The key to accurate forecasting is the model’s ability to accurately model reality. Validating the accuracy of the model is the first step in forecasting. iPool’s iView facility shows comparisons of actual vs. simulated prices for checking model validity. iPool has the following relevant features for market analysis and forecasting:
- Uses actual market provided bid-offer data files with minimal to no manual processing required.
- iPool’s Bid Aggregator creates typical bid offers for different calendar days from historical bid offers.
- iPool can auto-detect new incoming units and can extract relevant forecast parameters from historical scenarios that are user modifiable for forecasting future scenarios.
Intelligent Bid Behavior Modeling
Generator bid offers change from day to day and it can be dependent on the participant’s contract portfolio, the system demand, the system constraints and other market conditions. Using the actual historical bid offers may not be applicable for forecasting future scenarios. iPool captures the typical bids of all participants for different calendar days and with its iBid module can model the dynamic responses and adjust them to the prevailing market conditions during simulation.
iPool’s use of Object Oriented technology enables complex and flexible modeling of market events. These events can include changes in supply capacity, demand, limits, storage levels, price limits and even market rules. iPool:
- Optimizes planned maintenance
- Models planned and random Outages
- Models full and partial Outages
- Models mean time to Fail and Repair
- Calculates availability parameters from historical data
- Provides visual display of outages across time
Monte Carlo Simulation
The chronological sequential type of the Monte Carlo simulation of iPool, unlike other type of Monte Carlo simulation, can capture the very important tail-end part of the price duration curve. The simulation can be either market bid-based or non-market cost-based and it can model various stochastic variables. iPool does the:
- Modeling of random generator unit full and partial outages
- Modeling of random weather for wind generation and demand
- Modeling of participant bid responses to changes in capacity and market conditions