Predicting the Manual Regulating Market Direction in DK1 to Allow a BRP to profit from the Balancing Market
Authors
Spencer, Benjamin William Vennegaard ; Frandsen, Christian
Term
4. Term
Publication year
2014
Submitted on
2014-06-04
Pages
112
Abstract
This thesis examines whether a Balance Responsible Party (BRP) can leverage the Danish balancing market in the DK1 price area by predicting the direction of the manual regulating market. In Denmark, imbalances are settled hourly at a price set by the manual regulating market and only known after the operating hour, which typically leads BRPs to minimize exposure via the intraday market (Elbas). The thesis challenges the assumption that imbalances are usually settled unfavorably by developing a statistical model that uses regression analysis on inputs such as wind and consumption forecasts, actual production/consumption, temporal patterns, and the previous hour’s regulation. The model predicts the regulating direction up to two hours ahead with a 65% hit rate. Using these predictions, a speculative bidding strategy for consumption units is proposed that deliberately creates imbalances opposite to the system’s imbalance when regulation is needed, aiming to secure a favorable one-price settlement. In best-case calculations, the strategy yields approximately 30% higher profit compared with common approaches, and an initial assessment indicates it may reduce the need for manual regulation and support Energinet.dk’s objective of security of supply at the lowest socio-economic cost.
Denne afhandling undersøger, om en Balanceansvarlig (BRP) kan udnytte balancemarkedet i det vestdanske prisområde DK1 ved at forudsige retningen på det manuelle reguleringsmarked. I det danske elsystem afregnes ubalancer time for time til en pris, der fastsættes af det manuelle reguleringsmarked og først kendes efter driftstimen, hvilket normalt får BRP’er til at minimere ubalancer via intradagmarkedet (Elbas). Afhandlingen udfordrer antagelsen om, at ubalancer oftest afregnes til en ugunstig pris, ved at udvikle en statistisk model baseret på regressionsanalyse af blandt andet vind- og forbrugsprognoser, faktiske produktion/forbrug, tidsmønstre og forrige times regulering. Modellen kan forudsige reguleringsretningen to timer frem med en træfsikkerhed på 65%. På baggrund af modeludfaldet formuleres en spekulativ budstrategi for forbrugsenheder, der bevidst skaber ubalancer i modsat retning af systemets ubalance, når der er reguleringsbehov, for at opnå en gunstig afregningspris under én-prismodellen. I bedste fald viser beregningerne, at strategien kan øge profitten med cirka 30% i forhold til gængse strategier, og en indledende systemvurdering peger på, at adfærden kan reducere behovet for manuel reguleringskraft og dermed understøtte Energinet.dk’s mål om forsyningssikkerhed til lavest mulig samfundsøkonomisk omkostning.
[This apstract has been generated with the help of AI directly from the project full text]