Optimizing Size of Variable Renewable Energy Sources by Incorporating Energy Storage and Demand Response

Received June 3, 2019, accepted July 14, 2019, date of publication July 16, 2019, date of current version August 12, 2019.
Digital Object Identifier 10.1109/ACCESS.2019.2929297
Optimizing Size of Variable Renewable Energy
Sources by Incorporating Energy Storage
and Demand Response
MUHAMMAD FAIZAN TAHIR 1, (Student Member, IEEE), CHEN HAOYONG 1, (Fellow, IEEE),
ASAD KHAN2, MUHAMMAD SUFYAN JAVED3,4, NAUMAN ALI LARAIK 1,
AND KASHIF MEHMOOD5,6
1School of Electric Power, South China University of Technology, Guangzhou 510640, China
2School of Computer Science, South China Normal University, Guangzhou 510640, China
3Department of Physics, COMSATS University Islamabad, Lahore Campus, Punjab 54000, Pakistan
4Department of Physics, Jinan University, Guangzhou 510640, China
5School of Electrical Engineering, Southeast University, Nanjing 210096, China
6Department of Electrical Engineering, The University of Lahore, Lahore 54000, Pakistan
Corresponding author: Chen Haoyong (eehychen@scut.edu.cn)
This work was supported by the National Key Research and Development Program of China under Grant 2016YFB0900100.
ABSTRACT The electricity sector contributes to most of the global warming emissions generated from
fossil fuel resources which are becoming rare and expensive due to geological extinction and climate
change. It urges the need for less carbon-intensive, inexhaustible Renewable Energy Sources (RES) that
are economically sound, easy to access and improve public health. The carbon-free salient feature is the
driving motive that propels widespread utilization of wind and solar RES in comparisons to rest of RES.
However, stochastic nature makes these sources, variable renewable energy sources (VRES) because it brings
uncertainty and variability that disrupt power system stability. This problem is mitigated by adding energy
storage (ES) or introducing the demand response (DR) in the system. In this paper, an electricity generation
network of China by the year 2017 is modeled using EnergyPLAN software to determine annual costs,
primary energy supply (PES) and CO2 emissions. The VRES size is optimized by adding ES and DR (daily,
weekly, or monthly) while maintaining critical excess electricity production (CEEP) to zero. The results
substantiate that ES and DR increase wind and solar share up to 1000 and 874 GW. In addition, it also
reduces annual costs and emissions up to 4.36 % and 45.17 %.
INDEX TERMS Critical excess electricity production, demand response, EnergyPLAN, energy storage,
variable renewable energy sources.
I. INTRODUCTION
In this urbanized and industrialized era of technology,
the rapid growth of electricity is satiated by expensive
fossil fuels which are depleting quickly and polluting the
atmosphere. This leads to switch to clean, cheap and carbon-
free Renewable Energy Sources (RES) [1] that has been fur-
ther divided into controlled RES (hydropower, geothermal,
biomass) and Variable Renewable Energy Sources (VRES)
(wind and solar). VRES showing continuous growth in elec-
tricity production over the last decade, especially in Europe,
China and USA. China has recently become the world leader
The associate editor coordinating the review of this manuscript and
approving it for publication was Shuaihu Li.
in RES generation by surpassing the USA and Europe and
aims for 35% electricity consumption by RES till 2030 [2].
Howbeit, these resources have quasi variability because of
relying upon climate, season and location. It brings the uncer-
tainty that causes the system to become unstable and may
lead to cascaded failure [3]–[5]. This challenge can be over-
come by adding Energy Storage (ES), Demand Response
(DR) or both [6]–[8]. Electrical energy is stored in some form
and converted to electric power when VRES is insufﬁcient
to fulﬁl the demand. It diminishes uncertainty as well as
increases the reliability of the power system network [9], [10].
Nonetheless, its implementation cost is quite high as only
75-80% of stored energy are redeemable. In essence, DR
becomes an expedient alternative in easing the integration
VOLUME 7, 2019 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see http://creativecommons.org/licenses/by/4.0/ 103115

M. F. Tahir et al.: Optimizing Size of VRES by Incorporating Energy Storage and Demand Response
of VRES without having high implementation cost. DR is
defined as a change in the user’s energy pattern due to the
introduction of some incentive package or price signal gen-
erated by the electric market. DR programs are classified
in many types that have been illustrated in [11] and [12].
Therefore, energy saving and cost-effective alternatives can
be found by electricity network integration with other parts of
energy (like wind and solar), storage or DR and that creates a
smart energy network. Few of the sector- integrated solutions
are mentioned below:
The impact of Demand Side Management (DSM) strate-
gies in the penetration of RES are elaborated in [13]. These
strategies delay the installation of new generating stations
and improve the operation of existing plants by increas-
ing the capacity factor. However, CO2 emission costs and
energy efficiency measures are not considered in this work.
Energy storage role to integrate large-scale VRES in the
smart energy system is elaborated in [14]. Chauhan and
Saini [15] find the optimal size of a stand-alone renewable
energy system with the help of a DR strategy which relies
on appliances energy consumption schedules. Though this
DR strategy reduces the peak hour consumption of the study
area, but it is not included in the optimal design of RES.
Renewable energy integration is carried out in [16] by energy
trading method. Furthermore, geometric programming based
optimization is employed for balancing load and reducing
energy consumption from the grid. Nevertheless, RES price
and aggregator value are not taken into account. Incorporating
demand response in home energy management system is
proposed by Zunnurain et al. [17] that provides daily energy
saving of 3%. Modelling of converting the present electrical
system to 100% RES is elaborated in [18]. Energy storage
addition to address fluctuating nature of RES is demonstrated
in [19] and the optimal size of RES, DSM and ES have been
discussed in [20]. DR impacts in future electricity systems
and integrating high share or VRES are analyzed in [21].
Liu et al. [22] proposed load frequency control in small scale
power system to compensate renewable energy fluctuation by
using demand response and storage battery. Fan et al. [23]
proposed novel DR scheme to integrate high share of RES.
Modelling power system in software with the addition
of storages or DR may increase the electricity production
more than the transmission line capacity which is repre-
sented by Critical Excess Electricity Production (CEEP).
In the real world, CEEP must be zero otherwise breakdown
voltage will cause the power outage. Nevertheless, most of
the research while modelling and adding ES or DR ignore
the inclusion of CEEP while some Lund and Münster [24],
Lund [25] include CEEP but not limited it to zero. Therefore,
this work makes sure that it does not contradict the real
world scenario and that’s why CEEP is always limited to
zero. European Roadmap 2030 and 2050 has been exploring
different options to achieve sustainable and decarbonized
renewable energy [26] but such alternatives have been mostly
overlooked in Asia. China, the world’s biggest CO2 emit-
ter [27], [28] has signed the Paris agreement to reduce
CO2 emissions by 60-65% till 2030 [29]. Therefore, it is
imperative for China to integrate more renewable energy in
the system to develop green society. This study explores
different cases (DR and ES) to integrate renewable energy
in the system without risking the power system stability. This
work aims to be a first study to integrate daily, weekly and
monthly DR along with ES on such a large system of China
to integrate maximum renewable energy.
The major contributions of this work are:
i) The Electrical network system is modelled by collect-
ing dataset of different production units of China for the
year 2017. Further, this model is holistically invested by
EnergyPLAN software to determine annual costs, total fuel
balance or Primary Energy Supply (PES) and CO2 emissions.
ii) Technical optimizationis carried out by EnergyPLAN to
find the optimal case which has least CO2 emission, PES and
annual costs while integrating VRES in the system.
iii) ES addition and the effect of daily, weekly and monthly
DR in the system significantly reduces above prominent fac-
tors (CO2 emission, PES and annual costs) while restraining
CEEP to zero. Ultimately, the impact of different scenarios
with and without ES and DR are evaluated.
The rest of the paper is organized as follows: Section 2
elaborates the reference case and computation of the param-
eters necessary for modelling. Section 3 explains case stud-
ies and section 4 depicts results simulation. Finally, section
5 concludes the study and recommends future work.
II. SYSTEM MODEL
EnergyPLAN, Version 13 that released in September 2017 is
used for electrical network modelling to perform technical
optimization [30], [31]. It is done by adjusting the produc-
tion components optimally to reduce fossil fuel consump-
tion by lessening the PES. Additionally, it also minimizes
annual costs, CO2 emission and CEEP. EnergyPLAN has
been extensively applied to a variety of energy planning and
modelling such as integrating 100% RES in some European
countries [32]–[35], RES and combined heat and power man-
agement [36], [37], transport and renewable energy integra-
tion [38]–[40].
System model requires following inputs which serve the
role of decision variables.
i. Electricity demand.
ii. Electrical power production units.
iii. Simulation (defining technical limitations and con-
straints of the components if any)
iv. Costs (Determining investment costs per unit, lifetime
and percentage of investment of production units).
Technical optimization reduces the below mentioned per-
formance indicators:
a) PES
b) CO2 emission
c) Annual costs
Power system stability can be disrupted by under-
generation or over-generation that have a dramatic effect
on the system which may lead to cascaded failure [41].
103116 VOLUME 7, 2019

FIGURE 1. China electrical network 2017.
Hence, PP/Import problem, critical excess and grid stabiliza-
tion problem warnings in EnergyPLAN always assure power
system stability during simulating proposed energy model. It
should also be noted that the constraints of the components,
as well as CEEP, are within limits.
A. SYSTEM INPUTS
Following inputs are used to evaluate the system are given in
EnergyPLAN.
1) ANNUAL ELECTRICITY DEMAND
Annual electricity demand of the reference model for the year
2017 is given in Equation 1.
Pd(out)
e =
8784
t=1
Pd
e,t (1)
where t, P
d(out)
e represents the time in hours and annual
electricity consumption respectively.
2) PRODUCTION UNITS’ ELECTRICAL POWER
Total power demand (P
d(out)
e ) is fulfilled by Thermal Power
Plants (TPP), Nuclear Power Plants (NPP), constant RES
(hydropower) and VRES (wind and solar). It is represented
in Equation 2.
Pd
e = PTPP
e + PNPP
e + PHydro
e + PW
e + PS
e (2)
Central power plants tab in EnergyPLAN includes con-
densing power plant, nuclear and hydropower plants. Total
capacity and efficiency are assigned in the input tab for TPP,
NPP and hydropower plants but an additional input of water
supply is required for the hydropower plant. In addition,
the storage option in hydro plants provides ancillary services
to the grid.
Wind and Photo Voltaic (PV) supply in intermittent renew-
able electricity tab entails capacity, stabilization share and
correction factor. Stabilization share is the amount of VRES
that provide ancillary services for grid stability. Correc-
tion factor modifies the hourly distribution profile of VRES
that regulates supply between zero and full load but power
remains same at zero and full load. In this work, both factors
are inputted zero as indicated in Figure 1.
3) INVESTMENT COSTS
Investment costs (CInv
x ) can be determined by multiplying the
number of units (Px) with cost unit (Cunit−x) as represented
in Equation 3 [42].
CInv
x = Px × Cunit−x (3)
where x is the electricity production unit which can be wind,
solar, hydro and so on.
Costs units are defined in GCNY which is abbreviated as
Billions in Chinese Yuan Renminb. Hence, annual investment
costs are determined as:
CAnn
x = CInv
x ×
i
[1 − (1 + i)−n]
(4)
where i is the interest rate and n is the life period of the
production unit.
III. CASE STUDIES
EnergyPLAN developed by Henrik Lund in 1999 uses Delphi
Pascal (DP) programming which is a combination of object
oriented programming and integrated development environ-
ment [43]. Technical optimization using this freeware soft-
ware choose the suitable size of variable renewable energy
while maintaining CEEP to zero. Optimal wind and solar
sizes are determined for calculating CEEP against multiple
values of VRES to compute annual costs, fuel consumption
and CO2 emissions. VRES multiples values are assigned
in the output tab section of EnergyPLAN while in parallel
different CEEP regulation strategies can be activated. There
are nine different possibilities that can be activated according
to an order of priority to hold CEEP to zero. In this work,
CEEP regulation 71 is used to minimize critical excess which
means first VRES is reduced then VRES together with power
plants are decreased to limit CEEP to zero.
Hourly simulation at the national level for the speci-
fied year, 2017 is performed by EnergyPLAN. Hourly steps
demonstrate a greater degree of accuracy in intermittent
renewable energy sources. Electrical infrastructure compre-
hensive modelling in this energy balancing tool requires
input data set that contains energy demands, RES sources,
conversion units’ parameters detail, storage units’ limitations
and cost data set (annual costs of each production unit, life
time and percentage of investment). Input data set includ-
ing production capacities units and RES hourly distribu-
tion profile are collected from China energy balance data
VOLUME 7, 2019 103117

FIGURE 2. Flowchart of optimization of electrical network by using EnergyPLAN.
TABLE 1. Classification of case studies.
from International Energy Agency (IEA) [44], China energy
yearbook 2017 [45], China power industry statistics [46],
State grid corporation of China and various state-owned
energy corporations websites [47], [48]. Figure 1 depicts the
reference model of this research that meets China annual elec-
tric demand by various sources technologies. The ﬂowchart
of research work and case divisions are illustrated in
Figure 2 and Table 1.
103118 VOLUME 7, 2019

TABLE 2. Capacity and share of electrical production components of
reference model.
A. CASE 1 (REFERENCE CASE)
Energy demand forecasting is always hard to predict due
to tremendously increasing population, economy and many
other factors [49]. Many researchers and organizations pre-
dicted future energy demands by considering a lot of assump-
tions based on different scenario investigation [50]. As far
as, China future electricity demand is considered, it can be
estimated by considering 1.5% annually increased rate and
from 2014-2030 International Energy Agency (IEA) also
proposed this national energy demand rate for China [51].
However, this work uses 2017-year data for the reference
model which is elaborated in Fig. 1 and is considered as
case 1. Production components share of producing electricity
are depicted in Table 2.
FIGURE 3. Exclusions of nuclear supply from reference model.
TABLE 3. Capacity and share of electrical production components after
excluding nuclear supply.
B. CASE 2 (NUCLEAR SUPPLY EXCLUSION)
Compared to the reference model, NPP supply is excluded
in this case that results in the reduction of TPP supply
and enhances VRES share as demonstrated in Figure 3 and
Table 3. Nonetheless, TPP supply can be reduced to a certain
limit because of its share in stabilizing units. Additionally,
VRES share can also be increased up to a specific amount
because of CEEP constraint. Exclusion of most expensive
adding electricity storage.
nuclear supply has a considerable effect on cost that is
shown in the results section. Note that the hydropower supply
remains constant throughout this research.
Pd
e = PTPP
e + PHydro
e + PW
e + PS
e (5)
C. CASE 3 (ES ADDITION)
This case is the modification of case 2 with an additional
feature of electricity storage as shown in Figure 4.
FIGURE 4. Addition of electricity storage.
The well-established and mature ES technology pumped
hydro is used to increase the possibility of integrating more
VRES in the power system by eliminating the uncertainty
issues as this fixed storage capacity can be taken out at
any time. This storage capacity helps in reducing power
producing components capacity that proves its scintillated
performance by diminution PES, cost and emissions. Charg-
ing capacity (pump capacity), discharging capacity (turbine
capacity) and electric efficiency parameters are required for
modelling pumped hydro ES.
Pd
e = PTPP
e + PHydro
e + PW
e + PS
e + Sdis
e − Sch
e (6)
where, Sdis
e and Sch
e represents the discharging and charging
capacities respectively.
In the case of ES, it should not be recommended that charg-
ing and discharging occur simultaneously which may have a
negative impact on the system [19] (like increase in size and
capital costs). Equations (7) and (8) indicates that charging
and discharging cannot exceed their maximum limits and
introduction of the binary variable µe(0-1) in these equations
makes sure that two phenomena cannot occur simultaneously.
0 ≤ St
e,ch ≤ µeSmax
e,ch (7)
0 ≤ St
e,ch ≤ (1 − µe)Smax
e,dis (8)
VOLUME 7, 2019 103119

FIGURE 5. Addition of DR and observing its impact on the energy system.
Maximum and minimum energy stored is:
Enmin
e ≤ Ent
e ≤ Enmax
e (9)
After the end of the dispatch period, the final energy stored
must be equal to its initial value which is computed as:
En8784
x = En0
x (10)
D. CASE 4 (DR ADDITION)
In this work, DR in the form of load shifting is added to
the system as specified in Figure 5. DR is introduced as
flexible demand that has the ability to shift from one-time
interval to another but it needs the special equipment like
smart meter and Advanced Metering Infrastructure (AMI).
Other than energy demands and supply, EnergyPLAN also
requires cost analysis (that includes investment cost per unit,
lifetime and percentage of investment) to perform technical
optimization. Cost of all production units except instruments
related to DR activities are listed in the general cost section.
Therefore, to fully cover the impact of adding DR on the
system, the costs of such equipments are inserted in ‘addi-
tional costs’ section of EnergyPLAN. Though total power
consumption during this load management remains same but
load shifting from peak to valley interval causes a reduction
in annual costs. The DR amount is assumed to be 10 percent
of total electricity demand and EnergyPLAN gives the option
of adding this amount in the input section. EnergyPLAN also
facilitates the distribution of DR into daily, weekly and four
weeks over the yearly time period. Both, PES and RES elec-
tricity production share are increased in comparison to case
3. Further, monthly DR signifies maximum VRES integration
than daily and weekly DR as demonstrated in table 5.
Pd
e = PTPP
e + Phydro
e + PW
e + PS
e + PDR, SHDO
e − PDR, SHUP
e
(11)
E. CASE 5 (DR AND ES ADDITION)
This case assesses the performance indicators when append-
ing both ES and DR as delineated in Figure 6 and expressed
in equation 12.
Pd
e = PTPP
e + Phydro
e + PW
e + PS
e + . . . ,
. . . Se
dis − Se
ch + PDR, SHDO
e − PDR, SHUP
e (12)
adding demand response.
FIGURE 6. Addition of DR and observing its impact on the energy system.
The total amount of power shifted from one time period
(shift down, SHDO) period to other (shift up, SHUP) remains
same as shown in equation 13. Further, there is a limit up
to which DR can be shifted during one-hour interval. The
maximum and minimum limit is specified in Equation 15.
Usually, EnergyPLAN specified it to be 1000 MW but it is
not mandatory that all of this demand be shifted during the
specified hour.
8784
t=1
PDR,SHDO
e,t =
8784
t=1
PDR,SHUP
e,t (13)
PDR(min)
e ≤ PDR
e ≤ PDR(max)
e (14)
Similar to ES, binary variables are also introduced in this
case to make sure that two operations do not occur simultane-
ously. These shifting up (µSHUP
e ) and shifting down (µSHDO
e )
binary variables are shown in Equation 15.
0 ≤ µSHUP
e + µSHDO
e ≤ 1 (15)
The proposed work first observes the impact of these three
DRs without optimizing VRES. Just like ES, DR also facili-
tates the VRES integration, therefore, three DRs addition will
further increase the share of VRES that is depicted in Table 6.
ES and DR combination makes the system economically
viable than rest of the scenarios as it causes substantial reduc-
tion in carbon footprints along with annual costs.
103120 VOLUME 7, 2019

FIGURE 7. EnergyPLAN results for case 1 and case 2.
FIGURE 8. Finding optimum wind size without exceeding CEEP.
IV. SIMULATION RESULTS AND DISCUSSION
A. CASE 1 AND CASE 2
The data of the most recent year (2017) is set as a reference
model due to its availability that determines annual costs,
emissions and PES. Then, exclusion of nuclear supply with-
out altering VRES is carried out to discover its effect on
the above eminent factors. Hence, this exclusion not only
descends the PES but also slightly reduces annual costs
because NPP supply is considered to be the most expensive
source than all [52]. However, CO2 emissions are increased
due to the increment in the share of TPP electricity production
as sketched in Figure 7.
Two of the objective functions are alleviated without opti-
mizing and increasing the share of VRES. Hence, the exclu-
sion of NPP share can be given to VRES to assess if it
descends or ascends the annual costs in comparison to case 1.
Figures 8 and 9 show the results of case 2 to ﬁnd the optimal
amount of wind and solar when NPP supply is excluded.
However, optimal means that these sources do not cross
beyond transmission capacity while ﬁnding the least cost
solution with minimum emission and PES.
Though CO2 and PES are diminished to a minimum at
the maximum wind power supply (1100 GW) but both the
value of CEEP and annual costs increase. Annual costs are
VOLUME 7, 2019 103121

FIGURE 9. Finding optimum PV size without exceeding CEEP.
TABLE 6. Effect of DR and ES on production supply, before and after
optimizing VRES.
minimum at 450 GW supply of wind, but CO2 emissions
and PES are not minimized. Similarly, CO2 and PES are also
reduced to the minimum at maximum PV supply (1100 GW)
but annual cost is minimized at 550 GW supply and increases
beyond this value. The steep decline in annual costs of PV
at 580 GW is justified by learning curve theory of PV. This
theory discussed by Connolly et al. [30] which states that up
to a certain limit increase in PV capacity results to decrease
in costs.
Therefore, a compromise has to be made for either one of
the factors between emissions, PES and annual costs to get the
best-optimized results. Considering all the aforementioned
factors, final optimized values chosen for wind and PV are
550 GW and 203 GW respectively.
As discussed above that CO2 emissions increases in
case 2 (without optimization scenario) but after optimizing
VRES, this tradeoff is eliminated as now CO2 is reduced
from 2.40 (Gt) to 1.99 (Gt), PES 11.92 (PWh/year) to
10.96 (PWh/year) and annual costs from 2864 (GCNY) to
2815 (GCNY). Additionally, RES electricity production is
increased to 2.50 PWh/year and RES share of PES and
RES share of electricity is ascended to 22.8% and 39.7%
respectively.
B. CASE 3
Energy storage addition has become very popular choice
to address stochastic nature of renewable energy sources,
especially in Europe such as Madeira Island in Portugal [10],
Germany [53] and so on. The addition of ES dispenses more
opportunities for integrating VRES as it can store surplus
energy and detaches that energy when VRES supply is insuf-
ficient. Consequently, a tremendous reduction in all three
parameters is noted which is recorded in Table 7.
C. CASE 4
Besides energy storage, demand response is also viable alter-
native to add renewable energy in the system without jeop-
ardizing system stability [2], [54], [55]. In this work, DR is
divided into daily, weekly and monthly DR that distribute
flexible demand (0.63 PWh) into 365 days, 52 weeks and
12 months respectively. It checks to see if it can move the
demand over a specified period in order to improve the VRES
103122 VOLUME 7, 2019

TABLE 7. Summary performance and comparison results of three case studies.
TABLE 8. Summary performance of case 4 (after optimization).
TABLE 9. Summary performance of case 5 (before and after optimization).
integration. Case 4 (especially weekly DR) proves to be the
best case in terms of integrating VRES and decarbonizing
atmosphere as shown in table 8 but it does not turn out to
be economically viable due to increment in annual costs.
D. CASE 5
The last case includes both ES and DR but this case has
been further divided into six subcases. First, the impact of
daily, weekly and 4 weeks DR are highlighted separately
without increasing VRES share. Then, three DR scenarios are
evaluated again while integrating more VRES and observing
its effect on costs and other factors which are summarized
in Table 9.
As stated above, PES and CO2 are minimized more in case
4 in comparison to case 5 but ascending annual costs does
not make this case ideal. Therefore, the combination of ES
and DR are needed to make the system economically viable
while keeping the emissions minimum. In case 5, best results
before optimizing are achieved in 4 weeks DR scenario where
three objective functions have plummeted in comparison to
all aforementioned cases. Further, in the after optimizing
scenario, PES and CO2 are reduced to the minimum in
4-week scenario but annual cost increases than weekly and
daily scenario. Therefore, in this case, there is a tradeoff
between CO2, PES and annual costs. The overall comparison
of all seven cases have been summarized in Figure 10 and it
makes sure that CEEP for all the cases are zero. Moreover, it
demonstrates that reduction in cost and PES in case 2 (before
optimizing VRES size) in comparison to case 1 is justified
because now less energy is required to fulfill electric demand
and the most expensive supply has been excluded. However,
CO2 emissions are increased due to the increment in the share
of TPP electricity production. This tradeoff is eliminated
when the NPP supply share is given to VRES that reduces
all three factors. Likewise, alleviation of objective functions
in case 3 and 4 is more as compared to the first two cases
because ES and DR act as additional storage which integrates
more VRES. These two cases not only reduce dependency on
VOLUME 7, 2019 103123

FIGURE 10. Comparison of optimum results for PES, CO2 emissions & annual costs against 6 different cases.
fossil fuel sources but also minimize the annual costs more
than aforementioned scenarios. Finally, the combination of
ES and DR scenario proves to be the most economical option
to integrate VRES.
V. CONCLUSIONS AND FUTURE WORK
User-friendly, deterministic input/output EnergyPLAN tool
is used to model the electrical network of China as it takes
only a few seconds for the computation of the whole year
(2017). The research is divided into five case studies that
illustrate the optimal combination of VRES, ES and DR that
not only help to decarbonize the atmosphere by reducing
fuel consumption but also reduce the annual costs as well.
However, the integration of excess VRES increases produc-
tion more than the transmission capacity, which is repre-
sented in the form of CEEP. Consequently, the addition of
storage and DR (daily, weekly and monthly) in this research
not only enhanced the VRES integration without exceeding
CEEP but also achieved significant improvement in mini-
mizing the objective functions. The proposed work acquired
successive decrement in CO2 emissions (2.28 Gt to 1.12 Gt),
PES (11.99 PWh/year to 8.94 PWh/year) and annual costs
(2865 GCNY to 2740 GCNY) from case 1 to case 5. It is
concluded that case 4 (DR 4-week, after optimization sce-
nario) is best suited when the focus is to reduce greenhouse
gases. On the contrary, case 5 (ES and DR 4- week, before
optimization scenario) can be chosen when cost is the main
consideration.
This research can be further extended to multiple energy
carriers like heating, cooling and hydrogen demands in the
system by incorporating integrated demand response. It may
add complexity to the system but will increase the flexibility
and reliability while restraining the operating costs and emis-
sions to the minimum.
REFERENCES
[1] Z. Liang, H. Chen, X. Wang, I. I. Idris, B. Tan, and C. Zhang, ‘‘An extreme
scenario method for robust transmission expansion planning with wind
power uncertainty,’’ Energies, vol. 11, no. 8, p. 2116, Aug. 2018.
[2] Y.-Q. Bao, Y. Li, B. Wang, M. Hu, and P. Chen, ‘‘Demand response for
frequency control of multi-area power system,’’ J. Mod. Power Syst. Clean
Energy, vol. 5, no. 1, pp. 20–29, Jan. 2017.
[3] L. Bird, M. Milligan, and D. Lew, ‘‘Integrating variable renewable energy:
Challenges and solutions,’’ Nat. Renew. Energy Lab., Lakewood, CO,
USA, Tech. Rep. WE11.0820, 2013.
[4] A. K. Alsaif, ‘‘Challenges and benefits of integrating the renewable energy
technologies into the AC power system grid,’’ Amer. J. Eng. Res., vol. 6,
no. 4, pp. 95–100, 2017.
[5] M. FaizanTahir, ‘‘Optimal load shedding using an ensemble of artificial
neural networks,’’ Int. J. Elect. Comput. Eng. Syst., vol. 7, no. 2, pp. 39–46,
Dec. 2016.
[6] F. C. Robert, G. S. S. Sisodia, and S. Gopalan, ‘‘A critical review on
the utilization of storage and demand response for the implementation of
renewable energy microgrids,’’ Sustain. Cities Soc., vol. 40, pp. 735–745,
Jul. 2018.
[7] O. Ma and K. Cheung, ‘‘Demand response and energy storage integra-
tion study,’’ U.S. Dept. Energy, Office Energy Efficiency Renew. Energy,
Washington, DC, USA, Tech. Rep., 2016.
[8] M. Faisal, M. A. Hannan, P. J. Ker, A. Hussain, M. B. Mansor,
and F. Blaabjerg, ‘‘Review of energy storage system technologies in
microgrid applications: Issues and challenges,’’ IEEE Access, vol. 6,
pp. 35143–35164, 2018.
103124 VOLUME 7, 2019

[9] A. Habib and L. B. Ling Abbas, ‘‘Demand and application of energy
storage technology in renewable energy power system,’’ Amer. Sci. Res.
J. Eng., Technol., Sci., vol. 36, no. 1, pp. 75–84, Sep. 2017.
[10] M. Miguel, T. Nogueira, and F. Martins, ‘‘Energy storage for renewable
energy integration: The case of Madeira Island, Portugal,’’ Energy Proce-
dia, vol. 136, pp. 251–257, Oct. 2017.
[11] J. S. Vardakas, N. Zorba, and C. V. Verikoukis, ‘‘A survey on demand
response programs in smart grids: Pricing methods and optimization
algorithms,’’ IEEE Commun. Surveys Tuts., vol. 17, no. 1, pp. 152–178,
Mar. 2015.
[12] M. F. Tahir, H. Chen, I. I. Idris, N. A. Laraik, and S. Adnan, ‘‘Demand
response programs significance, challenges and worldwide scope in main-
taining power system stability,’’ Int. J. Adv. Comput. Sci. Appl., vol. 9, no. 8,
pp. 121–131, Jul. 2018.
[13] A. Pina, C. Silva, and P. Ferrão, ‘‘The impact of demand side management
strategies in the penetration of renewable electricity,’’ Energy, vol. 41,
no. 1, pp. 128–137, May 2012.
[14] H. Lund, P. A. Østergaard, D. Connolly, I. Ridjan, B. V. Mathiesen,
J. Z. Thellufsen, and P. Sorknæs, ‘‘Energy storage and smart energy sys-
tems,’’ Int. J. Sustain. Energy Planning Manage., vol. 11, pp. 3–14,
Oct. 2016.
[15] A. Chauhan and R. Saini, ‘‘Size optimization and demand response of
a stand-alone integrated renewable energy system,’’ Energy, vol. 124,
pp. 59–73, Apr. 2017.
[16] A. Chis, J. Rajasekharan, J. Lundén, and V. Koivunen, ‘‘Demand response
for renewable energy integration and load balancing in smart grid
communities,’’ in Proc. 24th Eur. Signal Process. Conf. (EUSIPCO),
Aug./Sep. 2016, pp. 1423–1427.
[17] I. Zunnurain, N. I. Maruf, M. Rahman, and G. M. Shafiullah, ‘‘Implemen-
tation of advanced demand side management for microgrid incorporating
demand response and home energy management system,’’ Infrastructures,
vol. 3, no. 4, p. 50, Nov. 2018.
[18] H. Lund, ‘‘Renewable energy systems—A smart energy systems approach
to the choice and modelling of 100% renewable solutions,’’ Chem. Eng.
Trans., vol. 39, pp. 1–6, Jan. 2014.
[19] D. Connolly, ‘‘The integration of fluctuating renewable energy using
energy storage,’’ Tech. Rep., 2011.
[20] D. B. Richardson and L. D. Harvey, ‘‘Optimizing renewable energy,
demand response and energy storage to replace conventional fuels in
Ontario, Canada,’’ Energy, vol. 93, pp. 1447–1455, Dec. 2015.
[21] S. Bouckaert and V. Mazauric, and N. Maïzi, ‘‘Expanding renewable
energy by implementing demand response,’’ Energy Procedia, no. 61,
pp. 1844–1847, Dec. 2014.
[22] L. Liu, H. Matayoshi, M. Lotfy, M. Datta, and T. Senjyu, ‘‘Load frequency
control using demand response and storage battery by considering renew-
able energy sources,’’ Energies, vol. 11, no. 12, p. 3412, Dec. 2018.
[23] S. Fan, G. He, K. Jia, and Z. Wang, ‘‘A novel distributed large-scale demand
response scheme in high proportion renewable energy sources integration
power systems,’’ Appl. Sci., vol. 8, no. 3, p. 452, Mar. 2018.
[24] H. Lund and E. Münster, ‘‘Management of surplus electricity-production
from a fluctuating renewable-energy source,’’ Appl. Energy, vol. 76, no. 1,
pp. 65–74, Sep./Nov. 2003.
[25] H. Lund, ‘‘Flexible energy systems: Integration of electricity production
from CHP and fluctuating renewable energy,’’ Int. J. Energy Technol.
Policy, vol. 1, no. 3, pp. 250–261, 2003.
[26] H. Wang, G. Di Pietro, X. Wu, R. Lahdelma, V. Verda, and I. Haavisto,
‘‘Renewable and sustainable energy transitions for countries with different
climates and renewable energy sources potentials,’’ Energies, vol. 11,
no. 12, p. 3523, Dec. 2018.
[27] Y. Shan, D. Guan, ‘‘Methodology and applications of city level CO2
emission accounts in China,’’ J. Cleaner Prod., vol. 161, pp. 1215–1225,
Sep. 2017.
[28] M. F. Tahir, C. Haoyong, K. Mehmood, N. Ali, and J. A. Bhutto,
‘‘Integrated energy system modeling of China for 2020 by incorporating
demand response, heat pump and thermal storage,’’ IEEE Access, vol. 7,
pp. 40095–40108, 2019.
[29] Y. Wang, ‘‘China signs Paris agreement on climate change,’’ Ph.D. disser-
tation, United Nations Framework Convention Climate Change, New York,
NY, USA, 2016.
[30] D. Connolly, ‘‘A user’s guide to EnergyPLAN,’’ Univ. Limerick, Denmark,
Tech. Rep., 2010.
[31] D. Neves, A. Pina, and C. A. Silva, ‘‘Demand response modeling: A com-
parison between tools,’’ Appl. Energy, vol. 146, pp. 288–297, May 2015.
[32] D. Connolly, H. Lund, B. V. Mathiesen, and M. Leahy, ‘‘The first step
towards a 100% renewable energy-system for Ireland,’’ Appl. Energy,
vol. 88, no. 2, pp. 502–507, Feb. 2011.
[33] R. Groenewoud, ‘‘Energy self-sufficient neighborhoods in the Nether-
lands: A technical framework on the energy storage & land usage require-
ment for intermittent renewable energy systems,’’ M.S. thesis, 2013.
[34] M. Child and C. Breyer, ‘‘Vision and initial feasibility analysis of a
recarbonised Finnish energy system for 2050,’’ Renew. Sustain. Energy
Rev., vol. 66, pp. 517–536, Dec. 2016.
[35] G. Krajačić, N. Duić, and M. da Graça Carvalho, ‘‘How to achieve a
100% RES electricity supply for Portugal?’’ Appl. Energy, vol. 88, no. 2,
pp. 508–517, Feb. 2011.
[36] H. Lund and W. W. Clark, ‘‘Management of fluctuations in wind power and
CHP comparing two possible Danish strategies,’’ Energy, vol. 27, no. 5,
pp. 471–483, May 2002.
[37] H. Lund and E. Münster, ‘‘Modelling of energy systems with a high
percentage of CHP and wind power,’’ Renew. Energy, vol. 28, no. 14,
pp. 2179–2193, Nov. 2003.
[38] H. Lund and E. Münster, ‘‘Integrated transportation and energy sector CO2
emission control strategies,’’ Transp. Policy, vol. 13, no. 5, pp. 426–433,
Sep. 2006.
[39] H. Lund and W. Kempton, ‘‘Integration of renewable energy into the
transport and electricity sectors through V2G,’’ Energy Policy, vol. 36,
no. 9, pp. 3578–3587, 2008.
[40] B. V. Mathiesen and H. P. Lund Nørgaard, ‘‘Integrated transport and
renewable energy systems,’’ Utilities Policy, vol. 16, no. 2, pp. 107–116,
Jun. 2008.
[41] M. F. Tahir, Tehzeeb-ul-Hassan, and M. A. Saqib, ‘‘Optimal scheduling
of electrical power in energy-deficient scenarios using artificial neural
network and Bootstrap aggregating,’’ Int. J. Elect. Power Energy Syst.,
vol. 83, pp. 49–57, Dec. 2016.
[42] Hydropower Status Report, Int. Hydropower Assoc., London, U.K., 2017.
[43] EnergyPLAN, and Advanced Energy Systems Analysis Computer Modeling
Documentation Version 14, Sustain. Energy Planning Res. Group, Aalborg
Univ., Aalborg, Denmark, 2017.
[44] World Energy Outlook 2017, IEA, Paris, France, 2017.
[45] Annual Report of China’s Power Sector Development, China Electricity
Council, China Electricity Press, Beijing, China, 2017.
[46] China Power Industry Statistic in 2017, China Electricity Council, Beijing,
China, 2017.
[47] China Three Gorges Corporation. Water Regimen Information, China
Three Gorges Corporation, Yichang, China, 2017.
[48] State Grid Corporation Power Supply and Sale Statistic, State Grid Cor-
poration, Beijing, China, 2017.
[49] Outlook to 2035, London, U.K., British Petroleum, 2016.
[50] L. Suganthi and A. A. Samuel, ‘‘Energy models for demand forecasting—
A review,’’ Renew. Sustain. Energy Rev., vol. 16, no. 2, pp. 1223–1240,
Feb. 2012.
[51] World Energy Outlook Special Report (2015), Int. Energy Agency, Paris,
France, 2017, p. 37.
[52] V. H. M. Visschers, C. Keller, and M. Siegrist, ‘‘Climate change benefits
and energy supply benefits as determinants of acceptance of nuclear power
stations: Investigating an explanatory model,’’ Energy Policy, vol. 39, no. 6,
pp. 3621–3629, Jun. 2011.
[53] S. Weitemeyer, D. Kleinhans, T. Vogt, and C. Agert, ‘‘Integration of
renewable energy sources in future power systems: The role of storage,’’
Renew. Energy, vol. 75, pp. 14–20, Mar. 2015.
[54] K. T. Ponds, A. Arefi, A. Sayigh, and G. Ledwich, ‘‘Aggregator of demand
response for renewable integration and customer engagement: Strengths,
weaknesses, opportunities, and threats,’’ Energies, vol. 11, no. 9, p. 2391,
2018.
[55] X. Wang, A. N. Palazoglu, and H. El-Farra, ‘‘Operational optimization
and demand response of hybrid renewable energy systems,’’ Appl. Energy,
vol. 143, pp. 324–335, Apr. 2015.
MUHAMMAD FAIZAN TAHIR received the
B.Sc. degree in electrical engineering from UET
Taxila, in 2011, and the M.S. degree in electrical
engineering from The University of Lahore, Pak-
istan, in 2015. He is currently pursuing the Ph.D.
degree in power system and automation with the
South China University of Technology. He was
a Lecturer with The University of Lahore, from
2012 to 2016. His research interests include renew-
able energy, demand response, load management,
and integrated energy system planning.
VOLUME 7, 2019 103125

CHEN HAOYONG received the B.S., M.S.E.,
and Ph.D. degrees in electrical engineering from
Xi’an Jiaotong University, in 1995, 1997, and
2000, respectively. He was an Influential Professor
of electrical engineering (in both academia and
industry). He has 14 years’ research and teaching
experience as a University Faculty Member of
electrical engineering. He is currently the Director
with the Institute of Power Economics and Elec-
tricity Markets, South China University of Tech-
nology. He has been the Principal of many key national projects of China
and has close connections with Chinese Power Industry. His current research
interests include excellent research record on modeling/optimization/control
of power systems and electricity markets.
ASAD KHAN received the bachelor’s degree in
applied mathematics from Government College
University Faisalabad (GCUF), Pakistan, in 2010,
the master’s degree in mathematical modeling and
scientific computing from Air University Islam-
abad (AIU), Pakistan, in 2012, and the Ph.D.
degree in image and video processing from the
University of Science and Technology (USTC),
in 2017. He joined USTC, in 2014. He is currently
a Postdoctoral Fellow and a Teaching Instructor
with South China Normal University, Guangzhou, China. His research inter-
ests include image processing, computer vision, deep learning, computa-
tional photography, hyperspectral imaging, and wearable computing. He is
an Active Reviewer for several top tier Journals and IEEE Transactions
including but not limited to Springer Nature, the IEEE TRANSACTIONS ON
IMAGE PROCESSING, the IEEE TRANSACTIONS ON NEURAL NETWORK AND LEARNING
SYSTEMS, Remote Sensing Letters, the IEEE TRANSACTIONS ON MULTIMEDIA,
the IEEE COMPUTERS, the IEEE SENSORS JOURNAL, IEEE ACCESS, Neural
Computing and Applications, Computer and Graphics, and IET Image Pro-
cessing.
MUHAMMAD SUFYAN JAVED received the
Ph.D. degree from the College of Physics, in 2017,
with distinction of outstanding International
Student and also received the Chinese Government
Outstanding International Student Scholarship,
in 2016 during his Ph.D. stay in Chongqing Uni-
versity, China. He was under the supervision of
Prof. C. G. Hu. He is currently a Postdoctoral
Researcher with Prof. W. J. Mai’s Group, Jinan
University, China. He has published more than
50 research papers in well reputed journals in the field of energy and mate-
rials science and their simulations. His current research interests include the
synthesis and direct growth of novel hierarchical nanostructured materials
for electrochemical energy storage and conversion devices.
NAUMAN ALI LARAIK received the B.Eng.
degree in electrical engineering from QUEST,
Nawabshah, Pakistan, in 2011, and the M.S.
degree in electrical engineering from Xi’an Jiao-
tong University, Xi’an, Shaanxi, China, in 2015.
He is currently pursuing the Ph.D. degree with
the School of Electric Power Engineering, South
China University of Technology, Guangzhou,
China. He was a Lecturer with DHA SUFFA Uni-
versity, Pakistan, for two years.
KASHIF MEHMOOD received the B.Sc. and
M.Phil. degrees (Hons.) in electrical engineering
from The University of Lahore, in 2011 and 2015,
respectively. He is currently pursuing the Ph.D.
degree with Southeast University Nanjing, China.
In 2012, he joined The University of Lahore,
where he has been with the Electrical Engineering
Department. His research interests include power
system operation and control, the application of
metaheuristic optimization (artificial intelligence)
techniques in power system’s problem, and advance flexible AC transmission
and distribution systems (FACTS).
103126 VOLUME 7, 2019

Optimizing Size of Variable Renewable Energy Sources by Incorporating Energy Storage and Demand Response

More Related Content

What's hot (18)

Similar to Optimizing Size of Variable Renewable Energy Sources by Incorporating Energy Storage and Demand Response (20)

More from Kashif Mehmood (10)

Recently uploaded (20)

Optimizing Size of Variable Renewable Energy Sources by Incorporating Energy Storage and Demand Response