Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Reduction of mismatch and shading loss by use


Published on

  • Be the first to comment

Reduction of mismatch and shading loss by use

  1. 1. INTERNATIONAL Electrical EngineeringELECTRICAL ENGINEERING International Journal of JOURNAL OF and Technology (IJEET), ISSN 0976 – 6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 3, October – December (2012), © IAEME & TECHNOLOGY (IJEET)ISSN 0976 – 6545(Print)ISSN 0976 – 6553(Online) IJEETVolume 3, Issue 3, October - December (2012), pp. 137-145© IAEME: ©IAEMEJournal Impact Factor (2012): 3.2031 (Calculated by GISI) REDUCTION OF MISMATCH AND SHADING LOSS BY USE OF DISTRIBUTED POWER ELECTRONICS IN GRID CONNECTED PHOTOVOLTAIC SYSTEMS A K PRADHAN1, S K KAR2, M K MOHANTY3 1 Gandhi Institute Technology And Management(GITAM),Bhubaneswar,Odisha 2 .S.O.A. UNIVERSITY, Bhubaneswar, Odisha 3 Orissa University of Agriculture and Technology, Bhubaneswar, Odisha ABSTRACT The energy efficiency can be improved by power electronics, which can reduce mismatch and shading losses in photovoltaic (PV) systems by using micro-inverters and DC-DC converters. Under partially shaded conditions, the use of the power electronics can recover between 10%– 30% of annual performance loss or more, depending on the system configuration and type of device used. Additional features may also increase the benefits like increased safety, reduced system design constraints and added monitoring and diagnostics by using per-panel power electronics, The economics of these devices will also become more favourable as production volume increases and integration within the solar panel’s junction box reduces part count and installation time. Some potential liabilities of devices include increased PV system cost, additional points of failure, and an insertion loss that may or may not offset performance gains under particular mismatch conditions. Keywords: Mismatch, shading, micro-inverters, converters. 1. INTRODUCTION Recent improvements in the field of power electronics as efficiency, reliability, and cost of the products have made them usable in many applications, from small residential installations to large commercial PV arrays. A number of new products have come to the market in the field of power electronics, which includes DC-DC converters and micro-inverters that are designed to either replace or work with traditional central PV inverters[1]. In general, the use of advanced power electronics at per unit-module or per unit-PV string basis can reduce the impact of module mismatch and partial shading. A traditional central inverter is having only a few input 137
  2. 2. International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 3, October – December (2012), © IAEMEchannels that independently track the maximum power point (MPP) of the PV system. Withlarge utility-scale inverters reaching up to half a megawatt (MW) in size, over 5,000 individualPV panels could potentially operate at one common maximum power point[2]. A reduction inthe output power of one or more of these PV panels can lead to mismatch in the maximumpower point between the various PV modules and strings. Possible causes of MPP mismatchinclude partial shading, soiling from dust, debris, and bird droppings, and module physicaldegradation. The impact of partial shading and mismatch can be reduced by increasing thenumber of independent MPP tracking channels in the PV system[3]. The improved MPPtracking method depends on the amount of mismatch throughout the system, the size andconfiguration of the system, and the characteristics of its PV modules, among other factors[4].A DC-DC converter sometimes called power boosters can provide improvement in systemperformance by working in conjunction with a central inverter. A DC-DC converter will trackthe maximum power point of solar module(s) connected to it and either increase (boost) ordecrease (buck) the output voltage to match the optimum voltage required by the centralinverter. Many currently available solar DC-DC converters use a separate enclosure for thepower electronics at each panel, typically attached to the PV module frame or rack. This savessome cost and labour over separate panel and power electronics. Another type of powerelectronics is the micro-inverter[5][6]. The current generation of micro-inverter productsappears to be achieving greater market penetration through improved efficiency, reduced cost,increased reliability, and diagnostic capabilities. The micro-inverters are installed on each PVmodule, replacing the use of a central inverter. Each PV panel’s DC power is converteddirectly to AC 240 V and grid-connected[7][8]. The output of each PV panel is thereforeeffectively in parallel, which eliminates power losses due to module mismatch. Thus theperformance improvements that arise from independently maximum power tracking PVmodules can be achieved with micro- inverters as well as with DC-DC converters. Anadditional benefit to micro-inverters compared with DC-DC devices is the reduction in DCbalance of system components, including the central inverter. Also, voltages tend to be lowerwith micro-inverter systems, which could be a safe benefit for rooftop systems [6]. Figure 1shows some example topologies for per unit-panel microelectronics, including DC-DCconverters and micro-inverters. 138
  3. 3. International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 3, October – December (2012), © IAEMEFig.1 Schematic diagram of conventional single-string PV system (top), DC-DC converter-equippedPV system(middle), and AC micro-inverter-equipped PV system (bottom).One common aspect of distributed power electronics is the effect of their long-term reliabilityon the complete PV system. In general, the probability of system failure increases with eachcomponent-level reliability, failure modes, and effect of failure on system availability will beimportant in assessing the overall value of distributed power electronics.2. PARTIAL SHADING AND MISMATCH LOSSESAccording to literature survey, the impact of shading and mismatch on PV systems, both withand without the use of DC-DC converters or micro-inverters been analysed [1-6]. Due to thedifferent possible string configurations and module characteristics in PV systems, it is difficultto generalize how mismatch will affect a given system. However, in most PV systems withconventional silicon panels, the presence of shade or mismatch will have a greater portion ofimpact on the system’s output. This is due to the serial nature of PV modules in strings, inwhich current reduction in one series-connected module causes mismatch losses in the rest ofthe string. Because of this potential for huge power losses in mismatched systems, solarmodule manufacturers typically include one or more bypass diodes in their modules, usuallylocated in the module’s junction box. The function of the bypass diode is to allow current to flowpast impaired sections of a module that are unable to produce as much current as the rest of thesystem[10][11]. To accomplish this, the module section is shorted out by the diode, producing nopower of its own. Since the bypass diode shorts out the partially shaded section, causing itsoperating voltage to fall to zero, the overall operating voltage of the PV string will be reducedaccordingly. 139
  4. 4. International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 3, October – December (2012), © IAEME.Fig. 2 Partial shading with bypass diode Fig. 3 Effect of partial shading in I-V curvesIf the PV system is composed of a single string, there is no additional impact due to the shadedstring’s reduced voltage. However, if multiple parallel strings are present in the system, anadditional source of mismatch loss occurs: voltage mismatch between parallel strings. In thissituation, the voltage needs to be equal between parallel strings of PV modules. This results inthe MPP tracking inverter to force unshaded modules in the affected string to operate at ahigher than optimum voltage to make up for the voltage drop from bypasses diodes elsewherein the string. This mismatch loss causes power losses in both shaded and unshaded modules.The impact of this voltage mismatch can range from an additional 60% power loss forunbalanced shade on two-string systems [5] to an additional 400% power loss for shadecovering 15%–20% of a utility-sized PV string [7]. It is clear that partial shading and othermismatch sources can result in performance losses much greater than the apparent scale of theshade itself. All analysis are based on the fig.2&3. The average performance showed thereduction of 22% due to shading from neighbouring trees and other elements. Of this lostpower, a majority (70%) was due to reduced irradiance and direct losses from shading. Only30% of the power loss was due to mismatch of current and voltage [12]. Therefore, theinstallation of DC-DC converters or micro-inverters would improve this particular system’sannual production by roughly 7% through the elimination of mismatch losses. The use ofdistributed MPP tracking equipment can also improve system performance if the PV panelshave mismatch in their operating conditions, particularly maximum-power current (Impp).Although values are the same for identical module model numbers, there can be some variationfrom panel to panel, as manufacturers typically bin the modules in 5 W to 10 W bins. Therecan therefore be a 5% difference or greater between the power output of modules with identicalmodel numbers. This can contribute to mismatch losses between module in the same string ifthe mismatch is between the Impp of series-connected modules. Judging from datasheets ofsilicon PV panels in the 200W to 240W range, the variation in Impp within a single bin istypically 2%–3%. However, this does not directly indicate that the power loss within a seriesstring is also 2%–3%, as seen in Figure 5. Because of the flatness of the I-V curve of a PVpanel in the neighbourhood of the maximum power point, a 2.5% change in operating currentnear the MPP only leads to a 0.5%–0.7% reduction in power for an average (0.72–0.74 fill 140
  5. 5. International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 3, October – December (2012), © IAEMEfactor) module. It is this 0.5%–0.7% series current mismatch loss that could be recoveredthrough the installation of per-module MPP tracking equipment. Of course, not every panel ina string can have below-average Impp, so the real mismatch loss is likely to be lower than thepreviously stated limit, and efficiency losses in DC-DC devices may further reduce this benefit. Fig. 4 power-current curve for a typical Si PV panel with fill factor = 0.71. [2]3. DC-DC CONVERTER AND ITS TOPOLOGIESSeveral different DC-DC converter device topologies are available for use with individual solarpanels, each with different strengths and operating uses. The simplest DC-DC converter uses asingle converter stage to either buck (reduce) or boost (increase) the output voltage of a PVpanel. In either case, the PV panel output voltage is MPP tracked by the control algorithm inthe device. A slightly more advanced DC-DC converter is the buck-boost converter, whichuses both buck and boost stages to allow the converter to either increase or decrease the outputvoltage of a PV panel. Fig 5: Cascading two switch Buck- Boost converterThe advantages of a buck-boost converter include an increased operating range and the abilityto correct for a greater amount of system mismatch. Since the device includes two conversion 141
  6. 6. International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 3, October – December (2012), © IAEMEstages rather than one, the increased flexibility may come at the cost of a slight efficiencyreduction as well as possible size and cost increases relative to single-stage devices[12]. In abuck-only DC-DC converter, the output voltage from a shaded panel is decreased, and theoutput current is increased to match the operating current of the unshaded modules in serieswith it. This type of converter works best in PV systems with limited mismatch, e.g., whereshade or mismatch occurs only on a few PV panels [5]. A boost-only DC-DC converteroperates by taking the input PV voltage (typically at the maximum power voltage of theparticular panel) and increasing it. This type of system is typically designed with every PVpanel in the system equipped with a boost converter. In some systems, the converter boosts thevoltage to a high constant value (~300 Vdc –550 Vdc), and all of the panels are placed inparallel. System mismatch is eliminated here because each panel contributes currentproportional to the amount of irradiance it receives. This system will work even with panelsfacing different directions, or at different tilt angles, because all of the converters are placed inparallel. The high constant-output voltage from the boost converter is chosen to maximize theefficiency of a fixed-input voltage inverter connected to the output of the converters. A systemusing buck-boost converters enjoy most of the benefits of both buck and boost converters. Forinstance, if shade is limited to only a few modules in the system, a buck-boost converter can beselectively installed on only the affected modules, and it will operate in buck mode to reducethe current mismatch between shaded and unshaded modules. Also, if the PV system includesmodules of different size, power rating, or orientation, a buck-boost converter can be placed onevery module in the series string, allowing for differences in the various module poweroutputs. If parallel strings are of different lengths, buck-boost converters on the shorter stringwill increase the operating voltage of the string to match the other longer strings. Buck-boostconverters can also be used with specialized fixed-input-voltage inverters that operate at aconstant high input voltage[15][16].4. ADVANTAGES OF DISTRIBUTED POWER ELECTRONICSAs stated above, there can be performance benefits to use perunit-panel distributed DC-DC andmicro-inverter products based on the reduction in panel current and voltage mismatch.Additional benefits include greater flexibility in system design and reduced time to place PVpanel in complicated rooftop designs. This can lead to lower levelized cost of energy (LCOE)and possible reduction in balance of systems (BOS) wiring cost in the case of boost convertersthat operate at higher voltage and lower current, or micro-inverters that do not require stringcombiner boxes. This is useful to the PV system installer as a remote diagnostic and warrantyrepair indicator, thereby helping to maximize system uptime. The owner also has morefeedback from the PV system to understand what conditions influence PV performance,possibly leading to better system maintenance and cleaning. In the case of micro-inverters, onespecific value-added benefit is the elimination of the single- point failure of a central inverter.A single micro-inverter failure does not cause the entire system to fail, power reduction duringa single failure is limited to the power of a single module. The overall lifetime of currentmicro-inverter products is also difficult to compare with central inverters, as these microinverters have not been available long enough to obtain field lifetime results[8]. In the case ofAC micro-inverters, these issues are mitigated because the system is de-energized when the 142
  7. 7. International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 3, October – December (2012), © IAEMEAC disconnect is thrown. The available DC voltage is limited to that of a single module, whichis considered benign. This arc-fault safety concern may also be mitigated by the use of per-module DC-DC converters, although further system tests would be required to verify thissafety benefit.5. PERFORMANCE ANALYSIS OF VARIOUS DC-DC TOPOLOGIESThe system is modelled either as a single-string installation or two parallel strings. The shadingon the system is somewhat more extensive than average, with an annual irradiance reduction of20% as measured by a detailed site survey. A reduced shading condition is considered as well,in which the annual irradiance loss due to shade is only 10%, concentrated entirely on one ofthe two strings. In addition to the residential rooftop shading simulations, larger systems withinter-row shading were also considered. For these systems, it was assumed that rows arespaced such that 3% of the annual irradiance is lost due to inter-row shading. A groundcoverage ratio of 0.54 and a module tilt of 18.5o are assumed. Modules are oriented inlandscape, with two modules stacked vertically and eight modules horizontally per row. Topredict the performance gains from the use of distributed electronics, substring-level I-Vcurves were calculated and summed based on the predicted irradiance and shade on a givenmodule substring. The performance of the DC-DC converter was modelled by a constantpower curve, as discussed in [9]. DC-DC converter efficiency was set equal to 0.99 for alldevices. The DC-DC efficiency loss will partly be offset by mismatch from soiling, aging, andmanufacturer distribution of PV panels, which are all mismatch terms neglected in thissimulation. Annual performance data is produced using the PV Watts engine, modified toallow for reduced irradiance due to partial shade[10]. For these particular installations, thepresence of shade led to performance losses of 3%–18%. The addition of DC-DC converterswith MPP tracking led to a recovery of 10%–20% of the annual loss due to partial shade, withmore gain coming from systems with more parallel strings and greater amount of shade. Theabove simulation results and additional results from literature are summarized in Table 1below:Table 1: Sources of mismatch loss cited in the literatureType of mismatch System loss Potential DC-DC gain Ref.Residential roof shade, 1string 5 – 15% 15 – 20% of loss [5]Residential rooftop tree shading – 5 – 20 % 20 – 30% of loss [5]multiple stringsResidential rooftop pole shading – 4 – 8 % 40 – 70 % of loss [6]multiple stringsCommercial system with inter – row 1- 5 % 0% of loss [4]ShadingResidential orientation mismatch 1–2% Appx. 100% of loss [11]within a stringImp distribution mismatch 0.2 - 1 % Appx. 100% of loss Fig, 4Soiling – CA and Southwest US 1.5 – 6.2 % + 15 – 40 % of loss [12] 143
  8. 8. International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 3, October – December (2012), © IAEME6. ECONOMIC ANALYSISThe relative benefit of distributed power electronics depends on the system configuration, theamount of current and voltage mismatch within the system, and the cost of the powerelectronics. A greater performance improvement could support a greater equipment cost. Themodel of pre-integrated power electronics, which reduce the components cost enough tobecome cost effective, works best with new rooftop installations that include the powerelectronics on every panel. The compatible technologies include micro-inverters, high-voltageboost converters, and buck-boost DC-DC converters. Buck converters may work the best as astandalone retrofit device, similar to the first-generation DC-DC converters available now.However, one advantage of selective installation on only a few modules is that performanceimprovement can be achieved with a much lower cost than installing the converter devices oneach module in a string. A separate AC branch circuit could be supplied to micro-inverters toincrease the capacity of an existing PV array. In the case of isolated shading from nearbyobstructions, individual modules could be fitted with retrofit buck- boost DC-DC converters. Incertain applications, this technology may be cost effective, particularly if the benefit of perunit-string monitoring is included, with its improvement in system up-time and reducedoperation and maintenance costs.7. CONCLUSIONDistributed power electronics have the potential to reduce PV performance loss due to partialshading and mismatch. Depending on the mismatch condition and system size, a variety ofproducts are available to improve the system performance. The benefits of unit module powerelectronics are greatest for multi-string residential installations with close-in shade obstructionsor mismatch from orientation or panel size. Value-added features of some devices includeperformance monitoring and emergency power-off, which may assist in market penetration,along with reduced cost through integration within PV panels’ junction boxes. Larger PVinstallations may prefer to install string-level DC-DC equipment to achieve some of themismatch reduction benefits of distributed power electronics without the part count and cost ofper-panel electronics.8. REFERENCES[1] P. Tsao, “Simulation of PV Systems with Power Optimizers and Distributed PowerElectronics,” IEEE Photovoltaic Specialists Conference, Honolulu, HI, 2010.[2] C. Deline , B. Marion (NREL),J. Granata , S. Gonzalez SNL “ A Performance and EconomicAnalysis of Distributed Power Electronics in Photovoltaic Systems”, Technical Report, NREL/TP-5200-50003 January 2011[3] A. Woyte, J. Nijs, R. Belmans, “Partial Shadowing of PV Arrays with Different SystemConfigurations: Literature Review and Field Test Results,” Solar Energy 74, pp. 217-233,2003. 144
  9. 9. International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 3, October – December (2012), © IAEME[4] N. Chaintreuil, F. Barruel, X. Le Pivert, H. Buttin, J. Merten, “Effects of Shadow on a GridConnected PV System,” 23rd European Photovoltaic Solar Energy Conference and Exhibition,p. 3417, 2008.[5] C. Deline, “Partially Shaded Operation of a Grid-Tied Photovoltaic System,” IEEEPhotovoltaic Specialists Conference, Philadelphia, PA, 2009.[6] C. Deline, “Partially Shaded Operation of Multi-String Photovoltaic Systems,” IEEEPhotovoltaic Specialists Conference, Honolulu, HI, 2010.[7] C. Deline, “Characterizing Shading Losses on Partially Shaded PV Systems,” PVPerformance Modeling Workshop, Albuquerque, NM. NREL Publication PR-5200-49504.[8] N. L. Curral, A. S. Araújo, N. F. Costa,”Microinverters for Photovoltaic Modules” MIEEC-2012.[9] B. Burger, B. Goeldi, S. Rogalla, H. Schmidt, “Module Integrated Electronics – AnOverview,” 25th European Photovoltaic Solar Energy Conference and Exhibition, Valencia,Spain, 2010.[10] B. Marion, M. Anderberg, “PVWatts – An Online Performance Calculator for Grid-Connected PV Systems,” ASES Solar 2000 Conference, June 16-21, 2000, Madison, WI.[11] S. MacAlpine, M. Brandemuehl, L.Linares, R. Erickson, “Effect of Distributed PowerConversion on the Annual Performance of Building-Integrated PV Arrays with Complex RoofGeometries,” ASME 3rd International Conference on Energy Sustainability, San Francisco,CA, 2009.[12] N.D. Kaushikaa, Anil K. Rai, “An investigation of mismatch loss in solar photovoltaic cellnetworks”, Energy, 32, pp755-759, 2007[13] Associacion de la Industria Fotovoltaica (ASIF), Informe annual 2008, available at[14] A. Abete et al., Analysis of photovoltaic modules with protection diodes in presence ofmismatching, Photovoltaic Specialists Conference, 1990., Conference Record of the TwentyFirst IEEE Volume , Issue , 21-25 May 1990 Page(s):1005 - 1010 vol.2[15] N.D. Kaushika and N.K. Gautam, Energy Yield Simulations of Interconnected Solar PVArrays, IEEE Transactions on Energy Conversion, Vol. 18, No. 1, March 2003[16] N.K. Gautam and N.D. Kaushika, An efficient algorithm to simulate the electricalperformance of solar photovoltaic arrays, Energy 27 (2002) 347-361 145