- Email: [email protected]
Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand
Energy xxx (2016) 1e11
Contents lists available at ScienceDirect
Energy journal homepage: www.elsevier.com/locate/energy
Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand Michael Emmanuel*, Daniel Akinyele, Ramesh Rayudu Smart Power & Renewable Energy Systems Group, School of Engineering and Computer Science, Victoria University of Wellington, Wellington 6140, New Zealand
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 June 2016 Received in revised form 28 October 2016 Accepted 19 November 2016 Available online xxx
This paper presents a performance analysis and economic viability of a 10 kWp grid-connected solar photovoltaic (PV) system installed at Maungaraki school, Wellington, New Zealand under the ”Dynamis Project”. The system consists of 40 panels and two units of 5 kW power converters with a communication capability while the distribution grid serves as a virtual storage. The excess power generated at low load condition during holidays and weekends is exported to the grid. With the system in operation since 2014, performance parameters based on International Energy Agency Photovoltaic Power System Programme (IEA PVPS), Clean Energy Council industry guide and IEC 61724 standard are evaluated. The final yield ranged from 1.1 to 4.9 h/d. The performance ratio (PR) varied between 76 and 79%, giving an annual PR of 78%. In addition, from the economic evaluation of this system at 4%, 6% and 8% discount rates, the levelized cost of energy are 12.1, 14.1 and 16.2 c/kWh respectively, with a simple payback period of 6.4 years. Overall, the total amount of power consumed annually from the grid reduced significantly by 32%. In monetary terms, the school's expenditure on power bills reduced effectively by 45% and was able save about NZD $4700 in 2014. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Photovoltaic Grid-connected Final yield Performance ratio Levelized cost of energy
1. Introduction The recent energy transition as a result of the growing presence of renewable energy resources (such as wind power, solar energy and hydropower) in the conventional power system is pivotal in dealing with climate change, enhancing power network resilience and provision of new economic opportunities [1e3,49,50]. One of the most common renewable energy resources is the photovoltaic (PV) system such as the rooftop residential and utility scale installations [4]. However, hydroelectric generation supplies the bulk of New Zealand's electricity demand, geothermal contributes about 17.2% while wind generation is about 5.4% of the country's power generation [5]. In a recent quarterly electricity generation report, solar power contributed 0.1% (as shown in Fig. 1) of the country's net electricity generation of about 42927.54 GWh at the end of 2015 [5]. The contribution of solar photovoltaic (PV) systems is quite insignificant when compared to the deployment of other renewables in New Zealand. However, in recent times, there is a wide-
* Corresponding author. E-mail address: [email protected] (M. Emmanuel).
spread deployment of rooftop grid-connected PV systems in residential apartments and schools as reported in this article. The proliferation of such systems on the electricity grid highlights the need to understand its operation through monitoring and thorough performance analysis [6]. In addition, utility interactive PV systems are essentially monitored to evaluate the final energy yield, detect possible design defects and avoid economic losses as a result of operational issues [6]. Another reason to carry out such analysis is to assess PV credibility, viability and ultimately to increase their penetration within the existing electric power network [7]. The IEA PVPS Task 2 has been able to analyse and publish 170 grid-connected PV systems installed in various countries of the world [8]. In addition, over the previous years, various authors from different countries have published results of the performance analysis of their respective grid-connected PV units. Pietruszko et al. [9] evaluated the performance of a 1 kWp a-Si PV system located at Warsaw, Poland. The performance ratio ranged from 0.6 to 0.8, annual system yield was 830 kWh and the efficiency of the PV system was in the range of 4e5%. Mondol et al. [10] presented the outcome of the performance analysis of a 13 kWp m-Si PV system installed in Northern Ireland. The evaluated annual final yield ranged from 1.61 to 1.76 h/d, with a PR which ranged from 0.6
http://dx.doi.org/10.1016/j.energy.2016.11.107 0360-5442/© 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
2
M. Emmanuel et al. / Energy xxx (2016) 1e11
Nomenclature
hsys l ma Si
a Si Aarray AC DC GI GI;ref h=d Lc Ls LSP LCOE
system efficiency power temperature coefficient(0:45%/ C) micromorph silicon amorphous silicon area of the array (m2 ) alternating current (A) direct current (A) in-plane solar irradiation (kWh=m2 ) reference irradiance (kW=m2 ) hour/day array capture losses system losses capacity factor levelized cost of energy
Fig. 1. Electricity generation from various energy sources in New Zealand [5].
to 0.62 and annual average system efficiency of 6.4%. Also, Chokmaviroj et al. [11] evaluated the performance of a 500 kWp gridconnected PV plant at Mae Hong Son province, Thailand. The plant was divided into two, 250 kWp, from a double glazed semi-Si PV modules. The plant generated about 383274 kWh and the efficiency of the PV array ranged from 9 to 12%. The final yield ranged from 2.91 to 3.98 h/d and the PR ranged from 0.7 to 0.9. Another performance study is that of Kymakis et al. [12]. The study evaluated a p-Si 171.36 kWp utility interactive PV park on the island of Crete, which was as a result of the favourable climatic condition and the recent incentivization of PV system installations in Greece. The yearlong evaluated PR was 0.67 and the final yield ranged from 1.96 to 5.07 h/d. Cherfa et al. [13] carried out an analysis of a mini-grid connected m-Si 9.54 kWp PV system which was a pilot project with the primary aim of acquiring experience in the design, monitoring and maintenance of such innovative technology in Algeria. The system performance was quantified over the monitored period which showed an annual 10981 kWh of energy injected into the grid. The average daily output energy was 30 kWh and PR ranged from 0.62 to 0.77. Ayompe et al. [14] presented results obtained from monitoring a m-Si 1.72 kWp rooftop gridconnected PV system in Ireland. Within the frame of the study, the monthly average daily capacity factor ranged from 5 to 15.5%, with an annual average of 10.1%. The PR and final energy were 0.82 and 2.4 h/d respectively. Another relevant study is that of Okello et al. [15] in South Africa. They presented the analysis of p-Si 3.2 kWp grid-connected PV system with a PR of 0.84 and final
m Si NPV p Si PO
mono crystalline silicon net present value poly crystalline silicon rated DC output power of the array under standard test conditions (Wp) PR performance ratio semi Si semi crystalline silicon TA ambient temperature ( C) TC cell temperature ( C) TR temperature rise for parallel-to-roof installation ( C) TSTC cell temperature at standard test conditions ( C) TLCC total life cycle cost YA array yield (kWh/kWp) Yf final yield (kWh/kWp) Yr reference yield (kWh/kWp)
energy yield of 4.9 h/d. Performance analysis of different utility interactive PV sizes have been conducted across various locations in India. For instance, Sharma and Chandel [16] investigated the performance of a p-Si 190 kWp PV plant connected to the grid in Punjab, northern India. The final yield, reference yield and PR ranged from 1.45 to 2.84 h/d, 2.29e3.53 h/d and 55e83% respectively. The annual average PR, capacity factor and system efficiency were found to be 74%, 9.27% and 8.3% respectively. Also, Padmavathi et al. [17] presented an analysis of a m-Si 3 MWp grid-connected PV plant in Karnataka, south western region of India. The annual average energy generated was 1372 kWh per kW of the installed capacity and the PR was 0.7. The annual average reference yield and final yield were 5.36 h/ d and 3.73 h/d respectively. In addition, Shiva Kumar and Sudhakar [18] analysed the performance of a p-Si 10 MWp PV plant connected to the grid at Ramagundam, southern India. The plant is one of the largest solar power plants with a seasonal tilt. Results revealed that the capacity utilization factor was 17.68% with an annual energy generation of 15 798.192 MWh. The final yield ranged from single-phase1.96e5.07 h/d, and annual PR was 0.86. Sundaram et al. [19] evaluated the performance analysis of a thin film a-Si 5 MWp grid-connected PV plant at Sivagangai district in Tamilnadu, south India. The annual actual generation was 8495.29 MWh with a system efficiency of 5.08%. The final yield, PR, average energy and exergy efficiency were found to be 4.81, 0.89, 6.08% and 3.54% respectively. Edalati et al. [20] presented a comparative performance analysis of m-Si (5.52 kWp) and p-Si (5.52 kWp) for utility interactive PV systems in dry climates. The annual average capacity factor for m-Si and p-Si were found to be 23.20% and 23.81% respectively. The annual final yield and PR for p-Si were 5.38 h/d and 0.83 respectively. Moreover, the annual final yield and PR for m-Si were 5.24 h/ d and 0.81 respectively. Sidi et al. [21] evaluated the performance analysis of the first utility-scale 15 MWp grid-connected PV plant in Mauritania, which has a high annual insolation between 1900 and 2200 kWh=m2 /year. The PV plant from a-Si/m a-Si was connected to the 33 kV electricity grid through inverters and nine transformers. Performances of two arrays from two different technologies, ma Si and a-Si were presented. The daily average array yield, PR, and capacity factor for ma Si were 4.38 h/d, 0.68 and 17.75% respectively. Moreover, the daily average array yield, PR, and capacity factor for a-Si were 4.79 h/d, 0.75 and 19.54% respectively. These performance metrics allow cross-comparison between different PV systems in terms of design, technology and diverse
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
M. Emmanuel et al. / Energy xxx (2016) 1e11
climatic conditions [7]. Also, the evaluation of these parameters help to assess product quality, determine future needs and detect failures in system components [7,8]. For instance, with the PR metric, system component malfunction such as inverter failure can be easily detected for utility interactive PV systems. In addition, other factors which could affect the PR include PV module soiling, shading and the ratio of the measured array efficiency to the nominal array efficiency [7,8]. Also, properly maintained PV systems have a high propensity to operate optimally with high PR value and availability [8]. Additionally, the analysis of the economic viability of utility interactive PV systems is very crucial in enhancing PV technology uptake as favourable renewable energy policies continue to evolve globally. Adaramola [22] presented the viability of grid-connected PV system in Jos, Nigeria. The study showed that the 80 kWp system was able to contribute 40.4% of the annual electricity demand with LCOE of $0.103/kWh. Mondal et al. [23] presented the financial viability of a proposed 1-MWp grid connected PV system in Bangladesh. For a project lifetime of 15 years, Mondal et al. reported energy production cost between $0.253/kWh and $0.282/kWh, which decreases with the increase of lifetime. Also, El-Shimy [24] presented a viability analysis of a 10 MWp grid-connected PV plant for 29 different sites in Egypt. The energy production cost ranged from $0.1989/kWh to $0.2424/kWh, and the equity payback varied between 4.9 years and 7.1 years. Adaramola [25] analysed the economic viability of a rooftop 2.07 kW grid-connected PV system in Norway with a feed-in-tariff of $0.356/kWh over 25 years of project lifetime. This resulted in $0.110/kWh premium over the LCOE of $0.246/kWh produced by the PV system. Edalati et al. [26] presented a techno-economic analysis of a 10 MWp utility interactive PV plant in Iran. The LCOE ranged between $0.1992/kWh in the south-eastern part of Iran to $0.3838/kWh in the northern part. Also, the sensitivity of NPV and payback time to six various annual inflation rate showed that inflation rate increase would lead to an elongated payback time and lower NPV. Orioli et al. [27] presented an economic analysis of a utility interactive PV installation using the ratio of NPV to the present value cost (NPV/PVC). A positive NPV/PVC value represents a profitable investment, while a negative value indicates that the revenue is than disbursements and a near zero value signifies a critical situation where little modifications in costs and benefits assessments abruptly impact the economic analysis. However, performance analysis of grid-connected PV systems in New Zealand are not available in the literature which is pivotal in understanding and quantifying their impacts on the traditional electricity network as an emerging alternative renewable power generation resource. Also, in order to develop the PV system as a sustainable energy resource and increase its uptake, it is important to assess its economic viability. This paper aims to fill these gaps by presenting a techno-economic analysis of a grid-connected 10 kWp PV system at Maungaraki school, Wellington in New Zealand as a case study. We have carried out our evaluation based on performance parameters specified by IEA PVPS and IEC 61724 standard for utility interactive PV systems. In addition, the Clean Energy Council (CEC) guide which represents industry best practice for the design and installation of grid-tied PV systems has been used for analysis in this study. Final system yield, energy yields, capacity factor and performance ratio are evaluated. Also, the economic viability of the PV system under consideration has been evaluated using metrics such as the NPV, LCOE and simple payback period. In addition, we present results from the Web Based PV System Monitoring and Reporting (WBPSMR) for the PV array.
3
2. Description of the grid-tied solar PV system The installation of the 10 kWp PV array was carried out under the ”Dynamis Project” with the aim of testing renewable energy technologies effectiveness and economic viability using schools as platforms in order to sensitize surrounding communities to become more energy self-sufficient [28]. The renewable energy for New Zealand schools under this project has successfully installed 10 kWp solar PV in two different schools. One of the installations was carried out at Maungaraki school (with a roll of about 250 pupils), as an attempt to pioneer a sustainable energy revolution. The school is located at a longitude of 174:9 E and latitude of 41:16 S, with a panel tilt angle of 41 . The PV panel tilt angle and the latitude of the corresponding site location are kept equal in order to obtain maximum solar radiation [29,30]. There are two identical parallel strings consisting of 20 panels each with modules rated at 250 Wp capacity, and tied to the grid via two 5 kW Enasolar single-phase inverters (as shown in Fig. 2). The energy meter is used to measure the amount of energy consumed and the excess power generated and exported to the grid which usually occurs during minimum load conditions that happens over weekends and holiday periods. The specification of the TNS250 module used is given in Table 1. In addition, the PV array was installed in an open space on a rooftop free of shadows or shading which can affect its performance. The PV panels are regularly maintained every six months to get rid of dust, birds' drops and other forms of impurities. Also, the cost of setting up the project was about NZD $28000 sourced through grants and donations. 2.1. Inverter with built-in communication interface The power converter has built-in cable (Ethernet) and wireless interfaces (as shown in Fig. 3) for web based PV system monitoring and reporting, which shows the energy production information such as daily and monthly output power, input and output voltages [31]. The inverter with a static or dynamically assigned internet protocol (IP) address can be connected to the existing home network via any of the available interfaces. Also, the inverters deployed have two independent maximum power point tracking PV inputs which can be connected to one solar string or two identical parallel solar strings [31]. In addition, the inverter has integrated lockable DC and AC switches for isolation purposes. The most relevant technical specifications of the inverter are presented in Table 2.
Fig. 2. Schematic block diagram of the PV system.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
4
M. Emmanuel et al. / Energy xxx (2016) 1e11
reference irradiance [8,32,34]. Therefore, the daily Yr is calculated as:
Table 1 PV module specifications. PV module
Specifications
Type of cell material Make Model Maximum power (Pmax ) Open circuit voltage (Voc ) Short circuit current (Isc ) Voltage at maximum power (Vmp ) Current at maximum power (Imp ) Number of cells in a module
Monocrystalline TNS Solar TN-60-6M 250 Wp 37.9 V 8.64 A 30.4 V 8.23 A 60
Z Yr ¼
24 t¼1
GI dt
GI;ref
(1)
where: GI ¼ total irradiation (kWh=m2 ) GI;ref ¼ reference irradiance (kW=m2 ). 3.2. PV array energy yield The ratio of the energy produced by the PV array to the rated PV capacity is referred to as the array yield, YA . Losses such as the ones due to manufacturing tolerance, temperature, dirt and dust are taken into consideration in estimating YA [8,32]. It is therefore, calculated as:
YA ¼
EA ½kWhDC PO ½kWpDC
(2)
3.3. Final system yield The final system yield, Yf , is the ratio of the net energy produced (ENET ) by the PV array to the rated DC array capacity (PO ) [32]. Also, it is the amount of energy supplied to the load per day (month or year) considering array capture and system losses. It quantifies the duration (e.g., yearly, monthly or daily) required by the PV array to operate at the rated DC power to supply an equal amount of energy and estimated as [7,12]:
Yf ¼
ENET ½kWhAC PO ½kWpDC
(3)
3.4. Performance ratio
Fig. 3. Inverter with built-in communication interface [31].
This is a dimensionless quantity used to indicate the monthly or annual impact of PV system losses on the rated array capacity. The system losses are as a result of PV array temperature, failure of system components and incomplete usage of the irradiation [32]. In addition, the performance ratio, PR, does not depend on location of PV installation and system size [8]. It is used to evaluate the quality of PV system installation and estimated as [35]:
3. System performance analysis This article considers the evaluation of the operational and reliability performance of the grid-connected PV system based on the IEC standards (61836 and 61724) and reports from International Energy Agency (IEA) photovoltaic power systems program (Task 2) [8,32,33]. The performance analysis of grid-connected PV systems (as shown in Fig. 4) is pivotal in assessing their operational performance under different climatic conditions and detection of operational issues [18]. It also enables the measurement of longterm variation in system performance and comparisons with other systems that differ in location, design and technology [7].
3.1. Reference yield The reference yield, Yr , is a function of in-plane irradiation and estimated as the ratio of the total irradiation per day (or year) to the
PR ¼
PV system actual energy yield Yf ¼ PV array ideal energy output Yr
(4)
3.5. Losses System losses, classified into DC and AC subsystems losses, are pivotal design considerations in providing realistic energy solutions [36]. However, in this study, losses due to shading or shadows is not considered because the location of the PV system is free of shadows during the day. The connecting cable (PV DC main cables) between the PV array and the inverter subsystems can lead to power loss which should be accounted for. For our case study, the DC cable losses is 3% which amounts to 0.97 de-rating factor [7,35]. However, AC subsystems losses is as a result of the connecting cable between the inverter
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
M. Emmanuel et al. / Energy xxx (2016) 1e11
5
Table 2 Electrical properties of the inverter. DC Input Number of inputs Max. open circuit voltage (Voc ) DC full power operating range Operating voltage range (Vmpp ) DC optimal operating voltage Max. input current (Impp ) Max. short circuit current (Isc ) Maximum usable input power (Pmax ) Maximum allowable input power Reverse polarity protection AC Output Nominal output voltage Output voltage range Output power (@ 50 Hz) Max. output current Power factor Max. efficiency Max. Euro. efficiency General system data Data interface Weight Operating temperature range Night-time consumption
2 Independent MPPT inputs 600 per DC input 235e500 V per DC input 120e500 V per DC input 350 V per DC input 15 A per DC input 16 A 3500 Wp per DC input 7000 W Inherent crowbar diodes 230 V AC single phase 202e259 V AC (New Zealand) 4990 W 21.5 A >0.98 >96.8% >95.4% IEEE 802.11 (Wi-Fi)/Ethernet 20 kg 30 C to þ50 C (full power), þ70 C (de-rated) <1.2 W
power produced by the inverter (PInv;AC ) to the DC power generated by the PV array unit (PPV;DC ) [17,18].
hinv ¼
PInv;AC PPV;DC
(8)
3.8. System efficiency The monthly system efficiency, hsys , is given as [16,39]:
Fig. 4. System performance analysis.
and the grid with a typical value of 1% and de-rating factor of 0.99 [7,35]. These two losses are components of the system losses, Ls , and estimated as:
Ls ¼ YA Yf
ENET Girrad Aarray
(9)
here, ENET is the monthly total AC PV energy output, Girrad is the monthly peak sun hour, and Aarray is the area of the array.
(5)
The array capture losses, Lc , refers to the normalised PV system losses which is as a result of PV array energy losses such as conversion losses, manufacturing tolerance and dirt [32,37]. In addition, Lc is calculated as:
Lc ¼ Yr YA
hsys ¼
(6)
3.9. Energy yield Energy yield of a PV array is a function of the meteorological data of the location where it is installed [40]. This is the PV system energy output estimated as [35,41,42]:
ENET ¼ PratedSTC mtemp mman mdirt Girrad hinv hpvinv hinvsb (10)
3.6. Capacity factor The capacity factor, LSP , is the defined as the ratio of the energy output from the PV system to the product of the operating duration and rated PV array output [38]. LSP is calculated annually as [12]:
LSP ¼
Yf ENET ¼ PO 8760 8760
(7)
3.7. Inverter efficiency The inverter efficiency, hinv , is defined as the ratio of the AC
where: ENET ¼ PV system energy yield (Wh). PO ¼ rated DC output power of the array under standard test conditions (Wp). mtemp ¼ de-rating factor due to temperature. mman ¼ de-rating factor for manufacturing tolerance (0.97). mdirt ¼ de-rating factor for dirt (0.97). Girrad ¼ irradiation value or peak sun hour (kWh=m2 ) hinv ¼ efficiency of the inverter. hpvinv ¼ efficiency of the sub-unit (connecting cables) between the PV unit and the inverter (0.97). hinvsb ¼ efficiency of the sub-unit (connecting cables) between the inverter and the switchboard (0.99).
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
6
M. Emmanuel et al. / Energy xxx (2016) 1e11
3.9.1. Temperature de-rating factor The de-rating due to temperature for this study is done in accordance with guidelines provided by the Clean Energy Council (CEC) for grid-connected solar PV systems without battery storage [41]. The cell temperature TC for a grid-connected PV is estimated as:
TC ¼ TA þ TR
(12)
where: l ¼ power temperature coefficient/ C (0:45%/ C) TSTC ¼ cell temperature at standard test conditions ( C)
The enhancement of the cost-competitiveness of the solar PV system over the years continues to drive its uptake as a viable power generation alternative in the global energy mix. In 2014, prices of PV modules decreased by 75% in comparison with their prices at the close of 2009 and as a result, improving PV system plug parity [43]. Important measures used in the economic evaluation of the PV system considered in this article are as follows: 4.1. Net present value (NPV) The NPV is a financial tool used to evaluate cash outflows and revenues, and investment characteristics and decisions especially for comparing mutually exclusive projects [44,45]. Also, it is the algebraic sum of the net cash flows over the project's life time to the present, discounted by an appropriate discount rate [46]. It is given as: N X m¼0
CFm ð1 þ rÞm
N m¼1
TLCC ¼
N X m¼0
TLCC
Em ð1 þ rÞ m
ICm ð1 þ rÞm
(14)
(15)
where: TLCC ¼ total life cycle cost. ICm ¼ investment cost in period m. Em ¼ energy output or saved in year m. In addition, LCOE which considers the current dollar value is known as the nominal LCOE while the real LCOE is a fixed dollar inflation-adjusted value. The real LCOE uses a discount rate without taking into account inflation rate and the nominal LCOE applies a discount rate with the inflation rate [45]. 4.3. Simple payback period
4. Economic analysis
NPV ¼
LCOE ¼ P
(11)
where: TA ¼ ambient temperature ( C) TR ¼ temperature rise for parallel-to-roof installation (35 C) However, TR is a function of PV array installation type which could be top-of-pole, parallel-to-top or rack-type mount [41]. The parallel-to-top array frames were deployed on site used for our case study with a TR of 35 C as recommended by CEC. With the estimated value of TC , the de-rating factor due to temperature, mtemp , is given by:
mtemp ¼ 1 þ fl*ðTC TSTC Þg
that reflects the average capital cost [43]. This metric is used to compare the cost of energy generated by a renewable technology with that of a traditional fossil fuel generating unit [45]. The LCOE is given as:
(13)
where: CFm ¼ net cash flow in year m. N ¼ analysis period. r ¼ annual discount rate. The NPV criteria include [45,47]: If the NPV is positive, the investment is economical. Negative NPV denotes the return are worth less than the initial investment. The financial viability of a project is uncertain with Zero NPV. Amongst independent projects, the higher the NPV value the better.
4.2. Levelized cost of energy (LCOE) This is the ratio of the lifetime costs to the lifetime electric power generation, both of which are discounted back to a base year
The simple payback (SPB) period is another pivotal financial tool used to estimate the number of years it will take to recover the project cost of an investment made [45]. Also, it is important to note that the SPB does not consider the time value of money and with a zero discount rate. The SPB is given as the first point in time when the following expression is satisfied [45]:
X m
DIICm
X
DSm
(16)
m
where: DIICm ¼ incremental investment costs at zero discount rate in period m. DSm ¼ sum total of the annual cash flows net annual costs at zero discount rate in period m. 5. Results and discussion This section presents the performance analysis of the gridconnected 10 kWp (40 250 Wp) PV system. We present results from our analysis, the National Renewable Energy Laboratory's (NREL) system advisory model (SAM) and Web Based PV System Monitoring and Reporting (WBPSMR). In order to monitor the behaviour of the PV system, it is pivotal to study the meteorological data recorded in the weather station covering the site location. Therefore, we have used full data set from National Institute of Water and Atmospheric research (NIWA) solarview tool [48] for our analysis. Fig. 5 shows the monthly averaged total in-plane irradiation and ambient temperature of the installation site. Significant peaks of in-plane insolation also known as Peak Sun Hour (PSH) occur during the summer period while the PSH dropped drastically during winter. Furthermore, the monthly average daily wind speed varied between 0.5 m/s in most parts of the months and 21.6 m/s in October. The relative humidity ranged from 38 to 97% and the maximum global irradiance was 1045W=m2 . In addition, Fig. 6 shows the module temperature as it varies across the entire year, with high temperatures during the summer period. Our analysis as depicted in Fig. 7 shows an interesting pattern of the monthly capacity factors and performance ratios over the entire year. The monthly capacity factors increased proportionately with the in-plane solar irradiation with an annual average value of 12.5%.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
M. Emmanuel et al. / Energy xxx (2016) 1e11
7
Fig. 5. Monthly averaged total in-plane irradiation and ambient temperature. Fig. 8. Monthly averaged daily final yield, PV array capture losses and system losses.
Fig. 6. Annual module temperature.
summer period and higher in winter [7]. In addition, the system and array capture losses increased as the in-plane insolation increases. Fig. 8 depicts the monthly mean daily final yield, array capture and system losses. The monthly average daily array yield ranged from 1.2 h/d (June) to 5.4 h/d (January), while the final yield varied between 1.1 h/d (June) to 4.9 h/d (January). The average annual final yield and reference yield were 2.99 h/d and 3.87 h/d respectively. The monthly averaged daily array capture losses ranged from 0.19 (June) to 1.14 h/d (January) while the system losses varied from 0.09 (June) to 0.42 h/d (January). The PR of the considered PV array experienced a slight variation within the range of 76e79%, and the annual average value was 78%. This is comparable to the range of values (0.6e0.8) reported by IEAPVPS Task 2 for grid-connected PV systems [8]. The monthly inverter efficiency ranged from 94.9% to 95.7% with higher values during the high PSH as shown in Fig. 9. In addition, the monthly system efficiency ranged from 11.71% to 12.19% with higher values during low in-plane solar insolation as shown in Fig. 9. The average annual system efficiency of the PV array is 11.96%. In addition, the 10 kWp PV array generated an average energy output of 1298 kWh during the summer period ranging from 978 kWh to 1546 kWh. During low PSH values in winter, the energy output ranged from 322.8 kWh to 816.3 kWh. From our analysis, the annual average energy output was 910.13 kWh. However, from the WBPSMR, the average energy output was 1058.33 kWh, with the monthly energy output from the two 5 kW inverters shown in Fig. 10. Low energy output in January was because the site was commissioned at the middle of the month. The two strings of the PV array are mounted in an open space close to the school's football field, and the amount of dust collected over
Fig. 7. Monthly capacity factors and performance ratios of the PV array.
The availability of a high solar resource as witnessed during summer period, led to increase in capacity factors (shown in Fig. 7) and the final yield (shown in Fig. 8). However, high values of PSH does not necessarily translate to high performance ratios as shown in Fig. 7. This is traceable to the impact of temperature losses on the overall performance of the PV array. As the temperature of the location increases, the PR values also dropped accordingly. Therefore, PR values are lower during
Fig. 9. Monthly system and inverter efficiencies.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
8
M. Emmanuel et al. / Energy xxx (2016) 1e11
Fig. 10. WBPSMR monthly PV output energy.
them varies, which causes variation in the array and inverter output as depicted in Fig. 10. Also, the daily power generated is shown in Fig. 11.
6. Comparison of PV systems performance Performance analysis metrics allow cross-comparison between different PV systems operating under diverse climatic conditions [7]. Table 3 shows performance of various utility interactive PV systems across different locations. The annual average final yield of the m-Si modules in this study is higher than the ones reported in Ireland and Northern Ireland. However, it is lower than the ones reported in Iran and India. Also, the annual average PR in this study falls within the range (0.25e0.9) reported by IEA PVPS Task 2 for 170 grid-connected PV systems in different countries of the world [8]. In addition, the PR for m-Si technology is lower than that of p-Si for most of locations with exception of Greece which ranged from 0.58 to 0.73.
7. Economic analysis The computation of the economic measures described is presented in this section using the NREL's SAM financial model. Although there is no feed-in tariff legislation for the PV system in New Zealand, its economic viability especially for schools is shown in this paper. A number of assumptions and inputs required in determining the financial metrics for the 10 kWp PV system are presented in Table 4. Also, the monthly energy consumption and peak demand are shown in Fig. 12. The total project cost was NZD 28000 (USD 19600) which included the system, balance of equipment (BOS), installation, installer margin and overhead costs. Therefore, the total installed cost per capacity was USD 1.96/Wdc. The cash flow is presented in Table 5 and results obtained are given in Table 6 with three discount rates. The discount rate has significant impacts on the NPV and LCOE, with 4% rate giving the highest NPV and lowest norminal LCOE values which makes it an appropriate rate for such a system. Fig. 13 shows payback cash flow with a simple payback period of 6.4 years.
Fig. 11. WBPSMR daily power generated.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
M. Emmanuel et al. / Energy xxx (2016) 1e11
9
Table 3 Grid-connected PV system performance evaluation in various locations. Location
Capacity
PV Technology type
PR
Yf (h/d)
Warsaw, Poland [9] Ballymena, Northern Ireland [10] Mae Hong Son Province, Thailand [11] Island of Crete, Greece [12] Algiers, Algeria [13] Dublin, Ireland [14] Khatkar-Kalan, India [16] Karnataka, India [17] Port Elizabeth, South Africa [15] Kerman, Iran [20] Kerman, Iran [20] Ramagundam, India [18] Wellington, New Zealand [Present study]
1 kWp 13 kWp 500 kWp 171.36 kWp 9.54 kWp 1.72 kWp 190 kWp 3 MWp 3.2 kWp 5.52 KWp 5.52 kWp 10 MWp 10 kWp
a-Si m-Si semi-Si p-Si m-Si m-Si p-Si m-Si p-Si m-Si p-Si p-Si m-Si
0.6-0.8 0.6e0.62 0.7e0.9 0.67 0.62e0.77 0.82 0.55e0.83 0.7 0.84 0.81 0.83 0.86 0.78
e 1.61e1.76 2.91e3.98 1.96e5.07 e 2.4 1.45e2.84 3.73 4.9 5.24 5.38 1.96e5.07 1.1e4.9
Table 4 Inputs and assumptions used in the analysis. System size (kWp) Operation and maintenance cost (Fixed cost by capacity) ($/kW-yr) System salvage value ($) Panel tilt (degrees) Panel azimuth (degrees) Annual panel degradation (%/year) Inflation rate (%/year) Real discount rate (%/year) Analysis time period (years) Load profile used Grid buyback rate (c/kWh) electricity retail rate (c/kWh) Corporate tax rate
10 20 0 30 0 0.5% 2.5% 4%, 6% and 8% 25 school load profile 15 25 NA
Fig. 13. Payback cash flow (simple payback period ¼ 6.4 years).
In addition, the electricity to/from the grid and electricity sales/ purchases across the entire year are shown in Figs. 14 and 15 respectively. During periods of low insolation, significant amount of electricity was purchased from the grid, while excess power was exported to the grid during periods with high PSH and minimum load.
Fig. 12. Monthly energy consumption and peak demand.
Table 5 Cash flow. Metric
Value
Annual energy (year 1) Energy yield (year 1) Electricity bill without system (year 1) Electricity bill with system (year 1) Net savings with system (year 1) Net capital cost Equity Debt
16158 kWh 1616 kWh/kW $8310 $5229 $3082 $19600 $19600 $0
Table 6 Economic measures evaluation of the 10 kWp system. Discount rate Real LCOE (c/kWh) Nominal LCOE (c/kWh) NPV ($)
4% 9.43 12.1 22000
6% 11.19 14.1 14600
8% 13.09 16.2 9100
7.1. Actual economic analysis The actual financial analysis of the power consumption before and after the installation of the 10 kWp PV system is presented in this section. The annual cost of the school's power consumption before the PV system integration was NZD $8380.86. After the interconnection, the school's energy usage cost reduced to NZD $3758.77. The consumption pattern (obtained from the data logging system) of the school is shown in Fig. 16 for pre- (without PV-2013) and post-installation (with PV-2014) of the PV system. It shows a dramatic drop in kWs used and amount of money spent on power bills from the utility. In addition, as shown in Fig. 16, the power consumed after the PV system installation represents the amount of power consumed minus the amount of excess power generated in a particular month. The excess power export to grid usually
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
10
M. Emmanuel et al. / Energy xxx (2016) 1e11
reduced significantly by 32% and 45% respectively at the end of the year 2014. In monetary terms, the school saved about NZD $4700 in 2014 on power bills, which matches closely with the payback period given by the SAM's financial model. 8. Conclusion In this study, the performance analysis of a 10 kWp gridconnected PV system installed at Maungaraki school, Wellington, New Zealand is presented. From our analysis, the main conclusions are given as follows:
Fig. 14. Annual electricity to/from the grid.
Fig. 15. Annual hourly electricity sales/purchase with/without PV system.
occurs at low load condition which happens during weekends and holidays. Also, the spike between November and December is due to readings carried over to the next month. Furthermore, from our analysis the total amount of power consumed and cost of power
The final yield, (Yf ), of the PV system ranged from 1.1 to 4.9 h/d, with an annual average value of 2.99 h/d. There is a direct proportional relationship between the peak sun hour (PSH) and Yf . Also, the array capture and system losses increased with high values of PSH. The availability of a high solar resource led to increase in the capacity factor as witnessed during the summer months. The average annual capacity factor was 12.5%. The performance ratio (PR) experienced a slight variation within the range of 76e79%, with an annual average value of 78%. Temperature has a significant impact on the overall performance of the PV system. The maximum cell temperature was 52 in the month of January and the minimum was 44.05 in July. The annual power generation efficiency of the PV system was 11.96%, which ranged from 11.71% to 12.19%. During summer months, the average energy output was 1298 kWh, which ranged from 978 kWh to 1546 kWh. However, the energy output during winter periods ranged from 322.8 kWh to 816.3 kWh. In addition, from the financial evaluation at 4%, 6% and 8% discount rates, the levelized cost of energy are 12.1, 14.1 and 16.2 c/ kWh respectively. Also, the net present value are USD 22000, 14600 and 9100 respectively with a simple payback period of 6.4 years. Over the monitored period (2014), the cost of grid power consumption reduced from NZD $8380.86 to NZD $3758.77, which resulted in savings of approximately NZD $4700 which matches closely with the simple payback period given by the SAM's financial model. Overall, the total amount of grid power consumed reduced significantly by 32%. The PV system saves the school up to half of its electricity bill during summer and a third in winter.
Fig. 16. School's energy usage and cost of energy for pre- and post-installation of the PV array.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
M. Emmanuel et al. / Energy xxx (2016) 1e11
Acknowledgments The authors gratefully acknowledge the support of Victoria University for this work through the VUW Research Trust. The authors gratefully acknowledge the assistance provided by Mike Rathbone to access the power generation data of the solar PV system at Maungaraki school. References [1] Renewables 2015 Global Status Report, http://www.ren21.net/wp-content/ uploads/2015/07/REN12-GSR2015_Onlinebook_low1.pdf. [2] Paudel AM, Sarper H. Economic analysis of a grid-connected commercial photovoltaic system at Colorado State University-Pueblo. Energy 2013;52: 289e96. [3] Lau K, Muhamad N, Arief Y, Tan C, Yatim A. Grid-connected photovoltaic systems for Malaysian residential sector: effects of component costs, feed-in tariffs, and carbon taxes. Energy 2016;102:65e82. [4] Electrical Power Research Institute, http://dpv.epri.com/index.html. [5] Ministry of Business, Innovation and Empowerment, http://www.mbie.govt. nz/info-services/sectors-industries/energy/energy-data-modelling/statistics/ electricity. [6] Jahn U, Mayer D, Heidenreich M, Dahl R, Castello S, Clavadetscher L, et al. International Energy Agency PVPS T13: analytical monitoring of gridconnected photovoltaic systems. Citeseer; 2000. p. 5. [7] Marion B, Adelstein J, Boyle K, Hayden H, Hammond B, Fletcher T, et al. Performance parameters for grid-connected PV systems. In: Photovoltaic specialists conference, 2005. Conference record of the thirty-first IEEE. IEEE; 2005. p. 1601e6. [8] Jahn U, Mayer D, Heidenreich M, Dahl R, Castello S, Clavadetscher L, et al. International energy agency PVPS Task 2: analysis of the operational performance of the IEA database PV systems. In: 16th European photovoltaic solar energy conference and exhibition. Glasgow, United Kingdom: Citeseer; 2000. p. 5. [9] Pietruszko S, Gradzki M. Performance of a grid connected small PV system in Poland. Appl energy 2003;74(1):177e84. [10] Mondol JD, Yohanis Y, Smyth M, Norton B. Long term performance analysis of a grid connected photovoltaic system in Northern Ireland. Energy Convers Manag 2006;47(18):2925e47. [11] Chokmaviroj S, Wattanapong R, Suchart Y. Performance of a 500kW P grid connected photovoltaic system at Mae Hong Son Province. Thail Renew Energy 2006;31(1):19e28. [12] Kymakis E, Kalykakis S, Papazoglou TM. Performance analysis of a grid connected photovoltaic park on the island of Crete. Energy Convers Manag 2009;50(3):433e8. [13] Cherfa F, Arab AH, Oussaid R, Abdeladim K, Bouchakour S. Performance analysis of the mini-grid connected Photovoltaic system at algiers. Energy Procedia 2015;83:226e36. [14] Ayompe L, Duffy A, McCormack S, Conlon M. Measured performance of a 1.72 kW rooftop grid connected photovoltaic system in Ireland. Energy Convers Manag 2011;52(2):816e25. [15] Okello D, van Dyk E, Vorster F. Analysis of measured and simulated performance data of a 3.2 kWp grid-connected PV system in Port Elizabeth, South Africa. Energy Convers Manag 2015;100:10e5. [16] Sharma V, Chandel S. Performance analysis of a 190 kWp grid interactive solar photovoltaic power plant in India. Energy 2013;55:476e85. [17] Padmavathi K, Daniel SA. Performance analysis of a 3 MW p grid connected solar photovoltaic power plant in India. Energy Sustain Dev 2013;17(6): 615e25. [18] Kumar BS, Sudhakar K. Performance evaluation of 10 MW grid connected solar photovoltaic power plant in India. Energy Rep 2015;1:184e92. [19] Sundaram S, Babu JSC. Performance evaluation and validation of 5 MW p grid connected solar photovoltaic plant in South India. Energy Convers Manag 2015;100:429e39. [20] Edalati S, Ameri M, Iranmanesh M. Comparative performance investigation of mono-and poly-crystalline silicon photovoltaic modules for use in gridconnected photovoltaic systems in dry climates. Appl Energy 2015;160: 255e65.
11
[21] Sidi CEBE, Ndiaye ML, El Bah M, Mbodji A, Ndiaye A, Ndiaye PA. Performance analysis of the first large-scale (15 MW p) grid-connected photovoltaic plant in Mauritania. Energy Convers Manag 2016;119:411e21. [22] Adaramola M. Viability of grid-connected solar PV energy system in Jos, Nigeria. Int J Electr Power Energy Syst 2014;61:64e9. [23] Mondal M, Sadrul Islam A. Potential and viability of grid-connected solar PV system in Bangladesh. Renew Energy 2011;36:1869e74. [24] El-Shimy M. Viability analysis of PV power plants in Egypt. Renew Energy 2009;34:2187e96. [25] Adaramola M. Techno-economic analysis of a 2.1 kW rooftop photovoltaicgrid-tied system based on actual performance. Energy Convers Manag 2015;101:85e93. [26] Edalati S, Ameri M, Iranmanesh M, Tarmahi H, Gholampour M. Technical and economic assessments of grid-connected photovoltaic power plants:Iran case study. Energy 2016;114:923e34. [27] Orioli A, Di Gangi A. Review of the energy and economic parameters involved in the effectiveness of grid-connected PV systems installed in multi-storey buildings. Appl Energy 2014;113:955e9369. [28] The Dynamis Project, http://www.dynamisproject.co.nz/. [29] Akinyele D, Rayudu R, Nair N. Global progress in photovoltaic technologies and the scenario of development of solar panel plant and module performance estimation- Application in Nigeria. Renew Sustain Energy Rev 2015;48:112e39. [30] Labed S, Lorenzo E. The impact of solar radiation variability and data discrepancies on the design of PV systems. Renew Energy 2004;29(7):1007e22. [31] Enasolar Grid-tied Solar Inverter, http://www.enasolar.net/Products/ EnaSolar-GT-Series-Solar-Inverter-v2-3f. [32] Solar Photovoltaic Energy Systems, IEC/TS 61836 (2007-12). [33] PV System Performance Monitoring, IEC 61724. [34] Dobaria B, Pandya M, Aware M. Analytical assessment of 5.05 kWp grid tied photovoltaic plant performance on the system level in a composite climate of western India. Energy 2016;111:47e51. [35] Design of grid connect PV systems, Tech. rep., Palau Public Utilities Corporation. [36] IEEE guide for array and battery sizing in stand-alone photovoltaic (PV) systems. 2008. [37] Akinyele D, Rayudu R, Tan RH. Comparative study of Photovoltaic technologies based on Performance, cost and space requirement: strategy for selection and application. Int J Green Energy 2016;13(13):1352e68. [38] Akinyele D, Rayudu R, Nair N. Development of photovoltaic power plant for remote residential applications: the socio-technical and economic perspectives. Appl Energy 2015;155:131e49. [39] Kim J-Y, Jeon G-Y, Hong W-H. The performance and economical analysis of grid-connected photovoltaic systems in Daegu, Korea. Appl energy 2009;86(2):265e72. [40] Tech. rep. Power system operation and augmentation planning with PV integration. International Energy Agency; 2014. http://www.iea-pvps.org/ index.php?id¼322. [41] Grid-connected solar PV systems-no battery, Design guidelines for accreditted installers. Clean Energy Council; 2013. [42] Akinyele D, Rayudu R. Community-based hybrid electricity supply system: a practical and comparative approach. Appl Energy 2016;171:608e28. [43] Renewable Power Generation Costs in 2014 (IRENA), http://www.irena.org/ documentdownloads/publications/irena_re_power_costs_2014_report.pdf/. [44] Palm T, Qayum A. Private and public investment analysis. South-Western Pub. Co.; 1985. [45] Short W, Packey DJ, Holt T. A manual for the economic evaluation of energy efficiency and renewable energy technologies. University Press of the Pacific; 2005. [46] A. Miller, A. Wood, M. Hwang, S. Lemon, E. G. Read, Economics of photovoltaic solar power and uptake in New Zealand. [47] Dorf RC. The technology management handbook. Heidelberg: Springer; 1999. [48] National Institute of Water and Atmospheric Research, http://www.niwa.co. nz/our-services/online-services/solarview. [49] Emmanuel M, Rayudu R. Evolution of dispatchable photovoltaic system integration with the electric power network for smart grid applications: a review. Renew Sustain Energy Rev 2017;67:207e24. [50] Emmanuel M, Rayudu R. Communication technologies for smart grid applications: a survey. J Netw Comput Appl 2016;74:133e48.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
Contents lists available at ScienceDirect
Energy journal homepage: www.elsevier.com/locate/energy
Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand Michael Emmanuel*, Daniel Akinyele, Ramesh Rayudu Smart Power & Renewable Energy Systems Group, School of Engineering and Computer Science, Victoria University of Wellington, Wellington 6140, New Zealand
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 June 2016 Received in revised form 28 October 2016 Accepted 19 November 2016 Available online xxx
This paper presents a performance analysis and economic viability of a 10 kWp grid-connected solar photovoltaic (PV) system installed at Maungaraki school, Wellington, New Zealand under the ”Dynamis Project”. The system consists of 40 panels and two units of 5 kW power converters with a communication capability while the distribution grid serves as a virtual storage. The excess power generated at low load condition during holidays and weekends is exported to the grid. With the system in operation since 2014, performance parameters based on International Energy Agency Photovoltaic Power System Programme (IEA PVPS), Clean Energy Council industry guide and IEC 61724 standard are evaluated. The final yield ranged from 1.1 to 4.9 h/d. The performance ratio (PR) varied between 76 and 79%, giving an annual PR of 78%. In addition, from the economic evaluation of this system at 4%, 6% and 8% discount rates, the levelized cost of energy are 12.1, 14.1 and 16.2 c/kWh respectively, with a simple payback period of 6.4 years. Overall, the total amount of power consumed annually from the grid reduced significantly by 32%. In monetary terms, the school's expenditure on power bills reduced effectively by 45% and was able save about NZD $4700 in 2014. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Photovoltaic Grid-connected Final yield Performance ratio Levelized cost of energy
1. Introduction The recent energy transition as a result of the growing presence of renewable energy resources (such as wind power, solar energy and hydropower) in the conventional power system is pivotal in dealing with climate change, enhancing power network resilience and provision of new economic opportunities [1e3,49,50]. One of the most common renewable energy resources is the photovoltaic (PV) system such as the rooftop residential and utility scale installations [4]. However, hydroelectric generation supplies the bulk of New Zealand's electricity demand, geothermal contributes about 17.2% while wind generation is about 5.4% of the country's power generation [5]. In a recent quarterly electricity generation report, solar power contributed 0.1% (as shown in Fig. 1) of the country's net electricity generation of about 42927.54 GWh at the end of 2015 [5]. The contribution of solar photovoltaic (PV) systems is quite insignificant when compared to the deployment of other renewables in New Zealand. However, in recent times, there is a wide-
* Corresponding author. E-mail address: [email protected] (M. Emmanuel).
spread deployment of rooftop grid-connected PV systems in residential apartments and schools as reported in this article. The proliferation of such systems on the electricity grid highlights the need to understand its operation through monitoring and thorough performance analysis [6]. In addition, utility interactive PV systems are essentially monitored to evaluate the final energy yield, detect possible design defects and avoid economic losses as a result of operational issues [6]. Another reason to carry out such analysis is to assess PV credibility, viability and ultimately to increase their penetration within the existing electric power network [7]. The IEA PVPS Task 2 has been able to analyse and publish 170 grid-connected PV systems installed in various countries of the world [8]. In addition, over the previous years, various authors from different countries have published results of the performance analysis of their respective grid-connected PV units. Pietruszko et al. [9] evaluated the performance of a 1 kWp a-Si PV system located at Warsaw, Poland. The performance ratio ranged from 0.6 to 0.8, annual system yield was 830 kWh and the efficiency of the PV system was in the range of 4e5%. Mondol et al. [10] presented the outcome of the performance analysis of a 13 kWp m-Si PV system installed in Northern Ireland. The evaluated annual final yield ranged from 1.61 to 1.76 h/d, with a PR which ranged from 0.6
http://dx.doi.org/10.1016/j.energy.2016.11.107 0360-5442/© 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
2
M. Emmanuel et al. / Energy xxx (2016) 1e11
Nomenclature
hsys l ma Si
a Si Aarray AC DC GI GI;ref h=d Lc Ls LSP LCOE
system efficiency power temperature coefficient(0:45%/ C) micromorph silicon amorphous silicon area of the array (m2 ) alternating current (A) direct current (A) in-plane solar irradiation (kWh=m2 ) reference irradiance (kW=m2 ) hour/day array capture losses system losses capacity factor levelized cost of energy
Fig. 1. Electricity generation from various energy sources in New Zealand [5].
to 0.62 and annual average system efficiency of 6.4%. Also, Chokmaviroj et al. [11] evaluated the performance of a 500 kWp gridconnected PV plant at Mae Hong Son province, Thailand. The plant was divided into two, 250 kWp, from a double glazed semi-Si PV modules. The plant generated about 383274 kWh and the efficiency of the PV array ranged from 9 to 12%. The final yield ranged from 2.91 to 3.98 h/d and the PR ranged from 0.7 to 0.9. Another performance study is that of Kymakis et al. [12]. The study evaluated a p-Si 171.36 kWp utility interactive PV park on the island of Crete, which was as a result of the favourable climatic condition and the recent incentivization of PV system installations in Greece. The yearlong evaluated PR was 0.67 and the final yield ranged from 1.96 to 5.07 h/d. Cherfa et al. [13] carried out an analysis of a mini-grid connected m-Si 9.54 kWp PV system which was a pilot project with the primary aim of acquiring experience in the design, monitoring and maintenance of such innovative technology in Algeria. The system performance was quantified over the monitored period which showed an annual 10981 kWh of energy injected into the grid. The average daily output energy was 30 kWh and PR ranged from 0.62 to 0.77. Ayompe et al. [14] presented results obtained from monitoring a m-Si 1.72 kWp rooftop gridconnected PV system in Ireland. Within the frame of the study, the monthly average daily capacity factor ranged from 5 to 15.5%, with an annual average of 10.1%. The PR and final energy were 0.82 and 2.4 h/d respectively. Another relevant study is that of Okello et al. [15] in South Africa. They presented the analysis of p-Si 3.2 kWp grid-connected PV system with a PR of 0.84 and final
m Si NPV p Si PO
mono crystalline silicon net present value poly crystalline silicon rated DC output power of the array under standard test conditions (Wp) PR performance ratio semi Si semi crystalline silicon TA ambient temperature ( C) TC cell temperature ( C) TR temperature rise for parallel-to-roof installation ( C) TSTC cell temperature at standard test conditions ( C) TLCC total life cycle cost YA array yield (kWh/kWp) Yf final yield (kWh/kWp) Yr reference yield (kWh/kWp)
energy yield of 4.9 h/d. Performance analysis of different utility interactive PV sizes have been conducted across various locations in India. For instance, Sharma and Chandel [16] investigated the performance of a p-Si 190 kWp PV plant connected to the grid in Punjab, northern India. The final yield, reference yield and PR ranged from 1.45 to 2.84 h/d, 2.29e3.53 h/d and 55e83% respectively. The annual average PR, capacity factor and system efficiency were found to be 74%, 9.27% and 8.3% respectively. Also, Padmavathi et al. [17] presented an analysis of a m-Si 3 MWp grid-connected PV plant in Karnataka, south western region of India. The annual average energy generated was 1372 kWh per kW of the installed capacity and the PR was 0.7. The annual average reference yield and final yield were 5.36 h/ d and 3.73 h/d respectively. In addition, Shiva Kumar and Sudhakar [18] analysed the performance of a p-Si 10 MWp PV plant connected to the grid at Ramagundam, southern India. The plant is one of the largest solar power plants with a seasonal tilt. Results revealed that the capacity utilization factor was 17.68% with an annual energy generation of 15 798.192 MWh. The final yield ranged from single-phase1.96e5.07 h/d, and annual PR was 0.86. Sundaram et al. [19] evaluated the performance analysis of a thin film a-Si 5 MWp grid-connected PV plant at Sivagangai district in Tamilnadu, south India. The annual actual generation was 8495.29 MWh with a system efficiency of 5.08%. The final yield, PR, average energy and exergy efficiency were found to be 4.81, 0.89, 6.08% and 3.54% respectively. Edalati et al. [20] presented a comparative performance analysis of m-Si (5.52 kWp) and p-Si (5.52 kWp) for utility interactive PV systems in dry climates. The annual average capacity factor for m-Si and p-Si were found to be 23.20% and 23.81% respectively. The annual final yield and PR for p-Si were 5.38 h/d and 0.83 respectively. Moreover, the annual final yield and PR for m-Si were 5.24 h/ d and 0.81 respectively. Sidi et al. [21] evaluated the performance analysis of the first utility-scale 15 MWp grid-connected PV plant in Mauritania, which has a high annual insolation between 1900 and 2200 kWh=m2 /year. The PV plant from a-Si/m a-Si was connected to the 33 kV electricity grid through inverters and nine transformers. Performances of two arrays from two different technologies, ma Si and a-Si were presented. The daily average array yield, PR, and capacity factor for ma Si were 4.38 h/d, 0.68 and 17.75% respectively. Moreover, the daily average array yield, PR, and capacity factor for a-Si were 4.79 h/d, 0.75 and 19.54% respectively. These performance metrics allow cross-comparison between different PV systems in terms of design, technology and diverse
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
M. Emmanuel et al. / Energy xxx (2016) 1e11
climatic conditions [7]. Also, the evaluation of these parameters help to assess product quality, determine future needs and detect failures in system components [7,8]. For instance, with the PR metric, system component malfunction such as inverter failure can be easily detected for utility interactive PV systems. In addition, other factors which could affect the PR include PV module soiling, shading and the ratio of the measured array efficiency to the nominal array efficiency [7,8]. Also, properly maintained PV systems have a high propensity to operate optimally with high PR value and availability [8]. Additionally, the analysis of the economic viability of utility interactive PV systems is very crucial in enhancing PV technology uptake as favourable renewable energy policies continue to evolve globally. Adaramola [22] presented the viability of grid-connected PV system in Jos, Nigeria. The study showed that the 80 kWp system was able to contribute 40.4% of the annual electricity demand with LCOE of $0.103/kWh. Mondal et al. [23] presented the financial viability of a proposed 1-MWp grid connected PV system in Bangladesh. For a project lifetime of 15 years, Mondal et al. reported energy production cost between $0.253/kWh and $0.282/kWh, which decreases with the increase of lifetime. Also, El-Shimy [24] presented a viability analysis of a 10 MWp grid-connected PV plant for 29 different sites in Egypt. The energy production cost ranged from $0.1989/kWh to $0.2424/kWh, and the equity payback varied between 4.9 years and 7.1 years. Adaramola [25] analysed the economic viability of a rooftop 2.07 kW grid-connected PV system in Norway with a feed-in-tariff of $0.356/kWh over 25 years of project lifetime. This resulted in $0.110/kWh premium over the LCOE of $0.246/kWh produced by the PV system. Edalati et al. [26] presented a techno-economic analysis of a 10 MWp utility interactive PV plant in Iran. The LCOE ranged between $0.1992/kWh in the south-eastern part of Iran to $0.3838/kWh in the northern part. Also, the sensitivity of NPV and payback time to six various annual inflation rate showed that inflation rate increase would lead to an elongated payback time and lower NPV. Orioli et al. [27] presented an economic analysis of a utility interactive PV installation using the ratio of NPV to the present value cost (NPV/PVC). A positive NPV/PVC value represents a profitable investment, while a negative value indicates that the revenue is than disbursements and a near zero value signifies a critical situation where little modifications in costs and benefits assessments abruptly impact the economic analysis. However, performance analysis of grid-connected PV systems in New Zealand are not available in the literature which is pivotal in understanding and quantifying their impacts on the traditional electricity network as an emerging alternative renewable power generation resource. Also, in order to develop the PV system as a sustainable energy resource and increase its uptake, it is important to assess its economic viability. This paper aims to fill these gaps by presenting a techno-economic analysis of a grid-connected 10 kWp PV system at Maungaraki school, Wellington in New Zealand as a case study. We have carried out our evaluation based on performance parameters specified by IEA PVPS and IEC 61724 standard for utility interactive PV systems. In addition, the Clean Energy Council (CEC) guide which represents industry best practice for the design and installation of grid-tied PV systems has been used for analysis in this study. Final system yield, energy yields, capacity factor and performance ratio are evaluated. Also, the economic viability of the PV system under consideration has been evaluated using metrics such as the NPV, LCOE and simple payback period. In addition, we present results from the Web Based PV System Monitoring and Reporting (WBPSMR) for the PV array.
3
2. Description of the grid-tied solar PV system The installation of the 10 kWp PV array was carried out under the ”Dynamis Project” with the aim of testing renewable energy technologies effectiveness and economic viability using schools as platforms in order to sensitize surrounding communities to become more energy self-sufficient [28]. The renewable energy for New Zealand schools under this project has successfully installed 10 kWp solar PV in two different schools. One of the installations was carried out at Maungaraki school (with a roll of about 250 pupils), as an attempt to pioneer a sustainable energy revolution. The school is located at a longitude of 174:9 E and latitude of 41:16 S, with a panel tilt angle of 41 . The PV panel tilt angle and the latitude of the corresponding site location are kept equal in order to obtain maximum solar radiation [29,30]. There are two identical parallel strings consisting of 20 panels each with modules rated at 250 Wp capacity, and tied to the grid via two 5 kW Enasolar single-phase inverters (as shown in Fig. 2). The energy meter is used to measure the amount of energy consumed and the excess power generated and exported to the grid which usually occurs during minimum load conditions that happens over weekends and holiday periods. The specification of the TNS250 module used is given in Table 1. In addition, the PV array was installed in an open space on a rooftop free of shadows or shading which can affect its performance. The PV panels are regularly maintained every six months to get rid of dust, birds' drops and other forms of impurities. Also, the cost of setting up the project was about NZD $28000 sourced through grants and donations. 2.1. Inverter with built-in communication interface The power converter has built-in cable (Ethernet) and wireless interfaces (as shown in Fig. 3) for web based PV system monitoring and reporting, which shows the energy production information such as daily and monthly output power, input and output voltages [31]. The inverter with a static or dynamically assigned internet protocol (IP) address can be connected to the existing home network via any of the available interfaces. Also, the inverters deployed have two independent maximum power point tracking PV inputs which can be connected to one solar string or two identical parallel solar strings [31]. In addition, the inverter has integrated lockable DC and AC switches for isolation purposes. The most relevant technical specifications of the inverter are presented in Table 2.
Fig. 2. Schematic block diagram of the PV system.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
4
M. Emmanuel et al. / Energy xxx (2016) 1e11
reference irradiance [8,32,34]. Therefore, the daily Yr is calculated as:
Table 1 PV module specifications. PV module
Specifications
Type of cell material Make Model Maximum power (Pmax ) Open circuit voltage (Voc ) Short circuit current (Isc ) Voltage at maximum power (Vmp ) Current at maximum power (Imp ) Number of cells in a module
Monocrystalline TNS Solar TN-60-6M 250 Wp 37.9 V 8.64 A 30.4 V 8.23 A 60
Z Yr ¼
24 t¼1
GI dt
GI;ref
(1)
where: GI ¼ total irradiation (kWh=m2 ) GI;ref ¼ reference irradiance (kW=m2 ). 3.2. PV array energy yield The ratio of the energy produced by the PV array to the rated PV capacity is referred to as the array yield, YA . Losses such as the ones due to manufacturing tolerance, temperature, dirt and dust are taken into consideration in estimating YA [8,32]. It is therefore, calculated as:
YA ¼
EA ½kWhDC PO ½kWpDC
(2)
3.3. Final system yield The final system yield, Yf , is the ratio of the net energy produced (ENET ) by the PV array to the rated DC array capacity (PO ) [32]. Also, it is the amount of energy supplied to the load per day (month or year) considering array capture and system losses. It quantifies the duration (e.g., yearly, monthly or daily) required by the PV array to operate at the rated DC power to supply an equal amount of energy and estimated as [7,12]:
Yf ¼
ENET ½kWhAC PO ½kWpDC
(3)
3.4. Performance ratio
Fig. 3. Inverter with built-in communication interface [31].
This is a dimensionless quantity used to indicate the monthly or annual impact of PV system losses on the rated array capacity. The system losses are as a result of PV array temperature, failure of system components and incomplete usage of the irradiation [32]. In addition, the performance ratio, PR, does not depend on location of PV installation and system size [8]. It is used to evaluate the quality of PV system installation and estimated as [35]:
3. System performance analysis This article considers the evaluation of the operational and reliability performance of the grid-connected PV system based on the IEC standards (61836 and 61724) and reports from International Energy Agency (IEA) photovoltaic power systems program (Task 2) [8,32,33]. The performance analysis of grid-connected PV systems (as shown in Fig. 4) is pivotal in assessing their operational performance under different climatic conditions and detection of operational issues [18]. It also enables the measurement of longterm variation in system performance and comparisons with other systems that differ in location, design and technology [7].
3.1. Reference yield The reference yield, Yr , is a function of in-plane irradiation and estimated as the ratio of the total irradiation per day (or year) to the
PR ¼
PV system actual energy yield Yf ¼ PV array ideal energy output Yr
(4)
3.5. Losses System losses, classified into DC and AC subsystems losses, are pivotal design considerations in providing realistic energy solutions [36]. However, in this study, losses due to shading or shadows is not considered because the location of the PV system is free of shadows during the day. The connecting cable (PV DC main cables) between the PV array and the inverter subsystems can lead to power loss which should be accounted for. For our case study, the DC cable losses is 3% which amounts to 0.97 de-rating factor [7,35]. However, AC subsystems losses is as a result of the connecting cable between the inverter
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
M. Emmanuel et al. / Energy xxx (2016) 1e11
5
Table 2 Electrical properties of the inverter. DC Input Number of inputs Max. open circuit voltage (Voc ) DC full power operating range Operating voltage range (Vmpp ) DC optimal operating voltage Max. input current (Impp ) Max. short circuit current (Isc ) Maximum usable input power (Pmax ) Maximum allowable input power Reverse polarity protection AC Output Nominal output voltage Output voltage range Output power (@ 50 Hz) Max. output current Power factor Max. efficiency Max. Euro. efficiency General system data Data interface Weight Operating temperature range Night-time consumption
2 Independent MPPT inputs 600 per DC input 235e500 V per DC input 120e500 V per DC input 350 V per DC input 15 A per DC input 16 A 3500 Wp per DC input 7000 W Inherent crowbar diodes 230 V AC single phase 202e259 V AC (New Zealand) 4990 W 21.5 A >0.98 >96.8% >95.4% IEEE 802.11 (Wi-Fi)/Ethernet 20 kg 30 C to þ50 C (full power), þ70 C (de-rated) <1.2 W
power produced by the inverter (PInv;AC ) to the DC power generated by the PV array unit (PPV;DC ) [17,18].
hinv ¼
PInv;AC PPV;DC
(8)
3.8. System efficiency The monthly system efficiency, hsys , is given as [16,39]:
Fig. 4. System performance analysis.
and the grid with a typical value of 1% and de-rating factor of 0.99 [7,35]. These two losses are components of the system losses, Ls , and estimated as:
Ls ¼ YA Yf
ENET Girrad Aarray
(9)
here, ENET is the monthly total AC PV energy output, Girrad is the monthly peak sun hour, and Aarray is the area of the array.
(5)
The array capture losses, Lc , refers to the normalised PV system losses which is as a result of PV array energy losses such as conversion losses, manufacturing tolerance and dirt [32,37]. In addition, Lc is calculated as:
Lc ¼ Yr YA
hsys ¼
(6)
3.9. Energy yield Energy yield of a PV array is a function of the meteorological data of the location where it is installed [40]. This is the PV system energy output estimated as [35,41,42]:
ENET ¼ PratedSTC mtemp mman mdirt Girrad hinv hpvinv hinvsb (10)
3.6. Capacity factor The capacity factor, LSP , is the defined as the ratio of the energy output from the PV system to the product of the operating duration and rated PV array output [38]. LSP is calculated annually as [12]:
LSP ¼
Yf ENET ¼ PO 8760 8760
(7)
3.7. Inverter efficiency The inverter efficiency, hinv , is defined as the ratio of the AC
where: ENET ¼ PV system energy yield (Wh). PO ¼ rated DC output power of the array under standard test conditions (Wp). mtemp ¼ de-rating factor due to temperature. mman ¼ de-rating factor for manufacturing tolerance (0.97). mdirt ¼ de-rating factor for dirt (0.97). Girrad ¼ irradiation value or peak sun hour (kWh=m2 ) hinv ¼ efficiency of the inverter. hpvinv ¼ efficiency of the sub-unit (connecting cables) between the PV unit and the inverter (0.97). hinvsb ¼ efficiency of the sub-unit (connecting cables) between the inverter and the switchboard (0.99).
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
6
M. Emmanuel et al. / Energy xxx (2016) 1e11
3.9.1. Temperature de-rating factor The de-rating due to temperature for this study is done in accordance with guidelines provided by the Clean Energy Council (CEC) for grid-connected solar PV systems without battery storage [41]. The cell temperature TC for a grid-connected PV is estimated as:
TC ¼ TA þ TR
(12)
where: l ¼ power temperature coefficient/ C (0:45%/ C) TSTC ¼ cell temperature at standard test conditions ( C)
The enhancement of the cost-competitiveness of the solar PV system over the years continues to drive its uptake as a viable power generation alternative in the global energy mix. In 2014, prices of PV modules decreased by 75% in comparison with their prices at the close of 2009 and as a result, improving PV system plug parity [43]. Important measures used in the economic evaluation of the PV system considered in this article are as follows: 4.1. Net present value (NPV) The NPV is a financial tool used to evaluate cash outflows and revenues, and investment characteristics and decisions especially for comparing mutually exclusive projects [44,45]. Also, it is the algebraic sum of the net cash flows over the project's life time to the present, discounted by an appropriate discount rate [46]. It is given as: N X m¼0
CFm ð1 þ rÞm
N m¼1
TLCC ¼
N X m¼0
TLCC
Em ð1 þ rÞ m
ICm ð1 þ rÞm
(14)
(15)
where: TLCC ¼ total life cycle cost. ICm ¼ investment cost in period m. Em ¼ energy output or saved in year m. In addition, LCOE which considers the current dollar value is known as the nominal LCOE while the real LCOE is a fixed dollar inflation-adjusted value. The real LCOE uses a discount rate without taking into account inflation rate and the nominal LCOE applies a discount rate with the inflation rate [45]. 4.3. Simple payback period
4. Economic analysis
NPV ¼
LCOE ¼ P
(11)
where: TA ¼ ambient temperature ( C) TR ¼ temperature rise for parallel-to-roof installation (35 C) However, TR is a function of PV array installation type which could be top-of-pole, parallel-to-top or rack-type mount [41]. The parallel-to-top array frames were deployed on site used for our case study with a TR of 35 C as recommended by CEC. With the estimated value of TC , the de-rating factor due to temperature, mtemp , is given by:
mtemp ¼ 1 þ fl*ðTC TSTC Þg
that reflects the average capital cost [43]. This metric is used to compare the cost of energy generated by a renewable technology with that of a traditional fossil fuel generating unit [45]. The LCOE is given as:
(13)
where: CFm ¼ net cash flow in year m. N ¼ analysis period. r ¼ annual discount rate. The NPV criteria include [45,47]: If the NPV is positive, the investment is economical. Negative NPV denotes the return are worth less than the initial investment. The financial viability of a project is uncertain with Zero NPV. Amongst independent projects, the higher the NPV value the better.
4.2. Levelized cost of energy (LCOE) This is the ratio of the lifetime costs to the lifetime electric power generation, both of which are discounted back to a base year
The simple payback (SPB) period is another pivotal financial tool used to estimate the number of years it will take to recover the project cost of an investment made [45]. Also, it is important to note that the SPB does not consider the time value of money and with a zero discount rate. The SPB is given as the first point in time when the following expression is satisfied [45]:
X m
DIICm
X
DSm
(16)
m
where: DIICm ¼ incremental investment costs at zero discount rate in period m. DSm ¼ sum total of the annual cash flows net annual costs at zero discount rate in period m. 5. Results and discussion This section presents the performance analysis of the gridconnected 10 kWp (40 250 Wp) PV system. We present results from our analysis, the National Renewable Energy Laboratory's (NREL) system advisory model (SAM) and Web Based PV System Monitoring and Reporting (WBPSMR). In order to monitor the behaviour of the PV system, it is pivotal to study the meteorological data recorded in the weather station covering the site location. Therefore, we have used full data set from National Institute of Water and Atmospheric research (NIWA) solarview tool [48] for our analysis. Fig. 5 shows the monthly averaged total in-plane irradiation and ambient temperature of the installation site. Significant peaks of in-plane insolation also known as Peak Sun Hour (PSH) occur during the summer period while the PSH dropped drastically during winter. Furthermore, the monthly average daily wind speed varied between 0.5 m/s in most parts of the months and 21.6 m/s in October. The relative humidity ranged from 38 to 97% and the maximum global irradiance was 1045W=m2 . In addition, Fig. 6 shows the module temperature as it varies across the entire year, with high temperatures during the summer period. Our analysis as depicted in Fig. 7 shows an interesting pattern of the monthly capacity factors and performance ratios over the entire year. The monthly capacity factors increased proportionately with the in-plane solar irradiation with an annual average value of 12.5%.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
M. Emmanuel et al. / Energy xxx (2016) 1e11
7
Fig. 5. Monthly averaged total in-plane irradiation and ambient temperature. Fig. 8. Monthly averaged daily final yield, PV array capture losses and system losses.
Fig. 6. Annual module temperature.
summer period and higher in winter [7]. In addition, the system and array capture losses increased as the in-plane insolation increases. Fig. 8 depicts the monthly mean daily final yield, array capture and system losses. The monthly average daily array yield ranged from 1.2 h/d (June) to 5.4 h/d (January), while the final yield varied between 1.1 h/d (June) to 4.9 h/d (January). The average annual final yield and reference yield were 2.99 h/d and 3.87 h/d respectively. The monthly averaged daily array capture losses ranged from 0.19 (June) to 1.14 h/d (January) while the system losses varied from 0.09 (June) to 0.42 h/d (January). The PR of the considered PV array experienced a slight variation within the range of 76e79%, and the annual average value was 78%. This is comparable to the range of values (0.6e0.8) reported by IEAPVPS Task 2 for grid-connected PV systems [8]. The monthly inverter efficiency ranged from 94.9% to 95.7% with higher values during the high PSH as shown in Fig. 9. In addition, the monthly system efficiency ranged from 11.71% to 12.19% with higher values during low in-plane solar insolation as shown in Fig. 9. The average annual system efficiency of the PV array is 11.96%. In addition, the 10 kWp PV array generated an average energy output of 1298 kWh during the summer period ranging from 978 kWh to 1546 kWh. During low PSH values in winter, the energy output ranged from 322.8 kWh to 816.3 kWh. From our analysis, the annual average energy output was 910.13 kWh. However, from the WBPSMR, the average energy output was 1058.33 kWh, with the monthly energy output from the two 5 kW inverters shown in Fig. 10. Low energy output in January was because the site was commissioned at the middle of the month. The two strings of the PV array are mounted in an open space close to the school's football field, and the amount of dust collected over
Fig. 7. Monthly capacity factors and performance ratios of the PV array.
The availability of a high solar resource as witnessed during summer period, led to increase in capacity factors (shown in Fig. 7) and the final yield (shown in Fig. 8). However, high values of PSH does not necessarily translate to high performance ratios as shown in Fig. 7. This is traceable to the impact of temperature losses on the overall performance of the PV array. As the temperature of the location increases, the PR values also dropped accordingly. Therefore, PR values are lower during
Fig. 9. Monthly system and inverter efficiencies.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
8
M. Emmanuel et al. / Energy xxx (2016) 1e11
Fig. 10. WBPSMR monthly PV output energy.
them varies, which causes variation in the array and inverter output as depicted in Fig. 10. Also, the daily power generated is shown in Fig. 11.
6. Comparison of PV systems performance Performance analysis metrics allow cross-comparison between different PV systems operating under diverse climatic conditions [7]. Table 3 shows performance of various utility interactive PV systems across different locations. The annual average final yield of the m-Si modules in this study is higher than the ones reported in Ireland and Northern Ireland. However, it is lower than the ones reported in Iran and India. Also, the annual average PR in this study falls within the range (0.25e0.9) reported by IEA PVPS Task 2 for 170 grid-connected PV systems in different countries of the world [8]. In addition, the PR for m-Si technology is lower than that of p-Si for most of locations with exception of Greece which ranged from 0.58 to 0.73.
7. Economic analysis The computation of the economic measures described is presented in this section using the NREL's SAM financial model. Although there is no feed-in tariff legislation for the PV system in New Zealand, its economic viability especially for schools is shown in this paper. A number of assumptions and inputs required in determining the financial metrics for the 10 kWp PV system are presented in Table 4. Also, the monthly energy consumption and peak demand are shown in Fig. 12. The total project cost was NZD 28000 (USD 19600) which included the system, balance of equipment (BOS), installation, installer margin and overhead costs. Therefore, the total installed cost per capacity was USD 1.96/Wdc. The cash flow is presented in Table 5 and results obtained are given in Table 6 with three discount rates. The discount rate has significant impacts on the NPV and LCOE, with 4% rate giving the highest NPV and lowest norminal LCOE values which makes it an appropriate rate for such a system. Fig. 13 shows payback cash flow with a simple payback period of 6.4 years.
Fig. 11. WBPSMR daily power generated.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
M. Emmanuel et al. / Energy xxx (2016) 1e11
9
Table 3 Grid-connected PV system performance evaluation in various locations. Location
Capacity
PV Technology type
PR
Yf (h/d)
Warsaw, Poland [9] Ballymena, Northern Ireland [10] Mae Hong Son Province, Thailand [11] Island of Crete, Greece [12] Algiers, Algeria [13] Dublin, Ireland [14] Khatkar-Kalan, India [16] Karnataka, India [17] Port Elizabeth, South Africa [15] Kerman, Iran [20] Kerman, Iran [20] Ramagundam, India [18] Wellington, New Zealand [Present study]
1 kWp 13 kWp 500 kWp 171.36 kWp 9.54 kWp 1.72 kWp 190 kWp 3 MWp 3.2 kWp 5.52 KWp 5.52 kWp 10 MWp 10 kWp
a-Si m-Si semi-Si p-Si m-Si m-Si p-Si m-Si p-Si m-Si p-Si p-Si m-Si
0.6-0.8 0.6e0.62 0.7e0.9 0.67 0.62e0.77 0.82 0.55e0.83 0.7 0.84 0.81 0.83 0.86 0.78
e 1.61e1.76 2.91e3.98 1.96e5.07 e 2.4 1.45e2.84 3.73 4.9 5.24 5.38 1.96e5.07 1.1e4.9
Table 4 Inputs and assumptions used in the analysis. System size (kWp) Operation and maintenance cost (Fixed cost by capacity) ($/kW-yr) System salvage value ($) Panel tilt (degrees) Panel azimuth (degrees) Annual panel degradation (%/year) Inflation rate (%/year) Real discount rate (%/year) Analysis time period (years) Load profile used Grid buyback rate (c/kWh) electricity retail rate (c/kWh) Corporate tax rate
10 20 0 30 0 0.5% 2.5% 4%, 6% and 8% 25 school load profile 15 25 NA
Fig. 13. Payback cash flow (simple payback period ¼ 6.4 years).
In addition, the electricity to/from the grid and electricity sales/ purchases across the entire year are shown in Figs. 14 and 15 respectively. During periods of low insolation, significant amount of electricity was purchased from the grid, while excess power was exported to the grid during periods with high PSH and minimum load.
Fig. 12. Monthly energy consumption and peak demand.
Table 5 Cash flow. Metric
Value
Annual energy (year 1) Energy yield (year 1) Electricity bill without system (year 1) Electricity bill with system (year 1) Net savings with system (year 1) Net capital cost Equity Debt
16158 kWh 1616 kWh/kW $8310 $5229 $3082 $19600 $19600 $0
Table 6 Economic measures evaluation of the 10 kWp system. Discount rate Real LCOE (c/kWh) Nominal LCOE (c/kWh) NPV ($)
4% 9.43 12.1 22000
6% 11.19 14.1 14600
8% 13.09 16.2 9100
7.1. Actual economic analysis The actual financial analysis of the power consumption before and after the installation of the 10 kWp PV system is presented in this section. The annual cost of the school's power consumption before the PV system integration was NZD $8380.86. After the interconnection, the school's energy usage cost reduced to NZD $3758.77. The consumption pattern (obtained from the data logging system) of the school is shown in Fig. 16 for pre- (without PV-2013) and post-installation (with PV-2014) of the PV system. It shows a dramatic drop in kWs used and amount of money spent on power bills from the utility. In addition, as shown in Fig. 16, the power consumed after the PV system installation represents the amount of power consumed minus the amount of excess power generated in a particular month. The excess power export to grid usually
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
10
M. Emmanuel et al. / Energy xxx (2016) 1e11
reduced significantly by 32% and 45% respectively at the end of the year 2014. In monetary terms, the school saved about NZD $4700 in 2014 on power bills, which matches closely with the payback period given by the SAM's financial model. 8. Conclusion In this study, the performance analysis of a 10 kWp gridconnected PV system installed at Maungaraki school, Wellington, New Zealand is presented. From our analysis, the main conclusions are given as follows:
Fig. 14. Annual electricity to/from the grid.
Fig. 15. Annual hourly electricity sales/purchase with/without PV system.
occurs at low load condition which happens during weekends and holidays. Also, the spike between November and December is due to readings carried over to the next month. Furthermore, from our analysis the total amount of power consumed and cost of power
The final yield, (Yf ), of the PV system ranged from 1.1 to 4.9 h/d, with an annual average value of 2.99 h/d. There is a direct proportional relationship between the peak sun hour (PSH) and Yf . Also, the array capture and system losses increased with high values of PSH. The availability of a high solar resource led to increase in the capacity factor as witnessed during the summer months. The average annual capacity factor was 12.5%. The performance ratio (PR) experienced a slight variation within the range of 76e79%, with an annual average value of 78%. Temperature has a significant impact on the overall performance of the PV system. The maximum cell temperature was 52 in the month of January and the minimum was 44.05 in July. The annual power generation efficiency of the PV system was 11.96%, which ranged from 11.71% to 12.19%. During summer months, the average energy output was 1298 kWh, which ranged from 978 kWh to 1546 kWh. However, the energy output during winter periods ranged from 322.8 kWh to 816.3 kWh. In addition, from the financial evaluation at 4%, 6% and 8% discount rates, the levelized cost of energy are 12.1, 14.1 and 16.2 c/ kWh respectively. Also, the net present value are USD 22000, 14600 and 9100 respectively with a simple payback period of 6.4 years. Over the monitored period (2014), the cost of grid power consumption reduced from NZD $8380.86 to NZD $3758.77, which resulted in savings of approximately NZD $4700 which matches closely with the simple payback period given by the SAM's financial model. Overall, the total amount of grid power consumed reduced significantly by 32%. The PV system saves the school up to half of its electricity bill during summer and a third in winter.
Fig. 16. School's energy usage and cost of energy for pre- and post-installation of the PV array.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107
M. Emmanuel et al. / Energy xxx (2016) 1e11
Acknowledgments The authors gratefully acknowledge the support of Victoria University for this work through the VUW Research Trust. The authors gratefully acknowledge the assistance provided by Mike Rathbone to access the power generation data of the solar PV system at Maungaraki school. References [1] Renewables 2015 Global Status Report, http://www.ren21.net/wp-content/ uploads/2015/07/REN12-GSR2015_Onlinebook_low1.pdf. [2] Paudel AM, Sarper H. Economic analysis of a grid-connected commercial photovoltaic system at Colorado State University-Pueblo. Energy 2013;52: 289e96. [3] Lau K, Muhamad N, Arief Y, Tan C, Yatim A. Grid-connected photovoltaic systems for Malaysian residential sector: effects of component costs, feed-in tariffs, and carbon taxes. Energy 2016;102:65e82. [4] Electrical Power Research Institute, http://dpv.epri.com/index.html. [5] Ministry of Business, Innovation and Empowerment, http://www.mbie.govt. nz/info-services/sectors-industries/energy/energy-data-modelling/statistics/ electricity. [6] Jahn U, Mayer D, Heidenreich M, Dahl R, Castello S, Clavadetscher L, et al. International Energy Agency PVPS T13: analytical monitoring of gridconnected photovoltaic systems. Citeseer; 2000. p. 5. [7] Marion B, Adelstein J, Boyle K, Hayden H, Hammond B, Fletcher T, et al. Performance parameters for grid-connected PV systems. In: Photovoltaic specialists conference, 2005. Conference record of the thirty-first IEEE. IEEE; 2005. p. 1601e6. [8] Jahn U, Mayer D, Heidenreich M, Dahl R, Castello S, Clavadetscher L, et al. International energy agency PVPS Task 2: analysis of the operational performance of the IEA database PV systems. In: 16th European photovoltaic solar energy conference and exhibition. Glasgow, United Kingdom: Citeseer; 2000. p. 5. [9] Pietruszko S, Gradzki M. Performance of a grid connected small PV system in Poland. Appl energy 2003;74(1):177e84. [10] Mondol JD, Yohanis Y, Smyth M, Norton B. Long term performance analysis of a grid connected photovoltaic system in Northern Ireland. Energy Convers Manag 2006;47(18):2925e47. [11] Chokmaviroj S, Wattanapong R, Suchart Y. Performance of a 500kW P grid connected photovoltaic system at Mae Hong Son Province. Thail Renew Energy 2006;31(1):19e28. [12] Kymakis E, Kalykakis S, Papazoglou TM. Performance analysis of a grid connected photovoltaic park on the island of Crete. Energy Convers Manag 2009;50(3):433e8. [13] Cherfa F, Arab AH, Oussaid R, Abdeladim K, Bouchakour S. Performance analysis of the mini-grid connected Photovoltaic system at algiers. Energy Procedia 2015;83:226e36. [14] Ayompe L, Duffy A, McCormack S, Conlon M. Measured performance of a 1.72 kW rooftop grid connected photovoltaic system in Ireland. Energy Convers Manag 2011;52(2):816e25. [15] Okello D, van Dyk E, Vorster F. Analysis of measured and simulated performance data of a 3.2 kWp grid-connected PV system in Port Elizabeth, South Africa. Energy Convers Manag 2015;100:10e5. [16] Sharma V, Chandel S. Performance analysis of a 190 kWp grid interactive solar photovoltaic power plant in India. Energy 2013;55:476e85. [17] Padmavathi K, Daniel SA. Performance analysis of a 3 MW p grid connected solar photovoltaic power plant in India. Energy Sustain Dev 2013;17(6): 615e25. [18] Kumar BS, Sudhakar K. Performance evaluation of 10 MW grid connected solar photovoltaic power plant in India. Energy Rep 2015;1:184e92. [19] Sundaram S, Babu JSC. Performance evaluation and validation of 5 MW p grid connected solar photovoltaic plant in South India. Energy Convers Manag 2015;100:429e39. [20] Edalati S, Ameri M, Iranmanesh M. Comparative performance investigation of mono-and poly-crystalline silicon photovoltaic modules for use in gridconnected photovoltaic systems in dry climates. Appl Energy 2015;160: 255e65.
11
[21] Sidi CEBE, Ndiaye ML, El Bah M, Mbodji A, Ndiaye A, Ndiaye PA. Performance analysis of the first large-scale (15 MW p) grid-connected photovoltaic plant in Mauritania. Energy Convers Manag 2016;119:411e21. [22] Adaramola M. Viability of grid-connected solar PV energy system in Jos, Nigeria. Int J Electr Power Energy Syst 2014;61:64e9. [23] Mondal M, Sadrul Islam A. Potential and viability of grid-connected solar PV system in Bangladesh. Renew Energy 2011;36:1869e74. [24] El-Shimy M. Viability analysis of PV power plants in Egypt. Renew Energy 2009;34:2187e96. [25] Adaramola M. Techno-economic analysis of a 2.1 kW rooftop photovoltaicgrid-tied system based on actual performance. Energy Convers Manag 2015;101:85e93. [26] Edalati S, Ameri M, Iranmanesh M, Tarmahi H, Gholampour M. Technical and economic assessments of grid-connected photovoltaic power plants:Iran case study. Energy 2016;114:923e34. [27] Orioli A, Di Gangi A. Review of the energy and economic parameters involved in the effectiveness of grid-connected PV systems installed in multi-storey buildings. Appl Energy 2014;113:955e9369. [28] The Dynamis Project, http://www.dynamisproject.co.nz/. [29] Akinyele D, Rayudu R, Nair N. Global progress in photovoltaic technologies and the scenario of development of solar panel plant and module performance estimation- Application in Nigeria. Renew Sustain Energy Rev 2015;48:112e39. [30] Labed S, Lorenzo E. The impact of solar radiation variability and data discrepancies on the design of PV systems. Renew Energy 2004;29(7):1007e22. [31] Enasolar Grid-tied Solar Inverter, http://www.enasolar.net/Products/ EnaSolar-GT-Series-Solar-Inverter-v2-3f. [32] Solar Photovoltaic Energy Systems, IEC/TS 61836 (2007-12). [33] PV System Performance Monitoring, IEC 61724. [34] Dobaria B, Pandya M, Aware M. Analytical assessment of 5.05 kWp grid tied photovoltaic plant performance on the system level in a composite climate of western India. Energy 2016;111:47e51. [35] Design of grid connect PV systems, Tech. rep., Palau Public Utilities Corporation. [36] IEEE guide for array and battery sizing in stand-alone photovoltaic (PV) systems. 2008. [37] Akinyele D, Rayudu R, Tan RH. Comparative study of Photovoltaic technologies based on Performance, cost and space requirement: strategy for selection and application. Int J Green Energy 2016;13(13):1352e68. [38] Akinyele D, Rayudu R, Nair N. Development of photovoltaic power plant for remote residential applications: the socio-technical and economic perspectives. Appl Energy 2015;155:131e49. [39] Kim J-Y, Jeon G-Y, Hong W-H. The performance and economical analysis of grid-connected photovoltaic systems in Daegu, Korea. Appl energy 2009;86(2):265e72. [40] Tech. rep. Power system operation and augmentation planning with PV integration. International Energy Agency; 2014. http://www.iea-pvps.org/ index.php?id¼322. [41] Grid-connected solar PV systems-no battery, Design guidelines for accreditted installers. Clean Energy Council; 2013. [42] Akinyele D, Rayudu R. Community-based hybrid electricity supply system: a practical and comparative approach. Appl Energy 2016;171:608e28. [43] Renewable Power Generation Costs in 2014 (IRENA), http://www.irena.org/ documentdownloads/publications/irena_re_power_costs_2014_report.pdf/. [44] Palm T, Qayum A. Private and public investment analysis. South-Western Pub. Co.; 1985. [45] Short W, Packey DJ, Holt T. A manual for the economic evaluation of energy efficiency and renewable energy technologies. University Press of the Pacific; 2005. [46] A. Miller, A. Wood, M. Hwang, S. Lemon, E. G. Read, Economics of photovoltaic solar power and uptake in New Zealand. [47] Dorf RC. The technology management handbook. Heidelberg: Springer; 1999. [48] National Institute of Water and Atmospheric Research, http://www.niwa.co. nz/our-services/online-services/solarview. [49] Emmanuel M, Rayudu R. Evolution of dispatchable photovoltaic system integration with the electric power network for smart grid applications: a review. Renew Sustain Energy Rev 2017;67:207e24. [50] Emmanuel M, Rayudu R. Communication technologies for smart grid applications: a survey. J Netw Comput Appl 2016;74:133e48.
Please cite this article in press as: Emmanuel M, et al., Techno-economic analysis of a 10 kWp utility interactive photovoltaic system at Maungaraki school, Wellington, New Zealand, Energy (2016), http://dx.doi.org/10.1016/j.energy.2016.11.107