RESHAPING NET LOAD OF HIGH RENEWABLE DISTRIBUTION FEEDERS WITH COMMERCIAL BUILDING MODEL PREDICTIVE CONTROL
Open Access
- Author:
- Ordorica, Salvador Adrian
- Graduate Program:
- Architectural Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- November 07, 2018
- Committee Members:
- Gregory Scott Pavlak, Thesis Advisor/Co-Advisor
- Keywords:
- MPC
Model Predictive Controlls
High Renewable Energy Pentration
Reshape Net Load - Abstract:
- The challenge associated with integrating higher levels of renewable energy sources into our existing electric grid systems is that these sources of energy are uncertain and their generation output is not in sync with the normal demands of both residential and commercial buildings. This study evaluates the effectiveness of using Model Predictive Control (MPC) to control a single medium office commercial building to reshape the net load in electric demand. This is done at the distribution feeder level in order to improve grid conditions for various levels of renewable energy penetration. This research evaluated 60 case studies that were simulated in a MATLAB simulation environment that covered five solar levels of energy ranging from 0 percent through 200 percent and four sky conditions: cloudy, sunny, ramp-up and ramp-down. Three building level metrics were tasked to the MPC as objective functions: peak-to-valley, load factor and system ramping. These building metrics were evaluated based on their effectiveness to reduce peak demand, increase load factor, and reduce system ramping in the net load observed while integrating higher levels of solar energy. The system ramping building metric was observed to consistently provide the best balance between building and grid level impacts. On average, for the optimized net load, peak-to-valley was reduced by 22.1 percent, load factor increased by 19.4 percent and system ramping reduced by 13.8 percent, thus achieving improvement in the overall grid level metrics. The system ramping building metric was observed to be the most energy efficient metric, capable of reshaping the net load by 7.39 percent on average less energy consumption as compared to the other two building metrics. Therefore, system ramping was observed as the superior building metric throughout all simulation case studies. In this research, it was found that using MPC within the tested simulation environment, a single building could be controlled to improve net load and reduce the severity of system ramping events that occur with high penetrations of solar energy. However, this comes at a consequence to the building level as it requires buildings to consume more energy which often leads to higher utility costs. Thus, markets or incentives would need to be created to compensate buildings for providing an improved load shape.