Uncertainty Quantification in Agrivoltaics System Sizing with Load Disaggregation and Decomposition

Document Type

Conference Proceeding

Publication Date

1-1-2025

Abstract

The lack of high-resolution metering data presents challenges in accurately designing photovoltaic (PV) and battery energy storage systems (BESS) for agricultural applications, i.e., agrivoltaics. This study employs a Monte Carlo-based load disaggregation and decomposition approach to generate hourly load profiles from monthly utility data, enabling uncertainty quantification in system sizing and financial assessment. A case study was conducted on three farms in Louisiana, where around 300 Monte Carlo-based stochastic hourly load profiles were generated per site to evaluate the impact of load uncertainty on PV and BESS design. The schedules, developed using site visits and interviews with farm owners to establish proper operational ranges, demonstrated significant variability while maintaining a monthly CV(RMSE) within the 15% threshold. Results indicate that system sizing is highly sensitive to load decomposition assumptions, leading to variability in PV and BESS capacity requirements across different scenarios. While the simple payback period remains stable at around 9 years, financial savings exhibit greater uncertainty, with potential variations of over 10%–15% due to fluctuating system configurations. These findings highlight the economic risks associated with load characterization uncertainty and emphasize the need for improved data resolution or modeling techniques to enhance the accuracy and reliability of renewable energy investment decisions.

Publication Source (Journal or Book title)

ASHRAE Transactions

First Page

256

Last Page

264

This document is currently not available here.

Share

COinS