The 2011 ANSWERS Seminar was held at the Wessex Hotel in Bournemouth, a location near to the superb Dorset coast and in close proximity to the town centre.
Over one hundred delegates attended all or part of the three-day meeting, with the programme covering the shielding, reactor physics and criticality technical areas. Programme topics included ANSWERS code developments, code applications, related general interest issues and software demonstrations. In addition to the seminar itself, parallel meetings of the UK Shielding Forum, the UK Working Party on Criticality and the UK Nuclear Science Forum were hosted at the same venue.
The 2011 ANSWERS Seminar started on Tuesday 24th May with Radiation Shielding. The host for the day was Pat Cowan, ANSWERS Applications Area Manager for Radiation Shielding. Pat introduced a variety of presentations on the recent developments in the ANSWERS Shielding codes and supporting tools and a number of applications using the ANSWERS codes.
The morning session started with a presentation by Geoff Dobson (Serco) entitled "Source Normalisation in MCBEND". He covered a number of aspects of source normalisation in cases with source rejection, and outlined new developments to support multiple source files. This was followed by a presentation by Ian Adsley (Nuvia), entitled "Plutonium measurement Issues", on the use of high resolution gamma monitors to measure the plutonium content of nuclear waste. Then Simon Richards (Serco) gave a talk on "Extension of Fission Spectra in MCBEND" in which he described a development to provide additional built-in spontaneous fission spectra from selected nuclides.
Following the morning break, David Picton (Serco) gave a presentation on "A Graphite Partitioning Study for the Hinkley Point B and Hunterston B Reactors". His study characterised the nuclear heating and damage in a moderator brick with a higher graphite loss than its neighbours. This was followed by a talk by Dr David Trnka (Areva) on "Detector Count Rate Determination Using Point-Kernel and Monte Carlo", in which a comparison of results calculated by MCNP and RANKERN was presented. The final presentation of the morning was by Steve Wall (Serco) on a "Whole Site Model for URENCO UK Limited". He described a study, using measurements in conjunction with MCBEND calculations, to determine dose rates at the URENCO site boundary and other locations on the site.
The afternoon session began with Adam Bird (Serco) giving an update on the current status of Visual Workshop version 2, including a demonstration of the latest features including Results Display and History Log. Then George Wright (Serco) gave a talk entitled "Update on BINGO", on the implementation of BINGO in MCBEND and new developments including support for run-time Doppler broadening and two-step coupled calculations. Next Aiden Robinson (Serco) described a study on "Dose Gradients within the Graphite Sleeves of Dungeness B Peripheral Channels". He described MCBEND calculations carried out to determine the sensitivity of the dose gradients to source intensity gradients in peripheral fuel elements.
Following the afternoon break a presentation was given by Albrecht Kyrieleis (Serco) on "Testing CAD (IGES) Models in MCBEND". Highlights of developments since the last Seminar were presented. This was followed by Chris Baker (Imperial College) giving a talk entitled "Radiation Transport Multiscale and Adaptive Methods", describing aspects of development work on the RADIANT code, including adaptive meshing. Finally, Pat Cowan ended the day by describing the current status and future plans for the ANSWERS Shielding Area.
The morning session began with an introduction to the reactor physics day by Tim Newton.
The first presentation was from Nigel Davies of Serco and described recent work to improve the treatment of burnup modeling in MONK. Nigel also discussed a new scheme to enable MONK to be coupled to a thermal-hydraulics solver for the purpose of thermal feedback calculations. The bases of the burnup and thermal feedback calculations are reaction rates and power distributions calculated on an XYZ mesh that is overlaid onto the MONK geometry model. The first presentation ended with discussion of coupling MONK to MCBEND for coupled neutron/gamma heating calculations and a new facility (called COWL) that can be used to simplify the transfer of irradiated material compositions between MONK and MCBEND calculations.
Pavel Mikoláš of Škoda gave the second presentation on WIMS modelling of fuel irradiation in three VVER cores and WIMS9/WIMS10 differences in estimates of reactor physics parameters for a VVER equilibrium cycle; the second part of the presentation also considered the effect of using JEFF3.1 instead of JEF2.2 data. Pavel concluded that the influence of using different JEF2.2 nuclear data sets has negligible effect on VVER reactor physics parameters, whereas there can be significant differences between values calculated with JEFF3.1 and JEF2.2.
The next presentation was from Jimmy Sudjana of Tractebel Engineering. Jimmy gave a brief overview of the PANTHER embedded supercell route currently being developed for the purpose of modelling mixed UOX/MOX PWR cores. The work described in this presentation was related to evaluating the requirement for including a control rod dependency in the nuclear data used by the embedded route. Jimmy presented results from the analyses and concluded that the use of a control rod dependency led to significant improvements in the estimates of rod worths, pin powers and assembly powers.
Hannah Woollard and Glynn Hosking of Serco presented details of recent developments of the CACTUS flux solver in WIMS. Hannah discussed the GMRES acceleration method, which can be used as an alternative to the Chebychev acceleration scheme currently used by CACTUS. Initial testing of GMRES acceleration indicates that the method shows promise, although further work is required to optimise the scheme. Glynn presented details of an explicit treatment of P1 scatter in CACTUS. The purpose of such a treatment is to improve calculation accuracy. Results from initial testing of CACTUS P1 solutions were compared with those from MONK, although it was noted that further validation of the treatment against other codes is required.
A presentation on uncertainty evaluation via sampling was then given by Albrecht Kyrieleis of Serco. The presentation described the new SPRUCE code that can be used to perform single and multi-variable sampling of calculation input data and thus ascertain uncertainties associated with those data. Albrecht gave an example of how the scheme had been applied to slab-geometry WIMS models to evaluate uncertainties in k-infinity due to variations in fuel density, coolant temperature and the density of a control rod material. For the purpose of evaluating uncertainties due to cross-section data, a case had been considered where 1000 different WIMS libraries had been used, each containing different U235 fission cross-sections that had been sampled from covariance data. Albrecht concluded that the study provides a useful platform for further investigations into uncertainty analysis through sampling.
The next presentation was given jointly by Emily Meese of Magnox North and Julie Martin of Serco. They discussed WIMS/MONK benchmarking of ARGOSY/PANTHER for vacancy modelling in Magnox reactors. Magnox have considered the transfer of fuel from a shutdown reactor into an operating reactor in order to extend plant lifetime and fuel utilisation. Doing so leaves vacancies in the shutdown reactor which can possibly lead to a reduction in its shutdown margin. The purpose of this study was to determine how accurately ARGOSY/PANTHER can model such vacancies in a shutdown core by using WIMS/MONK as a reference solution. When using ARGOSY data, the reactivity increase following vacancy introduction was lower than that predicted by WIMS/MONK. This is a consequence of the ARGOSY fuel data not accounting for interactions with vacancies. To allow for this corrections are applied to the ARGOSY thermal absorption and downscatter data. These corrections were validated against WIMS/MONK for a range of vacancy configurations and burnups. The outcome from this study has provided input for the safety case for performing inter-reactor transfer of fuel.
The final presentation of the morning session was given by Christopher Dean of Serco. Chris described the current status of the WIMS10 nuclear data libraries. The presentation also included some discussion of validation of the WIMS10 libraries. The effect of changing from JEF2.2 to JEFF3.1 data for a range of moderators and fuel enrichments was quantified. For the cases considered, agreement with experiment was improved when using JEFF3.1 data instead of either JEF2.2 or JEFF3.0 data. Chris also discussed work to validate the new ENDF/B-VII.0 library for WIMS10.
After lunch, Paul Smith gave a presentation of work undertaken by Fynn Scheben of Bath University for his PhD thesis. Fynn had considered iterative methods for neutron transport calculations. A mathematical framework was developed to help understand the convergence properties of transport solutions. Also, Fynn has considered the convergence of the method of perturbation as used in the MAX module in WIMS. It is hoped that the insight obtained from these studies will be useful in improving the solution methods used in WIMS.
Sheldon Hall, an EngD student from Imperial College, next gave a presentation. Sheldon has been working in conjunction with EDF Energy to develop a SPn solver for inclusion in PANTHER, thus providing a means for improved accuracy in the code via a transport solution. Results from the application of the SPn solver to a NEA benchmark and to modelling MOX/UOX interface effects were presented. Further developments of the solver will be made to consider P1 scatter effects.
The next presentation was given by Tim Newton of Serco who described new features in WIMS to improve QA functionality. These include facilities to allow comparisons to be made between the results of different WIMS executions and to enable "QA stamped" WIMS input data to be stored and retrieved. Tim also described the new EXTERNAL command, which allows any external program to be launched by WIMS. In future, this command could be augmented by an "open" WIMS interface to allow transfer of data between WIMS and other codes.
Paul Bryce of EDF Energy gave a presentation of a new PWR benchmark based on BOC1 of Sizewell B. This represents a benchmark for a large core, albeit with some geometric simplifications (e.g. modelled in 2D rather than 3D). Paul presented results from CACTUS and PANTHER solutions of the benchmark, both of which were able to predict pin powers to a RMS difference to measurement of within 2%. Paul noted that before the benchmark could be published it would be necessary to remove proprietary information.
Glynn Hosking returned to describe the recent development of the new MESH module in WIMS to help simplify the generation of input data for burnup and thermal feedback calculations. MESH accepts simple geometric descriptions of a model and automatically generates a set of WIMS materials required to model burnup and/or temperature feedback at the required level of spatial resolution. MESH also generates a flux solver geometry description for the model and WIMS input data to enable cross-sections for the model to be re-shielded. Facilities within MESH enable it to launch an external thermal hydraulics solver; updated temperatures and densities from this solver can then be mapped back onto the WIMS material set.
Tim Newton gave the last presentation of the day. This covered the current status of WIMS10 and summarised potential developments in future versions of the code. Customers are invited to submit their suggestions/requirements for the future development of WIMS.
The first session began with the standard safety talk from the hotel representative and an introduction to the criticality day by Malcolm Armishaw. The first presentation was from Ray Perry of Serco and described the ANSWERS Nuclear Data developments including an overview of the processing route, details and testing of the JEFF3.1 and ENDF/B-VII BINGO libraries using the MONK validation suite, improvements to the delayed neutron data, and the production of a CENDL3.1 library. The next presentation was from Ian Adsley of NUVIA and gave a fascinating update on the work being carried out to support the decommissioning of the Andreeva Bay dry storage units, and whether a criticality event at the bottom of a storage tube could be detected. The presentation included a number of pictures showing the state of the units and the difficulty of the work, as well as a hung-over polar bear and a 'surfing' penguin. The final presentation of the first session was from Nigel Davies of Serco who described the new coupled MONK burnup and Thermal Hydraulics route being developed based on the use of overlaid XYZ meshes to define the burn and thermal regions.
The second session began with a presentation by Jackie Martin of BE (EDF energy) on the comparison of safe mass limits using WIMS and MONK (with the ICASPA tool), showing that these preliminary results suggest the MONK/ICASPA methodology has greater flexibility and is more realistic. The paper also showed that the extrapolation length (required as an input to WIMS) could vary significantly with geometry and moderator. The second presentation of this session was given by Nigel Davies of Serco, who described a novel parallelization scheme based on Woodcock tracking and the use of separate computers to perform each of the functions normally found in a Monte Carlo burnup code. This approach assigned each function to one or more computers, allowing very large calculations to be performed, but the reducing memory requirements on any one machine, and providing the flexibility to allow machines to be reassigned as the workload varies during a many-step burnup calculation.
The third presentation of the session was from Elian Jones of AWE who described a study into practical storage configurations of enriched uranium metal spheres in cylindrical containers using a B4C based material. The study considered a number of packing arrangements, cylinder orientations and fault conditions. She showed that the B4C based material is an effective shield, fire was a particular hazard, and the leakage afforded by certain configurations should be studied further.
The final presentation before lunch was a short talk and a demonstration by Adam Bird and Tim Fry of Serco on the Visual Workshop 2 package, with the focus on results display, improved datsets handling, particle track display, and a 3D wire-frame mode for use with Nvidia 3D glasses.
The first presentation of the afternoon session was from Simon Richards of Serco in which he gave an update on the parallelization of the MONK burnup code described in Nigel Davies' first presentation of the day, noting that this task is still in its early stages. The results so far show that the speed scaling by number of nodes is good, but that there is an issue with the final keff that has yet to be resolved.
There followed a talk by Mark Goffin of Imperial College on the RADIANT code, which included a description of the multi (sub-grid) scale method, and several examples of its use, a multigrid solver, anisotropic spatial and goal based adaptivity, supermeshing, and angular discretisation. The final presentation of this session, by Jim Gulliford of the NEA-OECD, was an overview of the range of NEA nuclear science and data bank activities, covering their roles and the main areas of work and the main working methods in each activity. He showed how these activities linked into other NEA divisions, followed by a walk through of the various working parties, the ICSBEP and the IRPhE. He continued by listing recent nuclear science and data bank publications, a reminder of the forthcoming ICNC 2011 in Edinburgh, and some of the expert group meetings to be held in London in September.
Jim Closed by identifying several issues the working parties would be investigating, including uncertainity analysis, knowledge management, and the effect of Fukushima on work programmes and work programme priorities. There followed a separate presentation from Jim Gulliford giving much more detail on the forthcoming ICNC 2011, including the venue, hotels, registration, organising committee, draft programme, and technical tours.
The final session of the final day of the 2011 ANSWERS seminar began with a presentation from Brendan Tollit of Imperial College on FETCH, starting with an overview of the EVENT, BEANS and Fluidity components. He then reviewed the verification and validation of FETCH, the sub-grid scale diffusion approach, and the use of FETCH with the V/HTR using a range of mesh types, and finished by describing work done and results from coupling to a simplified 1D transient model. The next presentation was from Paul Smith of Serco on a comparison of quasi steady state scenarios in a post closure generic disposal facility (GDF) with the Oklo natural reactors in the Gabon. He began with a fascinating description of how the Oklo reactors were likely to have been formed, which included the need for atmospheric oxygen produced by photosynthesis: thus life was needed before these reactors could occur. The Neodymium, Uranium and Plutonium nuclide inventories were described, with the conclusion that some of the reactors had operated for several hundred thousand years, and the methods for such sustained criticality discussed. A mechanism in the GDF was proposed for the buildup of critical mass and this was compared with Oklo, suggesting that Oklo provides good support for these GDF studies. The final two presentations of the day were from Malcolm Armishaw and described a range of new features to be available in the next release of MONK, some of the long-term future plans for MONK, and a few hints and tips associated with embedded files and general 3D rotation of FG bodies.
Malcolm closed the day by thanking all those who had attended the seminar, in particular those who had taken the time and effort to prepare and present talks.