Advanced Core Engine (ACE) is a Westinghouse‑developed AI and machine‑learning tool for core loading pattern optimization that accelerates and improves nuclear fuel core design. Using a surrogate model trained on high‑fidelity neutronic simulations, ACE rapidly evaluates large design spaces to deliver faster insight, improved margins and stronger economic performance.
ACE is designed to reduce engineering cycle time, expand the search space beyond traditional methods and support more efficient in-core fuel management across multiple reactor types.
Optimizing core loading patterns is a computationally intensive, high‑dimensional problem with millions (or more) of possible configurations. Traditional approaches are:
These constraints can restrict optimization quality and lengthen reload design critical path, especially as plants pursue longer cycles, higher discharge burnup and tighter design constraints.
ACE Graphic User Interface (GUI)
ACE replaces brute‑force iteration with a machine‑learning surrogate model that rapidly predicts key core design metrics, enabling fast and scalable optimization.
Core elements include:
This architecture allows ACE to explore design spaces that would be impractical using conventional workflows.
ACE is built to optimize loading patterns while satisfying all applicable safety, operational and energy constraints, including:
The result is better‑balanced cores with improved margins and economic performance.
On widely accessible computing platforms — with scalable performance on higher‑end systems — ACE enables orders‑of‑magnitude more exploration in a fraction of the time:
This speed allows engineers to:
ACE currently supports, or is being extended to support, the following reactor technologies:
With ACE, customers can expect:
ACE complements existing Westinghouse core design tools and expertise, acting as a force multiplier rather than a replacement.
