3D activation calculation method continues success in nuclear plant decommissioning

Most recently for the Mülheim-Kärlich nuclear power plant: with the successful completion of the activation calculations for this nuclear power plant, SIEMPELKAMP NIS created the prerequisites for successful decommissioning. Mülheim-Kärlich joins a series of nuclear facilities (pressurized water and boiling water reactors) whose activity inventory has been successfully calculated using the novel 3D activation calculation method developed by SIEMPELKAMP NIS.

Decommissioning as a challenge

The planned shutdown of all nuclear facilities in Germany as a result of the energy transition, as well as the decommissioning of international nuclear power plants, presents operators with considerable technical and economic challenges. For the decommissioning of a nuclear power plant, precise knowledge of the radioactive inventory of all components resulting from decades of plant operation is required. On the one hand, this is needed for the decommissioning license; on the other hand, it is not economically feasible to carry out sampling measurements everywhere in the plant and for every single component. Therefore, in the context of safe and economical decommissioning planning, exact calculations of the radioactive inventory of all components are indispensable.

Innovative calculation method

To achieve this, SIEMPELKAMP NIS Ingenieurgesellschaft mbH developed a novel, innovative calculation method based on three-dimensional simulation software for nuclear facilities called MCNP, which was invented in Los Alamos, USA. It makes it possible to calculate the radioactive inventory of the power plant at the highest level of detail. In this new method, for the first time, in addition to the detailed, realistic modeling of the 3D geometry, the entire power history of the plants, which have been in operation for several decades, was included precisely with all fuel elements used. This method has now been implemented as a standard.

The calculation method makes it possible to make predictions regarding the radioactive inventory at any point in the plant, even in a long-term time forecast, with a level of accuracy that was previously unattainable. This makes it possible to determine even the smallest activations in the reactor pressure vessel, including all internals, down to below the level of naturally occurring radioactivity – including for components located further out, such as the “biological shield” and loop lines up to the steam generators.

SIEMPELKAMP NIS Ingenieurgesellschaft mbH has already successfully applied this new method internationally for the pressurized water reactor at the Krško nuclear power plant in Slovenia and, within Germany, for the two reactor units at the Biblis plant (RWE Power, pressurized water reactors) and the Isar 1 nuclear power plant (PreussenElektra, boiling water reactor) – and now also for the pressurized water reactor at the Mülheim-Kärlich nuclear power plant (RWE Power).

Accurate results, low total activity

It became apparent that the new 3D calculation method delivers significantly more accurate results than the previous 2D calculation methods: in the Mülheim-Kärlich nuclear power plant, total activities were lower by a factor of 4, and in the case of Biblis even lower by almost a factor of 10. “This is important because the nuclear disposal of nuclear power plants is very costly. The operation of a nuclear power plant leaves behind radioactive building structures and internals that must first be dismantled and then disposed of as radioactive special waste in specially designed containers,” explains Dr. Imrich Fabry, project manager for activation calculations at SIEMPELKAMP NIS.

High savings potential

Thanks to the new 3D activation calculation method, considerable savings in cast containers can be achieved in decommissioning planning compared with earlier planning calculations, since a much larger proportion can be classified as “medium-activated” or even as conventional waste. “Medium-” or “low-activated” waste is disposed of in significantly less expensive containers. In the case of Biblis, this corresponds to savings of approximately 30 million euros.

The SIEMPELKAMP NIS 3D activation calculation method – innovation thanks to the Monte Carlo method

The 3D activation calculation method newly developed by SIEMPELKAMP NIS is based on the so-called “Monte Carlo method” (see information box). The method performs computer calculations of the transport of nuclear radiation (in this case neutrons), which is generated during the operation of the nuclear facility, through complex building structures in three-dimensional geometry. This makes it possible to reproduce or simulate exactly the sometimes very long path of a single neutron through all complex building structures and different materials along its way – exactly as it actually occurs in nature.

Activity atlas

This makes it possible to access every location in the reactor model and to illuminate complex structures at high resolution – for example, the reactor pressure vessel, core internals, or structures located further out, such as the biological shield or main coolant lines. In particular, neutron scattering – also referred to as the neutron streaming effect – can be visualized in remote areas. This 3D calculation model forms the basis for an “activity atlas” that maps a kind of relief map of the entire reactor.

What was new was the combination of a high-resolution, complex 3D reactor model with nuclear co-calculation, i.e., the operating history of the plant. In the case of Biblis, this was 39 years since commissioning. With the successful linking of the operating side to the decommissioning side in the activation calculation, SIEMPELKAMP NIS broke new ground that no one had previously taken in this form. “Only this combination makes it possible to carry out calculations of a plant’s activity inventory with unprecedented accuracy. To make this work, it was necessary to intervene in the Monte Carlo program and partially reprogram the calculation code,” explains Dr. Fabry.

Experience leads to optimization

In such a calculation model, a reactor unit is spanned by approximately 100,000 “detector cells”. The neutron flux emerging from the reactor core during the years of operation is simulated, and the number of neutrons arriving in these detector cells is recorded over time. Using this information, the activity in each detector cell is calculated. The model of such a reactor to be calculated is developed in close coordination with the decommissioning experts at SIEMPELKAMP NIS. The experience gained in earlier decommissioning projects flows into the modeling phase of the 3D activation calculation method. For example, in the case of the reactor pressure vessel, the grid arrangement of the detector cells is harmonized with the dismantling and packaging strategy. As a result, this leads to optimization of the number of containers required.

Finally, the results of the activation calculations are stored electronically in an “activity database”. This makes it possible to determine the activity values for any location and point in time in the reactor.

Intelligent calculation methods and parallel processing save time

The Monte Carlo calculation method is so accurate because the entire transport of a single neutron with all its complex physical reactions through the entire plant is always simulated. However, this comes at a price: only when a very large number of such neutrons are started is the calculation result sufficiently accurate at the desired location. It is a stochastic method that combines elements of statistics with probability theory.

Highly advanced method

Because the aim is to calculate the radioactivity in the entire plant, this means that the computational effort of such a Monte Carlo calculation is immense. A single standard desktop PC would have to calculate for about two years for one reactor unit to complete this task. Only thanks to intelligent calculation methods (variance reduction) can the calculation time be reduced to a few weeks. With these highly advanced methods, the art lies in preferentially transporting neutrons to locations in the reactor building that would otherwise be illuminated only very rarely under normal circumstances. The results of the calculations are not changed in the process. Parallel processing, i.e., distributing the calculations across multiple computer cores, provides additional support.

This makes it possible to obtain results that cannot be achieved with conventional calculation methods; for example, the (radio)activity originating from the reactor core is attenuated outward to the outer end of the biological shield by as much as 15 orders of magnitude. Practically no measurable radiation escapes from the biological shield (for comparison: 15 orders of magnitude correspond to the ratio between the diameter of a hair and the diameter of the sun). On the other hand, the calculations also show that the reactor head and main coolant lines are activated from the outside by neutron streaming.

Minimizing dose exposure

With these results, it is possible to identify “hotspots” (areas with particularly high radioactivity) that decommissioning personnel can avoid. Therefore, in the upcoming decommissioning activities, both the dose exposure of the personnel involved in decommissioning can be minimized and the quantities of waste to be disposed of for final storage can be determined very accurately. This enables plant operators to achieve a high level of planning reliability with regard to costs. The newly developed method thus offers an effective tool for safe and economical decommissioning planning with great savings potential for operators.

The validation of the newly developed calculation method – i.e., the test of the validity of the calculation results – was carried out in 2017 with local dose measurements in Biblis A. An excellent agreement between calculation and measurement was observed.

Meeting of the “Women in Nuclear” at the SIEMPELKAMP NIS booth, KONTEC 2019

Monte Carlo method also for new-build projects

The experience gained from previous projects can even be used for new-build projects of nuclear power plants: thanks to the high-resolution calculations, it is possible to identify which, in some cases remote, areas in the reactor building are activated by neutron irradiation. This makes it possible to provide recommendations, for example, for better shielding or alternative building materials that are activated only to a small extent by neutron irradiation. As early as 2018, SIEMPELKAMP NIS was able to contribute its experience from activation calculations to the planning of the new PALLAS reactor (Netherlands), which is to produce important isotopes for medicine in the future, as part of an initial project. The NIS team presented this in March at KONTEC 2019, the International Symposium on the Conditioning of Radioactive Operational and Decommissioning Waste. With its 3D activation method, SIEMPELKAMP NIS is thus well positioned to support both German and international operators of nuclear power plants in cost-efficiency-enhancing decommissioning and even new-build planning.

What can “Monte Carlo” do?

Monte Carlo simulations are used when analytical formulas for evaluating processes in nature fail or their solution is too complex. Difficult issues in physics, in the world of finance, and much more can be answered using Monte Carlo methods and are therefore used in these areas.

The method originates from stochastics, a subfield of mathematics, and was developed by the physicists J. v. Neumann, Metropolis, and Ulam as part of the Manhattan Project (Los Alamos, USA). It is based on the very frequent execution of random experiments. With the help of probability theory, problems are solved numerically, relying on the law of large numbers: if a random experiment is carried out again and again under the same conditions, the relative frequency approaches the probability of the random experiment more and more closely. The random experiments (in the case of activation calculations, this is a large number of neutrons whose entire path is simulated) are carried out by generating computer-generated random numbers.

When there were no computers, sequences of random numbers from the Monte Carlo casino were used – hence the name of the method. If a Monte Carlo program such as the MCNP used by NIS is also combined with the physical data from scientific research over the last approximately 80 years, this method surpasses all others in terms of accuracy. The data is provided by international organizations such as the Nuclear Energy Agency (NEA) within the OECD and the International Atomic Energy Agency (IAEA).