A production simulation tool for systems with integrated photovoltaic energy sources.

Power system planning;probabilistic planning;renewable energy;solar photovoltaic modeling.

Bhana, Rajesh ; Gross ; George

关键词: Power system planning;    probabilistic planning;    renewable energy;    solar photovoltaic modeling.;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/29616/Bhana_Rajesh.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Climate change awareness, the drive to sustainability and the push for energy independence have resulted in the wider utilization of renewable energy sources. Photovoltaic (PV) power is a renewable solar energy source that is increasing in use as its costs decrease. We have developed an assessment and planning tool to quantify the longer-term variable effects of large-scale PV energy production on power systems.The tool consists of a probabilistic model for the representation of the power output of the PV resources, at single or multiple sites on the network, and the incorporation of the model into an extended probabilistic simulation framework for the evaluation of the longer-term variable effects of the resources. We develop the probabilistic model in a manner that captures the time-dependent, variable and intermittent nature of the PV resources and incorporate the model into the extended framework in a way that captures the correlation between the chronological load and the uncontrollable PV output. Because typical daily PV output patterns vary markedly over a year, we construct the probabilistic model on a seasonal basis, and because the duration and magnitude of PV output changes from day to day, we scale the daily PV output patterns, both in time and magnitude, into scaled-output characterizations that allow the meaningful comparison of different days of the season. We then classify the seasonal set of daily PV scaled-output characterizations into pattern cluster sets of days with similar daily output patterns before re-scaling the pattern clusters into class sets of daily PV output representations. From these class sets, we approximate the conditional probability distributions of the PV output random variables conditioned on each pattern class. We extend the conventional probabilistic simulation framework to incorporate the PV resources by using these approximations of the conditional probability distributions.By incorporating the probabilistic PV output model into the extended framework, we are able to quantify the longer-term variable effects of a high penetration of large-scale PV resources on a system in terms of reliability, economic-cost and environmental-impact metrics. To test the capability of the tool, we have applied the methodology on a variety of test system cases covering a wide span of load, system and resource characteristics. The application of the tool is useful in many areas of power system planning, including investment decision-making and policy formulation.