Leonardy Setyawan

Multilevel Energy Management System for Hybridization of Energy Storages in DC Microgrids

Hybridization of energy storages (ESs) with different ramp rates helps minimization of system bus voltage variation and extension of ESs lifetime in dc microgrids. Hybrid ES system (HESS) control is normally realized with centralized coordination. In this paper, HESS distributed control, which is independent of the communication link, is proposed to enhance system reliability. All ESs are configured as slack terminals to regulate system bus voltage with droop control. System net power decomposition and ESs power sharing are realized with localized low-pass filter (LPF) applied to ESs with low-ramp rates. The relationship between LPF cut-off frequency and ES ramp rate is elaborated in detail. However, bus voltage deviation and power tracking errors are the main drawbacks of HESS distributed control. Multilevel energy management system (EMS) is thus proposed to enhance system control accuracy. HESS distributed control is scheduled as the primary control. Bus voltage restoration and power sharing compensation are applied in secondary control to eliminate the voltage deviation and power tracking errors, respectively. In tertiary control, autonomous state of charge (SoC) recovery is implemented to limit SoC variation of ES with high-ramp rate. A lab-scale dc microgrid is developed to verify the proposed multilevel EMS for HESS control.

Design and stability analysis for an autonomous DC microgrid with constant power load

Fuel cell is a promising source in autonomous dc microgrid. Hybridization of fuel cell with battery is commonly implemented to overcome slow dynamic of fuel cell. Then battery compensates high frequency fluctuation and fuel cell provides consistent power at steady state. To achieve this objective, most control strategies require a centralized controller, which encounter reliability and scalability issues. This paper proposes a virtual capacitor droop controller to achieve autonomous dynamic power sharing at distributed level. An autonomous dc microgrid is designed to verify the proposed method. Considering the high penetration of constant power loads (CPLs) in dc microgrid and the destabilizing effect of CPLs, system stability is investigated with the implementation of the proposed controller. Both simulation and hardware experiment are conducted to validate the effectiveness of the proposed control method and analytical results.

Power-Capacity-Based Bus-Voltage Region Partition and Online Droop Coefficient Tuning for Real-Time Operation of DC Microgrids

Multiple-voltage-region control, in which the bus voltage range is divided into several regions, is usually implemented for DC microgrid operation in distributed manner. Voltage/power droop relationships are imposed for active power sharing among slack terminals. Conventionally, threshold voltages for voltage region partition are determined with fixed percentage of variation around the nominal value, which may result in unevenness of droop coefficients in different regions. If system droop coefficient is too high, significant bus voltage step change due to load variation will occur. On the other hand, significant power sharing error among slack terminals will be induced if the droop coefficient is too low. In this paper, a compromised solution with power capacity based bus voltage region partition is proposed to equalize the droop coefficients in different regions. However, the droop coefficients are determined based on the rated power capacity of system units. Bus voltage discontinuity appears when the power capacity reduces in actual implementation. To eliminate the voltage discontinuity, online droop coefficient tuning according to the real-time power capacity is implemented. Algorithms for local power capacity estimation of solar PV and battery energy storage have been proposed. A lab-scale DC microgrid has been developed for verification of the proposed methods.

Energy management system for control of hybrid AC/DC microgrids

Hybrid AC/DC microgrid is getting popularized due to the higher penetration of DC-compatible loads, energy sources and storages. To maintain system power balance in both AC and DC sub-grids, the bidirectional interlinking converter (BIC) is used. Distributed control of hybrid AC/DC microgrid is normally implemented to enhance system reliability. DC bus voltage and AC frequency are used for indication of system active power balance. BIC power flow is determined based on local information to equalize the normalized deviations of DC voltage and AC frequency. However, deviations of AC frequency and DC bus voltage degrade system power quality, which deteriorate the lifetime operation of system units. Multi-level energy management system (EMS) is thus proposed in this paper to enhance control accuracy while retaining system reliability. Distributed control of hybrid AC/DC microgrid is scheduled as the primary control. Restorations of AC frequency and DC bus voltage are implemented in secondary control. In tertiary control, the power references of system units are generated based on the comparison of marginal costs. Power sharing compensation is applied to minimize power tracking errors. MATLAB/Simulink model for hybrid AC/DC microgrid is developed for the verification of the proposed multi-level EMS.

Hierarchical control of active hybrid energy storage system (HESS) in DC microgrids

Energy storages can be characterized based on energy and power densities, ramp rate, etc. Hybridization takes advantages of all energy storages to enhance system performance. Hierarchical control which is comprised of both centralized and distributed control is proposed for HESS operation in this paper. In normal state, system operates with energy management system. The central controller is used to coordinate the power sharing among various energy storages based on their characteristics and operating statuses. In case of communication failure, distributed control which operates with local information is to be activated. The bus voltage is used as indicator for system power balance to realize distributed control of system units. MATLAB/Simulink is used for the verification of the proposed methods.

Implementation of sliding mode control in DC microgrids

Upon the increasing implementation of direct current (DC) power sources, energy storages and loads in power systems, DC microgrids have drawn more attention from global stakeholders. Reliable operation of DC microgrids mainly depends on proper control method to cope with bus voltage variations due to renewable sources intermittency and load changes. The main objectives of the control method are to maintain the operating bus voltage and the power balance. This control method is applied in battery energy storage (BES) component. By charging and discharging power to and from BES the control objectives can be achieved. Sliding mode control (SMC) method is implemented because of its robustness for sudden variations of power sources and loads. SMC is also of fast response in dynamics. Washout filter is applied to minimize the transient response. DC microgrid model is developed in Matlab/Simulink simulation to verify the performance of SMC. The comparison between SMC and the conventional proportional-plus-integral (PI) control is also carried out in the analysis.

Hybridization of energy storages with different ramp rates in DC microgrids

Hybridization of Energy Storages (ESs) with different characteristics is an effective and economic solution to enhance system performance. Hybrid Energy Storage System (HESS) in centralized coordination is normally implemented. To maintain system operation in case of communication failure, a novel algorithm is proposed for HESS distributed control in this paper. DC bus voltage is regarded as the global information carrier for indication of system power balance. All ESs are configured as slack terminals with droop control for obtainment of the reference bus voltages. Localized Low Pass Filters (LPFs) are added to ESs with low ramp rates to realize system net power decomposition and ESs power scheduling in distributed manner. To overcome the disadvantages of distributed control including voltage deviation and power sharing errors, multi-level Energy Management System (EMS) is applied. HESS distributed control is scheduled as the primary control. Bus voltage restoration and power sharing compensation are implemented in secondary control. Autonomous SoC recovery for ESs with high ramp rates is applied in tertiary control. Load shedding and generation curtailment are used to limit system net power range. MATLAB Simulink model is developed for the verification of the proposed method.

Optimal Depth-of-Discharge range and capacity settings for battery energy storage in microgrid operation

Battery energy storage (BES) plays an important role for mitigation of microgrids power imbalance induced by the intermittency of renewable sources and load changes. Due to high capital cost, optimal sizing of BES is crucial for economic operation of a microgrid. Conventionally, the optimal sizing of a BES is determined without considering the operating range of battery stored energy under varying system resource and load conditions. In this paper, both depth of discharge range and capacity are determined under the minimum system operation cost. Time varying resource and load conditions are considered in the optimization. The effects of BES lifetime degradation and energy not supplied are quantified and incorporated in system operation cost function. The BES lifetime degradation cost and customer interruption cost models are developed. The proposed method is implemented to a microgrid operation and verified using MATLAB simulation.

Power capacity based bus voltage region partition and online droop coefficient tuning for real-time operation of DC microgrids

Multiple-voltage-region control, in which the bus-voltage range is divided into several regions, is usually implemented for dc microgrid operation in distributed manner. Voltage/power droop relationships are imposed for active power sharing among slack terminals. Conventionally, threshold voltages for voltage region partition are determined with fixed percentage of variation around the nominal value, which may result in unevenness of droop coefficients in different regions. If system droop coefficient is too high, significant bus-voltage step change due to load variation will occur. On the other hand, significant power sharing error among slack terminals will be induced if the droop coefficient is too low. In this paper, a compromised solution with power-capacity-based bus-voltage region partition is proposed to equalize the droop coefficients in different regions. However, the droop coefficients are determined based on the rated power capacity of system units. Bus-voltage discontinuity appears when the power capacity reduces in actual implementation. To eliminate the voltage discontinuity, online droop coefficient tuning according to the real-time power capacity is implemented. Algorithms for local power capacity estimation of solar photovoltaic (PV) and battery energy storage have been proposed. A lab-scale dc microgrid has been developed for verification of the proposed methods.