In this study, an EMS is presented that assists building energy operators in achieving a logical and automated operation of DC systems by utilizing real-time cost and consumption information. This algorithm is provided for EMS-owned reserves, which include buildings such as the EV parking lot and incur expenses for the EMS to purchase, repair, and other related matters. Battery technology has been implemented for storage. The EV method employed in this study represents a sophisticated approach to energy management within the realm of smart buildings. Central to this methodology is the integration of electric vehicles as pivotal components in the optimization of energy consumption. EVs act as dynamic entities capable of not only drawing power from the grid but also contributing to the grid through strategic discharge during peak demand periods. Unlike conventional methods, our EV method goes beyond mere load balancing and incorporates a multi-objective model that considers not only economic factors but also risk considerations. This entails a robust framework where electric vehicles are not just passive consumers but active participants in the overall energy ecosystem.

The uniqueness of our EV method lies in its ability to adapt to real-world constraints, including production limits, flexible loads, and intricate cost-risk indicators. By leveraging a developed whale optimization algorithm, we ensure a more efficient and adaptive energy management system. This algorithm has been specifically enhanced to avoid converging to local optima, offering a higher likelihood of finding globally optimal solutions. Moreover, our EV method stands out by considering practical and nonlinear constraints, making it more aligned with the complexities of modern energy scenarios.

Before proceeding via this method, the required information is gathered. The load value, distribution network conditions, battery state of charge (SoC_Batt), and EV state of charge (SoC_EV) are included in these data. The primary distribution network is categorized into three modes of load, namely, low load, medium load, and high load, depending on its load level. Energy costs vary for each load level. If the combined power demand for charging the EV and the amount of load exceed the generated power, this approach utilizes either the energy stored in the battery or purchases power from the distribution network. It is possible that the battery’s stored power is inadequate to offset the fraction. In this instance, the necessary energy is procured concurrently from the distribution network and the battery. At the times that the SoC has been in the order SoC_min < SoC < SoC_max, the power of the work should be used. SoC_min and SoC_max are the minimum and maximum values, which specify the allowed amount of the storage level.

Constant connectivity exists between the building distribution system and the primary distribution network. The EMS is obligated to reduce the power it purchases from the distribution network and, if it possesses power generation sources such as renewable energy sources, sell the excess power to the national distribution network during periods of high electricity prices (typically during peak hours). Therefore, the utilization of an EV is employed in this particular scenario. The EV will be operational to the extent that the EMS will be able to discontinue charging or commence discharging the EV at specific times. If EV charging is halted, the EV will enter the standby mode; otherwise, it will enter the discharge mode.

To encourage EV owners to participate in energy management and utilize EVs as energy storage devices, the EMS operator must provide incentives. The charging and discharging model for EVs presented is predicated on the subsequent principles.

• Engagement in energy management activities is not required. The proprietor of an EV is free to sign a participation contract in energy management and configure SOC_min and SOC_discharging to their liking at any time. However, if they wish to cancel the contract prior to its expiration, they will incur a fine.

In accordance with the prescribed regulations governing EV involvement in energy management, the charging expense for every EV is computed using Eqs 1, 2, which can be utilized to compute the energy cost of all EVs when multiple EVs are engaged in energy management (Ghosh et al., 2022). E C i , E V = C E V − C ⋅ ∑ j = 1 k   P i , j , E V − C − C E V − D ⋅ ∑ j = 1 l   P i , j , E V − D , (1) E C total   − E V = C E V − C ⋅ ∑ i = 1 n   ∑ j = 1 k   P i , j , E V − C − C E V − D ⋅ ∑ i = 1 n   ∑ j = 1 l   P i , j , E V − D . (2)

Here, EC _{total_EV} represents the total cost of charging all EVs, while EC _i,EV represents the energy cost for charging the ith EV. The costs associated with charging and discharging 1 kW of power are denoted by C _{EV_C} and C _{EV_D} , respectively. Additionally, P _{i,j,EV_C} and P _{i,j,EV_D} represent the EV’s charge and discharge powers, respectively. If the EV is parked in the lot and is connected to the system, it will adhere to the EMS-provided charging and discharging procedure. The EV owner is responsible for the cost of all power consumed by the EV battery ( C E V − C ⋅ ∑ j = 1 k   P i , j , E V − C ). However, it is compensated for every unit of energy discharged by the EV ( C E V − D ⋅ ∑ j = 1 l   P i , j , E V − D ).

The EMS operator must determine the C _{EV_D} value. This task necessitates the consideration of numerous variables, including charging losses, EV battery life, and the amount of money saved by reducing the peak load (EV battery efficiency). If the C _{EV_D} value is accurately determined, field EVs will be able to manage and benefit from energy support.

The price of electricity influences the performance of EVs, and the price of electricity can impact the degree to which EV owners participate in energy management. Given the fluctuating electricity prices between peak and off-peak hours, the conduct of EV owners will likewise vary during these periods. The charging and discharging patterns of EVs, in relation to the load, are presented in Table 1.

The system is divided into three load modes, as shown in Table 1: high load, normal or medium load, and low load. Charge and discharge patterns of EVs vary between the three scenarios. When power levels are high or low, an EV is only charged if its SOC is below SOC_min. If the SOC value for any EV during charging falls within the range of SOC_min and SOC_discharging, the vehicle enters the standby mode, and neither charging nor discharging occurs. Moreover, the discharge of EVs is exclusively permitted during productive periods, provided that their SOC exceeds that of the SOC_discharging.

In Figure 1, the presented algorithm is shown for three modes: low load, medium load, and high load. The SOC_Smin and SOC_Smax parameters denote the minimum and maximum permissible values for the battery’s storage level, respectively. P_UG represents power exchangeable units. P_S and P_EV represent the power and storage capacity of an EV, respectively; positive values signify a charging state.

During the low-load mode, the EV’s SOC value is evaluated in order to ascertain the necessary course of action. The P_e value represents the EV’s charging rate. If the SOC value of the EV exceeds SOC_EVmin during the medium load stage, the system will enter the standby mode. If not, the EV enters the charging mode; this cycle will continue until its state of charge reaches SOC_EVmin. In this scenario, should the network encounter a power deficit, EVs with the highest SOC will enter the discharge mode (SOC_{EVdischarging}). Once the SOC of a vehicle reaches SOC_discharging, it will transition to the standby mode.

To conduct an economic analysis of the proposed energy management algorithm, it is imperative to derive a correlation that facilitates the computation of energy expenses. The determination of energy supply costs can be approached from three perspectives: perspective of the building manager, who may serve as the load collector or energy management operator; perspective of the consumers residing in the building; or from the standpoint of the distribution network. This study operates under the assumption that the building manager is responsible for EMS control and is tasked with reducing the overall cost of energy supply for the structure. Energy costs can be calculated using Eq. 3. E C total   = E C U G + E C total   _ E V . (3)

The total cost of energy supply for the building is denoted by EC _total . Additionally, E _CUG represents the expense incurred to procure electricity from the distribution network, while EC _{total_EV} represents the cost incurred to charge and discharge all EVs. The cost of purchasing energy from the distribution network is calculable with the aid of Eq. 4. E C U G = C U G ⋅ P tota l load   − ∑ i = 1 n   P i , E V − D . (4)

Here, P _{total_load} is the total load of the building and calculated using Eq. 5. P total _ load   = ∑ P D _ l o a d   + ∑ i = 1 n   P i , E V − D . (5)

C _UG represents the electricity purchased from the grid that costs one hundred dollars per kilowatt-hour. P _{D_load} represents the building load demand expressed in kilowatt-hours. P _{i,EV_C} and P _{i,EV_D} represent the quantity of charged power and discharged power, respectively, in ith EV in kW hours.

Assumed within the confines of this scenario are several critical parameters. First and foremost, the heating system is presumed to consistently operate at peak efficiency, ensuring optimal energy utilization. Simultaneously, the EV is envisioned to be charged to its maximum capacity upon connection to the charger, and all responsive devices are activated within their designated permitted operating range. Despite these proactive measures, due to the inherent uncontrollability of the load, the integration of EVs into the household load is presently deemed unfeasible. Figure 2 provides a comprehensive illustration of the building load demand diagram for this residential scenario, encapsulating the intricate dynamics of the load under the assumed conditions. The basis for the ToU pricing model is established at 268.87 cents, reflecting the electricity consumption throughout the day, as depicted by the load demand in Figure 2. Furthermore, the peak load consumption during this period is recorded at 6.80 kW.

On the same day, it is postulated that the house’s electricity consumption occurs within the timeframe spanning the 78th to the 85th time steps. Consequently, electrical service is anticipated to be intentionally disconnected from the residence at incremental intervals corresponding to these time steps. Notably, during these cut-off periods, both the television and personal laptop experience forced shutdowns, enduring a blackout period of 105 min. Additionally, specific measures are implemented to curtail energy consumption during this scenario.

In this particular scenario, the assumption is made that the household load remains uninterrupted, thereby framing load management as an exercise focused on minimizing co-payments throughout the day. Table 4 provides a comprehensive breakdown of performance intervals for responsive appliances, suggested performance intervals for non-responsive appliances, and the corresponding consumption levels of the heating system within each interval. These values are derived from solving the optimization problem, considering both scenarios with and without accounting for the capability of EVs to supply household loads.

An analysis of the outcomes reveals that device functionality adheres to the permitted time intervals during periods of reduced tariffs, showcasing the effectiveness of load management in minimizing costs. Table 5 elucidates the planning involved in EV charging and discharging during different time steps. For instance, during time steps 1–12, if the EV cannot supply the load, it is charged to ensure a fully charged battery before departing the residence. At time step 65, when the EV enters the residence, the battery is 2.7 kWh full, prompting a recharge at the end of each day to attain a pre-established battery charge level of 3.9 kWh.

Considering the EV’s capacity to supply the household load, the battery is strategically discharged during high tariff periods and charged during low tariff periods (as detailed in Table 5), resulting in an overall reduction in payment amounts. A comparative analysis between the two cases, one accounting for the ability of the EV to provide load and the other without, reveals a shared payment cost of 227.26 cents per day in the former and 238.16 cents per day in the latter.

Furthermore, Figure 3 visually depicts the household load dynamics both with and without the capability of energy supply via the EV. Notably, when EVs possess the ability to supply the household load, they are discharged during high tariff hours and charged more during low tariff hours, leading to an increase in the peak load at the end of the night, an effect absent in cases where the load supply capability is lacking. This nuanced exploration of load management strategies provides a comprehensive understanding of the interplay between EV capabilities and household load dynamics.

In this scenario, the household load experiences a planned interruption from time step 77 for a duration of 90 min, spanning 6 time steps, until time step 84. Prior to this interruption, up to time step 76, the household load management results mirrored those observed in the preceding scenario. At time step 77, the schedule undergoes an update to accommodate the planned interruption, and following the cessation of the load at time step 84, the household load is re-evaluated based on the most recent information regarding device operations. Noteworthy in this context is the consideration of a lost energy value for devices at 22 cents/kWh, surpassing the electricity tariff at any given time. Consequently, in the second mode, encompassing time steps 77 to 84, the EV strives to supply maximum electricity to devices that remain operational during this interval.

Given a switching time of 30 s, the cost associated with load interruption is calculated at 1.8 cents. This deviation in outage cost, as opposed to the cost calculated in the initial scenario, exemplifies the impact of the proposed household load management methodology on the reduction of outage-related expenses. Figure 4 visually represents the household load diagram in this third state, offering insights into the load dynamics during and after the interruption.

Based on the outcomes, the expected household peak load in this scenario is determined to be 6.29 kW, which is less than the peak load observed in the initial case. Moreover, the payment fee registers at 246.11 cents, slightly higher than that in the second case. This outcome underscores that, despite the occurrence of load interruptions and deviations from the optimal plan in terms of paid costs observed in the second case, the household load management strategy in the third case effectively prevented an excessive increase in overall payment costs. This nuanced analysis highlights the resilience and effectiveness of the proposed methodology under challenging conditions, contributing to a comprehensive understanding of its practical implications.

In this section, the effects of IBR pricing, as well as the combination of IBR and ToU, which have been considered in previous references, on the results of household load management, are attempted to be studied. Table 6 displays the results of the paid fee and peak load when comparing different tariffs.

The results show that when the IBR is used, the load is distributed throughout the day, and, as can be seen, the peak load is significantly reduced. As a result, the payment fee and peak load will be reduced by 7.9% and 16.4%, respectively, when the ToU tariff is considered.

To summarize the findings, the first scenario is considered in the ToU to be a state without household load management due to the existence of power outages and the occurrence of outages for various devices, as well as the uncontrollability of the load and the impossibility of supplying the household load by the EV. Regardless of the pricing method used, it results in the imposition of high costs on the subscriber. In the second scenario, unlike the first, it is not certain, owing to the result of load management planning by defining performance intervals for all devices, as well as checking the conditions of ability and inability to supply the household load by the EV, which shows that the supply ability is optimal. In this scenario, the load is supplied by an EV.

In comparison to scenario 1, the result is that the co-payment cost with the ToU tariff has been reduced even if the household load cannot be supplied by the EV. Domestic peak load has also declined since the first state. Outage is considered in the third scenario, but unlike the first, there is optimal planning for load management with the presence of the EV, which allows for a reduction in the outage cost when compared to the first. Furthermore, when compared to the first scenario, the domestic peak load and the cost of paying with the ToU tariff are lower in this case. However, despite the presence of load interruption in this case, the payment fee is only marginally higher than in the second scenario, owing to the effect of the proposed household load management. Finally, the goal is to compare different pricing on household load management planning, demonstrating that the combination of IBR and ToU results in a reduction in the payment cost and peak load when compared to the ToU pricing method.

To validate the outcomes of various tariffs, a qualitative comparison is conducted in this section. The multiple users and load priority (MULP) scenario is utilized by Abushnaf et al. (2015) to compare the outcomes of the multi-user algorithm and ToU. In the initial instance of this scenario, a comparison is made between the flat rate and ToU tariffs (without energy consumption control). The alterations in Figure 5A constitute a summer day, while the modifications in Figure 5B represent a winter day. The graphs additionally illustrate the load for both flat rate and ToU tariffs, along with the curves representing the costs incurred under each tariff option (flat rate and ToU).

As shown in Figure 5, the cost of payment is reduced by 0.83% when the ToU tariff is utilized during the summer, as opposed to the flat rate tariff. Conversely, when compared to the ToU tariff, the flat rate tariff will result in an 8% reduction in payment costs during the winter season. Conversely, energy consumption remains constant throughout the day and night, with no measures implemented to transfer superfluous loads to more cost-effective ToU periods. Nevertheless, during peak hours, the cost of electricity is considerably greater under this tariff compared to the flat rate tariff. This necessitates the promotion of a method for regulating peak-hour electricity consumption.

In this particular situation, the algorithm for home energy management transfers superfluous loads from periods of high demand to periods of lower cost with the intention of diminishing electricity expenses and overall consumption. In order to accomplish this, it is assumed that the demand limit of 1.7 kW during peak hours is constant for both seasons. The home energy management algorithm’s adaptable architecture enables subscribers to exercise diverse discretion in order to mitigate their daily energy usage and subsequently lower their electricity expenses.