Energy models represent a critical tool that can be used to examine future scenarios that consider resource availability and pricing, technology innovation, demand growth, and new energy and environmental policy. While such models cannot be used to predict the future, they can deliver insights that inform policy design. Despite their utility, energy models frequently suffer from one or more of the following limitations: closed model source code and data that prevents results verification, increasing model complexity over time, and insufficient quantification of future uncertainty. This project has focused on addressing these shortcomings by instituting a transparent process for model development and application; generating new insights into energy system development through the rigorous application of uncertainty analysis; and involving students and analysts in the process of model development and application.

To meet these objectives, this project has focused on the development and application of Tools for Energy Model Optimization and Analysis (Temoa), an open source framework for energy system modeling. Temoa is an instance of a particular type of energy model called an energy system optimization model (ESOM). ESOMs are widely used to model the system-wide impacts of energy development. ESOMs include detailed, bottom-up technology specifications and utilize linear programming techniques to minimize the system-wide present cost of energy supply by optimizing the installation of energy technology capacity and its utilization. The models are subject to a number of constraints that enforce system performance criteria as well as user-defined limits. Outputs include future estimates of technology capacity and utilization, marginal commodity prices, and emissions across the energy system.

Temoa has several characteristics that make it unique among ESOMs. First, it is open source, with input data and source code publicly available for download: https://github.com/TemoaProject/temoa. This allows interested parties to replicate our published model results. Second, Temoa was designed to operate in a high performance computing environment in order to enable rigorous uncertainty analysis. Given real world complexity, insights into energy system performance must account for future uncertainties related to both model input data and the simplified structure of the model itself. Third, a browser-based graphical user interface for Temoa has been developed. Interested modelers can test performance on http://temoacloud.com/input or download a free copy to run on their local machine. The user interface is intended to expand the base of model users and foster a more cohesive community of energy modelers around a set of open source tools.

Applications of Temoa have focused on different geographic areas. First, we constructed an open source US dataset at the national level. We are currently applying this national dataset to identify potential low carbon pathways that can be achieved without climate policy. Second, we have constructed an open source dataset for the State of North Carolina. Many states lack the in-house analytical capability to do energy systems modeling, and our efforts have helped inform ongoing discussions among stakeholders about the costs and benefits of a revised renewable portfolio standard. Third, in collaboration with World Bank colleagues, we have applied Temoa to examine electricity planning in fragile and conflict-prone countries. A case study of South Sudan has been developed and applied. Countries such as South Sudan need to develop planning strategies for power system development that explicitly consider the future risk of conflict and its effect on energy infrastructure. Temoa is well-suited to this analysis given its capability to perform uncertainty analysis. Such modeling efforts can yield useful planning insights for countries with little capacity to conduct their own modeling work.

In addition to the development and application of Temoa, this project has contributed to the development of standards within the international energy modeling community. Such efforts have included a call for greater transparency in energy modeling (DeCarolis et al., 2012) and an effort to establish best practice guidelines for the application of ESOMs (DeCarolis et al, 2017).

Dissemination of results has included peer-reviewed journal papers, conference papers and presentations, websites, and tutorials. This work has provided interdisciplinary research opportunities for several graduate and undergraduate research assistants, and has included collaboration with University College London. The topic of energy modeling has also been introduced to students at levels ranging from high school to graduate school. Future work to utilize and enhance the model are ongoing, and the expectation is that Temoa will be utilized for a long time to come.

Last Modified: 03/31/2017
Modified by: Joe Decarolis