Pressurised Water Reactor Main Steam-Line Break (MSLB) Benchmark

This benchmark incorporated full 3-D modelling of the reactor core into system transient codes for best-estimate simulations of the interactions between reactor core behaviour and plant dynamics and their testing on a number of transients of importance for plant behaviour and safety analysis. This included verifying the capability of codes to analyse complex transients with coupled core-plant interactions, to fully test the 3D-neutronics/thermal-hydraulics coupling, and to evaluate discrepancies between predictions of coupled codes in best-estimate transient simulations.

Publications and reports
Description of the problem

The benchmark was based on a well-defined problem concerning a pressurised water reactor main steam line break (MSLB), which may occur as a consequence of the rupture of one steam-line upstream of the main steam isolation valves. This event is characterised by significant space-time effects in the core caused by asymmetric cooling and an assumed stuck-out control rod during reactor trip. It is based on reference design and data from the Three Mile Island Unit 1 Nuclear Power Plant (TMI-1). It included a description of the event sequence with set points of all activated system functions and typical plant conditions during the transient.


It consisted of three exercises:

  1. Point kinetics simulation to test the primary and secondary system model responses
  2. Coupled 3-D neutronics/core thermal-hydraulics response evaluation using inlet and outlet core transient boundary conditions
  3. Best-estimate coupled core – plant transient modelling

The first two exercises helped to tune the models in the different codes in order to ensure they all solve the same problem; parametric studies and scenarios were developed to help understand the source of uncertainties. A suite of statistical methods had been applied to analyse code-to-code comparisons involving different types of data – single values, 1-D and 2-D distributions, and time histories. The statistical methods had been modified to correctly analyse relative normalized parameters.

Sponsors and participants

The exercise was co-organised by the OECD/NEA Nuclear Science Committee  and the Committee on the Safety of Nuclear Installations  and the US Nuclear Regulatory Commission. It involved around 70 experts from 15 countries representing 30 organisations. It brought together specialists in neutronics and thermal-hydraulics from universities, research centres, utilities, engineering companies and vendors. They worked together on this project. It was co-ordinated by the United States Pennsylvania State University Nuclear Engineering Programme Team.

Conclusions and recommendations

The exercise had shown that:

  • a proof of principle that coupling 3D neutronics with thermal-hydraulics is feasible and working
  • 3-D coupling provides more detailed insight into phenomena occurring in the core during transients, required for engineering simulations: power plant operators seek to know what happens in details during transients
  • best-estimate methods provide margins to safety limits, allowing more flexibility in plant operation
  • best-estimate methods can be used both for reactor operation and safety analysis and that tools common to both will emerge
  • its timely organisation had not only achieved a comparison of the performance of different codes but had driven the development of coupled 3-D neutronics/thermal-hydraulics codes, in particular, optimal coupling schemes through parametric studies
  • it also provided a template for multi-level benchmark methodology to be used for complex problems;
  • there is a need to develop a common approach for sensitivity/uncertainty analysis in neutronics and thermal-hydraulics.

Related data

Data related to the MSLB Benchmark can be requested from the NEA Data Bank:

External references

J.Bryce Taylor, Kostadin N Ivanov, "Statistical Methods used for Code-to-code Comparisons in the OECD/NRC PWR MSLB Benchmark", Annals of Nuclear Energy, Volume 27, Issue 17, 2000, pages 1589-1605,  doi:

B. Taylor, K. Ivanov, D. Ebert: Comparative Analysis of Best Estimate Solutions of the OECD/NRC MSLB Benchmark, Proceedings of the ANS/ENS Embedded Topical Meeting BE2000 – “Best Estimate Methods in Nuclear Installation Safety Analyses”, November 12-16, 2000, Washington D.C., USA,


Ian Hill