SHORT-TERM TERM WIND POWER FORECASTING: A REVIEW OF STATUS AND CHALLENGES Eric Grimit and Kristin Larson 3TIER, Inc. Pacific Northwest Weather Workshop March 5-6, 2010
Integrating Renewable Energy» Variable generation driven by the weather adds many challenges for meeting energy demand. Managing the reserve electric capacity and transmission systems becomes more difficult. Example: Wind power in the Bonneville Power Administration (BPA) energy balancing authority area (mostly in the Columbia Basin) Wind power ramp down opposite of load ramp up
Short-Term Wind Power Forecasting» State of the practice: Hours to Days Forecast Horizons» Numerical weather prediction with locally trained statistical post-processing (e.g., MOS) Minutes to Hours Forecast Horizons NW P On- Site Data Off- Site Data» Autoregressive statistical models and supervised machine learning techniques» Blending with short-term NWP model output» Adaptive predictor selection for large input data sets, including off-site meteorological observations» Regime-switching models Trained to minimize bulk errors for average power over all forecast intervals (e.g. RMSE over 1-hour periods for 1 yr).
Large Set Predictor Selection» Input Data Sources: On-site private wind project, met tower, and turbine data Nearby public met tower data Nearby private wind project and met tower data ASOS and meso-network (MADIS) stations within 300km o NWP model column output» Variables: Meteorological variables: o Wind speed Wind direction Temperature Pressure Large # of NWP model output fields Derived variables: Pressure differences Lags up to 24 hours o Time derivatives» Total Predictor Set: O(100,000) predictors, most useless and/or correlated Requires massive dimension reduction or memory-efficient sparse matrix methods: Dimension reduction: pick top few lagged predictors first for each measurement instrument separately o Memory efficient joint selection: use a LARS, Elastic Net approach
Regime Switching Models» Models applied to wind power forecasting Wind direction regime switching space-time model» Gneiting, Larson, et al. (2006) Trigonometric Direction Diurnal Model» Hering and Genton (2009) Markov switching auto- regressive e model» Pinson and Madsen (2009) Evolution of regime marginal probabilities using a Markov Switching Auto- Regressive (MSAR) statistical model for 10-minute, 1-step ahead prediction of wind power at Horns Rev (Denmark). (Pinson and Madsen, International Journal of Forecasting, 2009)
Standard Wind Power Forecast Skill» Location Wind Project A (Columbia Gorge, WA)» Time Period and Forecast Details Training Period: 03-May-2007 17-Aug-2008 Test Period: 18-Aug-2008 18-Aug-2009 Forecast Interval: 60 minutes (at HH:00) Forecast Leads: 120, 110,, 70 minutes Wind Project A Exp. # (input data) Train % Imp. RMSE 1(on-site) 10.0 10.2 2 (on-site, off-site) 15.5 14.6 3 (on-site, off-site, ASOS) 23.2 16.4 Test % Imp. RMSE RMSE skill score (% Imp.) measured relative to 1-hr persistence forecast
Wind Power Ramp Forecasting» Usual goal is to minimize large deviations forecasts are optimized to be conservative» Yields smooth forecasts under-prediction of ramps
An Alternative Metric Event-Based Scoring» With categorical (yes/no) forecasts of a binary event, there are four possible outcomes.» Costs can be associated with each outcome. Ramp Forecast Yes No Ramp Observed Yes No Hit False Alarm Miss Correct Negative
Example Wind Power Ramp Forecast Skill» Location Wind Project B (Columbia Gorge, WA)» Time Period and Forecast Details Test Period: 2006-2007 (6 months) Forecast Interval: 60 minutes (at HH:00) Forecast Lead: 1 day» Location Wind Project C (Columbia Gorge, WA)» Time Period and Forecast Details Test Period: 2007 Forecast Interval: 60 minutes (at HH:00) Forecast Lead: 1 hour Event Ramp Size: > 20% of capacity FCST YES NO OBS YES 281 1534 NO 637 5766 CSI (threat score) = Hits / ( Hits + Misses + Falses) = 0.11
Challenges for Ramp Event Prediction and Validation» Timing Example: Day-ahead forecasts of wintertime frontal passage on the west coast can be off by > 6 hr» Magnitude Example: Small errors in offshoreonshore pressure gradients can result in poor land/sea breeze intensity it» Location Example: Thunderstorm locations are difficult to forecast due to poor simulation of their initiation, development and decay processes» Frequency A ramp event is a rare event At the project level, events of 20% or more occur less than 10% of the time State-of-the-art forecasts have more false alarms and missed events than hits
Wind Power Ramp Event Capture Alberta Electric System Operator Wind Power Forecast Pilot Study (2007-08) Forecaster CSI (threat scores) Alberta Region Forecaster (Source: ORTECH Power, Wind Power Forecasting Pilot Project Part B: The Quantitative Analysis Final Report, 2008)
A Non-Deterministic Approach Goal: Quantify all the Uncertainties Example Ramp Event Probability Forecast and Observation Time Series DAY AHEAD RAMP EVENT PREDICTION 2-month hourly time series (Jan-Feb 2007)» System developed for BPA research project in 2007» Utilizes NWP ensembles for day-ahead ramp event prediction» Includes statistical calibration (e.g., Bayesian model averaging)
Integrating the Ramp Forecast» Goals: User-defined thresholds trigger alerts Full probability distribution input into stochastic unit commitment and economic dispatch system System operation can automatically optimize use of reserve capacity and avoid transmission i bottlenecks A proactive SmartGrid! Risk of down ramp event above users thresholdh
Challenges (Future Work)» Improved Ensemble NWP in Short-Range (0-6 hrs) Assimilation of more on-line data» Weather Regime Detection and Transition Prediction Concepts from low-frequency atmospheric regime transition studies» Large-Set Predictor Selection Dimension reduction Mutual information criterion Alternative metrics (asymmetric, event-based scores, etc.)» Forecast Integration and Utilization Packaging for energy management systems Human factors