Develop Uncertainty & Risk Functionality into RESOLVE Model Sample Clauses

Develop Uncertainty & Risk Functionality into RESOLVE Model. The Recipient shall: • Develop updated data ingestion processes to accept data from Task 4, capturing climate impacts on loads & resources, as well as enhancing representation of WECC and California geographic granularity in RESOLVE. o Develop updated temporal samples of the load, wind, solar, and hydro data developed in Task 4 for use in RESOLVE (for reference, the 2021 CPUC IRP RESOLVE Preferred System Plan used a sampled of 37 representative dispatch days). • Develop functionality to run RESOLVE resource portfolios through a range of alternative climate or uncertainty scenarios to assess cost impacts & reliability implications of different resource portfolios. • Extend the RESOLVE formulation to address uncertainty and risk-aversion. The current model defines an expansion plan for a single scenario with deterministic assumptions about uncertainty realization. To properly consider climate uncertainty, in this task, the Xxxxxxxx Berkeley National Laboratory (LBNL) team will develop an extension the RESOLVE formulation, which implies performing the following steps: o Develop functionality to incorporate uncertainty into RESOLVE’s modeling, dependent on the form of data received from Group 2: - If the probabilities associated with the scenarios can be readily estimated, the extension of the RESOLVE formulation will be performed in stochastic optimization fashion. - If no information about probabilities is available, a robust optimization approach will be developed. - If partial information about the probabilities is known (e.g., they can vary within a range), the model will be extended based on distributionally robust optimization, considering ambiguity sets. o Incorporate risk metrics, such as least-worst regret or conditional value at risk, into RESOLVE optimization objective function, enabling RESOLVE to develop “risk-averse” resource portfolios that are resilient against defined extreme events. - Allow risk-aversion in model to be adjusted by user for sensitivity analysis. • Develop reporting metrics to quantify the value of resilience (such as the change in expected system investment & operational costs when considering extreme events). • Implement improved solution methods and problem decomposition techniques to limit the problem size & solution time while preserving model fidelity: o Tests and validate the implementation via 2 steps: - Compare the decomposition results against the undecomposed ones for small data sets where an undecomposed model can ...