Project Title : Power Potantial Project 2
Whole System Operation
Balancing -Virtual Power Plant
Transmisssion led curtailment
Intertripping
ANM - Level 1 Controller
ANM - Level 2 Controller
ANM - Level 3 Controller
ANM - Level 4 Controller
Transmission Operator
Learning Objectives
Year : 2020 (Phase 2 VPP Online)
User Case : Whole System Operation (Virtual Power Plant)
Energy Market - Virtual Power Plant (VPP)
​
This project demonstrates an large distribution network. The system includes an interactive energy market where flexibility services (Active Power and Reactive Power) can be procured as part of an energy market for balancing services.
​
​
​
​
​
​
​
​
Generation Predictor
Main Functionality :
The Predictor provides short and long range predictions of Generation for use in the decision making processes. It uses neural forecasting to predict DER output based on historical data and current weather forecasts , date time and other factors. It supports an active learning process where the neural model is continually learning from historical data to optimise the accuracy of the predictions.
Main Functionality :
Virtual Power Plant
The Virtual Power Plant Module (VPP) can provide ongoing MW and MVAr support services at the point of delivery i.e. the Grid Supply Point. The VPP module determines the amount of MW and MVAr available in realtime from each of the participating DERs , allowing for network constraints and provides an aggregated MW and MVAr
Future Services / Congestion Forecasting
Main Functionality :
Congestion forecasting integrates demand and generation predictions with thermal and voltage constraint management. Driven by the NOVA Prediction engine and live weather forecasts this module identifies possible network constraints and the level of DER curtailment required to mitigate them. The results can then be reported to planning engineers or utilised by market based modules to procure services ahead of time.
Demand Prediction
Main Functionality :
The Predictor provides short and long range predictions of system loads for use in the decision making processes. It uses neural forecasting to predict demand based on historical data. It supports an active learning process where the neural model is continually learning from historical data to optimise the accuracy of the predictions.
