The 20th Anniversary ANSWERS Seminar was again held at the Wessex Hotel in Bournemouth, a location near to the superb Dorset coast and in close proximity to the town centre. Over sixty 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 and the UK Working Party on Criticality were hosted at the same venue.
The Technical Programme began with Radiation Shielding and an overview of LaunchPad, the graphical tool used to launch MCBEND, MONK and RANKERN, with particular emphasis on the ‘verification tool’ and ‘queue tool’ options within LaunchPad. This was followed by a more technical talk on how to run a Point Energy Adjoint calculation with MCBEND 10 – this being a new option within MCBEND 10. Two application talks followed – the first using MCBEND to model the response of detectors used to investigate the effects of material degradation in cooling ponds and the second to determine dose-rates following a hypothetical accident in cooling ponds. An interesting talk on the use of embedded files and other input utilities was given, which demonstrated how these could be used to improve QA when running a large number of similar calculations. Later in the morning two presentations were given on the RANKERN Point-Kernel code. The first giving advice and recommendations when using RANKERN for scatter and reflection calculations and the second on analysis of a skyshine experiment using RANKERN. The morning session ended with an update of validating the new BINGO neutron processor in MCBEND.
The afternoon started with the ever-popular demonstration of the new features of the VISTA suite of graphics codes. In particular the latest VISTA-TRACK was demonstrated and was very well received, as well as new features and current developments in VISAGE, VISTA-RAY and LaunchPad. Later in the afternoon the use of the ANSWERS Customer Web Pages were demonstrated. Between the demonstrations a talk on modelling irregular shaped sources using options within the new Unified Source module in MCBEND 10 was given. Application talks in the afternoon session included neutron flux calculations for Magnox decommissioning using MCBEND and the use of MCFANG for medical applications – including the generation of complex voxelised phantom models using the XYZMESH hole. The day ended with a presentation on a feasibility study undertaken on a CAD/MCBEND link and an overview of the current and future developments in the ANSWERS Shielding area. Overall there were a good variety of presentations, which gave a very good start to the Seminar.
The second day of the meeting, the Reactor Physics day, started with a presentation on work to provide a reference continuous energy burnup option in MONK. This was followed by a presentation on 20 years of progress in reactor physics methods at British Energy (BE). This presentation, specifically for the 20th anniversary of ANSWERS, took an at times humorous look at changes at BE, concluding that their greatest success was in the number of times they had managed to change their company name. Some serious points were made, however, showing the development and improvement in methods over the years. The final presentation before the morning coffee break was on the status of the JEFF3 project. Results using JEFF3.0 gave low k-effective predictions, particularly for LWRs. Results from JEFF3.1, employing a revised U238 cross-section, were an improvement and those from JEFF3.2, including several other cross-section revisions, gave good agreement with results from many different spectral reactor types.
In the second session, following the break, two guest presentations were made, the first by Belgonucleaire and the second by HMS Sultan. The first presentation by Benoit Lance covered an overview of the application of WIMS at Belgonucleaire and included a discussion of several different topics ranging from advanced fuel design to a novel application of the assessment of stress in MOX fuel pellets using the WIMS PROCOL module to determine heating in the granular structure of the fuel. Professor Beeley from HMS Sultan reviewed ten years of using ANSWERS codes in support of academic training for the naval nuclear propulsion programme. The final presentation before lunch was by Serco Assurance on tracking in the WIMS CACTUS module and its extension to three dimensions.
The afternoon started with a presentation on developments in the MAX Monte Carlo perturbation route including the removal of mesh structure dependence in the solution and an improved convergence behavior. This was followed by two guest presentations. The first by Pavel Mikolas of SKODA, discussed the influence of enrichment profile in the design of a gadolinium poisoned assembly for the VVER-440 reactor to reduce local pin power peaking and improve overall core performance. The second by Matthew Eaton of Imperial College described advances in methods for criticality and reactor transient analysis. The first part of the presentation gave an overview of the capabilities of the FETCH code. This concluded with a brief description of advances in methods for local solution refinement to remove artificial non-uniformity in solutions. This was followed by the afternoon coffee break.
Three further guest presentations followed the break. Jim Kuijper of NRG Petten described a study on the feasibility of burning first and second generation PWR UO2 recycled plutonium in pebble bed HTRs as part of studies within HTR-TN. Further studies on dynamic behavior are required to complete the investigation but conclusions so far are that Pu from first generation recycling can be burnt effectively while burning Pu from the second generation is less favorable. James Brushwood of HMS Sultan presented a study on a high temperature gas cooled reactor for marine application. The design exhibits generally good behavior in passive heat removal but requires decay heat removal in the case of a loss of forced convection accident. Christophe Schneidesch of Tractebel completed the guest presentations with the description of the analysis of the FWLB accident scenario using a coupled PANTHER-RELAP methodology.
The day was completed by a short presentation on future plans for the enhancement and development of the WIMS code. In the software demonstration; informal discussion; problem clinic session following the days presentations a demonstration was given on the structure and contents of the ANSWERS Internet site.
The final day of the 20th ANSWERS Seminar comprised the Criticality related technical presentations. Once again there was good representation from different areas of the criticality community: analysts, assessors, regulators and code developers.
The technical presentations began with a description of a new MONK development to model explicitly arrangements of dropped rods using a new hole geometry and some preliminary results. This was followed by a description of refurbishment projects at Springfields Fuels Ltd and examples of MONK and CODEMORE being used to analyse possible scenarios. In a change to the programme the MONK9A new features presentation was postponed until the afternoon.
After the morning break the presentations continued with a Nirex report on repository criticality safety assessments, including references to more realistic scenarios as well as the typical bounding case results. This was followed by a regulator’s view of the approval process for nuclear material, and included some explosive footage of transport flask testing, the unforgettable train test and the more rigorous drop tests. A nuclear data presentation followed that highlighted the improvements to be gained by using the latest JEFF3.2 U235 and U238 data in the BINGO libraries with MONK. The final talk before lunch was an overview of LaunchPad, the graphical tool used to launch MONK, with particular emphasis on the ‘verification tool’ and ‘queue tool’ options.
After lunch the programme continued with a description of a prototype Genetic Algorithm code that was successfully used with MONK8B to determine parameter values for specified target criteria in a small range of systems. This was followed by a presentation on the difficulty of determining concrete compositions used in plant and the effect on k-effective, and included a demonstration using debris from the beach and hotel car park. A presentation on the use of embedded files and other input utilities was given, highlighting how these could be used to improve both user efficiency and QA when running a large number of similar calculations. This session concluded with the overview of the new features of MONK9A (postponed from the morning session), and highlighted some recent new developments in hole geometries and input utilities.
The final session on the final day of the 20th ANSWERS seminar began with a talk on the use of CODEMORE to generate multiple MONK runs on a BEOWOLF cluster, and how this arrangement can be used easily to generate families of curves from which parameter sensitivity can be inferred and target criteria identified. The session closed with a summary of the current status and future plans for MONK, including details of the next code release and a request from ANSWERS to users for development ideas that will aid them in their work.