Energy Management System for Grid-Connected Nanogrid during COVID-19

Abstract

An effective energy management system (EMS) was designed based on the Stateflow (SF) approach for a grid-connected nanogrid (NG) composed of a photovoltaic (PV) array with a battery bank and supercapacitor (SC) energy storage system (ESS). The PV energy system, battery bank and SC (ESS), dual active bridge DC/DC converters, DC/AC inverters, control algorithms, and controllers were developed to test the operation of the NG. The average and high-frequency power components are separated using frequency division of the ESS power utilizing a low-pass filter; the average power is absorbed by the battery bank, while the high-frequency power is absorbed by the SC. The aim of this paper is to design an EMS to manage the energy of a grid-connected NG system considering the availability of the PV array, ESS, and demand requirements. Different scenarios of operation were tested to check the EMS behaviour during the day with a random demand profile, including: (1) a PV array with the grid supplying the load without an EMS; (2) a PV array, batteries, and the grid supplying the load with an EMS; (3) a PV array, batteries, an SC, and the grid supplying the load with an EMS; (4) a PV array, batteries, an SC, and the grid supplying the load with an EMS, with load profile reduction by 20% due to COVID-19. As per the simulation results, the proposed EMS enables the flow of power in the NG system and demonstrates the impact on the ESS by minimising carbon emissions via a reduction in grid consumption. Furthermore, the SF method is regarded as a helpful alternative to popular design approaches employing conventional software tools.

1. Introduction

Nanogrid (NG) and microgrid (MG) systems make use of numerous renewable energy sources (RES) such as solar photovoltaic (PV) and wind systems. Due to the intermittent nature of solar PV and wind systems, an energy storage system (ESS) such as a supercapacitor (SC), batteries, and a flywheel can be used to support the output power that is produced from the RES. The ESS is designed to be used with intermittent RES for output power levelling, load following, peak load shaving, and energy arbitrage [1,2].

In the past few years, energy management systems (EMS) of NG systems are regarded as an important research topic. Many researchers have presented different options that could be used for EMSs in different distribution systems. The EMS used in grid-connected and islanding NG systems helps to optimize energy generation, power flow, and usage amongst the distributed energy sources [3,4].

EMSs are used in distribution systems to control the rising demand for electric power, reducing future emissions and producing socio-economic benefits through sustainable growth [5].

The researchers in [6] developed an EMS for an NG that included PV, a battery and an SC for a house. The proposed EMS aim is to improve battery charging and discharging activity and alleviate the power fluctuation between the NG and main grid using a smoothing function.

EMSs based on the Stateflow (SF) method primarily cater to islanded and grid-connected MGs, which include diesel generators, PV arrays, and ESSs. The researchers put forward an integrated framework pertaining to MGs for dealing with the issue of load shedding, as well as for designing and integrating a resilient EMS. As per the simulation results, the designed EMS not only enables the supply of uninterrupted power during times of load shedding, but also helps to reduce grid burden by amplifying the utilisation of PV energy [7].

In [8], the researchers implemented an EMS by using an SF approach for grid-connected MGs composed of PV and battery storage. The objective of this work was to ensure power flow in the MG system. The simulation was carried out using a 20-s weather profile to demonstrate the efficiency of the proposed EMS.

In [9], an EMS was proposed for island MGs that includes PV, wind, a battery, and an SC. The aim of the paper is to minimize the fluctuation between the MG and loads, and to improve the stability of the grid voltage by using a combination of batteries and SC.

In [10], a new EMS was designed for a grid-connected MG that included a battery, PV energy, and an SC. The key benefits pertaining to the EMS included faster DC link voltage regulation, effective power sharing between the SC and battery, decreased charge/discharge actions pertaining to the battery during transient power fluctuation and steady state, seamless mode transitions, and enhanced power quality features of the AC grid. Both experimental and simulation studies validated the effectiveness of the EMS.

An optimal EMS was proposed in [11], coupled with a stochastic framework pertaining to uncertainties modelling concerning DC-isolated NGs, including battery storage, a PV array, a diesel generator, and wind energy, as well as electric vehicle charging infrastructure. A case study was put forward and various results were noted to confirm the formulation and analyse the effect of demand response programs pertaining to the grid operation. The obtained results validate the significance of flexible consumption for increasing the retailer’s monetary income and decreasing diesel consumption, and, consequently, CO₂ emissions. The role of public charging infrastructure has been remarkable with regards to the grid’s profitability and how demand response programs can impact viability.

In [12], an EMS was proposed pertaining to battery storage systems with regards to grid-connected MGs. The determination of a battery’s discharging/charging power is done in a way as to keep the overall energy consumption cost at the minimum level, considering variations in the grid tariff and load demand, as well as renewable power generation. The system was modelled as an economic load dispatch optimisation problem over a 24-h horizon, while mixed-integer linear programming (MILP) was employed to solve the problem. Thus, this formulation requires knowledge pertaining to the production of expected renewable energy power and load demand in the next 24 h. To achieve this, the authors proposed a long short-term memory (LSTM) network. As per the simulation results, the proposed real-time strategy was seen to be better than the offline optimisation strategy, decreasing the operating cost by 3.3%.

In [13], development of an optimal power scheduling controller was conducted with regards to the EMS of distributed energy resources in MG systems. The lightning search algorithm is employed to implement the developed optimised controller to address the uncertainties of MG energy management, as well as to offer optimum power delivery for loads by keeping the cost at its minimum. The proposed optimised controller’s objectives include: (i) to decrease the total operating cost pertaining to the distributed energy resources units, (ii) to develop an optimised controller with regards to MG energy management, and (iii) to minimise environmental emissions. The execution of the optimised controller was carried out based on real-time load-varying conditions as recorded in Perlis, Malaysia. It was observed that the optimised controller was able to successfully decrease the power consumption level from 971.65 MW to 364.3 MW, which, in turn, led to cost savings of RM 265432.06. The proposed scheduling optimised controller’s performance was compared with the search algorithm optimisation employed in validation. The results demonstrated that a cost-effective system was produced with the lightning search algorithm-based MG controller, achieving 62.5% cost savings and CO₂ emission reduction of 61.98%, which was much higher compared to the MG and the backtracking search algorithm-based MG optimisation.

In [14], an EMS is proposed with regards to DC MGs that includes PV, a microturbine (MT), fuel cells (FCs), a battery, and a diesel generator (DE), as well as Egyptian grid load profiles over the four seasons occurring in a year. This work aims to decrease CO₂ emissions, as well as the overall cost. The generation cost model includes the no-load cost and nonlinear behaviour pertaining to losses within distributed generation (DG). Furthermore, the authors accounted for the start/stop cost pertaining to DG, as well as the power losses occurring with the MG. The proposed optimisation problem becomes a non-convex mixed-integer nonlinear (MINL) problem, which is regarded as difficult to solve. The paper proposes employing the branch-and-reduce optimisation navigator (BARON) algorithm, which is regarded as a global optimisation technique.

The key goal of the above studies is to control and manage the power distribution in NG and MG systems, similar to this work’s key objective; however, the EMS proposed in this paper is different when compared to the earlier studies, as it includes the additional objective of reducing grid carbon emissions and consumption by using an ESS, and also takes into account the impact of the COVID-19 pandemic on the NG system, which was not accounted for in the above research works. Furthermore, this paper includes an EMS designed via the SF approach, which only few papers have previously reported. The SC has does not take into account the SF method as found in the literature, which is presented in this work. Modelling an EMS using SF offers various advantages such as decreasing the complexity, enhancing the ability to reconfigure operating conditions, and simplifying control tasks.

This paper aims to design an EMS catering to grid-connected NGs in order to efficiently manage energy distribution amongst battery banks, PV arrays, and SCs. This also offers flexible control pertaining to various scenarios of alterations in the PV array source and changes in load demand, as well as reductions in carbon emissions.

The paper is organised as follows. Section 1 presents an introduction of nanogrid EMSs. Section 2 describes the NG system design with an EMS in detail. Section 3 presents the results and discussion. The conclusions of this paper are presented in Section 4.

2. Nanogrid Description

The grid-connected nanogrid (NG) with an EMS is shown in Figure 1 It can be divided into a power circuit consisting of: (1) a PV array, battery bank, SC, DC/DC boost converter, dual active bridge (DAB) converters, DC/AC inverters, LCL filters. and three-phase dynamic load modelled in MATLAB Simulink; and the controllers, including (2) local controllers and an EMS designed using Stateflow (SF).

The (PV) array is operated with maximum power point tracking (MPPT); the batteries and the SC are used to maintain the energy balance in the system. By applying a low-pass filter, the batteries will handle the average power and the SC will handle the high-frequency power (HFP). An EMS is used to manage and control the power flow under different conditions to feed the load through the DC/DC and DC/AC converters.

2.1. Photovoltaic (PV) and Boost Converter

The PV array comprises of 14 PV panels connected in series and four PV strings connected in parallel, which are modelled in MATLAB. The parameters are set as maximum voltage output of PV panel V_m = 36.7 V, maximum current output of PV panel I_m = 8.18 A, open circuit voltage of PV panel V_oc = 45.3 V, short circuit current of PV panel I_sc = 8.68 A, maximum power of PV panel Pm = 300 W, and number of cells of PV panel = 72. The power curve of the PV array with irradiation data for 24 h at a constant temperature of 25 °C is shown in Figure 2. The power reaches 16.7 kW at its peak.

The maximum power of the PV array can be calculated by using:

Ppv, max = V_m I_m(1)

where V_m is the maximum voltage output of the PV array and I_m is the maximum current output of the PV array. To control the PV array’s power, MPPT is used based on the perturb and observe (P&O) algorithm. This algorithm calculates the duty cycle used by the DC/DC converter to reach the maximum power that the PV array can supply [15]. The boost converter is a highly effective power conversion device in which the input voltage is stepped up without the use of a transformer [16]. Figure 3 shows the basic diagram of the boost converter.

The inductance and capacitance parameters are calculated by using (2)–(4) for the modelling of the DC/DC boost converter [16].

(2)$L=\frac{V_{pv}\left(V_{out}-V_{pv}\right)}{\Delta I F_{sw boost}{V}_{out} }$

(3)$C_{in}=\frac{p_{PV,MAX}}{2fs{w}_{boost}{v}_{PV}\Delta v_{PV}}$

(4)$C_{out}=\frac{p_{PV,MAX}}{2fs{w}_{inverter}{v}_{out}\Delta v_{out}}$

where $L$ is the inductance, $Vpv$ is the PV array voltage, $Vout$ is the output voltage, $\mathsf{\Delta }I \mathrm{is} \mathrm{the}$ current ripple, $fs{w}_{boost}$ is the converter switching frequency, $p_{PV,MAX}$ is the maximum PV array power, $C_{in}$ is the input capacitor, $\mathsf{\Delta }vpv$ the voltage ripple of PV, $C_{out}$ is the output capacitor, $fs{w}_{inverter}$ is the inverter switching frequency, and $\mathsf{\Delta }{v}_{out}$ is the output voltage ripple. Table 1 shows the parameters used to design the boost converter.

2.2. Energy Storage System with Dual Active Bridge

In this work, the energy storage system consists of three lithium-ion batteries with a supercapacitor. Lithium-ion batteries have many features such as high-performance, high-energy density with low power duration (within one hour), and a long life cycle of more than 1000 cycles [17].

On the other hand, the SC has low-energy density with higher power duration (within seconds) compared to the batteries, and the SC shows a longer life span of more than 10⁵ cycles [18].

The batteries’ charge and discharge actions depend on the load demand, available power, and the state of charge (SOC) of the batteries. The energy constraint of the batteries is determined based on the SOC limits.

SOC_min ≤ SOC ≤ SOC_max(5)

where SOC_min and SOC_max are the lower and the upper limits of the battery SOC, respectively.

On the other hand, the SC charge and discharge actions depend on the load demand, available power, and the voltage (Vsc) of the SC. The energy constraint of the supercapacitor determined based on the Vsc limits.

Vsc_min ≤ Vsc ≤ Vsc_max(6)

where Vsc_min and Vsc_max are the lower and the upper limits of the supercapacitor Vsc, respectively.

The frequency division of power in the energy storage system utilizing a low-pass filter is applied to obtain the power average that is absorbed by the batteries and the high- frequency power that absorbed by the SC. The following equations have been used for this reason [19]:

(7)$P_{avg}=LPF(P_{load}-P_{pv}) = LPF\left(P_{ESS}\right)$

(8)$P_{avg}=\frac{f_{LPF}2\pi }{s+f_{LPF}2\pi }{P}_{ESS}$

(9)$P_{HFP}=P_{ESS}-P_{avg}$

where $f_{LPF}$ is the low-pass filter frequency, selected as 20 Hz, P_avg is the power average, $P_{HFP}$ is the high-frequency power, $P_{ESS}$ is the power of energy storage system, and S is the numerator coefficient. Figure 4 shows the frequency division of the power ESS using a low-pass filter. Table 2 shows the parameters of the ESS.

Figure 5 demonstrates the dual active bridge (DAB) converter, which was originally suggested in [20], with a more detailed analysis provided in [21]. The DAB can be defined as a bidirectional DC/DC converter with galvanic isolation that includes two active bridges that are interfaced and a high-frequency transformer placed between the bridges, as well as leakage inductance. The transformer provides galvanic isolation and also offers a high conversion ratio. For these reasons, a DAB is selected when there is a significant difference in output and input voltages [22]. This converter offers the benefits of high efficiency, soft-switching commutations, and automatic bidirectional power flow [23].

The leakage inductance and input and output capacitance parameters are calculated using the following equations for the modelling of the DAB converter [24].

(10)$L=\frac{\left(1-\left|d_{MAX}\right|\right)d_{MAX}{V}_{in}{V}_{out}}{2fs n P_{MAX}}$

(11)$C_{i_n}=\frac{p_{MAX}}{2fs v_{in }\Delta v_i}$

(12)$C_o=\frac{p_{MAX}}{2fs v_o \Delta v_o}$

where L is the leakage inductance, $C_{i_n}$ is the input capacitor, $C_o \mathrm{is} \mathrm{the}$ output capacitor, $d_{MAX}$ is the maximum duty cycle, $V_{in}$ is the input voltage, $V_{out}$ is the output voltage, $fs$ is the switching frequency, $n$ is the turn ratio of the transformer, $p_{MAX}$ is the maximum power, $\Delta v_i$ is the input voltage ripple, and $\Delta v_o$ is the output voltage ripple. Table 3 shows the parameters used to design the DAB converter.

2.3. DC/AC Inverters with LCL Filters

One of the most significant electronic components inside microgrid and nanogrid systems is the DC/AC inverter, which is required to convert DC electricity coming from the PV and ESS into AC electricity [25].

Figure 6 demonstrates the topology pertaining to DC/AC PV and ESS inverters along with the LCL filter. At the PV and ESS inverter outputs, LCL filters are employed to filter out the harmonic content that is not required. The procedure mentioned in [26] was used to design the LCL filter pertaining to the power converter.

Table 4 shows the system parameters. Based on these values, calculation of the parameters pertaining to the LCL filter is carried out by employing (13)–(19) [27].

(13)$Z_b=\frac{V_g^2}{P_n}$

(14)$C_b=\frac{1}{2\pi f_g{Z}_b}$

(15)$C=0.05C_b$

(16)$Ima{x}_g=\frac{p_n}{V_g}$

(17)$L_1=\frac{V_{DC}}{6f_{sw}\Delta I}$

(18)$k=\frac{1}{\left|1+r(1-L_{1}C{\left(2\pi f_{sw}\right)}^2)\right|}$

(19)$L_2=r{L}_1$

where $Z_b$ is the base impedance, V_g is the grid voltage, $P_n \mathrm{is} \mathrm{the}$ rated power, $C_b$ is the base capacitor, $f_g$ is the grid frequency, C is the filter capacitor, $Ima{x}_g$ is the grid maximum current, L₁ is the inverter-side inductance,$V_{DC}$ is the DC voltage, $f_{sw}$ is the switching frequency, $\Delta I$ is the current ripple. $k$ is the ripple attenuation, $r$ is the ratio, and $L_2$ is the grid side inductor.

2.4. Energy Management System (EMS) Using Stateflow (SF)

The Stateflow (SF) method is adapted for the EMS architecture. This method can be defined as an event-based modelling toolbox in MATLAB that is employed for modelling logic to dynamically control the EMS. To gain control of the power flow, it utilises the available energy information based on the battery bank with the SC, PV array and load requirements. It also offers control signals to the power conversion devices (inverters and DAB), which are generally employed in the integration of the NG system.

The EMS is the main controller that controls and coordinates all control action in the NG system. All the controllers of the DAB and inverters presented in previous sections operate based on the EMS algorithm. The DC/DC boost converter of the PV array is operated in MPPT mode. The DAB converters of the ESS operate in charging or discharging mode and keep the DC bus voltage constant. The power in the NG must be balanced under different operations of power from the PV array, the ESS, and load demand. The power balance equation is given as follows:

P_pv + P_batteries + P_sc = P_Load(20)

The various operation modes of the EMS are presented in Figure 7 and implemented by using the SF method in Figure 8.

There are different operation modes of the EMS. Each mode depends on four conditions; the first is the power generation from the PV array, the second is the SOC of the battery, the third is the Vsc of the supercapacitor, and the fourth is the high-frequency power.

When the power generated from the PV arrays is less than the load demand, the power is supplied by the batteries until the SOC reaches SOC_min, and by the SC until the Vsc reaches Vsc_min, while the batteries’ average power is shared among the three batteries (for example, if the average power is 3 kW, that means that each battery discharges 1 kW). On the other hand, the SC is managed by the high-frequency power. The SC can discharge a maximum of 5 kW; then, if the high-frequency power is greater than 5 kW, the grid will support the SC, while if the high-frequency power is less than 5 kW, the SC will supply the actual power required. Furthermore, in order to maintain power balance, the power grid will supply the load if the batteries reach SOC_min and if the SC reaches Vsc_min. On the other hand, the desired operation mode is when the power generated from the PV array is greater than the load demand. In this mode, the batteries and SC are charged until SOC_max and Vsc_max, respectively; if the SOC of the battery bank reaches SOC_max and Vsc reaches Vsc_max, then the surplus power will be sent to the grid. Note that a1, a2, and a3 are the statuses of batteries 1, 2, and 3, respectively (1 for discharge, −1 for charge, and 0 for idle). Psc_ref is the status of the SC (5000 for discharge, −5000 for charge, and 0 for idle).

3. Results and Discussions

This section discusses four different operation scenarios for testing the EMS’ behaviour during the day, with a random demand profile and irradiation data for the PV with a constant temperature of 25 °C, including: (1) a PV array with the grid supplying the load without EMS; (2) a PV array, batteries, and the grid supplying the load with EMS; (3) a PV array, batteries, a supercapacitor, and the grid supplying the load with EMS; (4) the PV array, batteries, a supercapacitor, and the grid supplying the load with EMS and load profile reduction by 20% due to COVID-19. Finally, this section discusses the impact of the energy storage system on the energy consumption of the grid and how it can minimize grid emissions.

3.1. The Scenarios

Four different operation scenarios are presented in this section.

3.1.1. First Scenario

This scenario consists of a PV array with the grid supplying the load without EMS during the day.

The objective of this case is to deliver power to the load for different generating conditions. Data for solar irradiation used for one day and changes every 30 min.

The power at different locations of the NG is shown in Figure 9. From midnight to 7 h, the grid is supplying power to the load, as the power generated by PV arrays is zero. After 7 h, the PV array starts to generate the load at different instants, while the surplus power from the PV array is exported to the grid until 17 h. After 17 h, the grid feeds the load until 24 h, as the power generated by PV array is zero.

3.1.2. Second Scenario

In this scenario, the EMS is applied for the PV array and batteries, with the grid feeding the load during the day.

The objective of this scenario is to deliver power to the load for different generating conditions with minimum supply from the grid.

The power at different locations of the NG is shown in Figure 10. Until 7 h, the batteries and the grid feed the load, as the power generated by the PV array is zero; after 7 h, the PV array starts feeding the load at different instants, while the surplus power from the PV array charges the batteries by same amount of power for each one until 17 h. After 17 h, the battery bank feeds the load until 24 h, as the power generated by the PV array is zero. Figure 11 shows the charging and discharging action of the batteries during the day.

3.1.3. Third Scenario

In this scenario, EMS is applied for the PV array, batteries, and SC, with the grid feeding the load during the day.

The objective of this scenario is to satisfy the load for different generating conditions with minimum supply from the grid.

The power at different locations of the NG is shown in Figure 12, Until 7 h, the batteries with the supercapacitor supply the load, as the power generated by the PV array is zero; note that the frequency division of the power and energy storage system utilizing a low-pass filter is applied to obtain the average power, which is absorbed by the batteries, and high-frequency power, which is absorbed by the supercapacitor. After 7 h the PV array begins feeding the load at different instants, while the surplus power from the PV array charges the batteries and the supercapacitor until 17 h. After 17 h, the battery bank with the SC feeds the load until 24 h, as the power generated by the PV array is zero. Figure 13 shows the charging and discharging action of the batteries and SC during the day.

3.1.4. Fourth Scenario

In this scenario, EMS is applied for the PV array, batteries, and SC, with the grid feeding the load during the day, with load profile reduction by 20% due to COVID-19.

The objective of this scenario is to satisfy the load for different generating conditions and to check the impact of the reduction of the load during COVID-19 on the NG.

The power at different locations of the NG is shown in Figure 14. Until 7 h, the batteries with the supercapacitor feed the load, as the power generated by the PV array is zero. After 7 h, the PV array starts feeding the load at different instants, while the surplus power from the PV array charges the batteries and the supercapacitor until 17 h. After 17 h, the battery bank with the SC feeds the load until 24 h, as the power generated by the PV array is zero. Because of the minimization of the load, the batteries charged more and discharged less than in the previous scenarios. Hence, the impact of COVID-19 could be positive for the NG system, as it reduced the charging and discharging behaviour of the ESS, which will increase the life span of the ESS. Figure 15 shows the charging and discharging action of the batteries and SC during the COVID-19 situation.

3.2. Minimization of Energy Consumption and Emissions from the Grid

The aim of this section is to find the impact of the ESS on the grid supply to the load during different operation conditions. Figure 16a shows the grid power for the first scenario, Figure 16b shows the grid power for the second scenario, and Figure 16c shows the grid power for the third scenario.

From above figures, it is observed that using batteries in NG can minimize the power imported from the grid. In addition, using the SC with batteries is considered the best option for minimizing power imports from the grid. On the other hand, from the above grid power curves, the energy consumption in kWh for each scenario was calculated, and is shown in Figure 17a–c, respectively.

The grid energy import is calculated by using following equation

(21)$Energy={\int }_{t_0}^t{P}_{grid}dt$

where $P_{grid}$ is the grid power; note that if the power is positive, the energy will increase, and if the power is negative, the energy will decrease. Figure 18 shows the energy calculation while MATLAB function is provided in Appendix A.

After calculating the energy consumption from the grid, we can note that the batteries and SC can effectively minimize the energy imported from the grid. By using the following equation, the grid carbon emission savings can be calculated:

CO₂ emission amount = emission factor in kg/kWh × energy in kWh(22)

The energy consumption from the grid in the first scenario is 14 kWh per day, in the second scenario, it is 1.9 kWh per day, and in the third scenario, it is 1.05 kWh per day. The grid emission factor in Malaysia is considered as 0.63 kg/kWh. Hence, we could save 7.623 kg of emissions per day by using batteries, and we could save 8.1585 kg of emissions per day by using batteries and an SC together. Table 5 shows the impact of batteries and an SC on minimize the grid emissions per day, month, and year.

The system is validated by existing work [28] and the EMS was improved. In summary, the advantages of the proposed EMS for NG system are as follows. The system significantly monitors power-sharing within the NG system by using an EMS algorithm. The proposed model reduces energy grid consumption and CO₂ emissions by efficiently controlling the batteries and SC, justifying the use of an SC to manage the high-frequency power, which minimized stress on the batteries. Finally, the impact of COVID-19 on NG systems has been investigated, showing how the decreased load affects the EMS. Future work will verify the EMS using experimental results and compare these with the simulation results.

4. Conclusions

An effective energy management system is designed using the Stateflow approach for a grid-connected nanogrid composed of a photovoltaic (PV) array, a battery bank, and a supercapacitor.

The nanogrid was modelled with a PV array, an energy storage system, a power conversion system, and an EMS. The simulation results with random demand profiles are discussed for four operation scenarios: (1) a PV array with the grid feeding the load without an EMS; (2) a PV array, batteries, and the grid feeding the load with an EMS’ (3) a PV array, batteries, an SC, and the grid feeding the load with an EMS; (4) a PV array, batteries, an SC, and the grid feeding the load with EMS, with load profile reduction by 20% due to COVID-19. The simulation results showed that the proposed EMS reduces the imported grid power and carbon emissions more efficiently when both batteries and a supercapacitor are used in the energy storage system. A further reduction of imported grid power and carbon emissions was obtained during the COVID-19 pandemic due to reduced load demand.

Author Contributions

Conceptualization, S.J. and J.P.; resources, S.J. and J.P.; writing original draft preparation, S.J. and J.P.; writing review and editing, S.J., J.P., N.A.R. and N.M.L.T.; supervision, J.P., N.A.R. and N.M.L.T.; project administration, J.P.; funding acquisition, J.P. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

Not applicable.

Acknowledgments

The authors would like to thank Ministry of Higher Education (MoHE), Malaysia for providing research grant with the code of FRGS/1/2022/TK08/UNITEN/01/3 to support this research and Universiti Tenaga Nasional (UNITEN), Malaysia.

Conflicts of Interest

The authors declare that they have no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Tables

Figure 1. Grid-connected nanogrid (NG) with energy management system (EMS).

Figure 2. The PV array power curve for one day.

Figure 3. DC/DC boost converter diagram.

Figure 4. Frequency division of the power energy storage system using a low-pass filter.

Figure 5. The dual active bridge (DAB) converter.

Figure 6. The schematic diagram of the LCL filter with a DC/AC inverter.

Figure 7. The EMS flowchart.

Figure 8. The EMS algorithm based on Stateflow.

Figure 9. PV and grid power curve in Scenario 1.

Figure 10. PV, batteries, and grid power curve in Scenario 2.

Figure 11. The batteries’ charge and discharge curve in Scenario 2.

Figure 12. PV, batteries, SC, and grid power curve in Scenario 3.

Figure 13. The energy storage system charge and discharge curve in Scenario 3.

Figure 14. PV, batteries, SC, and grid power curve during COVID-19 in Scenario 4.

Figure 15. The energy storage system charge and discharge curve in Scenario 4.

Figure 16. (a) Grid power import curve for Scenario 1. (b) Grid power import curve for Scenario 2. (c) Grid power import curve for Scenario 3.

Figure 16. (a) Grid power import curve for Scenario 1. (b) Grid power import curve for Scenario 2. (c) Grid power import curve for Scenario 3.

Figure 17. (a) Grid energy import curve for Scenario 1. (b) Grid energy import curve for Scenario 2. (c) Grid energy import curve for Scenario 3.

Figure 17. (a) Grid energy import curve for Scenario 1. (b) Grid energy import curve for Scenario 2. (c) Grid energy import curve for Scenario 3.

Figure 18. Energy calculation block.

Table 1

The boost converter parameters.

Parameters	Values
PV array voltage (Vpv)	513 V
Output voltage (Vout)	800 V
Maximum PV array power (Ppv, MAX)	16.7 kW
Current ripple (ΔI)	3.35 A
Voltage ripple of PV (Δvpv)	5.13 (10% of Vpv)
Converter switching frequency (fsw_boost)	25 kHz
Inverter switching frequency (fsw_inverter)	10 kHz
Inductance (L)	1.53 × 103 H
Input capacitor (Cin)	100 × 106 F
Output capacitor (Cout)	1 × 103 F
Output voltage ripple (Δv_out)	8 V (10% of Vout)

Table 2

The energy storage system parameters.

Parameters	Values
Operating voltage range for battery	42 V–54 V
Nominal voltage	48 V
Rated capacity	100 Ah
Initial SOC	40%
SOC_max	80%
SOC_min	20%
Rated capacitance of SC	177 F
Rated voltage	51 V
Rated power	5 kW
Number of series capacitors	18
Number of parallel capacitors	1
Initial voltage	50 V
Vsc_max	51 V
Vsc_min	38 V

Table 3

The dual active bridge parameters.

Parameters	Values
Input capacitance (Cin)	2000 × 106 F
Output capacitance (Co)	2000 × 106 F
Leakage inductance (L)	6 × 106 H
Input voltage ripple (Δvi)	5 V
Output voltage ripple (Δvo)	1.5 V
Maximum power (pMAX)	4.8 kW
Switching frequency (Fs)	25 kHz
Input voltage (Vin)	48 V
Output voltage (Vo)	800 V
Maximum duty cycle (dMAX)	0.35
Turn ratio of the transformer (n)	5

Table 4

The parameters of the LCL filter and inverter.

Parameters	Values of PV Inverter and LCL Filter	Values of ESS Inverter and LCL Filter
Rated power (Pn)	16.7 kW	20 kW
$\mathrm{DC} \mathrm{voltage} (V_{DC}$)	800 V	800 V
Grid voltage (Vg)	400 V	400 V
Grid frequency (Fg)	50 HZ	50 HZ
Switching frequency (Fsw)	10 kHZ	10 kHZ
$\mathrm{Base} \mathrm{impedance} (Z_b)$	4.79	4
$\mathrm{Base} \mathrm{capacitor} (C_b)$	$664.5 \mathsf{\mu }\mathrm{F}$	$795.77 \mathsf{\mu }\mathrm{F}$
Filter capacitor (C)	$32.22 \mathsf{\mu }\mathrm{F}$	$40 \mathsf{\mu }\mathrm{F}$
Grid maximum current (Imaxg)	48.4 A	58 A
Current ripple $(\Delta I$)	9% of Imaxg	9% of Imaxg
Inverter-side inductor (L1)	$3 \mathrm{mH}$	$2.55 \mathrm{mH}$
Grid-side inductor (L2)	$0.046 \mathrm{mH}$	$0.038 \mathrm{mH}$
Ratio (r)	0.0153	0.0152
Ripple attenuation (K)	20%	20%

Table 5

The carbon emission savings by using the ESS.

ESS	Emission Saving per Day (kg)	Emission Saving per Month (kg)	Emission Saving per Year (kg)
Batteries only	7.623	228.69	2744.28
Batteries with supercapacitor	8.1585	244.755	2937.06

Appendix A

The following MATLAB function was created to calculate the real-time energy in the simulation:

Figure 18. Energy calculation block.

References

1. Jamal, S.; Tan, N.M.L.; Pasupuleti, J. A review of energy management and power management systems for microgrid and nanogrid applications. Sustainability; 2021; 13, 10331. [DOI: https://dx.doi.org/10.3390/su131810331]

2. Zakaria, A.; Ismail, F.B.; Lipu, M.S.H.; Hannan, M.A. Uncertainty models for stochastic optimization in renewable energy applications. Renew. Energy; 2020; 145, pp. 1543-1571. [DOI: https://dx.doi.org/10.1016/j.renene.2019.07.081]

3. Young, B.; Ertugrul, N.; Chew, H.G. Overview of optimal energy management for nanogrids (end-users with renewables and storage). Proceedings of the 2016 Australasian Universities Power Engineering Conference (AUPEC); Brisbane, QLD, Australia, 25–28 September 2016; pp. 25-28.

4. Hussain, A.; Bui, V.H.; Kim, H.M. A Resilient and Privacy-Preserving Energy Management Strategy for Networked Mi-crogrids. IEEE Trans. Smart Grid; 2018; 9, pp. 2127-2139. [DOI: https://dx.doi.org/10.1109/TSG.2016.2607422]

5. Mbungu, N.T.; Naidoo, R.; Bansal, R.C.; Bipath, M. Optimisation of grid connected hybrid photovoltaic-wind-battery system using model predictive control design. IET Renew. Power Gener.; 2017; 11, pp. 1760-1768. [DOI: https://dx.doi.org/10.1049/iet-rpg.2017.0381]

6. Ding, Y.; Wang, Z.; Liu, S.; Wang, X. Energy Management Strategy of PV Grid-Connected Household Nano-Grid System. Proceedings of the 2019 IEEE Power & Energy Society General Meeting (PESGM); Atlanta, GA, USA, 4–8 August 2019; pp. 3-7.

7. Bakht, M.P.; Salam, Z.; Bhatti, A.R.; Anjum, W.; Khalid, S.A.; Khan, N. Stateflow-Based Energy Management Strategy for Hybrid Energy System to Mitigate Load Shedding. Appl. Sci.; 2021; 11, 4601. [DOI: https://dx.doi.org/10.3390/app11104601]

8. Kermadi, M.; Salam, Z.; Berkouk, E.M. Rule-based Power Management Controller using Stateflow for Grid-Connected PV-Battery Energy System supplying Household load. Proceedings of the 2018 9th IEEE International Symposium on Power Electronics for Distributed Generation Systems (PEDG); Charlotte, NC, USA, 25–28 June 2018; pp. 1-6.

9. Kumar, S.; Agrawal, A.; Gupta, R. Power Balance for WTG—Solar PV Fed DC Microgrids with Battery and Supercapacitor Support. Proceedings of the 2018 IEEE International Conference on Power Electronics, Drives and Energy Systems (PEDES); Chennai, India, 18–21 December 2018; pp. 1-6.

10. Manandhar, U.; Ukil, A.; Gooi, H.B.; Tummuru, N.R.; Kollimalla, S.K.; Wang, B.; Chaudhari, K. Energy management and control for grid connected hybrid energy storage system under different operating modes. IEEE Trans. Smart Grid; 2019; 10, pp. 1626-1636. [DOI: https://dx.doi.org/10.1109/TSG.2017.2773643]

11. Habeeb, S.A.; Tostado-Véliz, M.; Hasanien, H.M.; Turky, R.A.; Meteab, W.K.; Jurado, F. Dc nanogrids for integration of demand response and electric vehicle charging infrastructures: Appraisal, optimal scheduling and analysis. Electronics; 2021; 10, 20. [DOI: https://dx.doi.org/10.3390/electronics10202484]

12. Sigalo, M.B.; Pillai, A.C.; Das, S.; Abusara, M. An energy management system for the control of battery storage in a grid-connected microgrid using mixed integer linear programming. Energies; 2021; 14, 19. [DOI: https://dx.doi.org/10.3390/en14196212]

13. Roslan, M.F.; Hannan, M.A.; Jern Ker, P.; Begum, R.A.; Indra Mahlia, T.M.; Dong, Z.Y. Scheduling controller for microgrids energy management system using optimization algorithm in achieving cost saving and emission reduction. Appl. Energy; 2021; 292, 116883. [DOI: https://dx.doi.org/10.1016/j.apenergy.2021.116883]

14. Mosa, M.A.; Ali, A.A. Energy management system of low voltage dc microgrid using mixed-integer nonlinear programing and a global optimization technique. Electr. Power Syst. Res.; 2021; 192, 106971. [DOI: https://dx.doi.org/10.1016/j.epsr.2020.106971]

15. Kamran, M.; Mudassar, M.; Fazal, M.R.; Asghar, M.U.; Bilal, M.; Asghar, R. Implementation of improved Perturb & Observe MPPT technique with confined search space for standalone photovoltaic system. J. King Saud Univ.-Eng. Sci.; 2020; 32, pp. 432-441.

16. Pradhan, A.; Panda, B. A simplified design and modeling of boost converter for photovoltaic sytem. Int. J. Electr. Comput. Eng.; 2018; 8, pp. 141-149. [DOI: https://dx.doi.org/10.11591/ijece.v8i1.pp141-149]

17. Di Piazza, M.C.; Luna, M.; Pucci, M.; La Tona, G.; Accetta, A. Electrical Storage Integration into a DC Nanogrid Testbed for Smart Home Applications. Proceedings of the 2018 IEEE International Conference on Environment and Electrical Engineering and 2018 IEEE Industrial and Commercial Power Systems Europe (EEEIC/I&CPS Europe); Palermo, Italy, 12–15 June 2018; pp. 1-5.

18. Mandic, G.; Nasiri, A. Modeling and simulation of a wind turbine system with ultracapacitors for short-term power smoothing. Proceedings of the 2010 IEEE International Symposium on Industrial Electronics; Bari, Italy, 4–7 July 2010; pp. 2431-2436.

19. Lu, D.; Zhou, T.; Fakham, H.; Francois, B. Design of a power management system for an active PV station including various storage technologies. Proceedings of the 2008 13th International Power Electronics and Motion Control Conference; Poznan, Poland, 1–3 September 2008; pp. 2142-2149.

20. Qin, H.; Kimball, J.W. Generalized average modeling of dual active bridge DC-DC converter. IEEE Trans. Power Electron.; 2012; 27, pp. 2078-2084.

21. Shao, S.; Chen, L.; Shan, Z.; Gao, F.; Chen, H.; Sha, D.; Dragičević, T. Modeling and Advanced Control of Dual-Active-Bridge DC–DC Converters: A Review. IEEE Trans. Power Electron.; 2021; 37, pp. 1524-1547. [DOI: https://dx.doi.org/10.1109/TPEL.2021.3108157]

22. Ríos, S.J.; Pagano, D.J.; Lucas, K.E. Bidirectional power sharing for dc microgrid enabled by dual active bridge dc-dc converter. Energies; 2021; 14, 2. [DOI: https://dx.doi.org/10.3390/en14020404]

23. Henao-Bravo, E.E.; Ramos-Paja, C.A.; Saavedra-Montes, A.J.; González-Montoya, D.; Sierra-Pérez, J. Design method of dual active bridge converters for photovoltaic systems with high voltage gain. Energies; 2020; 13, 1711. [DOI: https://dx.doi.org/10.3390/en13071711]

24. Rodríguez, A.; Vázquez, A.; Lamar, D.G.; Hernando, M.M.; Sebastián, J. Different purpose design strategies and techniques to improve the performance of a dual active bridge with phase-shift control. IEEE Trans. Power Electron.; 2015; 30, pp. 790-804. [DOI: https://dx.doi.org/10.1109/TPEL.2014.2309853]

25. Mostafa, S.; Zekry, A.; Youssef, A.; Anis, W.R. Raspberry Pi Design and Hardware Implementation of Fuzzy-PI Controller for Three-Phase Grid-Connected Inverter. Energies; 2022; 15, 843. [DOI: https://dx.doi.org/10.3390/en15030843]

26. Liserre, M.; Blaabjerg, F.; Hansen, S. Design and control of an LCL-filter-based three-phase active rectifier. IEEE Trans.; 2005; 41, pp. 1281-1291. [DOI: https://dx.doi.org/10.1109/TIA.2005.853373]

27. Mahlooji, M.H.; Mohammadi, H.R.; Rahimi, M. A review on modeling and control of grid-connected photovoltaic inverters with LCL filter. Renew. Sustain. Energy Rev.; 2017; 81, pp. 563-5782. [DOI: https://dx.doi.org/10.1016/j.rser.2017.08.002]

28. Cabrane, Z.; Ouassaid, M.; Maaroufi, M. Battery and supercapacitor for photovoltaic energy storage: A fuzzy logic management. IET Renew. Power Gener.; 2017; 11, pp. 1157-1165. [DOI: https://dx.doi.org/10.1049/iet-rpg.2016.0455]