Arnold Sullivan

Mentor: Dr. Sean Meyn
College of Engineering
"I am very interested in working within the developing field of renewable energy, and Dr. Meyn's work seemed very interesting."


Electrical Engineering



Research Interests

  • Renewable energy integration
  • Demand response
  • Smart Inverters

Academic Awards

  • University Scholars Program 2016


  • Delta Upsilon
  • Power Engineering Club
  • IEEE


  • Reading Pals
  • Global Himalayan Expedition

Hobbies and Interests

  • Fencing
  • Biking
  • Reading
  • Live Music

Research Description

Using a Fridge as a Demand Response Resource
My research project is based on the creation of a “Smart Fridge” which will be capable of receiving a signal from a balancing authority and regulating its power consumption in order to provide ancillary service to the grid. The first step will be purchasing a fridge in order to create a model of its typical power usage in several situations (full fridge, empty fridge, increasing and decreasing the fridge and freezer temperature, etc.). This model will be compared to some of the existing models for fridge loads created at GE and NREL.Once the model has been created (estimated time line of 3-4 weeks to allow for different testing scenarios) we will create a board allowing us to interface with the fridge and vary its cooling cycle in terms of how much it is allowed to cool/heat back up and how quickly it cycles between on and off states, including defrost. This will be done automatically based on measurements of the BA signal and internal state measurements of the refrigerator. The algorithm to be implemented will be based on the randomized control architecture developed recently in Prof. Meyn’s lab. Based on these experiments we can obtain estimates of the cost of ancillary service from this approach. Two metrics to be investigated are increased cycles per month, and increased energy usage. We will compare the power consumption during these varied cycles to the power consumption of the fridge in its nominal state, and see what can be done to minimize power consumption while allowing the fridge to cycle as quickly as possible when regulation is required. This information will be used to investigate the frequency regulation possible with many thousands of refrigerators in a region, each controlled using the same local control algorithm. This part of the project will be based on simulations conducted on the the OPAL-RT grid simulator based in the lab of Prof. Bretas. The fridges will be controlled with randomized policies that focus on maximizing the possible frequency regulation while keeping the consumers from noticing any difference in their home appliances. We will be looking at the range of possible bandwidths over which our fridges could regulate the frequency and the magnitude of regulation available. At the same time, we will be refining the board used to control the fridge’s cycling as well as working on creating a standardized communications model by which the utility companies could send signals to the fridges. Our focus will be on one-way communication from grid operator to loads in order to avoid potential security flaws. Our eventual goal is to create a smart fridge that could be placed in consumer’s homes and contribute frequency regulation to the grid without any difference in power consumption or fridge performance.The concept can be applied to a wide range of loads beyond refrigerators.The combined contribution of all of these loads could provide valuable ancillary service to the grid, and help integrate variable power renewable energy sources into the grid without relying on expensive batteries.