Demand Response Management For Power Throttling Air Conditioning Loads In Residential Smart Grids

TL;DR

This paper proposes a demand response algorithm for residential smart grids that uses power throttling of air conditioners to reduce peak loads while managing user inconvenience, demonstrating significant load reduction with minimal states.

Contribution

It introduces a novel DRM plan modeling air conditioners as multi-state power throttling devices and analyzes the impact of additional power states on peak load reduction.

Findings

Adding one power throttling state reduces peak load significantly.

Peak load reduction diminishes with more than two power states.

Inconvenience duration and severity influence peak load reduction.

Abstract

In this paper we develop an algorithm for peak load reduction to reduce the impact of increased air conditioner usage in a residential smart grid community. We develop Demand Response Management (DRM) plans that clearly spell out the maximum duration as well as maximum severity of inconvenience. We model the air conditioner as a power throttling device and for any given DRM plan we study the impact of increasing the number of power states on the resulting peak load reduction. Through simulations, we find out that adding just one additional state to the basic ON/OFF model, which can throttle power to 50% of the rated air conditioner power, can result in significant amount of peak reduction. However, the peak load that can be reduced is diminishing with the increase in number of states. Furthermore, we also observe the impact of inconvenience duration and inconvenience severity in terms of peak load reduction. These observations can serve as useful guidelines for developing appropriate DRM plans.