Wind Plant Performance Prediction (WP3) Benchmarking Project
The Wind Plant Performance Prediction (WP3) Benchmark project is a validation exercise for wind farm pre-construction energy assessments. This project represents an unprecedented platform for data sharing and wind industry advancement including preconstruction and operational data from up to 100 or more modern operational wind projects.
- Generate accurate, independent benchmarks of pre-construction energy assessments
- Improve accuracy and reduce uncertainty in pre-construction energy estimates
- Create a platform for sharing data to advance the state-of-the-science for pre-construction energy assessments
For years, the wind industry has struggled with poor accuracy in energy estimates for new facilities. This has impacted our credibility with investors, and increases risk for project owners. Over time, large consultants with access to operating wind farm data have been able to validate their methods by comparing predicted to actual production. However, each consultant has a different dataset for validation and different methods and definitions for what should be eliminated from the study, making it hard for the rest of the industry to know what to make of the results. In addition, newer entrants to the WRA market have emerged but have been largely unsuccessful at gathering operating wind farm data in volumes sufficient to perform a meaningful validation exercise. Industry progress in the WRA field has slowed as a result - The average one standard error uncertainty for US wind projects is 7%. When a 3% deviation in energy from P50 means $17MM in NPV (typical 200MW project in Texas), this is not an acceptable level of uncertainty.
Additionally it is important to continue to reduce the Levelized Cost of Energy (LCOE) from wind plants due to increasing price pressure in US markets. One such way to reduce the LCOE is to reduce the risk profile for capital investors and thereby reduce the risk premium as well. Early estimates indicate that there is an opportunity to reduce the LCOE by ~10% through risk reduction activities.
As individual organizations, we can have a limited impact on this problem due to sample size and organizational risk. As a group, we can put a statistically significant number of projects in the hands of scientists that want to solve this problem of poor accuracy and high uncertainty.