Conférence sur les marchés énergétiques

À propos de
l'événement

Nous vous invitons à assister à 2 conférences données par les chercheurs norvégiens Stein-Erik Fleten et Sjur Westgaard de la Norwegian University of Science and Technology. Les conférences seront présentées en anglais et porteront sur les sujets suivants :

Conférence 1 : Real-option implied maintenance and switching costs for peaking power plants

Conférencier: Stein-Erik Fleten (travail conjoint avec des coauteurs)
Norwegian University of Science and Technology, NO–7491, Trondheim, Norway

We use nonparametric structural estimation to estimate the costs associated with shutting down, starting up, and abandoning peaking power plants, specifically simple cycle combustion turbines. This entails setting up a real options model of switching behavior of power plant operators. Estimates of switching costs are surprisingly difficult to obtain in practice. Our case study is made possible by the availability of detailed power plant data from the United States. Each year the owners of existing power plants must file Form 860 with the Energy Information Administration (EIA). From these data it is possible to determine whether an existing plant was shut down, started up, or abandoned. Our sample includes 8189 plant-year observations from the period 2001–2009. These data are augmented with time series of electricity prices and fuel prices, available from electricity market operators and the EIA.  Understanding shutdown, startup, and abandonment decisions is important for designing efficient mechanisms in electricity capacity markets. In an effort to provide incentive for firms to build and maintain sufficient peaking capacity, Independent System Operators in the United States recently have introduced capacity markets such as the Reliability Pricing Model (RPM) in the PJM system. Capacity markets provide revenue to plants for maintaining availability and therefore help to ensure system reliability. Participants in RPM bid an Avoidable Cost Rate (ACR). Avoidable costs are the incremental costs of being a capacity resource, i.e., the costs which could be avoided if a particular plant were shut down for a year. Owners of power plants may either develop estimates of these costs for each individual plant or use default rates provided by the market. From our switching and maintenance costs we estimate ACRs. Our estimates of ACRs are less than the default values used in PJM’s capacity market, implying that consumers may be paying too much for system reliability.

Conférence 2 : Modelling price distributions in the Californian electricity market using quantile regression with fundamentals

Conférencier: Sjur Westgaard, (travail conjoint avec co-auteurs)
Norwegian University of Science and Technology, NO–7491, Trondheim, Norway

In  this  paper, we ask how distributions of  electricity spot prices depend on fundamental factors such as fuel costs and renewable infeed. Taking the California day-ahead wholesale electricity market as our object of study, we find that natural gas prices, greenhouse gas allowance prices, load forecast, and forecast of solar and wind generation, as well as lagged prices and price volatility affect the price dynamics. We use a quantile regression model and investigate the zones SP15 for each trading period using data from 8 January 2013 to 24 September 2016. Natural gas prices have a positive effect over all quantiles and trading periods. Greenhouse gas prices has very little effect on the price formation. Load forecast has a positive effect, generally increasing with quantiles. Solar and wind production forecasts have a negative effect where the effect increase with higher quantiles. Lagged prices have a positive effect. The effect of volatility is positive for high quantiles and negative for low quantiles. We show how the results of the quantile regression models can be incorporated in scenario analysis and stress testing. By changing one or more variables from a base scenario, we can investigate its effect on Value at Risk for a given price area and trading hour. We claim that this will be a very helpful risk management tool for participants being long or short the electricity market.