The reduction of doses is a strategic objective for EDF. It is linked to productivity gains, nuclear acceptability and respect of regulation. The best way to reduce dose consists not only of improving the reactor shutdown organization (time spent in control area, biological shielding, and so on) but also in improving the radiological state of the unit and the efficiency of the source term reduction operations. The EDF source term reduction project was begun to:

• participate to the long term view about the radiation protection issue (international feedback analyses)

• develop contamination prediction tools (such as OSCAR) suitable for industrial needs (for both operating units and EPR development)

• develop scientific models useful for the understanding of contamination mechanisms in order to simplify strategic decision processes

• continue updating and analysing contamination measurement feedback in corrosion products (EMECC and CZT campaigns) and fission products and actinides

• continue with the operational support in the short or middle term by optimizing startup and shutdown processes, passivation or zinc injection procedures, and improving purification media or material characteristics.

During the last 15 years, EDF dosimetry has significantly decreased in:

• collective dosimetry: from 2.44 man*Sv in 1991 down to 0.66 man*Sv in 2008,

• personal dosimetry: in 1992, 1200 operators received doses higher than 20 mSv/y; in 2008, no operator received doses higher than 18 mSv/y.

In order to meet its project needs, the source term reduction project has been divided into six different parts: long-term view; calculation codes; modelling; operational processes; purification; materials.

The long-term view

EDF is continuously monitoring dose levels, for two reasons. This work provides a dosimetric level evaluation of French and foreign units and compares EDF performances with international results. It also monitors different practices, carried out by foreign operators, that are supposed to have a significant impact on source term and thus, on radiation protection (design and operation). This technical monitoring is mainly possible thanks to a collaboration with the Electric Power Research Institute and several missions and technical discussions with other operators specifically on primary chemistry in operation or during shutdown transients. Thanks to well-focused assessments about contamination and national and international feedback analysis, the knowledge accumulation over time highlights efficient units and the key factors of their efficiency. This leads to progress in detection and long-term planning.

Calculation codes

Considering the cost of specific experimentation and the difficulties encountered in analyzing the impact of a single parameter among all varying parameters that impact operation, calculation codes are an interesting and useful tool to predict contamination levels in the primary coolant. These codes allow modellers to aggregate the entire body of scientific knowledge obtained for more than 40 years. They give, in the short term, support to operating process optimization (start-up, shutdown, hydrogen concentration, pH, zinc injection impact) and to design studies (such as the EPR and SGR). They lead, in the long term, to knowledge transfer towards common tools used in operation.

For example, the OSCAR V1.1 code [2], developed by France’s nuclear energy research agency CEA for EDF and Areva NP, simulates, in a single code, corrosion and fission product behaviour in the primary coolant operating at high temperature level. OSCAR V1.1 features a more efficient thermochemical database (nickel solubility data at high and low temperature) and a complete release module for corrosion products. For fission products, OSCAR V1.1 benefits from a thermo-mechanical module called ALCYONE that offers more accurate calculation of the behaviour of fuel inside defective rods.

Further OSCAR developments are planned to :

• improve prediction of corrosion product contamination during

shutdown transients (in particular, temperature and chemical conditioning effects, oxygenation peak and purification) and to extend calculations to auxiliary systems such as the residual heat removal system

• take into account EPR design and operation characteristics (B/Li management, stellite reduction, etc.)

• improve prediction of fission product contamination by a model calibration procedure based on feedback available for defective fuel assemblies,

• integrate a module for actinide behaviour modelling. This module would include uranium oxide dissolution, actinide precipitation/dissolution mechanisms in the primary coolant and the link between actinides and corrosion products’ behaviour.

Modelling

R&D groundwork is carried out not only for future code development, but also to improve contamination mechanism knowledge for a possible short-term decision in case of incident in operation. This work includes modelling of metal species release with the help of the BOREAL test loop, characterization of colloidal particle and sorption research and kinetic measurements of nickel oxide solubility and dissolution in collaboration with the CEA.

Operational procedures

Operation, shutdown and start-up

Reactor operation procedures can have a significant effect on source term. It is well known that chemistry management has a relevant role on corrosion product contamination. A modified chemistry (based on a maximum authorized lithium concentration of 3.5 mg/kg) has been run on several French units for five years (the pilot unit is the 1300 MWe Cattenom 2, with the N4 reactors Chooz 1&2 and Civaux 1&2). In 2010, the modified chemistry will be extended to the 1300 MWe units on which the new Galice fuel management programme will be used. (The Galice programme fuel would be enriched to 4.5% and have a maximum burnup of 65 GWd/t, compared with 4.0% and 56 GWd/t in the current Gemmes fuel programme for 1300MW reactors authorised in 1999. The programme would run in either three or four cycles for 502 effective full power days, an increase of 73 over the three-cycle Gemmes programme). The feedback will be sufficient to analyze the impact of boron/lithium management and pH on source term and thus radiation protection and to compare them to international results.

Moreover, in addition to chemistry changes, EDF is also planning a study of recent fuel management programmes, such as Galice, and Alcade for N4 reactors (three batches of 4% enriched UO2 in 411 effective full power days with a maximum burnup of 57 GWd/t), and the MOX parity regime (four batches of mixed 3.7% enriched UO2 and MOX with total 8.65% Pu in 900 MWe PWRs for 310 effective full power days). The project would study the influence of the programme on contamination and dose rates (especially with respect to the extension of cycle length and impact of power increase). OSCAR calculations will also be considered with a view to consolidate the feedback analysis.

Furthermore, zinc injection has been implemented in France since 2004 at Bugey 2 cycle 22 in order to reduce dose rates (a curative aim). In 2005, zinc injection was also initiated at Bugey 4 cycle 23, during the previous cycle of the steam generator replacement (a preventive aim). The results [3] allow us to conclude that there is no contraindication to add zinc in terms of contamination and dose rates. The positive effect of zinc on contamination observed at Bugey 2 and Bugey 4 is not as significant as observed in other foreign units. An encouraging slightly decreasing trend can be discerned, but needs to be confirmed. Based on Bugey results and on international feedback, EDF decided to implement zinc injection on eight new units from 2009 to 2011, not only to control contamination and dose rates and to confirm Bugey results, but also to manage a potential axial offset anomaly (AOA) risk on future core design and to mitigate primary water stress corrosion cracking (PWSCC).

Finally, feedback about new practices, such as chemical dehydrogenation and bubble collapsing at 130°C (put in practice on EDF units since 2004) or fast cooling at –40°C/h (experimented on several 900 MWe series units) will be studied. The studies include a statistical analysis of dose index, collective dose and cadmium zinc telluride (CZT) campaign measurements.

Measurement campaigns

EDF strategy in terms of contamination monitoring is based on two complementary strategies. Usually, routine dose rate measurements are achieved by plant radiation protection teams on primary loops (cold, hot and crossover legs) in order to calculate a reactor cooling system (RCS) index. Measurements are also taken to be able to compare performances for each unit.

However, this current programme is not complete enough to allow an accurate analysis when specific experiments, including modified chemistry, zinc injection, new fuel management programmes, or steam generator replacement, are carried out. Nor is it complete enough in the case of specific contamination: 122Sb and 124Sb (in the secondary source assembly), 110mAg (in control rods and primary pumps) or 60Co (in mainly stellites). In that case, EMECC (Ensemble de Mesure et d’Etude de la Contamination des Circuits, or activity measuring device) campaigns [4] carried out by the CEA have been commissioned for more than 30 years on French fleet units to better characterize contamination mechanisms. At the same time, EDF has also commissioned and financed EMECC campaigns on foreign units (including Doel, Sizewell and Trillo during the last four years) with the contribution of several European operators to compare different good international practices.

In addition to EMECC campaigns, EDF has been carrying out a new dose rate measurement programme since 2006 based on a semi-conductor CZT probe. Comparison with EMECC results, which are more accurate but also more difficult to handle, shows that the CZT device is able to satisfactorily quantify the main radionuclides’ contribution to equivalent dose rate [5]. The first feedback analyses confirms that CZT is a pertinent tool to understand contamination mechanisms. Technical modifications are planned to improve sensitivity.

Finally, following actinide contamination at Cattenom 3 during cycle 8, several measurement campaigns have been planned in order to better understand actinides’ behaviour in the primary and auxiliary circuits. One goal is a new actinide behaviour model, which could be included in the OSCAR calculation code. The mini-CVCS (chemical volume control system) campaign, planned at Cattenom 3 after cycle 15, will give information about reliability of measurements achieved via the sampling circuit compared to RCS circuit.

Purification

In parallel with the control of source term, an important optimization programme for purification is underway at EDF, as much in terms of new equipment as in purification management optimization.

A technological survey of new purification devices available in the marketplace by means of meetings with filter and resin suppliers is in progress. These discussions are an efficient way to communicate EDF needs proactively with suppliers.

Several specific new filter improvement technologies are planned to be tested. For example, the possibility of using filters based on an ultrafiltration membrane technology is being studied as a possible replacement of pleated media filter currently in place in the CVCS purification system. Laboratory experiments are in progress at the CEA. In addition, a specific test directly in a nuclear plant is planned by the end of 2010 (for one month during operation). This technology shows promise in obtaining a thinner filtration threshold than is currently available. A further technical and economic feasibility study will be done on the basis of the test results to determine whether the technology will be used in EDF NPPs.

At the same time, silica-free filters are currently officially recognised for use in NPPs. Indeed, reduction of silica source term is important both for boron release and to avoid the negative impact of a possible zinc silicate precipitation on fuel cladding during zinc injection. These new filters are proposed to be trialled in an experiment in a nuclear plant before 2011.

In parallel, experiments with resins are also being carried out in the laboratory and in the nuclear plant. First of all, two types of cation exchange resins (gel and macroporous) were characterized in the laboratory (EDF/R&D) in 2008 [6]. A new study has been launched to improve the knowledge of anionic exchange resin performances (particularly in terms of exchange capacity, boron retention and iodine selectivity), as well as exchange kinetics and flow rate impact. Moreover, chemical conditioning (and more precisely peroxide hydrogen) impact on resin damage and lifetime is planned during this project.

Purification management is another important way to improve purification efficiency. For example, the filter and resin survey, launched in 2007 and 2008, includes several filter and resin manufacturers and shows the importance of stabilizing both assurance of supply and cost control. Moreover, operational guidelines for filter and resin operation have been written in order to provide to nuclear plant operators reference documents. The guidelines incorporate a description of filter and resin characteristics, a description of their role and the main operating conditions, the replacement criteria and the main material usage guidelines. These support documents are currently used in the French fleet and meetings are organized each year with chemistry teams of each unit in order to collect operational feedback and to update recommendations.

Furthermore, EDF R&D has been working for several years on a modeling tool (called OPTIPUR) with a view to optimize the management of operating auxiliary circuit demineralization devices. OPTIPUR will be used in the future for studying the best compromise between resin cost and lifetime with an eye to the renewal of the national resin market. It will also be used for other purposes.

Finally, the usual practice for shutdown purification consists of using a full resin volume deep bead column for several successive shutdown transients. The opportunity to use only one resin per shutdown transient with a lower resin volume is being studied (and experimentation had been planned for 2009). This practice would avoid releasing radioactive contaminants first retained during the previous shutdown (especially colloidal silver species).

Materials

For EDF, source term reduction is also closely correlated to the type of materials used in the primary and auxiliary circuits and to their surface state.

Particular attention has to be paid to the reduction of cobalt-based material (stellites). Although originally used for their excellent mechanical performances, nevertheless, they are at the root of 60Co contamination. As a part of this project, a specific design study for the EPR showed that a reduction of cobalt-based material could reduce dose rates by 15%. Similar work still needs to be achieved for the whole fleet. In the same way, other species are also present, such as antimony (in the secondary source assembly and primary pump buttering and bearings) and silver (in the silver-indium-cadmium (AIC) control rods, and pump seals). The impact of those sources on dose rates will be studied by a feedback analysis. Following these results, replacement material and their influence on dose rates will be examined in order to define a general suppression or reduction policy for EDF.

Considering the importance of steam generator tubing in contamination mechanisms (metal species release over a large exchange surface), an accurate control of the manufacturing process must be achieved in order to maintain a good behaviour regardless of process modifications. As a part of this project, the significant influence of SG tubes’ surface state on release has been demonstrated with the help of the BOREAL loop. This characterization facility can test process improvements or modifications with a view to raising alerts for nonconformity behaviour.

Moreover, the definition of SG tube specifications (particularly surface roughness, metal hardening, impurities, grain size, microstructure or oxide composition) could also be a good way to assure a low release rate. Nevertheless, studies carried out until now have not led to unequivocal specifications that link SG tube behaviour to its surface state. Further work will be carried out on surface state and oxide characteristics in order to have a better understanding of mechanisms involved in release (using the TITANE and BOREAL testing loops, and x-ray photoelectron spectroscopy (XPS) and secondary ion mass spectrometry (SIMS) characterizations).

SG tubes’ surface pre-oxidation has been identified for a long time as a means of dosimetry reduction. Optimal physicochemical conditions during hot functional tests (whether pre-critical or not) have been analytically studied by the EDF R&D department. These results, combined with experiments in the BOREAL loop, led to a proposal of a specific procedure for both startup after SG replacement and first startup of a new unit (for example, an EPR). This study will be completed by a theoretical international feedback (in collaboration with Sizewell B) and additional experimental tests in the BOREAL loop to definitively validate analytic conclusions.

At the same time, gaseous oxidation could be a positive technique to reduce metal species release in the primary circuit by generating a protective oxide layer similar to that created by a chemical process. An experimental test with the help of the BOREAL loop is also planned in 2010 in order to improve mechanisms and efficiency knowledge.


Author Info:

Gilles Ranchoux (gilles.ranchoux@edf.fr) and Luc Guinard: EDF/SEPTEN, 12-14 avenue Dutrievoz, 69628 Villeurbanne, France. Stéphane Taunier, Frédéric Gressier, and Florence Carrette: EDF/CEIDRE. Stéphanie Leclercq: EDF/R&D/MMC. Bernard Jeannin: EDF/Etat Major.

This article is an edited version of a paper given at the 2009 International ISOE ALARA Symposium in Vienna, Austria.

Related Articles
Sizewell B outage will extend into summer

References

[1] L. Guinard and al., “Review of the evolution of international collective radiation exposure and associated good practices, with special focus on Source Term Reduction strategies for EDF PWRs” ISOE European ALARA Symposium, Turku (Finland), 2008.

[2] F. Dacquait and al., Modalisation du transport des produits de corrosion dans le circuit primaire des REP. SFEN ST2 workshop, Paris (France), 2007.

[3] A. Tigeras and al., Complete Analyse of Zinc injection at BUGEY 2 and 4 NPC-08, Berlin (Germany), 2008.

[4] R. Eimecke et al., “Ensemble de Mesures et Etudes de la Contamination des Circuits”. International Conference on Radiation Shielding, Bournemouth (UK), 1988.

[5] A. Rocher et al., “New CZT measurement device” Comparison with EMECC measurements in EDF PWRs.” ISOE European ALARA Symposium, Vienna (Austria), 2009.