Solar Photovoltaic Systems
SPV systems fall broadly in two categories. Smaller systems used for home/office are known as off-grid/ grid interactive systems. These systems are installed typically on roof tops and have varying installed capacity, from < 1 kWp to few hundreds of kWp. Off-grid/grid interactive systems can have provisions for storage using a batteiy. However, the cost of SPV systems with storage increases depending on the amount of storage required. Apart from applications of providing electricity to homes/offices, smaller SPV systems are also used for water pumps, street lights, cell phone towers and many other such applications. A representative Single Line Diagram (SLD) of off-grid/grid interactive SPV system is shown in Figure 11.
An SPV array consists of numbers of solar modules connected in series and parallel, decided on the basis of the installed capacity and array design. The details of array design are described later. A Power Controller Unit (PCU) consists of a DC-DC converter, Maximum Power Point Tracking (MPPT) controller, Charge Controller (CC) and DC-AC inverter. The DC-DC converter and MPPT controller ensure that the DC power provided by the SPV array is always at MPP. A charge controller is required for management of the power flowing in and out of the batteiy. During a charging cycle, the power is stored in the battery. Although this primarily comes from the SPV, it is possible to charge the batteiy from the grid power. The power stored in the batteiy can be supplied to the load (through DC-AC inverter) when the SPV power is not available, e.g., after sunset. DC-AC converts the DC power to AC power for supplying to the load. In entirely off-grid systems, also known as standalone SPV systems, the grid power is not available. The load requirements are met from the SPV dining the day and from the battery dining the night. Part of the power generated by SPV dining the day has to be stored in a batteiy in addition to the power supplied to the load.
The charge controller is not required if the SPV system has no storage. The PCU consisting of only DC-DC converter along with MPPT and DC-AC inverter is known as “solar inverter”. Efficiency of solar inverters are significantly higher than PCUs with charge controllers. As the batteiy has a high cost, SPV systems are designed with no storage or minimum storage in order to cater to essential loads at night or dining power outages. The amount of power produced by an SPV array having a specific installed capacity depends on irradiance level and temperature, which varies throughout the day and the year. It is possible that even during daytime, the power required by the load is more than the power produced by an SPV array. In such situations, the excess required power is provided by the grid. Such systems are known as grid-interactive SPV systems.
The estimation of energy output of smaller SPV systems with a specific installed capacity is done using calculations based on “peak sun hour”, which is defined next. Such calculations are not very accurate and a long-term average is taken. It is to be noted that accurate instantaneous power calculations are not a requirement for smaller systems. For example, a household with a connected load of 4 kW, for example, does not require 4 kW all the time, as all the appliances are not switched on simultaneously. The average use of the load, which is also not precisely defined, is important. The SPV generation is also not fixed throughout the day as the irradiance and the ambient temperature vary. The estimation of peak sun hour of a particular place is done by plotting the irradiance versus the time of day, as shown in Figure 12. The irradiance increases from zero after the sun rises from the horizon and reaches its maximum when the sun is at the zenith position. During the afternoon, the intensity decreases and goes to zero at sunset. This is shown by the solid curve of Figure 12. The power produced by a solar panel varies as the irradiance changes. For example, a 265 Wp SPV module will produce 265 W only when the irradiance is 1000 W/m-, if the temperature effect is not considered. The temperature effect will be added later.
The area (A, in Figure 12) under the Irradiance-Time (between sunset and sunrise) curve is the energy received per day. An equivalent time has to be estimated for a normalized power of
Figure 11. SPY system with storage.
Figure 12. Estimation of peak sun hour
1000 W/iir. This is done by finding out an equivalent area corresponding to the power of 1000 W/nr (see Figure 12). The area (A2) of the rectangle ABCD is equal to Aj. The corresponding time (AD) is known as “peak sun hour”. This essentially means that the total energy received per day is (1000 W/m2 x Peak sun horn-). For example, if peak sun hour per day value at a particular place is 4 In s, a 265 Wp solar module will produce (265 W x 4) = Ю60 Wh of energy per day. As the peak sun hour per day value is not same throughout the year, an average value is taken for energy generation calculations. For example, a 265 Wp SPV module may produce less or more than the average value of 1060 Wh, or 1.06 kWh, depending upon the inadiance at that particular day. A long-term average is more accurate. For example, the 265 Wp SPV module is estimated to produce (1.06 x 365) kWh = 386.9 kWh energy per year. This estimation is more accurate. The peak sim horn' per day values for different locations are published by various agencies. For tracking-based systems, the peak sun hour value at a particular location is higher than that of fixed systems at that location. In fixed SPV systems, solar modules are kept at fixed orientation with an angle equal to the latitude of the place of installation. The modules face south in the northern hemisphere and face north in the southern hemisphere. In single axis tracking-based SPV systems, the solar modules are oriented in the east-west direction and rotate as per the sun movement during the day; from morning to evening. Dual axes tracking-based SPV systems also take seasonal sun movement into account. Duel axes tracking is not generally used for conventional SPV systems as the complexity overrides the benefits achieved. However, in Concentrated Photovoltaic (CPV) systems, the use of duel axes tracking is a must.
The temperature effect also has to be considered for the estimation of energy. A 265 Wp SPV module will produce 265 W at 1000 W/m2 only when the cell temperature is 25 °C. The ambient temperature varies during the day and throughout the year, depending on the weather conditions. Here also, a longterm average is considered in order to estimate the ambient temperature. As mentioned earlier, the cells of a module heat up during irradiation. Therefore, the cell temperature is more than the ambient temperature duiing operation. This heating due to inadiance is also not constant, as the irradiance itself varies. The estimation is typically done by assuming the average cell temperature is about 20 °C more than the ambient. Therefore, if the ambient (average) temperature at a particular location is 35 °C, then the cell temperature (average) is about 55 °C (35 °C + 20 °C). Temperature co-efficient (see section 6) is used to incorporate temperature effect into the energy estimation. If the cell temperature is 55 °C, which is 30 °C (55 °C-25 °C) more than STC, the power of the module changes by-12% (-0.4% x 30). This indicates 12% less power than STC. A 265 Wp module will produce 31.8 W (12% x 265 Wp) less power. The power produced by the module at an effective cell temperature of 55 °C is, therefore, 233.2 Wp (265 W-31.8 W). Considering both inadiance and temperature effect, the 265 Wp module produces 932.8 Wli (233.2 x 4) per day and about 340.5 kWh per year, if the peak sun hour and the ambient temperature of that location are 4 In and 35 °C, respectively, and with the assumption that the cell temperature is 20 °C higher than that of the ambient. The difference between the cell temperature and the ambient temperature is more for tracking systems than fixed systems.
The design of an SPV system starts with the determination of required installed capacity based on the load profile of the customer. The energy generation estimation of a specific installed capacity SPV plant can be done using the method described above. For example, based on the load profile of a household, it has been estimated that the household would require a total of 20 kWh energy per day. Out of that, 15 kWh is used during the day and the rest (5 kWh) dining the night. This can be estimated by listing all the electrical appliances (lights, fans, air conditioning units, laptops, TVs, fridges, etc.) being used, along with their wattage rating and the operating hours per day. A 4 kWp solar installation would be required in case the peak sun hour per day value at the location of installation is 5 Ins. The energy produced for this 5 kWp system is 20 kWh (4 kW x 5 Hrs). Out of this, 5 kWh is stored in the battery and the rest (15 kWh) is used to cater to the load during the daytime. During the night, the requirement is met from the stored energy in the batteiy. A simplistic view is presented here and several other issues, such as efficiency of PCU/Solar Inverter, efficiency of the batteiy during both charging and discharging cycles, provision for cloudy day (redundancy), etc., are not included in these design calculations.
In case 265 Wp modules are used, a 4 kWp SPV system would require 15 such modules. Few options are available for array design. It can have one string consisting of 15 modules (1 x 15) or 3 strings of 5 modules (3 x 5) or 5 strings of 3 modules (5 x 3). As discussed earlier, a string consists of series- connected modules and strings connected in parallel to complete the array. The string design is done by considering the voltage specifications, mainly the maximum voltage and operating voltage range of MPPT of the PCU/Solar Inverter. The maximum current is also seen, but not very important as the PCU/Solar Inverters have very high current ratings. The maximum voltage of a 1 x 15 array is 561.75 V (Voc x 15 = 37.45 V x 15), see Table 3 for V^ value. The operating voltage is 456.9 V ('m x 15 = 30.46V x 15). The operating current is 8.7 A (Im x 1 = 8.71 A x 1). The maximum current (Isc) is not considered here as the inverter does not come into the picture during a short circuit condition. A (3 x 5) array has a maximum voltage of 187.25 V, operating voltage of 152.3 V and operating current of 26.13 A. Similarly, a (5 x 3) array has a maximum voltage of 112.35 V, operating voltage of 91.38 V and operating current of 43.55 A. The array design must be made while keeping the PCU/Inverter specifications in mind. Apart from maximum voltage, both minimum and maximum of the MPPT voltage range must considered. For more accurate calculations, the temperature effects on Voc and Vm must be considered. In case more than one of the array configurations are found suitable, the final choice can be made to optimize the utilization of the available space.
The bigger SPV systems, starting from 1 MW and going up to hundreds of MW, are used for generating electricity which is directly exported to the grid. Such SPV systems are called “on grid” or SPV power plants. A representative Single Line Diagram (SLD) of such systems is shown in Figure 13.
Figure 13. SPY power plant.
In addition to the SPY array and the solar inverter, a transformer is required in order to convert the output of the inverter, which is typically at 415 V, to high voltage (11 KV, 33 KV, 66 KV, etc.), depending on the requirement of the electricity grid. The power is then evacuated to the grid at high voltage.
The energy estimation of SPY power plants has to be done accur ately. The cost of energy pumped to the grid is paid to the installer at a predecided price per unit (kWh). This is typically paid per year basis. Incorrect estimation of total energy to be produced per year' can upset the business model. Apart from total energy generated per year, the accurate estimation of energy generated during a shorter time-scale is also important. The power generated by SPY plants is combined with the conventional power and then supplied to the consumer. It is important, therefore, to know how much power is available from the SPV plant at different times, so that proper planning can be done. The data related to energy generation per hour or even per minute are important for such planning. The simple calculations which are used for off-grid/grid interactive SPV systems, described earlier, cannot be used for energy estimation for on grid SPV systems. The inaccuracies, particularly for the shorter duration energy generation associated with such calculations, are high.
Simulation tools are used for estimation of energy generation in SPY7 power plants. Apart from the energy estimation, such tools are capable of configuring string design, array design and simulating performance-related parameters, such as Performance Ratio (PR), which is defined later. There are some “free to use” simulation tools, which are widely used for academia and beginners. PY'WATT (www. pvwatts.nrel.gov) from NREL is one such example. Some reputed solar inverter companies also provide free-to-use simulation tools. The most widely-used simulator is PY7Syst (www.pvsyst.com). A brief description on PY'Syst simulation methodology has been given below.
One of the main inputs required for PY'Syst is the irradiance and the ambient data for the entire year. The accuracy of the energy estimation strongly depends on how accurate this data is. PY'Syst gets default irradiance and ambient data, which is freely available, from NASA. Data available from METEONORM (www.meteonorm.com) is known to have better accuracy. This is available at a price and can be seamlessly imported to PY’Syst for simulation. “Paid” and “free” data are available from other sources as well. Location-related input is provided by specifying the latitude of the place. Other important inputs are SPY7 module parameters, inverter specifications, soiling loss, DC and AC cable-related losses, transformer loss, transmission loss, etc. The simulator can provide design parameters, such as optimum tilt angle for SPY7 module, string/array design, etc. The loss diagram, the energy output and the PR are the primary output parameters.
The energy output per month, as shown in Figure 14, is most commonly used. In this, actual output of the plant is also shown for comparison. It is possible to obtain the energy output for an hour or for the entire day. However, the long-term average, such as for an entire year; is more accurate. For example, the energy output estimation done for an hour is less accurate than the energy output estimation done for a day. Similarly, energy output estimation done for a year is more accurate than the energy output estimation done for a month. A closer look at Figure 14 reveals that there are some inaccuracies between the simulated energy output and actual energy obtained for each month. The overall inaccuracy reduces significantly w'hen the energy estimation is done for a year'. The inaccuracy is as high as +4.5% for the
Figure 14. Simulated and actual energy of a SPY power plant per MWp mstallation.
month of April. However, the inaccuracy of the total energy estimated for the entire year is only +0.6%. This is a tracker-based SPV plant and the total actual energy output per year is 1,741,451 kWh per MWp installation. The corresponding simulated value of 1,730,277 kWh per MWp installation.
The Performance Ratio (PR), which is also estimated by PVSyst, is another important parameter. This is defined as given in equation (8).
where E is the energy obtained in kWh, Ir is the irradiance in kWh/т2, A is the total module area in in2 and q is the efficiency of the module.
PR essentially provides the energy output, normalized with respect to irradiance. The quality of SPV plants is judged on their PR values. Exactly identical plants installed in two different geographical locations produce different energy outputs, as the irradiance and the ambient conditions are different. However, the PR values are expected to be similar, as the difference of irradiances between these two locations are nullified for calculation of PR. It should be noted that the temperature effect is not normalized. Therefore, ideal PR of two identical SPV plants may be somewhat different. Good quality plants have PR values of 85% or more. In case the PR value of a SPV plant is less than 80%, it falls under the categoiy of poor quality. The PR estimation is typically done on an annual basis.
Due to the widespread development of solar PV, the cost has come down drastically and there are ample examples of economic viability of solar PV plants. In most places, it has crossed the grid parity milestone, which means that the cost of electricity generated from solar PV is less than the cost of electricity generated from conventional resources. As the subsidy is no longer required for such scenarios, accelerated growth is now expected. However, the major hindrance is still the requirement of a very large initial investment. The economic viability depends on the reliability of the components used, particularly the solar modules. The modules are expected to last 25 years or more and the Levelized Cost of Energy (LCOE) is calculated accordingly. However, if the modules last only 5 years, for example, due to poor reliability, the economic viability is no longer achieved. Although warranty commitment from the module manufacturers are intended to take care of this, it is not fool proof. Under the warranty, the manufacturers commit that the modules will generate power for at least 25 years with a typical degradation of 10% in first 10 years and another 10% in next 15 years. However, 25 years is a long time, and a particular manufacturer may not last that long. Moreover, due to the advancement of technology, the module specifications are continuously been upgraded, so if a module has to be replaced 10 years down the line, it may not be possible to get one with the same specifications. Additional safeguards, in terms of insurance coverage, are being introduced gradually. It is expected that there will be a continuous improvement of trust due to maturing of the technology and the introduction of more such safeguard mechanisms.
Exponential growth of solar PV may create another major issue related to coal-based power plants. As solar electricity is becoming cheaper than coal-based electricity, the former will be given preference. However, the entire supply during the night has to be provided by coal-based electricity. The Capacity Utilization Factor (CUF) of coal-based power plants is then compromised, putting stress on their economic viability. It is, therefore, prudent to have a planned growth strategy, keeping both solar and coal-based power plants in mind.
Ahmed, G.,Mandal, S., Вата, A.K., Bhattacharya, T.K. and Roy, J.N. 2017. Band offset reduction at defect-rich p/i interface through a wide bandgap a-SiO:H buffer layer. IEEE Journal of Photovoltaics 414.
Ahmed, G., Mandal, S., Ваша, A.K., Bhattacharya, T.K. and Roy, J.N. 2019. Mixed-phase nc-SiO:H interlayer to improve light trapping and shunt quenching m a-Si:H solar cell. IEEE Journal of Photovoltaics 18.
Antomadis, H., Jiang, F., Shan, W. and Liui, Y. 2010. All screen-printed mass-produced silicon ink selective emitter solar cells. Proc. ofPhotovol. Specialist Conf. (PYSC) 1193.
Bose, D.N. 2012. Organic Photovoltaics m Semiconductor Materials and Derices: New Age, Delhi.
Boyle, G. 2004. Renewable Energy: Oxford University Press.
Chuang, C.H.M., Brown, P, Bulovic, V. and Bawendi, M.G. 2014. Improved Performance and stability in quantum dot solar cells through band alignment engineering. Nature Materials 796.
Cusano, D A. 1963. CdTe solar cells and photovoltaic heterojunctions in II—VI compounds: Sol. St. Electron 217.
Foster, R., Ghassemi, M. and Cota, A. 2010. Solar Energy, Renewable Energy and Environment. CRC Press.
Geisza, J.F., Friedmann, D.J., Ward, J.S., Duda, A., Olavarria, W.J., Monarty, T.E., Kiehl, J.T., Romero, M.J., Norman, A.G. and Jones, K.M. 2008. 40.8% efficiency inverted triple-junction solar cell with two independent metamorphic junctions. Applied Physics Letters 123505-1-3.
Gilbert, M.M. 2013. Renewable and Efficient Electric Power Systems. IEEE Press John Wiley & Sons.
Gratzel, M. 2003. Dye sensitized solar cells. Journal of Photochemistry and Photobiology C. Photochemistry Reviews 145.
Green, M.A. 1995. Silicon Solar Cells-Advanced Principles and Practice. Bridge Piinteiy, Sydney.
Hamid, F. and Fatima, B. 2013. Characterization and modeling of CdS/CdTe heterojunction thin film solar cell for high efficiency performance. International Journal of Photoenergy 13.
Ikegami, S. 1988. CdS/CdTe solar cells by the screen-printing-sintering technique: Fabrication, photovoltaic properties and applications. Solar Cells 89.
Kiefer, K., Ulzhofer, C., Brendemuhl, X, Hairier, N.P, Brendel, R., Mertens, V, Bordihn, S., Peters, C. and Muller, J.W. 2011. High efficiency N-type emitter wrap through silicon solar cells. IEEE Journal of Photovoltaic 53.
Kronik, L., Cahen, D. and Schock, H.W. 1998. Effects of sodium on polycrystalline Cu (In,Ga)Se, and its solar cell performance. Advanced Materials 31.
Markvart, T. and Castaner, L. 2005. Practical Handbook of Photovoltaics: Fundamentals and Applications. Elsevier.
Magnone, P, Rose, R.D., Tonini, D., Frei, M., Zanuccoh, M., Belli, A., Galiazzo, M., Sangiorgi, E. and Fiegna, C. 2014. Numerical simulation on the influence of via and rear emitter in MWT solar cell. IEEE Journal of Pliotovolt 1032.
Min, C.,Nuofu, C., Xiaoli, Y., Wang, Y., Yiming, B. andXingwang, Z. 2009. Thermal analysis and test for single concentrator solar cells. J. of Semiconductors 4.
Nazeerusddin, M.K. and Snaith, H.K. 2015. Perovskite Photovoltaics: MRS Bulletin 40.
Nelson, V. 2011. Introduction to Renewable Energy. CRC Press.
Ohlsen, H.J., Bodegairi, M., Kylner, A., Stolt, L,, Hanskos, D., Ruckh, M, and Schock, H.W. 1993. ZnO/CdS/Cu(In,Ga)Se, thin film solar cells with improved performance. Proc. 23rd IEEE Photovoltaic Specialists Conference 364.
Roder, T.C., Eisele, S.J., Grabitz, P, Wagner, C., Kuluslnsh, G., Kohler, J.R. and Werner, J.H. 2010. Add-on laser tailored selective emitter solar cells. Progr. Photovolt. Res. Appl. 505.
Roy, J.N. and Bose, D.N. 2018. Photovoltaic Science and Technology. Cambridge University Press.
Sam-Shajing, S. and Niyazi Serdar, S. 2005. Organic Photovoltaics: Mechanisms, Materials, and Devices (Optical Engineering). CRC Press, Taylor & Francis Group, Boca Raton.
Sharps, P.R., Stan, M.A., Aiken, D.I., Clevenger, В , Hills, J.S. and Fatemi, N.S. 2003. Multi-junction cells with monolithic bypass diodes. World Conference on Photovoltaic Energy Conversion 626.
Street, R.A. 2005. Hydrogenated Amoiphous Silicon. Cambridge University Press.
Wang, Z.A. and Green, M.A. 1994. Series resistance caused by the localized rear contact in high efficiency silicon solar cells. Solar Energy Materials and Solar Cells 89.
Wang, Z.A., Altennatt, A.P and Green, M. 1995. Twenty four percent efficiency silicon solar cells with double layer antheflection coating and reduced resistance loss. Applied Physics Letter 3636.
Zhao, J., Wang, A. and Green, M.A. 1999. 24.5% Efficiency Silicon PERT Cells on MCZ Substrates and 24.7% Efficiency PERL Cells on FZ Substrate: Progress in Photovoltaic: Research and Applications 471.
Zanuecoli, M., Magnone, P, Sangiorgi, E. and Fiegna, C. 2015. Analysis of the unpact of geometrical and technological parameters on recombination losses in mterdigitated back contact solar cells. Solar Energy 37.