The 2017 ANSWERS Seminar was held once again at the Haven Hotel in Sandbanks, Poole, over the period 23rd to 25th May. It was attended by over 100 delegates from across the UK and Europe.
The three day event covered Shielding, Reactor Physics, and Criticality, with a technical programme which included presentations from both ANSWERS staff and customers.
The 2017 ANSWERS Seminar began on Tuesday 23rd May with the Radiation Shielding Day. Following a brief introduction and welcome by the ANSWERS Manager, Paul Smith, an overview of the Seminar and the Shielding Day was given by Adam Bird (Amec Foster Wheeler).
The first session began with Jayne Eratt (Rolls Royce) talking about modelling of ASPIS experiments for MCBEND11A validation using JEF3.1 Nuclear Data. The second talk was given by Christophe Murphy (Amec Foster Wheeler), a ‘how to do’ presentation on Heating Calculations in MCBEND. The next talk was an update on parallel and re-structuring developments in MCBEND, RANKERN and MONK given by Adam Bird and Pietro Santagati (Amec Foster Wheeler). The presentation covered some of the challenges faced in implementing an efficient parallel functionality, including the removal of legacy memory management code. Pietro followed this with an account of experiences implementing and using the Marsenne Twister random number generator, which will be available in future ANSWERS codes.
The first talk of the second session was an update on Visual Workshop developments by Tim Fry (Amec Foster Wheeler), describing features of the imminent 3C release. This was followed by a description of development work on the Random Hole by Geoff Dobson (Amec Foster Wheeler) which will improve the random nature and increase the achievable packing fraction of this Hole Geometry. The next talk was given by Mathew Parkinson (Bath University), and described a ‘Multilevel Monte-Carlo’ approach for Uncertainty Quantification - an interesting description of Monte-Carlo used outside the context of particle transport. The last presentation of the morning was an update on CAD import, including a demonstration of using Visual Workshop to translate an OBJ format CAD file into FG, given by Adam Bird.
The first afternoon presentation, after a great lunch provided by the Haven Hotel, was on Deterministic Methods for Shielding, given by Brendan Tollit (Amec Foster Wheeler). This covered existing and under-development deterministic and hybrid methods. The next two presentations covered the use of MCBEND for Advanced Gas-Cooled Reactor calculations: The first, presented by George Wright (Amec Foster Wheeler), with permission from EDF Energy, was a paper from the 16th International Symposium on Reactor Dosimetry, ‘A Case Study into the Effect of Excess Reactivity on the Results of Dosimetry Calculations Performed with Fixed Source Codes’. The second, presented by Dennis Allen (Amec Foster Wheeler), with permission from EDF Energy, was entitled ‘Using MCBEND to Derive Ex-Core Detector Responses for an Advanced Gas-Cooled Reactor’.
The next, visually pleasing and technically informative, presentation was given by Andrew Buchan (Imperial College), on the use of reduced order modelling with POD basis functions created for a particular class of problems for accelerating FETCH2. Finally, the Shielding Day was brought to a conclusion with a ‘Status and Future Plans’ presentation, given by Adam Bird.
The Reactor Physics day of the 2017 ANSWERS Seminar was held on Wednesday 24th May. The day was split into four sessions.
The first session began with a presentation from Tom Taylor (EDF Energy). Tom discussed AGR heating calculations using WIMS10. These considered use of the new WIMS heating models as a potential way to reduced observed channel power discrepancies. Pavel Mikolas (SKODA JS a.s.) gave the next presentation. He discussed initial WIMS analysis of a benchmark problem considering an in-core SPND in a VVER. Estimates of 103Rh concentrations and reaction rates in the SPND were made. The following presentation was given by Alan Charles (University of Cambridge). Alan presented an overview of Cambridge projects that use ANSWERS software; this included applications to PWR, BWR, MSR, SFR and THGR modelling. The final talk of the session was given by Peter Smith (Amec Foster Wheeler). Peter described new features in WIMS11, particularly those related to core-scale modelling and multiphysics. He also presented new features in Visual Workshop to support the display of geometries and results from the new modules in WIMS11.
The second session began with a presentation by Ben Lindley (Amec Foster Wheeler). Ben discussed recent developments to the WIMS/PANTHER route for BWR modelling. The route now contains all essential functionality and is ready to be validated against plant data. Next, Ed Wickens (University of Cambridge) described WIMS/PANTHER analysis of the IRIS SMR. Ed concluded that the WIMS/PANTHER route was capable of modelling such PW-SMRs. Eddie Birkett (EDF Energy) gave the following presentation; he discussed validation of loading graphite vacancies in end-of-life AGR cores. WIMS and PANTHER calculations were used to analyse the effects of introducing vacancies, with a view to optioneering end-of-life fuel management. The last presentation of the session was given by Brendan Tollit (Amec Foster Wheeler) and Alan Charles (University of Cambridge). They discussed recent developments and testing of the new SUBCHANNEL module in WIMS11. The testing had considered comparisons against other thermal-hydraulic codes for single-pin and bundle problems from an OECD benchmark. Good agreement was observed between SUBCHANNEL and the other codes.
The session following lunch contained three presentations from Amec Foster Wheeler staff on recent developments to the uncertainties toolkit in WIMS11. First, David Powney described new functionality in the HEAD module to allow perturbations to be made to infinite dilute cross-sections. The perturbations are made in a consistent manner, such that resonance integrals are adjusted in line with cross-sections. Glynn Hosking then described an application of the cross-section perturbation method, so that uncertainties of k-infinity to cross-section data could be assessed. This application makes use of new functionality in the PERT module to derive sensitivity coefficients and combine them with covariance data. In the final presentation of the session, Ben Lindley described a route in WIMS11 and Visual Workshop to propagate nuclear data uncertainties from WIMS to PANTHER. The route uses WIMS to generate sensitivity coefficients for flatfile parameters. Visual Workshop uses these sensitivity coefficients, alongside sampled microscopic cross-sections and covariance data, to generate sets of perturbed flatfiles for use in PANTHER calculations.
The final session began with a presentation by James Dyrda (OECD-NEA), who described the Nuclear Data Sensitivity Tool (NDaST). This makes uses of sensitivities calculated for IRPhEP and ICSBEP benchmarks to facilitate rapid scoping of the effects of nuclear data changes in a given system. Simon Richards (Amec Foster Wheeler) then gave an update on new reactor physics features in MONK10B. Next, Claire Heaney (Imperial College) gave a presentation on rapid reactor physics calculations using a reduced order modelling method. The aim of the work is to provide accurate real-time methods for use in a reactor simulator. The final talk of the day was from Glynn Hosking (Amec Foster Wheeler), and covered the current status of WIMS and potential future developments.
The Criticality Day was held on Thursday 25th May. It began with an overview of the day from Robert Mason (Amec Foster Wheeler), followed by the presentations, given in four sessions.
The first session began with Simon Richards (MONK Design Authority) describing the current status of MONK, developments and enhancements in the upcoming release (MONK10B) and future developments. This was followed by an insightful presentation by Anton Murfin (NNL) describing requirements to consider the effects of temperature in criticality safety assessments for transport applications, and how those requirements are being approached, including for low temperatures. This presentation prompted a lot of interest and several questions. Following in a similar theme, Tim Ware (Amec Foster Wheeler) provided a summary of work being carried out to consider lower temperature nuclear data, including challenges in obtaining such data, validating it and making sure that software can use it. The first session concluded with a presentation from Pietro Santagati (Amec Foster Wheeler), giving an overview of how an enhanced random number generator and a bespoke arrangement for parallelisation of the ANSWERS Monte Carlo codes are being considered and developed.
The second session began with Geoff Dobson (Amec Foster Wheeler) providing an update on research to use Maximum Likelihood Estimation for parameter estimation and also an overview of work considering how Extreme Value theory can be used to estimate the probability that keff is below a threshold. This was followed by an interesting presentation from Mark Henderson (EDF Energy), describing some of the criticality safety considerations when assessing dropped PWR fuel, including the challenge of whether an ‘obviously bounding’ model really is bounding. The third presentation of the session, given by Alfie O’Neill (NNL), provided a summary of two significant modelling projects using MONK, including validation studies at room and elevated temperatures, where good agreement with benchmarks was found, and the use of the TETMESH facility and some of the challenges faced. This prompted discussion on how the POLYBODY capability might be used for such systems. For the final presentation of this session Tim Fry (Amec Foster Wheeler) provided an overview of new features and developments in Visual Workshop, with a request for customers to feedback any comments or requests for further developments. Tim provided some guidance on specific queries after the presentation.
Following a break for lunch, the third session continued the overview of Visual Workshop developments, with David Long (Amec Foster Wheeler) giving a worked example of how the criticality search options can be used to seek a target value (for keff) given variability of a parameter, and how this can also be used with looping for a second parameter at specified values. The second presentation of this session, by James Dyrda (OECD-NEA), gave a thorough insight into studies to generate and understand benchmark correlations for ICSBEP and DICE, including a summary of some of the challenges and the progress being made. The third session was concluded with a presentation by Christophe Murphy (Amec Foster Wheeler) on developments to tools to understand source convergence, providing more information on the diagnostic tools, and building on the overview given by Simon Richards at the start of the day.
The fourth and final session of the day began with a presentation by Paul Smith (Amec Foster Wheeler), who described the new uncertainty sampling capabilities to be released with an update to Visual Workshop. Using an example, Paul showed how uncertainties in keff due to variability in input parameters and nuclear data could be understood. The second presentation of this session was given by Jim Gulliford (OECD-NEA), who provided an overview of NEA Nuclear Science and Data Bank Activities, including details of the Nuclear Education Skills and Technology (NEST) Framework which seeks to build and maintain nuclear expertise via established experts working alongside the next generation of young nuclear professionals. The final presentation of the session was given by Robert Mason, who described some of the limitations of the current version of the Random Hole capability and how a number of enhancements have been considered for the future. Paul Smith concluded the day, and the 2017 seminar, by thanking the attendees and presenters, wishing all a safe journey, and hoping to see everyone again in 2018.