The Swiss Energyscope – ETH (SES-ETH) model is a linear optimization model of the energy system. It determines the investment and operation strategies that minimize the total annual cost, given the end-use energy demand; the efficiency and costs of the conversion technologies; and the availability and costs of the energy resources. SES-ETH was developed at ETH Zurich based on the original model by Stefano Moret from EPFL (Moret, 2017).
SES-ETH represents the main energy demands: electricity, heating and mobility. SES is a snapshot model, that is, it models the energy system in a target year and it does not make any statements on the trajectory to reach this future state. We model this target year with an hourly resolution that allows us to represent the intra-day variations of the energy demand and resource availability.
SES-ETH is a simple representation of the energy system, it largely neglects all aspects of spatial resolution, and it reduces the temporal resolution by choosing typical days and clustering hours within a day. These simplifications do not hinder the ability of the SES-ETH model to make inferences; on the contrary, by reducing the dimensionality, we are able to analyze large sets of scenarios considering uncertainty of modelling inputs. What we lose in granularity we gain in model tractability and the ability to identify technologies that are very likely part of the future mix and to derive policy recommendations for today.
Figure 1.1 illustrates the SES-ETH structure. Imported and domestic resources (Chapter 2) can be converted with energy conversion technologies (Chapter 4) to satisfy end-use demand in energy services: electricity, lowand high temperature heat (LTHand HTH), and mobility (passenger and freight) (Chapter 3). The model represents the energy conversion processes and determines the optimal technology mix for a certain emissions target by minimizing the total system costs. In the following sections we describe the different components of SES-ETH: First, the objective function; second, we describe how we model the conversion processes through balancing inputs and outputs in every layer (Section 1.2; third, we describe the representation of CO2 streams; fourth, we present the modelling of seasonal and intra-day variations through typical days and intra-day clusters (Section 1.4); and finally we show some additional constraints that we use in the model.