Sammanfattning
Electricity prices are known to spike during peak hours, only to revert to normal levels during off-peak hours. We introduce a generalization of the time varying independent spike model commonly used to model the electricity spot price from daily data to hourly data to cope with this feature.
We let the probability of extreme prices depend on several variables, such as consumption, reserve margin or wind power. The model can then be used to forecast the risk of extreme prices.
More factors become relevant for predicting extreme events when moving to hourly data, but consumption is still the most important factor. The methodology is showcased by illustrating how extreme prices can be forecasted by predicting the consumption.
We let the probability of extreme prices depend on several variables, such as consumption, reserve margin or wind power. The model can then be used to forecast the risk of extreme prices.
More factors become relevant for predicting extreme events when moving to hourly data, but consumption is still the most important factor. The methodology is showcased by illustrating how extreme prices can be forecasted by predicting the consumption.
| Originalspråk | engelska |
|---|---|
| Sidor (från-till) | 85-90 |
| Tidskrift | Journal of Energy Challenges and Mechanics |
| Volym | 2 |
| Utgåva | 3 |
| Status | Published - 2015 |
Ämnesklassifikation (UKÄ)
- Sannolikhetsteori och statistik