Energy Reports 12 (2024) 3218–3242

Contents lists available at ScienceDirect

Energy Reports
journal homepage: www.elsevier.com/locate/egyr

Research paper

Empowering E-mobility: Day-ahead dynamic time of use tariff for electric

vehicle charging
Tafsir Ahmed Khan , Syed Abdullah-Al-Nahid , Silvia Tasnim , Md. Abu Taseen , Taskin Jamal *,
Tareq Aziz
Department of Electrical and Electronic Engineering, Ahsanullah University of Science and Technology, Dhaka, Bangladesh

A R T I C L E I N F O A B S T R A C T

Keywords: The greenhouse gases and preservation of the environment have been the most prioritized concern in many
Dynamic pricing countries and international organizations, leading to set aims and initiatives for lowering the dependency on
Time of use (ToU) fossil fuels and minimizing carbon emissions. As a part of many initiatives, great emphasis is given to promoting
Electric vehicle
electric vehicles (EVs) over their traditional carbon fuel-based counterpart. To support large-scale EV adoption, a
Particle Swarm Optimization
dynamic tariff scheme is required that addresses complex issues like user comfort, savings, and utility revenue. In
this research work, a dynamic pricing scheme is modeled for the centralized EV charging stations (EVCS) on a
residential feeder. This pricing scheme integrates the existing static time of use (ToU) pricing to an hourly dy-
namic ToU tariff that varies according to the demand level in the residential feeder. The proposed tariff plan
provides a day-ahead dynamic structure with a probable price variation range to aid the EV owners and the day-
ahead charging mechanism to schedule the EVs beforehand. The developed scheme calculates the returns and
adjusts the tariff including the demand charge and the recovery component on a daily, monthly, and yearly basis
for the unsatisfactory investment revenue. Also, user convenience is ensured by setting the minimum possible
tariff as well as satisfying the minimum revenue target with the integration of particle swarm optimization (PSO)
in designing the pricing structure. The simulation results demonstrate the effectiveness of the proposed dynamic
pricing scheme maintaining a trade-off between the interests of all parties involved. Moreover, the simulation
outcome of the test case indicates that EV owners can save more than 32 % on charging bills if they plan
correctly. On the other hand, the utility can earn more than 2.4 times the existing rate during the peak time of the
residential feeder.

2023a). The situation will get out of our hands sooner than we antici-
1. Introduction pate, as the number of private passenger vehicles will be close to 1.6
times by the year 2050 (International Energy Agency, 2023a). On the
1.1. Background study bright side, many countries and governments have already set their
target to achieve net-zero emissions by this time (Handayani et al., 2023;
A vital step in the transition to a sustainable environment will be International Energy Agency, 2023b). An effective way to achieve this
reducing our dependency on fossil fuels as much as possible (Li et al., net-zero emission would be the electrification of the transportation
2022). Fossil fuels are environmentally destructive and proved to be the sector (Abdullah-Al-Nahid et al., 2023; Lee, 2023; Achariyaviriya et al.,
largest contributor to the emission of greenhouse gases. To reduce the 2023). In this way, the existing vehicles that consume fossil fuel will be
emission of greenhouse gases from the usage of fossil fuels, the major replaced by Electric Vehicles (EVs), which use electrical energy from
users of these fuels should look into cleaner alternatives. For instance, cleaner resources (Zhou et al., 2020). In the past decade, EVs have
the current transportation sector, a major fossil fuel user, can go electric. become the center of research and development as the penetration of
The existing transportation sector accounts for more than one-third of this mode of transportation has significantly increased due to contin-
global carbon dioxide (CO2) emissions (International Energy Agency, uous promotion from automobile manufacturers and governments
2023a). More than 40 % of this carbon emission from the transportation (Dong et al., 2018). To increase the effectiveness of EVs and make them a
sector comes from passenger vehicles (International Energy Agency, viable replacement for fossil fuel-based vehicles, the charging

* Corresponding author.
E-mail address: [email protected] (T. Jamal).

https://doi.org/10.1016/j.egyr.2024.08.086

Available online 12 September 2024

2352-4847/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-
nc/4.0/).
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Nomenclature TElectricity(i,n) dynamic electricity pricing profile formed with

different “n”
EV electric vehicle ND-M number of days in different months of different years
EVCS electric vehicle charging station ND-Y number of days in different years
DA day-ahead DCkW per kW demand charge
RT real time DCRMonthly monthly demand charge that needs to be recovered
ToU time of use DCRDaily daily demand charge that needs to be recovered
DToU dynamic time of use DCDDaily deficit in demand charge recovery
PSO particle swarm optimization DCRDaily_adj daily demand charge after deficit adjustment
CO2 carbon dioxide DCARDaily daily demand charge actually recovered by the station
kW Kilowatt TDC_max maximum per kWh demand charge
kWh Kilowatt-hour TDemand dynamic demand charge per kWh
kVA Kilovolt-ampere pd previous day
PEV plug-in electric vehicle pm previous month
ICE internal combustion engine py previous year
RTP real time pricing EkWh-DA day-ahead amount of total energy consumption by EVs
CPP critical peak pricing ypayback total number of years in the intended payback period
CPPT pre-determined amount for CPP IRate varied inflation rates (projected)
PTR peak time rebate MRYearly the amount of money to be recovered in different year
DCFC DC fast charging MRDaily amount of money to be recovered in a day of different years
V2G vehicle-to-grid MRMonthly amount of money to be recovered in a month of different
V2V vehicle-to-vehicle years
NSP network service provider n distribution ratio
ND_RTP next day RTP component Umin minimum station utilization factor
LT low tension i, d, m, y ith time slot of dth day of mth month of yth year
MT medium tension TBLPB minimum threshold for the investment recovery
HT high tension component or the baseline payback tariff per kWh.
CLifetime lifetime cost TFPB final payback tariff /investment recovery component per
CAcquisition one time acquisition cost kWh
CO&M yearly operation & maintenance cost AdjDaily daily adjustment per kWh
CLandRent yearly land-related cost AdjMonthly monthly adjustment per kWh
AD maximum daily adjustment threshold AdjYearly yearly adjustment per kWh
AM maximum monthly adjustment threshold DDaily total daily deficit
AY maximum yearly adjustment threshold DMonthly total monthly deficit
x maximum deviation in forecasted day-ahead residential DYearly total yearly deficit
demand ADYearly aggregated yearly deficit amount
PAvailable available power in kW EkWh actual energy consumption by EVs in the considered day
PMaxCapacity maximum supply capacity of the feeder in kW xd numerical value of the daily adjustment per kWh
PDemand demand for residential load in kW xm numerical value of the monthly adjustment per kWh
PCharger power requirement of one charger xy numerical value of the yearly adjustment per kWh
NCharger number of installed chargers TFinalDynamicPricing final dynamic pricing profile for a certain day
NAC number of active chargers that can be fitted under the PRT_Demand_up real-time demand considering x% increment in kW
curve with decimal fraction (for upper limit)
NACs number of chargers after rounding down PRT_Demand_low real-time demand considering x% reduction in kW
TIstaticToU total income accumulated in a day by static ToU tariff (for lower limit)
TstaticToU static ToU tariff imposed by regulatory body TFDP_up upper limit of the day-ahead price variation range for a
TIDToU total income that has to be achieved in a day by dynamic certain day
ToU tariff TFDP_low lower limit of the day-ahead price variation range for a
TDToU dynamic ToU pricing. certain day
C DToU coefficient TElectricity_up upper limit of the calculated electricity pricing for
PRT_Available real-time available power in kW upper demand level
PRT_Demand real-time demand for residential load in kW TElectricity_low lower limit of the calculated electricity pricing for
NRT_AC number of active chargers that can be fitted under the real- upper demand level
time available power curve with decimal fraction TBDAP base day-ahead pricing of the variation range for a certain
NRT_ACs number of chargers in real-time after rounding down day.
TElectricity final dynamic electricity pricing profile. BERC Bangladesh Energy Regulatory Commission
TITElectricity(n) total income of the day calculated with final ARIMA autoregressive integrated moving average
electricity pricing profile formed with different “n” MILP mixed-integer linear program
TIDToU total income that has to be achieved in a day by dynamic USD United States Dollar
ToU tariff

3219
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

mechanism of EVs, starting from charge scheduling to final billing, 1.2. Tariff for EV Charging
should be flawless and friendly for all parties involved. To make EV
adoption economically feasible, a proper billing system or tariff should Tariffs determine the price that the owners pay for charging their
be developed and utilized (Danial et al., 2021). electric vehicles. The business model that the stockholders employ to
A suitable pricing scheme can improve economic efficiency, alleviate invest and recover infrastructure development costs with a profit de-
transformer and feeder overloading, normalize load curves, decrease pends on tariffs. To some extent, tariffs also regulate the electric vehi-
power losses, and bring consumer satisfaction (Khan and Aziz, 2022; cles’ mode of interaction with the grid. Tariffs include both the
Amin et al., 2020). The utility recovers the costs of delivering each unit electricity price and the component of recovery for charging infra-
of electrical energy to the EV by charging every customer a certain structure development. A visual representation of the categorization of
amount of bill depending on their energy usage (Khan and Aziz, 2022). A tariff components and pricing type is given in Fig. 1 (Amin et al., 2020;
well-designed pricing scheme or tariff properly recovers all the costs Tariff design, 2023; Limmer, 2019; Ansarin et al., 2020; Fridgen et al.,
involved in energy delivery with an appropriate profit to run the utility. 2018; Aguilar-Dominguez et al., 2020).
It should also provide cost-saving opportunities to the consumers as Electricity pricing denotes how electricity consumers, in this case, EV
well. There are many types of pricing schemes or tariffs used by utilities owners, are charged for the consumed electricity. Electricity pricing
worldwide, such as flat-rate, two-part, block rate, peak demand tariffs, includes an array of utility fees that are proposed to earn revenue from
etc (Khan and Aziz, 2022; Mehta and Mehta, 2006). Most existing the consumers to ensure the cost recovery for the utility. This cost
pricing schemes are designed by static plans and are not suitable enough generally includes the cost of energy production, transmission and dis-
to utilize the current smart grid technologies and ensure the benefits of tribution losses, and other operational expenditures. The electricity
both the utilities and the consumers (Khan and Aziz, 2022; Yang et al., pricing also includes the cost of service to various types of customers, in
2018). These plans are set based on the historical behaviour of the this context, electric vehicle charging stations (EVCS). Utility often de-
network. So, in static pricing schemes, the rates do not vary with the clares few pricing components to promote energy savings and other
electricity demand. On the other hand, the pricing schemes which are network-friendly behaviour. The common components of electricity
designed based on dynamic strategies (usually called “dynamic pricing”) pricing are energy charge (currency-unit per kWh), demand charge
can address the fluctuation in real-time demand by varying the rate (currency-unit per kWh or currency-unit per kVA), connection charge
alongside it (Khan and Aziz, 2022; Dutta and Mitra, 2017a). (currency-unit per month or per year), taxes and other fees (percentage)
The pricing schemes or tariffs designed to be used in EV charging (Tariff design, 2023). These components of electricity pricing are
have some similarities as well as some disparities with the tariffs used for applicable for EVCS as it is the primary customer for utility. However,
residential loads (Tariff Rate, 2023). There can be many kinds of tariffs the EVCS recovers these fees from EV owners when they get charged. EV
to be utilized in the billing system for EV charging. All of these tariffs owners usually pay for the energy the vehicles consume. So, all the
have their own advantages and limitations. In this paper, a dynamic components the EVCS pays to the utility must be included in the per kWh
pricing scheme is introduced to be used in EV charging. The current price of energy consumed by the EVs.
practices of dynamic pricing and the factors that affect the rates are also The EVCS and utility can design and implement different types of
determined through investigation. tariff structures for electricity (energy) pricing for EV owners. Electricity

Fig. 1. Categorization regarding tariff components and pricing type [Tk = Taka, currency of Bangladesh].

3220
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

pricing schemes can be both time-variant and invariant. Time-variant The study in (Muratori et al., 2019) analyzes the electricity pricing
pricing schemes are mostly dynamic and more advanced. Some of the schemes for DC fast charging enabled public electric vehicle charging
popular electricity pricing schemes are flat-rate tariff (Ansarin et al., stations in the USA. The majority of the considered electricity pricing
2020), time of use (ToU) tariff (Amin et al., 2020; Tariff Rate, 2023; schemes have a ToU arrangement. The study has emphasized the effect
Fridgen et al., 2018; Australian consumers, 2023; Shaolun Zeng et al., of demand charges on tariffs. Three different arrangements of the de-
2008), real-time pricing (RTP) (Amin et al., 2020; Fridgen et al., 2018; mand charge: no demand charge, flat-rate demand charge, and ToU
Cao et al., 2021; Berkow et al., 2023), critical peak pricing (CPP) (Amin demand charge, are investigated. The study also shows the utilization
et al., 2020; Fridgen et al., 2018; Berkow et al., 2023; Newsham and factor of the EVCS has a great impact on the final electricity prices,
Bowker, 2010; Utility rate design initiative, 2023), peak time rebate especially for cases with tariffs that include demand charges. However,
(PTR) (Dutta and Mitra, 2017b; Bejan et al., 2021) etc. These pricing the study has not considered the demand variation in the impact anal-
schemes have their own pros and cons over others. However, a few ysis. A dynamic ToU pricing mechanism is proposed for charging
pricing schemes can exhibit the real costs of electricity generation and plugged-in EVs in parking decks (Guo et al., 2014). In this mechanism,
other services more appropriately than a simple, easy-to-implement and 2-block rates are proposed based on demand level and seasonal impact
widely used flat-rate tariff. This means those pricing schemes yield to promote smart charging and maximize the total revenue of the
better performance in terms of economic efficiency. The ToU is simple parking deck. The rates are varied with an interval of 1 hour, and the
yet very popular and much better than flat-rate and PTR in terms of timings are set for the next 24 hours. The dynamic pricing scheme also
fairness (Amin et al., 2020). However, the static nature of traditional has a provision of providing parking fee rebates if the customers charge
ToU can be a drawback for this scheme. The RTP is unstable in terms of their vehicles in the parking deck.
bill steadiness, yet it is very flexible with the highest fairness (Amin According to (Emodi et al., 2022), the pricing schemes for charging
et al., 2020). Consumers can regulate their electricity consumption and electric vehicles in Australia are mostly static ToU, except for a few
make economic decisions while charging electric vehicles. The RTP also exceptions. A little dynamicity is available in this ToU in the sense that
offers great economic efficiency to the utility as the scheme reflects the the rates become high in the afternoons and first half of the evenings of
actual cost of energy supply better than any other scheme (Dutta and weekdays while rates are really low in the middays and later part of the
Mitra, 2017a). nights. These ToU structures are not directly related to the demand level
It may seem that the RTP pricing scheme is the best solution over of the grid. However, the utilities in Australia apply different mecha-
other schemes, but because of its unpredictability, RTP may be proven nisms to manage consumer loads to address real-time spikes in demand
dangerous and unpopular among consumers because of its quickly levels. In South Korea, a seasonal ToU tariff is utilized in EV charging
changing behaviour. An alternative to the RTP pricing scheme can be a (Korea Electric Power Corporation, 2023). This ToU has three block
dynamic time of use (DToU) scheme (Khan and Aziz, 2022). In terms of rates for three different seasons. This means the rates are different for
economic efficiency and fairness, the ToU pricing scheme stays in the different times of the day of different seasons. In this tariff, the included
middle of the two extreme tariffs, flat and RTP; this can be a great option demand charges and the energy rates vary for different classes of EVCS.
for EV charging, provided a certain level of dynamicity is adopted In a study (Abdullah-Al-Nahid et al., 2022), an assumption-based ToU
(Ansarin et al., 2020). Instead of keeping the time blocks static, they can tariff is utilized for day-ahead scheduling vehicles seeking charge in a
be made shiftable based on demand level. To make it more fair, more centralized EVCS of Bangladesh. The rates are somewhat dynamic and
blocks of rates can be introduced so that the pricing scheme can imitate vary based on the available power of the feeder. The critical peak con-
the demand curve. This will take the fairness and economic efficiency of ditions are managed by making the tariff unavailable in those critical
ToU to RTP’s level. Other schemes, such as CPP, can selectively be in- peak hours, which means no EV can get a charge. However, the pre-
tegrated into this DToU scheme for their advantages. sented ToU tariff is hypothetical and the unit values are imaginary. In
addition, the real-time variation of the demand is not addressed in the
1.3. Dynamic Pricing: Current Practices and Influencing Factors study.
In some studies, vehicle battery utilization by EVCS through vehicle-
1.3.1. Current Practices in EV Charging to-grid (V2G) or vehicle-to-vehicle (V2V) technology is brought into
Dynamic pricing is thought to be a competent method to overcome consideration by price mechanisms. For managing the operation, EVCS
many technical challenges created by the high penetration of EVs. It can may need to use the batteries of the EV via bidirectional charging. For
positively react to variations in the operating conditions of the grid instance, in the study in (Ghosh and Aggarwal, 2018), a model for
system and increase the flexibility of the system (Limmer, 2019). Due to bidirectional charging of EVs is considered. A menu-based pricing plan
these reasons, the development and use of dynamic pricing schemes is shown to the incoming EVs before arriving at the EVCS. Every EV
have recently earned great attention in the field of electric vehicle owner has to select a particular plan from the menu to get a charge. Each
research. plan on the menu has a different price against a certain level of battery
The report in (Hove and Sandalow, 2023) shows the EV charging utilization, a deadline, and an approximated amount of charged energy.
policies and tariffs used in China and the USA for charging EVs of In (You et al., 2017), a optimal charging strategy is proposed and tested
different categories. The most common tariffs that are adopted by these under a dynamic electricity pricing environment for a smart EVCS with a
two countries for charging EVs are some variants of traditional ToU. bidirectional charging facility. According to the study, the EVCS buys
Some incentive programs are also incorporated with these tariff struc- and stores energy in the EVs with real-time energy prices. Then, the
tures. However, according to the report, the dynamic ToU tariff is stored energy is discharged to another EV at a different time. The pricing
adopted only in a few states of the USA as a pilot program. An hourly scheme adjusts the values based on financial cost analysis and optimi-
pricing program is introduced for Chicago, USA (Zethmayr and Kolata, zation considering battery degradation due to bidirectional charging.
2019). This dynamic pricing program is tested on level 2 and level 3 DC However, no clear ideas can be obtained from these two studies about
fast charging (DCFC)-enabled 3 different types of EVs with an intention whether the tariffs respond to the demand variation or not. Also, the
to evaluate how much a driver/ owner can save on charging cost payback component for recovering the investment and costs of EVCS is
compared to the utility’s flat-rate tariff. The study has also considered 4 not mentioned in these studies.
different types of drivers: light, average, heavy, and Uber. The results
show that the proposed dynamic pricing program can yield up to 59 % 1.3.2. Influencing Factors Affecting Dynamic Pricing
savings on energy costs if the drivers plan ahead. However, the proposed There are many factors that can affect the rates of dynamic electricity
pricing scheme has not shown any proof of addressing the real-time pricing for EV charging. Some of these factors are directly associated
demand fluctuation and preserving the utilities’ interest. with utility and demand. However, many of these factors are related to

3221
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

EVs and customers. Depending on these factors, the tariff for charging an revenue collection and increasing customers’ convenience. Demand
EV can vary. Some of these factors are forecasted demand level (Khan charge and payback components for recovering the investment and
and Aziz, 2022), real-time demand fluctuation (Dong et al., 2018; Khan other EVCS costs were also missing.
and Aziz, 2022), critical peak condition (Khan and Aziz, 2022), A well-designed dynamic pricing scheme will be a must to respond
congestion of EVs (Guo et al., 2014), demand charge (Guo et al., 2014), well to the upcoming EV penetration worldwide. In this paper, a dy-
utilization factor of the EVCS (Guo et al., 2014), flexibility in charging namic ToU-based pricing scheme is developed to charge DCFC-enabled
schedule (Limmer, 2019; Daneshzand et al., 2023), battery utilization EVs by transforming the existing static ToU rates. The scheme provides a
(Limmer, 2019; Ghosh and Aggarwal, 2018; You et al., 2017; Richard- day-ahead pricing plan to the customers and the charging stations to
son, 2013), flexibility in the amount of energy (Limmer, 2019), charging complete the day-ahead charge scheduling. This day-ahead pricing plan
duration flexibility (Limmer, 2019), flexibility in charging location incorporates a critical peak pricing component to deal with critical peak
allocation (Limmer, 2019), seasonal impact (Muratori et al., 2019; Guo events. The dynamic tariff responds to the real-time variation in demand
et al., 2014; Korea Electric Power Corporation, 2023), customer class by adjusting the prices in real time. The dynamic tariff also includes a
(Zethmayr and Kolata, 2019; Meng et al., 2023) etc. demand charge component to recover the monthly demand charge and a
payback component that recovers the investment and all costs of EVCS
1.4. Knowledge Gap and Problem Statement infrastructure development. The prices of this dynamic tariff structure
should be optimized through a good metaheuristic algorithm to keep a
Conventional pricing schemes, like flat-rate and fixed time-of-use trade-off between the customers’ satisfaction and the utility’s interest.
(ToU) tariffs, often fall short of optimizing economic efficiency and This research uses a version of the particle swarm optimization (PSO)
fairness. A thorough examination is necessary to enhance the pricing algorithm to solve optimization issues.
structure, aiming to create a dynamic tariff system that accurately re-
flects energy generation costs by adjusting rates in response to real-time 1.5. Contributions and Novelty
demand. This revised tariff framework should consistently prioritize
user convenience while also aligning with the economic interests of A comprehensive dynamic pricing scheme for charging electric ve-
Electric Vehicle Charging Stations (EVCS) and investors. hicles is developed in this research by restructuring an existing static
Despite the significant contributions, the studies mentioned above ToU tariff for the centralized charging stations, which will operate under
and a few others have some shortcomings, which can be summarized as a coordinated charging scheme. The core novelty and contributions of
follows: this research are given as follows:

• The studies in (Hove and Sandalow, 2023; Zethmayr and Kolata, • Formation of a dynamic pricing scheme by remodelling the existing
2019; Muratori et al., 2019; Emodi et al., 2022; Ghosh and Aggarwal, static ToU tariff for charging EVs under a centralized charging
2018; You et al., 2017) have not addressed demand levels and scheme. This will help the utilities or other bodies to implement a
real-time variations to reflect the real cost of energy generation. dynamic tariff structure in a newly developed charging station for a
• The studies in (Hove and Sandalow, 2023; Zethmayr and Kolata, locality that already has a static ToU or a flat-rate tariff for EV
2019; Muratori et al., 2019; Guo et al., 2014; Emodi et al., 2022; charging.
Ghosh and Aggarwal, 2018) have not considered the critical peak • The pricing scheme addresses user convenience by keeping the tariff
reduction in the tariff. minimum whenever possible.
• The studies in (Hove and Sandalow, 2023; Zethmayr and Kolata, • It addresses the high peak energy generation cost by increasing the
2019; Muratori et al., 2019; Emodi et al., 2022; Korea Electric Power tariff and introducing a critical peak pricing component.
Corporation, 2023; Abdullah-Al-Nahid et al., 2022; Ghosh and • It addresses the real-time variation in demand by adding a real-time
Aggarwal, 2018) have not used any sort of optimization algorithm in pricing component with the day-ahead profile while billing.
designing the tariff. • It recovers the monthly demand charge by adding a dynamic demand
• The studies in (Hove and Sandalow, 2023; Zethmayr and Kolata, charge component with the tariff.
2019; Emodi et al., 2022; Abdullah-Al-Nahid et al., 2022; Ghosh and • The pricing scheme recovers the investment and other costs of EVCS
Aggarwal, 2018) have not considered including the demand charge by adding a payback component with the tariff. The payback
in the tariff structure. component is optimized to have the minimum possible price to
• The studies in (Hove and Sandalow, 2023; Muratori et al., 2019; Guo ensure user convenience.
et al., 2014; Abdullah-Al-Nahid et al., 2022; Ghosh and Aggarwal, • The pricing scheme offers a day-ahead price variation range to make
2018) have not prioritized user convenience in the tariff structure. the user experience much smoother.
• None of the studies in (Hove and Sandalow, 2023; Zethmayr and
Kolata, 2019; Muratori et al., 2019; Emodi et al., 2022; Korea Electric Table 1 contains representative literature related to existing tariff
Power Corporation, 2023; Abdullah-Al-Nahid et al., 2022; Ghosh and structures for EV charging and emphasizes the novelty of this research
Aggarwal, 2018) has done a proper economic analysis on the tariff to over existing studies. The summary also indicates the superiority of the
ensure the payback of all investments and costs of EVCS. proposed scheme in the considered areas of comparison.
• The studies in (Hove and Sandalow, 2023; Muratori et al., 2019; Guo
et al., 2014; Emodi et al., 2022; Korea Electric Power Corporation, 1.6. Organization of Paper
2023; Ghosh and Aggarwal, 2018) are not fully dynamic in terms of
the response to real-time variation of demand as the considered ToU The rest of the article contains the following sections- Section 2 de-
pricing has fewer blocks. scribes the methodology of the proposed pricing scheme, Section 3
displays the simulated outcomes along with essential analysis, and
Most importantly, none of the examined studies shows the transition Section 4 depicts the important concluding notes with a few suggestions
process of adopting a dynamic pricing scheme from an existing static for future work.
ToU or a flat-rate tariff, which most countries use in EV charging. A
preliminary investigation of this research was carried out in (Khan and 2. Methodology
Aziz, 2022), where a day-ahead dynamic pricing scheme was presented
that also varies with the real-time demand curve. However, the scheme The supervised charging schemes must adopt a dynamic tariff
in the preliminary investigation was not optimized for ensuring proper structure that varies with the power demand and can recover all the

3222
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Table 1
Key parameter comparison with some related studies.
Key parameters (Hove and (Zethmayr and (Muratori (Guo et al., (Emodi (Korea Electric Power (Abdullah-Al-Nahid Proposed
Sandalow, Kolata, 2019) et al., 2019) 2014) et al., Corporation, 2023) et al., 2022) Scheme
2023) 2022)

EVCS type L3 DCFC L2 & L3 DCFC L3 DCFC PEV All LT & HT L3 DCFC L3 DCFC
parking lot
Tariff type ToU ToU ToU ToU ToU Seasonal ToU ToU DToU
No. of tariff blocks 2–3 Indefinite 2 2 NM 3×3 Indefinite Indefinite
Dynamicity Low High High Low Low Low High High
Rate changing NM 1 hour NM 1 hour NM NM 1 hour 1 hour
interval
Addressed demand No No No Yes No NM Yes Yes
level and variations
Addressed real time No NM No No No NM No Yes
demand variations
Consideration of No No Yes No No NM No Yes
demand charge
Price optimization No Yes No Yes No NM No Yes
Optimization N/A NM N/A MILP N/A N/A N/A Modified
technique PSO
Consideration of user No Yes No No N/A NM No Yes
convenience
Critical peak No No No No No NM Yes Yes
reduction
consideration
EVCS costs recovery No No Yes Yes NM NM No Yes
Proper economic No No No Yes No NM No Yes
analysis
Provision of incentive Yes No No Yes No No No Yes
program
NM: Not mentioned; N/A: Not applicable

investment with a suitable profit. The tariff scheme should also address forecasting mechanism. Subsequently, it assembles the dynamic elec-
real-time demand fluctuations and compensate for any loss or excess tricity pricing profile or component for a specific day based on the
income. The dynamic tariff structure should benefit both the utility and collected data, taking into account any residual data from the previous
customers in some ways. It should compensate the utility for higher day if applicable. Next, the methodology constructs the dynamic de-
demand and benefit the customers when the demand is low. In a usual mand charge component, considering the daily energy requirements for
static time of use or ToU tariff scheme (or in a flat-rate tariff), the utility charging EVs.0
doesn’t get paid higher if the peak is not in the anticipated period. On Following that, the methodology formulates the dynamic investment
the other hand, customers might have to pay at the peak rate even if recovery component for the specified day, which may involve adjust-
there is no peak during the fixed peak hour. This section describes the ments or compensation based on residual data from the prior day,
proposed methodology of the dynamic pricing scheme for charging EVs month, or year. To ensure customer convenience, a day-ahead time slot-
with mathematical modelling. based price variation range is established. This range provides customers
Fig. 2 displays the step diagram of the overall workflow of the pro- with maximum and minimum limits as well as a base price, aiding them
posed method to form a dynamic pricing scheme for centralized EV in scheduling their EV charging.
charging stations (EVCS). The proposed dynamic tariff has three major Finally, the methodology combines the dynamic electricity pricing,
components: dynamic electricity pricing component, dynamic demand dynamic demand charge, and dynamic investment recovery components
charge component and dynamic investment recovery component. The for a particular time on a given day to generate the final dynamic
recipient of the first two components is the Energy Network Service pricing. This daily dynamic pricing remains within the day-ahead price
Provider (NSP) or utility and the recipient of the third component is a variation range, which is communicated to customers 24 hours in
public or private investor in the charging station. The third component is advance of their charging sessions.
a payback tariff component that covers investment, operation and
maintenance costs, all rents, and a suitable profit.
The methodology operates in six distinct steps. It commences by 2.1. Development of Different Pricing Components
gathering essential data from the utility, charging station, and cus-
tomers. The essential data includes forecasted day-ahead and real-time As shown earlier in Fig. 2, the methodology works in six steps.
demand data that comes from a well-established and separate However, the most important three steps are for forming three major
components: dynamic electricity pricing component, dynamic demand

Fig. 2. Step diagram of the overall workflow of the proposed methodology.

3223
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

charge component and dynamic investment recovery component. day-ahead demand curve. While distributing, an additional 10 % profit
is included with the calculated income. This additional amount can be
2.1.1. Dynamic Electricity Pricing Component used to subsidize for any undesired situation such as lower utilization,
The development method of the dynamic electricity pricing compensate for any unwanted deficit, and fund any incentive-based
component has two stages: forming a day-ahead dynamic time of use pricing activities. The income distribution is done according to the
(DToU) tariff with a critical peak pricing (CPP) and developing the real- pattern of the load curve and the number of vehicles in every slot. A
time adjustment named real-time pricing (RTP). The flowchart for the dynamic time of use (DToU) pricing profile is back-calculated from the
methodology of dynamic electricity pricing is illustrated in Fig. 3. redistributed income. DToU pricing profile will follow the load pattern,
which means the price will increase if the load increases and vice versa.
2.1.1.1. Day-ahead dynamic time of use (DToU) tariff profile. The first After forming the day-ahead DToU pricing profile, a supplementary
stage of the dynamic electricity pricing profile development is the day- pricing component is recommended for slots with critical peak condi-
ahead DToU tariff which is formed before a day begins. This will be tions, denoted as Critical Peak Pricing (CPP). While declaring CPP, the
utilized in day-ahead scheduling for EV charging. The tariff formation position of slots that have available power just below the requirement
starts by gathering the projected load curve for the upcoming day. The (less than 100 %, but no lesser than 80 %) of a single charger is iden-
day will be divided into 24 time slots, each with a 1-hour time period. tified. If any slot is found, a CPP of a pre-determined amount named
Additional information, such as existing static ToU or flat-rate tariff “CPPT” is proposed. This pre-determined amount can be a value around
rates and EV charging station parameters (feeder capacity, number of the usual peak price of the existing static ToU tariff. If the available
vehicle chargers, power requirement, etc.) are also required. The power of a slot is more than 100 % of the power requirement of a single
available power curve is created from the gathered information, and the charger, CPP is not required and, hence, not proposed. The inclusion of
highest number of chargers that can be fitted under the curve is CPP means a vehicle can still get a charge with lesser power or overload
determined. the feeder from the distribution transformer marginally for a very brief
The total income of the charging station is then calculated, which can period. In return, the EV owner will repay with a much higher fee. This
be collected by a static ToU tariff imposed by the existing regulatory CPP component is merged with the DToU pricing profile to complete the
body. Here, it is assumed that all active chargers that can be fitted under day-ahead electricity pricing profile, which will be utilized in day-ahead
the available curve are engaged in charging. This is done by adding the electric vehicle charge scheduling. The day-ahead electricity pricing
slot-wise products of static ToU tariff, the power requirement of each profile may also include an adjustment component from the previous
charger, and the concurrent number of chargers that can be fitted under day (denoted as ND_RTP).
the curve (Khan and Aziz, 2022). After that, the calculated income is
distributed among the 24-time slots according to the pattern of the 2.1.1.2. Real-Time Pricing (RTP) components. The second stage is the

Fig. 3. Flowchart for the methodology of dynamic electricity pricing.

3224
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

formation of a real-time pricing (RTP) component, which is an adjust- be used for charge scheduling.
ment for the day-ahead DToU tariff. The load and available power may
vary from the day-ahead forecasted value in real-time. This variation 2.1.2. Dynamic demand charge component
should have an impact on dynamic electricity pricing. To address this Usually, the utilities charge the EV charging stations a fixed monthly
variation in demand, an adjustment profile named RTP is proposed in amount per kW of sanction load. However, this fixed amount needs to be
the methodology, which will be added to the day-ahead DToU pricing recovered from the EV owners against their energy consumption (per
profile before billing. The process of developing the RTP component kWh) while charging EVs. Fig. 4 illustrates the methodology for devel-
starts with collecting the forecasted demand data, which is done oping a dynamic demand charge profile. The methodology starts by
10 minutes before each slot. The demand data shows the new available collecting charging station parameters (number of EV chargers, power
power curve, and the number of EV chargers that can be fitted under the ratings etc.), demand charge information (per kW monthly charge),
curve is re-estimated. number of days in months of a particular year, and day-ahead and real-
Here, if the number of chargers rises due to less residential demand time energy usage by EVs. Then, the monthly demand charge that the
and more available power, the EV charging station will be able to earn charging station will pay to the utility is calculated. After that, the
more by charging an additional EV. While giving back this unplanned amount that the charging station needs to recover each day of the month
extra income, the respective or affected slot (‘ith’) shares n% (n = dis- is found. Any deficit from the previous day is also adjusted with this
tribution ratio) of the profit and the next 23 slots get the rest (1− n)% of daily amount. Then, the dynamic demand charge per kWh is calculated
the profit. In such a case, the RTP component will have negative values. during day-ahead scheduling considering day-ahead forecasted energy
On the other hand, if the number of chargers decreases due to the rise in usage and minimum utilization factor (Umin) so that the maximum per
residential demand and, hence, a reduced amount of available power, kWh amount remains under a limit. All EVs are then charged with the
the charging station will face a deficit for charging one less EV. While calculated value of per kWh demand charge throughout the day. At the
recovering this unexpected deficit, the respective or affected slot (‘ith’) end of the day, the actual recovery and deficit amounts are determined.
bears n% of the deficit and the next 23 slots bear the rest (1− n)% of the If any deficit is found, then that amount is added to the daily amount
deficit. In such a case, the RTP component will have positive values. In that needs to be recovered the next day. If a month ends, the total
both cases, some of the values from the RTP component will go to the recovered amount is calculated, which will be paid to the utility.
next day. Those values will be enlisted to the “Next Day RTP (ND_RTP)
Component”. This “ND_RTP” component will be added to the day-ahead 2.1.3. Dynamic investment recovery or payback component
DToU profile of the next day. The third component of the proposed dynamic pricing scheme is the
After forming the RTP components for the current and the next day, dynamic investment recovery or payback component. The recipient of
the value of “n” is optimized, and the final dynamic electricity pricing is the income from this component is a public or private investor in the
completed. The value of “n” has an impact on final billing, the daily charging station. It covers investment, operation and maintenance costs,
income of the charging station and users’ comfort. So, an act of opti- all rents, and a suitable profit. This tariff component ensures the payback
mization is required to determine the near-optimum value of “n”. The of the investment in the charging station. The value of this tariff
RTP component (with near-optimum value of “n”) for the current day is component is a fixed amount for all time slots of a particular day.
added to the day-ahead pricing profile of the day, whereas the ND_RTP However, this fixed amount can be different for different days in the
will be added to the day-ahead pricing profile of the next day, which will considered payback period. After the dynamic investment recovery or

Fig. 4. Flowchart for the formation of dynamic demand charge component.

3225
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

payback component is calculated, it is added to the other two compo- the beginning through direct payment by a lump sum amount. No
nents and the dynamic pricing scheme is completed. The flowchart for money is borrowed; hence, no interest rate is added to the considered
the methodology to calculate the dynamic investment recovery or varying inflation rate. The operation and maintenance cost for electric
payback component is illustrated in Fig. 5. It starts by collecting the vehicle charging stations is different from conventional ICE vehicle
charging station parameters (such as the total number of EV chargers, refuelling stations (Danial et al., 2021). Usually, the cost of energy
power ratings, etc.), all cost information, profit margin, intended period consumed by the EVs during charging is considered in the operation cost
of payback and the information about the particular day of a particular of the EV charging station. However, in this study, the energy cost is
month of a particular year. Then the lifetime cost (CLifetime) is calculated already covered by the dynamic electricity pricing component. The
where the net cash amount is calculated in the “Present-Worth Method”. operation & maintenance cost covers the costs of running the charging
Three main cost components are considered in this study to calculate station, such as the cost of measurement and monitoring systems, sub-
the lifetime cost, such as acquisition cost (CAcquisition), operation and scription fees of different software and services, cost of charging scheme
maintenance cost (CO&M), and land-related cost (CLandRent). The acqui- implementation, cost of station-customer communication, the cost of
sition cost includes all equipment costs along with their installation, labour during the period and the cost of annual maintenance. Lastly,
taxes, supplementary materials, permits, and decoration (Danial et al., land should be taken into possession to establish an EV charging station.
2021; Nicholas, 2019). It is assumed that the acquisition cost is paid at The land can be owned by the investor or by any entity. In many cases,

Fig. 5. Flowchart for the formation of dynamic investment recovery or payback component.

3226
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

the land can be bought before the installation solely for the EV charging
station. However, this study assumes that the land is rented for the entire
payback period in exchange for yearly rent. It is also assumed that the
rent is fixed in terms of value but not in terms of net amount because of
varying inflation rates.
All the future costs should be converted to the present worth and
then added. However, in this paper, it is already assumed that all the
future costs will have the same present value. The net amount will be
different for different years because of the varying inflation rates. Since
the operation, maintenance, and land-related costs are periodic and
have the same values as the first year, these can be added and multiplied
by the total number of years in the intended payback period. Then, this
product can be added to the fixed acquisition cost to form the total
lifetime cost. After determining the total lifetime cost in present worth,
the amount of money to be recovered in the first year is estimated. While
calculating, the profit margin should also be considered here. After
Fig. 6. Illustration of the day-ahead price variation range.
finding the amount to be recovered in the starting year, the future
amounts for the rest of the years should also be calculated. These future
amounts should be calculated in the “Future-Worth Method”. Then, the the price variation range.
amounts of money to be recovered each day and each month of different
years are determined. The consideration of having a leap year should be 2.3. Mathematical modeling
included here.
When the amount of money to be recovered in a day of a particular The required mathematical modelling for three major components
month of a particular year is already found out, the minimum/ baseline (dynamic electricity pricing, dynamic demand charge and dynamic in-
payback tariff per kWh of energy is determined for that particular day. vestment recovery (payback) components) of the proposed dynamic
This value is calculated considering the full occupancy of all the chargers pricing scheme is developed and displayed below.
at all hours of the day. This is a highly improbable case for an EV
charging station that is established in a residential complex and powered 2.3.1. Mathematical modeling for dynamic electricity pricing component
by the same feeder of the complex. The EV charging station will prob- The most important component of the proposed dynamic pricing
ably not be able to supply the maximum possible energy in a day due to scheme is the dynamic electricity pricing profile. This is a process that
the fact that the station is designed with the maximum number of has to be done every day of operation because of the dynamicity of the
chargers that can be fitted under the maximum available power. That is tariff structure.
why the calculated payback tariff is the minimum and hence identified
as the baseline. This value will remain fixed for all days in a year. 2.3.1.1. Mathematical modeling for day-ahead dynamic time of use
The methodology has three important adjustments for this payback (DToU) Tariff. The day-ahead DToU tariff is formed before a day begins,
component: daily, monthly, and yearly adjustments. All these adjust- which will be utilized in day-ahead scheduling for EV charging. The
ments are added to the proposed baseline payback tariff to form the tariff formation starts by gathering the necessary information. From the
dynamic investment recovery or payback tariff. This final payback tariff gathered information, the available power curve is created by using Eq.
is dynamic as it is being adjusted on a daily, monthly, and yearly basis. (1) (Abdullah-Al-Nahid et al., 2020). After generating the available
However, this will remain constant for all time slots of the day. For the power curve, the highest number of chargers that can be fitted under the
starting day these adjustments have zero values. So, EVs are charged curve is determined by Eq. (2) (Khan and Aziz, 2022). In most cases, this
with the baseline payback tariff in the first day. After each day, the daily will be a fractional number. To round down the number Eq. (3) is used
return and deficit are calculated and the daily adjustment profile is (Khan and Aziz, 2022). The rounded integer denotes the highest number
recalculated. Similarly, after each month and year, the monthly and of EVs that get a full charge.
yearly adjustment profiles are also recalculated based on the monthly PAvailable (i) = PMaxCapacity − PDemand (i) (1)
and yearly deficit amounts. The daily, monthly, and yearly adjustments
are kept under limits so that these cannot increase the final payback PAvailable (i)
tariff abnormally which may create pressure on the customers. The NAC (i) = (2)
PCharger
limits are set by three thresholds (AD, AM, and AY) which are optimized
to ensure users’ comfort and proper recovery of the investment. NACs (i) = Round down{NACs (i)} (3)

2.2. Generation of price variation range where, PAvailable = available power in kW; PMaxCapacity = maximum sup-
ply capacity of the feeder in kW; PDemand = demand for residential load in
To increase customers’ convenience, a day-ahead price variation kW; NAC = number of active chargers that can be fitted under the curve
range is generated by the proposed methodology. This day-ahead price with decimal fraction; NACs = number of chargers after rounding down
variation range will help EV owners schedule their EVs for charging. and PCharger = power requirement of one charger.
This will also aid the coordinated charging scheme to operate more The total income of the charging station is then calculated, which can
efficiently. In this day-ahead price variation range, a calculated “Base be collected with existing static ToU tariff by Eq. (4) where the slot-wise
Day-ahead Pricing” will be offered along with a probable upper limit and products of static ToU tariff, the power requirement of each charger, and
a lower limit. The day-ahead price variation range is visually illustrated the concurrent number of chargers that can be fitted under the curve are
in Fig. 6. The upper and the lower limits of the variation range are added (Khan and Aziz, 2022). After that, the calculated income from Eq.
determined by considering the maximum deviation (in both positive and (4) is distributed with an additional 10 % profit among the 24 time slots
negative manners, ± x%) in real-time residential demand compared to by Eq. (5) and Eq. (6).
the forecasted day-ahead residential demand. Two dynamic pricing
profiles are formed considering both the positive and the negative de-
viation in residential demand, which are the upper and lower limits of

3227
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

24
∑ ⎧
{ }
TIstaticToU = TstaticToU (i) × PCharger × NACs (i) (4) ⎨ 0; NAC (i) ≥ 1
i=1 CPP(i) = CPPT ; 0.8 ≤ NAC (i) ≤ 1 (10)
⎩
∞; NAC (i) < 0.8
TIDToU = 1.1 × TIstaticToU (5)
where, CPP = Critical Peak Pricing component and CPPT = pre-
24 {
∑ } determined additional amount for critical peak time.
TIDToU = TDToU (i) × PCharger × NACs (i) (6)
i=1
2.3.1.2. Mathematical Modeling for Real-Time Pricing (RTP) Component.
where, TIstaticToU = total income accumulated in a day by static ToU From the real-time demand data, which are forecasted 10 mins before
tariff; TstaticToU = static ToU tariff imposed by a regulatory body; TIDToU each slot, the new available power curve and the number of EV chargers
= total income that has to be achieved in a day by dynamic ToU tariff that can be fitted under the curve are re-estimated by Eqs. (11), (12), and
and TDToU = dynamic ToU pricing. (13).
To make the pricing profile imitate the load profile, a coefficient “C”
PRT_Available (i) = PMaxCapacity − PRT_Demand (i) (11)
is declared, which is shown in Eq. (7) (Khan and Aziz, 2022). This co-
efficient “C” will give the DToU pricing profile after being multiplied by
PRT_Available (i)
slot-wise demand. From Eqs. (5), (6), and (7), the following equations NRT_AC (i) = (12)
PCharger
can be obtained in which the value of “C” is determined. After deter-
mining from Eq. (8) & (9), the value of “C” is put into Eq. (7) and the NRT_ACs (i) = Round down{NRT_ACs (i)} (13)
day-ahead dynamic time of use (DToU) pricing profile is back calculated
(Khan and Aziz, 2022). where, PRT_Available = real-time available power in kW; PRT_Demand = real-
TDToU (i) = C × PDemand (i) (7) time demand for residential load in kW; NRT_AC = number of active
chargers that can be fitted under the real-time available power curve
24
∑ { } with decimal fraction and NRT_ACs = number of chargers in real-time
TIDToU = C × PDemand (i) × PCharger × NACs (i) (8) after rounding down.
i=1
The equations to form the RTP and the Next-Day RTP (ND_RTP)
components are illustrated below. If NRT_ACs(i) > NACs(i), the RTP and
ND_RTP components will have negative values, which can be obtained
by Eq. (14) and Eq. (15).

⎧
⎨ RTP (k) − n × TDToU (i),
0
⎪ [here, k = i, (‘i’ is the affected slot)]
1
RTP (i) = (1 − n) × T DToU (i) (14)
⎩ RTP0 (k) −
⎪ , [here, k = ʹʹi + 1ʹ to 24, (remaining slots)]
23

TIDToU
C= 24 {
∑ }
PDemand (i) × PCharger × NACs (i) (1 − n) × TDToU (i)
ND_RTP1 (j) = ND_RTP0 (j) − , [here, j
i=1 23
1.1 × TIstaticToU = 0 toʹʹi − 1ʹʹ, (of next day) (15)
= 24 {
(9)
∑ }
PDemand (i) × PCharger × NACs (i)
i=1 where, RTP = real-time pricing component for the current day; ND_RTP
= real-time pricing component for the next day; n = distribution ratio
After forming the day-ahead dynamic time of use (DToU) pricing
(percentage of share of RTP allocation in affected slot) and i = position
profile, a supplementary pricing component (Critical Peak Pricing
of the affected slot of the current day.
(CPP)) is recommended by Eq. (10) for slots that have critical peak
If NRT_ACs(i) < NACs(i), the RTP and ND_RTP components will have
conditions. While declaring CPP, the position of slots that have available
positive values, which can be obtained by Eq. (16) and Eq. (17).
power just below the requirement (less than 100 %, but no lesser than

⎧
⎨ RTP (k) + n × TDToU (i),
0
⎪ [here, k = i, (‘i’ is the affected slot)]
1
RTP (i) = (1 − n) × T DToU (i) (16)
⎪ RTP0 (k) +
⎩ , [here, k = ʹʹi + 1ʹʹ to 24, (remaining slots)]
23

80 %) of a single charger is identified. If any slot is found, a CPP of a pre-

determined amount named “CPPT” is proposed. This pre-determined
amount can be a value around the usual peak price of the static ToU
(1 − n) × TDToU (i)
tariff imposed by a regulatory body. ND_RTP1 (j) = ND_RTP0 (j) + , [here, j
23
= 0 to ʹʹi − 1ʹʹ, (of next day) (17)

3228
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

The RTP component for the current day is added to the day-ahead
DCARDaily (d, m, y) = TDemand (d, m, y) × EkWh (d, m, y) (27)
pricing profile of the day. In contrast, the ND_RTP will be added to
the day-ahead pricing profile of the next day using Eq. (18) to get the
final dynamic electricity pricing profile.
TElectricity (i) = TDToU (i) + CPP(i) + ND_RTPfrom previous day (i) + RTP(i) (18) where, TDemand = dynamic demand charge per kWh; EkWh-DA = day-
ahead amount of total energy consumption by EVs; TDC_max =
where, TElectricity = final dynamic electricity pricing profile. maximum per kWh demand charge; Umin = minimum station utilization
factor and DCARDaily = daily demand charge actually recovered by the
2.3.1.3. Mathematical Modeling for Distribution Ratio (Value of “n”) station.
Optimization. A very important step would be optimizing the distribu-
tion ratio (value of “n”) in the RTP components. Eq. (19) defines the 2.3.3. Mathematical Modeling for Dynamic Investment Recovery
objective function of the optimization problem. The constraint function Component
is shown in Eqs. (20) and (21).
2.3.3.1. Mathematical Modeling for Cash Recovery. Total lifetime cost
Objective function : Max[n] (19)
(CLifetime) is calculated using Eq. (29), where net cash amount is calcu-
TITElectricity (n) ≥ TIDToU (20) lated in the “Present-Worth Method”. Then, the amount of money (with
profit) to be recovered in the first year is estimated using Eq. (30). The
24 {
∑ } future amounts for the rest of the years should also be calculated in the
TITElectricity (n) = TElectricity (i, n) × PCharger × NACs (i) (21) “Future-Worth Method” using Eq. (31). The amounts of money to be
i=1 recovered each day and each month of different years are determined
using Eq. (32) and Eq. (33) respectively.
where, TITElectricity(n) = total income of the day calculated with final
electricity pricing profile formed with different “n”; TIDToU = total in- CLifeTime = CAcquisition + ypayback × (CO&M + CLandRent ) (29)
come that has to be achieved in a day by dynamic ToU tariff and TEle- ( )
ctricity(i,n) = dynamic electricity pricing profile formed with different CLifeTime × 1 + PMargin
MRYearly (1) = (30)
“n”. ypayback

2.3.2. Mathematical Modeling for Dynamic Demand Charge Component MRYearly (y[2≤y≤ypayback ] ) = MRYearly (y − 1) × {1 + IRate (y − 1)} (31)
The monthly demand charge that the charging station will pay to the
utility is calculated by Eq. (22). The amount that the charging station MRYearly (y)
MRDaily (y) = (32)
needs to recover each day of a month can be found by Eq. (23). Any ND− Y (y)
deficit from the previous day is adjusted with the daily amount in Eq.
(24). MRMonthly (m, y) = MRDaily (y) × ND− M (m, y) (33)

DCRMonthly (m, y) = DCkW (m, y) × PCharger × NCharger (22) where, CLifetime = total lifetime cost; CAcquisition = one-time acquisition
cost; CO&M = yearly operation & maintenance cost; CLandRent = yearly
DCRMonthly (m, y) land-related cost; ypayback = total number of years in the intended
DCRDaily (d, m, y) = (23)
ND− M (m, y) payback period; MRYearly = the amount of money to be recovered in
different year; IRate = varied inflation rates (projected); y = respective
DCRDaily_adj (d, m, y) = DCRDaily (d, m, y) + DCDDaily (previous day) (24) year; MRDaily = amount of money to be recovered in a day of different
years; ND-Y = number of days in different years; MRMonthly = amount of
where, DCRMonthly = monthly demand charge that needs to be recovered;
money to be recovered in a month of different years and ND-M = number
DCkW = per kW demand charge; NCharger = number of installed chargers;
of days in different months of different years.
DCRDaily = daily demand charge that needs to be recovered; ND-M =
number of days in different months of different years; DCRDaily_adj = daily
2.3.3.2. Mathematical Modeling for Baseline Payback Tariff. The mini-
demand charge after deficit adjustment and DCDDaily = deficit in de-
mum threshold for the investment recovery component or the baseline
mand charge recovery.
payback tariff per kWh of energy for a particular day of a year is
The dynamic demand charge per kWh is calculated by Eq. (25). The
determined using Eq. (34). This value will remain fixed for all days in a
maximum per kWh amount is limited by Eq. (26), considering a mini-
year.
mum station utilization factor Umin. The actual amount that the charging
station recovers each day of a month is determined by Eq. (27). By using MRDaily (y)
TBLPB (y) = (34)
Eq. (28), any deficit in demand charge recovery is calculated. 24 × NCharger × PCharger
⎧
⎨ DCRDaily_adj (d,m, y), if TDemand (d,m,y) < TDC_max (m, y)
⎪
where, TBLPB = minimum threshold for the investment recovery
TDemand (d,m,y) = EkWh− DA (d,m,y)
⎪ component or the baseline payback tariff per kWh.
⎩
TDC_max (m,y), if TDemand (d,m,y) ≥ TDC_max (m, y)
(25) 2.3.3.3. Mathematical Modeling for Adjustment Profiles. The baseline
payback tariff is the minimum amount which needs to be adjusted
DCRDaily (d, m, y) further on a daily, monthly, and yearly-basis to recover the required
TDC_max (m, y) = (26)
Umin × PCharger × NCharger × 24hr money properly without any deficit and to restrict the final payback
tariff from fluctuating abnormally. All three adjustments are added to

{
DCRDaily (d, m, y) − DCARDaily (d, m, y), if DCRDaily (d, m, y) > DCARDaily (d, m, y)
DCDDaily (d, m, y) = (28)
0, if DCRDaily (d, m, y) ≤ DCARDaily (d, m, y)

3229
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

the proposed baseline payback tariff to form the dynamic investment 12 ND−∑
M (j,y)
∑
recovery component (final payback tariff) which is shown in Eq. (35). DYearly (y) = MRYearly (y) − {TFPB (i, j, y) × EkWh (i, j, y)} (42)
j=1 i=1
TFPB (d, m, y) = TBLPB (y) + AdjDaily (d, m, y) + AdjMonthly (d, m, y)
(35) {
+ AdjYearly (d, m, y) 0, if DYearly (py) ≤ 0
AdjYearly (d, m, y) = (43)
xy (d, m, y), if DYearly (py) > 0
where, TFPB = final payback tariff /investment recovery component per
kWh; AdjDaily = daily adjustment per kWh; AdjMonthly = monthly
adjustment per kWh and AdjYearly = yearly adjustment per kWh.

⎧
⎪ D (py) × {1 + IRate (y − 1)} DYearly (py) × {1 + IRate (y − 1)}
⎪ Yearly
⎪
⎪
⎪
⎪ N D− Y (y) ND− Y (y)
⎪
⎪ , if ≤ AY × TBLPB (y)
⎨ EkWh (d, m, y) EkWh (d, m, y)
xy (d, m, y) = (44)
⎪
⎪
⎪ DYearly (py) × {1 + IRate (y − 1)}
⎪
⎪
⎪ ND− Y (y)
⎩ AY × TBLPB (y), if > AY × TBLPB (y)
⎪
⎪
EkWh (d, m, y)

On the starting day, these adjustments are assumed to have zero where, DMonthly = total monthly deficit; xm = numerical value of the
value and the EVs are charged with baseline payback tariff. Deficits are monthly adjustment per kWh; pm = previous month; AM = maximum
calculated on a daily, monthly, and yearly-basis. From the 2nd day of the threshold for monthly adjustment (in percentage); DYearly = total yearly
1st year, daily adjustment may be a non-zero value based on the pre- deficit; xy = numerical value of the yearly adjustment per kWh; py =
vious day’s deficit. The daily deficit is calculated using Eq. (36). The previous year and AY = maximum threshold for yearly adjustment (in
daily adjustment profile is determined by Eq. (37). To restrict the next- percentage).
day adjustment from increasing the payback tariff more than a limit, a
threshold AD (a percentage of the baseline payback tariff) is applied 2.3.3.4. Mathematical Modeling for Adjustment Thresholds (Values of AD,
which is shown in Eq. (38). AM and AY) Optimization. The daily, monthly, and yearly deficits are
recovered by adding adjustments. The values of these adjustments
DDaily (d, m, y) = MRDaily (y) − TFPB (d, m, y) × EkWh (d, m, y) (36)
directly depend on the maximum adjustment thresholds (AD, AM, and
{ AY). If these values are larger, the recovery of investment with a profit
0, if DDaily (pd) ≤ 0
AdjDaily (d, m, y) = (37) will be faster as the final payback tariff increases, which may create
xd (d, m, y), if DDaily (pd) > 0
pressure on the customers. On the other hand, if the values of the
⎧
DDaily (pd) DDaily (pd) threshold constants are smaller, the recovery of investment with a profit
⎪ if ≤ AD × TBLPB (y)
⎪
⎨ EkWh (d, m, n),
⎪
EkWh (d, m, y) will be slower as the final payback tariff decreases. There will be less
xd (d, m, y) = (38) than the required return on investment at the end of the set payback
⎪ DDaily (pd)
⎪
⎩ AD × TBLPB (y),
⎪ if > AD × TBLPB (y) period, which is a highly unexpected situation for investors. So, the
EkWh (d, m, y) values must be optimized so that these are as minimum as possible yet
large enough to recover the investment with a profit within the expected
where, DDaily = total daily deficit; EkWh = actual energy consumption by
payback period. Eq. (45) defines the objective function of the optimi-
EVs in the considered day; xd = numerical value of the daily adjustment
zation problem. The constraint function is shown in Eqs. (46) and (47).
per kWh; pd = previous day and AD = maximum threshold for daily
adjustment (in percentage). Objective function : Min[AD , AM , AY ] (45)
Similarly, monthly and yearly adjustments are made from 2nd month
and 2nd year, respectively. The monthly and yearly deficits are calcu- Npayback
∑
lated using Eq. (39) and Eq. (42). The monthly and yearly adjustment ADYearly = DYearly (y) ≤ 0 (46)
y=1
profiles are determined by Eq. (40) and Eq. (43). Like daily adjustment,
limiting thresholds AM and AY (a percentage of the baseline payback AD ≥ AM ≥ AY (47)
tariff) are also applied which is shown in Eq. (41) and Eq. (44).
ND−∑
M (m,y)
where, ADYearly = aggregated yearly deficit amount.
DMonthly (m, y) = MRMonthly (m, y) − {TFPB (i, m, y) × EkWh (i, m, y)}
i=1 2.3.4. Mathematical modeling for final dynamic pricing
(39) The final dynamic pricing profile for a day of a month of a year is
{ determined by adding all three components using Eq. (48), which will be
0, if DMonthly (pm) ≤ 0 used during final billing.
AdjMonthly (d, m, y) = (40)
xm (d, m, y), if DMonthly (pm) > 0
TFinalDynamicPricing (i,d,m,y) = TElectricity (i,d,m,y)+TDemand (d,m,y)+TFPB (d,m,y)
⎧
⎪ DMonthly (pm) DMonthly (pm) (48)
⎪
⎪
⎪
⎪ ND− M (m, y) ND− M (m, y)
⎪
⎨ EkWh (d, m, y),
⎪ if ≤ AM × TBLPB (y) where, TFinalDynamicPricing = final dynamic pricing profile for a certain day
EkWh (d, m, y)
xm (d, m, y) = (41) and (i,d,m,y) = ith slot of the dth day of mth of yth year.
⎪
⎪
⎪ DMonthly (pm)
⎪
⎩ AM × TBLPB (y), if ND− M (m, y) > AM × TBLPB (y)
⎪
⎪
⎪ 2.3.5. Mathematical modeling for price variation Range
EkWh (d, m, y)
The upper and the lower limits of the variation range are determined

3230
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

by considering the maximum deviation in real-time residential demand design. The battery size of this particular EV is 55 kWh, which a 50-kW
compared to the forecasted day-ahead residential demand (e.g., ± x%), fast charger can charge in nearly 60 mins (Tesla Model 3 3, 2021). It is
which are shown in Eqs. (49) and (50). Then the upper and lower limits also assumed that all EVs will come to the charging station through a
are calculated for the upper and lower demand levels, respectively. The centralized and coordinated charging scheme. This is why it is also
base day-ahead pricing of the variation range is determined by Eq. (53). assumed that the charging slots are occupied with 100 % maximum
possible occupancy considering concurrent available power. The simu-
PRT_Demand_up (i) = (1 + x) × PDemand (i) (49)
lation results for all three components of the dynamic pricing are dis-
played, analyzed, and corrected after any adjustment is made. The tariff
PRT_Demand_low (i) = (1 − x) × PDemand (i) (50)
structure for a random day is illustrated alongside the first day to vali-
TFDP_up (i, d, m, y) = TElectricity_up (i, d, m, y) + TDemand (d, m, y) + TFPB (d, m, y) date the methodology. The pricing components are demonstrated in
Bangladeshi Taka or Tk. (1 Tk. = 0.0091 USD).
(51)

TFDP_low (i, d, m, y) = TElectricity_low (i, d, m, y) + TDemand (d, m, y) + TFPB (d, m, y) 3.1. Dynamic electricity pricing
(52)
The simulation results for the dynamic electricity pricing are shown
TBDAP (i, d, m, y) = TDToU (i, d, m, y) + CPP(i, d, m, y) + ND_RTP(i, d, m, y) in Fig. 7 to Fig. 41. The considered parameters for this simulation are
+ TDemand (d, m, y) + TFPB (d, m, y) presented in Table A.1. The dynamic electricity pricing profiles are
formed in two stages: day-ahead dynamic time of use (ToU) profile
(53)
formulation and real time pricing (RTP) adjustment. Day-ahead dy-
where, PRT_Demand_up = real-time demand considering x% increment in namic ToU has a critical peak pricing (CPP) component in it. Day-ahead
kW (for upper limit); PRT_Demand_low = real-time demand considering x% dynamic ToU is formed by restructuring the existing static ToU tariff of
reduction in kW (for lower limit); ± x = maximum deviation in fore- Bangladesh based on the forecasted load curve of the residential com-
casted day-ahead residential demand (in percentage); TFDP_up = upper plex. It assumed that the considered centralized charging station takes
limit of the day-ahead price variation range for a certain day; TFDP_low = its required power from the same transformer that the residential
lower limit of the day-ahead price variation range for a certain day; complex has been using. Real-time pricing (RTP) component is modeled
TElectricity_up = upper limit of the calculated electricity pricing for upper based on the real time variation in the demand curve.
demand level; TElectricity_low = lower limit of the calculated electricity
pricing for upper demand level and TBDAP = base day-ahead pricing of 3.1.1. Day-ahead dynamic ToU profile formulation
the variation range for a certain day. The day-ahead dynamic time of use (ToU) profile formulation pro-
cess for the starting day (day-1 of year-1) is illustrated in Fig. 7 to
3. Results and analysis Fig. 14. The process starts by taking the existing static ToU tariff rates
given by the regulatory body and the forecasted demand curve of the
The simulation results of the proposed dynamic pricing scheme are feeder which powers the residential complex and charging station. Fig. 7
shown and analyzed in this segment for justification. The simulation is shows the existing static ToU tariff rates for a battery charging station at
designed in the MATLAB “R2021a” application installed in a 64-bit a medium tension (MT-11 kV) voltage level, which is imposed by the
Windows 10 operating system. The simulation is designed with 9 Bangladesh Energy Regulatory Commission (BERC) (Tariff Rate, 2023).
years of random data to imitate the cash flow for the investment in an EV It has three fixed block rates: Super Off-peak, Off-peak, and Peak, which
charging station in Bangladesh. It is assumed that all investments and are placed at different predefined time slots of the day.
costs must be recovered with a substantial amount of profit within this 9- Fig. 8 displays the day-ahead (forecasted) demand and available
year payback period. In the simulation, 3 fast EV chargers, each with a curves of a particular feeder that provides energy to a residential com-
50-kW power rating, are considered. This means the station can charge a plex with a centralized EV fast charging station at an 11 kV voltage level
maximum of 72 EVs, and the maximum energy usage will be 3600 kWh inside Dhaka city. The demand curve, shown in Fig. 8(a), is obtained
in a day. However, the station will not be able to exploit the full energy from Ref (Abdullah-Al-Nahid et al., 2022). after a few modifications.
use capacity for most of the day due to residential loading. In some The demand curve is shown in a 24-hour format, meaning the load levels
occasions, a slightly slow charge is also considered when the available are assumed to remain constant for an entire hour (denoted by the time
power is at least 80 % of the power rating of a single charger. All EVs are slot). The slot-wise value of the demand is then subtracted from the
assumed to be fully charged in 60 mins. In the simulation, the Tesla maximum supplied power by the distribution transformer through the
Model 3 (2021) standard electric vehicle is taken into account for tariff considered feeder. Fig. 8(b) illustrates the available power curve
calculated by this subtraction. This available curve is necessary to
determine the maximum number of active chargers that can be used in a
slot.
Fig. 9 displays the highest number of active chargers that can be
fitted in each slot below the available power curve. Then, the slot-wise
number of chargers, kilo-watt (kW) rating of each charger, and slot-wise
existing tariff rates from BERC are multiplied and recorded against each
slot. This gives the collected slot-wise income which is generated by the
existing static tariff. Fig. 10 illustrates this slot-wise collected income for
the first day of the first year.
The slot-wise incomes generated by the existing static ToU tariff are
then added which yields the total usual income of the station. This in-
come is what the charging station would usually get if it uses the existing
static ToU tariff. Then, this income is restructured across all time slots
according to the day-ahead demand curve and the number of active
chargers in each slot. Fig. 11 shows the restructured equivalent charging
Fig. 7. Existing static ToU tariff rates for battery charging station (MT-11 kV) income. An extra 10 % profit is included here that can be awarded back
in year-1. later to the users as a part of a subsidy or an incentive-based program

3231
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Fig. 8. Day-ahead forecasted load curves of a residential complex of the considered area on starting day (Year-1, Day-1): (a) Demand (Abdullah-Al-Nahid et al.,
2022), (b) Available power.

Fig. 9. Number of chargers that can be fitted under available power curve
(Year-1, Day-1).
Fig. 12. Back calculated day-ahead DToU pricing (Year-1, Day-1).

Fig. 10. Slot-wise collected income by charging EVs if existing static ToU tariff
is used (Year-1, Day-1). Fig. 13. Forecasted Critical Peak Pricing (CPP) component (Year-1, Day-1).

Fig. 11. Restructured equivalent (with 10 % extra profit) charging income by Fig. 14. Day-ahead electricity pricing profile (Year-1, Day-1).
DToU tariff (Year-1, Day-1).

3232
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Fig. 15. Real-time demand and available power curves of the residential complex of the considered area on starting day (Year-1, Day-1): (a) Demand, (b) Available.

The proposed Critical Peak Pricing (CPP) component is then added to

the formed day-ahead Dynamic Time of Use (DToU) pricing profile and
completes the day-ahead electricity pricing. The combined day-ahead
electricity pricing profile is illustrated in Fig. 14. It can be perceived
that the peak value of the combined tariff is about 2.5 times the peak
price of the existing static ToU tariff. This combined day-ahead elec-
tricity pricing profile (DToU + CPP) is ready to be used for day-ahead EV
charge scheduling for the first day of operation.

3.1.2. Real-Time Pricing (RTP) component formulation

It is safe to assume that the forecasted day-ahead demand curve will
have considerable variation in real time. This should affect the elec-
tricity pricing if a tariff scheme is said to be dynamic. A true dynamic
Fig. 16. Number of active chargers that can be fitted under the real-time pricing scheme or tariff must address the real time variation in fore-
available power curve of the starting day (Year-1, Day-1). casted demand. In the proposed methodology, this real-time variation in
forecasted demand is addressed by an adjustment named the Real Time
such as peak-time rebates (PTR). From the restructured income profile, a Pricing (RTP) component. The formulation process of the Real Time
day-ahead tariff is back-calculated, which is dynamic in nature as it Pricing (RTP) component is illustrated in Fig. 15 to Fig. 18. In these
varies with the demand curve. This tariff scheme, shown in Fig. 12, has figures, adjustment is made to the day-ahead pricing profile of the
different values for different time slots and is named Dynamic Time of
Use (DToU) here.
A relatively large peak pricing is included in the proposed method-
ology for peak hours in which charging an EV may overload the trans-
former slightly. This huge pricing component is named Critical Peak
Pricing (CPP) and is shown in Fig. 13. From the figure, it can be observed
that the pricing at the 23rd slot is significantly high, in this case, equal to
the peak pricing of considered static ToU. This will make the final
pricing much higher when added to the DToU profile, discouraging most
customers from scheduling their electric vehicles in that slot. However,
the critical peak pricing will not be applicable for all peak situations.
This will only be imposed if the available power is marginally below the
power rating of a charger. For simplicity, a threshold of 80 % is applied
while designing the CPP. This potentially means an electric vehicle can
connect to a charger and get a charge at a slow rate or by overloading the
distribution transformer marginally. Fig. 18. Final dynamic electricity pricing profile (Year-1, Day-1).

Fig. 17. Real Time Pricing (RTP) components for current and next day with “n = 0.5” (Year-1, Day-1): (a) RTP component for current day, (b) Next day RTP
component which will be added to day-ahead electricity pricing profile.

3233
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Fig. 19. RTP components and dynamic electricity pricing profiles with different values of “n” (Year-1, Day-1): (a) RTP component for current day, (b) Next day RTP
component, (c) dynamic electricity pricing profiles.

starting day (day-1 of year-1).

Fig. 15 shows the real-time demand of the considered feeder of the
residential complex with its subsequent available curve. The real-time
demand for a particular slot is re-forecasted just 10 minutes before the
slot starts. The real-time demand of the starting day is illustrated in
Fig. 15(a). The corresponding available power is obtained from the
demand data, which is shown in Fig. 15(b). In this simulation, up to
±10 % variations from day-ahead values are considered in real-time.
The real-time variations will not affect the operation unless it changes
the number of maximum chargers in a slot. Fig. 16 displays the highest
number of chargers that can be fitted below the new real-time available
power curve. In comparison with day-ahead numbers from Fig. 9, the
number of chargers that can provide charge is lowered and grown in
some slots. For instance, the number of chargers is lowered in the 5th,
6th, and 20th slots, while the number rises in the 12th and 19th slots. In
other slots, the number remains the same. The charging station can Fig. 21. Final dynamic electricity pricing profile with “n = 0.253” (Year-1,
accommodate more or less EVs than calculated. This potentially means Day-1).
the day-ahead estimation for income will be affected because of real-
time variations in demand.
The income will increase if the station charges more vehicles than the
day-ahead estimation because of residential demand reduction and
decrease if the opposite happens. The lesser income will hurt the utility
because users are paying at the same rate that was designed for lower
demand. On the other hand, the added income should be returned to the
customers as they were charged at the same rate that was planned for a
higher demand level. That’s why the Real Time Pricing (RTP) compo-
nent is designed which is shown in Fig. 17. To make the income the same
as day-ahead estimation, this RTP component adds a positive value
when the income decreases due to the reduction in the number of active
chargers. However, if the methodology tries to recover the whole deficit
from the affected slot, then it will create a huge pressure on the other
vehicles in that slot that are already taking charge. That is why the
proposed methodology distributes this deficit and recovers it with a “n: Fig. 22. Additional 10 % profit from selling electricity at Final dynamic elec-
(1− n)” ratio from the affected slot and the next 23 slots. From Fig. 17(a) tricity pricing (Year-1, Day-1).
it can be seen that the 5th slot has a positive value in RTP components.

Fig. 20. Optimization of the distribution ratio (Year-1, Day-1): (a) Constrain: day-ahead forecasted income vs. other generated incomes with different values of “n”,
(b) Near-optimum value of “n”.

3234
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Fig. 23. Additional 10 % profit from selling electricity at Final dynamic electricity pricing when less EVs are being charged (Year-1, Day-1): (a) Actual number of
active chargers in real time, (b) Value of additional 10 % profit.

This will make the final pricing higher than the day-ahead pricing. This between day-ahead forecasted income and other generated incomes of
occurs due to the unexpected and real-time increment of demand in the starting day using different electricity pricing profiles formulated
those slots. In this case, only 50 % (n = 0.5) of the deficit is recovered with 6 different values of “n”. It can be observed that the total income
from the 5th slot and the rest of the 50 % ((1− n) = 0.5) is recovered decreases with the increment of the value of “n”. At some point, it even
from the next 23 slots. This happens for the 6th and 20th slots of the day goes below the day-ahead forecasted income. To prevent this, the value
too. On the other hand, the 12th and 19th slots have negative values in of “n” should be as small as possible. However, it will be unethical to
RTP components due to real time decrement in demand and this will recover or award one slot’s deficit or profit from other slots completely.
make the final pricing lower than the day-ahead pricing. In this case, The affected slots should bear the maximum amount of effect of the
50 % (n = 0.5) of the corresponding extra income is awarded back to deficit or profit. In this sense, the value of “n” should be as large as
these slots and the rest of the 50 % ((1− n) = 0.5) is given back to the possible. This contradictory requirement is settled by using the Particle
next 23 slots. This means an adjustment profile must be made for the Swarm Optimization (PSO) algorithm. The attainment of the near-
next day as well. This adjustment profile (named as Next Day RTP or optimum value of “n” for the starting day (day-1 of year-1) is shown
ND_RTP) is a part of the RTP component, which will be added to the day- in Fig. 20(b). The converged value of “n” for the first day is 0.253 for the
ahead electricity pricing profile of the upcoming day (shown in Fig. 17 simulated data set. After determining the near-optimum value of “n” for
(b)). The day-ahead electricity pricing profile from Fig. 14 was for the the day, RTP components are re-formed and added to the day-ahead
starting day in which there was no ND_RTP profile from the previous electricity pricing profile to form the final profile which is shown in
day. However, there might be one for the subsequent days. Fig. 21.
Finally, the attained RTP component for the current day is added to
the day-ahead pricing profile that will be used for the final billing pro- 3.1.4. Impact of fewer EVs on final electricity pricing
cess of EV charging as a tariff. Fig. 18 shows the final dynamic electricity In real-world scenarios, chargers may remain unoccupied despite the
pricing profile for the starting day that has a day-ahead DToU, a day- availability of needed energy because of the reduced number of EVs that
ahead CPP, and a real time RTP component (for current day). arrive at EVCS to get a charge. This will reduce the amount of additional
profit that EVCS was supposed to make with the static ToU tariff. Fig. 22
3.1.3. Optimizing the distribution ratio (value of “n”) shows the additional 10 % profit from selling electricity to EVs from
The effects of “n” (distribution ratio) in the RTP components and Fig. 16 at final dynamic electricity pricing (from Fig. 21). However, if
final electricity pricing profile are illustrated in Fig. 19. The value of “n” the number of EVs reduces from the maximum possible number as
is varied from 0 to 0.5 with an interval of 0.1 here. From Fig. 19(a) and shown in Fig. 23(a), the profit also reduces in respective slots as illus-
Fig. 19(b), it can be seen that the larger value of “n” has a bigger impact trated in Fig. 23(b).
on the respective slot of RTP components and a smaller impact on As the number of EVs reduces in some slots, it indicates these slots
subsequent slots. The opposite happens for the smaller value of “n”. are less popular among EV owners. The proposed scheme offers a sub-
Fig. 19(c) illustrates the real-time variation of final dynamic electricity sidy to these slots and hence the electricity pricing is reduced which may
pricing. be beneficial to the EV owners who are willing to take those slots. The
Finding the appropriate value of this distribution ratio (“n”) for a day amount of the subsidy for one unutilized charger (one less EV) is equal to
is purely an optimization problem. Fig. 20(a) presents a comparison the additional profit of a particular slot that is earned by charging one
EV. Fig. 24 illustrates the dynamic electricity prices before and after the
subsidy.

3.2. Dynamic demand charge

The dynamic demand charge proposed in this paper depends on the

money which needs to be recovered daily. Ideally, this daily amount is a
fixed value for all days in a month. However, this daily amount may
fluctuate significantly based on the real-time recovery and deficit.
Fig. 25(a) shows the daily demand charge which needs to be recovered
before and after deficit adjustment. The day-ahead dynamic demand
charge per kWh is shown in Fig. 25(b), which is generated based on the
daily amount after deficit adjustment. A 25 % of minimum station uti-
lization factor is considered here so that the maximum per kWh demand
charge remains under a limit. The maximum demand found in the 1st
Fig. 24. Final dynamic electricity pricing profile after adding the subsidy year is 0.22372 Tk./kWh for the considered data set.
(Year-1, Day-1). Fig. 26 shows the deficit and recovery in demand charge for the 1st

3235
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Fig. 25. Dynamic demand charge (Year-1): (a) demand charge to recover in a day, (b) day-ahead dynamic demand charge.

Fig. 26. Deficit and recovery in demand charge (Year-1): (a) daily deficit, (b) monthly recovery.

Fig. 27. Amounts of money to be recovered (all years): (a) each year, (b) per day of each year, (c) per month of each year.

year. In Fig. 26(a), the daily deficit profile is illustrated. This deficit 3.3. Dynamic investment recovery or payback component
occurs after using the day-ahead dynamic demand charge from Fig. 25
(b). In Fig. 26(b), monthly recovery amounts are shown from where it The designed payback tariff simulation results are shown in Fig. 27 to
can be seen that the recovered monthly demand charges are marginally Fig. 38. The receiver of the income from this tariff component is the
above the required amount. This validates the effectiveness of the pro- investor in the charging station. The considered parameters for this part
posed methodology for a dynamic demand charge. of the simulation are also presented in Table A.1. Before the formation of

Fig. 28. Baseline payback tariff per kWh: (a) year – 1, (b) all years.

3236
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Fig. 29. Daily deficit with only baseline payback tariff: (a) year – 1, (b) all years.

Fig. 30. Final payback tariff per kWh after daily, monthly & yearly adjustments: (a) year – 1, (b) all years.

Fig. 31. Daily deficit after daily, monthly & yearly adjustments: (a) year – 1, (b) all years.

the dynamic investment recovery or payback component starts, the based on the projected inflation rate. The baseline payback tariff per
variable inflation rates for the entire payback period are forecasted. The kWh for the first year is shown in Fig. 28(a), whereas Fig. 28(b) illus-
inflation rates are predicted for the years 2023–2032 using the ARIMA trates the baseline payback tariff of all years. The values of the baseline
(12,1,2) model based on the historical data set of the years 1987–2022. payback tariff for the 1st, 5th, and 9th year are 0.56486 Tk, 0.73708 Tk,
and 0.89077 Tk, respectively. These values imply that the final payback
3.3.1. Money to recover per year, per day and per month tariff will never go below these values.
The amounts of money to be recovered in each year are shown in Since the baseline payback tariff is the minimum possible tariff, the
Fig. 27(a). These amounts rise due to the effect of the inflation rate. The target of daily income will not be fulfilled on most days. This means
per-day quantities for those years are determined from the per-year there will be some amount of deficit on those days. The daily deficit
amounts, which are displayed in Fig. 27(b). Here, for calculating the profiles are shown in Fig. 29. From this figure, it can be seen that the
per-day quantities, the effect of the leap year is considered for the years yearly aggregated deficit is significant and needs to be reduced by
2024 and 2028. From the per-day amounts, the per-month quantities for adjustments.
the months of those years are determined, which are exhibited in Fig. 27
(c). These quantities are necessary to go further in the development 3.3.3. Payback tariff and deficits after daily adjustment
process of payback tariff and will be cross-examined for performance Since the proposed baseline payback tariff creates deficits in income
evaluation at the end of this paper. every day, three adjustments are made on a daily, monthly and yearly-
basis. These adjustments are from the second day, the second month and
3.3.2. Baseline payback tariff and deficits the second year, respectively. The payback tariffs and the deficit profiles
At first, the methodology finds the baseline payback tariff per kWh of after these adjustments are illustrated in Fig. 30 and Fig. 31, respec-
energy for a day of a particular year from the amount of money to be tively. By imposing the daily adjustment, the daily deficit from the
recovered on that day. This baseline payback tariff will be revised yearly baseline tariff is recovered the next day while charging the EVs. The

3237
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Fig. 32. Determination of near-optimum solution with PSO algorithm. Fig. 34. Final Dynamic Pricing profile for the first day (Year-1, Day-1).

daily adjustment increases the tariff for the next day. However, to keep investment with the payback tariff formulated by the proposed
this increment in check, an adjustment threshold (AD) is set to limit the methodology.
tariff. The value of AD is 0.5 here, so only 50 % of the baseline is added as
an adjustment from the second day. Because of this limiting threshold, 3.3.4. Optimization of the AD, AM and AY thresholds using PSO algorithm
the payback tariff after only daily adjustments can be saturated, which A version of the Particle Swarm Optimization algorithm is used to
results a deficit in the daily income on most days. The remaining deficit optimize the values of AD, AM, and AY thresholds. At first, the total
(if any) must be recovered by adding the monthly adjustment from the lifetime deficit amount is found for different values of AD, AM, and AY
2nd month of the starting year. This monthly adjustment increases the thresholds. In this case, the value of AD is varied from 0.1 to 3 (0–300 %)
tariff for the next month, provided there is a positive deficit in the with an incremental step of 0.1. However, the values of AM and AY
current month. However, to keep this increment in check, a monthly varied from 0.1 to 2 (0–200 %) with the same incremental step. This
adjustment threshold (AM = 0.4 here) is set to limit the tariff, just like the produces 12000 different combinations. Then the total lifetime deficit
daily adjustment. It can be observed that the tariff rates are higher amount is recorded along with the values of AD, AM, and AY thresholds
compared to the tariff with only daily adjustments except for the first against a solution number. This set of 12000 solutions is considered as
month of the first year. This is because there is no monthly adjustment in the search space in this optimization problem to be solved with the PSO
the first month, as the monthly deficit cannot be calculated until a algorithm.
month is passed. This is why the first month of the first year has a lower Since the search space is enormous, 1000 particles were considered
payback tariff (shown in Fig. 30) and higher deficit (shown in Fig. 31). as the swarm size. The considered PSO coefficients for this optimization
Even after daily and monthly adjustments, a positive deficit is found at are ω = 0.2, C1 = 0.9, and C2 = 0.5. Fig. 32 shows the determined near-
the end of the first year (Fig. 31(a)). This happens due to the saturation optimum solution of the optimization problem using the PSO algorithm.
in payback tariff after adding monthly adjustments with the limiting The found solution to the optimization problem is the 641st out of
threshold (AM). Any amount of deficit in income after a year must be 12000, where the minimum value of AD is 50 %, AM is 40 % and AY is
recovered in the next year. So, an adjustment is made to the payback 10 %.
tariff, which is denoted as the yearly adjustment. This yearly adjustment
increases the tariff for the next year provided there is a positive deficit in 3.4. Final dynamic pricing formation
the current year’s income. It is important to mention that the current
year’s deficit is reevaluated with the effect of the inflation rate before After formulating the “Dynamic Electricity Pricing”, “Dynamic De-
calculating the value of the yearly adjustment. Like the previous, an mand Charge” and “Dynamic Investment Recovery or Payback” com-
adjustment threshold (AY = 0.1 here) is set to limit the increment in the ponents, the final dynamic pricing profile can be obtained. The tariff rate
tariff for yearly adjustment. of a particular time slot of a particular day of a particular month of a year
These payback tariff profiles (shown in Fig. 30(b)) are the final prices can be determined by adding the electricity pricing of that slot with the
to recover the investment with a suitable profit. The analysis of deficits dynamic demand charge and the dynamic investment recovery or
can be an evaluation of the performance of the proposed tariff scheme. payback tariff of that day. The dynamic electricity pricing profile for the
The deficit profiles after all three adjustments are illustrated in Fig. 31 first day of the starting year is illustrated in Fig. 33(a), where the lowest
(b). The lifetime deficit is also indicated, which is − 307 Tk. for the price to charge an electric vehicle is 5.27837 Tk./kWh and the highest
considered data set, which indicates the successful recovery of the price is 26.6705 Tk./kWh. Compared to existing rates from BERC

Fig. 33. Dynamic Pricing components for the first day (Year-1, Day-1): (a) dynamic electricity pricing, (b) dynamic demand charge, (c) dynamic investment recovery
or payback tariff.

3238
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Fig. 35. Price variation range for the first day (Year-1, Day-1): (a) day-ahead, (b) actual real time billing price inside the range.

Fig. 36. Day-ahead forecasted and real time demand and available power curves of the residential complex (Year-1, Day-68): (a) Day-ahead, (b) Real time.

Fig. 37. Number of chargers that can be fitted under available power curves (Year-1, Day-68): (a) Day-ahead, (b) Real time.

(shown in Fig. 7), an EV can get a charge at 32 % less price considering EV owners can get high benefits from this dynamic electricity pricing
the lowest prices and discounting the 10 % extra profit. On the other scheme, which is already beneficial for the utility. Dynamic demand
hand, the utility can get paid at a much higher rate (around 2.4 times the charge and dynamic investment recovery or payback component (final
peak rate) when residential demand is at its peak. These percentages can payback tariff) for the first day of the first year is shown in Fig. 33(b) and
be even bigger if the variation range is considered. So, if planned well, Fig. 33(c), respectively. The payback tariff is low because no adjustment

Fig. 38. Day-ahead profiles: (a) ND_RTP from 67th day (to be added to the day-ahead electricity pricing profile of 68th day), (b) Day-ahead electricity pricing profile
of 68th day.

3239
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Fig. 39. RTP components (Year-1, Day-68): (a) RTP component of 68th, (b) ND_RTP of 68th day (to be added to the day-ahead electricity pricing profile of 69th day).

Fig. 40. Dynamic Pricing components (Year-1, Day-68): (a) dynamic electricity pricing, (b) dynamic demand charge, (c) dynamic payback tariff.

Fig. 41. Price variation range for the first day (Year-1, Day-68): (a) day-ahead, (b) actual real time billing price inside the range.

component is available on the first day of the first year. This price is that can be fitted under the available power curves will be different in
fixed for all the slots of the day. All three of these components are added some slots. The comparison of the numbers of chargers that can be fitted
to form the final dynamic pricing profile, which is illustrated in Fig. 34. under available power curves is shown in Fig. 37.
This pricing profile is then used for the final billing. Based on the number of chargers that can be fitted under the day-
ahead forecasted available power curve, the day-ahead electricity pric-
3.5. Day-ahead price variation range ing profile of the day is formulated. In this day-ahead profile, the left-
over from the previous day’s RTP component is also added, which is
Fig. 35(a) illustrates the day-ahead price variation range for the first denoted as “ND_RTP” (shown in Fig. 38(a)). The day-ahead electricity
day of the starting year, whereas Fig. 35(b) shows the actual real-time pricing profile for the 68th day is shown in Fig. 38(b). Based on the
billing price, which remained within the range. number of chargers that can be fitted below the real-time forecasted
available power curve, the RTP components for 68th (RTP) and 69th
3.6. Case study (Day-68, Year-1) (ND_RTP) days are created in real-time. Fig. 39 displays the RTP com-
ponents that are generated on the 68th day. Fig. 39(a) displays the RTP
The formation of dynamic pricing for a day other than the first day is component for the 68th day, which will be adjusted to the final elec-
shown in this case study. For this case study, the 68th day of the 1st year tricity pricing profile, whereas Fig. 39(b) shows the ND_RTP component
is chosen randomly. The formation process of the dynamic pricing that will be added to the day-ahead electricity pricing profile of the 69th
profile is illustrated with all necessary components. The forecasted and day.
the real-time demand and available power curves of the considered After combining the RTP component for the 68th day (shown in
residential complex for the 68th day of the 1st year are shown in Figs. 36 Fig. 39(a)) with the day-ahead electricity pricing profile of the same day
(a) and 36(b). Since there are significant variations in real-time available (shown in Fig. 38(b)), the final dynamic electricity pricing is obtained
power curve compared to the forecasted one, the number of chargers (shown in Fig. 40(a)). This dynamic electricity profile is then combined

3240
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

with the dynamic demand charge (shown in Fig. 40(b)) and the dynamic they plan right. On the other hand, the utility can earn more than 2.4
payback tariff (shown in Fig. 40(c)) to form the final dynamic pricing times the existing peak rate when residential demand is at its peak. In
profile, which will be the actual billing price for the different slots of the the future, the pricing scheme will incorporate several additional sig-
68th day. Fig. 41(a) displays the day-ahead price variation range with nificant parameters, including EV congestion, battery utilization factor,
the base pricing (combination of Fig. 38(a) & Fig. 38(b)) for the 68th seasonal impact, customer classification, charging duration, and
day. The final dynamic pricing profile remains inside this variation charging location allocation. Additionally, the adjustment thresholds for
range as it was supposed to be (illustrated in Fig. 41(b)). daily, monthly, and yearly rates will be made dynamic and subject to
regular revisions, ideally on an annual basis. Lastly, additional empirical
4. Conclusions testing across different geographical regions will be carried out in future
works to validate its universal applicability and adaptability.
In this study, a dynamic pricing scheme is developed for centralized
EV charging stations on a residential feeder. The dynamic pricing CRediT authorship contribution statement
scheme has three components: the first one relates to the cost of energy
delivered to the EVs, the second one is responsible for the demand Tareq Aziz: Writing – review & editing, Writing – original draft,
charge, and the third one is for recovering the cost of EV charging station Visualization, Supervision, Project administration, Investigation,
(EVCS) infrastructure development and operation. In this scheme, the Conceptualization, Formal analysis, Methodology, Resources, Software,
existing static ToU pricing is restructured to a dynamic ToU profile Validation. Taskin Jamal: Writing – review & editing, Writing – orig-
based on the demand level of the residential feeder. The scheme reduces inal draft, Visualization, Validation, Supervision, Methodology, Inves-
the price when the demand level is low and increases the price when tigation, Formal analysis. Md. Abu Taseen: Visualization, Data
demand is high to address the high generation cost during peak hours. curation, Formal analysis, Resources. Silvia Tasnim: Visualization,
The pricing policy offers a day-ahead dynamic plan for ease of sched- Data curation, Formal analysis, Resources. Syed Abdullah-Al-Nahid:
uling. The customers can schedule their vehicles based on their Writing – original draft, Investigation, Formal analysis, Data curation,
preferred time slot and day-ahead pricing. The day-ahead pricing also Resources, Visualization, Writing – review & editing. Tafsir Ahmed
comes with a price variation range, which brackets the dynamic prices of Khan: Writing – original draft, Visualization, Validation, Software,
EV charging to assure the customers that the price will not go outside the Methodology, Investigation, Formal analysis, Data curation, Conceptu-
range, which can make decision-making a lot easier. The pricing is alization, Resources, Writing – review & editing.
calculated assuming that all available charging slots of the EVCS will be
utilized. The price components of this dynamic pricing scheme are Declaration of Competing Interest
optimized by a PSO algorithm to maximize or minimize different pa-
rameters so that the tariff becomes minimum provided the tariff satisfies The authors declare that they have no known competing financial
the constraints of minimum revenue collection. This confirms that the interests or personal relationships that could have appeared to influence
proposed dynamic pricing scheme preserves the interests of all parties the work reported in this paper.
involved. The minimum tariff benefits the EV owners whereas the con-
straints ensure that the EVCS recovers the revenue that is equal to or Data Availability
more than with existing static ToU or flat-rate tariff. In simulation, it is
found that, an EV owner can save more than 32 % on charging bills if No data was used for the research described in the article.

Appendix

Table A.1
Considered simulation parameters.

Parameter Value or Name

Type of EV charger DC Fast Charger (DCFC)

Number of DC Fast Charger in EVCS 3
Considered electric vehicle Tesla Model 3 (2021) standard
Power rating of each charger 50 kW
Maximum capacity of the feeder or the transformer 200 kW
Number of charging slots in a day 24
Additional critical peak price Tk. 10
Maximum real time variation from day-ahead demand ±10 %
Slow charging threshold 80 %
Acquisition cost Tk. 1000000 (one time)
Operation & Maintenance cost Tk. 170000 (1st year)
Land rent Tk. 90000 (1st year)
Expected payback period 9 years
Profit margin (total profit after the payback period ends) 100 % in 9 years

References International Energy Agency, “Transport – topics,” IEA. [Online]. Available: https://
www.iea.org/topics/transport. [Accessed: 18-Feb-2023a].
International Energy Agency, World energy outlook 2021 – analysis,” IEA. [Online].
Li, X., Xu, C., Meng, J., 2022. Dynamic nonlinear impacts of Fossil Energy on renewable
Available: https://www.iea.org/reports/world-energy-outlook-2021. [Accessed: 18-
energy stocks: a quantile perspective. Energy Rep. 8, 15511–15523. https://doi.org/
Feb-2023a].
10.1016/j.egyr.2022.11.129.

3241
T.A. Khan et al. Energy Reports 12 (2024) 3218–3242

Handayani, K., Overland, I., Suryadi, B., Vakulchuk, R., 2023. Integrating 100% Cao, G., Fang, D., Wang, P., 2021. The impacts of social learning on a real-time pricing
renewable energy into electricity systems: A net-zero analysis for Cambodia, Laos, scheme in the electricity market. Appl. Energy vol. 291, 116874. https://doi.org/
and Myanmar. Energy Rep. 10, 4849–4869. https://doi.org/10.1016/j. 10.1016/j.apenergy.2021.116874.
egyr.2023.11.005. M.B. Berkow, M. Centore, K. Mohlin, and B. Spiller, A primer on time-variant electricity
International Energy Agency, Net zero by 2050 – analysis,” IEA. [Online]. Available: pricing,” Environmental Defense Fund. [Online]. Available: https://www.edf.org/
https://www.iea.org/reports/net-zero-by-2050. [Accessed: 18-Feb-2023b]. sites/default/files/a_primer_on_time-variant_pricing.pdf. [Accessed: 18-Feb-2023].
Abdullah-Al-Nahid, S., Jamal, T., Aziz, T., Bhuiyan, A.H., Khan, T.A., 2023. Additive Newsham, G.R., Bowker, B.G., 2010. The effect of utility time-varying pricing and load
linear modelling and genetic algorithm based electric vehicle outlook and policy control strategies on residential summer peak electricity use: A Review. Energy
formulation for decarbonizing the future transport sector of Bangladesh. Transp. Policy vol. 38 (7), 3289–3296.
Policy vol. 136, 21–46. Utility rate design initiative: Analysis Narative,” Alliance to Save Energy, 12-May-2016.
Lee, K., 2023. Bold fiscal policies for a net-zero EU: Promotion of electric vehicles and [Online]. Available: https://www.ase.org/sites/ase.org/files/rdi_analysis_narrative.
expansion of green energy. Energy Rep. 10, 2944–2949. https://doi.org/10.1016/j. pdf. [Accessed: 18-Feb-2023].
egyr.2023.09.138. Dutta, G., Mitra, K., 2017b. A literature review on dynamic pricing of electricity. J. Oper.
Achariyaviriya, W., Suttakul, P., Phuphisith, S., Mona, Y., Wanison, R., Phermkorn, P., Res. Soc. vol. 68 (10), 1131–1145.
2023. Potential reductions of CO2 emissions from the transition to electric vehicles: Bejan, I., Jensen, C.L., Andersen, L.M., Hansen, L.G., 2021. Inducing flexibility of
Thailand’s scenarios towards 2030. Energy Rep. 9, 124–130. https://doi.org/ household electricity demand: The overlooked costs of reacting to dynamic
10.1016/j.egyr.2023.08.073. incentives. Appl. Energy vol. 284, 116283. https://doi.org/10.1016/j.
Zhou, K., Cheng, L., Lu, X., Wen, L., 2020. Scheduling model of electric vehicles charging apenergy.2020.116283.
considering inconvenience and dynamic electricity prices. Appl. Energy vol. 276, A. Hove and D. Sandalow, “Electric vehicle charging in China and the United States -
115455. https://doi.org/10.1016/j.apenergy.2020.115455. Center on global energy policy at Columbia University: SIPA,” Center on Global
Dong, X., et al., 2018. A charging pricing strategy of electric vehicle fast charging stations Energy Policy at Columbia University | SIPA, Feb-2019. [Online]. Available: https://
for the voltage control of electricity distribution networks. Appl. Energy vol. 225, www.energypolicy.columbia.edu/publications/electric-vehicle-charging-china-and-
857–868. https://doi.org/10.1016/j.apenergy.2018.05.042. united-states/. [Accessed: 18-Feb-2023].
Danial, M., Azis, F.A., Abas, P.E., 2021. Techno-economic analysis and feasibility studies Zethmayr, J., Kolata, D., 2019. Charge for less: An analysis of hourly electricity pricing
of electric vehicle charging station. World Electr. Veh. J. vol. 12 (4), 264. for Electric Vehicles. World Electr. Veh. J. vol. 10 (1), 6.
Khan, T.A., Aziz, T., 2022. Developing A day-ahead dynamic pricing scheme for charging Muratori, M., Kontou, E., Eichman, J., 2019. Electricity rates for electric vehicle direct
electric vehicles in Bangladesh,” 2022 In. ternational Conf. Green. Energy, Comput. current fast charging in the United States. Renew. Sustain. Energy Rev. vol. 113,
Sustain. Technol. (GECOST), Miri Sarawak, Malays. (Virtual) 440–445. https://doi. 109235.
org/10.1109/GECOST55694.2022.10010568. Y. Guo, X. Liu, Y. Yan, N. Zhang and W. Su, “Economic analysis of plug-in electric vehicle
Amin, A., Tareen, W.U., Usman, M., Ali, H., Bari, I., Horan, B., Mekhilef, S., Asif, M., parking deck with dynamic pricing,” 2014 IEEE PES General Meeting | Conference &
Ahmed, S., Mahmood, A., 2020. A review of optimal charging strategy for electric Exposition, National Harbor, MD, USA, 2014, pp. 1-5.
vehicles under dynamic pricing schemes in the distribution charging network. Emodi, N.V., Dwyer, S., Nagrath, K., Alabi, J., 2022. Electromobility in Australia: tariff
Sustainability vol. 12 (23), 10160. design structure and consumer preferences for Mobile Distributed Energy Storage.
Mehta, V.K., Mehta, R., 2006. “Tariff,”. Principles of Power System. S. Chand & Sustainability vol. 14 (11), 6631.
Company, pp. 87–100. Korea Electric Power Corporation, “Electric Rates Table: Electric Vehicle Charging
Yang, Y., Wang, M., Liu, Y., Zhang, L., 2018. Peak-off-peak load shifting: Are public Tariff,” Kepco, 01-Jan-2021. [Online]. Available: https://home.kepco.co.kr/kepco/
willing to accept the peak and off-peak time of use electricity price? J. Clean. Prod. EN/F/htmlView/ENFBHP00108.do?menuCd=EN060201. [Accessed: 18-Feb-2023].
vol. 199, 1066–1071. Abdullah-Al-Nahid, S., Khan, T.A., Taseen, M.A., Jamal, T., Aziz, T., 2022. A novel
Dutta, G., Mitra, K., 2017a. A literature review on dynamic pricing of electricity. J. Oper. consumer-friendly electric vehicle charging scheme with vehicle to grid provision
Res. Soc. vol. 68 (10), 1131–1145. supported by genetic algorithm based optimization (Jun). J. Energy Storage vol. 50,
Tariff Rate: Effect from March 2023”, Bangladesh Power Development Board- 104655 (Jun).
Government of the People’s Republic of Bangladesh, https://bpdb.gov.bd (accessed Ghosh, A., Aggarwal, V., 2018. Menu-based pricing for charging of electric vehicles with
Jun. 18, 2023. vehicle-to-grid service. IEEE Trans. Veh. Technol. vol. 67 (11), 10268–10280.
Tariff design, Greening the Grid. [Online]. Available: https://www.greeningthegrid.org/ P. You, Z. Yang, M.-Y. Chow, and Y. Sun, “Optimal Cooperative Charging Strategy for a
electric-vehicle-toolkit/electric-vehicle-building-blocks-guidebook/tariff-design smart charging station of electric vehicles,” 2017 IEEE Power & Energy Society
[Accessed: 18-Feb-2023]. General Meeting, 2017.
Limmer, S., 2019. Dynamic pricing for Electric Vehicle Charging—a literature review. Daneshzand, F., Coker, P.J., Potter, B., Smith, S.T., 2023. Ev smart charging: how tariff
Energies vol. 12 (18), 3574. selection influences grid stress and carbon reduction. Appl. Energy vol. 348, 121482.
Ansarin, M., Ghiassi-Farrokhfal, Y., Ketter, W., Collins, J., 2020. The economic https://doi.org/10.1016/j.apenergy.2023.121482.
consequences of electricity tariff design in a renewable energy era. Appl. Energy vol. Richardson, D.B., 2013. Encouraging vehicle-to-grid (V2G) participation through
275, 115317. premium tariff rates. J. Power Sources vol. 243, 219–224.
Fridgen, G., Kahlen, M., Ketter, W., Rieger, A., Thimmel, M., 2018. One rate does not fit Meng, F., Ma, Q., Liu, Z., Zeng, X.-J., 2023. Multiple dynamic pricing for demand
all: An empirical analysis of electricity tariffs for residential microgrids. Appl. Energy response with adaptive clustering-based customer segmentation in smart grids. Appl.
vol. 210, 800–814. https://doi.org/10.1016/j.apenergy.2017.08.138. Energy vol. 333, 120626. https://doi.org/10.1016/j.apenergy.2022.120626.
Aguilar-Dominguez, D., Dunbar, A., Brown, S., 2020. The electricity demand of an EV Nicholas, M., 2019. Estimating electric vehicle charging infrastructure costs across major
providing power via vehicle-to-home and its potential impact on the grid with U.S. metropolitan areas” (Aug). Int. Counc. Clean. Transp. (ICCT), Work. Pap. (Aug).
different electricity price tariffs. Energy Rep. 6, 132–141. https://doi.org/10.1016/j. Abdullah-Al-Nahid, S., Khan, T.A., Taseen, M.A., Aziz, T., 2020. A consumer-friendly
egyr.2020.03.007. electric vehicle charging scheme for residential consumers. 2020 Int. Conf. Smart
Australian consumers’ likely response to cost- reflective electricity …” [Online]. Grids Energy Syst. (SGES), Perth, Aust. (Virtual) 893–897. https://doi.org/10.1109/
Available: https://publications.csiro.au/rpr/download?pid=csiro: SGES51519.2020.00164.
EP152667&dsid=DS2. [Accessed: 18-Feb-2023]. Tesla Model 3 (2021) charging guide,” Pod Point, 19-Apr-2022. [Online]. Available:
Shaolun Zeng, J. Li, and Yulong Ren, “Research of time-of-use electricity pricing models https://pod-point.com/guides/vehicles/tesla/2021/model3. [Accessed: 18-Feb-
in China: A survey,” 2008 IEEE International Conference on Industrial Engineering 2023].
and Engineering Management, 2008.

3242