Uncertainty Quantification in Wind Plant Energy Estimation

Anna Craig

Research output: Contribution to conferencePaper

Abstract

Data characterizing a wind power plant's production is important for the validation, calibration, and refinement of the wake loss models used to simulate performance for pre-construction energy estimates. In this process, there is the potential for significant variation in the methods used by an analyst for, first, the preparation of the operational production data for comparison to model data and, second, the generation of analogous production data from simulations. In this work, the uncertainty introduced by these method variations is quantified by considering ensembles of analysis methods. Specifically, 1920 methods variations were developed for preparing the operational data as a predictable production benchmark and simulations taking as input sample sets of the observed wind resource ranged in size from covering 4% to 125% of a full year were run using 3 engineering wake loss models in FLORIS simulations. The benchmark and simulation input uncertainties place lower bounds on the precision of the calibration between the simulation model estimates and operational benchmark data. Even within subsets of method variations (representative of what might be seen within a single company), the uncertainty in the predictable expected production is estimated to be between 1% and 3% the median predictable expected production while the uncertainty in the simulated expected production remains at best on the order of 2%. To put this into context: the simulated expected production is very closely related to the annual energy production (AEP) metric commonly used to quantify a plant's predicted performance and a 3% error in the pre-construction AEP estimate can result in a $17M loss due to additional financing costs for a typical 100MW Texas plant.
Original languageAmerican English
DOIs
StatePublished - 2019
EventAIAA SciTech 2019 Forum - San Diego, California
Duration: 7 Jan 201911 Jan 2019

Conference

ConferenceAIAA SciTech 2019 Forum
CitySan Diego, California
Period7/01/1911/01/19

Bibliographical note

See NREL/CP-5000-72741 for preprint

NREL Publication Number

  • NREL/CP-5000-74482

Keywords

  • energy analysis
  • model calibration
  • uncertainty quantification
  • variations
  • wind plant

Fingerprint

Dive into the research topics of 'Uncertainty Quantification in Wind Plant Energy Estimation'. Together they form a unique fingerprint.

Cite this