Agent-Based Modeling of Solar Power Adoption by Los Angeles County Residents
Agent-based modeling -- which simulates a population of virtual people having unique attributes, preferences and decision-making rules -- can provide valuable assistance in the selection and design of the appropriate mix of policies for Los Angeles. To evaluate policies intended to drive adoption of residential photovoltaic systems in Los Angeles, this project develops an agent-based simulation model customized to the region. By running hundreds of simulations to understand the project aims to uncover likely patterns of technology adoption under various policies. The resulting model provides policymakers a rigorous means to evaluate likely scenarios flowing from alternative policy options. A variety of existing local, state and federal policies designed to drive adoption are already in place such as the federal Residential Renewable Energy Tax Credit, the California Solar Initiative rebates, state and federal low interest loans, and LADWP net metering. These policies operate by altering the economics of the adoption decision. Other potential policies such as streamlined permitting processes, demonstration sites, and outreach programs focus on reducing non-economic barriers to adoption. The existing mix of regulatory, economic, technological, geographic, and other aspects of adoption of residential PV systems is far too complex to be analyzed using traditional economic models.
School of Law
Anderson School of Management
Empirical Research Group
Civil and Environmental Engineering (University of Wisconsin, Madison)
Progress and Results
After conducting exploratory interviews with solar installers and completeling a literature review, the team now looks to move forward on a survery. The survey, used to collect the data needed for the agent-based modeling (ABM) has been developed and approved and will be launched to residents in Los Angeles County.
Once completed, the data will inform the ABM to model adoption of solar power by L.A. County homeowners. The teams aims to model individual homeowner behavior using utility models and accounting for social networks. Additionaly, they will run base-case scenarios for the present time period to ensure the model is working as intended. Then, a series of simulation experiments will compare the adoption trends under various policy alternatives to a baseline scenario
Other Information of Interest
Empirical Research Group
University of Wisconsin, Madison