In the last few years, however, we have been faced with a different type of design bases problem; that of designing and implementing a configuration management system during the design and construction of a new nuclear production facility. It has been discovered that existing CMS guidance documents including NEI 97-04, INPO AP-929, and TECDOC 1335 do not address the full range of configuration management concerns predominant in new construction.

A New Challenge

In 2005, work was started at the request of an overseas client to design a functioning configuration management system based on state-of-the-art practices in the industry and complying with US regulatory requirements. Data-gathering and plant visits were conducted to determine the current state of configuration management in the United States.

What was determined was that while plants were actively implementing the existing guidance, there was wide variation in the degree, nature, and level of detail in the implementation. There was a general uniformity of process implementation, but with a wide degree of variance in scope and mechanism.

Consequently, it was determined that we should design a configuration management system and process from scratch that would comply with US regulatory requirements and be suitable for new plant design and construction. We immediately ran into difficulties.

There were multiple critical problems. First, existing methodologies did not provide a mechanism for requirements to flow through documentation into design and actual plant configuration. Secondly, the terminology used was not well defined, resulting in ambiguity and dissent among implementers. Third, the scope of information managed under the configuration management program was not well defined. Fourth, there was no justification for failing to manage non-regulatory requirements with the same rigor we use to manage regulatory requirements.

Over the next few years, we had to revise everything we knew about configuration management at nuclear power plants. We, in essence, had to reconstitute the design bases of the configuration management processes and practices themselves to understand what was valuable to keep and what made sense to modify.

We discovered that the configuration management practices of the day had evolved to respond to evolving regulatory requests constrained by the significant costs of reconstituting design bases information. This is significant - it told us that the configuration management practices have not been designed for configuration management, but have evolved to maximize regulatory compliance while minimizing cost.

It doesn't mean the practices are ineffective, just that they may not account for all design considerations for new plants going forward, which is exactly the problem we were running into.

A New Design

Where did we go from there? To produce a new design we first established a context in which to examine the practice of configuration management. We chose to start with the linguistics, the speech acts that are a fundamental constituitive component of how we as human beings make work happen. We examined configuration management from the perspective of the request-promise cycle, and the conversations for action developed by Winograd and Flores.

From that, we recognized that while the CM processes outlined in contemporary guidance documents were fundamentally sound, we could substantially improve the practice with clarification of terms and mechanisms for connecting SSCs to the design requirements.

This led to the establishment of the Williamson-Merritt Taxonomy for Nuclear Configuration Management, which gave us not only well defined terminology and a mechanism for requirement connection through design documentation to SSCs, but addressed multiple concerns regarding margin management as well. We were confident that by grounding the design of the taxonomy in the way that humans produce work, we were establishing a solid foundation for the future of nuclear configuration management.

The Future is Now!

Finally, simultaneous forces came together making the time right to move forward with the development of the SCMRM. While we were completing the taxonomy, other vested parties were expanding the concept of standardizing the data structures for CM related information to enable data interoperability, the ability to transfer standardized data forms between all interested parties.

Completion of the SCMRM will open the door to a future where an owner-operator can be confident about their compliance with requirements, where EPCs can strip cost from the design and construction of nuclear power facilities, where licensees can eliminate overhead required to reconstitute design bases continually, where vendors and suppliers can provide standardize components at reasonable and profitable prices, and where the plant can be designed, constructed, operated, maintained, and decommissioned just as we promised at the very beginning, over a one hundred year life time of the facility.

It will be a future where the public can trust us to do what we said we would do, that we will provide safe, reliable, inexpensive nuclear generated electricity to the communities around us, the communities which we call home.