Impact of Scheduling Flexibility on Demand Profile Flatness and User Inconvenience in Residential Smart Grid System

Abstract

The objective of this paper is to study the impact of scheduling flexibility on both demand profile flatness and user inconvenience in residential smart grid systems. Temporal variations in energy consumption by end users result in peaks and troughs in the aggregated demand profile. In a residential smart grid, some of these peaks and troughs can be eliminated through appropriate load balancing algorithms. However, load balancing requires user participation by allowing the grid to re-schedule some of their loads. In general, more scheduling flexibility can result in more demand profile flatness, however the resulting inconvenience to users would also increase. In this paper, our objective is to help the grid determine an appropriate amount of scheduling flexibility that it should demand from users, based on which, proper incentives can be designed. We consider three different types of scheduling flexibility (delay, advance scheduling and flexible re-scheduling) in flexible loads and develop both optimal and sub-optimal scheduling algorithms. We discuss their implementation in centralized and distributed manners. We also identify the existence of a saturation point. Beyond this saturation point, any increase in scheduling flexibility does not significantly affect the flatness of the demand profile while user inconvenience continues to increase. Moreover, full participation of all the households is not required since increasing user participation only marginally increases demand profile flatness.

Home energy management systems: A review of modelling and complexity

“…For example, the works [12,15,25,61,74,75,91,93] state that refrigerators are uncontrollable loads, while they are considered uninterruptible loads in [62,63,68], interruptible loads in [22,24], and regulating loads in [21,[47][48][49]67,90]. This may be explained by the fact that refrigerators have relatively short cycling characteristics (20-45 min) and thus may be perceived as an uncontrollable load at an hourly resolution, while being controllable at the minute resolution.…”

“…Quadratic deviation penalty is quadratic with respect to the deviation from the ideal operation of a device, such as in [34,93]. Note that in mathematical optimization problem for state-dependent device models, this implies at least quadratic programming (QP) if the penalty is included within the objective, and quadratically constrained quadratic programming (QCQP) if it is part of the set of constraints.…”

“…Table 10 summarizes the multi-objective optimization methods used in HEMS literature. In 32 of the 35 cases within Table 10, a priori approaches are presented, while a posteriori approaches are used in [58,93,105]. The a priori methods may be preferable as they allow consumer preferences to be defined once at the outset, and do not require ongoing inputs from the consumer.…”

“…Consumption and well-being [93] Cost, consumption, emissions [16] Cost, well-being, emissions, consumption [43] requirements, as well as act based on signals received from the HEMS. Devices with these improved capabilities are called smart appliances or smart devices [21,77,106,117].…”

Locational Pricing to Mitigate Voltage Problems Caused by High PV Penetration

“…First of all, the DSO does not need to know the actual uncontrollable load of the households to approximate the voltage caused by the uncontrollable load alone. Since this could contain privacy-sensitive information, the customers preferably do not share this information in a central instance [10]. If the DSO knows the consumption of the PV panels and of the flexible loads during the voltage measurement, he can calculate the effect that these have on the voltage measurement with the voltage sensitivity factors.…”