Forecasting day-ahead electricity prices: Are next-day weather forecasts good day-ahead price indicators in Nord Pool Spot market?
Vänttinen, Ville (2017-01-25)
Forecasting day-ahead electricity prices: Are next-day weather forecasts good day-ahead price indicators in Nord Pool Spot market?
Vänttinen, Ville
(25.01.2017)
Tätä artikkelia/julkaisua ei ole tallennettu UTUPubiin. Julkaisun tiedoissa voi kuitenkin olla linkki toisaalle tallennettuun artikkeliin / julkaisuun.
Turun yliopisto. Turun kauppakorkeakoulu
Kuvaus
siirretty Doriasta
Tiivistelmä
Electricity markets have several unique features. Unpredictable and irregular demand and supply shocks have stronger effect in case of power markets, compared with other energy commodity markets. Electricity is an energy commodity with few substitutes and its demand is quite inelastic for price changes. Nearly all electricity needs to be consumed straight away after generation as electricity is a commodity with limited storability. This means that power markets supply and demand need to be balanced in real-time. Electricity’s end-user demand has various different demand profiles, and they all show strong variability by temporal effects, and so influence market price formation.
In short term, price forecasting is an important tool for power market participants as it helps in electricity delivery contract planning in modern real-time power market. In longer term, the profitability of power industry related investments, and viability of whole industry’s value chain, depends highly on current and future electricity prices.
The purpose of this thesis is to evaluate and estimate the proportional contribution that weather forecasts and other temporal power price features add for electricity price forecasting. The key electricity price features are identified and analyzed, and the main issues in electricity price forecasting are underlined. Particularly the role of weather forecasts as future power price determinant is inspected. Thesis’s input data includes weather forecasts and day-ahead daily spot prices of three Elspot bidding areas. AR-models are fitted with OLS method, and model residuals are used as indicators of model suitability. Overall results of this thesis show pretty clearly that electricity price prediction can be significantly improved when stochastic and deterministic price features and weather impacts are considered in price forecasting model.
In short term, price forecasting is an important tool for power market participants as it helps in electricity delivery contract planning in modern real-time power market. In longer term, the profitability of power industry related investments, and viability of whole industry’s value chain, depends highly on current and future electricity prices.
The purpose of this thesis is to evaluate and estimate the proportional contribution that weather forecasts and other temporal power price features add for electricity price forecasting. The key electricity price features are identified and analyzed, and the main issues in electricity price forecasting are underlined. Particularly the role of weather forecasts as future power price determinant is inspected. Thesis’s input data includes weather forecasts and day-ahead daily spot prices of three Elspot bidding areas. AR-models are fitted with OLS method, and model residuals are used as indicators of model suitability. Overall results of this thesis show pretty clearly that electricity price prediction can be significantly improved when stochastic and deterministic price features and weather impacts are considered in price forecasting model.