Mechanical Engineering ETDs

Publication Date

Fall 12-12-2020

Abstract

Solar Insolation Micro-Forecasts (SIMF) are used by Independent Service Operators (ISOs) and other grid operators to maintain constant and stable electrical grid frequency, voltage, power factor and waveform across their transmission infrastructure. Intermittent, thick, dense and typically cumulus clouds negatively impact the electrical grid by quickly turning on and off power production from large solar photovoltaic (PV) fields, causing grid stability problems between generation and load. Forecasting insolation values over large PV fields allows operators the chance to anticipate and proactively implement mitigation strategies like engaging spinning reserve from gas turbines, deploying generators, buying power in the spot market, and or engaging grid tied battery/energy storage.

This research has built and fielded a deployable SIMF system. We utilized new sensors to alter the way clouds are imaged. We also employed a machine learning code (AI) called LAPART, that learns and generates five-minute accurate predictions of PV insolation values based on the specific spatial configuration of individual fields.

Keywords

solar irradiance forecasts, very short-term solar forecasting, intra-hour solar irradiance forecasts, solar micro forecasts, solar PV power integrations, solar cloud occlusion

Degree Name

Mechanical Engineering

Level of Degree

Doctoral

Department Name

Mechanical Engineering

First Committee Member (Chair)

Professor Andrea Mammoli

Second Committee Member

Professor Peter Vorobieff

Third Committee Member

Professor Svetlana V. Poroseva

Fourth Committee Member

Professor Manel Martinez-Ramon

Fifth Committee Member

Professor Emeritus Thomas Caudell

Document Type

Dissertation

Language

English

Download

Share

COinS