Configuration Management is the process of identifying and documenting the characteristics of a facility’s structures, systems and components (including computer systems and software), and of ensuring that changes to these characteristics are properly developed, assessed, approved, issued, implemented, verified, recorded and incorporated into the facility documentation.

Without Configuration Management, there is no confidence that the nuclear power plant has been designed, constructed, and is being operated in accordance with design requirements, and changes to the plant configuration are consistent with those requirements. Without this confidence in plant configuration, regulators, governmental authorities, and the public will not allow continued operation of an industrial facility that carries the inherent perceived risk of a nuclear power plant.

Configuration Management maintains a precise correlation between all design requirements, regardless of origin, the documentation that supports plant design, maintenance, and operation, and the physical configuration of the plant.

Currently, there is a great degree of variance in the understanding and implementation of Configuration Management in the United States. The current state of the art has evolved based on the cost-prohibitive reconstruction of design requirements, and consequently works with an artificially limited subset of plant design requirements. While considered adequate by US regulatory authorities, it is generally considered inadequate management of design requirements by contemporary Configuration Management experts.

Configuration Management has evolved in the US as it was being discovered that original design requirements for most plants had been lost over time. While some plants were actually discontinued and decommissioned as a result of the loss of design bases, the majority of others pursued expensive reconstruction of design bases. As the fledging CM recovery effort evolved, practices and standards were developed that were constrained by the high cost of CM recovery. Consequently, the current state of CM does not represent a good design for CM, but rather a cost-effective compromise for existing plants. It does not mean the existing NEI taxonomy is ineffective, but rather that the NEI’s taxonomy is valuable only for pre-existing nuclear power plants.

It means that new plants and new design efforts must adopt a new CM taxonomy for effective and low cost CM. Constraining a new design to the existing taxonomy will result in a great deal of duplication of effort, confusion over CM scope, and inadequate design criteria coverage. Using a new taxonomy also means that new software technology can be leveraged to optimize maintenance of the design bases and accompanying documentation.

Currently, the recommended solution would utilize a Configuration Management System (CMS) software system designed around the Williamson-Merritt Taxonomy for Configuration Management.

The Williamson-Merritt taxonomy is a scheme of classification for objects and relationships in nuclear power plant configuration management. It is a hierarchal structure that relates design requirements regardless of origin with individual structures, systems, and components (SSC) within the power plant. It incorporates regulatory, engineering, and owner requirements and provides for management of the relationships to each individual SSC. It also incorporates a Design Bases Ruleset which gives the user the ability to limit the scope of the CM project in a controlled, consistent manner.

Configuration Management System software is a digital repository for storing, maintaining, and evaluating Configuration Management data. It stores not only the data itself, but every change and modification to that data so that historical connections can be evaluated and maintained. It allows multiple sites, locations, and users to simultaneously access great volumes of complex engineering data and make it instantly available and usable.

CMS has the potential to accommodate all aspects of the Margin Management Information System including automatic evaluation and identification of low margin parameters, and automatic generation of the Margin Management Health report.

The CMS Propositum software demonstrates that it is possible to integrate object and relationship management to demonstrate connections between design requirements and facility Systems, Structures and Components. Previously this had not been possible, explaining partially why nuclear power plants had not undertaken this project earlier. It also successfully integrated the Williamson-Merritt taxonomy for Configuration Management as an integrated part of its design.

The CMS Propositum software could not support production usage for many reasons. One, it was a relatively fragile proof-of-concept that did not incorporate either end-user functional requirements or robust validation and verification testing. Two, the interface was, by necessity of low cost, a minimally usable interface, with no human factors or design evaluation incorporated. It was not designed to be used in a production environment, it was designed to demonstrate that the functionality was possible.

The most visible benefit of configuration management is increased confidence in nuclear safety. Effective configuration management ensures that the plant is operated as designed and licensed and that the plant operator can anticipate and manage the operation and maintenance of the plant to preserve the approved facility configuration. These combine to provide fewer challenges by the regulatory authority, and increased confidence on the part of public and governmental authorities that the plant is being operated safely.

From an operational perspective, CM concerns become a thing of the past for the new owner-operator, with associated dramatic reductions in CM related cost, design bases recovery, and regulatory and public operating restrictions.

Margin Management is a subset of effective Configuration Management. Margin Management deals specifically with Design Parameters, and the margins to design and licensing commitments, which are just a fraction of the issues managed in configuration management.