PHOTOVOLTAICS Photovoltaic conversion is the direct conversion of sunlight into electricity with no in- fervening heat engine. Photovoltaic devices are solid state; therefore, they are rugged and simple in design and require very little maintenance. Perhaps the biggest advan- tage of solar photovoltaic devices is that they can be constructed as stand-alone sys- {ems to give outputs from microwatts to megawatts. That is why they have been used as the power sources for calculators, watches, water pumping, remote buildings, com- munications, satellites and space vehicles, and even megawatt scale power plans, Photovoltaic panels can be made to form components of building skin, such as roof shingles and wall panels. With such a vast array of applications, the demand for photo- voltaics is increasing every year. In 1995, 80 MW, of photovoltaic panels were sold for the terrestrial markers, In the early days of solar cells in the 1960s and 1970s, more energy was required to produce a cell than it could ever deliver during it lifetime. Since then, dramatic imn- provements have taken place in the efficiencies and manufacturing methods. In 1996, the energy payback periods were reduced to about two and a half to five years, depend ing on the location of use [11], while panel lifetimes were increased to over 25 years, ‘The costs of photovoltaic panels have come down to $5-10 per peak watt over the last ‘two decades and are targeted to reduce to around $1.00 per peak watt. Historically, the photoelectric effect was first noted by Becquerel in 1839 when light was incident on an electrode in an electrolyte solution [I]. Adams and Day first observed the effect in solids in 1877 while working with selenium. Early work was done with selenium and copper oxide by pioneers such as Schottkey, Lange and Grandahl. In 1954 researchers at RCA and Bell Laboratories reported achieving effi- ciencies of about 6 percent by using devices made of p and n types of semiconductors, ‘The space race between the USA and the Soviet Union resulted in dramatic improve- ‘ments in the photovoltaic devices. Ref. [3] gives a review of the early developments in photovoltaic conversion.1 t a ‘PRINCIPLES OF SOLAR ENGINEERING 9.1 SEMICONDUCTORS ‘A basic understanding ofthe atomic structure is quite helpful in understanding the be- havior of semiconductors and their use as the photovoltaic energy conversion devices ‘Any fundamental book on physics or chemistry generally gives adequate background orbasie understanding. Ref. [2] presents an in-depth treatment of « numberof topies in semiconductor physics. For any atom, the electrons arrange themselves in orbitals around the nucleus so as to resll in the minimum amount of energy: Table 9.1 shows the distribution of the “leetrons in various shells and subshels in light elements. In elements that have elec- ftens in multiple shells, the innermost electrons have the minimum energy and there fore, require the maximum amount of externally imparted energy to overcome the a traction ofthe nucleus and become free. Electrons in the outermost band of subshells tie the only ones that participate in the interaction of an atom with its neighboring atoms. If these electrons are very loosely attached to the atom, they may attach them- selves with a neighboring atom to give that atom a negative charge, leaving the orig- thal atom as a positive charged ion, The postive and negatively charged ions become attached by the force of attraction ofthe charges thus forming fonic Donds. Ifthe elec. irons in the outermost band do not fil the band completely but are not loosely attached tither, they arrange themselves so that neighboring atoms can share them to make the ‘Gutermnost bands full The bonds thus formed between the neighboring atoms are called covalent bonds. ‘Since electrons in the outermost band of an atom determine how an atom will - actor join with a neighboring atom, the outermost band is called the valence band. ‘Some electrons in the valence band may be so energetic that they jump into a still pigher band and are so far removed from the nucleus that a small amount of impressed Tasce would eause them to move away from the atom. Such electrons are responsible for the conduction of heat and electricity, and this remote band s called a conduetion band. 'The difference in the energy of an electron in the valence band and the inner- ost aubshell ofthe conduction band i called the Band Gap, or the forbidden gap. ‘Materials whose valence bands are full have very high band gaps (>3eV). Such saterials are called insulators. Materials, onthe other hand, that have relatively empty ‘Valence bands and may have some electrons inthe conduction band are good conduc- ors. Metals fallin this category. Materials with valence bands party filled have iner- ‘podiate band gaps (=3eV). Such materials are called semiconduetors (Fig. 9.1), Pure semiconductors are called intrinsic semiconductors, while semiconductors doped with ‘ery small amounts of impurities are called extrinsic semiconductors. I the dopant ‘ratorial has more electrons inthe valence band than the semiconductor, the doped ma- {arial is called an n-type of semiconductor. Such a material seems to have excess elec- trons available for conduction even though the material js electronically neutral. For xampe, silicon has four cletrons in the valence band. Atoms of pure silicon arrange themselves in such a Way that, to form a stable structure, each atom shares two elec trons with each neighboring atom with covalent bands. If phosphorous, which has five valence electrons (one more than Si) i introduced as an impurity in silicon, the doped raterial seems to have excess electrons even though itis electrically neutral, Such a ‘Table 9.1". Electronic structure o anne " Principal quantum nmber 2 ‘Avimtha quantum number! ‘Letter designation of state Z Symbol Bement 1H Hydrogen 2 He __ Helium > Eihiom 4 Be Benylim 5B Boron 6 © Carbon 7 ON Nitrogen 8 0 Oxygen oF Fluorine [a ee Na Sodium 12 Mg Magnesiom 3 Al Aluminum Si Silicon iP Phosphorus bs salt 17 CL Chlorine 18 A Anon 1 K Poussiom 2 Ca Cakiom 21 Se —_‘Seandium 2 Titanium 2 ¥ Nanaium 24 Ce Chromium 25 Mn Manginese 26 Fe kon 27 Co Cobalt ON Nickel 23 Gu Copper 300 Zn Zine 31 Ga Gallium 32. Ge Germania 33 As Anenic Mo Se ‘Selenium 35 Br Bromine 36 Ke Krypton ane ‘From Charlot E. Moore, Atomic ards Circa 467, 1982PHoTOVvoLTAIcs a3 ‘Table 9.1°. Electronic structure of atoms Pci quanto mambern we 3 witli. ‘Amati quanam number ee ee Letter designation of state 5 | > | | »| 3 | | «| Z__ Symbol Flement —_¥, vot Ho Hyaeogen 1360 | He Helm 2458 | 2 Li Litho «839 1 Be Benyiam 9.32 2 B Boron 530 eer c can tts. |B] 2 | 2 N Nios st | g | 2 | 3 © owen =r | g | 2 | F Fone 1742 | & | 2 | 5 Ne Neon 2136 2/6 Ne Sodio Sa 1 Mg Magnesia 7.64 2 Al Aluminum 598 a}4 Si Sitcom 818 | Neoneoe | 2 | 2 Phosphor: 1055 at Ss Saltur 1036 2)4 Cl Ohlre 13.07 alte A Aron 15.75 ee 1K Poastum ase H 20 Ce Cake 61 2 21 Se Scandium 6.36 1f2 2 TT Tianam 6a 22 2 Y——Vanadum 674 ar 2 Ch Chromium 6.75 sa 25 Ma Manganese 7.43 s|2 2 Fe hom 790 s|2 Co Cotmk 786, ‘Argon Cove 7] 2 2% ON Nickel 783 a2 2 Co Copper am 1 30 Zn Zine 939 ww} 2 3 Gs Gallum 600 w}2 |1 32 Ge Oeemaniam 788 w}2]2 33 AS Aenie oat w}2|3 34 Se Seleniam 9:75 wo} 2 |4 35 Br Bromine 188 w}2|s 36 Ke Krypton 400 wl2 fo ‘From Charlot E, Moore, Atomic energy level, Vol. Il, Washington, D.C. National Burea of Stan- ards Circular 467, 1982a4 PRINCIPLES OF SOLAR ENGINEERING raat Sentetn ot 008 | tae is ert Figure 9.1. Electrical conduction is described in terms of allowed and forbidden energy bands Band gap for insulators (athe highest, followed by semiconductors c) and metal (B),respectively. doped material is called n-type silicon. If on the other hand, silicon is doped with Boron which has three valence electrons (one less than Si), there seems to be a positive hole (missing electrons) in the structure, even though the doped material is electrically neutral. Such material is called p-type silicon, Thus, n- and p-type semiconductors ‘make it easier forthe electrons and holes respectively to move in the semiconductors. 9.141 p-n Junetion As explained earlier, an n-type material has some impurity atoms with more electrons than the rest of the semiconductor atoms. If those excess electrons are removed, the impurity atoms will fit more uniformly in the structure formed by the main semicon. ductor atoms; however, the atoms will be left with positive charges. On the other hand, ' p-type material has some impurity atoms with fewer electrons than the rest of the. semiconductor atoms. Therefore, these atoms seem to have holes that could accommo. date excess electrons even though the atoms are electrically neutral (Fig. 9.2). If addi- tional electrons could be brought to fill the holes, the impurity atoms would fit more uniformly in the structure formed by the main semiconductor atoms, however, the atoms will be negatively charged. The above scenario occurs at the junction when a p and an n type of material are joined together as shown in Fig, 9.3. As soon as the two materials are joined, “excess” 00000 ®0000 oogo0o ®O000®8 00000 00000 @0000 00000 90000 00000 (a) (b) Figure 9.2. Representation ofm and pp semionductn0mype showing ets” ecu a dos) pypestowing “ece pole bls. n-type Q. Q Gag Co @ p-typ Figue9.3 “Excess cacuoes om sm juntion, leaving the n-side ofthe ‘harees electrons from the 1 layer jump to filth junction, the material has positive cha side. The negative charges on the p sid from the n side to the p side, while the side to the m side is made easier because side. This restriction makes the p-n jum ofa p-n junction is made use of in solar 9.1.2 The Photovoltaic Effect When a photon of light is absorbed by s electron is increased by the amount of & is equal to or more than the band gap energy will jump into the conduction & photon energy is less than the band gap jump into the conduction band. In this verted to excess kinetic energy of the e ture. Ifthe absorbed photon had more the band gap simply increases the kines photon can free up only one electrom es band gap. This fact is a big reason for devices. The key to using the photovs ‘channel the free electrons through am ¢ the holes. This is achieved with the hel Figure 9.4 shows a schematic of a erated in the 7 layer by the action of pk cireuit, recombine with positive holes layer. The negative charges in the p la in that direction. Ifthe n layer is made and, therefore, the probability of recorPHOTOvOLTAICS ais n-type | junction ddddd p-type Figure 9.3, “xces” electrons from n-matrial amp to il excess" hoes om the psd of _-m junction, leaving the n-sde of the junction positively charged and the p-side negatively charged. s from the n layer jump to fill the “holes” in the p layer. Therefore, close to the jon, the material has positive charges on the 1 side and negative charges on the p ‘The negative charges on the p side restrict the movement of additional electrons the m side to the p side, while the movement of additional electrons from the p to the n side is made easier because of the positive charges at the junction on the n . This restriction makes the p-n junction behave like a diode. This diode character {a p-n junction is made use of in solar photovoltaic cells as explained below. 12 The Photovoltaic Effect ‘a photon of light is absorbed by a valence electron of an atom, the energy of the is increased by the amount of energy of the photon. Ifthe energy of the photon equal to or more than the band gap of the semiconductor, the electron with excess ‘will jump into the conduction band where it can move freely. If, however, the ‘energy is less than the band gap, the electron will not have sufficient energy to {nto the conduction band. In this case, the excess energy of the electrons is con- to excess kinetic energy of the electrons which manifests in increased tempera- If the absorbed photon had more energy than the band gap, the excess energy over ‘band gap simply increases the kinetic energy ofthe electron. It must be noted that a can free up only one electron even if the photon energy is lot higher than the ‘sap. This fact is a big reason for the low conversion efficiency of photovoltaic s. The key to using the photovoltaic effect for generating useful power is to J the free electrons through an external resistance before they recombine hholes. This is achieved with the help of the p-n junction Figure 9.4 shows a schematic of a photovoltaic device. As free electrons are gen- in the » layer by the action of photons, they can either pass through an external t, recombine with positive holes in the lateral direction or move toward the p - The negative charges in the p layer at the p-n junction restrict their movement ‘that direction. Ifthe m layer is made extremely thin, the movement of the electrons therefore, the probability of recombination within the m layer is greatly reduceda6 PRINCIPLES OF SOLAR ENGINEERING photon i c~_n=type 4 +O 4 0,8 404 FO, ewer SS Se ioe j p-type 4 Figure 94, Schematic o photovoltaic device Unless the extemal circuit is open. Ifthe external circuit is open, the electrons gener- ated by the action of photons eventually recombine with the holes, resulting in an in- crease in the temperature of the device. In atypical crystalline silicon cel, the n layer is about 0.5 jm thick and the p layer is about 0.25 mm thick. As explained in Chapter 2, energy contained in a photon E, is tsiven by E, = ho, 9.) ‘where is the Planck's constant (6.625 X 10-¥ J-sec) and v is the frequency which is related to the wavelength A and the speed of ight c as, Therefore, 02) For silicon, which has a band gap of 1.11 eV, the following example shows that photons of solar radiation of wavelength 1.12 um or less are useful in creating elec- tron-hole pairs. This spectrum represents a major part of the solar radiation, Table 9.2 lists some candidate semiconductor materials for photovoltaic cells along with their band gaps. Example 9.1. Calculate the wavelength of light capable of forming an electron- hole pair in ‘Table 9.2. Energy gap for some car Material Band gap (eV) si Ltt sic 260 CAs, 1.00 cate 14s case 178 cas 202 CAs00, 290 Gahs 40 GaP 224 cas. 180 cud “a Case 1 Calas, 150 Culnse, rol Solution, The band gap energy Fore = 3 x 108 m/sec, h = 6.625 equation gives the required wavelem (6.625 x 10-843 4 pd Example 9.2. A monochromat ‘of 638 nm is incident on a silicon so the number of photons per seco ' the maximum possible efficiem Solution 4, The intensity of light in the las init. If the number of photons 1,= Np 1X 107°W = NyPHoTovoLtaics a7 >. Energy gap for some candidate materials for photovoltaic cells (7). Bandgap @V) Material Band gap (eV) culate, 090 ine 127 nyt 120 2340 160 ate 20 z8e 260 AP 243 AIS 163 Anse, 60 Sb,Se, 120 Ge 067 Se 160 Solution, The band gap energy of silicon is 1.11 eV. From Eg. (9.2) we can write he 4 & For c= 3 x 108 m/sec, h = 6.625 X 10-! J - sec, and 1 eV = 1.6 % 10-¥ J, the above equation gives the required wavelength as: (6.625 10°. seo) 108 m/sec) _ a (LUIML6 X 10-3) peda Example 9.2. A monochromatic red laser beam emitting 1 mW at a wavelength ‘of 638 nm is incident on a silicon solar cell. Find: ‘a. the number of photons per second incident on the cell; and b. the maximum possible efficiency of conversion of this laser beam to electricity. Solution a. The intensity of light in the laser beam ([,) is equal to the energy of all the photons init If the number of photons is N,,, then I= NaE, 3) 1X 109W = Ny E,W = 11/300)a8 PRINCIPLES OF SOLAR ENGINEERING E,= hela = (6.625 x 10™ J — sec)-3 x 10" m/sec mr = 312 x 10-95 1X 107 I/see = 339 x 19-07 ~ 321 X 10! photons /sec b. Assuming that each photon is able to generate an electron, a total number of ‘Njq electrons will be generated. Therefore, the electrical output will be equal to Vn’ (Band Gap). Therefore, the maximum possible efficiency is WNy(BG) _ BG. Wa)E, _ LAL x 1.6 x 10-5 312 x 10°F Ma 0.4) = 569 oF 56.9%. From the above examples, itis clear that fora silicon solar cell, none of the pho- tons of the sunlight over 1.12 zm wavelength will produce any electricity. However, * oe et 2s Soe a | cas. ve ° | aay Figure 9.5. The maximum solar energy conversion efficiency a a function ofthe energy gap ofthe semiconductor. The measured efficiency for various materials denoted by a solid square enetsy gap of the material The curve as been callated for an del junction outside the atmosphere photons of sunlight at a wavelength ‘maximum efficiency of 100 percent, ‘verted at lower efficiencies. The over integrating the efficiency at each wan In addition, other factors such as pro theoretical maximum achievable eff ison of the maximum energy conven Figure 9.5 shows that the optin 1seV. 9.2 ANALYSIS OF PHOTOVE ‘This section presents an electrical in the design of photovoltaic devi to the expressions for the number d at a temperature 7 will not be pres to books such as Refs. [1] and [2p jjunetion a current is generated call the net current due to the current . ‘current or the reverse saturation rent J, from the n side to the p si of electrons in the p material ex ceasily move to the m side to fill the rent J,, Normally, the electrons ocet in the n material do not have enoug! side, However, if a foward bias ve due to the action of the photons of ‘enough energy to cross over and re rise (0 a light-induced recombina given by 4 = charge of an electron 1,602 x 10-"° Coulom Boltzman’s Constant The junction current J, isthe mPHOTOVOLTAICS aw of sunlight at a wavelength of 1.12 sm may be converted to electricity at « ‘efficiency of 100 percent, while photons at lower wavelengths will be con- lat lower efficiencies, The overall maximum efficiency of a cell can be found by fing the efficiency at each wavelength over the entire solar spectrum: f naan a es) fra ition, other factors such as probability of electron-hole recombination reduce the sal maximum achievable efficiency of a solar cell. Figure 9.5 shows a compar- the maximum energy conversion of cells using different materials ‘Figure 9.5 shows that the optimum band gap for terrestrial solar cells is around Vv. ANALYSIS OF PHOTOVOLTAIC CELLS -seotion presents an electrical analysis of a photovoltaic cell which will be useful design of photovoltaic devices for various applications. The physics leading ‘expressions for the number density of electrons and holes in m and p materials ‘emperature T will not be presented here. For such details, the reader is referred sks such as Refs. [1] and (2). It would suffice to point out here that at the p-n na current is generated called the junction current. The junction current J, is ‘pet current due to the current J, from the p side to the m side (called the dark t or the reverse saturation current) and a light induced recombination cur- 4J, from the n side to the p side. Based on the temperature 7, a certain number ‘electrons in the p material exist in the conduction band. These electrons can fly move to the n side to fill the holes created at the p-n junction, generating a cur: J, Normally, the eleetrons occupying the conduction band due to the temperature the’n material do not have enough potential energy to cross the p-n junction to the p . However, if a foward bias voltage V is applied, which in a photovoltaic cell is to the action of the photons of light, some of the electrons thus generated have \gh energy to cross over and recombine with the holes in the p region. This gives to a light-induced recombination current J,, which is proportional to J, and is nby Texple,V/kD, 06) ‘where ¢, = charge of an electron {602 X 10-' Coulombs or J/V, k = Boltzman’s Constant = 1.381 % 10-23 J/K ‘The junction current J, is the net current due to J, and J,20 PRINCIPLES OF SOLAR ENGINEERING ° 5 Tipe Jn junetion 7 et p= type RL (a) fees ° on yy dL Ss fs R (b) ° Figure 9.6. Equivalent circuit of photovoltaic ell yrs d; 1 fexp(e,V/KT) — 1} Referring to Fig. (9.6), itis clear thatthe current generated in the cell has two par- allel paths, one through the junction and the other through the extemal resistance R,, Figure (9.6) shows an equivalent circuit of a photovoltaic cel. It must be pointed out here that the current generated in a photovoltaic cell, including the junction current, is proportional to the area of the cell. Therefore, itis appropriate to analyze in terms of the current density J (current per unit area) instead of the current /. The relationship between the two is: vA (0.8) Referring to Fig, (9.6), we can write: J ae 9.9) = Jylexple,V/kD) ~ 11, ‘where J, is the short cireuit current, For short circuit, V = Oand J, = J, For open circuit, J, which gives Figure (9.7) shows a typical ps ‘output is the product of the load eur tance where A is the area of the cell. ‘The power output exhibits a m power output (P.,.). differentiate P exple,Vq/l where V,, stands for voltage at max Jj», and the maximum power P... tively. 010 020 v Figure 9.7. Typical caren, voltage andPHOTOVOLTAICS a gives Jo lexp(eVo/kD) ~ 1h 0.10) v= Hull) Figure (9.7) shows a typical performance curve (I-V) of a solar cell. The power is the product of the load current and voltage and is a function of the load resis- P= ALV=1Y eon) STR Ais the area of the cell. ‘The power output exhibits a maximum. To find the condition for the maximum. ‘output (P,,) differentiate P with respect to Vand equate it to zero: Vn) wy + 2s exote altn (1 + Se) = 1 4 (12) V,, stands for voltage at maximum power. The current at max power condition and the maximum power P,,,, can be found from Eqs. (9.9) and (9.11), respec i‘ | ose 10 5 QD Max power i 0.90) \\ point 0.40 x iS oso £ : i Eos H on & a 0.10 O10 020 090 040 050 060 070 V = Voltage (volts) Figure 9.7. Typical curren, voltage and power characteristics ofa solar cell Adapted from (6)See "flaca 7H pz sof A dtmurer/ pene | | uur voltage, vats Figure 94. Typical cumentvoltge characteristics of silicon cell showing the effect of illumination evel ad ‘earn, Jun = J, = Jy lesple,Va/kD) ~ 1) Combining Eqs. (9.12) and (9.13), J, s found tobe eVolKT TH (e,Vq/Ty * 20> afer 7 eV q/ kD" So Jum . Maximum efficiency f. The cell area required for an « 900 Wim?, Solution. Given: and Tr a, Using Eg, 9.10), Since ¢ / will be needed for 1.60 clk 80 Therefore, 0.13) (9.14) ». Voltage at maximum power co of trial and error solution: (0.15) exp(37.06V,\1 Figure (9.8) shows the effect of illumination intensity and the load resistance on the performance of a silicon cell. Temperature also affects the performance in such ; ‘way that the voltage and thus the power output decreases with inereasing temperature, cc. Current density at maximum po Example 9.3. The dark current density for a silicon solar cell at 40°C is 1.8 10-* A/m? and the short circuit current density is 200 A/m?. Calculate: 4 Open circuit voltage b. Voltage at maximum power . Current density at maximum power 4. Maximum powerPHOTOVOLTAICS oy ‘Maximum efficiency ‘The cell area required for an output of 25 W when exposed to solar radiation of 900 Wim. Solution. Given: J, = 18% 10-¢A/m? J, = 200.A/m?, T= 40°C = 313K. Using Eq. (9.10), vg = Ean 41) Since ¢ /&T will be needed for other parts also it can be evaluated separately as 1.602 x 10-"1/V e/kT = 2 AON = 37.06"! KT G3a1 x 10°? 1/K)GI3K) ~*7°¥ Therefore, 1 200 = snaatliax 0) = 0.624 Voltage at maximum power condition can be found from Eq. (9.12) by an iterative or trial and error solution: 200 expG7, 37.06V,) = 1+ p37. 06V,)(1 + 37.06¥,) = 1 + eS or Vj, = 542V. ‘. Current density at maximum power ean be found from Eq. (9.14): eVq/ KT 1+ Vq/kD * 7? (67.06) (0.542) “T+ G7.06).(0:543) 190.5A/m = (200 + 1.8 x 10-9)A/m? 1m4 PRINCIPLES OF SOLAR ENGINEERING a. Pie Waeinds P, ; AN = (0.542¥)-(190.5A/m?) = 103.25 W/m?, 103.25 W/m? _ e. ta = 900 Wat = SH f, Cell area required: hx Fim 25w Ppo/A 108.25 W/m? = 420m, 9.2.1 Efficiency of Solar Cells ‘Theoretical limitation on the efficiency ofa solar cell can be calculated using Eq, (9.5), ‘These efficiency limitations and the practical efficiencies of some of the cells are shown in Figure 9.5. Some of the reasons for the actual efficiency being lower than the theoretical limitation are: 1. Reflection of light from the surface of the cell, This can be minimized by anti- reflection (AR) coating. For example, AR coatings can reduce the reflection from ‘Si cell to 3% from more than 30% from an untreated cell 2. Shading of the cell due to current collecting electrical contacts, This can be mini mized by reducing the area of the contacts and/or making them transparent, how- ever, both of these methods will increase the resistance of the cell to current flow. Internal electrical resistance of the cell 4. Recombination of electrons and holes before they can contribute to the current. This effect can be reduced in polycrystalline and amorphous cells by using hydro- gen alloys. 9.2.2. Multijunction Solar Cells ‘The limits imposed on solar cells due to band gap can be partially overcome by using ‘multiple layers of solar cells stacked on top of each other, each layer with a band gap higher than the layer below it, For example (Fig. 9.9), if the top layer is made from a cell of material A (Band gap corresponding to A,), solar radiation with wavelengths, less than A,, would be absorbed to give an output equal to the hatched area A. The solar radiation with wavelength greater than A,, would pass through A and be converted by the bottom layer cell B (Band gap corresponding to A,) to give an output equal to the hatched area B. The total output and therefore the efficiency of this tandem cell would Intensity W/m" Figure 9.9. Energy conversion fom at be higher than the output and the e ciency would increase with the mun ‘must be as thin as possible, which straint on crystalline and polyerys this concept is being investigated m At present, a triple junction a of layers of cells made from a-Si,C: loy) with a band gap of 2.0 eV, a-Si band gap of 1.75 eV, and a-Si, Ge gen alloy) with a band gap of 23 about 50 percent higher than a com tal device structure of a triple june 9.2.3 Design of a Photovoltaic Solar Cells may be connected in se and current, When similar cells or and current are as shown in Figure currents as shown in Figure 9.12. acteristics, the output characterist nected to form modules, modules nected to form arrays. Principles s connections.PHoTOvoLTaIcs os Intensity W/m* Wavelength Energy conversion from a we-layered stacked cel higher than the output and the efficiency of each single cell individually. The effi- -y would increase with the number of layers. For this concept to work, each layer be as thin as possible, which puts a very difficult if not an insurmountable con- int on crystalline and polycrystalline cells to be made multijunction. As a result, ‘concept is being investigated mainly for thin film amorphous solar cells. ‘At present, a triple junction a-Si solar cell is under development. This cell consists ayers of cells made from a-Si,C:H (an amorphous Silicon, Carbon and hydrogen al- ) with a band gap of 2.0 eV, a-Si:H (an amorphous Silicon and hydrogen alloy) with Spand gap of 1.75 eV, and a-Si, Ge:H (an amorphous Silicon, Germanium and hydro ‘alloy) with a band gap of 2.3 eV. The efficiency of a multijunction cell can be '50 percent higher than a corresponding single cell. Figure 9.10 shows a concep- device structure of a triple junction a-Si cell Design of a Photovoltaic System Cells may be connected in series, parallel, or both to obtain the required voltage ‘current. When similar cells of devices are connected in series, the output voltages current are as shown in Figure 9.11. A parallel connection results inthe addition of mts as shown in Figure 9.12. Ifthe cells or devices 1 and 2 have dissimilar char- istics, the output characteristics will be as shown in Figure 9.13. Cells are con- ted to form modules, modules ate connected to form panels and panels are con- fed to form arrays, Principles shown in Figures 9.11 and 9.12 apply to all of these tions.26 PRINCIPLES OF SOLAR ENGINEERING Bandgap Materiol Deposition Thickness (ev) Process (nm) Bake, 1.8 = 21 mm Sn0p:F Textured tin oxide APCVD | 500 - 800 ~1.95 | o-SiC:H:8 PecvD | 10 172 Fi O-Si:H, Hpdiluted PECVD | 60 - 80 - 172 [ay ue-SEHiP PECvO | 10 ~1.95 [2 o-SiG:HB. PEcvo | 10 1.72 | ig o-Si:H, Hpdiuted PECVD | 300 - 450 1.72. [Tn we SEP PECVD | 10 ~1.95 [Ps o-SIGH:B PEOVD m14a [ig o-SiGe:H, Hadiluted PECVD | 160 — 190 1.72 [ng Me~SiHiP Pecvd | 10 - 50 Magnetron} 80 — 100 Sputtered: ‘Ag or All Sputtered Polyurethane: Figure 9.10. Typical wipte-unction ai cll structure Source: 4.oo seus om 30 SPAIN “TS AHA “ous u pore 1 A28 PRINCIPLES OF SOLAR ENGINEERING 1 hth i a 1 seem ° ® vows we r= 4k oe Figure 9.12. Characteristics of wo similarclls connected in parallel Ie 1th, 1 | = te te I 1 h : WVe vate Y WV Ye (a) (o) gare 913, Charteris of wo isinlr cells connected in (series, nd (para Example 9.4. An application solar cells from Example 9.3 each Solution, Assuming that the « tions, the voltage and current from Vi, = 0542V, Power/Cell = Number of cells ‘Number of cel in ser Number of rows of 52 ee Since the number of rows must be rows which will give 303 W out ‘Abblocking diode is used ins {rom flowing backward, for exam tions. A bypass diode is use in par if it is shaded. A photovoltaic sy shown in Figure 9.15, Figure 9.14. Connection of cls in rowsPHOTOVOLTAICS 29 Example 9.4, An application requires 300 Watts at 28V. Design a PV panel using solar cells from Example 9.3 each with an area of 6em® Solution. Assuming that the cells will be operated at maximum power condi- tions, the voltage and current from each cell are Vy, ostav, y= (1005 8) 6x 10m mm 0.11434, Power/Cell = 0.542 * 0.1143 = 0.062W, 300W Number of cells required = > oxy joo = 4840+ her of cells required = 5 oer joan 4840 System Voltage __28V Voltage/cell — 0.542V— 4840 $240 931. 52 az ‘Number of cells in series ‘Number of rows of $2 cells connected in parallel Since the number of rows must be a whole number we may increase the number to 94 rows which will give 303 W output (Fig. 9.14. "A blocking diode is used in series with a module or an array to prevent the current from flowing backward, for example, from the battery to the cells under dark condi- tions. A bypass diode is used in parallel with a module in an array to bypass the module if it is shaded. A photovoltaic system may be connected to a DC or an AC toad as shown in Figure 9.15. 52 cells eat 303W @ 28V 9.14, Connection of ells insows and columas for Problem 9420 PRINCIPLES OF SOLAR ENGINEERING = l B7 el es) ea) te l Figure 9.18. Schematic ofa photovoltai sytem, 9.3 MANUFACTURE OF SOLAR CELLS AND PANELS ‘Manufacture of crystalline silicon solar cells is an outgrowth of the manufacturing, ‘methods used for microprocessors. A major difference is that silicon used in micro- processors is ultra pure, which is not needed for photovoltaic cells. Therefore, a major source of feedstock for silicon solar cells has been the waste material from the micto- electronics industry. Solar cells are also manufactured as polycrystalline and thin films, Below are some of the common methods of manufacture of silicon solar cells 9.3.1 Single Crystal and Polycrystalline Cells ‘Single crystal silicon cells are produced by a series of processes: (1) growing crys- {alline ingots of p-silicon, (2) slicing wafers from the ingots, (3) polishing and cleaning the surface, (4) doping with n material to form the p-n junction, (5) deposition of elec- trical contacts, (6) application of anti-reflection coating and (7) encapsulation, Figure 9.16 illustrates the process. Czochralski Method (Fig. 9.17(a)) is the most common method of growing sin- tle crystal ingots. A seed crystal is dipped in molten silicon doped with a p-material (Boron) and drawn upward under tightly controlled conditions of linear and rotational speed, and temperature. This process produces cylindrical ingots of typically 10 cm di- ‘ameter, although ingots of 20 em diameter and more than 1 m long can be produced for other applications. An alternative method is called the float zone method (Fig 9.17(b)). In this method a polycrystalline ingot is placed on top of a seed crystal and the interface is melted by a heating coil around it. The ingot is moved linearly and ro: Woiten Soler | | Pelyeryatat ingots] | Wafer Slicing Potishing |Grade Siticon|__.| D0 0 and Lapping ras with | — Seo Waters Reposition of ing with | woes | [nam | [EP*) [= pes | [nom] (Eee pec & a ae es | Figure 9.16. Series of processes forthe manufacture of crystalline Lpoyerystallne cells (@) Czochralski Method Figure 9.17. Crystalline silicon ingot tationally, under controlled condit cost Polycrystalline ingots are prod (rectangular) as shown in Figure 9 cone direction in order to orient th ferred direction. In order to achiew ‘rater than 0.5 mm are needed an lar to the wafer. Ingots as large as method. Ingots are sliced into wafers b pregnated with diamond abrasive wastage of valuable crystalline sil Alternative methods that red films. Some ofthe thin film prog 9.19), Edge-Defined Film-fed Gro © [igure 9.18, Polycrystalline ingot producPHOTOVOLTAICS, a Let Polerytaline Sai erst silicon ali Motte sone SWS 7 Movable Heater Winding /_— Single Crystal Seed Molten silicon persists ethos, (b) Float Zone Method 9.17. Crysalline silicon ingot production mets ly, under controlled conditions. This process has the potential to reduce the cell Polycrystalline ingots are produced by casting silicon in a mold of preferred shape at) as shown in Figure 9.18. Molten silicon is cooled slowly in a mold along direction in order to orient the crystal structures and grain boundaries in a pre~ direction. In order to achieve efficiencies of greater than 10 percent, grain sizes than 0.5 mm are needed and the grain boundaries must be oriented perpendici to the wafer, Ingots as large as 400 cm 40 em 40 cm can be produced by this Ingots are sliced into wafers by internal diameter (ID) saws or multiwire saws im- xd with diamond abrasive particles. Both of these methods result in high of valuable crystalline silicon. ‘Alternative methods that reduce wastage are those that grow polycrystalline thin ‘Some of the thin film production methods include Dendric Web Growth (Fig. ), Edge-Defined Film-fed Growth (EFG) (Fig. 9.20), Ribbon against drop (RAD) Cy) ott igure 9.18. Polyerystllne ingot peodution,432 PRINCIPLES OF SOLAR ENGINEERING ‘Supporting dendrite “Molten aiicon Figure 9.19. Thin lm production by deni web growth, Figure 9.20. Thin lm production by edge defined filmed growth (EFC). |, Supported Web method, a 921). ‘A pn junction is formed in th jorous) in the top layer ofa p ‘of phosphorous in the vapor pl to prevent the diffusion of w a solid layer ofthe dopant mate (800-900") diffusion. Electrical contacts are attached sro more than 10 percent of the the back surface, The front grid ps jon through a mask or by screen pe form modules, Sllcon waters | | wane | Figure 9.21. Schematic of RAFT processPHOTOVOLTAICS 3 |, Supported Web method, and Ramp assisted foil casting technique (RAFT) 9.21). 'A pn junction is formed in the cell by diffusing a small amount of m material norous) inthe top layer ofa p-sificon wafer. The most common method is diffu- ‘of phosphorous in the vapor phase. In this case, the back side of the wafer must be zed to prevent the diffusion of vapors from that side. An alternate method is to de- ‘ca soli layer of the dopant material on the top surfaces followed by high tempera- (800-900°) diffusion. Electrical contacts are attached to the top surface of the cell in a grid patter 10 -cno more than 10 percent of the cell surface and a solid metallic sheet is attached the back surface. The front grid pattern is made by either vacuum metal vapor depo- through a mask or by sereen printing. Figure 9.22 shows how cells are connected form modules. stleon waters | il | 7 Thepection Regener Figure 9.21. Schematic of RAFT processing4 PRINCIPLES OF SOLAR ENGINEERING Figure 9.22. Assembly of solr cell to form a module Antireflection (AR) coatings of materials such as silicon dioxide (Si0,), titanium dioxide (TiO,) and tantalum pentaoxide (Ta,0,) are deposited on the cell surface to re- duce reflection from more than 30 percent for untreated Si to less than 3 percent. AR Coatings are deposited by vacuum vapor deposition, sputtering or chemical spraying Finally, the cells are encapsulated in a transparent material to protect them from the en vironment. Encapsulants usually consist of a layer of either polyvinyl butyryl (PVB) or ethylene vinyl acetate (EVA) and a top layer of low iron glass. 9.3.2 Amorphous Silicon and Multijunction Thin Film Fabrication Amorphous Silicon (a-Si) cells are made as thin films of a SiH alloy doped with phos- ‘Phorous and boron to make n and p layers respectively, The atomic structure of ana Si cell does not have any prefered orientation. The cells are manufactured by depositing A thin layer of a-Si on a substrate (glass, metal or plastic) from glow discharge, sputter. ing or chemical vapor deposition (CVD) methods. The most common method is by an RF glow discharge decomposition of silane (SiH, on a substrate heated toa tempera ture of 200-300°C. To produce p-silicon, diborane (B,H,) vapor is introduced with the silane vapor. Similarly phosphene (PH) i used to produce n-Silicon. The cell consists of an n-layer, an intermediate undoped a-Si layer, anda p-layer on a substrate. The cell thickness is about um. The manufacturing process can be automated to produce rolls of solar cells from rolls of substrate. Figure 9.23 shows an example of roll-to-rol a-Si cell manufacturing equipment using a plasma CVD method, This machine can be used ‘o make multjunction or tandem cells by introducing the appropriate materials at dif- ferent points in the machine 9.4 DESIGN FOR REMOTE PHOTOVOLTAIC APPLICATIONS, Photovoltaic power may be ideal for a remote application requiring a few watts to hun- dreds of KW of electrical power. Even where a conventional electrical grid is avai able, for some applications, where-uninterruptible or emergency standby power is nec- essary, photovoltaic power would be cations include water pumping for p mote houses, stret lighting, battery lay stations, and weather stations electrical utility switching stations, p quality is aconcern, data acquisition lation fans and vaccine refrigeration. ‘The design of a PV system is base |. Which is more important, the da . Is back-up energy source need Is energy storage important? Wh Is the power needed as AC or DC ‘There are three basic steps in the Estimation of load and load prof Estimation of available solar rah .. Design of PV system, including. and electrical system schematic. Each of these steps will be explaines based on Refs. [12] and [13]. 9.4.1 Estimation of Loads and | Precise estimation of loads and their tems since the system is sized as thePHOTOVOLTAICS Stainless Steet (anode) RF Generator 13.56 MHz Figure 923, A schematic diagram ofa rll plasma CVD machine (8). essary, photovoltaic power would be appropriate. Some examples of remote PV appli- cations include water pumping for potable water supply and irrigation, power for re- ‘mote houses, street lighting, battery charging, telephone and radio communication re- lay stations, and weather stations. Examples of some other applications include, electrical utility switching stations, peak electrical utility power where environmental quality is a concern, data acquisition systems and speciality applications such as venti- lation fans and vaccine refrigeration, ‘The design of a PV system is based on some basic considerations forthe application. Which is more important, the daily energy output or the power (average or peak)? Isa back-up energy source needed and/or available? Is energy storage important? What type—Battery, pumped water etc.? Is the power needed as AC or DC? What voltage? ‘There are three basic steps inthe design of a PV system: Estimation of load and load profile; Estimation of available solar radiation; and Design of PV system, including area of PV panels, selection of other components, and electrical system schematic, Each of these steps will be explained in the following examples, These examples are based on Refs. [12] and [13]. 9.4.1 Estimation of Loads and Load Profiles Precise estimation of loads and their timings (load profile) are important for PV sys- tems since the system is sized as the minimum required to satisfy the demand over a436 PRINCIPLES OF SOLAR ENGINEERING " day. For example if power is needed for five diferent appliances requiring 200, 300W,SOOW,L000W and 1S00W respectively so that ony one apliane son at any foes inns one ime and each appliance is on form average of I hour a day the PV stom Would ae be sized based on 1300W peak power and 3S00Watt hous (Wh of daly energy te Lana wae quirement. The multiple loads on a PV system are intentionally staggered to use the Lights (AC) ous xt sialest possible system, since the capital oss of PV sytem are the most impetant —— as opposed othe energy costs in a conventional fel based system, ee Soe Pando ono BS ve " Example 9.5. Daily /oad calculations, How much energy per day is used by a re- Beers GAS ‘mote weather station given the following load characteristics? Average DC load: (770 + 2807 = 1 Pe an ‘Average AC load: [3220 + 1092 + 8 io Taira an Taine toy Pacis Carp Comet 20 7 Peak AC load: (1000) + max (4923 Dan pho to ; ieee et ate 9.4.2 Estimation of Available S Ta a ad lane eal ae ere ae Methods of estimation of available » {erm measured solar radiation values can be used to estimate the average day can be used along with percent (2.0W)(8 hr) + (OW) hr) + (1.5W)22.5 br) + GOW)2.5 hr) system a de@iion ia inade VA 106.75 Wh 5 ‘will be fixed ata certain tilt and aziem Daily Energy use is about 107 Wh per day. +15° works best for winter and lait round an angle equal to the latitude Solution. Daily energy Example 9.6. Load calculations. An owner of a remote cabin wants to install a 9.24), PV power system. The loads in the home are described below. Assume that all lights and electronics are powered by AC. Find the daily and weekly peak and average en- ergy use estimates. The system used is a 24-volt DC system with an inverter. 9A:3 EV System Sista If meeting the load at all times is not Ligh 4, 23.yat compact Muoreeent bulbs On atnghe Fors hoors average values of ergy cua Ligh 6 13-wat compact orescent balls 2hours each daytime) ee Stereo 110 watts (amplifier, 1S watts (other) ‘On for 8 hours per week ‘system designed by this approach wi Water pump SS wats (375 amp stat caren) ns foe 2hoors per dy start for a detailed design. A detalles Computer 280 wats monitor included) (On for 1 ours daily cies of the components depending (reeked nights only) thee ths sya i GOS Bathyoom fan 40 watts (3.5 amp start curen) On for | hour pe day whether heaven Soa Microwave 550 wats (AC)-1000 W surge On for 30 mints per dy ‘mode. Example 9.7. Heuristic appros Solution, Loads need to be broken down according to: (1) run-time, (2) peak watt, 12-volt panels with 6-volt, 125 power, (3) night or day use, and (4) AC or DC louds, The load profile is as follows: Farmington, NM, with a daily load cam be assed that he pomp a ‘ici assumed that the night and dy lighPHOTOVOLTAICS ‘Run ime (hows) nergy (Wh) Day Weck Day en wan © (ox13) ai) ax (yS5) GB 75AK24V) (yas) ——_cx2s0) (4) SAK24V) (x50) Cx1000) erage DC load: [770 + 280/7 = 150 Whiday. erage AC load: [3220 + 1092 + 880 + 750 + 1925] 7 = 1124 Whiday. DC load: max {{(3.5)(24) + 40) =: ((3.75)(24) + 55)] = 145 watts {AC load: (1000) + max [(4)(23):: (6,13)] + 250 + 100 = 1442 wa 2 Estimation of Available Solar Radiation is of estimation of available solar radiation are descried in Chapter 2. If long ‘measured solar radiation values are available at a location, Eqs. (2.56) to (2.64) a be used to estimate the average solar radiation per day. Otherwise, data for clear can be used along with percent sunshine data (if available). For designing a PV sm, a decision is made whether the PV panel will be operated as tracking the sun or ill be fixed at a certain tlt and azimuth angle. For fixed panels, a tilt angle of latitude 15° works best for winter and latitude ~ 15° for summer. To keep the panel fixed year ‘an angle equal to the latitude provides the maximum yearly energy (see Figure 124). 943 PY System Sizing ‘Wimeeting the load at all times is not critical, PV systems are usually sized based on the “average values of energy and power needed, available solar radiation, and component iciencies. This is known as the heuristic approach. It is important to note that a sm designed by this approach will not give the best design but may provide a good ‘Sart for a detailed design. A detailed design accounts for the changes in the efficien- ‘cies of the components depending on the load and the solar radiation availability and ‘whether the system is operating in a PV-to-load, PV-to-storage, or storage-to-load ‘mode. Example 9.7. Heuristic approach to PV system sizing. A PV system using 50- ‘watt, 12-volt panels with 6-volt, 125 amp-hour batteries is needed to power a home in Farmington, NM, with a daily load of 1700 watt-hours. System voltage is 24 volts 'ivcam be assumed that the pomp and fan wil ot tr precisely athe same instant, siti assumed thatthe night and daylighting loads will ot be on simultaneous38 PRINCIPLES OF SOLAR ENGINEERING Tt = Lot 48 rocking Setor ‘Raatation Feed ot tt Tine of Day Wont Figure 9.24 Solarsaition on panes a iferent tit anges ‘There are an average of 5 daylight hours in the winter. Specify the collector and stor- age values forthe system using the heuristic approach, Solution. Load = 1700 Whiday, Daylight hours Average Panel output 1700 Wh/Day Number of panels. = a (5 har/day)(SOW/panel) 8, round off to 7 panels. Since the system voltage is 24V, but each panel produces only 12 volts an even ‘number of panels will be needed. Therefore number of panels = 8. Farmington, NM, is a very sunny location, so 3 days of storage is plenty. Assum- ing a battery efficiency of 75 percent and a maximum depth of discharge 70 percent, Storage (1700)(3)4(0.75 x 0.7) = 9714 Wh, Number of Batteries = (9714/Wh)/(125Ah x 6¥) = 13 (Rounded off to next whole number). Since the system voltage is 24V, and each battery provides 6V, the number of batteries is increased to 16, Ina detailed design, the efficiencies of battery storage, inverter, and, the balance of system (BOS) must be accounted for. The following example shows hhow these efficiencies increase the energy requirements of the PV panel Example 9.8. System operating efficiency. Using the cabin electrical system from Example 9.6, calculate the overall system efficiency for each operating mode possible for the system. Estimate the amount of energy required per day for the system. When load timing (day or night), assume half of the load runs during the day and half runs at night. The inverter used has a component efficiency of 91 percent, the battery effi- ciency is 76 percent, and the distribution system efficiency is 96 percent. Solution, From the example, the loads are: Average DC load: 150Wh/day, ‘Average AC load: 1124 Whiday. ‘The various system efficiencies are: PVtoload (DC): 0.96 Battery to load (DC): (0.76)(0.96) = PVtoload (AC): (0.9640.91) = Battery to load (AC): (0.76091 [Expected day and night loads are: Day (DC): (0.5110) + 0.540) Night (DC): (0.5)(110) + 0.540) Day (AC); (156) + (0.5880 * 7 Night (AC): (460) + (0.5)(880 +7 Without considering system efficiene Egy = (150) ‘The expected daily energy requirem Egy = (15)/00.96) + TSG Fay ‘The actual energy requirement is 35 9.4.4 Water Pumping Applica ‘Water pumping for drinking water plication of PV. For a simple sche pump water at a volumetric rate V water, gis the acceleration due to head H, is (A +B). In case the (A+B + ©). The pump must wo namic head H, aso, ‘where His the frictional head loss water at the pipe outlet. The pum flowrate) and is available as a chaPHOTOVOLTAICS: erage DC load: 1S0Whiday, age AC load: 1124 Whiday. various system efficiencies are toload (DC: 0.96 (day, DO), ry load (DC): (0.76\.0.96) (night, DO), toload (AC): (0.96)(0.91) (day, AQ), tery to load (AC): (0.760.91)(0.96) = 0.664 (nighAC). spected day and night loads are: (DC): (0.5)(110) + (0.5)(40) = 75 Whiday, (0.5)(110) + (0.5)(40) = 75 Whiday, (156) + (0.5880 + 750) /7 + (0.5)(275) = 409.9 Whiday, (460) + (0.5)(880 + 7507 + (0.5)(275) = 713.9 Whiday. thout considering system efficiency, the daily energy requirement is: Egy = (150) + (1124) = 1274 Wh/day. he expected daily energy requirement is: Egy = (15)/{0.96) + (75)/(0.73) + (409.9)/(0.874) + (713.9)/(0.664) 1725 Wh/day actual energy requirement is 35 percent higher than simple calculation. 4 Water Pumping Applications ter pumping for drinking water or irrigation at remote locations isan important ap- ation of PV. For a simple schematic shown in Figure 9.25 the power needed to imp water ata volumetric rate Vis given by P = pVeHn, where pis the density of er, i the aceleration due to gravity, and 1, isthe pimp efficiency. The static dH, is (A + B). Incase the water level is drawn down, te static head would be +B + C). The pump must work against the total head H which includes the dy- mic head H, also, e y= Hy +3 ‘where His the frictional head loss inthe pipe and the bends and w i the velocity ofthe ‘water at the pipe outlet. The pump efficiency 7, is a function of the load (head and flowrate) and is available as a characteristic curve from the manufacturer. For general“0 PRINCIPLES OF SOLAR ENGINEERING st Lorry] Storage Lets) Tank [_] Submersible Pump Figure 9.25. Water pumping using submersible pomp design purposes typical values given in Table 9.3 may be used. The table lists two basic types of pumps, centrifugal and positive displacement. These pumps can be dri- ven by AC or DC motors. DC motors are preferable for the PV applications, because they can be directly coupled to the PV array output. Centrifugal pumps with sub- ‘mersible motors are the optimum for PV applications because of their efficiency, reli bility and economy. However, for deep wells Jack pumps may be necessary. Jack ‘pumps are the piston type of positive displacement pumps that move chunks of water Table 9.3. Typical range of pump performance parameters. Wire-wo-water Type pump efficiency () os Cent 15.25 om Centrifugal with ee 10-20 Sobmerble 20-30 Sobmersie 30-40 Jack pump 3045 100 Jack pump 35.30 ‘with each stroke. They require through batteries Example 9.9. A PV system is: El Paso, TX. The following infor Solution. Since the required needed. Assuming a fricti 105) = 122 m. We select a 7SV D Daily Energy Required = = Since the system will be used from ‘optimum. Daily solar radiation at E June 11.51 kWh/m*day July 10.58 -KWh/m?-day. ‘August 10.02 KWhim?-day Since August has the minimum ins for this month. We select PV panels may be obtained from a manufactur Voltage V,. = 21.7 Current 15358 ‘Temperature Correction Factor for ‘Assuming that the panels will ope 60°C, the voltage at the highest expPHOTOVOLTAICS an each stroke. They require very large currents, therefore they are connected batteries. Example 9.9. A PV system is designed to pump water for livestock in the vicinity Paso, TX. The following information is available: Near EL Paso, TX, 32 20'N 10640" W 1670m, ~stoasc (Cased Borehole 15 emt 106 m 8m 8325 Uday, Jane vo Avg, Solution. Since the required head is very high, a deep well jack pump with DC is needed. Assuming a friction factor of 0.05 the total head = 106 + 8 + (106) 122m. We select a 75V DC Jack pump with an average efficiency of 0.45. Daily Energy Required = pV gH/1, Seat to 1 = (8321/4 wo ‘ (esi Byazame ts the system will be used from June to August, a tilt angle of lat. - 15° would be Daily solar radiation at El Paso for | axis tracking and tilt angle of 17° is: LSI kWhim?-day 10.98 _KWh/m2-day 10.02 KWhim?-day August has the minimum insolation, the panel area will be based on insolation ‘month, We select PV panels with the following specifications (this information be obtained from a manufacturer): V,, @ 25° = 174V ' 134A sre Correction Factor for voltage = 1 ~ 0.0031 (T, — T,) ing that the panels will operate at a maximum temperature of 4: the voltage at the highest expected temperatureefor Problem 99 Figure 9.26. A system schen Voltage Required Therefore, number of panels in ser Power output/pane! wim? 31Ax 1 Power output/panel/d ‘Assuming an overall efficiency of Number of panels r ‘The array will consist of 3 parallel Asa check, the daily water pumps = X485 (os = 8798 I/day ‘Therefore, the system will meet th PROBLEMS 9.1 Find the wavelength of rai ‘gap of GaAs. 9.2, What is the theoretical maxi Jength 0.45 jm is incident 9.3, Find the theoretical maxima 9.4, The reverse saturation curre short circuit current when ¢ compute a. open circuit voltage, ’. max. power output ofthePHOTOVOLTAICS = 17.4 [1 = 0.00331 (60 ~ 25)) = 155V Voltage Required = 75V. “Therefore, number of panels in series = 75/15.5 ~ 5. ‘Power outpuv/panel = 3.1 A X 15.5 V = 48.05 W at a standard insolation of 1000 Wim? 10.02KWh/m? — day Power output/panel/day = 48.05W x wife oy ia iogehcugbe IKW/m? = 481.5 Wh/day. “Assuming an overall efficiency of 90 percent due to insolation times, wiring ete 6,150 Wh/day Number of panels required = 55 = ber of panels required = 9.55 481.5 Wh/day he array will consist of 3 parallel rows of 5 panels each in series. ‘As acheck, the daily water pumping rate for August 5 X 0.9 Wh/day-3600 sec/br 15 x 481.5 X 0.9 Wh/tday-3600 sec/br 5 45 ‘981 Jc Kg/(122 m 8798 I/day ‘Therefore, the system will meet the requirement. A system schematic is in Fig. 9.26. PROBLEMS 9.1 Find the wavelength of radiation whose photons have energy equal to the band ‘gap of GaAs. 9.2. What is the theoretical maximum efficiency of conversion if blue light of wave- Jength 0.45 jum is incident on a GaAs solar cell? 93. Find the theoretical maximum overall efficiency of Gas solar cells in space, 4, The reverse saturation current I, ofa silicon cell at 40°C is 1.8 X 10-7 amp. The short circuit current when exposed to sunlight is 5 amp. From this information ‘compute a, open circuit voltage, 'b. max. power outputof the cell, and