High resolution output power estimation of large-scale distributed PV systems

TLDR: In this article, the authors presented a simple, yet effective method to obtain fine-time resolution output data for a very large number of PV systems using only power measurements from a significantly smaller number of solar PV inverters, and without requiring any weather station information.

Abstract: With high penetration of residential, roof-top photovoltaic (PV) systems, it has become important for the electric utilities to conduct various impact analyses to ensure proper operation and control of power distribution systems. A key requirement for these studies is the data on PV power output at fine time resolution, often in the order of minutes, and for some transient studies, even in the order of seconds. This paper presents a simple, yet effective method to obtain fine time resolution output data for a very large number of PV systems using only power measurements from a significantly smaller number of PV inverters, and without requiring any weather station information. The method uses simplified models of power generation in solar PV systems. Physical information such as latitude/longitude, orientation and tilt of PV panels, electric properties such as different PV panel parameters and measurements from a small subset of PV systems are used to estimate fine time resolution data for all the PV systems in a distribution feeder. The details of the proposed method including the simplified model and algorithms to determine the most suitable PV measurement to use for estimating the output of a given PV system are described. Extensive validation of the proposed methods using measurements from an actual high penetration PV implementation in a distribution feeder in Arizona is presented. Even though the results shown in this study are at the resolution of minutes, the method can be readily extended to finer time resolutions with the data set from the small number of PV inverters used in the estimation updated at the finer time resolution.