The ANSWERS Seminar, 20th-22nd May 2014

Over one hundred and twenty attendees from the UK and Europe attended the 29th ANSWERS Annual Seminar, a three day event covering Shielding, Reactor Physics and Criticality. The programme included presentations from ANSWERS staff and customers on the development and application of ANSWERS software.

Day 1: Radiation Shielding

The seminar started on Tuesday 20th May with Radiation Shielding. The day was split into four sessions with separate hosts for each session. Session 1 was hosted by Pat Cowan, Session 2 by George Wright, Session 3 by Dave Picton and Session 4 by Kim Vignitchouk.

Pat Cowan, the ANSWERS Applications Area Manager for Radiation Shielding, introduced the programme for the Shielding Day, including 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 first session started with a presentation given by James Watson entitled MCBEND Modelling in Support of the Berkeley Chute Silo Retrieval Project. He described the modelling strategies used, including Leakage Files and Source Contribution capability. The second presentation entitled BINGO Efficiency Investigations was given by Geoff Dobson. He described the implementation of BINGO in ANSWERS codes and potential improvements on performance. The last talk of the session was given by Kim Vignitchouk, entitled Using MCBEND GRID option for Maze Problem. This presentation described the use of MCBEND11A GRID option in a case where traditional acceleration methods were not appropriate.

After the morning break, the second session started with a presentation by Ian Adsley from Nuvia, on The Remediation of the Olympic Park Site. He described the methodology developed to assay the contamination of the site and its remediation prior to construction works. The second talk entitled Interaction between Monte Carlo sample over-splitting and geometry defects was given by Geoff Whiley, from EDF Energy. This described the HRA/HYA core-octant model and the impact of over-splitting and geometry defects on results. The final talk of the morning sessions was a presentation given by Tim Fry on Visual Workshop. This provided an overview of the current status and new features of Visual Workshop 3A as well as a brief description of future developments.

The first afternoon session started with a talk by George Wright entitled MPC Modelling. The presentation focused on the use of MCBEND as a tool in scoping studies and detailed shielding assessments. The next presentation was given by Adam Bird on Importing Geometries from CAD and described the latest progress in CAD link capability. The third talk, before the afternoon break, was given by Oisin O’Sullivan, entitled Graphite Dosimetry within AGR Cores. He described a methodology using 'Power Partitioning method' combined with 'Phantom Fuel Correction Factors' to predict the dose in the vicinity of vacant fuel channels.

Following the afternoon break, the second afternoon session started with a presentation given by Mark Goffin from Imperial College on Adaptive Spatial and Angular Resolution using a Wavelet Angular Basis. He described the development of an adaptive method which could be used to create importance maps in MCBEND. The second talk given by Adam Bird, was entitled Container Confusion. The presentation focused on the distinction between three keywords: 'contained', 'confined' and 'container'. The final presentation of the shielding day was given by Pat Cowan and briefly described the current status and future plans of the MCBEND shielding code.

Day 2: Reactor Physics

The Reactor Physics day was split into four sessions with separate hosts for each session. Session 1 was hosted by Glynn Hosking, Session 2 by Tim Newton, Session 3 by Brendan Tollit and Session 4 by Dave Powney.

The first session began with a presentation by Christophe Schneidesch from Tractebel Engineering. Christophe discussed the testing of methodologies for uncertainty analysis in the context of core physics calculations, specifically with the aim of meeting new safety criteria. Ben Lindley from Cambridge University gave the next presentation, which considered the modelling of reduced-moderator LWRs with WIMS, PANTHER and PARCS/RELAP. Ben had performed a wide range of feasibility studies for hypothetical reactor designs using uranium, plutonium and thorium fuels. The final talk of the session was given by Tim Fry from AMEC. Tim described new WIMS-specific features in Visual Workshop, including the display of geometry for a range of different WIMS modules. Visual Workshop can now also produce results displays from CACTUS calculations, including flux, reaction rate and power maps.

After the morning break, the second session started with a presentation by Brendan Tollit from AMEC on the new features that are included in WIMS10. Paul Bryce from EDF Energy then gave a presentation on reactivity validation evidence from Sizewell B. Paul described work to compare critical boron concentration estimates from WIMS/PANTHER calculations with measured values; he also highlighted the importance of fully understanding the basis of the experimental data being used in such comparisons. The next presentation was by Jack Blake from Bath University, who talked about the development of a new approach to diffusion synthetic acceleration. This new approach aims to improve the efficiency of the acceleration by using domain decomposition to split a model into diffusive and non-diffusive regions. The final presentation of the morning sessions was given by Dave Powney from AMEC. He discussed recent WIMS validation studies, particularly work to validate the new gamma heating route in WIMS10. Included in this work was an analysis of the NESSUS experiment

The first afternoon session started with a joint talk by Dave Powney and Tim Newton from AMEC. They described the basis of the PERSEUS calculation method in WIMS and studies to try to improve its accuracy. Martin Knight from EDF Energy gave the next presentation. This considered the implementation of a predictor-corrector burnup scheme in WIMS. By using this method it is possible to achieve target accuracy with fewer burnup timesteps. Implementation of the scheme was done entirely in terms of the WIMS input language. The last presentation before the afternoon break was given by Glynn Hosking from AMEC. This presentation discussed the prototyping of a new feature in CACTUS to allow it to treat hole geometries (as used by other ANSWERS codes like MONK and MCBEND).

Following the afternoon break, the second afternoon session started with a presentation given by Peter Smith from AMEC. Peter described the implementation and testing of a new treatment in the WIMS PROCOL module for particulate absorbers in a slab geometry. Brendan Tollit next gave a presentation on the development and testing of an SP3 transient solver in WIMS. A benefit of SP3 solutions is that they can provide accuracy improvements over diffusion theory without the computational overhead of a full transport solution. This new functionality forms the first transient analysis tool to be available in WIMS. The final presentation of the reactor physics day was given by Glynn Hosking and briefly described the current status of WIMS and future plans for its development.

Day 3: Nuclear Criticality

The first presentation of the Criticality day was Using the New Features in MONK10A, presented by Chris Baker of ANSWERS. During this presentation Chris gave a short guide to using a range of the new features in MONK10. This was followed by a presentation on Modelling Continuous Dissolver Geometries presented by Nigel Davies of ANSWERS. Nigel described the La Hague continuous rotary dissolver and the ORNL continuous dissolver. He then went through the MONK model of the La Hague dissolver in some detail, demonstrating how a MONK model for a relatively complicated geometry can be reduced to three PowerPoint slides

The next presentation was in two parts: first Paul Smith (ANSWERS) talked about Sampling Methods; and then Chris Baker (ANSWERS) talked about Optimisation. Paul gave an overview of the SPRUCE tool which allows the user to launch multiple calculations (e.g. MONK) that require algorithmic sampling, and to analyse the results, and showed how this was one way of determining the uncertainty on calculated results arising from the uncertainties on inputs, data and methods. Chris then took over and described a range of classical and heuristic optimisation algorithms which have been coded in JAVA and used with MONK to optimise both a single parameter and multiple parameters.

The final presentation in the first session was Comparison of Neutronic Properties of Solvents by Laura Carruthers of AWE. Laura discussed the reasons for considering alternative solvents to those previously considered (ethanol and trichloroethylene). She then proceeded to present an assessment of the proposed replacement solvents in terms of their moderating and reflecting properties. She showed how hand calculations were a useful tool for determining moderator effectiveness and then presented a series of safe volume curves for the two proposed solvents which were generated using MONK.

The second session stared with a presentation on CRITEX Validation by Paul N Smith (ANSWERS). Paul introduced a code called CRITEX intended to calculate fission heating, temperature rises, power variation and fissile material burn-up in uranyl nitrate or plutonium nitrate systems (with some restrictions on enrichment). Paul showed that CRITEX generally achieves good agreement with experiments.

The next paper was An Advanced Multiphysics Model of an Aqueous Homogeneous Reactor by Chris Cooling (Imperial College). Chris described a physically intuitive and comprehensive point kinetics model of an aqueous homogeneous reactor capable of calculating key physical core variables enabling a large number of scenarios to be modelled. Results from various calculations were presented.

Mary Erlund (NNL) then gave an interesting talk on Americium 241 Criticality Safety Studies. Mary explained that there was interest in using Am241 for providing a radioisotope power source for manned space travel beyond Mars. She described a potential source of Am241 was via nuclear fuel reprocessing and Pu241 decay and discussed a number of criticality safety and regulatory issues. Finally Adam Bird of ANSWERS gave a presentation on, and demonstration of, Visual Workshop 3A for MONK10A.

The first presentation after the lunch break was Geological Disposal: Using ICASPA with MONK to calculate critical masses for hypothetical post-closure criticality events, by Robert Mason (AMEC). Robert introduced the programme of work undertaken by AMEC for Radioactive Waste Management Ltd (RWM), a subsidiary of NDA, describing how MONK was used to calculate data for transient models used for consequence analysis.

This was followed by a presentation by Dan Ayres of Imperial College on Uncertainty Quantification in Criticality Calculations using Methods of Polynomial Chaos. Dan described the method he has developed to quantify the effect that nuclear data errors have on keff. The method considers errors in cross sections and nu-bar data for which mean and covariance data are available. The approach is non-intrusive and can be applied to any deterministic code/method. The final talk in the session was Nuclear Data Libraries for MONK10A by Ray Perry (ANSWERS). Ray provided details of the ongoing International Nuclear Data Projects and the current versions of libraries for each.

The final session of the Criticality day started with an interesting presentation by Nigel Davis from ANSWERS who presented A Brief Review of Source Convergence. This talk reviewed various methods that can be used to accelerate and monitor source convergence. This was followed by a related talk by David Long, also from ANSWERS, on Investigating the use of Shannon entropy for source convergence determination in MONK. David concentrated on methods to determine whether the Shannon Entropy is converged and the source has settled. The talk concluded that although these methods could indicate that the source has converged there is still is no definitive method to do this and so the determination of source convergence is still an open issue.

Next was a presentation from Mark Henderson from EDF Energy on Sizewell B New Fuel Containers: Examples of Simplification to MONK input. The talk focused on modernising older MONK inputs to encourage a more user friendly input and to increase the quality assurance. Mark raised awareness of the OVERLAP Part, and used it to good effect in simplifying his model. Finally, Simon Richards presented MONK status and future plans, which closed the ANSWERS seminar for 2014.

Thank you for letting us use cookies to improve your experience of this website. Please read our cookie statement for information on how we use cookies and how to restrict their use.