Optimization of On-Grid Hybrid Renewable Energy System: A Case Study on Azad Jammu and Kashmir

Abstract

Expansion of modern power systems due to increasing energy demands face the challenges of grid reinforcement cost, size and complexity, transmission losses, and environmental factors. Placement of renewable energy sources (RES) based generation systems addresses these challenges. However, the size and placement location of RES-based system require optimization of installation and operational cost with better return on investment and reduction of greenhouse gas emissions. This paper presents an optimized solution for RES-based generation system to be installed with the existing power system of Azad Jammu and Kashmir (AJK) region that is facing power shortfall and load shedding. The weather and climate data from NASA and National Renewable Energy Laboratory (NREL) have been used and various models of on-grid hybrid renewable energy system (HRES) are compared to highlight their techno-economic benefits. An optimal hybrid photovoltaic, wind, and hydroelectric energy-based generation system is proposed with a significant reduction in cost of energy, net present cost, initial costs, and GHG emissions. Installation of the proposed hybrid RES-based generation system guarantees reduction in system power losses and line flows with an improved voltage profile of the system.

1. Introduction

Electricity is essential for life, innovation, invention, and progress. Many regions in developing countries face a shortage of electrical energy due to generation shortage, unavailability of transmission and distribution networks, or their limited capacity to transmit the power. In many regions, despite extensive availability of renewable-based electricity generation options, the advantage cannot be taken due to either unavailability or limited capacity of the transmission and distribution grids. Installation of additional lines and grid reinforcement methods for increasing the capacity are costlier options which take several years to complete [1]. Therefore, on-site generation i.e., generation near the demand centers to avoid the need of transmission networks is considered to be the most viable solution under such conditions. On-site generation also helps in reducing the transmission losses, network congestion, and capital investment requirements. However, such benefits can be fully exhausted if the optimally sized renewable-based generation is optimally placed within the grid [2].

The interest in renewable energy sources (RES) based DGs has increased due to their smaller footprint, quicker installation, and environmental impacts. The regions with power shortage challenges can be energized in shorter duration with DGs. However, it is noteworthy that due to unplanned and non-optimal placement of such generation resources, the power network and power system face challenges such as, but not limited to, power quality, power congestion over the network, and reverse power flow [3]. The energy utilization in a power grid can be maximized by reducing the energy wastage due to power losses and better energy management techniques. For example, the sustainable reconfiguration of distribution networks [4] and enhancement in network planning [5] using various optimization methods can improve the performance of modern power grids. Recently, a Robust Model Predictive Control for maximizing the user comfort and minimizing the economical cost is discussed in [6]. Moreover, the energy harvesting can be maximized if the RES-based DGs are installed after studying the availability of primary energy, power requirements, land availability, and cost of the DG systems and the cost incur on connecting these generators to the main grid. It is, therefore, necessary to optimize the planning of DG installation based upon these factors.

This work focuses upon mitigating the problem of power shortage by proposing an on-grid Hybrid Renewable Energy System (HRES). The hybrid system is preferred to minimize the challenges posed due to intermittent nature of most renewable resources. In order to make the study implementable in real scenario, a real power system of Azad Jammu and Kashmir (AJK) has been considered. AJK is mostly energized from the national grid of Pakistan [7]; however, it has faced severe electricity shortage in recent years. In the summer, almost 4–6 h of power cuts hamper economic growth. Consequent to the ever-increasing energy demand for the AJK, the nation has aggressive plans for the transition towards renewable energy as AJK has abundant availability of hydroelectric, wind, and solar energy. The optimal locations, sizes, and type of DGs are found based upon real data from various sources.

With 8–10 h of daily sunshine, 95% of the land gets consistent solar insulation of approximately 5–7 ${\mathrm{kWh}/\mathrm{m}}^2/\mathrm{day}$ (200–250 ${\mathrm{W}/\mathrm{m}}^2$) for 85% of days in a year [8]. The yearly estimate of solar irradiation in the most humid zones of AJK is over 150 ${\mathrm{W}/\mathrm{m}}^2$ (Nov–Jan), whereas the northern and Kashmir areas received over 200 ${\mathrm{W}/\mathrm{m}}^2$ (Feb–Oct) [9]. The nation’s hydroelectric generation capacity is approximately 60 GW [10], with approximately 20 GW from AJK [11]. Along the coastal belt of 9700 ${\mathrm{km}}^2$ with wind energy at the power density of 4 ${\mathrm{MW}/\mathrm{km}}^2$, Pakistan has the potential of generating 43 GW [12]. Being agricultural and dairy farming rich country, 0.71 million tonnes of dung is produced daily by 172 million cattle, only 50% of which is enough for generating 1.9 GW daily [13,14]. AJK has a total of over 15 million livestock animals capable of producing around 4 million tonnes of biomass every year [15]. Despite its potential, most of Pakistan’s renewable energy sources remain untapped or unexplored. Non-dispatchability due to the intermittent nature of most of these resources has been a major challenge for maximizing their utilization.Pakistan is ranked sixth among countries affected by global warming and extreme weather patterns [16]. The purpose of this research article is to design a resilient hybrid power system employing renewable energy resources to meet the local problem of a dependable and robust power system that supplies electricity in the prevailing weather conditions and reduces outages.

Environmentally friendly hybrid systems that combine two or more renewable energy sources have the potential to reduce demand on the main grid by a significant amount [17]. Much of the research has focused on using HRES in remote and challenging terrains where grid expansion is impossible or uneconomical. In recent years, several studies on HRES integration have been conducted. Increasing interest in studying HRES system for AJK has been witnessed in recent years [18,19].

Renewable-based HRES has been proposed for residential consumers of Kotli region of AJK in [18]. The actual system is simulated in HOMER Pro, utilizing the available resources and total load demand from residents. Various configurations and combinations of available energy generating systems are compared to get the most optimum system design. It contains 40 kW PV, a 25-kW biomass generator, and two 40 kW hydro kinetic turbines each accumulating to approximately 94% of annual demand. In comparison to power purchasing from grid, the system is cost effective as well efficient. However, the biomass-based systems are less efficient and require more maintenance along with higher greenhouse gas emission [20].

Geothermal energy-based low cost HRES is proposed for Tattapani region of AJK to fulfill the perennial base load needs of tiny community [21]. Considering the approximate conversion rate of $USD1=PKR178$, the net present cost (NPC) is PKR 234.11M for an on-grid hybrid geothermal, PV, and wind system at Tattapani. The system can provide 7350 kWh/day of average load demand, with excess exported to the grid. The suggested system’s cost of energy (COE) is 7.50 PKR/kWh. The system is expected to save the emission of 1.8 million kg of CO₂ and other pollutants. However, the system expandability is limited due to short availability of geothermal energy; hence, the system design cannot be extended to other regions.

An on-grid HRES for the University of Azad Jammu and Kashmir has been proposed in [19] with COE of 0.251 PKR/kWH which is significantly lower than the purchase from national grid i.e., at the rate of 27.32 PKR/kWh. The system produces approximately 1000 kWh per day and has a peak output of 200 kW. The system hybridizes the PV, battery and grid, which has led to higher initial cost and more dependency upon the grid when solar energy is unavailable.

Refs. [22,23] proposed PV/wind/biomass and PV/wind/biogas/battery HRES, respectively, for electricity dispatch to residential houses via demand response. An HRES system for the electrification of Korkadu village in India is proposed in [24]. This system utilizes two different approaches: one proposed the indigenous resources-based hybrid configuration while the second proposed different storage backup media-based configuration. Utilization of energy storage system for handling the challenges posed by the integration of stochastic and variable DG is also discussed in [25]. A hydroelectricity, PV, and wind-based hybrid system is proposed in [26] for Ethiopia to determine the dependability and viability. Barsoum [27] built a standalone solar and biomass energy system in Sarawak, with the goal of developing an optimal, dependable, viable, and standalone system for satisfying the region’s energy needs. A standalone power producing concept for Cameron’s remote settlements has been proposed in [28]. The Sundargarh region of Orissa in India was replicated with the use of PV, wind turbines, micro hydropower, as well as diesel generators [29]. The system is only affected by the variations in wind power and household requirements. A techno-economic study of a hybrid energy system was carried out to meet the residential load demand of a distant area in Pakistan [30]. An HRES that incorporates conventional and non-conventional energy sources has been proposed in [18] for Nooriabad, Sindh, Pakistan. It has been found that a single renewable energy-based system cannot meet the whole demand and is also not financially sustainable.

All these studies are proposed for the smaller regions. As the whole of AJK has significant RES, it is highly needed to design a system for complete grid so that the dependency upon the national grid for energy needs as well as CO₂ footprint of the nation may be reduced. A study highlighting the techno-economic feasibility of HRES for sustainable rural electrification in Benin region is presented [31]. Optimization, simulation, and sensitivity analysis are done using HOMER Pro. After studying various combinations of HRES, the PV/Diesel Generator/battery system-based system has been finalized for Banin. This system provided stable electricity with 70% decrease in battery needs as compared to only PV/battery systems, and reduced CO₂ emissions by 97%.

Several studies have been undertaken globally on using local natural resources instead of the grid. This research aims to exploit the potential of AJK’s natural resources (PV/Wind/Hydro) to generate electricity. The goal is to reduce grid stress while sustaining energy supply by using renewable resources that can generate power remotely. Hybrid systems can handle the intermittent nature of renewable resources. Grid backup is often used to reduce system size and costs, as well as to sell or acquire excess electricity [32]. The optimum HRES is found using a load following method with the lowest cost of energy. This study provides a solution for comparable climatic settlements, and the proposed system may be used anywhere using local renewable energy resources and load. This research studied various combinations of renewable-based energy systems to find the best optimal results using HOMER. HOMER is used to design and evaluate HRES that employ renewables from a technological and financial (economic viability and cost comparison) standpoint. This software can examine on and off-grid systems by executing three primary actions: simulation, optimization, and sensitivity analysis [33]. Considering Net Present Cost (NPC), Cost of Energy (COE), reduced GHG emission, lower amount of excess energy, and renewable factor with low purchases from grid, the on-grid PV/wind/hydro-based system is finalized. A feasible HRES is developed for a selected location and the whole AJK network has been simulated in DIgSILENT PowerFactory.

The problem of optimal DG placement has been extensively addressed in the literature considering variety of variables and objectives. In [34], the technical and economic considerations like load point, DG prices, energy loss, dependability indices, and DG portability are considered for cost/worth analysis. However, the proposed method is tested upon an hypothetical network available in digsilent (RTS). The aim to reduce the costs while maximizing total system benefit is achieved in [35]. The optimal location and size are determined by solving the overall cost minimization by mathematical approach in the first stage, followed by the optimal DG investment payback time results solved in the second stage. There is an appropriate location and size for each DG cost feature and investment payback time. The results are provided as five different choices, among which any can choose. However, this study lacks the practical system, many practical variables such as primary energy availability, availability of enough land at the proposed locations, and other such factors affecting the practical implementation of the proposed system. It is also noteworthy that the system considered static loads, which are not near to practical scenarios. Moreover, intermittency in the generation from DGs is not considered, nor this issue is addressed. Loss minimization on a redial distribution network using DG units was addressed in [36]. To explore the highly confined search space and determine the best placements and sizes for DG units, an enhanced GA technique was used. A basic constraint handling method was presented and used, in which the number of constraint violations was added instead of the quantity of violations. It is noteworthy that GA has some inherent issues of sticking to the local optimal, making the results a bit controversial in comparison to modern techniques. Moreover, the simplification in the penalty function may also have serious implication on the quality of results. Additionally, the practical system for testing, types of DGs, hybridizing of DG for reducing the intermittency, and other such precautions are not taken in order to make the results practically implementable. The load profile considered is static, which is not the practical scenario. Moreover, the test for futuristic load increase are not mentioned. The research in [37] offered a strategy for improving the process of DG resource placement and scaling. The technique considers both technical and economic criteria and uses suitable indices to assign monetary values to all of the variables. This strategy enhances the probability of findings convergence and, as a consequence, is suitable for use in real-world networks with several factors. However, this study does not considers the types of DGs, and hence, the associated intermittency related issues. It is also unable to discuss the addressing of varying loads over the time and load expansion in the future.

Based upon the discussion presented so far, it can be summarized that most researches considered the variety of optimization algorithms for a set of one or other variables. However, the practical variables are rarely considered. The practical variables include (1) availability of primary energy for renewable-based DGs, (2) intermittency in the generation of the renewables-based DGs, (3) the load variation in real power system, (4) availability of land for installation of DG plants, (5) testing and optimization upon actual power system, (6) cost of installation and connection of newly installed generation station with existing power system, (7) environmental impacts, (8) cost of purchase of electricity to mitigate the intermittency, (9) capital investment, (10) increasing the renewable factor, and (11) recovery of various costs translated as return on investment. The presented work is a novel study considering all these factors into single work.

This article examines the issue of insufficient resilience in hybrid power systems that include solar, wind, and hydro, with the goal of reducing blackouts and power outages in the AJK area. Power quality and economics of system design, as well as system size and system efficiency, take center stage in this setting. Thus, the initial objective is to develop an economically feasible system, followed by an analysis of the hybrid system’s socio-technical impact on the community. The main contributions of the proposed HRES are as follows:

• The optimization algorithm involved in this work is novel as it contradicts with the conventional approaches presented so far in the recent literature. The optimization is performed in two steps. The algorithm proposed for optimal location selection followed by finding optimal sizing. This work selects the optimal locations for placement of RES-based DGs based upon non-conventional variables which are in line with real world power system scenario. These include: (1) considering the availability of primary energy, (2) land for installing DGs, (3) costs incurred upon the installation as well as operation of the DGs, (4) power and energy loss minimization, and (5) availability of power grid. Afterwards, the sizes are optimally found using the HOMER Pro optimization algorithm which contains detailed information about the climate and weather as well as environmental factors. Such systematic and novel approach is missing in the literature, including the mentioned articles.
• The presented work is first of its kind that proposes the design of a viable cost-effective electrification modeling for AJK to make it self-sufficient in terms of energy demands by using indigenous renewable resources.
• This research also provides additional real-world performance optimization for hybrid systems and provides a guideline on energy-efficient, sustainable, and cost-effective design in other regions.
• Determine the effects of renewables on the environment and the economy of AJK.
• To utilize renewable energy sources, such as wind and solar, that can solve the issue of power shortages of region, allowing a continuous power supply.
• To reduce the use of fossil fuels and associated greenhouse gas, while boosting the use of renewable energy.

The paper is organized as follows: Section 2 explains the details about AJK region and site selection parameters, renewable potential, and load demand in AJK. HOMER-based HRES along with relevant mathematical details are presented in Section 3. The RES shortlisted in Section 3 are detailed in Section 4 for their cost and performance analyses as well as finalized component models and their description. Simulation of complete HRES in DIgSILENT PowerFactory and the techno-economic results are discussed in Section 5. Finally, the conclusion of the study is presented in Section 6.

2. Study Region

Azad Jammu and Kashmir is a self-governing region administered by Pakistan and has an area of about 13,297 sq. km. The region has diverse availability of sunlight in its southern parts whereas the central and northern parts being the lower part of Himalayas, receive plenty of winds, rain, and rivers. The region has extreme potential of renewable electrical generation due to extensive availability of primary energy. Although the electrical network is available in the region, some far regions are still not electrified. Moreover, the region faces the electricity shortfall of approximately 7 MW [38]. This study aims to compensate this shortfall in the electric power grid of AJK through optimal placement of RES. The region has seven electricity distribution substations, the locations for which are pinned in Figure 1.

2.1. Site Selection

This study aims to provide a detailed insight into the electrification potential of whole region of AJK through RES. The considered RES are wind, solar PV, and hydro. To make the proposed system most cost efficient, and to reduce the time for installation and starting operation, it has been emphasized that the locations for installation of RES-based HRES are selected based upon the factors including:

• Availability of primary distribution network.
• Availability of primary energy.
• Load concentration factor [39] in different regions.

The assessment of wind and solar resources are collected from Sustainable and Renewable Energy Development Authority of AJK [40], NASA meteorology [41], and accuweather [42]. The load profile for a community near selected location is obtained from respective grid stations.

Figure 2 shows the potential of PV in terms of average monthly temperature and wind speed availability in different regions nearby the available distribution substations in AJK.

The average monthly temperature in Muzaffarabad and Palandari districts is at its highest levels throughout the year. The monthly average solar Direct Normal Irradiance (DNI) for Palandri and Muzaffarabad regions is given in Table 1. Palandari also receives highest wind speeds for most time in the year. The potential for hydal energy is higher in Muzaffarabad due to availability of rivers Neelam and Jehlam, respectively.

2.2. Electrical Load Demand

Electrical load is categorized as residential, commercial, or industrial. In AJK, there is no industrial load, and the commercial load is relatively smaller with shorter activity time. In this work, total load demand is categorized as residential load only. The data regarding load demand is obtained from the electricity department of AJK. Currently, the peak demand for AJK is 246.2 MW, and the total active power in the network is almost 239 MW. According to the obtained data, AJK is facing an electricity shortage. The peak demand is recorded in the summer (July), with a maximum deficit of 7 MW. Load data at different distribution substations is shown in Figure 3. Districts Muzaffarabad and Palandari have highest load demands over the year.

Based upon the factors discussed, Muzaffarabad (34.3551° N, 73.4769° E) and Palandari (33.7145° N, 73.6860° E) are selected as potential sites for installation of HRES. Muzaffarabad and Palandari are termed as Loc-1 and Loc-2 in rest of this article, respectively. The average wind speed at selected sites is recorded to be 7–9 m/s [43]. The maximum temperature is recorded to be 39 °C and 35 °C in Muzaffarabad and Palandari, respectively [44]. The melting of snowfall (March–April) and monsoon rains (June–August) lead water flows to increase, reaching a high of 894 m³/s at Muzaffarabad [45].

Table 1. Monthly average Solar Direct Normal Irradiance (DNI) [9,19,45].

Months	Palandari (kWh/m2/day)	Muzaffarabad (kWh/m2/day)
January	3.480	2.950
February	4.450	3.540
March	5.500	4.550
April	6.200	5.880
May	6.700	6.590
June	7.600	7.460
July	6.700	6.600
August	6.500	5.940
September	6.100	5.700
October	5.800	4.890
November	4.400	3.690
December	3.700	2.790

Table 1. Monthly average Solar Direct Normal Irradiance (DNI) [9,19,45].

Months	Palandari (kWh/m2/day)	Muzaffarabad (kWh/m2/day)
January	3.480	2.950
February	4.450	3.540
March	5.500	4.550
April	6.200	5.880
May	6.700	6.590
June	7.600	7.460
July	6.700	6.600
August	6.500	5.940
September	6.100	5.700
October	5.800	4.890
November	4.400	3.690
December	3.700	2.790

3. HOMER-Based HRES Design

HRES is modeled using the Hybrid Optimization of Multiple Energy Resources (HOMER) program. It has been used in many studies throughout the world to evaluate design both on- and off-grid connected power systems. When compared to other similar software (MATLAB, PSCAD), it has some distinguishing features such as a broader scope of renewable resource input combinations under varying constraints, greater selection of system architecture and dispatch, simulation, optimization, and sensitivity analysis with limited input, while recommending the best system design [33]. It recommends the best-optimized model architecture based upon Net Present Cost (NPC). It replicates system functioning by doing energy balance calculations for a year. Afterwards, it calculates the cost of installing and running the system during the project’s lifetime [46].

The NPC is HOMER’s primary value against which all system configurations are ranked in the optimization findings, and the foundation for calculating COE. The NPC covers the capital cost (CC), replacement costs (RC), operation and management cost (O&M), fuel cost (if any), electricity purchase and sale from the grid ($P_{grid}$), salvage value (SV), and penalties associated with air pollution.

$$T_{NPC}=CC+RC+O\&M+SV+P_{grid}$$

(1)

Lowering COE while maintaining acceptable levels of reliability is the primary goal in the optimization task; COE is defined as estimated cost per kWh of usable electrical energy and can be calculated as.

$$COE=\frac{T_{NPC}\ast C_{RF}}{P_d}$$

(2)

The capital recovery factor ($C_{RF}$) is utilized to convert the present value into the equal annual cash flows and can be computed as follows:

$$C_{RF}=\frac{n{(1+n)}^y}{n{(1+n)}^y-1}$$

(3)

where n is interest rate and y is lifetime of project in years [31].

For a given HRES, a power deficit is inevitable when the total available power generation is less than the load demand. In such case, Loss of Power Supply Probability (LPSP) is a widely used index to quantify the stability of HRES on a yearly basis. LPSP is defined as the ratio of the total unmet load to the total electric load demand. It can be described by the following equation.

$$LPSP=\frac{{\sum }_1^T{P}_{unmet}\left(t\right)}{P_d}$$

(4)

$P_{unmet}\left(t\right)$: Unmet load in hour (h), $P_d$: electric load demand

$$P_{unmet}\left(t\right)=\left\{\begin{array}{cc}0& E_l\left(t\right)<E_{tot}\left(t\right)\\ E_l\left(t\right)-E_{tot}\left(t\right)& E_l\left(t\right)>E_{tot}\left(t\right)\end{array}\right.$$

(5)

$E_l$: Load demand at time t, $E_{tot}$: total available power generation at time t.

LPSP can either be 0 or 1 if no unmet load exists or there exists an unmet load, respectively, during the simulation period.

$$LPSP=\left\{\begin{array}{cc}0& P_{unmet}\left(t\right)=0\\ 1& P_{unmet}\left(t\right)>0\end{array}\right.$$

(6)

HOMER minimizes the $T_{NPC}$ and LPSP while selecting the optimal RES mix for given system and conditions, hence:

$$OF:min(T_{NPC},LPSP)$$

(7)

It is noteworthy that HOMER provides a list of various combinations of RES mix without considering the technical aspects of grid such as line flows, power losses, and voltage profile. Therefore, following constraints are considered while choosing the combination of RES from the list provided by HOMER. Final selection of RES mix for HRES is subject to these constraints.

$$P_{tot}\le (P_{PV}+P_{WT}+P_{Hyd}+P_{grid})$$

(8)

$$(RF=\frac{P_r}{P_d})>80\%$$

(9)

$$GHG=(F_{em}\ast F_{cons})+(F_{em}\ast G_{net})<lowest$$

(10)

$$EEand{G}_{net}<lowest$$

(11)

where: $P_r$ is power generated using renewable, $F_{em}$ is emission factor, $F_{cons}$ is the fuel consumption, $G_{net}$ are net grid purchases, EE is excess energy. The constraint in Equation (8) ensures that the total generation should not exceed the power shortfall in AJK. Equation (9) ensures increased renewable penetration factor, whereas Equations (10) and (11) attempts to reduce the GHG emission, excess energy, and grid purchases. In general, all these constraints attempt the cost-effective operation and generation from proposed system.

4. Cost and Performance Description of Selected RES

The component summary in terms of cost and performance is shown in

Table 2. Component description and cost.

Item	PV Panel [47]	Wind Turbine [48,49]	Hydro Turbine [18]
Manufacturer	Sharp-ND-500 W	Norvento nED100 (IEC-IIIA)	Kinetic Turbine (HKT)
Rated Power (kW)	0.5	100	1600
Initial cost (×${10}^3\$$)	1.5	40	500
Replace. cost (×${10}^3\$$)	1.2	30	250
Maintenance cost ($)	5	500	500
Lifetime (years)	25	25	20
Efficiency (%)	22	70	90
Pipe losses (%)	—	—	15
Operating Temp.	25 °C	—	—
Rated speed (m/s)	—	10.0	—
Cut-out-speed	—	20.0	—
Rated speed (m/s)	—	10.0	—
Rotor diameter (m)	—	22, 23	24

for the PV, wind, and hydro turbines.

4.1. PV Array

Sharp-ND-500WPV panel with rated power of 500 W is used in this work, selected based on efficiency. The capital and replacement costs are USD 1500 and USD 1200 per kW, respectively, whereas the O&M cost is USD 5 per year.

4.2. Wind Turbine

A NorventonED 24 (100 kW) wind turbine is used in this work, having a rated capacity of 100 kW. The lifetime is 20 years as per the manufacturer. The wind turbine power curve increases gradually with a wind speed starting from 0 m/s to 25 m/s. The power output starts to increase from 3 kW with a wind speed of 3 m/s to 100 kW with a speed of 10 m/s.

4.3. Hydro Turbine

The Kinetic Turbine (HKT) model with a rated capacity of 100 kW is selected in this work. The lifetime is 20 years as per the manufacturer. The hydro turbine power curve increases gradually with a water intake. The initial cost for the hydro turbine is USD 0.5 M with a replacement cost of USD 0.25 M. The operation and maintenance cost are USD 500 per year. The average water flow is set as 10 m³/s, the flow ratio is set at 50–105%, the hydro turbine efficiency is 85%, the available head is 24 m, and pipe losses are 15%.

4.4. Inverter

Power from PV needs to be converted from DC to AC using a bi-directional converter. Converters are extensively studied in recent years for enhancing their efficiency and power handling capacity [50,51]. In the simulation, generic converter is used, which has 95% efficiency and an initial cost of USD 1000 per kW.

4.5. Grid Extension

To enhance the cost efficiency of the proposed system, on-grid mode of operation is suggested in this work. For the current scenario, the purchase rate from the grid is set to be 0.1 USD/kWh and the sale rate is 0.05 USD/kWh. The break-even grid extension distance($D_{grid}$) is given as:

$$D_{grid}=\frac{(C_{NPC}{C}_{rf}-C_p{E}_d)}{(C_{cap}{C}_{rf}+C_{om})}$$

(12)

where: $C_{cap}$ is the capital cost of grid extension ($\mathrm{USD}/\mathrm{km}$); $C_{om}$ is the O&M cost of grid extension ($\mathrm{USD}/\mathrm{year}/\mathrm{km}$); $C_{pr}$ is the cost of power from the grid ($\mathrm{USD}/\mathrm{kWh}$); $E_{dis}$ is the total annual electrical demand (kWh/year) $C_{npc}$; and $C_{rf}$ is the total NPC of the standalone power system in USD.

5. Results and Discussion

5.1. Simulation System Design

The real power distribution network of AJK has been implemented and simulated in DIgSILENT PowerFactory. The system data is collected from National Transmission and Dispatch Company (NTDC), Pakistan [52]. Total active and reactive power demands are 246.2 MW and 58.21 MW, respectively. The transmission system operates at 132 kV whereas the load is served at 11 kV. Highest load appears at Muzaffarabad substation whereas Palandari substation serves the second highest load in AJK. According to the load flow, the maximum and minimum voltages appear at Muzaffarabad and Kotli substations, respectively. The PowerFactory model of AJK power distribution network is shown in Figure 4.

HOMER performs simulation with respect to the load and provides a mix of selected types of RES for the optimal HRES. The actual power deficit is assigned as an input load to HOMER. The finalized RES combination for HRES from HOMER are installed in PowerFactory network model. In this work, due to availability of national grid at most locations, battery backup installation is uneconomical. Furthermore, keeping in view the economic conditions of AJK, it is attempted to keep the initial capital cost as low as possible. Therefore, battery backup is not preferred, yet considered. However, it can be included in future. The HOMER HRES schematic is shown in this Figure 5 for both locations. DC/AC converters are included due to PV systems. The flowchart detailing the complete systematic set of steps involved in this study is given in Figure 6.

5.2. Analysis Case Studies

The key objective of the research is to meet the load deficiency of AJK with the maximum utilization of renewable and reduced GHG emissions. To analyze better, following cases have been formed and studied based on their NPC, COE, RF, and EE.

• Case 1: PV/DGen/Grid.
• Case 2: Wind/Hydro/DGen.
• Case 3: PV/Wind/Battery.
• Case 4: PV/Wind/Hydro/Grid.

HOMER optimization results for RES are analyzed in three categories: architecture, costs, and system variables such as renewable factor (RF) and excess energy (EE), as shown in

Table 3. Optimization of HRES in HOMER.

Architecture									Cost			System
Loc	PV (MW)	Wind (GWh)	Gen (MW)	Hydro (MW)	Batt (MWh)	Grid (MWh)		Conv. (MW)	COE ($/kWh)	NPC (M$)	IC (M$)	RF (%)	EE (%)
						Purchase	Sale
Case 1: PV/Dies-gen/Grid
Loc-1	40	—	2.4	—	—	5186.7	29304.4	9.57	0.152	8.6	14.9	85.2	55.9
Loc-2	11.67	—	1	—	—	6018.5	20552.4	16.38	0.046	2.69	8.41	81.9	28
Case 2: Wind/Dies-gen/Hydro
Loc-1	—	108.63	4.1	1.06	—	—	—	—	0.158	35.7	11.6	79.2	88.5
Loc-2	—	97.15	4.6	1.06	—	—	—	—	0.204	38.3	12.1	74	85.5
Case 3: PV/Wind/Battery
Loc-1	7.87	42.72	—	—	29.2	—	—	3.61	0.140	31.7	16.0	100	78.4
Loc-2	9.12	24.88	—	—	37.1	—	—	4.82	0.171	32.4	17.8	100	67.6
Case 4: PV/Wind/Hydro/Grid
Loc-1	0.70	9.87	—	1.06	—	1878.8	6678.1	0.61	0.00338	0.93	1.89	91.1	0
Loc-2	0.70	13.63	—	—	—	3402.8	5883.1	0.63	0.00815	2.69	1.60	81.8	0.103

. The system in case 1 has a COE of 0.198 USD/kWh. This system comprises PV of 51.667 MW and DGen (i.e., diesel generator) of 3.4 MW. The annual grid purchase and sale is approximately 11.2 GWh and 49.8 GWh, respectively. The model has a high NPC and IC of USD 11.29 M and 23.31 M, respectively. The system has the RF of 83.55% and EE of 41.95%.

The system in case 2 is more cost efficient due to its lower COE i.e., 0.362 USD/kWh as compared to case 1 but it has higher NPC and IC of USD 74M and 23.7M, respectively. Moreover, the EE of 87% in this case can cause system instability. The wind turbines in this case generate 205.7 GWh of energy annually. The DGen of 8.7 MW and hydropower plant of 2.12 MW are proposed. Due to higher generation from DGen, the RF in this case is lower with exact value of 76.6%.

Case 3 has the highest RF of 100%. Although, model has higher COE of 0.311 USD/kWh compared to the previous cases. The system comprises PV of 16.99 MW, wind turbines in this case generate 67.6 GWh of energy annually, battery backup of 66.267 MWh, and converter of 9.43 MW. The model has a high NPC and IC of USD 64.1 M and 33.8 M, respectively, with the EE of 73%. Due to higher NPC, IC and EE, this system has lesser preference like previous cases.

Case 4 has the lowest COE of 0.011 USD/kWh with the lowest NPC and IC of USD 3.62 M and 3.49 M, respectively. The system comprises PV of 1.4 MW, 55 wind turbines capable of generating 23.5 GWh annually, and hydropower of 1.06 MW. The annual grid purchases and sales of 5.2 GWh and 12.5 GWh are recorded, respectively. The system has highest RF of 86.45% and the least EE of 0.103% among all the cases. Add sentence about grid purchase and sale comparison.

By comparing all the available combinations, Case 1 has a low NPC and COE because of the availability of the grid. Case 2 has the highest NPC and COE because of the cost of wind turbines. Case 3 has a renewable factor of 100%; however, the NPC and COE are higher because of the cost of battery. Case 4 has the lowest NPC and COE among all the cases because there is no battery and diesel generator. Hence, case 4 is the most optimal HRES in meeting the load demand while improving all the parameters (NPC, COE, RF, and EE), and is taken for further analysis.

The impact of increased energy demand upon the NPC, COE, and RF has been observed in HOMER. If the energy demand increases from 75 GWh to 80 GWh daily, the NPC increases from USD 3.62 M to USD 5.67 M, COE increases from USD 0.011 to 0.015 and the RF decreases slightly to 84.45%. The analysis conducted above indicates that the hybrid generation mix in case 4 is the optimal hybrid system, generating 42.43 GWh of energy yearly. The wind turbines, PV, hydro, and grid contribute 59.45%, 10.13%, 18.45%, and 11.34%, respectively. Figure 7 illustrates the optimal HRES monthly average power production at a COE of 0.011 USD/kWh for Loc-1 and Loc-2, respectively.

If the demand rises from 75 MWh to 80 MWh per day, the COE increases from 0.011 USD/kWh to 0.015 USD/kWh, and the NPC increases from USD 3.62M to 5.67M.

5.3. Cost Analysis

A grid-connected system enables the system’s surplus energy to be returned to the grid. Furthermore, when renewable energy sources are not available, electricity from the grid is used to satisfy demand, while maintaining the system cost efficiency.

Table 4. Grid purchases and sales for case 4.

Loc	Purchase (MWh)	Cost ($\times {10}^3\$)$	Sales (kWh)	Revenue ($\times {10}^3\$$)
Loc-1	1878.8	187.8	6678.1	333.9
Loc-2	3402.8	340.2	5883.1	294.1
Total	5281.7	528.1	12,561.2	628.06
Saving			99960 $

summarizes the yearly purchases and sales of HRES at Loc-1 and Loc-2. For AJK, the annual saving of USD 99960 is expected.

As shown in

Table 5. Economic comparison for case 4.

Metric	Value
Present worth ($)	12,462,415
Annual worth ($/year)	840,080
Return on investment (%)	69.5
Internal rate of return (%)	74.2
Simple payback (year)	1.36
Discounted payback (year)	1.435

, the base case system chosen for economic comparison is DG. The difference in investment of USD 12.462 M between ordinary DGen and the proposed HRES (case 4) system may be repaid in a very short period of 1.36 years. This translates to a higher rate of return of 74.2% and a return on investment of 69.5%.

5.4. GHG Emissions

For meeting the shortfall of 7 MW, it assumed that the consumers use the DGen at domestic level. Such approach damages the environment due to emission of harmful GHGs. Two different scenarios are discussed for comparison sake. In first scenario, it is complete load is met through the generation from traditional diesel generators (DGen), while in second scenario, the proposed HRES is considered. The comparison in

Table 6. GHG emission comparison.

Emissions	DGen	Proposed HRES	Red. (%)
CO2 (kg/year)	6,036,196	1,883,715	68.8
CO (kg/year)	38,049	0	100
SO2 (kg/year)	14,781	8167	44.7
NO (kg/year)	35,743	3994	88.8
Unburned Hydrocarbons	1660	0	100
Particulate Matter	231	0	100
Total	6,126,660	1,895,876	69.1

demonstrates that the proposed HRES is an efficient way to alleviate the global pollution problem due to significantly lower amount of GHGs emission. Approximately 69%, 45%, and 89% reduction in CO₂, SO₂, and NO is expected annually, respectively. It is expected that the emission of CO, unburned hydrocarbons and particulate matter can be completely avoided. The overall reduction of 69.1% is expected annually.

5.5. Electrical System Performance Evaluation

AJK has seven major power stations that serve the entire area; the demand load for the year is shown in the Figure 8. February has the lowest load of 158.2 MW and July has the maximum load of 246.26 MW.

When renewable distributed generation was integrated with distribution lines with increased capacity, voltage profile improvement was observed. Power losses were reduced throughout the network. With the infeed of renewables into the network, the overall network losses reduced from 16.29 MW to 12.58 MW that accumulated to 22.7% reduction for peak load month of July, as shown in Figure 9. Distributed generation that are linked to the grid reduce the total load on the network by powering some of the loads locally [3]. Along with addressing the issue of power shortage with HRES, the proposed model has improved the voltage profile due to onsite injection of power demand [53,54]. As can be seen in

Table 7. Bus voltage (with and without RES).

Bus Bar	MZD (B1)	Nauseri (B2)	Hattian (B3)	Bagh (B4)	RWT (B5)	Palandari (B6)	Kotli (B7)
Pl (MW)	47.56	26.4	37.1	34.2	33.4	40.1	27.5
Ql (Mvar)	10.4	3.31	9.38	4.04	8.02	10.18	4.46
Bus Voltages and Line Currents Without RES
Voltage (kV)	130.9	130.9	124.5	124.5	124.3	124.1	124.0
p.u.	0.99	0.99	0.94	0.94	0.94	0.94	0.94
Bus Voltages and Line Currents With RES
Voltage (kV)	131.1	131.1	127.7	127.7	127.6	127.6	127.6
p.u.	0.99	0.99	0.97	0.97	0.97	0.97	0.97

, the bus voltage at the Kotli grid station was enhanced from 124.0 kV (0.94 pu) to 127.6 kV (0.97 pu) for given loads at different substations.

The line flows are the major cause of line losses, which have been reduced significantly by the proposed HRES. The major infeed to AJK is from Rampura substation in Punjab. The line flow data is shown in

Table 8. Line flows (with and without RES).

Line	Without RES I (kA)	With RES I (kA)	Red. (%)
Rampura-Hattian	1.082	0.719	33.5
Hattian-Bagh	0.530	0.500	1.88
Bagh-RWT	0.422	0.402	4.7
RWT-Palandari	0.230	0.210	8.69
Palandari-Kotli	0.110	0.101	8.18
Rampura-MZD	0.353	0.338	4.24
MZD-Nauseri	0.141	0.141	0%

. The line flow on the main line i.e., Rampura–Hattian has reduced by 33.5%. The line flows are also reduced on other lines such as by up to 8.7% on the Rawalakot–Palandari line and 8.2% on the Palandari–Kotli line. It is expected that the future load expansion can be met effectively due to reduced line flows.

5.6. Economic Viability of Proposed HRES

The cost of power losses in the transmission and distribution networks is commonly reflected in electricity tariffs. The longer the transmission lines, the higher the losses; hence: increased tariffs [55]. Therefore, it is believed that the distributed generation (DG), as proposed in this HRES, has the potential to cut utility costs while providing electricity more reliably and efficiently. Moreover, in modern electrical grids, the grid codes penalize the companies for increased emissions. The related costs of environmental penalties will be reduced if low-emissions renewable energy technologies are extensively installed. The cost and time for installation of RES-based generation facilities is comparatively lower; hence, such systems are preferred worldwide to meet the demand instantly. It is also worth noting that the investors can be incentivized for installation of proposed HRES to reduce the financial impact upon the government in terms of capital cost. DGs are independent of each other; hence, they can operate well in case of malfunctioning or failure of one or some of the DGs. Furthermore, if the concept of unbundled power system and energy markets is implied, the energy cost can be reduced significantly.

6. Conclusions

This study examined the long-term economic and technical feasibility of HRES-based on-grid electrification for meeting the power deficit in AJK. For providing the region with renewable energy, a PV/wind/hydro-based hybrid system is proposed in conjunction with the existing electrical grid. The following are the most important outcomes of this study:

• The selected sites have a tremendous potential for renewables. The climate, geography, and habitat provide the site with enough potential to exploit the available renewable resources.
• PV/Wind/Hydro mix can provide reliable electricity supply with 0% unmet power demand.
• Switching to an on-grid system architecture eliminates the need for batteries, reducing the various cost significantly.
• Compared to other HRES and standalone power system, the proposed system is cheaper with exceptional return on investment of 69.5%. The system is repaid in a noticeably short period of 1.36 years with a higher internal rate of return of 74.2%.
• With 0% surplus of power, the proposed system is extremely viable.
• The proposed HRES has a renewable factor of 86.45% and reduces overall GHG gases to 69.1%.
• With the proposed HRES, the total losses is reduced by 22.7% and total line infeed from external grid is reduced by 33.5%. Moreover, the overall voltage profile of all the network is improved as well.

The presented results highlight the potential for renewable-based electrical generation to overcome the electrical power deficit for whole region of AJK. Given AJK abundant energy resources, this concept can be utilized to power distant rural regions. For future electrification activities in AJK, the present research recommends an on-grid hybrid PV/wind/hydro system.