|dc.description.abstract||Together with the other Nordic transmission system operators Statnett is currently investi- gating the possibility of implementing the so called Flow Based Market Clearing approach. Flow Based Market Clearing is today the preferred market clearing method by the European Commission, and is already introduced in the Central Western European countries. The ap- proach maximizes social surplus while introducing a simplified grid model as a part of the market clearing algorithm. This is the main difference compared to today s Net Transfer Capacity allocation method, where the physical network flows are not implemented in the optimization problem. The new market clearing approach gives a more efficient system, increases the network flow, and keeps the system security at a satisfactory level.
As a foundation for the Flow Based Market Clearing model the transmission system oper- ator has to make estimations of future production and consumption in each bidding area. These estimations form the areas net position, i.e. the areas net export. This projects main objective is to develop a methodology for a new method for estimation of net positions.
The developed estimation approach assumes that the Day-ahead market curves mainly con- sists of two parts; One part represented by price dependent bids, and one part represented by non-price dependent bids. It is further assumed that the flexible part of the market curves is roughly the same from one day to the next, and the changes that do occur are related to marked bids that are not effected by price changes, e.g wind and solar power production. The change in these bids will therefore cause parallel shifts to the market curves. The devel- oped method aims to identify these price in-depended bids and uses them to make forecasts of future net positions. The forecasted net positions are found with the market clearing al- gorithm Euphemia, which is offered through the simulation tool "Simulation Facility".
Two initial test were carried out to test the new estimation approach. One with up-to-date data, and one with a data foundation that is two weeks old.
The results from the initial tests indicate that the new estimation method can give positive results compared to other possible methods. In order to keep the system security at a satis- factory level, it is important to reduce large errors. The simulation results indicate that large error are reduced, however a few very large error do occur with the new method for the tested case.||