Assessing the impact of power plant location decisions on the long term development of the electricity sector

Assessing the impact of power plant location decisions on the long term development of the electricity sector

Paling, J.

Herder, P.M. (mentor)
Chappin, E.J.L. (mentor)
Cunningham, S.W. (mentor)
Bollinger, L.A. (mentor)

Technology, Policy and Management

Energy & Industry

System Engineering, Policy Analysis & Management

2013-10-25

This research presents a new way to assess the impact of location decisions for new power plants by power producers. It builds on an existing agent-based model that is able to simulate market and investment behavior of power produc- ers in a north western European electricity market (EMlab-generation). This study adds a module that is based on game theory concepts to simulate both the location selection by power producers and the required spatial permit procedure. In the end this model could help to give insights in possible future developments of power plant siting in a specific country (Netherlands in this case). This could help TSOs and policy makers assess their siting regimes and anticipate for po- tential future new generation capacity. The main research question is formulated as: How does the set of factors considered in location decisions for new power plants affect the future development of the technological and spatial distribution of power generation in the Netherlands?. This research starts to investigate what aspects play a role in the location decision for new power plants. From literature and empirical sources it is found that besides the technical requirement of power generating technologies (Cooling water, feedstock connection etc. ), the permit procedure plays a major role in the decision. Permit risk as power producer call it could significantly decrease the profitability of a project by delays and is thus a very important factor. Permit risk is connected with the potential for local activism against the siting of the new power plant. Activism is simplified and conceptualized to three main factors: Population density, wealth and attitude towards a specific power generation technology. The model addition contains two main parts. First the power producer se- lects a technology and finds the best locations according to its own selection method. This could be conceptualized and modelled easily with utility func- tions. The second part of the permit procedure is more difficult as it involves negotiations with the local government and potentially several local parties (ac- tivists). The game theory concept of the nucleolus is used to conceptualize the permit procedure. First the government is payed a monetary environmental compensation fee (based on legal obligation), afterwards the local parties are payed a compensation fee, the unhappiest party first. The compensation is payed to ease the negative attitudes towards the new power plant and to lower the chance parties go to court and potentially delay the project, threatening profitability. This in essence is the model and the data input is based on data of potential power plants sites in the Netherlands (SEVIII) and regional statistics by the bureau of the statistics. The simulation results show that the model is able to produce significantly different results for different location selection methods by power producers, both in geographical distribution of power plants as the technology mix. Addi- tionally the permit procedure and the amount of compensation required by the local parties also significantly influences the spatial and technological distribu- tion of power plants. Finally other location related factors such as the potential for CCS at a location were also incorporated and showed significantly different results. The results are subject to some limitations, these mainly have to do with the choices for simplifications and the data used as model inputs. This is something to always consider when reviewing the results of this study and when reviewing agent-based modelling studies in general. The analysis of the results contain several different types of contributions: • The current EMlab-generation model is extended and is now able to in- corporate location decision and the permit procedure. Additionally the research presents a lot of valuable location data of power plants. The model shows similarities with the described behavior in the literature and empirical data. • This study contains methodological contributions in the area of agent- based modelling and game theory. The concept of the Nucleolus has not been used in an agent-based environment and have not been used to sim- ulate a specific kind of permit procedure. Under specific conditions and assumptions we showed that it is possible to use the nucleolus and this enables new applications in many different research fields that consider social negotiations. • A start has been made to show the potential of the model for real world applications. Scenarios of Tennet for the generation capacity in 2030 have been compared with our model results, showing differences and similari- ties. The amount of data and detail of this research’s model could help improve these scenarios and help Tennet anticipate for future generation capacity developments. Additionally some observations with regards to policies are given and the possibilities of this model to policy analysis is discussed.

Agent-based modelling
Electricity market simulation
Game Theory
Location Aspects
Permit procedure

http://resolver.tudelft.nl/uuid:68bbbdff-27ec-47be-8eff-9cd685aa5d83

Student theses

master thesis

(c) 2013 Paling, J.