Solar PV/Thermal including Beam Splitting Dr. Albert Lin & Ahmad Mojiri School of Engineering [email protected] [email protected] Building 57, Level 3, R12Solar energy conversion • Utilisation of solar energy traditionally divided into two fields 1) Solar thermal 2) Photovoltaics RMIT University©2016 AMME 2 SSUNLIGHT Solar thermal PVT PV HEAT ELECTRICITYTraditional solar conversion technologies 1. The solar thermal collector > Solar energy to thermal energy RMIT University©2016 3 Fig 1. Diagram of a flat plate solar thermal collector AMMETraditional solar conversion technologies 2. The photovoltaic (PV) collector > Single step energy conversion process RMIT University©2016 AMME 4 Fig 2. An array of photovoltaic modulesPhotovoltaic Effect R) A schematic of solar cell structure L) How a solar cell converts sunlight into electricity RMIT University©2016 AMME 5PV – basic principles contd. • Photons absorbed in solar cells create electronhole pairs inside the solar cell. • Movement of electrons and holes generates electricity. • Free electrons migrate throughout the cell by metal contacts to the external load. For more information: www.pveducation.org Solar cells: Operating principles, technology, and system applications, Martin Green RMIT University©2016 AMME 6Photovoltaic systems: some facts • typically ONLY 10-20% of incidental solar energy is utilised for electricity conversion in crystalline silicon PV. • majority of remaining energy is absorbed by the cell and converted to heat. • Operating temperature has a strong effect on solar cell performance! RMIT University©2016 AMME 7 Fig 3. Conversion of sunlight into electricity and heat by a solar cell as a function of wavelength.Temperature effects on PV behaviour • The physics behind the effects of temperature on PV behaviour is beyond the scope of this lecture but in summary: 1) Marginal increase in short circuit current (Isc) with temperature. 1 𝐍 𝐵 𝐵 𝐵 𝐵 ≈ 0.033 % 𝐍 2) Linearly proportional drop in open circuit voltage (Voc) with temperature. 𝐵 𝐵 𝐵 = −2.45𝐠𝐠𝐍 RMIT University©2016 AMME 8Temperature effects on PV currentvoltage relationship RMIT University©2016 AMME 9 Fig 4. Effect of elevated temperature on the electrical performance of a PV cell. Picture is from: www.pveducation.orgTemperature effects on PV power-voltage relationship RMIT University©2016 AMME 10 Fig 5. Effect of temperature on solar cell power outputTemperature effect on Pmax RMIT University©2016 AMME 11 Fig 6. Effect of temperature on solar cell Pmax valueEffect of temperature on PV continued • Overall effect on PV conversion efficiency (Evans and Florschuetz, 1977). 𝐵𝐠= 𝐰 1 − 𝐵𝐵(𝐵 − 𝐵𝐵 η0 is conversion efficiency at ref. conditions, βref decay coefficient, Tref is temperature at reference conditions (typically 25oC), Tc is cell temperature, and ηpv is actual conversion efficiency. Approximate linear drop in efficiency with temperature. RMIT University©2016 AMME 12Published decay coefficient values RMIT University©2016 AMME 13Example: Temperature induced PV efficiency drop Q: A mono-crystalline (mc-Si) solar cell has a conversion efficiency of 15% at 25 degrees C. If the temperature coefficient for the mc-Si solar cell is taken as 0.004 (oC-1) what is the PV conversion efficiency when operating at 60oC? Solution η0 = 0.15, βref = 0.004, Tref = 25oC, Tc = 60oC, ηpv = ? ηpv = 0.15[1 - 0.004(60-25)] ηpv = 0.129….An approximate 15% drop! RMIT University©2016 AMME 14 …The negative impact of temperature can be significant!So how hot can a PV module become? • A number of methods exist to approximate PV temperature. • PV temperature can be simply estimated by, 𝐵𝐠= 𝐵𝐵 + 𝐵 • Where factor k is dependent on; 1) PV mounting style (open rack, building integrated etc.) 2) Wind speed 3) Module type RMIT University©2016 AMME 15Two common PV installation types RMIT University©2016 AMME 16 Fig 7. Rack mounted Fig 8. Building integrated Will these two systems operate at the same temperature?PV operating temperature continued RMIT University©2016 AMME 17 Fig 9. Differential temperature of PV modules i) roof integrated, and ii) rack mounted Why MPP temperature is lower than open circuit temperature?Motivation behind the PVT collector • Photovoltaic panels will inherently heat up while in operation > reduction in PV yield! • To mitigate the negative effect of temperature on PV output we can cool the PV panel > Improve PV yield! • By implementing active heat recovery we can make use of the captured heat > Sunlight = heat + electricity, Improve overall collector efficiency! RMIT University©2016 AMME 18PV/T collector definition • A single integrated solar collector which can simultaneously convert sunlight into electricity and heat. • Photovoltaic panel used for electricity generation and also as a thermal absorber. • Concept originally proposed by Wolf in the US in the mid 70’s and investigated further by MIT. RMIT University©2016 AMME 19Main features of the PVT Collector RMIT University©2016 20 Fig 10. Main features of the PVT Collector AMMEAdvantages • dual purpose: simultaneous conversion of heat and power. • (potentially) increase PV efficiency. • heat collected from cells can be actively recovered and purposefully used. • potential cost reduction. • increase energy yield per unit roofing area. • architectural uniformity (building integration). RMIT University©2016 AMME 21Typical applications • Domestic hot water (PVT/water) • Space heating (PVT/air) RMIT University©2016 AMME 22 Fig 11. Example of a PVT/water system Fig 12. Example of a PVT/air systemPV performance improvement • PV performance may be improved due to cooling. • Introduction of glass cover however will introduce reflection losses >> reduce PV yield • A glass cover + air gap >> increase PV temp RMIT University©2016 AMME 23 Fig 13. Energy flow diagram of a PV/T collectorThermal performance of a PV/T collector • Thermal performance of a PV/T collector is lower than a traditional solar thermal collector. • Lower absorption due to conversion of electrical energy by PV cell. 𝐵𝐵𝐠= 𝐵 − 𝐵𝐵 • High longer wavelength reflection losses from cell. AR coatings optimised for photocurrent generation. RMIT University©2016 AMME 24Energy yield per unit area of roofing Collector type Area (m2) Thermal (kWh) Electrical (kWh) Solar thermal 1 520 - Photovoltaic 1 - 72 PVT collector 2 700 132 RMIT University©2016 AMME 25 Fig 15. Traditional solar collectors tested alongside a PVT collector (Zondag et al., 2002)Aesthetics RMIT University©2016 AMME 26 Fig X. A mix of traditional solar collectors. Fig X. The integrated PV/T collector.Based on these advantages substantial research has been carried out globally in this area. RMIT University©2016 AMME 27PV/T Collector types RMIT University©2016 AMME 28 Fig 16. Classification of flat plate PV/T collectorsPVT/air collector • Principal purpose is space heating. • Advantages 1) No freezing or boiling issues 2) No damage if leakage occurs • Disadvantages 1) Low thermal capacity/conductivity > low heat transfer 2) Low density > high circulation volume flow rate needed 3) High losses possible due to leakage RMIT University©2016 AMME 29PVT/water collector • Principal applications: space heating / domestic hot water. • Advantages 1) Higher thermal capacity 2) Greater heat transfer from PV > Improved PV output 3) Lower volume flow rates needed. • Disadvantages 1) Leakage and frost issues. 2) Costs associated with fluid network material/installation. RMIT University©2016 AMME 30Companies manufacturing such systems • SOLIMPEKS • TES • SOLARUS • Cogenra • ... RMIT University©2016 AMME 31 Picture from: http://solimpeks.com.auTESZEUS® Photovoltaic-Thermal Hybrid Solar Collector RMIT University©2016 AMME 32 http://www.tessolarwater.com/index_en.html?zeuspv-t.html&2Cogenra Solar T14 RMIT University©2016 AMME 33 http://www.cogenra.com/product The T14 system includes integrated waste heat recovery for optional cogeneration. For large heat load applications, the T14 captures and delivers the by-product heat up to 248 °F/120 °C.. SYSTEM SPECIFICATIONS 13.1 – 17 KW Length: 145‘– 4” (44.4 m) Max Width: 10' (3.05 m) Ground coverage: 3-5 acres / MWSRB Energy RMIT University©2016 AMME 34Continue RMIT University©2016 AMME 35Part II: PV/T Modelling RMIT University©2016 AMME 36Introduction to photovoltaic modelling • Mathematical modelling allows us to simulate the behaviour of a PV cell under varying environmental conditions (irradiance, temperature, etc.) >> forecasting! • The current-voltage behaviour of a PV cell is non-linear. RMIT University©2016 AMME 37 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.0000 0.2000 0.4000 0.6000 Current (A) Voltage (V)The equivalent circuit model of a PV cell • A number of models currently exist ranging in accuracy and complexity. • The 5 parameter - single diode equivalent circuit model is commonly employed (see figure below). • Strikes a good balance between accuracy and complexity. RMIT University©2016 AMME 38 Fig 19. Equivalent circuit diagram of the single diode PV cell.Single diode characteristic equation • Applying the relevant Kirchhoff's laws to the equivalent circuit model we obtain a non-linear implicit characteristic equation. • Where I ph is the photogenerated current, V is voltage, Rs is the series resistance, Rsh is the shunt resistance, n is the diode factor, I0 is the reverse saturation current, and V th is the thermal voltage (kT/q). RMIT University©2016 AMME 39 0 1 s th V IR s nV ph sh V IR I I I e R                    The I-V Curve of a PV cell • Electrical data generally provided by the manufacturer at STC: [Isc, Voc, Imp, Vmp, Pmax]. RMIT University©2016 AMME 40 0.000 1.000 2.000 3.000 4.000 5.000 6.000 0.000 1.000 2.000 3.000 4.000 5.000 6.000 7.000 8.000 9.000 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 Power (W) Current (A) Voltage (V) Short circuit current (Isc) Open circuit voltage (Voc) Maximum power point (MPP) Max current (Imp) Max voltage (Vmp)The fill factor, FF • The fill factor term is commonly used in PV. • It is a measure of the junction quality of a cell. Mathematically defined as, • 𝐵 = 𝐵𝐵𝐵 𝐵𝐵𝐵 • As FF approaches 1, the higher the cell quality. RMIT University©2016 AMME 41 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 Current (A) Voltage (V)Modelling a PV device • To apply the above equation, and generate the IV curve we must know five parameters {Rs, Rsh, n, I0, Iph}. • Generally not provided by cell/module manufacturers. • Modelling accuracy strongly influenced by parameter values. • R s, Rsh, and n particularly influential on shape of I-V curve. RMIT University©2016 AMME 42 0 0 ( ) ( ) ( ) sh s ph s th th s sh sh ph s sh s s sh R R I R I V LambertW nV V nV R R R I I i R R R R R                Rs and R sh • The series and shunt resistances represent various ohmic losses within the PV cell. • The series resistance >> resistances introduced by cell solder bonds, cell-interconnect busbars, cell metallisation, and resistance within the emitter base regions. • R sh >> high-conductivity paths across the solar cell p-n junction created as a result of crystal damage and impurities in and near the junction • Ideally Rs is very low, and Rsh is very high. In reality they will vary significantly from manufacturer to manufacturer. RMIT University©2016 AMME 43Influence of R s on I-V curve RMIT University©2016 AMME 44 Fig 20. Effect of varying Rs on the IV curve of a PV cell (van Dyk, 2004)Influence of R sh on I-V curve RMIT University©2016 AMME 45 Fig 21. Effect of varying Rsh on the I-V curve of a PV cell.Effect of modelling parameters • Series and shunt resistances clearly influence the behaviour of the PV model. • Similarly the diode factor, n, will alter the shape of the I-V curve particularly around the “knee” of the IV curve. • To model the behaviour of a cell, we need to determine the parameters specific to the cell/device under study. • A number of methods currently exist to achieve this. RMIT University©2016 AMME 46Approximating solar cell modelling parameters • Solar cell modelling parameters may be determined either analytically or numerically. • Analytical methods are commonly used as they are fast and relatively simple to carry out. • By introducing several simplifications, the non-linear nature of governing equations are reduced to an analytically solvable form. • Numerical methods limit these simplifications and can therefore improve accuracy. But more difficult and time consuming to carry out. • Here we will discuss the analytical approach only. RMIT University©2016 AMME 47Analytical calculation of cell parameters • We can analytically approximate Iph, Rs, Rsh, n, and I0 using data provided by the manufacturer [Isc, Voc, Imp, V mp]. • Manufacturer data obtained at standard test conditions (STC) where T = 25oC, and G = 1000 W/m2. • A PV device will predominantly work outside this range! • We can determine parameter values at these reference conditions and then adjust to suit environmental conditions. RMIT University©2016 AMME 48Series Resistance, Rs • The value for R s will actually depend on the operating point along the IV curve. • As a PV system is directly coupled with MPPT hardware, we can assume that a PV system is operating at MPP. 1) Approximate Rs0 by (and assuming n = 1), 𝐵0 = 1 𝐍 𝐵 𝐍 𝐵 − 𝐠𝐵 − 𝐵𝔎𝐵𝐍 𝐍 𝐵 𝐵𝔎 2) The approximate Rs 𝐍 𝐠= 1 𝐍 𝐵 𝐍 𝐵 − 𝐠𝐵 − 𝐵𝔎 − 𝐵𝔎𝐵𝐍 𝐍 𝐵 + 𝐵𝔎 − 𝐵𝐵𝐰 𝐵𝔎 RMIT University©2016 AMME 49Shunt resistance, Rsh • Once R s has been found we can then calculate Rsh. 𝐵ℎ = (𝐠𝐵 − 𝐵𝔎 (𝐠𝐵 − 𝐵𝐵𝐍 𝐍 𝐵 − 𝐵𝐠𝐠𝐵 − 𝐵𝐵𝐠− 𝐵𝐵𝐵ℎ • Here we again assume the diode factor, n, is equal to one. RMIT University©2016 AMME 50Photogenerated current, Iph • Photogenerated current at STC calculated as a function of I sc, Rs, and Rsh. 𝐍 𝔎,𝐵𝐠≈ 𝐍 𝐵(𝐵 + 𝐵ℎ 𝐵ℎ • I ph will depend on incidental irradiance and temperature. 𝐍 𝔎 = 𝐍 𝐵𝐵 𝐍 𝔎,𝐵𝐠+ 𝐵𝐵(𝐵 − 𝐵,𝐵𝐍 Where Gref = 1000 W/m2, Tc,ref = 25 deg C, and µisc = short circuit current decay coefficient (from manufacturer). RMIT University©2016 AMME 51Reverse saturation current, I0 • Reverse saturation current at reference conditions 𝐰,𝐵𝐠= 𝐵𝐬𝐵𝐵 −𝐵𝐵,𝐵𝐍 𝐵𝐵,𝐵𝐍 • I 0 influenced by temperature. We adjust I 0,ref by 𝐰 = 𝐰,𝐵𝐍 𝐵 𝐵 ,𝐵𝐍 3 𝐍 𝐵𝐍 𝐵 1 𝐍 𝐬𝐵𝐍 − 1𝐵 RMIT University©2016 AMME 52 We now have all five modelling parameters!Using the model – effect of irradiance RMIT University©2016 AMME 53 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.00 0.10 0.20 0.30 0.40 0.50 0.60 Current (A) Voltage (V) Current G = 700 W/m^2 Current G = 800 W/m^2 Current G = 900 W/m^2 Current G = 1000 W/m^2 Power G = 700 W/m^2 Power G = 800 W/m^2 Power G = 1000 W/m^2 Power G = 900 W/m^2Using the model – effect of temperature RMIT University©2016 AMME 54 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.00 0.10 0.20 0.30 0.40 0.50 0.60 Power (W) Current (A) Voltage (V) T = 25 deg C (Current) T = 35 deg C (Current) T = 45 deg C (Current) T = 55 deg C (Current) T = 25 deg C (Power) T = 35 deg C (Power) T = 45 deg C (Power) T = 55 deg C (Power)PV/T Performance calculation • Thermal calculations >> Based on the modified Hottel-Whillier analysis • Electrical calculations >> non-linear current voltage analysis RMIT University©2016 AMME 55Thermal analysis of PV/T collector • Analysis similar to standard solar thermal absorber. • Modified to include PV panel effects. • The Hottel-Whillier-Bliss (HWB) one dimensional approach. • Useful energy gain by working fluid calculated by, Where A = collector area, FR is the collector heat removal factor, G = irradiance, Uloss = overall heat loss coefficient, TI = inlet fluid temp, Ta = ambient temp, and τα = transmissionabsorption coefficient. RMIT University©2016 AMME 56FR calculation • Eqn. to calculate FR, • Where m is the mass flow rate of fluid, Cp is the specific heat of fluid, and F’ is the heat removal efficiency factor • 𝐲 = 1 𝐵 𝐍 1 𝐵[ 𝐫(𝘒𝐠𝑝+𝔎𝐵𝐠1 +𝐵 1ℎ𝐍 RMIT University©2016 AMME 57 Accounts for heat resistance between cells and absorberFin efficiency, F calculation RMIT University©2016 AMME 58 • Fin efficiency primarily dependent on fin geometry Fig 17. Schematic of a fin and tube absorber typically utilised in a traditional solar thermal collectorFin efficiency calculation • But also material thermal conductivities. • Co-efficient M must take into account the hybrid nature of the PVT collector (i.e. the thermal conductivities of PV and absorber materials). • K abs = absorber thermal conductivity, Labs and LPV are the absorber and PV thicknesses respectively, and Uloss is the overall heat loss coefficient. RMIT University©2016 AMME 59Calculation of overall heat loss, UL • Similar to a traditional solar thermal collector, the PVT collector is subject to heat loss. • Overall heat loss, UL must be found by summing individual losses. UL = Utop + Ubtm + Uedge • Ubtm and Uedge loss may be calculated using standard procedure for FP solar thermal collector • Top heat loss, Utop, may be calculated for a glazed PVT collector by the eqn. RMIT University©2016 AMME 60Top loss coefficient, Utop • Standard Duffie and Beckman (2006) eqn. RMIT University©2016 AMME 61 N: number of glass covers 𝐺 Collector Tilt (deg) 𝐵 : emittance of glass (0.88) 𝐍 𝐺 emittance of absorber plate 𝐵 : ambient Temperature 𝐍 𝐵: mean plate temperature ℎ 𝐺 wind heat transfer coefficientUL Calculation • BUT a PVT collector may be unglazed! • In this case, the previous eqn. can not be used. • We must individually calculate contributions due to forced/natural convection and radiation. RMIT University©2016 AMME 62 Fig. 18 Example of an unglazed PVT collectorUnglazed PVT: Utop calculation • The radiation heat loss coefficient may be calculated by (Eicker, 2003), • Where ε p is plate emissivity, Tpm is mean plate temperature and Ts is the sky temperature. • Sky temperature is calculated by using the modified Swinbank equation (Fuentes, 1987). RMIT University©2016 AMME 63Unglazed PVT: Utop calculation contd. • Now to consider forced and natural convection loss components. • Forced convection may be calculated using the Watmuff et al. (1977) correlation. • Where v is wind velocity in ms-1. • Natural convection is calculated from mean plate and ambient temperature (Eicker, 2003). RMIT University©2016 AMME 64Unglazed PVT: Utop calculation contd. • Combining hw and hnat, we may calculate the overall convection heat transfer coefficient, hc. ℎ 𝐠= 3 ℎ𝐳 + ℎ𝐵𝐳 • By taking the summation of radiation and convection heat losses, we may now compute Utop. • Calculation of U top is somewhat different to a traditional flat plate solar collector. RMIT University©2016 AMME 65Part III: PVT collector design aspects RMIT University©2016 AMME 66PVT design concepts • Currently a number of PVT designs are being investigated. • In this section we will briefly discuss the major design variants. • Discuss thermal and PV aspects but also briefly consider manufacturability. RMIT University©2016 AMME 67 Fig 22. Examples of two PV/T designs. Absorber design is a significant design parameterThermal aspects: PVT absorber designs • Principal designs may be categorised by their absorber design. • Thermal aspects: improve i) absorption, ii) heat transfer from PV/absorber to HX fluid. RMIT University©2016 AMME 68 Fig 23. (clockwise) i) Sheet and tube absorber, ii) channel flow, iii) free flow, iv) two absorber designThe sheet and tube absorber • Well known technology (traditional flat plate collector). • Integrate a PV channel directly onto a fin and tube type absorber. • Thermal losses can be reduced by the addition of covers. Typically 0, 1, and 2 covers have been looked at. RMIT University©2016 AMME 69 Glazing 2 Glazing 1 Absorber Glazing 1 Absorber Fig 23. Glazing options for a PVT absorber.Sheet and tube absorber continued • Addition of covers increases reflection losses >> Reduce PV output. • Suppressing thermal losses increases absorber and PV temperature >> Reduce PV efficiency. • Compromise must be made between thermal and electrical output. RMIT University©2016 AMME 70 Fig 24. Components of the sheet and tube PVT collectorThermal performance: sheet and tube designs RMIT University©2016 AMME 71 Fig 25. Thermal efficiencies for a PVT collector with two glazing options (Zondag, 2008)PVT Thermal performance • The reduced thermal efficiency of a PVT collector compared to traditional flat plate collector explained by – 1) Reduced absorption of PV due to increased reflection loss at various layers. 2) PV is not spectrally selective and therefore suffers greater thermal radiation loss. 3) Greater heat resistance between PV and working fluid. 4) Lower thermal yield due to electrical conversion. • Reflection, spectral, and thermal resistance losses all contribute to reducing thermal output. RMIT University©2016 AMME 72Electrical performance: sheet and tube design RMIT University©2016 AMME 73 Fig 26. PV electrical efficiency of a PV panel and PVT collector with 0 and 1 cover (Zondag, 2008)PVT Reflection loss • The relatively low transmission-absorption factor of PVT collector is an important loss mechanism. • (τα)ST ≈ 0.95 • (τα)PVT ≈ 0.75 - 0.85 (depends on PV and absorber) RMIT University©2016 AMME 74 Fig 27. Effect of absorption on thermal efficiencyReflection loss continued • Five areas investigated to improve PVT absorption 1) Reduce reflection loss at top cover (glazing). 2) Reducing reflection at top surface of PVT absorber. 3) Reducing reflection at PV top grid. 4) Increasing PV absorption. 5) Increasing absorption of the opaque surface below PV. RMIT University©2016 AMME 75Reflection loss at top cover • Low iron glass typically used τ ≈ 0.91 – 0.92 • Recently reported glass with τ ≈ 0.96. Ideal for solar applications! • Plastic covers generally not used for PVT collectors due to lower optical performance, thermal expansion, and UV degradation. • Plastic covers may suffer damage under high operating temperatures (e.g. stagnation). • Early work at MIT found at high temperatures, plastic covers outgassed, reducing transmission by 10-15% RMIT University©2016 AMME 76Reflection loss at PV top contact • PV Cell typically consists of a contact grid at top surface. • High reflection losses as a result of this grid. • Represents a small % of total area >> secondary loss. • Tackled two ways; 1) Reduce reflection from grid. 2) Reduce contact grid area. • Reduce contact grid most promising. RMIT University©2016 AMME 77Thermal resistance • Thermal resistance between the PV cells and HX fluid should be minimised. RMIT University©2016 AMME 78 Fig. 28 Assembly explosion of A proposed PVT collectorThermal resistances in PVT absorber • Low heat transfer from PV cell to working fluid will result in high temperature gradients and high PV temp. • PVT absorber assembly should consist of: 1) thin layers. 2) layers should have a high thermal conductivity. 3) provide electrical insulation! A little challenging to achieve all… • Early PVT designs reported a ΔT (i.e. Tpv – Tfluid) of over 30 degrees >> Very poor heat transfer! RMIT University©2016 AMME 79Effect of thermal resistance on FR RMIT University©2016 AMME 80 Fig 29. Effect of thermal resistance on heat removal factor (FR) on a glazed and Unglazed PVT collector (Zondag, 2008)Absorber to hx fluid heat transfer • Conventionally sheet and tube design absorber used. • Easy to manufacture and high heat removal factor. • Fin efficiency improved by reducing W/D (i.e. tube spacing and diameter). RMIT University©2016 AMME 81 Fig 30. Conventional fin and tube design.Absorber design continued • Channel design investigated. • Thin channels can potentially increase heat transfer over fin and tube variant. • Issues raised: pressure loss, difficult to manifold, and flow distribution issues. • One study revealed a yearly yield improvement of only 2% over fin and tube design (de Vries, 1998) RMIT University©2016 AMME 82 Fig 31. Box channel type absorber designCollector heat loss • to maximise thermal yield we must minimise heat loss to ambient. • heat may be lost from top, sides, and bottom of collector. • loss from top of absorber is largest contributor. • heat loss via convection and radiation heat transfer modes. • glazing principal method to supress top heat loss. RMIT University©2016 AMME 83Electrical efficiency • PV performance of a PVT collector is typically lower than traditional PV collector due to – 1) reflection losses introduced by glazing (~8% loss) 2) increased operating temperature (if glazed) • Electrical output dependent on PV type (a-Si, c-Si, pc-Si, etc.) • PV type will also influence thermal output! RMIT University©2016 AMME 84Type of PV • Crystalline and amorphous silicon type PV only used up until now. • Crystalline silicon most popular due to its higher conversion efficiency. • Some studies however have revealed a greater combined efficiency achieved with a-Si cells. RMIT University©2016 AMME 85PV type and thermal efficiency • Study by Tripanagnostopoulos et al. (2002) RMIT University©2016 AMME 86 Fig 32. Thermal efficiency curves for air and water PVT collectors with a-Si and pc-Si PV cells.a-Si photovoltaic cells • Low cost solar cells. • Low temperature decay coefficient ~ -0.1%/K compared to crystalline silicon cells ~ -0.4%/K. • May be directly deposited onto various substrates (eg. glass). • Improved thermal efficiencies obtained in several studies using a-Si solar cells. • Low electrical efficiency ~6% RMIT University©2016 AMME 87PV type and electrical efficiency RMIT University©2016 AMME 88 Fig 33. Electrical efficiency curves for air and water PVT collectors with a-Si and pc-Si PV cells.Shading • Shading from the collector tray is a small issue for a traditional solar thermal collector. • For a PVT collector this is not the case as PV cells are electrically connected. • One shaded cell can reduce entire output of series connected string! • Air gap between PV cells and glazing therefore reduced. • Will increase thermal loss to ambient. • Not an issue for unglazed PVT collector RMIT University©2016 AMME 89PV operating temperature • Improve heat transfer from PV to fluid! • Reduce thermal resistance between PV and fin. RMIT University©2016 AMME 90 Fig 34. Effect of heat transfer on PV temperature.Temperature distribution • Typical PV panel will operate at uniform temperature. • Temperature gradients will inherently exist in a PVT collector. • Cells at higher temperature will operate under less voltage. • An issue for parallel connected strings. RMIT University©2016 AMME 91Temperature distribution continued • Consider a PVT collector cooled by a header/riser fluid network. • How will the temperature vary over the absorber surface? i.e. what is T(x,y)? RMIT University©2016 AMME 92 Fig. X Example of an unglazed PVT Collector cooled by a header/riser Fluid network.Flow distribution of heat exchange fluid • Typical solar thermal collector analysis assumes uniform flow of fluid through absorber. • Somewhat complex to model as fluid flow, heat transfer, and the electrical output from PV (non-linear!) must all be considered. • Flow distribution influenced primarily by fluid network geometry (size of array, manifold/riser diameters) and operational parameters (flow direction in manifolds, mass flow rate, etc.) RMIT University©2016 AMME 93Heterogeneous array temperature • Linear temperature gradient in direction of fluid flow. • For series connected cells operating under the same voltage this is ok. • Temperature gradients among parallel strings will drive down entire array output! RMIT University©2016 AMME 94Collector reliability issues • As PVT collectors are relatively new, reliability issues have not been extensively looked at. • Major areas of concern: 1) Stagnation 2) Thermal shock 3) Electrical insulation RMIT University©2016 AMME 95Part IV: Beam Splitting RMIT University©2016 AMME 96High Temperature Hybrid Receivers • The temperature of thermal output is over 100°C. This means that in traditional hybrid collectors, the temperature of the solar cell will be even higher than this. • The efficiency of a cell as a function of temperature in general is as below: 𝐨𝐠= 𝐠(𝐽25℃ 1 − 𝐵 So one solution is thermally decoupling the cell from the thermal collector. RMIT University©2016 AMME 97Solar Spectrum; thermal radiation from the surface of the sun RMIT University©2016 AMME 98 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 250 500 750 1000 1250 1500 1750 2000 2250 2500 2750 3000 3250 3500 3750 4000 Spectral Irradiance W m-2 nm -1 Wavelength nm ASTM G173-03 Reference Spectra Extraterrestrial W/m^2/nm Global W/m^2/nm Near Visibl Mid UV The area under the blue curve is the amount of energy received from the sun at earth surface which almost equal to 1000 𝐍 𝐲 for a surface normal to the sunlight at solar noon. Photothermal receivers can capture the whole spectrum but photovoltaic receivers can’t!!!Photons as Energy Carriers; • In visible range not only the number of emitted photons (photon flux) is higher but also the energy of them is higher. • Energy of a photon (𝐩 is a function of its frequency (𝐩 or wavelength (𝐩 as below: 𝐠= ℎ𝐠= ℎ𝐍 𝐍 ℎ is Planck’s constant and is equal to: 6.626 × 10−34 𝐮 𝐍 RMIT University©2016 AMME 99Band Gap in Solar Cells • In order to generate current in a cell (made of a semiconductor), electrons must absorb energy (from a photon) and jump from conduction band to valance band. • The minimum energy required for this process is equal to the energy difference between these two bands. This energy is called “band gap” energy. e.g. for silicon cell it is almost 1.1 eV. RMIT University©2016 AMME 100Photon Energy Example • Knowing that the band gap of a typical silicon solar cell is 1.1(e.V), calculate the longest wavelength that a photon can have and still be able to generate electrical current in the cell. • Solution: The minimum energy required to excite an electron across the band gap in the cell is equal to the band gap energy; 1.1 𝐵 = 1.1 × 1.6022 × 10−19 𝐠= ℎ𝐍 𝐠⇒ 𝐍 = 1,130𝐵 RMIT University©2016 AMME 101External Quantum Efficiency • Each photon with energy higher than the band gap excites just one electron and generates just one electron-hole pair no matter how much extra energy it has. • External quantum efficiency is the ratio of the number of collected current carriers (electronhole pairs) to the number of radiation energy carriers (photons); RMIT University©2016 AMME 102External Quantum Efficiency • External quantum efficiency is zero for photons with energies lower than the band gap; for the rest it is ideally one, but not practically due to recombination losses. RMIT University©2016 AMME 103 The figure is from: http://pveducation.org/pvcdrom/solar-celloperation/quantum-efficiencySilicon Cell’s Spectral Response • Spectral Response is the ratio of the short circuit current (𝐵𝐩 generated by the cell to the radiation power incident on the cell (𝐵𝐩.Its dimension is A/W: 𝐵 = 𝐍 𝐵 𝐵𝐍 • SR includes the energy of photon in calculating the efficiency of the cell; hence SR curve is slightly different from the EQE curve. RMIT University©2016 AMME 104Spectral Response • Photons with shorter wavelengths have higher energy; such photons generate electron-hole pairs but the excess energy thermalise the cell (below: SR of Si cells). RMIT University©2016 AMME 105 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 200 400 600 800 1000 1200 1400 Spectral Response (A/W) Wavelength (nm)Cell Efficiency • Cell efficiency is the ratio of the electric power generated by the cell to the incident radiation on the cell: 𝐵𝐵𝐠= 𝐵 𝐵 𝐵𝐍 𝐵 = 𝐵𝐵 𝐵𝐵𝐵 & 𝐵 = 𝐵𝐍 𝐵𝐍 Rearranging the above equations: 𝐵𝐵𝐠= 𝐠𝐵 × 𝐵 × 𝐵 RMIT University©2016 AMME 106Using Spectral Splitting for Hybrid Collectors • A spectral splitter is used to divide the spectrum into several bands and direct the most suitable band to the PV receiver. In concentrated PV this helps to avoid excess heating of the cells • The cut-off wavelengths of the spectral splitter depends on the PV type. RMIT University©2016 AMME 107How to Split the Spectrum Two main methods: • Using dichroic mirror; such mirrors can be designed and fabricated to reflect or transmit just a certain range of light. The cut-off wavelengths can be tuned in the design stage. • Using selective absorbers to absorb a certain range of light and transmit the suitable range to PV receivers. RMIT University©2016 AMME 108Spectral Splitters… • (a) shows a dichroic mirror and (b) shows a selective absorber RMIT University©2016 AMME 109Concluding remarks • Solar energy may be simultaneously converted to heat and electricity using a PVT hybrid collector. • As PV collectors are negatively influenced by temperature, PVT collectors offer the possibility of improving electrical conversion by cooling. • Principal advantages include: 1) Improved solar conversion efficiency by including active heat recovery. 2) Greater energy output per unit area of roofing. 3) Improved aesthetics. 4) Potential cost reduction. RMIT University©2016 AMME 110Concluding remarks continued • Currently however market penetration of PVT collectors is very small. A number of technical issues must be overcome such as; 1) Maximise heat transfer from PV to HX fluid. 2) Improve PV absorption characteristics to better suit heat collection. 3) Reliability concerns regarding stagnation need to be addressed. 4) Currently no specific standards exist for the PVT collector. RMIT University©2016 AMME 111Concluding remarks continued • Beam splitting is a promising approach to better utilise sunlight considering the operational characteristics of PV and solar thermal collectors. • Matching incidental light with the spectral response of a PV cell, substantial improvement in efficiency can be obtained. • Principal challenge is cost associated with the required optics. RMIT University©2016 AMME 112Thank you [email protected] [email protected] RMIT University©2016 AMME 113