Applied Energy, Vol.210, 1037-1050, 2018
Modeling constraints to distributed generation solar photovoltaic capacity installation in the US Midwest
This paper presents a model for estimating the amount of distributed generation solar photovoltaics (DGPV) that can be accommodated by an electrical system, limited by the ability of existing generation infrastructure to change output. The model is applied to a region of seven states in the U.S. Midwest, showing the potential for temporary curtailment of installed DGPV to mitigate those constraints, and the associated reductions in coal- and natural gas-fired electricity generation. Scenarios considered are those under which 100%, 99.5%, and 95% of available solar can be utilized, and the point at which DGPV curtailment can no longer mitigate constraints because of the inability of conventional generation to keep pace as solar output declines in the late afternoon. The model is developed using historical hourly fuel use data for electricity generation, hourly solar capacity factor, and carbon emissions factors. The model includes a generalized linear model, constructed by regression analysis, used to allocate potential reductions in conventional generation in response to demand changes caused by DGPV. Limits on output and ramping of conventional generation are considered and initially mitigated, where possible, by changes in hydroelectric generation. Results indicate that a substantial amount of DGPV can be accommodated without any changes to current generation infrastructure, and that small amounts of curtailment enable greater DGPV capacity and generation. In a scenario without curtailment, DGPV could provide about 6% of total generation in the modeled region. With just 0.5% curtailment at DGPV installations, roughly 9% of generation could be provided by DGPV. Ultimately, up to 14% of generation could be achieved through DGPV installation before technological limitations would diminish the ability to meet all demand with further increased DGPV.
Keywords:Solar photovoltaic (PV);Distributed generation;Power system modeling;Renewable energy;Carbon emissions;Electricity