Unit 1 Return To Service - experiences in managing a $1 billion project

Traditionally, OPG-N has used task work-down curves to measure performance against plan. This approach was used on the Unit 4 Return to Service project and did not prove useful in predicting the schedule slippage that occurred. For the Unit 1 Return to Service (RTS) project, the decision was made to use earned value reporting and analysis to enhance the existing performance reporting. The Unit 1 RTS project was broken down into fourteen sub projects along equipment or physical plant areas and five major work group areas. This formed the basis of weekly reporting for both schedule and cost performance index (SPI and CPI). The earned value reporting provided a number of benefits over the existing performance reporting methods. SPI provided an early predictor of potential schedule slippage and initiated management corrective actions long before critical path analysis indicated any potential problems. Both SPI and CPI provided better indicators of project performance than seen by task work-down curves and were the basis for consistent reporting to executive level management. Earned value reporting was extended into more non-traditional areas and the design engineering work program was developed and reported using SPI and CPI. As can be expected on projects of this nature, where equipment has laid dormant for a number of years, there was some scope growth that necessitated adjustment of the original baseline schedule. A decision was made and a methodology developed to renormalize the Unit 1 RTS schedule after the scope growth reached 5% (after six months of work). Renormalization allows the additional scope to be added to the schedule without losing the project performance history. The overall use of earned value analysis and reporting resulted in improved understanding of project status, initiated early corrective actions where necessary and provided a consistent reporting format for executive overview. This approach is also applicable to Unit planned outages. (author)