detection limits in the decontamination solution as well as in the gas phase.
The total amount of radioactive waste (ion exchange resins) was less than 12.75m3 at any unit.
The FSD procedure was planned and executed under the control and attendance of the regulator and the consultants responsible.
After completion of the last decontamination cycle and the purge process, components were inspected. The integrity and function of the system components were analysed according to the requirements under specific operating conditions.
Author information: Anna Prüllage, Chemical and process engineer at Siempelkamp NIS; Markus Thoma, Head of department, special projects at Siempelkamp NIS; Laura Schneider, Chemical and process engineer at Siempelkamp NIS; Hartmut Runge, Radiological physicist at Siempelkamp NIS; Markus Becker, Project manager, chemistry at Siempelkamp NIS