Introduction

Heat‑pipe microreactors (HPMRs) represent a class of advanced microreactor concepts designed to deliver reliable and flexible power in remote, mobile, or austere environments. These systems leverage passive heat‑removal technology based on heat pipes (sealed, two‑phase thermal devices that transport heat through the evaporation and condensation of a working fluid). Heat pipes enable extremely high effective heat conductance with minimal temperature gradients and require no pumps, moving parts or external power, providing exceptional reliability and simplicity. Their passive nature makes them particularly well‑suited for microreactors, which often must operate autonomously for long periods.

Heat‑pipe microreactors typically consist of a core and an array of heat pipes embedded throughout the structure that extract heat directly from the fuel region and transport it to a heat sink or power conversion system. Because heat removal does not rely on coolant flow or complex thermal‑hydraulic loops, HP‑MRs can achieve high reliability, compactness and ease of transport, making them attractive for use in remote installations, disaster‑relief operations, and potentially extraterrestrial outposts.

To support the international community in advancing modelling and simulation capabilities for these systems, the NEA has launched the Heat‑Pipe Microreactor (HP‑MR) benchmark. Conducted under the Working Party on Scientific Issues and Uncertainty Analysis of Reactor Systems (WPRS) and its expert groups (EGPRS, EGTHM, EGMUP), the benchmark provides a structured framework for comparing neutronics‑only and fully coupled multiphysics predictions on a realistic, openly published 2 MW_th HP‑MR design. The design includes TRISO fuel compacts, metal‑hydride moderator pins, graphite matrix and sodium heat‑pipes. The model has been openly published through the Virtual Test Bed, enabling broad participation within OECD countries.

The benchmark strengthens collaboration, helps identify modelling gaps and supports the development of best practices for analysing these emerging reactor systems.

Objectives

The HP‑MR benchmark aims to:

• Compare neutronics‑only and coupled multiphysics predictions across a realistic microreactor configuration.
• Identify best practices for steady‑state and transient multiphysics modelling of heat‑pipe microreactors.
• Identify gaps or weaknesses in coupling strategies and modelling capabilities.
• Facilitate international collaboration on microreactor analysis and enable knowledge transfer to universities, research organisations, vendors, and regulators.
• Provide materials that can support training and education activities within existing NEA programmes.

Scope and schedule

The activity spans 2026–2029 and is organised into three phases:

Specification development (2025-2026)

• Preparation and distribution of detailed benchmark specifications.
• Iteration with participants to ensure clarity and feasibility.

Phase 1: Year 1

Phase 1A: Neutronics‑only analysis

• Comparison of k‑eff and power distributions for several core configurations.
• Overseen by EGPRS.

Phase 1B: Heat‑pipe analysis

• Comparison of thermal and heat‑pipe behaviour.
• Overseen by EGTHM.

Phase 2: Year 2

Steady‑state multiphysics comparison

• Coupled neutronics and thermal‑hydraulics.
• Overseen by EGMUP.

Phase 3: Year 3

Transient multiphysics comparison

• Includes load‑following scenarios and heat‑pipe failure events.
• Overseen by EGMUP.

A final consolidated report will be published at the end of the project.

Co-ordinating organisations

• Lead organisation: Argonne National Laboratory (United States)

Participation

Institutes interested in participating in the HP‑MR benchmark may request access to the working area and mailing list by filling out the HP-MR conditions form and returning it to the WPRS Secretariat. Contact: wprs@oecd‑nea.org

Participation is open to OECD member research institutions, universities, vendors, and regulators, with both active contributors and observers welcome.

Working area

Benchmark specifications, input data, templates and results are available in a password‑protected NEA working area.

Details, including specifications and results templates, can be found in the working area: 

HP‑MR Benchmark (requires password | reminder)