(690h) Open-Source Capacity Expansion Model with a Focus on Accessibility, Usability, and Complete Life-Cycle Assessment

As the global energy economy electrifies, the use and pervasion of the power sector increases. Correspondingly, the importance of understanding its capabilities, limitations and resiliency are invaluable as decarbonization hinges on these features. Capacity expansion models (CEMs) are used to simulate and stress-test tailored electricity sector transitions based on user-input information (electricity demand, renewable resources, regional targets, and technology properties).

There are many CEMs available with a wide range of features and focuses. This tool, SESAME – Power Greenfield, is the first to have extensive pre-processed data available, comprehensive LCA, and an intuitive user interface (UI) making it accessible to a wide range of audiences. More specifically, firstly, Power Greenfield leverages its’ connections within MIT’s Sustainable Energy System Analysis Modelling Environment (SESAME) to source technoeconomic analysis (TEA) and life-cycle assessment (LCA) values from within the SESAME pathway analysis tool. Also, load and variable renewable energy (VRE) profiles have been sourced for NERC regions within the contiguous USA meaning that the user can run a variety of case studies without the entry-barrier of extensive data collection and pre-processing. Secondly, emissions equivalencies are tracked from cradle-to-grave, including extraction of raw materials, manufacturing, product use, and retirement of resources. This is in contrast to most models which only consider emissions released from fuel burned. While this simplification is an accurate assumption for thermal generators, emissions from other life-stages become nonnegligible in highly renewable systems. Lastly, Power Greenfield’s comprehensible UI allows the user to run a variety of scenarios including, but not limited to fine-tuning carbon taxes, imposing carbon intensity limits, adjusting generator contributions, and altering renewable, storage or fuel costs.

Methods:

An overall diagram of model structure and information flow is shown in the attached figure. As indicated in the attached figure, this linear model is formulated in Python with Pyomo and solved with Gurobi. With these parameters, the model completes optimizations and displays results in under 30 seconds. Power Greenfield operates under a series of simplistic assumptions. Each NERC region is analyzed as a 1-nodal system with an assumed transmission and distribution (TD) efficiency loss, and TD cost consistent across all analysis regions. As the name suggests, this is a greenfield model, so does not consider current generator or transmission resources. Lastly, this model is deterministic, meaning that the model operates under an assumption of perfect foresight in demand and VRE profile predictions.

Results and Implications:

This model minimizes overall cost based on user inputs and outputs generator, transmission, and energy storage installations required, as well as tracks energy flow within the system. The results of the system can be broadly categorized into three categories: scalars, vectors, and breakdowns. Scalar results list equipment sizing parameters which can be used to orient the user to the general magnitude of the system elements. Vector results show electricity flows over a user-defined time-period, so the user can understand system operations. The time period can be manipulated in length and season to explore temporal variation in trends. Lastly, breakdowns show how different sources contribute to important LCA and TEA values. TEA is broken down by equipment technologies and by cost categories (fixed, operational, fuels, etc.), and LCA is broken down by technologies. It is important to note the value of these all these outputs. The user can obtain and explore insightful results all within the app interface, and without any data post-processing.

Future studies are planned to explore the value of this tool even further, including, but not limited to the following studies. Firstly, highlighting the contrast between the decarbonization plans in the different regions due to variability in load and renewable resources. Second, exploring the effectiveness of a carbon tax vs. imposed carbon intensity limits. Lastly, exploring weaknesses in system resiliency from highly-renewable systems and quantifying necessary dispatchable resources to ensure grid security.