The microgrid is a power distribution system that supplies power from distributed generation to end-users. Demonstration projects and R&D regarding microgrids are currently in development in several advanced countries. In South Korea, renewable energy-based microgrid demonstration projects are carried out mainly as island or university campus grids. These R&D efforts aim to popularize microgrid systems in South Korea while considering the limited land availability, which impedes the widespread distribution of photovoltaic systems and the microgrid market’s growth. This study presents a floating photovoltaic system configured as an island microgrid combined with a hybrid power system. The floating photovoltaic system is configured on an idle water body integrated with an existing pumped hydroelectric system. The integration of a current pumped hydroelectric system minimizes a battery energy storage requirement, which compensates for the renewable energy sources’ intermittent power output. We evaluate the optimal power flow of the setup using a reliability index to ensure a stable power supply within the standalone microgrid and maximize the supply power range according to the demand response.

Massive electrical load exhibits many patterns making it difficult for forecast algorithms to generalise well. Most learning algorithms produce a better forecast for dominant patterns in the case of weekday consumption and otherwise for less dominant patterns in weekend and holiday consumption. In view of this, there is the need to cluster the load patterns, so learning algorithms can focus on the patterns independently to produce forecasts with better accuracy for all cases. However, clustering time-series data breaks the time-series dependency, making model training difficult. This paper presents a novel sequence-to-sequence cluster framework to reform time-series dependency after clustering; this enables independent clusters to be modelled using Convolutional Neural Network-Gated Recurrent Unit, which learns spatiotemporal features for future forecasts. A real-world dataset by the Korea Power Exchange composed of nationwide consumption is used for case studies and experiments. Experimental results verify that the proposed study effectively improves the accuracy of electric load forecasting by about 50%, with a WAPE of 0.67%. The proposed method also speeds up the training process of the forecast algorithm by about 35%, given that only a subset of the dataset is trained due to clustering. Korea Water Resources Corporation has implemented the proposed method for load forecasting and system marginal price estimation.

A demand side management technique is deployed along with battery energy-storage systems (BESS) to lower the electricity cost by mitigating the peak load of a building. Most of the existing methods rely on manual operation of the BESS, or even an elaborate building energy-management system resorting to a deterministic method that is susceptible to unforeseen growth in demand. In this study, we propose a real-time optimal operating strategy for BESS based on density demand forecast and stochastic optimization. This method takes into consideration uncertainties in demand when accounting for an optimal BESS schedule, making it robust compared to the deterministic case. The proposed method is verified and tested against existing algorithms. Data obtained from a real site in South Korea is used for verification and testing. The results show that the proposed method is effective, even for the cases where the forecasted demand deviates from the observed demand.

Robust operation of load management control for a building is important to account for the uncertainty in demand as well as any distributed sources connected to the building. This paper discussed an online load management control solution using distributed energy storage (DES) while considering uncertainties in demand as well as DES to reduce peak demand for economic benefit. In recent years’ demand-side management (DSM) solutions using DES such as stationary energy management system (BESS) and plugged-in electric vehicles (PEV) have been popularised. Most of these solutions resort to deterministic load forecast for the day ahead energy scheduling and do not consider the uncertainties in demand and DES making these solutions vulnerable to uncertainties. This study presents an online density demand forecast, k-means clustering of PEV groups and stochastic optimisation for robust operation of BESS and PEV for a building. The proposed method accounts for uncertainties in demand and uncertainties due to mobile energy storage as presented by PEVs. For a case study, we used data obtained from an industrial site in South Korea. The verified results as compared to other methods with a deterministic approach prove the solution is efficient and robust.

In South Korea, Renewable Energy Sources (RES) have been increasing with the application of energy policies, such as Feed in Tariff (FIT) and the Renewable Portfolio Standard (RPS). However, a rapid increase in RES supply leads to an uncertain power supply due to the intermittent output of RES. A representative example is the curtailment of Wind Turbines (WT), which frequently occurs in Jeju Island, South Korea. The proportion of RES power on Jeju Island is 67%, and there are cases where WT is curtailed among the operable sections to maintain the balance of power supply and demand. This paper applies Power-to-Gas (P2G) technology to hydrogenate, store, and utilize unused power to solve this problem. In this paper, Aewol-eup in Jeju Island is selected as a target site for case study. An Integrated Energy System (IES) for various energy operations is designed to control RES output. This paper proposes the optimal facility configuration and finally drives the optimal design and operation solution of IES by analyzing the objective functions and focusing on the Independent Power Producer (IPP) perspective.

A Demand-side management technique are deployed along with battery energy-storage systems (BESSs) to lower the electricity cost by mitigating the peak load of a building. Most of the existing methods rely on manual operation of the BESS, or even an elaborate building energy-management system resorting to a deterministic method that is susceptible to unforeseen growth in demand. In this study we propose a real-time optimal operating strategy for BESS based on density demand forecast and stochastic optimization. This method takes into consideration uncertainties in demand when accounting for an optimal BESS schedule, making it robust compared to the deterministic case. The proposed method is verified and tested against existing algorithms. Data obtained from a real site in South Korea is used for verification and testing. The results show that the proposed method is effective, even for the cases where the forecasted demand deviates from the observed demand