《IEA PVPS：2025漂浮式光伏電站：發電量、可靠性與維護綜述報告（英文版）（67頁）.pdf》由會員分享，可在線閱讀，更多相關《IEA PVPS：2025漂浮式光伏電站：發電量、可靠性與維護綜述報告（英文版）（67頁）.pdf（67頁珍藏版）》請在三個皮匠報告上搜索。

1、 carlos BayWa r.e.Task 13 Reliability and Performance of Photovoltaic Systems Floating Photovoltaic Power Plants:A Review of Energy Yield,Reliability,and Maintenance 2025 PVPS Report IEA-PVPS T13-31:2025 Task 13 Performance,Operation and Reliability of Photovoltaic Systems Floating PV Power Plants:A

2、 Review of Energy Yield,Reliability,and Maintenance What is IEA PVPS TCP?The International Energy Agency(IEA),founded in 1974,is an autonomous body within the framework of the Organization for Economic Cooperation and Development(OECD).The Technology Collaboration Programme(TCP)was created with a be

3、lief that the future of energy security and sustainability starts with global collaboration.The programmes are made up of 6,000 experts across government,academia,and industry dedicated to advancing common research and the application of specific energy technologies.The IEA Photovoltaic Power System

4、s Programme(IEA PVPS)is one of the TCPs within the IEA and was established in 1993.The mission of the programme is to“enhance the international collaborative efforts which facilitate the role of photovoltaic solar energy as a cornerstone in the transition to sustainable energy systems.”To achieve th

5、is,the Programmes participants have undertaken a variety of joint research projects in PV power systems applications.The overall programme is headed by an Executive Committee,comprised of one delegate from each country or organisation member,which designates distinct Tasks,that may be research proje

6、cts or activity areas.The IEA PVPS participating countries are Australia,Austria,Belgium,Canada,China,Denmark,Finland,France,Germany,India,Israel,Italy,Japan,Korea,Malaysia,Morocco,the Netherlands,Norway,Portugal,South Africa,Spain,Sweden,Switzerland,Thailand,Turkey,the United Kingdom,and the United

7、 States of America.The European Commission,Solar Power Europe and the Solar Energy Research Institute of Singapore are also members.Visit us at:www.iea-pvps.org What is IEA PVPS Task 13?Within the framework of IEA PVPS,Task 13 aims to provide support to market actors working to improve the operation

8、,the reliability and the quality of PV components and systems.Operational data from PV systems in different climate zones compiled within the project will help provide the basis for estimates of the current situation regarding PV reliability and performance.The general setting of Task 13 provides a

9、common platform to summarize and report on technical aspects affecting the quality,performance,reliability and lifetime of PV systems in a wide variety of environments and applications.By working together across national boundaries,we can all take advantage of research and experience from each membe

10、r country and combine and integrate this knowledge into valuable summaries of best practices and methods for ensuring PV systems perform at their optimum and continue to provide competitive return on investment.Task 13 has so far managed to create the right framework for the calculations of various

11、parameters that can give an indication of the quality of PV components and systems.The framework is now there and can be used by the industry who has expressed appreciation towards the results included in the high-quality reports.The IEA PVPS countries participating in Task 13 are Australia,Austria,

12、Belgium,Canada,Chile,China,Denmark,Finland,France,Germany,Israel,Italy,Japan,the Netherlands,Norway,Spain,Sweden,Switzerland,Thailand,and the United States of America.DISCLAIMER The IEA PVPS TCP is organised under the auspices of the International Energy Agency(IEA)but is functionally and legally au

13、tonomous.Views,findings and publications of the IEA PVPS TCP do not necessarily represent the views or policies of the IEA Secretariat or its individual member countries.COPYRIGHT STATEMENT This content may be freely used,copied and redistributed,provided appropriate credit is given(please refer to

14、the Suggested Citation).The exception is that some licensed images may not be copied,as specified in the individual image captions.SUGGESTED CITATION Selj,J.,Wieland,S.,Tsanakas,I.(2025).Selj,J.,Jahn,U.,Maugeri,G.(Eds.),Floating Photovoltaic Power Plants:A Review of Energy Yield,Reliability,and Main

15、tenance(Report No.T13-31:2025).IEA PVPS Task 13.https:/iea-pvps.org/key-topics/floating-pv-plants/COVER PICTURE PV Floating System in Zwolle,Netherlands.Copyright BayWa r.e.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and M

16、aintenance 3 INTERNATIONAL ENERGY AGENCY PHOTOVOLTAIC POWER SYSTEMS PROGRAMME IEA PVPS Task 13 Reliability and Performance of Photovoltaic Systems Floating Photovoltaic Power Plants:A Review of Energy Yield,Reliability,and Maintenance Report IEA-PVPS T13-31:2025 April 2025 Task 13 Reliability and Pe

17、rformance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 4 AUTHORS Main Authors Josefine Selj,IFE,Kjeller,Norway Stefan Wieland,Fraunhofer ISE,Freiburg,Germany Ioannis Tsanakas,CEA INES,Le Bourget-du-Lac,France Contributing Authors Wilfried van

18、Sark,Utrecht University,Utrecht,the Netherlands Nathan Roosloot,IFE,Kjeller,Norway Gaute Otnes,IFE,Kjeller,Norway Vilde S.Nysted,IFE,Kjeller,Norway Minne de Jong,TNO,Eindhoven,the Netherlands Jan Kroon,TNO,Petten,the Netherlands Leonardo Micheli,Sapienza University of Rome,Italy Sara Golroodbari,Utr

19、echt University,Utrecht,the Netherlands Christian Reise,Fraunhofer ISE,Freiburg,Germany Anna Heimsath,Fraunhofer ISE,Freiburg,Germany Andreas Beinert,Fraunhofer ISE,Freiburg,Germany Christian Gertig,Everoze,Bristol,United Kingdom Josh Stein,Sandia,Albuquerque,United States of America Kevin Anderson,

20、Sandia,Albuquerque,United States of America Emilio Muoz,University of Jan,Jan,Spain Matthew Berwind,Fraunhofer ISE,Freiburg,Germany Adam R.Jensen,Technical University of Denmark,Copenhagen,Denmark Carlos D.Rodrguez-Gallegos,RINA Consulting,Australia Oktoviano Gandhi,Solar Energy Research Institute o

21、f Singapore,Singapore Lokesh Vinayagam,Solar Energy Research Institute of Singapore,Singapore Theodoros Makris,RWE,Germany Ingrid Hdrich,Fraunhofer ISE,Freiburg,Germany Christian Reichel,Fraunhofer ISE,Freiburg,Germany Suraj Ravindrababu,Fraunhofer ISE,Freiburg,Germany Editors Josefine Selj,IFE,Kjel

22、ler,Norway Giosu Maugeri,RSE,Italy Ulrike Jahn,Fraunhofer CSP,Halle,Germany Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 5 TABLE OF CONTENTS Acknowledgements.6 List of abbreviations.7 Executive summary.9 Int

23、roduction.12 FPV Energy Yield.16 2.1 Meteorological input data for FPV.17 2.2 FPV technology overview.17 2.3 Energy production estimates for FPV.20 2.4 Modelling yield for FPV systems.32 2.5 Uncertainty analysis.35 Reliability of Floating PV.37 3.1 Defining degradation.38 3.2 Driving degradation des

24、cribing FPV specific stressors.39 3.3 Understanding and quantifying degradation sources of information.41 Operation and Maintenance of Floating PV.49 4.1 Instrumentation.49 4.2 Main O&M actions,importance and best practices.50 4.3 Failure modes and effects analysis in O&M:example for FPV.54 4.4 O&M

25、budgeting Cost aspects.57 4.5 Outlook:O&M challenges and opportunities.58 Conclusion.60 References.61 Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 6 ACKNOWLEDGEMENTS This report received valuable contributio

26、ns from several IEA-PVPS Task 13 members and other international experts.The contributors to the report have received funding of their work through several projects and funding bodies,as listed below.This report is supported by the German Federal Ministry for Economic Affairs and Climate Action(BMWK

27、)under contract no.03EE1120B.This report is supported by the Research Council of Norway,through the projects HydroSun 328642 and Predict 344524.This report is supported by the Danish Energy Agency through grant no.134223-496801.This report was supported by the U.S.Department of Energys Office of Ene

28、rgy Efficiency and Renewable Energy(EERE)under the Solar Energy Technologies Office Award Number 52788.Sandia National Laboratories is a multimission laboratory managed and operated by National Technology&Engineering Solutions of Sandia,LLC,a wholly owned subsidiary of Honeywell International Inc.,f

29、or the U.S.Department of Energys National Nuclear Security Administration under contract DE-NA0003525.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 7 LIST OF ABBREVIATIONS BOS Balance of system CAPEX Capital

30、Expenditures CFD Computational Fluid Dynamics EU European Union EYA Energy yield assessment FEM Finite element method FMEA Failure modes and effects analysis FPV Floating photovoltaics FRP Fiber-reinforced plastic GPV HDPE Ground-based photovoltaics High-density polyethylene HJT Heterojunction IAV I

31、nter-annual variability IEA IEC I-V International Energy Agency International Electrotechnical Commission Current-Voltage LCOE LeTiD LSLR Levelized cost of electricity Light and elevated temperature induced degradation Least squares linear regression MPPT Maximum power point tracker NIR Near infrare

32、d OLS Ordinary least squares O&M Operation and Maintenance PERC Passivated emitted and rear contact PID Potential induced degradation PLR Performance loss rate POA Plane-of-Array irradiance PV Photovoltaics RH Relative humidity ROV Remotely operated vehicles RPN Risk priority number SC Short circuit

33、 SCADA Supervisory Control and Data Acquisition SR Shading ratio SSC Sea State Codes STL STC Seasonal and trend decomposition Standard test conditions TopCON Tunnel oxide passivated contact TRL Technology readiness level UAV Unmanned Aerial Vehicle UV Ultraviolet VIS Visible Task 13 Reliability and

34、Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 8 WIL Wave-induced losses WIML Wave-induced mismatch losses WVTR Water vapor transmission rate WS Wind speed YoY Year-on-Year Task 13 Reliability and Performance of Photovoltaic Systems

35、Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 9 EXECUTIVE SUMMARY Photovoltaic(PV)systems are essential for the transition to sustainable energy,reducing fossil fuel dependence and mitigating climate change.Although PV requires minimal land area PV can meet the Europe

36、an Unions energy needs using only 0.26%of its land space for deployment is often scarce in densely populated regions.Floating photovoltaics(FPV)offer an effective solution to land-use challenges by installing PV systems on floating structures in water bodies.FPV is a growing niche within PV with a c

37、umulative installed capacity reaching 7.7 GW globally by 2023.Almost 90%of the installed FPV capacity is in Asia,with close to 50%of in China alone,while the Netherlands and France are the largest markets outside Asia 3.FPV shows strong potential to support climate targets,but still faces challenges

38、 like regulatory barriers,cost competitiveness compared to ground-based PV(GPV),and uncertainties about environmental impacts and system reliability.FPV systems are currently installed mainly on sheltered inland waters,such as quarry lakes,irrigation ponds and reservoirs.FPV technical standards are

39、still being developed.Guidelines have been published by the World Bank,DNV,and Solar Power Europe,and emerging national standards from South Korea,China,and Singapore address design,components,and safety.The International Electrotechnical Commission(IEC)is working on formal standards for floats,moor

40、ing systems,and electrical connectors.However,the published best practices lack quantitative guidance for yield modelling and reliability,which this report aims to address.It provides data-driven insights,models,and parameters essential for accurate energy yield,reliability,and maintenance predictio

41、ns over FPV systems lifetimes.ENERGY YIELD ASSESSMENT The report provides guidelines and quantitative recommendations for the accurate energy yield assessment(EYA)of FPV systems,a key factor for determining the levelized cost of electricity and project profitability.Current models for EYA are insuff

42、icient,lacking reliable data for critical parameters like module temperature,wave-induced losses,soiling losses,and performance loss rates.Standard modelling tools do not adequately cover FPV-specific needs,and existing meteorological databases often exclude sea and coastal areas,which limits FPV yi

43、eld estimation.This chapter identifies essential parameters and highlights knowledge gaps in meteorological data,energy production modelling,and uncertainty analysis that distinguish FPV EYA from that of GPV.1.Meteorological Data Requirements:The report highlights the need for improved meteorologica

44、l data tailored to FPV,as the water-based environment affects variables like irradiance,wind,and temperature.It is uncertain how this affects prediction accuracy for FPV.2.Thermal Losses:Thermal performance depends on the FPV system design.Modelling tools,such as PVsyst and pvlib,need to incorporate

45、 these specifics for more accurate yield estimates.3.Wave-Induced Losses:Wave motion affects irradiance by altering module tilt and creating irradiance non-uniformity.No complete model for wave-induced losses currently exists and the report encourages field data collection to improve accuracy in yie

46、ld modelling.4.Soiling Losses:FPV systems may experience unique soiling challenges,including bird droppings and other debris from surrounding ecosystems.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 10 The re

47、port underscores that FPV yield estimation tools and methods are still evolving,and it encourages improved empirical studies and data-sharing to refine modelling approaches and align them with the distinctive characteristics of FPV installations.RELIABILITY When assessing the reliability of an FPV s

48、ystem,one faces important knowledge gaps and challenges.First,the stress profiles experienced by components in a FPV installation are neither well understood nor quantified and will vary a lot depending on float technology and water body conditions.Second,open information and systematic studies on o

49、bserved degradation and field failures remain scarce,as are studies of performance loss rates(PLR).And third,as a result of the first two points,there is no accelerated stress testing protocol developed for component reliability evaluation.In the following,each of these three topics will be further

50、discussed.The term degradation denotes the gradual process of change in characteristics with operational time of a material/component/system triggered by stress impact.Typically,we distinguish between three types of degradation:reversible degradation,irreversible degradation,and failures.For FPV,the

51、 balance of system components may be even more critical than the PV modules.Junction boxes,cables,connectors,and related protecting materials may suffer from additional stress compared to GPV systems.The report provides an overview of environmental stressors in the operating environment of FPV syste

52、ms,and finally discusses three different sources for quantification of degradation effects:1.Field data:collection of long-term field data is indispensable for accurate identification of failure modes and the design of appropriate testing protocols.As the available field data on failures and degrada

53、tion is very limited,performance stability is measured through long-term trends in historical production data.Three commonly used statistical methods are deployed to calculate the PLR through historical PV performance and climatic data:Ordinary Least Squares(OLS),seasonal and trend decomposition usi

54、ng locally weighted scatterplot smoothing(STL),as well as Year-on-Year(YoY).These methods are based on determining trends in the historical data.The major drawback of statistical methods is that they do not trace the correlation of the evaluated degradation rates with the climatic variables and degr

55、adation processes.Despite the significant number of FPV systems that have now been operating for several years,long-term FPV performance studies are rare.A study by SERIS using three years of data from a large FPV test bed found PLRs between-0.7%and-0.5%per year,like those of nearby rooftop PV.2.Lab

56、oratory:in the lab environment,accelerated stress tests enable reliability screening of key components in short timeframes,to identify and mitigate quality issues before they manifest as problems in actual installations.A challenge with laboratory testing for FPV is that there are no standards on re

57、liability testing of FPV components,and few field measurements of stressors and field degradation.IEC standards that can be relevant for FPV components are summarized in this section.3.Simulation Models:simulations are one convenient option to overcome the lack of experimental(long-term)data,and to

58、capture the correlations between the degradation rates and the stressors/climatic variables.However,we emphasize the importance of using validated simulation models to obtain reliable results.For FPV,four types of simulations are of particular interest to study the influence of single stressors:a)Wi

59、nd loads through Computational Fluid Dynamics(CFD)and mechanical simulations,b)Moisture ingress through mass transport simulations,c)Hotspot formation through electrical and thermal simulations,d)Thermally Induced Stress through thermal and mechanical simulations.Task 13 Reliability and Performance

60、of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 11 OPERATIONS&MAINTENANCE There are currently no standards available that describe the recommended sensors and procedures for monitoring of FPV power plants.Instrumentation requirements for GPV powe

61、r plants,including requirements with respect to accuracy and number according to the size of the plant,can be found in IEC 61724-1.The report introduces a preliminary failure mode and effects analysis of technical and operational challenges,and how these impact operation and maintenance(O&M).Availab

62、le data is limited,and one can only anticipate that the occurrence and degree of severity for the different events may change as more data is collected.A list of key aspects and considerations when budgeting for FPV O&M projects is also provided.FPV technology faces key R&D challenges,especially as

63、installations scale up and offshore projects expand.Major areas include:Monitoring and Remote Sensing:Remote FPV sites,especially offshore,struggle with data transmission reliability and high communication costs.Advanced solutions using drones and satellites can enhance monitoring and reduce O&M cos

64、ts.Expert Dependence:FPVs complexity requires specialized experts(e.g.,divers,marine engineers)for maintenance and inspections,increasing costs and time.AI-driven data analytics,Unmanned Aerial Vehicle(UAV)based inspections,and autonomous systems offer potential to reduce human intervention.Extreme

65、Weather and Degradation:Marine environments introduce severe stressors like corrosion and UV exposure,accelerating FPV component degradation.R&D in advanced materials,protective designs,and robust emergency-response plans is crucial to improve FPV durability.Environmental Impact and Regulations:Conc

66、erns on FPV effects on aquatic ecosystems,such as water quality and habitat shading,call for eco-friendly designs and regulatory standards that minimize harm and adapt O&M practices for sustainability.CONCLUSIONS FPV offers a promising solution for expanding renewable energy without increasing land-

67、use pressures.However,the absence of regulatory frameworks and limited long-term data creates uncertainty for developers,regulators,and investors,slowing FPV adoption.Rapid innovation in the field often prioritizes confidentiality,even though the industry would benefit from open data sharing.This re

68、port aims to support FPV development by building a knowledge base on energy yield,reliability,and O&M areas where FPV diverges from GPV.Key research priorities include understanding FPV-specific stressors,improving predictive models,automating O&M,and assessing environmental impacts.Addressing these

69、 gaps can lead to a more mature,sustainable FPV industry,ready for broader deployment.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 12 INTRODUCTION PV is a cornerstone in the transition to sustainable energy,

70、and accelerated deployment is essential to reduce reliance on fossil fuels and mitigate global warming.While PV systems occupy relatively small amounts of land for instance,meeting the current energy demand of the European Union(EU)with PV would require only 0.26%of its total land area 1 land availa

71、bility for solar deployment is often limited in densely populated areas.One solution is to deploy PV systems on water bodies.Floating Photovoltaics refers to mounting solar photovoltaic systems on structures that float on water.It is a relatively novel,but rapidly growing technology,exhibiting promi

72、sing synergies with other usage of water bodies.Figure 1.Annual and cumulative growth in deployment of FPV by installed capacity(MWp)and by percentage of total global PV installations(%).Figure 1 shows the cumulative and annual installed capacity of FPV since the first deployment in 2007.The deploym

73、ent has grown from just above 1.6 GW at the end of 2018 to 7.7 GW by the end of 2023 2.Almost 90%of the installed FPV capacity is in Asia,with close to 50%of in China alone 3,followed by Taiwan,India,Israel,Japan and South Korea.However,FPV also holds potential to support the EUs climate neutrality

74、goals,with the Netherlands and France currently hosting the 7th and 10th largest FPV capacities 3.As the technology matures,Figure 2.Categorization of FPV as suggested by Solar Power Europe 1 and the WMO sea state codes 4.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power

75、Plants:A Review of Energy Yield,Reliability,and Maintenance 13 FPV deployment is expected to accelerate further,but several barriers persist.Legislative hurdles,cost competitiveness relative to ground-based PV,and uncertainties surrounding environmental impact and reliability could impede global ado

76、ption.The commercial deployment of FPV systems today is on sheltered waters.However,there is currently no consensus regarding how deployment on different types of water bodies should be classified.Solar Power Europe 1 suggests a division between onshore(or inland)FPV and marine FPV,where onshore is

77、further separated into static freshwater bodies,inner waters,and larger inner waters,while marine FPV is separated into nearshore FPV and offshore FPV,as shown in Figure 2.Another option is to use the Sea State Codes(SSC)of the World Meteorological Organization(WMO).This scale spans from no waves(SS

78、C 0)to wave height 14 m(SSC 9).The two may also be used in combination,although the suggested wave height in 1 must be modified to fit with established WMO SSC codes.In this report,we use the term inland FPV for deployment of FPV on freshwater bodies,while the term nearshore FPV is used for deployme

79、nt of FPV at sea,but close to the coastline.This report is limited to address inland and nearshore FPV,as some technologies may be suited for both applications.The report does not cover offshore FPV applications,as both the challenges it faces,and the FPV technologies under development for these con

80、ditions differ significantly from inland and nearshore FPV.The majority of current FPV installations in Europe are deployed on quarry lakes,sandpit lakes,irrigation ponds and other man-made water bodies.Figure 3 shows installations on a)a mining and quarry lake,b)a sandpit lake and c)a hydropower da

81、m.The installation and operational costs will depend significantly on the category of application,but focusing on sheltered waters(corresponding to FPV deployed at Static Freshwater bodies and Inner waters),a cost premium of 20-25%in Europe and USA respectively is estimated compared to GPV 4,5.The C

82、APEX is highly affected by float costs,which are influenced by wind and snow loads,as well as the efficiency of the PV modules.The CAPEX estimates Figure 3.Installation of FPV on quarry lake,sandpit,and hydropower dam in Europe.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV

83、Power Plants:A Review of Energy Yield,Reliability,and Maintenance 14 are also in line with a median CAPEX cost of 1.18 USD/W(2022),recently reported by SERIS based on their comprehensive FPV database 3.Of all projects installed in 2022 or earlier,30 projects have reported CAPEX lower than 1 USD/W an

84、d the lowest CAPEX reported for any FPV project is 0.41 USD/W for 36 MW FPV in India 6.The development of technical standards for(any type of)FPV systems is currently limited but actively pursued by various national and international organisations.These efforts aim to address the unique challenges a

85、ssociated with deploying solar PV systems on water bodies.The first comprehensive guideline was the Floating Solar Handbook for Practitioners published in 2019 7.The technical considerations covered by the handbook include site selection and assessment,FPV system design and components,electrical saf

86、ety measures,as well as operation and maintenance procedures.Also published in 2019 were national standards for PV modules in FPV applications by South Korea 8 and for the high-density polyethylene(HDPE)floats and structures by China 9,10,11,12.While the Floating Solar Handbook for Practitioners is

87、targeted towards a more general audience and includes non-technical aspects,more normative standards are provided by the DNVGL-RP-0584 Recommended Practice 13 and the Singapore Technical Reference TR 100:2022 14.DNVGL-RP-0584 was developed in a joint industry project with FPV developers,investors,in

88、stallers,and equipment suppliers.Meanwhile,Singapores TR 100:2022 modified the GPV standard IEC TS 62738:2018 15 for FPV contexts by addressing the specific nature of FPV systems over water;also referencing other existing standards from adjacent technologies such as the marine industry.Efforts are o

89、ngoing in the IEC Technical Committee 82,Working Group 3(TC82 WG3)to formalise an international standard for FPV systems.This standard will also have links to two other Recommended Practices by DNV,which are currently underway to develop better guidance on the floats(various types)and the anchoring&

90、mooring systems.Similarly,IEC TC82 WG2 plans to develop a standard for PV modules when deployed in FPV applications.Generally,the key focus areas addressed in the technical standards include proper design,component selection and implementation of FPV systems to ensure long-term durability and reliab

91、ility,including(but not limited to)cables,floats,anchors,and mooring systems,taking into account the exposure of components to high humidity,their possible submergence into the water and the dynamic environmental conditions(combined wind,wave,and possibly tidal forces).Ultimately,clear and comprehen

92、sive guidelines will boost FPV system design and installation quality,as well as investors confidence and bankability.While the published best practice reports from DNV 13,the World Bank 7,and Solar Power Europe 1 provide valuable contributions to different areas within FPV,they fall short of offeri

93、ng quantitative recommendations for energy yield modelling or addressing reliability and field failures.The aim of this report is to address these gaps by focusing on topics at the forefront of FPV research.Specifically,we provide observations,models,and quantitative parameters to support the effici

94、ent and economically viable deployment of FPV systems.The report examines factors influencing the energy yield over the lifetime of FPV plants,along with the associated O&M requirements,incorporating quantitative values where available.Sustainability of FPV power plants represents a very wide and mu

95、ltifaceted topic,encompassing everything from environmental impacts,carbon footprints,and recycling,to socio-economic and socio-cultural impacts.With respect to environmental impacts,a wide range of both potential benefits and potential adverse impacts have been discussed in the scientific literatur

96、e.Case studies detailing the Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 15 impact of FPV power plants at specified sites are starting to appear 16,17,18 as are guidelines for monitoring 19 and impact asses

97、sment 20.However,the impact depends on the sensitivity of the water body itself,local climate,and FPV design and coverage,making it is difficult to generalize the findings.Overall,the knowledge base on the environmental impact of FPV power plants remains limited,and more research is needed.IEA PVPS

98、Task 12 published a report on Carbon Footprint Analysis of FPV in 2024 21,concluding that the two studied FPV technologies had a slightly greater carbon footprint than the GPV references,but seven times lower than that of the grid mix in the countries where they were installed(Germany and the Nether

99、lands).The largest contribution to the carbon footprints is from the manufacturing of the PV module(60%to 70%,depending on the system).The scientific literature provides very little information on social impacts,and acceptance,of FPV.Most large waterbodies provide extensive ecosystem services to loc

100、al communities including fishery,irrigation water,recreation,tourism and transportation.While it is often claimed that installing FPV instead of GPV will reduce the level of conflict for space,there are few studies that can support these arguments 22.A detailed discusson on sustainability of FPV is

101、comprehensive and beyond the scope of this report.We recommend that these aspects are dealt with in a separate report.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 16 FPV ENERGY YIELD The purpose of this chap

102、ter is to provide concrete guidelines and quantitative recommendations for a selection of parameters essential for accurate energy yield assessment(EYA)of various FPV technologies.EYA,or modelling of the average yearly expected energy production,is crucial for determining the levelized cost of elect

103、ricity(LCOE)and,consequently,the profitability of a project.The absence of published and validated values for parameters used in EYA for FPV can hinder the development and investment in new FPV projects.The toolbox for EYA of FPV is currently inadequate.Models that accurately estimate module tempera

104、ture and wave-induced losses(WIL)are not available in standard modelling software.Soiling losses and performance loss rates(PLR)should ideally be based on published and validated empirical values,but such values are currently unavailable.The validation of models and empirical values is challenging d

105、ue to the scarcity of high-quality data series that include both production and weather data.Additionally,while meteorological data suitable for energy production estimates for PV is readily available,coverage of sea and coastal areas is missing from many databases and PV modelling tools.An importan

106、t effort in this report is hence to provide an overview of both what is known and current gaps in knowledge needed for accurate EYA of FPV.EYA can be divided into three components:meteorological input data,energy production estimates,and uncertainty analysis.As good procedures for EYA exists for GPV

107、,the focus in this chapter is on the topics that separate EYA for FPV from GPV.Section 2.1 provides a brief overview of current knowledge and knowledge gaps related to meteorological data sets above water bodies.Sections 2.2,2.3,and 2.4 are concerned with energy production estimates.A challenge for

108、energy production estimates for FPV is the diversity of FPV technology.In Section 2.2,we give a brief introduction to FPV technologies and classification schemes for the different technologies.The quantitative losses and the models to describe them will depend on both FPV technology and site/sea sta

109、te,inferring that a broad set of data series spanning different type of technologies and sites will be necessary to gain a good understanding and validate models and values in different conditions.We also address the impact of technology design choices on various losses.In Section 2.3,we describe th

110、ree loss mechanisms that are important input to energy production estimates,and which differ from their GPV counterparts in terms of value and/or origin:Thermal losses,WIL,and soiling losses.The impact of these losses on different types of FPV technologies is,to the extent possible,quantified.The PL

111、R is also an important input to energy production estimates.However,as it represents a sum of various degradation mechanisms,it is covered in Chapter 3 on Reliability.Section 2.4 provides an overview of modelling options for FPV in the industry standard modelling software,PVsyst,and in pvlib,commonl

112、y used in research and development.The third part of an EYA is uncertainty analysis.To date,no published efforts are known for quantifying uncertainties in crucial FPV modelling steps such as soiling or WIL.Section 2.5 summarizes the main methods to model uncertainty and provides advice on how to de

113、al with uncertainty analysis for EYA of FPV.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 17 2.1 Meteorological input data for FPV The long-term and short-term solar resource variability represents the single

114、 greatest uncertainty in a solar power plants predicted performance.Satellite-derived,and high-quality historical solar radiation data sets covering at least 10 years are usually considered necessary for the site selection of large solar energy systems 23.Solar resource data above water bodies has n

115、ot been of interest to the providers of meteorological data for solar power plants until recently.There are currently no dedicated measures undertaken to ensure that the meteorological parameters used in PV yield assessments are accurate above inland water bodies.For nearshore-and offshore locations

116、,the meteorological parameters are lacking in several of the databases commonly used for solar power plants,including SARAH2-PVGIS,ERA5-PVGIS,and Meteonorm(8)for PVsyst.In the new release of PVGIS,the databases will be updated to cover nearshore 25 km towards the sea-to enable assessment of FPV clos

117、e to the coastlines.The spatial resolution for the coastal areas will be the same as for inland areas,SARAH3s native resolution at 0.05 x 0.05,and ERA5 interpolated to the resolution of ERA5-Land at 0.1 x 0.1(A.Martinez,personal communication,May 17,2024).Historical data based on satellite imagery o

118、f meteorological data above sea can be obtained from the NASA POWER service 24.Comparing the irradiance on land with irradiance 57 km offshore,Golroodbari et al.25 finds that in 70%of the locations,the average value for irradiation at the offshore site is higher than at the land-based site.For inlan

119、d water bodies,it will also be of interest to evaluate,and likely improve,the accuracy of temperature and wind data,as this will be affected by the water body.It is also worth noting that easier access to data assessing the sea states of water bodies would facilitate planning and implementation of F

120、PV.2.2 FPV technology overview The term FPV encompasses a wide range of technologies,with a broad range of different FPV floating system manufacturers currently in operation.Figure 4 provides an overview of the major FPV technology manufacturers and their market share measured by installed capacity.

121、Each manufacturer naturally aims to differentiate their products with unique characteristics,making any categorization scheme prone to inaccuracies and oversimplifications.This diversity poses a challenge when modelling EYA for FPV.Early reviews and reports often overlooked this critical information

122、,leading to confusion about the expected performance of various FPV technologies.However,to address the different types of FPV technologies Figure 4.Market share of FPV technologies by installed capacity.Data based on 3.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Pl

123、ants:A Review of Energy Yield,Reliability,and Maintenance 18 efficiently without delving into specifics,we find it beneficial to use the categorization shown in Figure 5.Category 1 is pure-floats,which,as the name indicates,are solutions fully composed of floats,typically made of high-density polyet

124、hylene(HDPE).Category 2 combines metal or fiber-reinforced plastic(FRP)with floats or pipes,with ZIM Float being a well-known European example.Category 3 encompass other types of FPV technologies such as platforms,ferro-cement structures,and membrane technology.The latter is patented by the company

125、Ocean Sun.Technologies submerged in water and FPV systems that utilize active water cooling(water is pumped and sprayed onto the modules)are not specifically addressed in this report.2.2.1 Pure pontoon-based float technology(pure-floats)Pure-float FPV technology was the first to gain commercial trac

126、tion,and the three dominating FPV technologies world-wide,developed by Sungrow,Ciel&Terre,and Northman Energy Technologies are all pure-floats.According to SERIES FPV database,these technology providers have a market share of more than 50%(27%,12,7%and 14,9%respectively)3.The category also encompass

127、es many other,smaller technology providers,covering a range of different float designs.The development of the float technologies over time can also be substantial,as an example,Figure 6 shows the iterative development of Ciel&Terres Hydrelio technology.Naturally,the properties of the floats will cha

128、nge with the design,but also with choices made with respect to tilt angle and electrical configuration.Figure 5.Categorization of FPV system based on float type.Illustration based on 3,8,121.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield

129、,Reliability,and Maintenance 19 Figure 6.Development of Ciel&Terre Hydrelio technology.The numbers refer to accumulated installed capacity for Ciel&Terre up until the given year.Copyright:Ciel&Terre International.2.2.2 Pontoon floats+metal or FRP structures Another type of structure that is utilized

130、 as mounting for FPV is based on a combination of floats(or pipes)and metal or FRP structures.Here,the floats do not support the individual PV modules directly,as with the pure-float technology,but instead support the metal or FRP structure that the PV modules are in turn mounted on to.Examples of t

131、his type of FPV technology include the European ZIM Float by Zimmermann PV-Steel Group and the Korean Scotra.The ZIM Float technology is currently the most widely deployed FPV technology in Europe with more than 250 MW installed predominantly in Germany and the Netherlands.There are currently two ve

132、rsions of the ZIM Float technology,ZIM Float1 and the second-generation ZIM Float2.The South-Korean company Scotra has developed FPV systems for more than a decade and installed their first commercial FPV plant(500 kW)in 2012.There exist many generations of the float with quite substantial developme

133、nts over the years.ZIM Float and Scotra have world-wide market shares(by the end of 2022)of 3.6%and 3.4%respectively 3.An FPV power plant with ZIM Float technology is depicted in Figure 7.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Re

134、liability,and Maintenance 20 Figure 7.FPV power plant with ZIM Float technology installed in 2021 with a capacity of 13.7 MW at Lippe Gabrielsplas in the Netherlands.2.2.3 Non-pontoon-based floats One alternative technology involves deploying PV modules on a thin membrane that floats directly on the

135、 waters surface.This solution patented by the company Ocean Sun 26,27,uses a ring of HDPE material to provide buoyancy,with the PV panels fastened using keders welded onto the membrane.PV modules deployed with this technology experience different environmental impacts compared to pure-floats.Irradia

136、nce conditions and wave movement effects vary,the dominant heat dissipation mechanism is different,and soiling and cleaning will be influenced by the horizontal mounting solution.Consequently,the models and parameters used to describe the yield must also be adapted.2.3 Energy production estimates fo

137、r FPV Estimating the energy production of FPV systems introduces additional challenges and complexities compared to GPV.This section delves into three loss mechanisms that differ significantly from their GPV counterparts.For a comprehensive overview of PV yield and associated losses,refer to the IEA

138、-PVPS Task 13 Report“Performance Modelling Methods and Practices”28.Section 2.3.1 introduces the most relevant parameters for describing the heat exchange between an FPV system and its environment.Section 2.3.2 examines the effects of irradiance non-uniformity caused by wave-induced dynamic tilt var

139、iations.Section 2.3.3 explores the issue of soiling in FPV systems.In each section the loss mechanism is described in general,before quantitative values for different FPV technologies are discussed.The maturity and understanding of these topics vary significantly,which is also reflected in the text.

140、2.3.1 Thermal losses Temperature impacts both the instantaneous and long-term performance of PV modules.As cell temperature increases,PV efficiency decreases.Additionally,high temperatures and BayWa r.e.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of

141、Energy Yield,Reliability,and Maintenance 21 frequent temperature cycles can accelerate various degradation mechanisms.Therefore,minimizing operating temperature and thermal cycling can enhance power generation and potentially extend the lifetime of a PV system 29.The operating temperature of PV modu

142、les mounted in various FPV systems has been widely discussed.Several studies indicate that FPV systems often operate at lower temperatures compared to co-located GPV installations 30,31.However,other studies report similar or even higher operating temperatures for FPV compared to GPV 32.Thus,improve

143、d thermal performance is not an inherent advantage of all FPV installations;it depends critically on system design and local climate conditions.The temperature of a PV module in the field depends on numerous factors,including irradiance,ambient and sky temperature,circulation and humidity of the sur

144、rounding medium,mounting structure,and materials.To(try to)differentiate the impact of all these local climatic conditions from the impact of the technical installation itself,it is convenient to use so-called heat loss coefficients,or U-values.These coefficients measure a systems capacity to exchan

145、ge heat with the environment.Higher U-values indicate better heat exchange and thus leading to cooling and,therefore,lower operating temperatures.Forced convection increases heat transfer between the module and ambient medium compared to free convection,both because a greater temperature difference

146、is maintained and because the heat transfer coefficient increase with turbulence.Hence,wind will have a significant effect on the cooling and therefore an explicit dependence on wind is often introduced by splitting the U-value into a constant term and a windspeed()dependent term:=+WS.In a meteorolo

147、gical context,and hence in most databases that include wind measurements,wind velocity is provided at 10 m above land,in a free environment.These will not be representative values for the wind experienced by a PV system,and it is therefore common to neglect wind effects in EYA of PV.U-values reporte

148、d as a single constant is also default in PVsyst.However,reporting a single U-value inhibits the possibility of finding the explicit dependency of wind.Single U-values are therefore less useful to describe the thermal properties of a specific design than U-values with an explicit term for the wind d

149、ependency.Although U-values are convenient,both when modelling energy yield and comparing the thermal properties of different technologies and PV mounting solutions,they also introduce a wide range of potential pitfalls.The most serious pitfall,perhaps,is that while U-values are often perceived to d

150、escribe a property of the(F)PV system,there is still a significant dependency on several environmental parameters that are not accounted for in the equations.In addition,a U-value is not uniquely defined.To understand why one U-value may not be directly comparable to another,and why the equations de

151、scribing module temperature as a function of U-values take different forms,it is illustrative to look at how the most common models are derived.The UThe U-value value how it is derived and why it is not uniquely definedhow it is derived and why it is not uniquely defined The commonly used models to

152、estimate PV module temperatures are based on a steady-state thermal balance:=0(1)where is the energy flux from the sun,is the energy flux extracted as electrical power,and the remaining three terms are the heat losses by radiation,convection and conduction(all Task 13 Reliability and Performance of

153、Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 22 in W/m2).Often,both the radiative and conductive heat loss are assumed to be negligible,and the convective heat loss is a function of the difference in temperature(T)between the module(m)and its sur

154、roundings(a,ambient),hence ()=0(2)Rearranging to calculate the module temperature Eq.(2)becomes =+(3)The U-values in the literature referenced in this section are derived with models based on Eq.(3).Another simplification that is often implemented,is that the electrical output power scales with the

155、energy flux from the sun.The simplest way to express the scaling would be =,(4)where is the module efficiency.Note that inaccuracies are introduced with this scaling,because the module efficiency is temperature dependent and because is defined as the energy flux entering the module,while the module

156、efficiency is normally defined as a fraction of the incoming irradiance,including light reflected from the surface of the module.It is convenient to use variables that are commonly measured in the field or that are known from module data sheets.In PVsyst,is expressed as,i.e.the incoming radiation,mu

157、ltiplied by the absorption coefficient,(defined as 1 ,where is reflection).can then be expressed as =(5)Inserting this expression for and in Eq.(3)we get =+(1 )(6)This equation,with the choice of using either a single value or the+WS term,is used in PVsyst.Note that in the PVsyst manual the cell tem

158、perature,is used in this equation,while in the Faiman model 33,and in IEC 61853-2 34,module temperature is used.An alternative to expressing as a function of is to express both and as functions of =(7)(8)Inserted in Eq.(3)this gives =+=+()(9)In this version of the equation the use of the module effi

159、ciency is in accordance with the standards for measuring module efficiencies(i.e.reflection is included).There are also other ways of expressing and which lead to small variations in the equations used(such as in Liu et al.35,where is expressed as=,where is the transmittance of the glass and is the

160、absorption of the PV layer).These nuances in the equations will generally lead to negligible differences in the resulting temperature or U-values.However,it must be noted that U-values derived using Faiman or IEC 61853-2 will not Task 13 Reliability and Performance of Photovoltaic Systems Floating P

161、V Power Plants:A Review of Energy Yield,Reliability,and Maintenance 23 be directly comparable to U-values derived using Eq.(6)and Eq.(9).In Faiman/IEC 61853-2 nomenclature,=+0+1=+0+1(10)the optical losses and electrical efficiency of the module are integrated in the U-values.This is easily overlooke

162、d.None of the U-values cited in Table 1 have the module efficiency term included in the U-values.Finally,there are also models that include the radiative term in Eq.(1).This could be of relevance to improving the accuracy of the temperature models in general,and of particular relevance if night-time

163、 temperature effects are of interest,e.g.for degradation models.Driesse et al.36 have suggested how the equations above can be altered to include a radiative term.Eq.6,9,or 10 can be used to derive U-values based on field measurements and module parameters.AirAir-cooled FPV systemscooled FPV systems

164、 Most FPV systems installed so far can be called Air-cooled FPV systems,meaning designs where the modules are not in direct contact with water and their only ambient medium is air.In this configuration the thermal behaviour of the FPV system is influenced by the same parameters as a GPV system:irrad

165、iance,air temperature,wind speed,humidity,and mounting design.Such FPV systems are predominantly cooled by air.The water temperature does not have a direct effect on the operating temperature of the module,but can influence the ambient air temperature,and hence indirectly the operating temperature o

166、f the module.The mounting structure and the local wind conditions are therefore the most critical parameters to determine the cooling efficiency of air-cooled FPV systems.Figure 8.Heat loss coefficients(U-values)based on the different PV structures.Source:Liu et al.2018 35 Task 13 Reliability and Pe

167、rformance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 24 As illustrated in Figure 8,mounting structures that obstruct the rear surface of the modules(i.e.they have a“large footprint”on the water surface)will also usually lead to less efficien

168、t cooling by the wind and greater operating temperatures than mounting structures that are more open(i.e.they have a“small footprint”on the water surface).In large footprint configurations,the operating temperatures of FPV may exceed those of well-ventilated GPV installations 29,32,37.Liu et al.publ

169、ished one of the first papers quantifying U-values for systems categorized as free-standing,small-footprint,and large-footprint 37.Figure 9.The wind direction impacts the best fit U-value.The graph is based on measurements performed on a Ciel&Terre FPV system in Marlenique,South Africa 38.Given the

170、same wind velocity,it is evident that the FPV modules are more efficiently cooled by wind coming from the rear than from the front.Local wind conditions are affected by topography,vegetation,and infrastructure.Generally,the wind speed will increase over open areas,such as water bodies.The size of th

171、e water body,vegetation/buildings on the shore,and the placement of the FPV system on the water body,will influence the wind conditions experienced by the system.The wind direction may also be of importance for the operating temperature of the PV modules.Figure 9 illustrates how wind direction impac

172、ts the U-value(and hence operating temperature)of a Ciel&Terre system deployed in Marlenique,South Africa.In this system,the wind cools the modules more efficiently when coming in from the rear side of the modules(which are all oriented in the same direction)38.WaterWater-cooled FPV systemscooled FP

173、V systems FPV designs where the modules can be mounted horizontally,directly in contact with the water surface,also exists.We will denote these systems as water-cooled.In this configuration,the modules can benefit from the higher heat transfer coefficient of water compared to air and of the more lim

174、ited thermal cycles water basins experience.In this configuration,the water temperature becomes the dominant parameter affecting the cell temperature.Water-cooled FPV systems are likely to achieve lower operating temperatures,compared to air-cooled installations.The circulation of the ambient water

175、significantly influences the operating temperature of such FPV systems,with increasing water velocities leading to more efficient heat transport,equivalent to the effect of wind velocity.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Rel

176、iability,and Maintenance 25 PurePure-f floatsloats An FPV module mounted on a pure-float structure will be cooled by the ambient air and wind,fundamentally like a module mounted on a GPV system.The most important feature of the pure-float with respect to thermal losses,is therefore to what extent it

177、 exposes the PV module to wind.Even within the individual categories introduced by SERIS,there will be substantial differences in the thermal properties of the individual floating structures.Even though FPV systems consisting of a version of pure-float technology is dominating on the world market,re

178、markably,only a few publications with quantitative assessment of U-values from such systems exist.In the categorization by Liu et al.the pure-float belongs to the large footprint tag and has a reported U-value of around 30 W/m2K 37,similar to a well-ventilated rooftop system.The pure-float systems i

179、n reference 37 is the Hydrelio Classic FPV technology from Ciel&Terre.Another analysis performed on unspecified large footprint systems in the Netherlands and in Singapore reports median U-values of 37 W/m2K and 36 W/m2K respectively 29.The wind-dependent U-values of the same systems are =24.4+6.5 a

180、nd =34.8+0.8.From a Ciel&Terre pure-float system in Marlenique in South Africa,U-values of =19.3+6.2 and =20.2+3.0 are reported 38,depending on the prevailing wind direction.Note that the floats in the pure-float category have been continuously developed.The footprint of the floats has decreased and

181、 the accessibility of wind underneath the mounted module has increased with development of the pontoons,as exemplified in Figure 6.No absolute criterions for the footprint classification have been published,but it may well be that the most recent pure-float models no longer fit in the large footprin

182、t classification.In much of the scientific literature available today,only the type/category the FPV system belongs to is provided,while the specific technology and producer is not detailed.With respect to accessibility and transparency of information it would be beneficial to report the specific FP

183、V technologies tested,rather than only the type/classification.Metal or FRP structures+floats/pipesMetal or FRP structures+floats/pipes As with the pure-floats,the metal/FRP structure mounted on floats or pipes will be cooled by the ambient air and wind.This type of structure will usually allow more

184、 air to flow beneath the PV modules and will often be categorized as a medium footprint structure.In 29 data from this type of structure is fitted to provide a median(single)U-value of 41 W/m2K,or,with the wind term included,U=18.9+8.9.The single U-value is 5 W/m2K greater than the median U-value of

185、 the large footprint system in the same study(see Table 1).However,comparing the U-values with explicit wind terms,the wind dependent term is significantly greater for the Metal/FRP structure,implying that a greater differences in the operating temperature of these two technologies will be expected

186、at windy sites.ZIM Float,the FPV technology with the highest installed capacity in Europe,is Category 2,but there are also several other technologies in this category.There is ongoing work to publish performance analysis of FPV power plants with ZIM Float technology,but the only work currently publi

187、shed with a confirmed Category 2 FPV technology is 29(O.Gandhi,SERIS,personal communication,August 6,2024).Membrane FPV technologyMembrane FPV technology When PV modules are installed on a thin membrane directly floating on the water surface,it alters the dominating heat dissipation mechanism.While

188、PV modules mounted on pontoons will be predominantly cooled by air convection,PV modules mounted in thermal contact with a thin membrane will predominantly release heat through the membrane.The membrane in turn Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Re

189、view of Energy Yield,Reliability,and Maintenance 26 is cooled by water convection.The circulation of water underneath the membrane is therefore important for efficient cooling of the membrane and modules.For water bodies with significant circulation,a simplified Computational Fluid Dynamics(CFD)calc

190、ulation suggests that almost all heat absorbed by the PV module is conducted down into the water 39.Measurements performed at a pilot site in Skaft,Norway,show that U-values are in the range of 70-80 W/m2K for this technology 31.The modules in this technology effectively have air as ambient medium o

191、n their front side and water as ambient medium on their rear side(assuming that the membrane is thin).Hence,the model for heat dissipation should not use only air as ambient.Currently,only air is available as ambient medium in the commercial modelling tools.It is recommended to include water tempera

192、ture data to accurately calculate module operating temperatures or U-values for FPV technologies where the PV modules are in contact with water.Figure 10 shows an installation with Ocean Sun technology at the Magat dam in the Philippines.Figure 10.Ocean Sun FPV system in Magat in the Philippines.Cop

193、yright:Ocean Sun Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 27 Table 1:Summary of U-values reported in literature.Cooling mechanism Configuration Location W/m2K+W/m2K Ref Air-cooled Free standing GPV array

194、 Default,PVsyst 29 29.0+0 WS 40 Air-cooled Free standing GPV array PVUSA 25.0+1.2 WS 40 Air-cooled Hydrelio Classic,Ciel&Terre Tengeh Reservoir,in Singapore,near the sea 30 NA 37 Air-cooled Tracked,small footprint and open structure Inland lake(NL),near the sea 57 24.4+6.5 29 Air-cooled 17-tilted,Ea

195、st-West oriented,large footprint and closed structure Inland lake(NL),near the sea 37 25.2+3.7 29 Air-cooled 7-tilted east,Large footprint and close structure,SG Tengeh Reservoir,in Singapore,near the sea 36 34.8+0.8 29 Air-cooled 12-tilted east,medium footprint and close structure,SG Tengeh Reservo

196、ir,in Singapore,near the sea 41 18.9+8.9 29 Air-cooled 10-tilted east,free standing and open structure,SG Tengeh Reservoir,in Singapore,near the sea 55 35.3+8.9 29 Air-cooled Horizontal module,3.2-cm off a floating membrane Fjords Inner branch on(NO)west coast 46 NA 31 Air-cooled SolarisFloat azimut

197、hal tracking FPV system Lake Oostvoorne(NL)39.5 24.7+3.9 30 Air-cooled Current Solar 15-tilted modules mounted on high-density PE pipes Small pond,Kilinochchi,Sri Lanka 33.2 25.7+2.8 30 Air-cooled 15-tilted modules mounted with Ciel&Terre FPV system Small pond,Marlenique(SA)NA 20.2+3.0 38 Air-cooled

198、 15-tilted modules mounted with Ciel&Terre FPV system Small pond,Marlenique(SA)NA 19.3+6.2 38 Water-cooled Horizontal module on a floating membrane Fjords Inner branch on Norwegian west coast 711 NA 31 1 As the cooling mechanism infers,this U-value has been calculated using water temperature,and not

199、 air temperature,as ambient medium temperature.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 28 2.3.2 Wave-induced losses FPV systems are floating structures that move with wind and waves,causing fluctuations

200、 in the effective tilt angle and orientation of the PV modules.This movement affects the incident irradiance on the modules and gives rise to wave-induced losses(WIL).The WIL of an FPV system encompass an irradiance loss(or gain)due to a difference in average tilt of all the modules compared to a st

201、atic system with the same tilt,and a mismatch loss due to different irradiance conditions on the individual modules.Currently,only a limited number of studies have been published on this topic,and terminology is still evolving.Drenkmper et al.introduced the term wave-induced mismatch loss(WIML),defi

202、ning it as the difference in power output between a system with fixed tilt and one with tilt that varies due to wave motions 41.In this definition,WIML consists of two components:one due to different irradiance conditions on the individual modules(mismatch loss),and one due to a difference in averag

203、e tilt of all the modules(irradiance loss or gain).Chen et al.42 use the terms mismatch induced loss and insolation induced loss to address the two components separately.For practical purposes,modelling software like PVsyst may treat WIL as a combined“effective”mismatch value.However,to fully unders

204、tand WIL,it is helpful to address the mismatch and irradiance components separately and use the term wave-induced losses(WIL)to encompass both.All PV systems are affected by mismatch losses.Mismatch losses within a string of series connected PV modules are caused either by differences in the experie

205、nced conditions or in the rated power of the individual modules.The output of the string will be limited by the module with the lowest current.A wave-induced mismatch loss will come in addition to the mismatch loss induced by differences in rated power(or degradation).The WIL depends significantly o

206、n a large set of parameters,including both environmental parameters and the FPV structure itself.Currently,measured values for WIL are not accessible.Modelling of WIL requires three basic modelling steps:1)modelling of the interaction between the FPV structure and waves,2)modelling of the irradiance

207、 on the array,and 3)modelling the electrical response to the irradiance.Results from modelling approaches are scarce and only validated to a limited extent,and complete validation requires high resolution measurements of(at least)sea state/waves,module/float movement,design plane-of-array(POA),and i

208、rradiance at module-level.There are,however,modelling efforts published that provide insight to the sensitivity of WIL to different parameters.In DNVs recommended practice 13 the current recommendation is to use one of three methods to estimate WIL:wave tank measurements,numerical analysis or engine

209、ering judgement.It may not be feasible to model WIL in detail for each new FPV project,but with increasing number of publications and experience,the engineering judgements will become more precise.Providing generic values for WIL is not possible due to the significant technology dependence.However,a

210、 table with WIL values for the most established FPV technologies for different sea states would be useful,but this information is not currently available.Impact of FPV technology on WILImpact of FPV technology on WIL Different aspects of FPV technology will be decisive for various loss mechanisms.Nu

211、mber of modules per float.Intuitively,the number of modules per float impacts WIL significantly.When all modules that are connected to the same maximum power point tracker(MPPT)are situated on the same rigid float,they all have the same orientation Task 13 Reliability and Performance of Photovoltaic

212、 Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 29 and do not experience wave-induced mismatch losses.Note that the system can still experience changes in the effective tilt,and hence irradiance and production,due to waves.Systems with one or a few modules per

213、float experience different orientations due to wave motion,leading to increased mismatch losses(if connected to the same MPPT).The extent to which floats follow wave movement also affects WIL.Designs like membrane-based systems,which move freely with waves,are expected to experience higher mismatch

214、losses.Pure-float systems are somewhat more resistant to wave influence(depending on how they are connected),while semi-closed structures or systems using rigid materials(metal or FRP with floats/pipes)show lower mismatch due to a more rigid structure.The electrical configuration and the length of t

215、he PV string has also been shown to affect the mismatch loss.For commercial size FPV systems with more than 20 PV modules in a string,the mismatch loss will saturate 41,43.It can be understood by looking at the probability of(at least)one module positioned in the“worst”possible angle towards the sun

216、.Impact of environmental parameters on WILImpact of environmental parameters on WIL The severity of WIL also depends heavily on environmental parameters such as the sea state and the irradiance angle of incidence 13,41.There is currently no published literature with measured values for WIL.A few pap

217、ers publish modeled values of WIL 41,42,43 that explore the effects of the sea state(wave period,wave height,wave direction),latitude and time of year on the resulting WIL.The impact of the sea state and latitude is summarized below.Sea state.The combination of wave height and wave period is importa

218、nt for the magnitude of the WIL 41,42,43.Steep waves will induce larger differences in tilt between the modules and hence larger mismatch losses.The shorter the wave period for a specific wave height,the higher the power loss of the system influenced by waves compared with the static system.In addit

219、ion,increasing wave height will also lead to greater WIL.An important conclusion from this is that the effect of WIL on FPV systems deployed at lakes,dams and reservoirs with calm water is small or moderate even for the most affected,one-module-per-float or membrane systems.Latitude and seasonal dep

220、endence.Reflection increases nonlinearly with angle of incidence.Therefore,larger angles of incidence infer both a greater irradiance loss and greater differences between the irradiance absorbed by modules(at a given absolute difference in tilt).FPV designs with non-optimal tilt for a given latitude

221、 will therefore experience higher WIL(everything else being equal).For the typical,relatively horizontal FPV system,this implies that WIL is greater at higher latitudes,and that it has a seasonal dependence,with the highest relative yield losses when the tilt angle is least optimal(i.e.winter for Eu

222、rope)41,43.The simplest approach to model the sensitivity of different parameters(environmental and technology specific)on WIL is to assume that the floats are“slave-to-the-waves”,i.e.that the floats do not dampen the waves,and that each individual float is not constrained in its movements.In Figure

223、 11 and Figure 12,this approach has been taken to model the sensitivity of a one-module-per-float technology,based on Nysted et al.43.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 30 Figure 11.Modelling of th

224、e sensitivity of WIL for a)significant wave height,b)peak period and c)string length for a one-module-per-float type technology,based on 43.Figure 12.Modelling of WIL for different times of year for two latitudes,0N and 50 N for a system with a design tilt of a)0 and b)15,based on the modelling meth

225、od in 43.PurePure-f floatsloats With respect to WIL,the most important adaptation to account for different FPV technologies will be in the wave-float interaction.The mismatch model published by Drenkmper et al.41 represents a floating structure that resembles the pure-floats.The tilt angle is 12,and

226、 the float is according to the paper“in-line with commercial products of single floats on the market today”.Two different movement models were used,the first assumes that the PV modules perfectly follow the waves and that the floats do not dampen the waves.These are the same basic assumptions as in

227、the paper by Nysted et al.43.The other movement model is a mechanical simulation of movements of interconnected floats on the water surface.The model accounts for gravity,buoyancy,wave forces,interconnection forces and inertia.A comparison of both models confirms the expectation that with the mechan

228、ical simulation,the movements are dampened compared to the wave-following model,and therefore the calculated wave-induced losses are lower.Metal or FRP structures+floats/pipesMetal or FRP structures+floats/pipes There is currently no published literature reporting on modeled or measured WIL for a me

229、tal/FRP type of structure.For this type of structure,the PV modules mounted on the same rigid float will have the same orientation,reducing or eliminating mismatch losses between the Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliabi

230、lity,and Maintenance 31 modules on the same float.The changes in irradiance compared to a fixed-tilt system will also likely be reduced(which may be a loss or a gain).In sum,it is expected that FPV systems with rigid materials(metal or FRP with floats/pipes)will be less affected by WIL than pure-flo

231、ats given the same sea state.Membrane FPV technologyMembrane FPV technology As previously mentioned,in the membrane FPV technology the PV modules are expected to follow the local wave motion,with only limited damping.Therefore,although the impact of the membrane size and the fasting mechanism of the

232、 modules are not known,it can be argued that the slave-to-the-waves approach is a good starting point to model WIL for this FPV Category.Nysted et al.43 use linear irregular waves and a response model where the PV modules are assumed to follow the local wave motions exactly and not dampen the energy

233、 of the incoming waves.Using these assumptions,in addition to the more general modelling steps for mismatch described in section 2.3.2,the dependency of WIL on a combination of significant wave height and peak period is shown in Figure 11.In this section,Figure 11 is used to illustrate general trend

234、s for the sensitivity of WIL to different parameters.The values can also be interpreted as a worst-case scenario for WIL for membrane FPV technology,and one-module-per-float technologies.2.3.3 Soiling losses Although FPV has the potential to provide numerous advantages,it can also lead to new challe

235、nges,such as potentially more severe soiling losses.These losses are produced by particles or objects that accumulate on top of the solar panel,which block the irradiance reaching the solar cells,and thus reducing their power output.Furthermore,if the soiling is concentrated in a particular area of

236、the solar panel,e.g.soiling due to bird droppings,this can produce hotspots,as seen in Figure 13,which can accelerate the panel degradation and result in higher O&M costs.Pre-construction bird surveys may help to identify such possible issue during the project development phase.To reduce the soiling

237、 losses,studies have shown that a higher panel tilt is advantageous 44.FPV systems,however,are typically kept with a tilt below 20,regardless of the geographical location.This typically low tilt is a trade-off between optimizing POA irradiance and keeping capital expenditure(CAPEX)low and power dens

238、ity high.CAPEX will increase with increasing tilt due to increased requirements for the float,mooring and anchoring system,while the power density will reduce due to interrow shading.In addition,numerous FPV systems are installed at locations surrounded by diverse fauna and as a result,bird dropping

239、s can become a challenge due to local(nesting)and migratory birds.Potential solutions to tackle this challenge are bird deterrents such as shiny reflective surfaces,ultrasound devices,lasers and scare techniques like water sprayers,scarecrows and fake falcons.However,the use of bird deterrents shoul

240、d be evaluated with respect to sustainability,and the use of such systems may be regulated.On the other hand,as these systems are not installed on land,they are expected to experience lower soiling from dust in comparison to GPV systems.Moreover,the water required to clean them is directly available

241、(although it needs to be assured that clean fresh water is used).In addition,there may be more water reaching the solar panels surface,e.g.via waves and wind influence,thus potentially reducing soiling losses.If the soiling effect is considerable for a particular project,manual or automatic cleaning

242、 using robots might have to be done periodically.Nevertheless,as these systems are surrounded by Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 32 water,cleaning can be challenging and thus,the FPV system need

243、s to be properly designed to allow for these O&M tasks to take place in an adequate way.Too little open information is currently available to establish a span of expected soiling rates for various FPV technologies in different climates.Expected FPV soiling losses of 1-3%has been reported by 7,while

244、it is also noted that this depend on the site and cleaning schedule.Scientific reports on soiling rates for FPV systems would be of importance to improve accuracy of EYAs.Figure 13:Higher local temperature due to hotspot induced by bird dropping.2.4 Modelling yield for FPV systems There is a range o

245、f tools capable of modelling,with different degrees of accuracy,the energy yield for GPV.The list includes,but is not limited to,PVsyst,SAM,PV*Sol,Homer,PVGIS,PVWatts,PVCase,and pvlib.To our knowledge,none of these modelling tools have implemented features to deal specifically with the loss factors

246、of FPV.One could also imagine that other FPV related features such as modelling of evaporation or irradiance reaching the water surface could be included in the same modelling tool.This,however,is not within the scope of this report to discuss.In the following section,we will focus on the modelling

247、capabilities and shortcomings of two important modelling tools,pvlib and PVsyst,with respect to FPV losses and yield.pvlib and PVsyst are chosen because they are widely used for research and development of PV projects.Modelling the performance of any PV system follows certain predefined steps regard

248、less of what software is used.These steps include calculation of:POA irradiance,effective irradiance,module(and cell)temperature,array IV curve(or Pmp),and inverter efficiency.For an FPV system,some of these steps need to be adjusted to account for system and site-specific conditions.For example,FPV

249、 arrays are not fixed and can move as waves pass through the array,therefore,calculation of POA irradiance needs to account for a moving array and may require wave conditions as input.Another example is module and cell temperature modelling.As discussed in Section 2.3.1,FPV systems are affected by i

250、rradiance,air temperature and wind speed,as are terrestrial systems.However,the effects of water temperature,relative humidity and direct splashing of water onto the modules may also need to be considered.Furthermore,FPV systems where the modules are mounted on a membrane that floats directly on the

251、 water will have a much different thermal signature than modules suspended over the water surface.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 33 2.4.1 FPV modelling in PVsyst The solar industrys most widely

252、 used software simulation tool for assessment of a PV systems bankability is PVsyst.This is in part due to the actual modelling capabilities,interface and databases of PVsyst,and in part due to the relatively long trustworthy history of the software.Temperature Temperature modellingmodelling in in P

253、VsystPVsyst Numerous models have been proposed for simulation of the module temperature.In PVsyst,module temperature is estimated using the Faiman model(Eq.(11)33.=+(1 )+(11)where Ta is the ambient temperature,is the plane-of-array irradiance,is the absorption coefficient,is the electrical efficienc

254、y of the module and is the wind speed.and are the constant and convective heat transfer components,with unit W/(m2K)and W/(m2Km/s),respectively.Often,a simplified version of the equation with a lumped heat loss factor,is used:=+.Note that IEC 61853-2 34 also recommends the Faiman model for yield ass

255、essment.PVsyst recommends U-values of =29 W/(m2K),=0 W/(m2Km/s)for free standing(open-rack)systems.For modules with a fully insulated back side,PVsyst recommends a value of =15 W/(m2K),=0 W/(m2Km/s)with the argument that only the front side contributes to the convective heat exchange.The default val

256、ue proposed by PVsyst for new projects lies between these two,=20 W/(m2K),=0 W/(m2Km/s),which is considered representative for typical rooftop systems.For utility scale PV plants,single U-values are also commonly used as input,and wind data is not taken into account.Fundamentally,the Faiman equation

257、 will be adequate to model temperatures of FPV systems when the main heat exchange mechanism is convective heat transfer to the ambient air,as it is for GPV.For membrane floats(Category 3 in Figure 5)given that they are in thermal contact with the membrane(negligible air gaps),and other FPV technolo

258、gies where the PV panels are in contact with water,the modelling of heat transfer should consider that the ambient on the rear side of the module effectively is water.This results in a new expression for the heat balance,and hence an altered equation to calculate cell temperature 31,39.PVsyst does n

259、ot provide the possibility to change the model for the cell temperature calculations,and hence cannot accurately model the temperature of this type of FPV technology.If PVsyst,or similar,is used for EYA of this type of FPV,it is shown that using water temperature as the ambient temperature will impr

260、ove the module temperature model 31.Mismatch Mismatch modellingmodelling in in PVsystPVsyst In PVsyst,mismatch is treated as a constant loss factor,valid for the whole simulation.Hence,it is not a detailed model of the mismatch under the simulated conditions,simply an estimate of the total effect of

261、 non-identical module and string parameters.The value will usually be dominated by the dispersion of the module efficiencies(power class for new modules,power class+degradation for older modules).According to PVsyst,the mismatch value for GPV systems is usually set to 2%.The origin of the mismatch f

262、or FPV is discussed in Section 2.3.2.As there is no actual modelling of mismatch performed in PVsyst,such modelling must be performed outside of PVsyst,and the aggregated result must subsequently be input to the PVsyst model.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Pow

263、er Plants:A Review of Energy Yield,Reliability,and Maintenance 34 2.4.2 FPV modelling in pvlib pvlib python is a community-developed open source toolbox for simulating the performance of PV systems 45.The python library contains more than 100 PV performance modelling functions,which aligns with the

264、core mission of providing open,reliable,and reference implementations of PV system models.The functions cover all the modelling steps for estimating energy yield.The current version of pvlib(v0.11.2)does not feature major FPV-specific capabilities,however,due to pvlibs flexible design,it is possible

265、 to account for many of the modelling aspects specific to FPV.Temperature Temperature modellingmodelling in pvlibin pvlib pvlib python supports several PV temperature models developed for GPV,including Fuentes,SAPM,NOCT SAM,Faiman,PVsyst,Ross,and a generic linear heat loss factor model.As reported i

266、n Section 2.3.1,most FPV systems are predominantly air-cooled with a thermal behaviour similar to GPV systems.For such systems,the existing models in pvlib are straightforward to use for modelling PV module and cell temperature.Similar to PVsyst,it is possible to use water temperature instead of air

267、 temperature as the heat sink temperature used by the aforementioned temperature models.In any case,appropriate heat loss coefficients should be used,as discussed in Section 2.3.1.The pvlib documentation contains an example of how to calculate FPV module temperature.Mismatch Mismatch modellingmodell

268、ing in pvlibin pvlib Although constant loss factors are easily applied in pvlib as well,pvlib also offers functionality needed for modelling electrical mismatch in full fidelity.This is made possible by two main capabilities:no limitation on the maximum number of module orientations,and no fixed or

269、minimum simulation time interval.This is particularly beneficial for FPV systems,as variation in module orientation caused by waves occurs on a short time scale on the order of seconds 43.Given module orientation values from an external wave model,pvlib can simulate the incident irradiance and tempe

270、rature of each module individually.These module-specific operating conditions can then be passed into pvlibs electrical functions(e.g.the CEC and PVsyst single-diode models)to simulate I-V curves for each module.The user can then process these I-V curves using general-purpose numerical python packag

271、es to combine the module-level I-V curves into string-and array-level curves,allowing calculation of the wave-induced electrical mismatch loss.Alternatively,streamlined simulation of array-level I-V curves is possible using SunPowers Python package PVMismatch 46.The string-or array-level curves can

272、then be passed into any of pvlibs existing inverter models.Note that some of pvlibs inverter models are capable of accounting for multiple maximum power point tracking(MPPT)inputs,a required consideration when simulating multi-MPPT inverters connected to mismatched strings.Possibilities and shortcom

273、ingsPossibilities and shortcomings As with existing commercial PV simulation software,pvlib lacks dedicated capabilities for modelling FPV systems.To achieve more accurate yield simulations,FPV-specific temperature models should be added,particularly temperature models which are capable of accountin

274、g for heat transfer to both air and water.Auxiliary functions for determining the inputs to these FPV models,such as module orientation variation due to wave action and wind speed adjustments,may also be added.Information on water albedo was added in the latest release of the software.Task 13 Reliab

275、ility and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 35 In terms of mismatch modelling,it should be noted that PVMismatch is not actively developed/maintained.Additionally,its simulation functionality relies on a particular elect

276、rical model for which PV module parameters are not readily available.For these reasons,it is desirable that pvlibs own electrical simulation capabilities be extended to facilitate the process of combining mismatched I-V curves.2.5 Uncertainty analysis Uncertainty in annual EYA arises from two main c

277、ategories of uncertainty:random(aleatory)uncertainty and lack of knowledge(epistemic)uncertainty.On the one hand,random uncertainty is inherent in annual energy estimates,with the inter-annual variability of the solar resource being one of the largest sources of uncertainty in annual energy modellin

278、g 47.The latter is for example quantified in 47 as the coefficient of variation of the annual GHI,and given a range of 2%to 6%.On the other hand,epistemic uncertainty includes uncertainty in all data measurements,modelling parameters,and models that are imperfect representations of the physical syst

279、em.In theory,this uncertainty,could be reduced through improved data measurement and accurate performance models 48.Literature mainly considers uncertainty propagation in models of generic PV systems 48,49,50,51.This usually entails identifying the main loss factors in the PV model chain,together wi

280、th their underlying uncertainty-either through measurement or educated guesses.The first steps are often to distinguish the involved type(s)of uncertainty,the affected model input(parameter or variable)and settle on an uncertainty measure,such as standard deviation.With that at hand,mainly three app

281、roaches are commonly pursued,and combined if needed:i)Independent input variables.For simple models,such as linear combinations of input variables or products of independent input variables,uncertainty propagation from input to output can be done analytically by calculating variances.Even though the

282、re are interdependencies in the PV modelling process,standard literature 48,50 often represent the annual PV yield as a product of independent loss factors to be able to utilize this relatively simple,analytical model to calculate the output averages and variances 48.However,it shifts the challenge

283、towards interpreting and determining the uncertainties of the loss factors.ii)Small and uncorrelated input uncertainties.The so-called Gaussian law of error propagation relates the variances of a models input and output when the model can be represented by an analytical equation.It simplifies uncert

284、ainty propagation by assuming small and uncorrelated input uncertainties2.Additionally,this approach provides a more accurate calculation of average output values for nonlinear models,where the average output generally differs from the result obtained by merely averaging the inputs.iii)Correlated in

285、put variables.One can use Monte-Carlo simulations that stochastically sample input uncertainty to calculate uncertainty in important output variables such as the annual yield.The advantage of this strategy is that it can trace uncertainty propagation along the full PV simulation chain,and in doing s

286、o can also account for correlated input variables.This,however,comes at a cost,most notably the need for a lot of sampling of input statistics to obtain reliable output statistics.Also,it is difficult to draw analytic conclusions from computed output statistics.To address the uncertainty with respec

287、t to the selection of competing component models in the PV chain(accounting for the variability in predictions of,e.g.,the Faiman and Sandia 2 However,input covariances can easily be incorporated.Note also that,despite the name,Gaussianity of input probability distributions is no prerequisite.Task 1

288、3 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 36 module temperature model),an ensemble of model chains can be utilized,with each ensemble member being a sequence of different component models.This so called“poor ma

289、ns ensemble”approach can also mimic epistemic uncertainties by assigning respective standard values to the parameters of each considered component model.To date,no published efforts are known for quantifying uncertainties in crucial FPV modelling steps such as soiling,fluctuating POA irradiance or m

290、ismatch losses.For best results,modelers are advised to combine educated guesses with a critical reading of existing literature on uncertainty propagation.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 37 RELI

291、ABILITY OF FLOATING PV The economic viability of PV power plants is fundamentally linked to their lifetime energy yield.Factors such as degradation effects and the overall lifespan of the power plant directly impact electricity production and the levelized cost of energy(LCOE),consequently influenci

292、ng profitability 45,p.11.In Section 3.1,we begin by defining degradation and exploring methods to quantify it.Conducting an effective reliability analysis is essential for minimizing failures and ensuring a stable power supply 52.However,while evaluating the reliability of an FPV system,several sign

293、ificant knowledge gaps and challenges arise,which will be addressed in this chapter.Firstly,climatic and environmental factors play a major role in degradation and are by nature location specific.The stress profiles experienced by components in an FPV installation are neither well understood nor qua

294、ntified and will vary a lot depending on float technology and water body conditions.An overview of what is currently known regarding climatic stressors for FPV is provided in Section 3.2.The second knowledge gap relates to the scarcity of information on and systematic studies of observed field failu

295、res as well as of degradation/performance loss rates and is addressed in Section 3.3.1.And third,as a result of the first two points,there is no accelerated stress testing protocol developed for component reliability evaluation,complicating the process of gaining relevant insight from indoor testing

296、.Existing knowledge and current efforts are summarized in Section 3.3.2.A final source of insight into FPV degradation effects is simulations,which will be discussed in Section 3.3.4.The correlations between the main topics of the Chapter are illustrated in Figure 14.Figure 14.Figure adapted and rep

297、rinted under under a CC BY 4.0 license from 53.Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 38 3.1 Defining degradation 3.1.1 Definitions The term degradation denotes the gradual process of change in charact

298、eristics with operational time of a material,component,and/or system triggered by stress impact.Typically for PV,this aging process causes a decrease in performance,and hence a power loss 54,p.15.Degradation is caused by stressors,such as physical,chemical,or mechanical stress acting on a material,c

299、omponent or system.Examples include temperature,irradiation ultraviolet(UV),visible(VIS),near infrared(NIR),water/moisture,electrical potential as well as mechanical stresses such as compressive or tensile impact.The sum of the local stressors that a PV module experiences during operation is specifi

300、c to its design and exact location and surrounding,e.g.,stress induced from mounting or variations in temperature and shading.The microclimate can be inhomogeneous even within a PV module(different humidity or cell temperature)54,pp.16,17.Typically,three different kinds of degradation are distinguis

301、hed:reversible degradation,irreversible degradation,and failures.Reversible degradation is either related to accumulated soiling,including the growth of algae and loot,which will not be removed by natural rainfall,but may be removed with dedicated cleaning actions.Other effects of reversible degrada

302、tion include electric phenomena like polarization(potential induced degradation,PID)or light and elevated temperature induced degradation(LeTiD),from which PV modules may(partly)recover naturally,e.g.during nighttime,or driven by specific electrical devices during nighttime.Irreversible degradation

303、is mostly related to material ageing both within the solar cell and with the embedding materials.Solar cells may degrade by a variety of diffusion and corrosion processes,which again may be related to changing properties of the embedding materials.Moisture ingress through the back sheet or the edge

304、seal may increase with time,and embedding polymers may change their transparency with time.Mechanical stress may lead to an increasing number of micro-cracks in solar cells which can lead to a reduced power output.Based on IEC 60050-191 55,the definition of failure is the termination of the ability

305、of an item to perform a required function.For a PV module,this means that the module needs to be replaced.While that is relatively clear from a safety perspective,it is less so from a performance perspective.However,a performance of 80%of the initial value is often used as a threshold for failure.Fa

306、ilures are commonly subdivided into early life(1-2 years)failures,often related to poor design or manufacturing errors,failures during the steady-state life,often either random or the result of technology limitations,and wear-out failures,caused by mechanisms that degrade the performance gradually u

307、ntil the device does not function anymore.Ideally,wear-out failure only occurs after the warranty period expires 56.Regarding FPV,the balance of system(BOS)components may be even more critical than the PV modules.Junction boxes,cables,connectors,and related protecting materials may suffer from addit

308、ional stress compared to GPV systems,while additionally,there are FPV-specific BOS components like anchors and floats whose reliability also needs to be considered.3.1.2 Quantification of degradation PV module and system degradation may be detected in several ways.Some of them are preventive and sho

309、uld be part of standard O&M procedures(see Chapter 4),others are retrospective and related to the fulfilment of contractual obligations.Standard evaluations rely Task 13 Reliability and Performance of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance

310、39 on yield data and meteorological data only.More advanced assessments look for specific data signatures and are combined with onsite visual and infrared inspections and sometimes even with lab testing of selected components.As a typical degradation measure,the performance loss rate(PLR)is often us

311、ed.It is stated in percentage per year and quantifies the temporal decline of a PV systems power output in relation to its irradiation input.Accurate long-term PLR information is vital for predicting electricity output,improving system reliability and anticipating maintenance requirements.On the eco

312、nomic side,PLR has been shown to be one of the most influential factors of LCOE.Beyond the irreversible physical degradation of PV modules,the PLR also captures performance-reducing events,which may be reversible or even preventable through good operations and maintenance(O&M)practices 57,58,59,60.3

313、.1.3 Distinction of degradation effects As mentioned above,observed degradation effects may be divided into component failures,reversible degradation,and irreversible degradation.To evaluate irreversible degradation in a correct manner,most or all reversible effects must be known(and ideally repaire

314、d or removed).In the process,also shading losses must be accounted for,which may increase with time when caused by vegetation,decreasing the overall system yield.As degradation assessments typically deal with long-term observations,i.e.large sets of operational data,methods of automated failure dete

315、ction may play an important role.These algorithms,often run concurrently,are generally grounded in expert knowledge,assisted by statistical and machine learning methods,validated with field data,and designed to be flexible and adaptable to the characteristics of PV systems and O&M service providers.

316、They can include communication faults,inverter outage and inverter late wake up,check on open-circuit conditions,several relative comparisons between individual strings and inverters,and all kinds of sudden changes in performance indicators 61.Beside the distinction of“real”degradation from failures

317、 or reversible effects,further(algorithm-driven)differentiation between degradation mechanisms is desirable.This,however,needs further research,as“real”degradation is often related with very small changes of performance with time,and as new cell and module technologies may even show-previously unkno

318、wn degradation modes.3.2 Driving degradation describing FPV specific stressors A PV system will experience a combined effect of multiple stressors over its lifetime in the outdoor environment.Knowledge about these stressors is a precondition for the creation of meaningful predictions of degradation

319、and service life.The stressors include temperature,humidity,UV radiation,wave and wind loads,tidal variations,temperature fluctuations 62,63,high voltages,corrosive compounds,soiling,abrasive loads,shading,and flora&fauna.The stressors are design-and water body-dependent and,when co-occurring,can at

320、tenuate or have amplified effects.When comparing FPV to conventional GPV,one might intuitively expect some of these stressors(e.g.,humidity and mechanical loads)to have more significant differences in stress profiles than others(e.g.,UV radiation).For example,the design and placement of a mooring sy

321、stem for FPV platforms must account for water-level variations,soil conditions,bathymetry,and location.In deep water,the construction of a mooring system can consequently present significant challenges and incur substantial costs 64,65.For components common to both Task 13 Reliability and Performanc

322、e of Photovoltaic Systems Floating PV Power Plants:A Review of Energy Yield,Reliability,and Maintenance 40 GPV and FPV systems,there is however little openly available quantitative information on FPV specific stress profiles.If available,such stress profiles could be compared with available stress p

323、rofiles from GPV environments,to assess to which extent,e.g.,qualification testing and module design considerations originating from the conventional GPV industry are applicable or in need of modification.BiofoulingBiofouling and soilingand soiling Biofouling is an often discussed problem for FPV in

324、stallations 66 and can act as a stressor by affecting corrosion processes or inducing weight and stability issues for the floating structure.Biofouling removal will lead to increased O&M costs but can also in itself act as a stressor by increasing wear on various components.Soiling can block incomin

325、g light and thus decrease the performance of PV modules 67.FPV installations have been seen to attract birdlife 37,and bird droppings impact the short-term performance of FPV systems more than for GPV 68 and lead to hotspots.Bird droppings is an example of FPV stress profiles having a major dependen

326、ce on both float technology(e.g.,plastic pontoon versus membrane versus metal platforms,see Figure 5)and water body(e.g.,of different Sea State Codes as discussed in the Introduction).Different float technologies will also yield significant differences in,e.g.,wave-induced mechanical loads or the ex

327、posure to water for various components.While stress profiles on calm inland water bodies will likely not be too different from a conventional GPV system,the differences are expected to increase when going to larger water bodies and eventually to near-shore or offshore conditions.In case of high wind

328、s and wave forces,the floating structure may experience drifting or deformation,and materials may leach into the water because of damage 69.Waves can also affect the electrical components of FPV systems,leading to disconnections or damage,which impacts the systems electrical integrity and performanc

329、e.HumidityHumidity A few reports have compared humidity measurements from meteorological stations at FPV systems on inland and nearshore water bodies and found modest differences in average relative humidity;on the order of 0-10 percentage points higher relative humidity for average daily values 37,

330、70,71.However,these results are hard to generalize.Further,relative humidity values aggregated to a daily timescale carry limited relevant information in this context,as the coupling of humidity and temperature is of central importance for moisture ingress and related degradation phenomena.In additi

331、on to the sources of mechanical loads present for GPV,FPV has waves as an additional factor,including potential wave-slamming and-breaking.An FPV system typically involves more complex and dynamic structures than a GPV system to provide mechanical support for electrical components,making it essentia

332、l to assess factors such as fatigue.Some reports exist in the literature examining FPV specific mechanical loads,so far mostly focused on the mechanical integrity of the float system 72,and less on the PV module and other electrical components.Lower module operating temperatures for FPV have been wi

333、dely discussed in the context of increased module efficiency(Section 2.3.1),but also has the potential to significantly influence reliability.Most degradation processes are thermally activated;thus,a lower operating temperature will dampen the effect of other stressors such as UV and humidity.How significant such an effect might be has so far not been properly studied.SaltSalt When deployed in nea