Transition to a low-carbon electricity system — investment decisions under heterogeneity, uncertainty and financial feedback
Licentiate thesis, 2021

A transition to a low-carbon electricity system will require a substantial increase in the investment rate in low-carbon technologies. This calls for a better understanding of investment decisions and their impact on the pace of the transition. We explore the transition to a low-carbon electricity system by developing an agent-based model, focussing on agents’ decisions when investing in new power plants. This work addresses three characteristics associated with investment decisions—heterogeneity, uncertainty, and financial feedbacks—which many energy system models do not take into account.

In this study, heterogeneity is represented by the agents’ different levels of risk aversion (represented by the hurdle rate employed by the agent) and their different expectations for the future carbon price. Uncertainty is reflected by the agents’ imperfect foresight. They have limited information on future electricity prices, carbon prices, electricity demand, etc. Also, there is stochasticity in fuel prices and electricity demand.Financial feedback in the model is designed so that an agent’s previous investment decisions will impact its future ability to invest. An agent’s new investment impacts the capacity mix and electricity price, which, in turn, impact the revenue the agent receives. If the investment turns out to be profitable, then the agent has more capital to make further investments; if unprofitable, investing may be hampered in the future.

This study has analysed the transition on two levels—the system level and the level of the individual agents. On the system level, we explored system dynamics such as the development of the capacity mix, electricity price, and emission trajectories over time. We have also analysed the competition among low-carbon technologies, the value of wind, and returns on investments for different technologies. On the level of the individual agents, we investigate different investment criteria and outcomes.

Under a scenario with an increasing carbon tax, our model exhibits a transition to a low-carbon electricity system. Wind, together with gas-fired power plants, competes with nuclear power in the capacity expansion. The value of wind will drop relatively, but not absolutely. On the agent level, results show that agents who use lower hurdle rates are more willing to make investments and therefore accelerate the transition. But as every investment is associated with risk, these agents also face a greater risk of going bankrupt, especially when they are less financially constrained. Agents who expect carbon prices that are too low or too high compared to the actual development also have a higher tendency to go bankrupt, and the return on equity for these agents is generally lower than for agents with a more accurate carbon price expectation.

This study illustrates the importance of including heterogeneity, uncertainty, and financial feedback in models of the energy transition. When the model is run with homogeneous agents or has different degrees of uncertainties, or no financial feedback mechanisms, results differ on both the system level and the level of the individual agents.

agent-based modelling

investment decisions

electricity system transition

investment financing

Online(Zoom), password: 123
Opponent: Assoc. Prof. Emile Chappin, Department of Technology Policy and Management, Delft University of Technology


Jinxi Yang

Chalmers, Space, Earth and Environment, Physical Resource Theory, Physical Resource Theory 2

Yang, J., Azar, C., Lindgren, K., 2021. Modelling the transition towards a carbon-neutral electricity system - investment decisions and uncertainty.

Yang, J., Azar, C., Lindgren, K., 2021.Financing the transition towards carbon neutrality – an agent-based approach to modelling investment decisions in the electricity system


European Commission (EC) (765515), 2017-10-01 -- 2021-09-30.

Subject Categories

Energy Systems


Chalmers University of Technology

Online(Zoom), password: 123


Opponent: Assoc. Prof. Emile Chappin, Department of Technology Policy and Management, Delft University of Technology

More information

Latest update