Received October 20, 2020, accepted November 2, 2020, date of publication November 9, 2020, date of current version November

19, 2020.
Digital Object Identifier 10.1109/ACCESS.2020.3036660

Techno-Economic and Sizing Analysis of Battery

Energy Storage System for Behind-the-Meter
Application
CHIH-TA TSAI , ERICA M. OCAMPO , TEKETAY MULU BEZA , AND CHENG-CHIEN KUO
Department of Electrical Engineering, National Taiwan University of Science and Technology, Taipei 10607, Taiwan
Corresponding author: Cheng-Chien Kuo ([email protected])

ABSTRACT As the cost of the battery energy storage system (BESS) is lower, the penetration rate of
battery storage is rising in the behind-the-meter (BTM) market. BESS with time-of-use rates (TOU) for
charge and discharge scheduling can be used to reduce electricity costs. This research uses 6,600KW contract
capacity for industrial customers as the study case. Through the BESS techno-economic simulation in Hybrid
Optimization Models for Energy Resources (HOMER) Grid software, the optimal capacity planning can be
obtained from the electricity demand. The results show that the contract capacity can be adjusted to 6,100kW
with a project benefit of $68,557, levelized cost of energy (LCOE) of 0.09338 $/kWh, net present cost (NPC)
of -$24,768,508, internal rate of return (IRR) of 3.72%, return of investment (ROI) of 2.16%, and discounted
payback of 8.37 years in the 500kW/1065kWh required BESS capacity case. The sensitivity of the electricity
price, key components cost, and real interest rate are evaluated and the method will be applied as a reference
for planning the BTM BESS of other utility customers.

INDEX TERMS Techno-economic analysis, battery energy storage, energy arbitrage, peak shaving, BTM
application.

I. INTRODUCTION used. In FTM application, the energy storage system offers

Battery energy storage (BESS) has the potential to improve bulk energy services, ancillary services, grid support, and
electric power grid performance, stability, and resilience. renewable energy integration [4], [5]. A survey shows that
Nowadays, electric power generation must always be in the BTM application accounts for 30% of global Li-ion
a dynamic balance with load demand. With the growing BESS market in 2019 [6] and it mainly provides customer
role of variable power generation and rising load demand energy management services to utility customers [7]–[10].
profiles, grid operators still have inadequate resources to The incentive includes energy storage system and time-of-
sustain this dynamic balance. BESS, if properly deployed, use rates(TOU) in order to decrease electricity expenses,
can provide the grid operators a flexible fast-responding increasing self-consumption of distributed generation, reduc-
reserve to efficiently manage the variation from generation ing electricity demand, offering emergency power supply,
and demand [1], [2]. Installing BESS so that load demand can and increasing power stability and demand response. It also
be supplied whenever needed would signify a major improve- brings benefits to utilities [11], [12]. For example, the load
ment in electricity generation and transmission. Modern factor of utilities can be improved and the variation between
developments and progress in energy storage and power elec- on-peak and off-peak can also be decreased. The integration
tronics technologies are making the application of BESS a of renewable energy and BESS from the user terminal can
more feasible solution for increasing flexibility and improv- reduce the impact of intermittent output of renewable energy
ing reliability of power delivery [3]. on the grid directly.
As the cost of BESS is lower, the front-the-meter (FTM) Related literature that describes the feasibility of using
and the behind-the-meter (BTM) applications are widely BESS for the FTP and BTM application is presented as
follows. Zurfi et al. [13] analyses the residential appli-
The associate editor coordinating the review of this manuscript and cations where behind the meter battery energy storage is
approving it for publication was Zhiyi Li . installed at the customer’s locations and used for daily cycling

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
203734 VOLUME 8, 2020
C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

under TOU and demand charge plans for saving on monthly the energy storage system and analyze the most economical
electricity bills. Trovato et al. [14] tries to compare BESS BESS capacity in different contract capacity conditions. The
connected BTM with a large scale renewable source for opti- proposed methods and discussions are as follows:
mal delivery of energy to maximize profits. In [15], [16] an • A survey has been taken of the power consumption of
agent-based model for techno-enviro-economic assessment industrial customers in 2019, and statistics on Taiwan’s
of household and community BESS is proposed. According interest rates, inflation rates, TOU rates, and various
to the study, both household BESS and community BESS equipment costs, etc. were compiled as basic data for
can significantly minimize the grid peak power import and simulation analysis.
export. On the other hand, Roberts et al. [17] discussed the • According to the load conditions of industrial cus-
impact of shared BESS on photovoltaic self-consumption tomers with no installation of the energy storage system,
and monthly electricity costs in apartment buildings. The the electricity cost expenditure is regarded as a bench-
study indicates that central battery energy storage of 2-3kWh mark for benefit comparison.
per apartment building can increment solar self-consumption • On the conditions of load, TOU rates, and meeting the
up to 19%. Some studies also indicated that the decreases electricity demand, the BESS capacity allocation, elec-
in the cost of PV and battery energy storage encouraged trical and economic results were simulated for different
utility customers to invest in PV battery prosumage [18]–[21]. contract capacities.
Pandzic et al. [22] tried to develop a model to govern the • The BESS capacity with the highest project benefit was
optimal energy and power capacity of a stationary BESS to obtained from the results mentioned above.
reduce electricity bills. Papers related to peak shaving and • From the economical contract capacity and BESS capac-
financial viability of storage systems had been discussed ity, the research evaluates the sensitivity of the electric-
in [23]–[27]. ity price, key components cost, and real interest rate.
Having the literature discussed above in mind; Finally, the results can be used to analyze the degree of
• This study uses HOMER Grid software to propose eco- impact on economic benefits
nomic benefit evaluation methods and procedures for
applying BESS to behind-the-meter, so that can be used A. SIMULATION SOFTWARE DESCRIPTION
in practical situations. HOMER Energy LLC is one of the world’s leading
• The analysis of economic evaluation indicators used in distributed generation and micro-grid modeling software
this study: levelized cost of energy (LCOE), net present companies. The software was originally developed by
cost (NPC), internal rate of return (IRR), return of invest- NREL (National Renewable Energy Laboratory), which was
ment (ROI), discounted payback period (DPP), and con- acquired by Underwriter Laboratories Inc. in December
duct sensitivity analysis, the results can be used as a 2019. HOMER Grid is one of the HOMER Energy anal-
reference basis for investors to make decisions. ysis software series. HOMER Grid combines engineering
• After evaluation and analysis, through the charging and and economics information in a holistic model. It quickly
discharging adjustment of the BESS, in this case study executes complex calculations to compare numerous compo-
and Taiwan’s electricity price plan, it can effectively nents and design results, enable to select points at which dif-
reduce the electricity expenditure of users, so that users ferent technologies come to be cost-competitive, and consider
can obtain the minimized NPC, that is the largest project different options for reducing project risk and minimizing
benefit. energy expenses [28].
• The results through the HOMER Grid Software simula-
tion can be used to evaluate energy arbitrage revenue. B. LOAD PROFILE
This case study is for industrial customers located in
II. METHODOLOGY northern Taiwan. The period of the load profile is from
This research analyzes the economic benefits of BESS energy January 2019 to December 2019. The data was sampled
arbitrage, and the profit is mainly through the charging every 5 minutes for an annual electricity consumption
and discharging adjustment of BESS to reduce electricity of 27,045,579 kWh/year and an average electricity consump-
expenses. The method considers two main points: The first tion per month of 2,253,798 kWh/month. The order quan-
is following the TOU rates, to discharge during periods tity of the factory is different every month, so the monthly
of high electricity prices, and to charge during periods of electricity consumption is also different. Fig. 1 shows that
low electricity prices to obtain the benefits from the elec- the highest power demand is 2,488,690 kWh/month in May.
tricity price difference. The second is during the period of Because Taiwan New Year is in February, the production
high power consumption, through BESS discharge, reducing will be the least, that is why the electricity consumption of
the user demand for electricity purchase during the period, the month is minimum, which is 1,786,387 kWh/month. The
thereby reducing the annual contractual capacity charge. This daily load profile is shown in fig. 2 wherein the on-peak time
research uses an industrial customer as a test case. HOMER is from 8 AM to 12 PM and 1 PM to 4 PM. The annual average
Grid software is used to simulate the techno-economic of power consumption is 74,098 kWh/day, the maximum power

VOLUME 8, 2020 203735

 C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

TABLE 1. Types of Electricity Price and the Application in Taiwan.

FIGURE 1. The power demand of each month in 2019.

FIGURE 2. Daily load profile.

FIGURE 4. The electricity expenses of each month in 2019.

corresponding to different contract capacity costs, and two

or three stages of TOU rate can be selected. Accord-
ing to TPC regulations, if the average power consumption
FIGURE 3. Load power consumption frequency distributio. is below 10% of the contract capacity every 15 minutes,
the basic electricity charge should be doubled. If it exceeds
10% of the contract capacity, the basic electricity charge
intake is 6,579kW, and the average power consumption is should be tripled. As shown in Table 2, this research
3,087kW. The proportion of the power consumption fre- case is the fifth type, with an extra-high voltage (EHV)
quency in the range of 1,750kW∼3,300kW which is 55.55%, of 69kV, a three-stage TOU rate, and a contract capacity
as shown in fig. 3. The frequency of 2,750kW is the highest, of 6,600kW.
is 13.09%. The usage frequency of 6,000kW is less than The contract capacity pricing for basic electricity in sum-
1%, so it can provide short-term electricity demand through mer months is 1.35 times that of non-summer months.
the energy storage system to reduce the contract capacity The three-stage energy charge during the on-peak period is
expenditure. 3.57 times that of the off-peak period. In the non-summer
months, the mid-peak period charge is 2.28 times that of
C. CALCULATION METHOD OF ELECTRICITY CHARGE the off-peak period. The 2019 monthly electricity charge
Table 1 shows information about electricity charges from for this case study is as shown in fig. 4. The total annual
Taiwan Power Company (TPC). It can be divided into 5 types. electricity charge is 2,535,531$/year. The energy charge is
The second and third types have non-TOU rate options. The 2,069,581$/year accounting for 81.62% of the total annual
fourth and fifth types are based on different voltage levels, electricity charge, while the remaining 18.38% is from the

203736 VOLUME 8, 2020

C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

TABLE 2. Three-Stage Energy Charge of Extra High Voltage.

contract capacity charge with 465,949$/year. The highest and cycle.

lowest months of electricity charges are in July and February,
CF 1 CF 2 CF 3
respectively, wherein the highest is 1.78 times the lowest. NPC = −CF 0 + { + +
The electricity consumption throughout the year meets the (1 + i) 1
(1 + i) 2
(1 + i)3
contractual capacity specification. CF N
+ ... +
(1 + i)N
D. ECONOMIC ASSIGNMENT CRITERIA XN CF t
= −CF 0 + (2)
The descriptions and formulas of the economic evaluation t=1 (1 + i)t
indicators used in this study are as shown in subsections
where CF t is the cash flow of the t-year (In the definition of
1 ∼ 8 [28], respectively.
the HOMER Grid software: the expenditure is negative and
the income is positive.) ($); i is the annual real interest rate
1) ANNUAL REAL INTEREST RATE
(%); N is the project lifetime (year); t is the number of years
Annual real interest rate (i) is used to convert between (year); CF 0 is the initial capital cost ($).
one-time cost and annualized costs. HOMER Grid software
uses the annual real interest rate to compute discount factors 3) CAPITAL RECOVERY FACTOR
and to calculate annualized costs from present costs.
The capital recovery factor (CRF) is the equation used to
i0
−f compute the present value of the annuity during the project
i= (1) lifetime.
1+f
i(1 + i)t
where i is the annual real interest rate (%); i0 is the nominal CRF (i, t) = (3)
(1 + i)t − 1
interest rate (bank board rate) (%); f is the expected inflation
rate (%). where t is the number of years (year); i is the annual real
interest rate (%).
2) NET PRESENT COST
The total NPC value represents the cost of the system life 4) LEVELIZED COST OF ENERGY
cycle in HOMER Grid software. Equation (2) indicates the The definition of levelized cost of energy (LCOE) in this
summation of the cash flow of the t-year over the factor and study is the average electricity cost per kWh of load con-
the initial capital cost. The costs contain capital cost, oper- sumption. It is computed by dividing the total annualized
ation cost, replacement cost, maintenance cost, etc. Income cost (TAC) by the total annualized load consumption. The
contains electricity selling and salvage value after the life LCOE unit is $/kWh. The TAC is the annualized value of

VOLUME 8, 2020 203737

 C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

NPC, and its unit is $/year. The equation is as follow:

TAC = |NPC| × CRF(i, N ) (4)

TAC
LCOE = (5)
Eprim

where Eprim is the total annualized load consumption

(kWh/year); N is the project lifetime (year).

5) PROJECT BENEFIT
The project benefit is the difference between the NPC of
the current case and the reference case, and its unit is in $.
The sign of the project benefit describes whether the current
case compares favorably as an investment option with the
reference case: a positive value indicates that the current FIGURE 5. Schematic diagram of the grid-connected battery energy
system saves money over the project lifetime compared to the storage system.

base case system. The equation is as follow:

8) INTERNAL RATE OF RETURN
XN CF c,t The internal rate of return (IRR) is the discount rate where the
PB = (−CF c,0 + )
t=1 (1 + i)t present value of cash flow is exactly equal to the initial invest-
XN CF ref ,t
− (−CF ref ,0 + ) ment amount. The IRR is an indicator of the profitability of
t=1 (1 + i)t a series of cash flows. Therefore, according to (6), the IRR
= NPC c − NPC ref (6) equation is defined as shown in (9).

where, NPC c is the NPC of the current case ($); NPC ref is
XN CF c,t − CF ref ,t
the NPC of the reference case, which is a scenario without a 0 = −CF c,0 + CF ref ,0 + (9)
t=1 (1 + IRR)t
battery energy storage system installed ($).
III. GRID-CONNECTED BATTERY ENERGY STORAGE
6) RETURN ON INVESTMENT SYSTEM DESCRIPTION
The return on investment (ROI) is the yearly cost savings A. GRID-CONNECTED BATTERY ENERGY STORAGE
relative to the initial investment. The software computes the SYSTEM SCHEMATIC
return on investment using (7). Fig. 5 is the schematic diagram of grid-connected BESS
and it consists of a grid storage system power conversion
ROI" # system (PCS) and load. The power demand of the load is
PN PN provided by the grid. The energy storage system charges at
(−CF c,0 + t=1 CF c,t )−(−CF ref ,0 + t=1 CF ref ,t )
= off-peak and discharges at on-peak according to the TOU rate,
N ×(Ccap,c −Ccap,ref )
in order to reduce the user’s electricity expenses.
× 100% (7)
B. STORAGE SYSTEM
where Ccap,c is the capital cost of the current case ($); Ccap,ref
The research used lithium ferrous phosphate (LFP) with high
is the capital cost of the reference case ($).
energy density and safe to analyze economic benefits. The
rated specification of the battery cell is 3.2V/280Ah, and
7) DISCOUNTED PAYBACK PERIOD
the continuous charging/discharging current is 140 A. The
The discounted payback period (DPP) is the measure of the battery module is made up of 14 battery cells connected in
time it takes for the cumulative discounted net cash flow to series, the rated specification is 44.8V/280Ah. The battery
offset the initial investment in the asset or project. Its unit is rack is composed of 17 battery modules connected in series,
the year. The equation is as shown in (8). the rated specification is 761.6V/280Ah, the rated capacity
is 213.2kWh, and the working voltage is between 666.4V
Ccap,last
DPP = tfull + CF c,tfull +1 CF ref ,tfull +1
(8) and 856.8V. The minimum state of charge (SOC) is 20% and
t +1 − tfull +1 the round trip efficiency is 92%. The C rate is 0.5C and the
(1+i) full (1+i)
lifetime throughout is 967,765kWh.
where tfull is the time before the accumulated cash flow
becomes positive (year). Ccap,last is the remaining unre- C. POWER CONVERSION SYSTEM
covered capital cost after accumulated cash flow becomes The power conversion system (PCS) used is a bidirectional
positive ($). converter (DC to AC and AC to DC). In this simulation

203738 VOLUME 8, 2020

C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

FIGURE 6. Flowchart of BESS operation.

analysis, the conversion efficiency was set to 97.3%, while if the number of battery rack is 1 or 2, the price per kWh
the working life was set to 10 years. is 340 $/kWh, which can be installed in a 10-foot con-
tainer. If the number of battery rack is between 3 and 6,
D. SYSTEM DISPATCH STRATEGY the price per kWh is 290 $/kWh, which can be installed in
The BESS charge and discharge control method used in this a 20-foot container. If the number of battery rack is between
study adopts the default mode of HOMER Grid software. The 7 and 10, the price per kWh is 280 $/kWh, which can
BESS operation flow is shown in fig. 6 and uses the state of be installed in a 40-foot container. Each container includes
charge (SOC) of BESS as the basis for the charge and dis- DC and AC switchboards, firefighting facilities, air condi-
charge judgment. If the SOC is greater than or equal to 20% tioning systems, lighting and energy management systems,
and the load demand is greater than or equal to the constraint etc.
of contract capacity, BESS will immediately discharge. If this In the PCS, the prices are 186 $/kW, 123.2 $/kW, 100 $/kW,
condition is not met, BESS discharges during the period of and 81.57 $/kW in 100 kW, 250 kW, 500 kW, and 700kW,
high electricity prices. When the SOC is less than 20%, BESS respectively. In the simulation analysis of HOMER Grid, the
stops the discharge mode and charges during periods of low capacity of PCS can be derated so that it can correspond to
electricity prices. different capacity intervals. The BESS operation and mainte-
nance (O&M) cost was set to 4.25 $/kWh/year. The price of
IV. COMPONENT COST AND FINANCIAL ASSUMPTION BESS includes transportation, installation, and miscellaneous
A. SYSTEM COMPONENT COST expenses. The project lifetime of the simulation was set to
The information on component cost was taken from Tai- 10 years, and the residual value of the equipment was not
wan system integration company. In the battery system, included.

VOLUME 8, 2020 203739

 C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

TABLE 3. Annual NPC Without BESS Installed.

FIGURE 7. Interest rate information for the Taiwan from January 2015 to
December 2019.

HOMER Grid software. The results can be used as the ref-

erence for the subsequent analysis of the economic benefit
evaluation after installing energy storage system. As depicted
in table 3, the total NPC for 10 years is -$24,837,065,
which is the total electricity cost for 10 years. The LCOE is
0.09376 $/kWh.

B. THE CONDITION WITH BESS INSTALLATION

Energy arbitrage was obtained through charge and discharge
management of BESS. The user’s contract capacity was
reduced by using the off-peak electricity price to charge,
FIGURE 8. Inflation rate information for the Taiwan from January 2015 to and the on-peak electricity price to discharge in that way
December 2019. the total electricity cost can also be reduced. This simulation
analyzed the BESS capacity configuration with the highest
B. INTEREST RATE AND INFLATION RATE project benefit of different contract capacities from 6,500kW
Fig. 7 shows the statistics of interest rate data publicized by to 5,900kW and the optimal contract capacities scheme.
the Central Bank of the Republic of China (Taiwan) in the First, the BESS capacity combination of the NPC closest
past 5 years. The interest rate for the one-year deposit in to the positive value which represents the least expenditure
December 2019 is 1.04%. The average of 5 years is 1.11%. was determined. Then, the contracted capacity from each
The maximum and minimum are 1.36% and 1.04%, respec- level was compared to find the contracted capacity with
tively. Fig. 8 shows the inflation rate information for Taiwan the highest project benefit after the reduction and the cor-
from January 2015 to December 2019 announced by the responding BESS capacity allocation. The simulation result
Directorate General of Budget, Accounting, and Statistics, is as shown in table 4. The best scheme was found to be
Executive Yuan, R.O.C. The inflation rate is 1.14% with 6,100 kW contract capacity, 500kW/1065kWh BESS capac-
a 5-year average of 0.73%. The maximum and minimum ity which can get the highest project benefit of $68,557 and
inflation rate are 2.41% and -0.94%, respectively. It can be the lowest LCOE of 0.09338 $/kWh, with -$24,768,508 NPC,
seen that the commodity price in Taiwan is long-running 3.72% IRR, 2.16% ROI, 8.37 years for discounted payback,
stable. In this paper, the simulation used the average value $24,365,321 electricity charge reduction, and 1.9% total elec-
of the statistical period as the analysis parameter. The real tricity charge saved. The total installation cost of BESS is
discount rate calculated by (1) is 0.38%. $358,850, of which the cost of the storage system and PCS are
$308,850 and $50,000, accounting for 87.07% and 13.93%,
V. SIMULATION RESULTS AND DISCUSSION respectively. The cost for 10 years of O&M is $44,337.
A. THE CONDITION WITHOUT BESS INSTALLATION As shown in fig. 9, the larger the capacity of BESS, the
According to the simulation parameters, the 10-year elec- more electricity costs can be saved. However, if the cost
tricity cost without installing BESS was analyzed through of BESS equipment and maintenance increases, the project

203740 VOLUME 8, 2020

C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

TABLE 4. Result of Economic Characteristics in Different Contract Capacity.

benefit decreases. Take the contract capacity of 5,900 kW as TABLE 5. Sensitivity Results of Electricity Price.
an example, the utility power capacity is limited to 5,900 kW
so that it can ensure stable operation. In this condition, the
PCS capacity of BESS is 700kW, and the battery capacity
must be increased to 2,130kWh. The investment cost of
equipment in the early stage was greatly increased, which is
1.84 times higher than 500kW/1065kWh. The electricity cost
only saved about 1%, while the project benefit is −$9,362,
which is not in line with economic benefits. Another exam-
ple is when the contract capacity is increased to 6,500 kW.
Although the capacity of BESS combines 100kW/213 kWh
is the lowest for the energy arbitrage, the project benefit is
also less than the installation and maintenance costs of BESS.
Therefore, it is not in line with economic benefits.
Using BESS 500kW/1065kWh with a summer electricity
consumption charge and discharge on June 17 and 18, as an
example, and the limit of electricity consumption of 6,100kW.
The lowest SOC of the energy storage battery was set to
20% giving a usable capacity of 852 kWh. The charging
starts at 22:30 (off-peak and low electricity price period)
on June 17. With a rating of 500 kW, the BESS was able
to fully charge in about 1.7 hours, as shown in fig. 10
(a). The discharge starts at 10:00 on June 18 during the
on-peak high electricity price to reduce the cost of electric-
ity. As shown in fig. 10 (b), power consumption exceeds charge was reduced to 2,487,372 $/year. The energy charge
6100kW at 9:45 and 10:15. In addition to satisfying the and contract capacity charge reduce to 2,056,722 $/year and
electricity demand, discharging of the BESS also avoided 465,949 $/year, respectively, that is a decrease of about 0.62%
penalties due to the electricity consumption exceeding the and 7.58%. The contract capacity charge decreased more than
contract capacity. When this system is compared with the the energy charge. The BESS has energy conversion loss of
system without the BESS installed, the annual electricity charging and discharging, so the power consumption of users

VOLUME 8, 2020 203741

 C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

FIGURE 9. The project benefit and electricity cost saving percentage in different contract capacity.

FIGURE 10. The charge and discharge operation for 500kW/1065kWh. a) BESS charge during off-peak electricity price, discharge during
on-peak electricity price and b) BESS discharge mode to reduce contract capacity.

rises to 27,081,633 kWh/year, which is a 0.13% increase. The economic benefits through the 500kW/1065kWh BESS sim-
main reason for power consumption rise is the loss of energy ulation with different electricity prices, BESS capital cost,
storage batteries of 21,754 kWh/year, and the loss of PCS and real interest rate.
of 14,300 kWh/year
A. ELECTRICITY PRICE
In this part, a sensitivity analysis was carried out by adjust-
VI. SENSITIVITY ANALYSIS ing the electricity price, which was done by comparing the
According to the analysis of results in the condition with situations wherein the electricity price was not adjusted for
BESS installation, this section discusses the impact on the 10 years, adjusted of 3% every 3 years, adjusted of 3% every

203742 VOLUME 8, 2020

C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

TABLE 6. Sensitivity results of bess capital cost.

2 years, and adjusted of 3% annually. From the sensitivity and PCS capital cost. Through the 10% change in each step
analysis in table 5, it can be seen that although NPC, LCOE, and for change range of 1-0.5 times, the impact on economic
BESS energy arbitrage, ROI, and IRR increased with the fre- benefit was analyzed. Fig. 11 shows that the capital cost of the
quent adjustment of electricity price, the discounted payback storage system is greater than that of PCS. By reducing stor-
shorten. In the case of adjusting 3% every year, the IRR is age system capital cost, the economic benefits obtained were
already higher than 5%, which means that the BESS ben- significantly higher than that of reducing PCS capital cost.
efit can be more prominent in the case of high electricity According to the analysis results of table 6, when the storage
prices. system’s capital cost decreased by 10%, the project benefit
increases by 45.05%. When the PCS capital cost decreased
B. BESS CAPITAL COST by 10%, the project benefit only increases by 7.29%. Finally,
The sensitivity analysis of this part is divided into three parts: when the total BESS capital cost decreased by 10%, the
including BESS total capital cost, storage system capital cost, project benefit increases by 52.34%, the IRR increases

VOLUME 8, 2020 203743

 C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

FIGURE 11. Effect of varying key components cost on the project benefit.

TABLE 7. Sensitivity Results of Real Interest Rate.

to 5.88%, and the discounted payback is shortened to VII. CONCLUSION

7.52 years. According to the analysis of the case study, the highest
economic benefit of the BESS capacity sizing and the cor-
C. REAL INTEREST RATE responding contract capacity can be calculated by using
In this simulation, the real interest rate, initially from 0.38%, HOMER Grid software. The conclusions of the simulation
was decreased to 0% and increased to 5%, with increments result are summarized as follow:
of 1%. From the analysis results of equations (1) to (5) and • Setting the load, electricity price data, BESS equip-
Table 7, it can be seen that the higher the real interest rate, ment cost, and limited conditions with BESS in the
the lower the present value of the cash flow in each future simulation as the comparison criteria and through the
year. Therefore, the real interest rate is inversely proportional HOMER Grid software simulation, the most economical
to the NPC and project benefit. LCOE and discounted pay- BESS capacity is 500kW/1065kWh with a correspond-
back increases when the real interest rate was increased. The ing contract capacity scheme of 6,100kW with a 10-year
analysis results of ROI and IRR can be obtained from (8) project benefit of $68,557, LCOE of 0.09338 $/kWh,
to (9), which are calculated using the annual cash flow IRR of 3.72%, ROI of 2.16%, and discounted payback
value without considering real interest rate, so these two of 8.37 years. When the capacity of BESS is increased,
values have nothing to do with the change in real interest the contract capacity can be more reduced, and more
rate. According to the results of Table 7, when the real electricity expenses can be saved.
interest rate is 4%, the NPC with the BESS installed is • According to the analysis of results of installing or
lower than that without the BESS installed, so the project not installing BESS, although the electricity cost can
benefit becomes negative and the payback period exceeds be reduced by installing BESS, the power consump-
10 years. tion increases because of the loss of energy conversion

203744 VOLUME 8, 2020

C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

between charging and discharging. In the 500 kW/ [11] International Renewable Energy Agency, ‘‘Innovation landscape brief:
1065 kWh BESS study case, the power consumption of Behind-the-meter batteries,’’ Irena, vol. 2019, p. 24, 2019.
[12] Y.-J. Kim, G. Del-Rosario-Calaf, and L. K. Norford, ‘‘Analysis and exper-
the user increased to 27,081,633 kWh/year which is an imental implementation of grid frequency regulation using behind-the-
increase of 0.13%. This increase in power consumption meter batteries compensating for fast load demand variations,’’ IEEE
includes the 21,754 kWh/year loss from energy storage Trans. Power Syst., vol. 32, no. 1, pp. 484–498, Jan. 2017.
[13] A. Zurfi, G. Albayati, and J. Zhang, ‘‘Economic feasibility of residential
batteries and 14,300 kWh/year loss from PCS. behind-the-meter battery energy storage under energy time-of-use and
• Using 500kW/1065kWh BESS for sensitivity analy- demand charge rates,’’ in Proc. IEEE 6th Int. Conf. Renew. Energy Res.
sis, when the price of electricity increases more often, Appl. (ICRERA), Nov. 2017, pp. 842–849.
[14] V. Trovato and B. Kantharaj, ‘‘Energy storage behind-the-meter with
the project benefits, ROI, and IRR of BESS energy arbi- renewable generators: Techno-economic value of optimal imbalance man-
trage increases, while the discounted payback becomes agement,’’ Int. J. Electr. Power Energy Syst., vol. 118, Jun. 2020,
shorter. Above all, it means that the BESS benefit can be Art. no. 105813.
[15] S. Dong, E. Kremers, M. Brucoli, R. Rothman, and S. Brown, ‘‘Techno-
more prominent in the case of high electricity prices. enviro-economic assessment of household and community energy stor-
• In the cost sensitivity analysis, the proportion of cap- age in the U.K.,’’ Energy Convers. Manage., vol. 205, Feb. 2020,
ital cost increase for energy storage system is more Art. no. 112330.
[16] S. van der Stelt, T. AlSkaif, and W. van Sark, ‘‘Techno-economic analysis
compared to the increase for the PCS. In terms of of household and community energy storage for residential prosumers with
the degree of economic benefit, the economic benefit smart appliances,’’ Appl. Energy, vol. 209, pp. 266–276, Jan. 2018.
obtained from reducing the storage system capital cost is [17] M. B. Roberts, A. Bruce, and I. MacGill, ‘‘Impact of shared battery energy
storage systems on photovoltaic self-consumption and electricity bills in
significantly higher than that of reducing the PCS capital apartment buildings,’’ Appl. Energy, vol. 245, pp. 78–95, Jul. 2019.
cost. When the storage system capital cost decreased by [18] M. Beuse, M. Dirksmeier, B. Steffen, and T. S. Schmidt, ‘‘Profitability
10%, the project benefit can increase by 45.05%. If the of commercial and industrial photovoltaics and battery projects in South-
East-Asia,’’ Appl. Energy, vol. 271, Aug. 2020, Art. no. 115218.
total BESS capital cost decreased by 10%, the project [19] P. Unahalekhaka and P. Sripakarach, ‘‘Reduction of reverse power flow
benefit can increase by 52.34%, the IRR can increase to using the appropriate size and installation position of a BESS for a PV
5.88%, and the discounted payback can be shortened to power plant,’’ IEEE Access, vol. 8, pp. 102897–102906, 2020.
[20] K. Say, W.-P. Schill, and M. John, ‘‘Degrees of displacement: The impact of
7.52 years. household PV battery prosumage on utility generation and storage,’’ Appl.
• In the sensitivity analysis of the real interest rate, Energy, vol. 276, Oct. 2020, Art. no. 115466.
the relationship between the real interest rate and NPC is [21] M. Fisher, J. Apt, and J. F. Whitacre, ‘‘Can flow batteries scale in the
behind-the-meter commercial and industrial market? A techno-economic
inversely proportional. As the real interest rate increases, comparison of storage technologies in California,’’ J. Power Sources,
project benefit decreases, LCOE increases, and dis- vol. 420, pp. 1–8, Apr. 2019.
counted payback extends. Therefore, the case study of [22] H. Pandžić, ‘‘Optimal battery energy storage investment in buildings,’’
Energy Buildings, vol. 175, pp. 189–198, Sep. 2018.
low real interest rates can easily confirm the economic [23] S. Dong, E. Kremers, M. Brucoli, R. Rothman, and S. Brown, ‘‘Improving
benefits of installing BESS. the feasibility of household and community energy storage: A techno-
enviro-economic study for the U.K.,’’ Renew. Sustain. Energy Rev.,
REFERENCES vol. 131, Oct. 2020, Art. no. 110009.
[1] R. H. Byrne, T. A. Nguyen, D. A. Copp, B. R. Chalamala, and I. Gyuk, [24] J. Mostert and B. Bekker, ‘‘Estimating financial viability of behind-the-
‘‘Energy management and optimization methods for grid energy storage meter peak shaving based on load profile shape: A shopping mall case
systems,’’ IEEE Access, vol. 6, pp. 13231–13260, 2018. study,’’ in Proc. Int. SAUPEC/RobMech/PRASA Conf., Jan. 2020, pp. 1–6.
[2] M. Uddin, M. F. Romlie, M. F. Abdullah, S. A. Halim, A. H. A. Bakar, and [25] T. Simpkins, K. Anderson, D. Cutler, and D. Olis, ‘‘Optimal sizing of a
T. C. Kwang, ‘‘A review on peak load shaving strategies,’’ Renew. Sustain. solar-plus-storage system for utility bill savings and resiliency benefits,’’
Energy Rev., vol. 82, pp. 3323–3332, Feb. 2018. in Proc. IEEE Power Energy Soc. Innov. Smart Grid Technol. Conf. (ISGT),
[3] D. Wu, M. Kintner-Meyer, T. Yang, and P. Balducci, ‘‘Economic analysis Sep. 2016, pp. 1–6.
and optimal sizing for behind-the-meter battery storage,’’ in Proc. IEEE [26] R. Martins, H. Hesse, J. Jungbauer, T. Vorbuchner, and P. Musilek, ‘‘Opti-
Power Energy Soc. Gen. Meeting (PESGM), Jul. 2016, pp. 1–5. mal component sizing for peak shaving in battery energy storage system
[4] I. Naidji, M. B. Smida, M. Khalgui, A. Bachir, Z. Li, and N. Wu, ‘‘Efficient for industrial applications,’’ Energies, vol. 11, no. 8, p. 2048, Aug. 2018.
allocation strategy of energy storage systems in power grids considering [27] A. B. Culaba, A. J. R. Del Rosario, A. T. Ubando, and J. S. Chang,
contingencies,’’ IEEE Access, vol. 7, pp. 186378–186392, 2019. ‘‘Optimal design of an integrated renewable-storage energy system in a
[5] R. Dufo-López and J. L. Bernal-Agustín, ‘‘Techno-economic analysis mixed-use building,’’ Int. J. Energy Res., pp. 1–13, Apr. 2020.
of grid-connected battery storage,’’ Energy Convers. Manage., vol. 91, [28] Homer Grid. (2020). [Online]. Available: https://www.homerenergy.
pp. 394–404, Feb. 2015. com/products/grid/docs/1.7/index.html
[6] Homer Energy. (2020). [Online]. Available: https://www.homerenergy.
com/index.html
[7] R. Irany, K. Aancha, G. Daniel, and G. Arwa, ‘‘Energy storage monitor—
Latest trends in energy storage,’’ World Energy Council, vol. FEL-100,
pp. 1–32, 2019.
[8] H. T. Dinh, J. Yun, D. M. Kim, K.-H. Lee, and D. Kim, ‘‘A home energy
CHIH-TA TSAI received the B.S. and M.S. degrees
management system with renewable energy and energy storage utilizing
in electrical engineering from Kun Shan Univer-
main grid and electricity selling,’’ IEEE Access, vol. 8, pp. 49436–49450,
2020.
sity and National Cheng Kung University, Taiwan,
[9] H. Hesse, R. Martins, P. Musilek, M. Naumann, C. Truong, and A. Jossen, in 2003 and 2005, respectively. He is currently pur-
‘‘Economic optimization of component sizing for residential battery stor- suing the Ph.D. degree with the National Taiwan
age systems,’’ Energies, vol. 10, no. 7, p. 835, Jun. 2017. University of Science and Technology. He was
[10] B. P. Bhattarai, K. S. Myers, and J. W. Bush, ‘‘Reducing demand charges working with the Industrial Technology Research
and onsite generation variability using behind-the-meter energy stor- Institute as a Senior Engineer. His research inter-
age,’’ in Proc. IEEE Conf. Technol. Sustainability (SusTech), Oct. 2016, ests include photovoltaic systems, energy storage
pp. 140–146. systems, energy management, and micro grid.

VOLUME 8, 2020 203745

 C.-T. Tsai et al.: Techno-Economic and Sizing Analysis of Battery Energy Storage System for Behind-the-Meter Application

ERICA M. OCAMPO received the B.S. degree in CHENG-CHIEN KUO received the B.S., M.S.,
electrical engineering from the University of Santo and Ph.D. degrees from the National Taiwan
Tomas Manila, Philippines, in 2008, and the M.S. University of Science and Technology (NTUST)
degree in electrical engineering from the National in 1991, 1993, and 1998, respectively. He has
Taiwan University of Science and Technology been with St. John’s University, from 1994 to
(Taiwan Tech), Taipei, Taiwan, in 2018, where she 2015, and then join NTUST, since 2015, where
is currently pursuing the Ph.D. degree in electrical he is currently a Professor and an Assistant Head
engineering. From 2009 to 2019, she is affiliated with the Department of Electrical Engineering.
with the University of Santo Tomas (UST) Manila, His research interests include fault diagnosis, con-
Philippines, as an Instructor with the Faculty of ditional monitoring system design, distribution
Engineering and a Research Staff in a project of UST with the Department automation, partial discharge measurement, and optimization techniques.
of Science and Technology. Her research interests include optimal power
flow and application of machine learning and artificial intelligence in power
systems.

TEKETAY MULU BEZA received the B.Sc. degree

in electrical engineering from Arba Minch Univer-
sity, in 2007, and the M.Sc. degree in electrical
power systems engineering from Bahir Dar Uni-
versity, Ethiopia, in 2011. He is currently pursuing
the Ph.D. degree with the National Taiwan Uni-
versity of Science and Technology (Taiwan Tech),
Taiwan. He has been a Lecturer with Bahir Dar
University, from 2011 to 2017. His research inter-
ests include micro grid and renewable energy anal-
ysis, distributed power generation, energy management, load forecasting,
and power system optimization techniques.

203746 VOLUME 8, 2020