Demand-side Management and Supply Issues in Energy Consumption

Following recent developments in demand-side management are presented according to the issues/problems for which the state-of-the-art solutions and their results have been provided by the researchers.

A Large Number of Energy Consumers

Within a distribution network, there are often a large number of residential electricity consumers, it is really difficult to include them all in the DSM. Amarasekara et al. concentrated on serving residential consumers with smart grid DSM at a large-scale distribution level. Since the huge level of users was the main barrier to integration, they introduced an Aggregated Method (AM) to reduce the severity of the problem. In general, by defining some laws, they generated micro-grids that aggregate the physical device and cost constraints of the underlying users. Such values should be shared between individual energy consumers after making aggregated decisions for a model. Their findings show that energy customers receive more advantages when fulfilling their energy requirements and operating conditions when using their model. Their overall analyses found that the conceptual model for developing flexible DSM programs can be readily implemented by the electricity market operators.

Minimizing Electricity/Operating Cost

The problem of reducing electricity costs for smart grids with unpredictable renewable energy generation and bi-directional sharing of energy was discussed by Liu and Hsu. Firstly, a two-stage robust optimization (RO) framework was developed, and two solution methods, Column-and-constraint generation (C&CG) and Scalable and robust demand-side management (SRDSM) were introduced to solve the problem. Results showed that the solutions produced by the SRDSM algorithm result in 5–6 percent higher energy costs; however, their calculation time is only 2–3 percent of that needed by the C&CG algorithm when the number of consumers varied from 10 to 30. Nevertheless, the C&CG algorithm failed to produce a response within 24 hours when the number of consumers rose from 100 to 500. In comparison, in a relatively short time, the SRDSM algorithm provided high-quality solutions (PAR < 1.005). This implies that while the C&CG approach provides optimum solutions but when the problem size is large, the SRDSM algorithm is much more robust and efficient.

Zheng et al. developed a load shifting algorithm based on economic linear programming with model predictive control to reduce the operational cost of a microgrid system based on biomass combined heat and power. The model regulates both electrical and thermal energy supply and demand as decision variables at the same time. Furthermore, the simulation of Monte Carlo was then used to produce probability distributions of cost to show investment risk. Results indicate that the suggested method for load shifting decreases operating costs by 6.06 percent and raises the fraction of renewable energy by 6.34 percent relative to the situation with the same micro-grid configuration but without demand-side management. Probability distributions created by cost simulation from Monte Carlo allow the investor in the smart grid to think about trade-offs between high investment and potentially increased operational threat.

Under a scenario with uncertainty, Yang presented a two-stage Real-Time DSM (RDSM) approach for response executors (REs) in a microgrid. In the first stage, a rolling dynamic optimization model based on model predictive control (MPC) was developed to minimize the daily total cost and preserve the supply-demand balance. A rule-based real-time power allocation control (RTAC) algorithm was implemented during the second stage to modify the first stage optimization results together with each RE's scheduled ability (SA) value. Many test cases were tested and the findings indicate that their approach was capable of effectively obtaining economic benefits for both MG and REs as well as increasing the MG system's net load characteristics. The approach was also active in increasing the supply-demand balance in order to increase the external power system's stability.

Hybrid Renewable Energy Systems

Recently, among researchers working in the field of stand-alone hybrid renewable energy systems, DSM has drawn great interest. Thiaux addressed the study of stand-alone hybrid PV / Diesel / Battery systems including a new DSM strategy called 'load shaping & authorization dispatching,' which can be described in a multi-subscriber configuration as an improved and optimized direct load control for domestic loads. In addition, a modern probabilistic model of consumer behavior focused on the Bayesian network and Monte Carlo simulation was applied to capture the demand's real-time and stochastic nature. The statistical results showed that the implementation of DSM could provide the hybrid PV system with a significant improvement in economic and ecological efficiency. In reality, for an unchanged daily consumption (18.5 kWh / day), consumers may experience the benefits of a reduced energy price of 11.3 percent.

Dynamic Charging of Electric Vehicles (EVs)

Electric vehicles (EVs) dynamic charging have the ability to greatly reduce the battery size needed for an acceptable range. Nonetheless, with reducing carbon emissions being the main driver of progressing EV development, attention is required as to how a dynamic charging system will affect such emissions. Humfrey, Sun, and Jiang introduced a demand-side management approach to allocating resources for dynamic charging of electric vehicles (EVs) including local renewable generation incorporation. In order to consider individual users, an energy retailer, and a regulator as players with conflicting interests, a multi-objective optimization problem was developed. Through modeling a regulator in the multi-objective optimization problem, the system works to reduce the power from the grid if carbon emissions are high per unit of energy. As this typically occurs at peak times, there was a reduction in demand during these periods, resulting in a smoother, more stable daily profile. The DSM approach lowered CO2 emissions by 22 percent and 42 percent respectively when compared to a first-come-first-served (FCFS) allocation method, with and without incorporated renewables.