Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Optimal design of industrial energy systems under future uncertainty with multi-stage stochastic programming
Due to climate change, industrial energy systems need to make investments. They face short-term and long-term uncertainties while making investments that can hardly be changed after realization. To consider those uncertainties, stochastic optimization is used to solve the multi-stage scenario tree.
Industrial energy systems are designed to cover the industries’ supply for electricity, heating, cooling, steam, etc. However, to mitigate climate change, industrial energy systems need to satisfy CO2 targets. To comply with CO2 targets, industrial energy systems need to invest into their system for the future requirements. Based on a design optimization via mathematical optimization, an optimal system design can be found subject to constraints.
Meanwhile, industry faces massive short-term (e.g. price fluctuations) and long-term uncertainties (e.g. changing regulatory conditions, long-term price uncertainties) while making investment decisions that can hardly be changed after realization. To consider both short-term and long-term uncertainties, methods from stochastic opti-mization need to be utilized to solve the multi-stage scenario tree.
Industrial energy systems are designed to cover the industries’ supply for electricity, heating, cooling, steam, etc. However, to mitigate climate change, industrial energy systems need to satisfy CO2 targets. To comply with CO2 targets, industrial energy systems need to invest into their system for the future requirements. Based on a design optimization via mathematical optimization, an optimal system design can be found subject to constraints. Meanwhile, industry faces massive short-term (e.g. price fluctuations) and long-term uncertainties (e.g. changing regulatory conditions, long-term price uncertainties) while making investment decisions that can hardly be changed after realization. To consider both short-term and long-term uncertainties, methods from stochastic opti-mization need to be utilized to solve the multi-stage scenario tree.
Within the multi-stage stochastic programming model, scenarios for uncertain parameters can be included. Due to computational intractability, the scenario tree needs to be designed carefully by using sampling and reduction techniques. Finally, the optimization framework is used to answer the following questions amongst other things:
- How can an optimal transition pathway of an industrial energy system be identified considering future uncertainty?
- What actions need to be done now despite future uncertainty?
- Which future long-term and short-term uncertainty need to be included into the model?
- What level of detail needs to have the scenario tree?
- Which future technologies need to be considered?
- What is the trade-off between the consideration of many short-term scenarios vs. many long-term scenarios?
Within the multi-stage stochastic programming model, scenarios for uncertain parameters can be included. Due to computational intractability, the scenario tree needs to be designed carefully by using sampling and reduction techniques. Finally, the optimization framework is used to answer the following questions amongst other things: - How can an optimal transition pathway of an industrial energy system be identified considering future uncertainty? - What actions need to be done now despite future uncertainty? - Which future long-term and short-term uncertainty need to be included into the model? - What level of detail needs to have the scenario tree? - Which future technologies need to be considered? - What is the trade-off between the consideration of many short-term scenarios vs. many long-term scenarios?
ETH Zurich
Niklas Nolzen
Doctoral student
Energy & Process Systems Engineering
CLA F 19.3
Tannenstrasse 3
8092 Zurich, Switzerland
nnolzen@ethz.ch
ETH Zurich Niklas Nolzen Doctoral student Energy & Process Systems Engineering CLA F 19.3 Tannenstrasse 3 8092 Zurich, Switzerland