Modellering av marked for elsertifikater i Samkjøringsmodellen

Modelling of el-certificates in the EMPS model In 2012 a common market for el-certifikates was introduced in Norway and Sweden. This is a market incentive with the goal of promoting the development of 26.4 TWh new renewable energy in the two countries. For both countries this is part of the requirements to fulfill their commitments according to the renewable directive. Investors wanting to participate in the development of these renewable projects it is important to have a reliable prediction of the price development of the el-certificates. This is also crucial information for other participants in this market. For the authorities it is important to be able to evaluate if the market is working as intended. As a consequence it is important to have a tool that can predict how the market will behave depending on different scenarios and conditions. The certificate market is tightly coupled to the electricity market The EMPS (no. samkjøringsmodellen) is a well know, fundamental model for the Nordic power market used to determine future power prices, water values and investment analysis. The model contains a detailed empirical representation of the variance in key climatic parameters (run-off, temperature and wind), and computes and optimal management of water. Since the model is already in use among several of the actors in the market it is quite natural to base a model of the el-certifikate market on the same code. The certificate model is implemented as a virtual reservoir representing a storage of certficates. Power production from power plants entitled to el-certificates will fill up the reservoir. The quota curved defined in the bilateral agreement together with the actual power consumption determines the number of certificates that will be cancelled each spring. This represents the draw-down of the certificate reservoir. Important benefits of the model are: ? It is built on a modelling framework that the actors are familiar with * There is a tight coupling between the power system, the power market and the certificated market and this is well represented in the modell * All the main data for the Nordic power market is already implemented in the model * The model contains a well proven method of handling stochastic events Climatic conditions will impact both the number of certificates that will be issued each year as well as the amount of certificates that the consumers must buy. P recipitation and wind speeds will significantly impact the power production while the temperature is has a large influence on the power consumption. During the year the both the power price and certificate price will develop depending on the balance between the power consumption and production of certificates. Utilities that does not possess sufficient amount of certificates at a specific date will be penalized with a charge that is 150% of the average price the last 12 months. A correct implementation of this scheme has been crucial for the correct behavior of the model. This is particularly important when simulating scenarios with shortage of certificates in the market. The project had duration of 2 years and was started in spring 2013. The conceptual design was finished in 2013 and a first prototype was made available to the user group in February 2014. Testing revealed several issues when moving from SINTEF Energys simplified test data set to the operative data set of the users. This has been handled and corrected subsequently in the project. Towards the end a larger scenario study was undertaken. This has shown that under normal circumstances the discounted certificate price is fairly stable. This is consistent with theoretical considerations for a well formed market where future variations are utilized in arbitrage. Considering that this is an output from the model, not a condition, this gives a good indication that the model gives a consistent description of liquid market. Towards the end of the project a workshop was organized and participants from market analysts, other research institutes and trading companies where invited to present and discuss their models. This was a fruitful opportunity to present the project to the market in addition to scientific publication.

Det norsk-svenske markedet for elsertifikater startet i 2012 og vil sørge for at det kommer inn 26,4 TWh ny fornybar kraftproduksjon fordelt mellom Norge og Sverige innen 2020. Pga. markedets størrelse er det viktig at markedet fungerer effektivt. En tren ger derfor verktøy både til prognosering og til overordnet analyse av systemet. Sertifikatmarkedet er tett integrert med kraftmarkedet. Sertifikatene kjøpes av leverandører, tildeles produsenter og betales til slutt på strømregningen. I prosjektet har vi tenkt å inkludere sertifikatmarkedet i Samkjøringsmodellen, som er en modell for kraftmarkedet. Denne modellen har også en detaljert empirisk representasjon av usikkerhet i klimavariable, og en metode for optimal håndtering av et lager. Dette er viktige faktorer for kortidsprognosering av sertifikatmarkedet. Viktige forskningsmessige utfordringer er anvendelse av vannverdiberegning på et sertifikatlager, dynamisk fastsettelse av straffeavgiften for manglende sertifikater og fastsetting av en sluttverdi f or lagrede sertifikater dersom planleggingshorisonten er kortere enn 2035. Prosjektet vil starte med en designfase, og så implementeres et marked for elsertifikater som i størst mulig grad samsvarer med det gjeldende systemet. Deretter testes funksjonal iteten av brukerne i prosjektet. Det lages også egne case-studier både for prognosering og for overordnet analyse av systemet. Modellen justeres deretter basert på brukernes egen testing og vurdering. Endelig programvare lisensieres slik at andre også kan få tilgang. Metode og eksempelstudier publiseres i egnet tidskrift / internasjonale konferanser. I prosjektet deltar både systemansvarlig (NVE) og registeransvarlig (Statnett) myndighet for sertifikatsystemet i Norge, i tillegg til produsentinteresser (S tatkraft og Energi Norge) og en analysebedrift som lager prognoser for kraft- og sertifikatmarkedet (SKM Market Predictor).