Modeling intermittent renewable electricity technologies in general equilibrium models Karen Tapia-Ahumada a, ⁎, Claudia Octaviano b , Sebastian Rausch c,b , Ignacio Pérez-Arriaga d,a a MIT Energy Initiative, MA Institute of Technology, USA b Joint Program on the Science and Policy of Global Change, MA Institute of Technology, USA c Department of Management, Technology, and Economics, ETH Zurich, Center for Economic Research at ETH (CER-ETH), Switzerland d Institute for Research in Technology, Comillas Pontifical University, Spain abstract article info Article history: Accepted 7 August 2015 Available online 1 September 2015 Keywords: Renewable energy Electricity Intermittency General equilibrium Top-down modeling Bottom-up modeling Economy-wide top-down (TD) equilibrium models have traditionally proved to be valuable tools for assessing energy and climate policies. New modeling challenges brought about by intermittent renewable energy sources, however, require a careful review of existing tools. This paper presents an overview of TD modeling approaches for incorporating renewable energy and describes in detail one approach, using the MIT USREP model, to identify critical parameters and assumptions underlying the general equilibrium for- mulation. We then quantitatively assess its performance regarding the ability to correctly estimate the par- ticipation of intermittent renewables in the electricity sector as predicted by a bottom-up electricity sector model, which is designed to analyze the expansion and operation of a system with a large penetration of wind and which is integrated within an economy-wide general equilibrium framework. We find that a properly specified TD approach to modeling intermittent renewable energy is capable of roughly replicating the results from the benchmark model. We argue, however, that the general equilibrium approach is highly sensitive to key parameters which are a priori typically unknown or at least highly uncertain. Our analysis suggests that traditional TD simulation tools have to be enhanced to avoid potentially misrepresenting the implications of future climate policies where presumably renewable energy could participate at large scale. Detailed power system models that capture system reliability and adequacy constraints are needed to prop- erly assess the potential of renewable energy. © 2015 Elsevier B.V. All rights reserved. 1. Introduction Macro-economic “top-down” (TD) equilibrium models are widely used analytical tools to investigate the impacts of energy and climate policy in terms of technological pathways, environmental impacts (i.e., greenhouse gas emission reduction potentials) and their social costs and benefits. 1 While these models are used to derive policy recom- mendations, the “current generation” of TD approaches seems to lack the required detail and model features to adequately represent intermittent renewable energy sources. 2 Intermittent wind and solar energy resources require detailed temporal and spatial analyses, as well as, the study of operational implications such as the need for addi- tional reserve requirements, storage and transmission capacity. General Economic Modelling 51 (2015) 242–262 ⁎ Corresponding author at: MIT Energy Initiative, 77 Massachusetts Ave., Cambridge, MA 02139, USA. Tel.: +1 617 715 5367. E-mail addresses: katapia@mit.edu (K. Tapia-Ahumada), claus@mit.edu (C. Octaviano), srausch@ethz.ch (S. Rausch), Ignacio.Perez@iit.upcomillas.es (I. Pérez-Arriaga). 1 See, for example, the TD equilibrium models used in inter-model comparison activities such as the Stanford Energy Modeling Forum (e.g., Fawcett et al., 2014 and the work to ex- pand the GTAP dataset for energy and climate policy analysis Nijkamp et al., 2005). 2 Traditional modeling approaches, both in the domains of economy-wide TD equilibri- um as well as engineering-type “bottom-up” (BU) models, have proven to generate ade- quate and reliable model-based approximations of real-world energy (and electricity) production for systems characterized predominantly by fossil-based energy sources and technologies. TD models typically represent energy production technologies through highly aggregated (often smooth) production functions. While the strength of these models is to include energy supply and demand decisions within an internally consistent macro-economic framework, they typically lack the technological, spatial and temporal resolution. BU models, on the other hand, typically feature a highly resolved and technology-rich representation of energy (supply and demand) technologies but fail to in- clude interactions with the broader economic system due to their partial equilibrium na- ture. Importantly, BU models are hence not capable of incorporating macro-economic determinants of energy demand and supply and they cannot assess policies in terms of their social cost (e.g., GDP or consumption impacts). See, for example, Hourcade et al. (2006) for a more in-depth overview and discussion of both modeling paradigms. http://dx.doi.org/10.1016/j.econmod.2015.08.004 0264-9993/© 2015 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Economic Modelling journal homepage: www.elsevier.com/locate/ecmod