Session: 02-02: Advances in Green Energy Modeling and Innovative Technologies

Paper Number: 130561

130561 - Wind and Solar Renewable Energy System Estimation With Batteries Using the Monte Carlo Sampling Approach 

Abstract: 

The factors involved in the declining costs of electricity produced from renewable energy are government regulations, policy obligations, and incentives and the goal to reduce the carbon footprint in the energy sector. Over the past decade, the concept of Distributed energy systems and decentralized energy generation has been growing steadily to make the electricity flow bi-directional and secure energy availability.  Therefore, owning renewables can be considered an investment in the energy industry, but determining the proper sizes for large-scale renewable energy can be challenging due to their inherent intermittency and potential for year-to-year variations in the face of climate change. These concerns can be mitigated by installing energy storage systems (ESSs) of a suitable size, but care should be taken to avoid unnecessary costs. Therefore, methods are needed to optimize renewable plus storage system sizes for various applications to help ensure cost-effectiveness.  Wind and solar resources depend on atmospheric dynamics such as wind speed and solar irradiance that are highly time and seasonal-dependent, so a robust installation optimization method to account for these effects is presented here.

To make the optimization program estimations reliable the uncertainties associated with wind speed and solar irradiance must be analyzed, studied, and integrated into the algorithm.  Here we present how uncertainty can be quantified year-to-year and the Monte Carlo method used to quantify the effect of uncertain renewable generation on least-cost system sizing of renewables and ESSs over a given time horizon. And to determine the optimal ESSs and to manage the high renewable penetration scenario.  The proposed algorithm is shown through a case study of a utility near Oklahoma City, which has significant renewable energy potential.  To perform Monte Carlo sampling, historical wind speed and solar irradiance data collected for the last 15 years from an atmospheric radiometric measurements (ARM) user facility is used. Probability distribution fitting is performed on this historical data for corresponding 15-min intervals year-to-year and probability distributions are selected using the sum square error method. These probability distributions are then used in conjunction with an optimization program with the objective function defined to minimize electricity costs over a 10-year period, including the system installation costs.  Monte Carlo sampling of optimal sizing helps to characterize the impact of uncertain generations on planning algorithms, particularly in high-renewable penetration scenarios.

This proposed optimization framework is applicable to estimate the system sizes for residential and utilities that are interested in minimizing the electricity cost with renewable energy and BESS.

Presenting Author: Yogesh Manoharan The University of Memphis

Presenting Author Biography: Yogesh Manoharan is a PhD student from The University of Memphis. He is working on Energy systems valuations and modeling. His interests are optimizing the energy storage system sizes, load dispatches to increase potential savings considering the investment and lifetime of the system. He is currently working on an optimization model that is aware of energy systems degradation.

Authors:

Yogesh Manoharan The University of Memphis
Alexander John Headley The University of Memphis