Carpe diem: A novel approach to select representative days for long-term power system modeling. (1st October 2016)

• Record Type: Journal Article
• Authors: Nahmmacher, Paul; Schmid, Eva; Hirth, Lion; Knopf, Brigitte
• Is Part Of: Energy. Volume 112(2016)
• Page Start: 430
• Page End: 442
• Publication Date: 2016-10-01
• Subjects: Power system modeling -- Variability -- Renewable energy sources -- Time slices; CC combined cycle -- CCS carbon capture and storage -- CSP concentrated solar power -- GT gas turbine -- LDC load duration curve -- PV photovoltaic -- RMSE root mean square error -- SSE sum of squared errors -- VRE variable renewable energy; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2016.06.081
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: With an increasing share of wind and solar energy in power generation, properly accounting for their temporal and spatial variability becomes ever more important in power system modeling. To this end, a high temporal resolution is desirable but due to computational restrictions rarely feasible in long-term models that span several decades. Therefore many of these models only include a small number of representative 'time slices' that aggregate periods with similar load and renewable electricity generation levels. The deliberate selection of the time slices to consider in a model is vital, as an inadequate choice may significantly distort the model outcome. However, established selection methods are only based on demand variations and are not applicable to input data with a large number of fluctuating time series, which is a drawback for models with high shares of renewable energy. In this paper, we present and validate a novel and computational efficient time slice approach that is readily applicable to input data for all kinds of power system models. We illustratively determine representative days for the long-term model LIMES-EU and show that a small number of model days developed in this way is sufficient to reflect the characteristic fluctuations of the input data. Highlights: We present a novel approach to efficiently cover wind & solar variability in models. It allows simultaneously accounting for multiple variable energy sources & regions. We validate our approach and apply it to the long-term power system model LIMES-EU. The developed method is readily applicable to other power system models.