Optimizing Geographic Allotment of Photovoltaic Capacity in a Distributed Generation Setting

A multi-objective optimization was performed to allocate 2 MW of photovoltaic (PV) among four candidate sites on the island of Lanai, Hawaii, such that energy was maximized and variability in the form of ramp rates was minimized. This resulted in the Pareto-optimal set, an optimal solution set that provides a range of geographic allotment alternatives for fixed PV capacity. Within the Pareto-optimal set, a trade-off was found between energy produced and variability experienced, whereby a decrease in variability always necessitates a simultaneous decrease in energy. With this development, system designers have a method to select the preferred combination of energy generation and variability within the set of optimal alternatives to meet their needs. A design point within the optimal set was selected for study that decreased extreme ramp rates by more than 50% while decreasing annual energy generation by only 3% above the maximum generation allocation. To quantify the allotment mix selected, a new metric called the "ramp ratio" was developed. It compares ramping magnitude when all capacity is allotted to a single location to the aggregate ramping magnitude in a distributed scenario. The ramp ratio quantifies simultaneously how much more smoothing a distributed scenario would experience than single-site allotment and how much a single site is being underutilized for its ability to reduce aggregate variability. This paper creates a framework for use by cities and municipal utilities to reduce variability impacts while planning for high penetration of PV on the distribution grid, thereby maximizing the value of investments.