Where Sun Meets Water FLOATING SOLAR MARKET REPORT This report was researched and prepared by the Solar Energy Research Institute of Singapore (SERIS) at the National University of Singapore (NUS), under contract from the World Bank, with inputs and editing from staff and consultants at the World Bank and the International Finance Corpo- ration (IFC). Authors and contributors to the report were, from SERIS: Thomas Reindl, Celine Paton, Abhishek Kumar, Haohui Liu, Vijay Anand Krishnamurthy, Ji Zhang, Stephen Tay, Yanqin Zhan; from the World Bank: Zuzana Dobrotkova, Sandra Chavez, Chris Jackson, Oliver Knight, Sabine Cornieti, Pierre Audinet, Gailius Draugelis, Surbhi Goyal, Pierre Lorillou; and from IFC: Stratos Tavoulareas, Dzenan Malovic, Hemant Mandal, Jean-Francois Mercier, Ishan Purohit. The work was funded by the Energy Sector Management Assistance Program (ESMAP), The Government of the Kingdom of Denmark as represented by the Ministry of Foreign Affairs, and the World Bank, and also benefited from in-kind contributions from SERIS. © 2019 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW | Washington DC 20433 | USA 202-473-1000 | www.worldbank.org This work is a product of the staff of the World Bank with external contributions. The findings, interpre- tations, and conclusions expressed in this work do not necessarily reflect the views of the World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guar- antee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. RIGHTS AND PERMISSIONS The material in this work is subject to copyright. Because the World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; pubrights@worldbank.org. ESMAP would appreciate a copy of or link to the publication that uses this publication for its source, addressed to ESMAP Manager, World Bank, 1818 H Street NW, Washington, DC, 20433 USA; esmap@worldbank.org. All images remain the sole property of their source and may not be used for any purpose without written permission from the source. Attribution—Please cite the work as follows: World Bank Group, ESMAP and SERIS. 2019. Where Sun Meets Water: Floating Solar Market Report. Washington, DC: World Bank. Front Cover: © SERIS Back Cover: © Pixbee/EDP S.A. Where Sun Meets Water FLOATING SOLAR MARKET REPORT EXECUTIVE SUMMARY Energy Sector Management Assistance Program (ESMAP) The Energy Sector Management Assistance Program (ESMAP) is a global knowledge and technical assistance program administered by the World Bank. ESMAP assists low- and middle-income countries to increase their know-how and institutional capacity to achieve environmentally sustainable energy solutions for poverty reduction and economic growth. ESMAP is funded by Australia, Austria, Canada, Denmark, the European Commission, Fin- land, France, Germany, Iceland, Italy, Japan, Lithuania, Luxemburg, the Netherlands, Norway, the Rockefeller Foundation, Sweden, Switzerland, the United Kingdom, and the World Bank. Solar Energy Research Institute of Singapore (SERIS) The Solar Energy Research Institute of Singapore (SERIS) at the National University of Singa- pore, founded in 2008, is Singapore’s national institute for applied solar energy research. SERIS is supported by the National University of Singapore, National Research Foundation (NRF) and the Singapore Economic Development Board. It has the stature of an NUS Univer- sity-level Research Institute and is endowed with considerable autonomy and flexibility, including an industry friendly intellectual property policy. SERIS’ multi-disciplinary research team includes more than 160 scientists, engineers, techni- cians and PhD students working in R&D clusters including (i) solar cells development and simulation; (ii) PV modules development, testing, certification, characterization and simula- tion; (iii) PV systems, system technologies, including floating PV, and PV grid integration. SERIS is ISO 9001 & ISO 17025 certified. SERIS has extensive rich knowledge and experience with floating PV systems, including having designed and operating the world’s largest floating PV testbed in Tengeh Reservoir, Singapore, which was commissioned by PUB, Singapore’s National Water Agency, and the Economic Development Board. Launched in October 2016, this testbed compares side by side various leading floating PV solutions from around the world. Through detailed monitoring and in-depth analysis of performance of all the systems, SERIS accumulated deep insight into floating solar and SERIS’ objective is to disseminate the best practices in installation and operation of floating solar pants as well as help to formulate standards for floating PV. CONTENTS E EXECUTIVE SUMMARY  1 1 WHY FLOATING SOLAR?  15 1.1. The benefits of floating solar  18 1.2. The effects of floating installations on water bodies  20 1.3. Technological advantages of floating solar  21 1.4. Challenges  23 1.5. Comparison with ground-mounted systems  25 2 TECHNOLOGY OVERVIEW  31 2.1 Key components and system designs  32 2.2 Novel FPV concepts   41 2.3 Hybrid operation with hydropower plants  48 3 GLOBAL MARKET AND POTENTIAL  55 3.1 Availability of floating solar resource  55 3.2 Current market status  58 4 POLICY CONSIDERATIONS AND PROJECT STRUCTURING  81 4.1 Financial incentives and support mechanisms in selected countries  81 4.2 Supportive governmental policies  82 4.3 Other policy and regulatory considerations  84 4.4 Business models and project structuring  85 5 COSTS OF FLOATING SOLAR  91 5.1 Recent disclosed FPV costs  91 5.2 Calculating the levelized cost of electricity  99 5.3 Sensitivity analysis  103 5.4 Risk assessment  104 5.5 Conclusion 104 6 SUPPLIERS OF FLOATING PV SYSTEMS  109 6.1 General overview  109 6.2 Providers of floating technology solutions for inland freshwater applications  110 6.3 Providers of floating technology solutions for offshore or near-shore applications  120 EXECUTIVE SUMMARY  • iii ACRONYMS AC alternating current ADB Asian Development Bank AGC automatic generation control BBC British Broadcasting Corporation CAPEX capital expenditure C&T Ciel & Terre International CIESIN Columbia University Center for International Earth Science Information Network DC direct current DR discount rate ESMAP Energy Sector Management Assistance Program EPCI equity project cost investment EJ exajoules FiT feed-in tariff FPV floating PV GIS geographic information system GWp gigawatt-peak GHI global horizontal irradiance GRanD Global Reservoir and Dam Database GWSP Global Water System Project HDPE high-density polyethylene IC insurance cost IEA International Energy Agency IFC International Finance Corporation IRENA International Renewable Energy Agency IEI inverter warranty extension investment km kilometers kV kilovolt kWh kilowatt-hour kWp kilowatt-peak LCOE levelized cost of electricity LP loan payments LSIS LS Industrial Systems MWh megawatt-hours MWp megawatt-peak NHI Natural Heritage Institute TU Delft Netherlands’ Delft University of Technology iv •  FLOATING SOLAR MARKET REPORT O&M operation and maintenance PERC passivated emitter rear cell PR performance ratio PV photovoltaic REC renewable energy certificate SEAC Solar Energy Application Center SECI Solar Energy Corporation of India SERIS Solar Energy Research Institute of Singapore SMART Solar Massachusetts Renewable Target km2 square kilometers SMCC Sumitomo Mitsui Construction Co., Ltd. SCADA supervisory control and data acquisition SDR system degradation rate MOIT Vietnam’s Ministry of Industry and Trade Wp/m 2 watt peak per square meter WACC weighted average cost of capital WEC World Energy Council All dollar figures denote U.S. dollars unless otherwise noted ACRONYMS   • v CHINA © Sungrow vi •  FLOATING SOLAR MARKET REPORT EXECUTIVE SUMMARY FLOATING SOLAR MARKET REPORT Why floating solar? Other potential advantages of FPV include: Floating solar photovoltaic (FPV) installations open • Reduced evaporation from water reservoirs, as the up new opportunities for scaling up solar generating solar panels provide shade and limit the evapora- capacity, especially in countries with high population tive effects of wind density and competing uses for available land. They • Improvements in water quality, through decreased have certain advantages over land-based systems, algae growth including utilization of existing electricity transmission infrastructure at hydropower sites, close proximity to • Reduction or elimination of the shading of panels demand centers (in the case of water supply reser- by their surroundings voirs), and improved energy yield thanks to the cooling • Elimination of the need for major site preparation, effects of water and the decreased presence of dust. such as leveling or the laying of foundations, which The exact magnitude of these performance advantag- must be done for land-based installations es has yet to be confirmed by larger installations, across multiple geographies, and over time, but in many cases • Easy installation and deployment in sites with low they may outweigh any increase in capital cost. anchoring and mooring requirements, with a high degree of modularity, leading to faster installations. The possibility of adding FPV capacity to existing hydropower plants is of particular interest, especial- An overview of floating solar ly in the case of large hydropower sites that can be technology flexibly operated. The solar capacity can be used to The general layout of an FPV system is similar to that boost the energy yield of such assets and may also of a land-based PV system, other than the fact that help to manage periods of low water availability by the PV arrays and often the inverters are mounted on allowing the hydropower plant to operate in “peak- a floating platform (figure E.1). The direct current (DC) ing” rather than “baseload” mode. And the benefits electricity generated by PV modules is gathered by go both ways: hydropower can smooth variable solar combiner boxes and converted to alternating current output by operating in a “load-following” mode. Float- (AC) by inverters. For small-scale floating plants close ing solar may therefore be of particular interest where to shore, it is possible to place the inverters on land— grids are weak, such as in Sub-Saharan Africa and that is, just a short distance from the array. Otherwise, parts of developing Asia. both central or string inverters on specially designed floats are typically used. The platform, together with its anchoring and mooring system, is an integral part of any FPV installation. EXECUTIVE SUMMARY  • 1 1 FIGURE E.1  Schematic representation of a typical large-scale FPV system with its key components Transmission Central Lightning protection inverter (from other arrays) PV modules system (connected to metal frames supporting modules and grounded) Floats/pontoons Transformer Combiner box Mooring lines Anchoring Source: Solar Energy Research Institute of Singapore (SERIS) at the National University of Singapore. Currently most large-scale FPV plants are deployed and depth), soil conditions, and variation in water lev- using pontoon-type floats, with PV panels mounted at el. Bank anchoring is particularly suitable for small a fixed tilt angle. Typically, the floating structure can be and shallow ponds, but most floating installations are made of so-called pure floats or floats that are com- anchored to the bottom. Regardless of the method, bined with metal trusses (figure E.2). A pure float con- the anchor needs to be designed so as to keep the figuration uses specially designed self-buoyant bodies installation in place for 25 years or more. Mooring to which PV panels can be directly affixed. This config- lines need to be properly selected to accommodate uration is the most common. It is available from several ambient stresses and variations in water level. suppliers and has an installed capacity worldwide of several hundred megawatts. Another type of design The current global market for uses metal structures to support PV panels in a man- floating solar ner similar to land-based systems. These structures are fixed to pontoons whose only function is to provide The first FPV system was built in 2007 in Aichi, buoyancy. In this case, there is no need for specially Japan, followed by several other countries, includ- designed floats. The floating platform is held in place ing France, Italy, the Republic of Korea, Spain, and by an anchoring and mooring system, the design of the United States, all of which have tested small- which depends on factors such as wind load, float scale systems for research and demonstration pur- type, water depth, and variability in the water level. poses. The first commercial installation was a 175 kWp system built at the Far Niente Winery in Cali- The floating platform can generally be anchored to fornia in 2008. The system was floated atop a water a bank, to the bottom, to piles, or to a combination reservoir to avoid occupying land better used for of the three. The developer selects a design suitable growing grapes. to the platform’s location, bathymetry (water profile 2 •  FLOATING SOLAR MARKET REPORT Medium-to-large floating installations (larger than 1 Netherlands, Norway, Panama, Portugal, Singapore, MWp) began to emerge in 2013. After an initial wave of Spain, Sweden, Sri Lanka, Switzerland, Thailand, Tunisia, deployment concentrated in Japan, Korea, and the Turkey, the United Kingdom, and Vietnam, among others. United States, the FPV market spread to China (now Projects are under consideration or development in the largest player), Australia, Brazil, Canada, France, Afghanistan, Azerbaijan, Colombia, Ghana, the Kyrgyz India, Indonesia, Israel, Italy, Malaysia, Maldives, the Republic, Myanmar, and Pakistan, among others. FIGURE E.2   The most common float types: pure float, Indonesia (top) and pontoons with metal structures, India (bottom) Source: © Ciel & Terre International. Source: © NB Institute for Rural Technology. EXECUTIVE SUMMARY  • 3 Recently, plants with capacity of tens and even hun- and annual new additions are growing exponentially dreds of megawatts have been installed in China; (figure E.3). more are planned in India and Southeast Asia. The first plant larger than 10 MWp was installed in 2016, As of December 2018, the cumulative installed capaci- and in 2018 the world saw the first of several plants ty of floating solar was about 1.3 gigawatt-peak (GWp), larger than 100 MWp, the largest of which is 150 MWp. the same milestone that ground-mounted PV reached Flooded mining sites in China support most of the in the year 2000. If the evolution of land-based PV is largest installations (box E.1). With the emergence of any indication, floating solar could advance at least these new markets, cumulative installed FPV capacity as rapidly, profiting from all the decreases in costs FIGURE E.3  Global installed floating PV capacity and annual additions 1,400 1,314 1,200 1,000 786 800 MWp 528 600 400 359 200 169 68 101 0 0.5 1 1.5 2.2 3.4 5.7 11 0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Annual installed FPV capacity Cumulative installed FPV capacity Source: Authors’ compilation based on media releases and industry information. BOX E.1 China’s collapsed coal mines turned into a solar opportunity There are dozens of flooded coal mines in China. as solar panel assemblers and maintenance person- Spurred by China’s “Top Runner” program, solar nel. They are earning better wages and are no longer developers are turning these environmental and exposed to harmful mine conditions known to cause social challenges into an opportunity. Anhui Province lung disease. is home to the world’s largest floating solar installa- Producing solar power in mining regions while scal- tions to date, ranging from 20 megawatts (MW) to ing back coal-based power production is one way to 150 MW per site. improve local air pollution in several regions of China. Local people who just a few years ago worked underground as coal miners are now being retrained Source: Authors’ compilation based on Mason (2018) and BBC (2018). 4 •  FLOATING SOLAR MARKET REPORT attained by land-based PV deployment. Most of the combined power output, have started to appear (box installations to-date are based on industrial basins, E.2). In these installations, special attention needs to drinking water reservoirs, or irrigation ponds (figure be paid to possible effects on the downstream flow E.4), but the first combinations with hydropower res- regime from the reservoir, which is typically subject ervoirs, which bring the added benefits of better utili- to restrictions related to water management (in case zation of the existing transmission infrastructure and of cascading dams), agriculture, biodiversity, naviga- the opportunity to manage the solar variability through tion, and livelihood or recreational uses. FIGURE E.4  Floating solar installations in Malaysia (top), and Japan (bottom) Source: © Ciel & Terre International. Source: © Ciel & Terre International. EXECUTIVE SUMMARY  • 5 BOX E.2 Hydropower-connected solar PV systems The development of grid-connected hybrid systems The PV power plant is directly connected through that combine hydropower and floating photovoltaic a reserved 330 kilovolt (kV) transmission line to the (PV) technologies is still at an early stage. Only a small Longyangxia hydropower substation. The hybrid sys- system of 220 kilowatt-peak (kWp) has been deployed tem is operated so that the energy generation of the in Portugal (see photo) (Trapani and Santafé 2015). But hydro and PV components complement each other many projects, and of much greater magnitudes, are (Choi and Lee 2013). After the PV plant was added, being discussed or developed across the world. the grid operator began to issue a higher power dis- The largest hybrid hydro-PV system involves ground- patch set point during the day. As expected, on a mounted solar PV. This is the Longyangxia hydro- typical day the output from the hydro facility is now connected PV power plant in Qinghai, China (Qi 2014), reduced, especially from 11 a.m. to 4 p.m., when PV which is striking for its sheer magnitude and may be generation is high. The saved energy is then request- considered a role model for future hybrid systems, both ed by the operator to be used during early morning floating and land-based. and late-night hours. Although the daily generation The Longyangxia hydropower plant was commis- pattern of the hydropower has changed, the daily sioned in 1989, with four turbines of 320 megawatts reservoir water balance could be maintained at the (MW) each, or 1,280 MW in total. It serves as the major same level as before to also meet the water require- load peaking and frequency regulation power plant in ments of other downstream reservoirs. All power gen- China’s northwest power grid. The associated Gonghe erated by the hybrid system is fully absorbed by the solar plant is 30 kilometers (km) away from the Long- grid, without any curtailment. This system shows that yangxia hydropower plant. Its initial phase was built hydro turbines can provide adequate response as and commissioned in 2013 with a nameplate capacity demand and PV output varies. of 320 megawatt-peak (MWp). An additional 530 MWp Source: Authors’ compilation based on Trapani and Santafé (2015); was completed in 2015. Qi (2014); and Choi and Lee (2013). First-ever hydropower-connected FPV operation, Montalegre, Portugal Source: © Pixbee/EDP S.A. 6 •  FLOATING SOLAR MARKET REPORT Marine installations are also appearing. The deploy- Policy and regulatory ment of FPV technologies near shore may be of strong considerations interest to populous coastal cities. Indeed, it may be the only viable way for small island states to generate Currently, even in countries with significant FPV devel- clean solar power at scale, given the limited availability opment there are no clear, specific regulations on of land suitable for ground-mounted PV installations. permitting and licensing of such plants. Processes for the moment are assumed to be the same as for Still at a nascent stage, near-shore solar PV is concep- ground-mounted PV, but legal interpretation is need- tually similar to deployment on inland water bodies. But ed in each country. In some countries, drinking water the offshore environment poses additional challenges: reservoirs or hydropower reservoirs are considered national-security sites, making permitting more com- • Water surface conditions are much rougher (larger plex and potentially protracted. waves and higher winds) • Mooring and anchoring become even more critical As highlighted in this report, FPV deployment is expect- amid large tidal movements and currents ed to be cost-competitive under many circumstanc- es and therefore not to require financial support. • Salinity tests the durability of components Nevertheless, initial projects may require some form • The accumulation of marine organisms on equip- of support to overcome barriers associated with the ment (“bio fouling”) can interfere with functionality. industry’s relatively limited experience with this tech- nology. The harsher near-shore environment imposes strin- gent requirements on floats, anchors, moorings, and So far, a number of countries have taken different components. Alternative design and technological approaches to FPV. Typical policies currently support- solutions may be required, drawing on the rich experi- ing FPV installations can be grouped into two cate- ence of existing marine and offshore industries. Com- gories: pared to the open sea, coastal areas such as lagoons and bays are relatively calm and thus more suitable Financial incentives: for FPV, however installations must still be able to with- stand waves and high winds. On the other hand, some • Feed-in tariffs that are higher than those for ground- lagoons and bays can be environmentally sensitive, mounted PV (as in Taiwan, China) which may limit the possibility for FPV deployment in • Extra bonuses for renewable energy certificates certain areas. (as in the Republic of Korea) The biggest uncertainties are long-term reliability and • A high feed-in tariff for solar PV generally (as in cost. Marine-grade materials and components are Japan) critical for these installations, which must withstand • Extra “adder” value for FPV generation under the the prevailing environmental conditions. Operation compensation rates of state incentives program and maintenance costs for near-shore PV are also (as in the U.S. state of Massachusetts). expected to be higher than for inland installations. In the Maldives, near-shore solar PV is powering a Supportive governmental policies: tourist resort; in Norway, a large fish farm (figure E.5). • Ambitious renewable energy targets (as in Korea Future systems will likely fulfill needs that are additional and Taiwan, China) to energy production, such as the generation of hydro- gen or the solar-based desalination of water. • Realization of demonstrator plants (as in the Indian state of Kerala) 7 EXECUTIVE SUMMARY  • 7 FIGURE E.5  Near-shore floating installations in the Baa Atoll of the Maldives (left), and off the west coast of Norway, (right) Source: © Swimsol. Source: © Ocean Sun. • Dedicated tendering processes for FPV (as in • Special considerations for hydro-connected plants: Taiwan, China and India) – Whether the hydropower plant owner/operator • Openness on the part of the entities managing the is allowed to add an FPV installation water bodies, such as tenders for water-lease con- – Whether the hydropower plant owner/operator is tracts (as in Korea). allowed to provide a concession to a third party to build, own, and operate an FPV plant However, for most countries hoping to develop a – Management of risks and liabilities related to well-functioning FPV segment as part of their solar PV hydropower plant operation and weather events market development, the following policy and regula- that can affect the solar or hydropower plants tory considerations need to be addressed: – Rules of dispatch coordination of the solar and • Unique aspects of permitting and licensing that the hydropower plants’ outputs. necessitate interagency cooperation between ener- gy and water authorities. This also includes environ- mental impact assessments for FPV installations. Market opportunities There are more than 400,000 square kilometers (km2) • Water rights and permits to install and operate of man-made reservoirs in the world (Shiklomanov an FPV plant on the surface of a water body and 1993), suggesting that FPV has a theoretical potential anchor it in or next to the reservoir. on a terawatt scale, purely from the perspective of the • Tariff setting for FPV installations (which could be available surface area. The most conservative estimate done as for land-based PV, for example, through of FPV’s overall global potential based on available feed-in tariffs for small installations and tenders or man-made water surfaces exceeds 400 GWp, which auctions for large ones). is equal to the 2017 cumulative installed PV capacity globally. Table E.1 provides a summary of the man- • Access to existing transmission infrastructure: made freshwater bodies supporting this very conser- – How will this be managed? vative estimate. Considering global irradiance data on – Who will be responsible? significant water bodies, and assuming 1 percent to – What permits/agreements will be required? 10 percent of their total surface area as used for FPV deployment, an estimate of potential peak capacity 8 •  FLOATING SOLAR MARKET REPORT TABLE E.1 . Peak capacity and energy generation potential of FPV on freshwater man-made reservoirs, by continent Possible annual energy Total Number FPV potential (GWp) generation (GWh/year) surface area of water Percentage of Percentage of available bodies total surface area used total surface area used Continent (km2) assessed 1% 5% 10% 1% 5% 10% Africa 101,130 724 101 506 1,011 167,165 835,824 1,671,648 Middle East and Asia 115,621 2,041 116 578 1,156 128,691 643,456 1,286,911 Europe 20,424 1,082 20 102 204 19,574 97,868 195,736 North America 126,017 2,248 126 630 1,260 140,815 704,076 1,408,153 Australia and Oceania 4,991 254 5 25 50 6,713 33,565 67,131 South America 36,271 299 36 181 363 58,151 290,753 581,507 Total 404,454 6,648 404 2,022 4,044 521,109 2,605,542 5,211,086 Source: SERIS calculations based on the Global Solar Atlas © World Bank Group (2019) and the GRanD database, © Global Water System Project (2011). Note: GWh = gigawatt-hour; GWp = gigawatt-peak; km2 = square kilometers; PV = photovoltaic. was derived using the efficiency levels of currently Costs of floating solar and available PV modules and the surface area needed for project structuring their installation, operation, and maintenance. Then, to estimate potential electricity generation, the capacity Capital costs estimate was multiplied by the expected specific ener- The capital costs of floating PV are still slightly high- gy yield, with local irradiance used alongside a con- er or comparable to those of ground-mounted PV, servative assumption of an 80 percent performance owing chiefly to the need for floats, moorings, and more ratio. These estimates use very low ratio of coverage resilient electrical components. The cost of floats is of the reservoir. In reality, many existing projects imple- expected to drop over time, however, owing to better mented on industrial or irrigation reservoirs cover economies of scale. much more significant portions of the reservoirs, after environmental studies confirm no expected impact on Total capital expenditures for turnkey FPV installations the aquatic life in the reservoirs. The situation from one in 2018 generally range between $0.8–1.2 per Wp (fig- reservoir to another can differ significantly, however. ure E.6), depending on the location of the project, the depth of the water body, variations in that depth, and There are individual dams on each continent that the size of the system. China is the only country that could theoretically accommodate hundreds of mega- has yet built installations of tens to hundreds of mega- watts or, in some cases, gigawatts of FPV installa- watt-peak in size. The costs of smaller systems in other tions. Examples of such reservoirs are provided in regions could vary significantly. table E.2. While hydropower and solar capacity do not provide the same type of power production (solar As reflected in figure E.6, Japan remains a region with typically has a lower capacity factor and generates relatively high system prices, while China and India variable power), the table compares the surface achieve much lower prices, a pattern that can also be needed for a PV plant having the same peak capacity seen in ground-mounted and rooftop solar systems as the hydropower reservoir. when compared to the global average. EXECUTIVE SUMMARY  • 9 TABLE E.2 . Reservoir size and estimated power generation capacity of selected hydropower dams, and potential of FPV to match the dams’ hydropower capacity Percentage of reservoir area required for FPV to match dam’s Dam/reservoir Country Reservoir size (km2) Hydropower (GW) hydropower capacity (%) Bakun Dam Malaysia 690 2.4 3 Lake Volta Ghana 8,500 1.0 <1 Guri Dam Venezuela 4,250 10.2 2 Sobradinho “Lake” Brazil 4,220 1.0 <1 Aswan Dam Egypt 5,000 2.0 <1 Attaturk Lake and Dam Turkey 820 2.4 3 Narmada Dam India 375 1.5 4 Source: Authors’ compilation. Note: GW = gigawatt; km2 = square kilometer; PV = photovoltaic FIGURE E.6  Investment costs of FPV in 2014–2018 (realized and auction results) UK—0.2 MWp Sheeplands (2014) 1.14 Japan—2 MWp Shiroishi Saga (2015) 3.12 Portugal—0.2 MWp EDP Hydro (2016) 2.31 UK—6.3 MWp Queen Elizabeth II (2016) 1.22 China—20 MWp Anhui Xinyi (2016) 1.48 Japan—2.4 MWp Noma Ike (2017) 2.93 China—40 MWp Anhui Sungrow (2017) 1.13 India—0.5 MWp Kerala (2017) 2.84 Japan—1.5 MWp Mita Kannabe (2017) 2.93 Japan—13.7 MWp Yamakura Dam (2018) 0.97 India—2 MWp Andhra Pradesh (2018) 0.92 China—150 MWp Three Gorges (2018) 0.99 India—5 MWp West Bengal Auction Lowest Price (2018) 0.83 India—5 MWp West Bengal Auction Avg Price (2018) 1.14 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.8 1.2 $/Wp Source: Authors’ compilation based on media releases and industry information. Note: Using the 2017 $ annual exchange rates, as released by OECD. PV = photovoltaic; $/Wp = U.S. dollars per watt-peak. Levelized costs of electricity, including a conservative and optimistic scenario. This result sensitivity analysis holds at a range of discount rates, as shown in table Calculated on a pretax basis, the levelized cost of E.3. Both projects have the same theoretical financial electricity (LCOE) for a generic 50 MW FPV sys- assumptions and irradiance. However, the main dif- tem does not differ significantly from that of a ferentiating factors are system price (a floating sys- ground-mounted system. The higher initial capital tem is considered 18 percent more expensive), and expenditures of the floating system are balanced performance ratio (5–10 percent higher for floating by a higher expected energy yield—calculated for systems). 10 •  FLOATING SOLAR MARKET REPORT TABLE E.3. Results of (pre-tax) calculations of the LCOE of FPV vs. ground-mounted PV Floating PV 50 MWp LCOE Ground-mounted PV 50 Conservative Optimistic ($cents/kWh) MWp (+5% PR) (+10% PR) Tropical WACC 6% 6.25 6.77 6.47 8% 6.85 7.45 7.11 base case 10% 7.59 8.28 7.91 Arid/desert WACC 6% 4.52 4.90 4.68 8% 4.96 5.39 5.15 10% 5.51 6.01 5.74 Temperate WACC 6% 6.95 7.53 7.19 8% 7.64 8.30 7.93 10% 8.49 9.26 8.85 Source: SERIS calculations. Notes: kWh = kilowatt-hour; LCOE = levelized cost of electricity; MWp = megawatt-peak; PV = photovoltaic; WACC = weighted average cost of capital. The bold LCOE values are the “more likely” cases per type of climate. The LCOE calculation represents only a break-even Given their small size (except in China), most FPV analysis—that is, if the tariff were set at the LCOE, the projects are financed in local currencies and main- net present value of the project would be zero.1 Equity ly by local or regional banks. Japan, Taiwan, China investors would presumably require a higher tariff from and a few other economies have seen an increased the offtaker to make the project economically viable for involvement of local commercial banks seeking to them, assuming debt financing was accessible. take advantage of favorable long-term feed-in tariffs available for FPV. The involvement of large internation- If the performance ratio of an FPV project is assumed al commercial banks, and of multilateral development to be 10 percent higher than that of a ground-based finance institutions in developing countries, is expect- project (instead of 5 percent), a sensitivity analysis ed to grow as larger projects become more common shows that the LCOE is only 3-4 percent higher than in areas outside China. the one for the ground-mounted system. Project structuring Challenges To understand how FPV projects are typically financed, While enough large-scale projects have been imple- it is useful to classify them into two main categories: mented to show the commercial viability of FPV, there those with an installed capacity of 5 MWp or lower, are remaining challenges to its deployment—among and and those with an installed capacity greater than 5 them the lack of a robust track record; uncertainty MWp. Table E.4 summarizes typical financial structures surrounding costs; uncertainty about predicting envi- for these categories, which are similar to financial struc- ronmental impact; and the technical complexity of tures for land-based PV deployment. To gain trust in the designing, building, and operating on and in water technology, public grants are often provided to finance (especially electrical safety, anchoring and mooring R&D and pilot projects (<1 MWp), which are often run issues, and operation and maintenance). The experi- by universities or public research institutions. ence of other technologies operating in aquatic envi- ronments, including near-shore environments, offers The discounted payback period is 20 years, and the equity internal 1.  rate of return is set at the discount rate. good lessons in some of these areas. EXECUTIVE SUMMARY  • 11 TABLE E.4 . Financing structure vs. size of FPV system System size (MWp) Business model Ownership Financing structure ≤ 5 Self-generation Commercial Pure equity and/or corporate financing (or “on balance and industrial sheet” financing). Owner would typically be an energy- companies  intensive commercial or industrial company with ponds, lakes, or reservoirs on its premises and willing to install an FPV system for its own use. > 5 Power sold to Independent power Mix of debt and equity (typically 80:20); on balance sheet the grid producers and or non-recourse project finance. The latter is still rare, public utilities  however, because such project finance structures make sense only for projects of a certain size (generally larg- er than 10 MWp). Future large projects will likely have financing structures similar to the ones used for utility-scale ground-mounted PV projects. Source: Authors’ compilation. In addition to the technical aspects, challenges relat- output, while making better use of existing transmis- ed to permitting and commercial aspects include: sion assets, and this could be particularly beneficial a lack of clarity on licensing/permitting (especially in countries where grids are weak. concerning water rights and environmental impact assessment); difficulties in selecting qualified suppli- When combined with other demonstrated bene- ers and contractors; difficulties in designing insurance fits such as higher energy yield, reduced evapora- policies that include for example liabilities for potential tion, and improved water quality, FPV is likely to be damage of hydropower plant (when combined with an attractive option for many countries. Although the such plant); and uncertainties about the adequacy of market is still nascent, there is a sufficient number of warranties of the performance or reliability of critical experienced suppliers to structure a competitive ten- components. In most countries, the policy and regula- der and get a commercial project financed and con- tory framework needs to be adjusted to provide more structed, and the additional costs appear to be low clarity in some of these areas. and are falling rapidly. The priority over the next few years should be to carry Conclusions and next steps out strategic deployments of FPV at sites where it is already economic, while applying the “precautionary FPV deployment appears likely to accelerate as the principle” when it comes to possible environmental technologies mature, opening up a new frontier in the or social impacts. This may include initial limits on global expansion of renewable energy and bringing the portion of the water surface that is covered and opportunities to a wide range of countries and mar- efforts to avoid installations in the littoral zone near kets. With a global potential of 400 GW under very shore, where plant and animal life may be more abun- conservative assumptions, FPV could double the dant. In addition, development of the constituent existing installed capacity of solar PV but without the technologies and knowledge of positive and negative land acquisition that is required for ground-mount- impacts will be greatly enhanced if early installations ed installations. At some large hydropower plants, are diligently monitored, which will entail some public covering just 1–4% of the reservoir with FPV could expenditure. The need for monitoring, added to the double the installed capacity, potentially allowing possible additional capital costs of FPV over those of water resources to be more strategically managed by ground-mounted systems, and the risk profile of FPV, utilizing the solar output during the day. Additional- given its early stage of deployment, make early instal- ly, combining the dispatch of solar and hydropower lations in developing countries a strong candidate for could be used to smooth the variability of the solar concessional climate financing. 12 •  FLOATING SOLAR MARKET REPORT To support market development, an active dia- In addition to the financing of public and private logue among all stakeholders, public and private, is investments, the World Bank Group is committed to required to further global understanding of FPV tech- supporting the development of floating solar as well nologies and to spread lessons learned from early as hydro-connected solar by generating and dissem- projects across a wider area. Through this market inating knowledge. Publications and tools planned for report and an upcoming handbook for practitioners, the Where Sun Meets Water series are: the World Bank Group and SERIS hope to contribute • An FPV market report executive summary to this goal, and we look forward to working with gov- ernments, developers, and the research community • An FPV market report to expand the FPV market by bringing down costs, • An FPV handbook for practitioners supporting grid integration, maximizing ancillary ben- efits, and minimizing negative environmental or social • Global mapping of FPV potential (a geospatial tool) impacts. • Proposed technical designs and project structur- ing for hydro-connected solar. References BBC (British Broadcasting Corporation). 2018. “Solar Farm Means ‘I Can Breathe More Easily.’” Video story, BBC News, April 24. https://www.bbc.co.uk/news/av/business-43881280/solar-farm-means-i-can-breathe-more-easily. Choi, Y.-K., and N.-H. Lee. 2013. “Empirical Research on the Efficiency of Floating PV Systems Compared with Overland PV Systems.” CES-CUBE 25: 284–89. Global Solar Atlas: https://globalsolaratlas.info/, World Bank Group (2019). Lehner, B., C. Reidy Liermann, C. Revenga, C. Vörösmarty, B. Fekete, P . Crouzet, P. Döll, M. Endejan, K. Frenken, J. Magome, C. Nilsson, J. C. Robertson, R. Rodel, N. Sindorf, and D. Wisser. 2011a. “Global Reservoir and Dam Database, Version 1 (GRanDv1): Reservoirs, v1.01.” NASA Socioeconomic Data and Applications Center (SEDAC), Palisades, NY. http:// sedac.ciesin.columbia.edu/data/set/grand-v1-reservoirs-rev01. ———. 2011b. “High-Resolution Mapping of the World’s Reservoirs and Dams for Sustainable River-Flow Management.” Frontiers in Ecology and the Environment 9: 494–502. Mason, Pauline. 2018. “Meet the Ex-Miners Who Are Now Walking on Water.” BBC News, April 27. https://www.bbc.co.uk/ news/business-43864665. Qi, S. 2014. The Analysis of Complementation in PV Grid-Connected Part of Longyangxia 320 MWp, in Engineering. Xi’an: University of Technology. Shiklomanov, Igor A. 1993. “World Fresh Water Resources.” In Water in Crisis: A Guide to the World’s Fresh Water Resourc- es, edited by Peter H. Gleick. New York: Oxford University Press. Sungrow: https://en.sungrowpower.com/reference?id=22&ref_cate_id=30. Swimsol: https://swimsol.com/. Trapani, K., and M. Redón Santafé. 2015. “A Review of Floating Photovoltaic Installations: 2007–2013.” Progress in Photovol- taics: Research and Applications 23 (4): 524–32. EXECUTIVE SUMMARY  • 13 REPUBLIC OF KOREA © LSIS 1 WHY FLOATING SOLAR? The amount of solar energy reaching the earth is tre- (AC) to be fed into the power grid. The typical config- mendous. At noon, it can be more than 1,000 watts per uration of a grid-connected PV power plant is shown square meter (W/m ). The total solar energy received 2 in figure 1.1. over the course of a year is about 3,400,000 exajoules (EJ). This is roughly 7,500 times the world’s annual The PV industry is developing fast. Thanks to techno- primary energy consumption of about 450 EJ (WEC logical advancement and an increasing scale of pro- 2013). There are two main ways to harvest solar ener- duction, the cost of solar cells and modules has come gy, via solar heat and solar photovoltaics (PV). In PV down drastically over recent years. Solar PV modules generation, a device called a solar cell is used to turn were more than 80% cheaper in 2017 than they were light directly into electricity. The direct current (DC) in 2009 and the cost of electricity from solar PV fell by generated by an array of solar modules (solar cells almost three-quarters in 2010–2017 and continues to grouped and packaged together) then goes into an decline (IRENA 2018). Fueled by falling prices, the inverter, where it is converted to alternating current cumulative installed capacity of PV grew significantly FIGURE 1.1  General configuration of a photovoltaic power plant Sunlight Utility grid Solar modules LV/MV voltage set up Mounting racks AC utility Transfers Inverter & meter DC electricity DC/AC to inverter disconnects Transfers the AC service converted AC panel electricity Source: Adapted from IFC 2015. Note: AC = alternating current; DC = direct current; LV = low voltage; MV = medium voltage. CHAPTER 1:  WHY FLOATING SOLAR?  • 15 1 over the past years and by the end of 2018 stood at shallow water bodies. This report will only briefly dis- about 500 GWp (IEA 2018). Record new photovoltaic cuss this type of system. capacity was added in 2018, breaking the 100 GWp barrier for the first time (BNEF 2019). The vast major- Floating solar has the potential to become a third pil- ity of installations are either ground-mounted (often in lar of PV deployment and application, complementing large solar farms of tens to hundreds of megawatts, ground-mounted (or land-based) PV and rooftop PV. MW) or on rooftops of commercial/industrial buildings There are more than 400,000 square kilometers (km2) (where installations are also often on a megawatt scale) of man-made reservoirs in the world (Shiklomanov or private residences (with kilowatt scale installations). 1993), suggesting that FPV has a deployment poten- tial on a terawatt scale. Apart from saving land, bene- Spurred by the high cost or limited availability of land fits include greater efficiency and cost savings. These in countries such as Japan, the Republic of Korea, and other benefits will be discussed in this chapter, and Singapore, the PV industry has started to look along with a number of challenges that remain to be into using water bodies for PV applications. This has addressed. the added benefit of allowing the deployment of large PV installations near load centers, thereby reducing the The first FPV system was built in 2007 in Aichi, Japan. cost of transmission infrastructure. Since then, many such projects have been installed, with the largest to be found in China, Japan, and Korea, The term floating PV (FPV) may be used to refer to any and also Taiwan, China, the United Kingdom, India, the type of PV system installed on water bodies, such as United States, and Cambodia. Smaller systems (with lakes, reservoirs, hydroelectric dams, mining ponds, peak capacity below 2 megawatts, or megawatts-peak, industrial and irrigation ponds, water treatment ponds, MWp) have also been installed in countries such as and coastal lagoons. Figure 1.2 shows two examples Australia, Belgium, Brazil, Chile, Colombia, France, of FPV systems. In most cases, PV panels are usually Germany, Indonesia, Israel, Italy, Malaysia, the Nether- mounted on a pontoon-based floating structure. The lands, Panama, Philippines, Portugal, Singapore, Spain, floating platforms are anchored and moored at a fixed Sweden, Thailand, and Tunisia. Many of these smaller location. In this report, we distinguish FPV from anoth- systems were set up for research and demonstration er form of PV deployment that may be called “PV over purposes. Novel arrangements such as submerged water” and involves mounting PV panels on piles above and concentrated PV systems are also being tested in FIGURE 1.2  Examples of floating photovoltaic systems: 150MWp in Guqiao, China (left) and 10 kWp system in Kunde winery, California, United States (right) Source: © Sungrow. Source: © World Bank. Note: kWp = kilowatt-peak; MWp = megawatt-peak. 16 •  FLOATING SOLAR MARKET REPORT some places. The first commercial installation of FPV and especially in China and Southeast Asia (PV-Tech is a 175 kilowatt-peak (kWp) system set up at the Far 2017a, 2017b; Maisch 2017) but many tenders for Niente Winery, California, United States, in 2008. This floating solar are also announced in India (Saurabh utilizes an irrigation pond to avoid occupying land better 2016, Kenning 2017, Kenning 2018, Sivakumar 2019). used for growing grapes. Good reviews of early-stage Eastern China, for example, is highly populated and FPV projects and technologies can be found in several has limited available land but abundant water bodies. studies, including those of Connor (2009), Trapani and In Southeast Asia, meanwhile, FPV could unlock the Santafé (2015); Patil, Wagh, and Shinde (2017); and huge additional capacity of the region’s many existing Sahu, Yadav, and Sudhakar (2016). Beginning in 2013, hydropower plants, for example, along the Mekong Riv- FPV installations larger than 1 MWp started to emerge, er. As these and other new markets emerge, cumulative mainly in Japan and Korea, then in China. Interest in installed FPV capacity and new additions are growing FPV has since grown rapidly: large FPV plants (i.e., with rapidly (figure 1.3). Table 1.1 summarizes some import- peak capacity in the tens and even hundreds of mega- ant milestones in FPV installations’ early development watts) are being installed or planned around the world, stages (Planair and PITCO 2017). FIGURE 1.3  Global installed floating PV capacity and annual additions 1,400 1,314 1,200 1,000 786 800 MWp 528 600 400 359 200 169 68 101 0 0.5 1 1.5 2.2 3.4 5.7 11 0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Annual installed FPV capacity Cumulative installed FPV capacity Source: Authors’ compilation based on media releases and industry information. TABLE 1 .1  Milestones in the early development of FPV installations Milestones Installation First FPV installation 20 kWp in Aichi Province, Japan (2007) First nonresearch FPV installation 175 kWp at Far Niente Winery, United States (2008) First tracking FPV installation 200 kWp Petra Winery (rotating system), Italy (2010) First MW-scale FPV installation 1,180 kWp in Saitama Prefecture, Japan (2013) First FPV installation using micro-inverters 300 kWp in Fukuoka Prefecture, Japan (2016) First FPV combining solar and hydro 220 kWp at the Alto Rabagão Dam, Portugal (2017) Source: Planair and PITCO 2017. Note: FPV = floating photovoltaic; kWp = kilowatt-peak; MW = megawatt. CHAPTER 1:  WHY FLOATING SOLAR? • 17 1.1.  The benefits of floating solar or recreational purpose. In many cases, these can be used at low or no cost, unlike land, which must typically 1.1.1.  Land-use advantages of FPV be leased or purchased. For countries where land is scarce or unsuitable for PV installations, the cost of building ground-mounted PV To discourage solar PV farms from competing with land power plants is driven upward by high land prices and for other uses (such as agriculture) and to encourage high opportunity costs. This is particularly pertinent the utilization of idle water bodies, some countries or to small countries, or where a mountainous terrain is regions offer financial incentives for deploying PV on unsuitable for the deployment of PV or the opportunity water. For example, the government in Taiwan, China cost of using land is high (because of agriculture or has implemented a feed-in-tariff regime that favors float- urban development, for example). In countries such as ing installations over ground-mounted PV. In Massachu- Japan or Singapore, large-scale ground-mounted PV setts (United States), an extra “adder” value under the installations take up precious real estate. compensation rates of state incentives program is avail- able for FPV systems, while in Korea a higher renewable Even where large swathes of open land are avail- energy certificate (REC) weighting is given to FPV sys- able for the installation of PV modules, these may tems than for ground-mounted ones. Such incentives be far from populated areas where energy demand will be discussed in further detail in chapter 4. is high. To deploy PV in remote areas requires trans- mitting energy over long distances, using high volt- Possibility to utilize hard-to-access terrain 1.1.2.  age transmission lines, to the residential or industrial Ground-mounted PV may not be possible to deploy, areas where energy is actually needed. This is both for example, in mountainous regions but even here costly and inefficient, especially since a certain per- floating systems can be set up on man-made lakes or centage of the solar energy is lost in transmission. reservoirs. This is the situation in western China, where wind and solar resources are abundant and land is available at For instance, a test installation will be constructed on almost no cost, but the generated solar power cannot a hydropower dam in the Swiss Alps, at an altitude of be utilized in nearby regions and requires long trans- 1,800 meters (figure 1.4). The potential for boosting mission lines that are often only partially built. Cur- the performance of dams at high altitudes is great. tailments of up to 20–22 percent have been reported PV panels can benefit from the usually clear skies in Gansu and Xinjiang, for instance (National Energy seen at these altitudes, the extreme cooling effect, as Administration 2017), affecting investors’ returns and well as snow reflection. A land-based test installation confidence alike. On the other hand, deploying PV in the Alps produces an estimated 50 percent more in or near populated areas is costly, since land here power than it would at a lower altitude (Romande has a higher opportunity cost, and a higher real cost. Energie, 2018). In Taiwan, China, ground-mounted PV projects are restricted by the government because they compete However, similarly to a ground-mounted PV system in with agriculture. harsh operating environment, a floating system must be designed to withstand the challenges that can be In such cases, the advent of FPV offers a viable and brought by such environment, including a frozen water much-needed solution. It utilizes water surfaces that surface, snow coverage of panels, and possible large otherwise may not serve any economic, ecological, fluctuations in water levels. 18 •  FLOATING SOLAR MARKET REPORT FIGURE 1.4. Visualization of a future pilot plant on a hydropower dam in the Swiss Alps in winter (top) and summer (bottom) Source: © Romande Energie. Source: © Romande Energie. CHAPTER 1:  WHY FLOATING SOLAR? • 19 The industry expects that in the future offshore FPV will The effects of floating 1.2.  be integrated not only with fish farming but also with installations on water bodies other offshore applications such as water desalination, 1.2.1.  Integration with aquaculture and other oil and gas exploration, shipping, data centers’ cool- applications ing, or even hydrogen production. Floating solar installations can improve the econom- 1.2.2.  Reduced water evaporation ic value of water bodies, in particular in reservoirs Evaporation represents a significant loss of water where the water surface is left unused. In other cases, resources worldwide, with reported values as high as FPV can be combined with other productive uses, to 40 percent (Helfer, Lemckert, and Zhang 2012; San- increase profit and efficiency. tafé and others 2014). Reducing water evaporation is critical, especially in countries where water is scarce. For example, FPV can be added to pond-based or oth- Covering parts of a water body’s surface with FPV er types of fisheries, where it can replace the diesel panels is an efficient way to reduce evaporation from generators typically used for auxiliary services (e.g., drinking water reservoirs or irrigation ponds, even as it oxygen pumps, lighting). “PV over water” installations, generates green electricity (figure 1.7). where panels are mounted on piles, are favored in China in installations that combine PV and aquacul- The shade provided by floating panels not only reduces ture (figure 1.5). Adding PV to aquaculture has been the amount of solar radiation reaching the water, but also accomplished at several sites of the so-called Top limits the effects of wind on the water surface, which are Runner program, initiated by the Chinese government part of the evaporation process. However, quantifying to demonstrate and explore advanced PV technolo- the extent of evaporation reduction is difficult, especial- gies. This option is restricted to shallow waters, and ly since FPV plants typically cover only part of the entire for the purposes of this report is not considered FPV. water surface. Rigorous studies are needed that utilize long-term reservoir operation data, including water lev- Combining floating solar with fish farming is explored el variations, rainfall, inflow, and outflow. in Norway and Singapore in near-shore conditions (figure 1.6). Power supply to fish farms provided by Water quality and other potential 1.2.3.  floating solar presents various advantages as fishes environmental impacts only eat during the day and less during winter, periods which are highly compatible with solar power output to Floating solar is considered environmentally benign. provide the required electricity to the fish feeders. Most floats used to support PV panels are made of FIGURE 1.5. A “PV over water” installation FIGURE 1.6. Floating solar for fish farming in Singapore Source: © Jinko Power. Source: © Ocean Sun. 20 •  FLOATING SOLAR MARKET REPORT FIGURE 1.7. FPV system for covering of entire light for photosynthesis and less heat penetrating the reservoir surface to reduce water evaporation water surface. However, there may be adverse environmental impacts of blocking sunlight. Local studies are needed to understand the possible interactions between FPV installations and the water environment. Implications will also differ by the use of the water body. For exam- ple, natural lakes have a higher rate of bioactivity than industrial ponds used to cool water. In cases where an adverse impact is expected, the maximum surface to Source: © ISIGENERE. be covered by floating panels should be limited. a plastic material called high-density polyethylene 1.3. Technological advantages (HDPE), which is used in drinking water applications of floating solar (e.g., pipes) and does not degrade or contaminate the water. However, manufacturers’ claims should Besides saving land resources and potentially better always be tested, especially where a drinking-wa- use of water surfaces, FPV has some attractive tech- ter reservoir is involved. There are no international, nical benefits. These are mainly related to design and standardized testing procedures for floats, but some deployment, and power system performance/yield. countries (e.g., China) are developing their own cer- Note that while some advantages have been proven to tification programs. Many manufacturers conduct rel- a certain degree, others remain conceptual. evant tests during the product design phase and can provide the relevant test results. Related best practic- 1.3.1.  Increased energy yield es will be described in the next publication of Where One of the important advantages of deploying PV on Sun Meets Water series1, to follow shortly after this water is arguably the performance benefit derived from publication. the operating environment. There are four key elements of this: In addition, FPV discourages algae growth and could, in certain cases, improve water quality. Algae growth • The evaporative cooling effect of the water tends is significant in many reservoirs, and can increase to lower the operating temperatures of the PV the cost of treating water. Uncontrolled growth can modules. A study of an FPV testbed in Singa- have severe consequences for a lake’s ecological pore indicates that the ambient air temperature balance. For example, in Lake Taihu, China, 2007, on water is lower by about 1°C to 3°C than the significant algae growth lowered water quality and adjacent land environment (Liu and others 2018). led to the death of aquatic life (Qin and others 2010). This enables a lower operating temperature for PV City residents, too, were affected by the foul smell of modules. The cooling of PV modules is also more the drinking water drawn from that lake. Algae growth effective thanks to a higher temperature gradient is affected by several factors, such as water tempera- and thus faster heat transfer. As a result, module ture and light intensity. It is reasonable to assume that temperatures were observed to be lower by 5°C to by covering part of a reservoir’s surface with PV pan- 10°C, depending on the air ventilation underneath els, its growth can be curbed since there will be less the floating structures. In cases where the module is in good thermal contact with water, the cooling 1.  World Bank Group, ESMAP and SERIS. 2019. “Where Sun Meets effect can be even greater. Since elevated module Water: Floating Solar Handbook for Practitioners.” Forthcoming. Washington DC: World Bank. temperatures are a major loss factor for PV sys- CHAPTER 1:  WHY FLOATING SOLAR? • 21 tems in many climates, a reduction in temperature Complementary operation with 1.3.3.  can significantly increase the energy yield of a giv- hydropower en installed PV capacity. There is great potential worldwide for the combined • The wind speeds over open water tend to be higher and integrated operation of hydropower stations and than over land, thus facilitating module cooling. FPV. Usually dry seasons with less water flow corre- spond to periods of high solar insolation and vice ver- • Plants on water bodies are rarely shaded by near- sa. A hybrid of the two would thereby reduce seasonal by objects or buildings. Since the tilt angles of FPV variations in power production. In addition, the natural arrays are usually kept low to reduce wind loads, variability of solar radiation can be largely compensat- the inter-row shading is also reduced. ed by fast-responding hydro turbines. This improves • Water bodies tend to be less dusty than the arid power quality and reduces power curtailment. Also, a desert locations where solar farms are often con- hybrid system can optimize the diurnal cycle by lever- structed, thus minimizing the effects and complica- aging more solar power during the day and hydropow- tions of dust gathering on panels. er at night. Some early FPV projects reported an improved In hybrid systems, a reservoir is basically used as a energy yield of more than 10 percent over that of giant storage facility for the variable, nondispatchable ground-mounted PV systems (Trapani and Santafé solar power. Retrofitting existing hydropower plants 2015; Choi and Lee 2013). It is reasonable to expect with new FPV projects would benefit from (i) skilled that this benefit is highest in warm climates. More staff on site, and (ii) supervisory control and data details about the cooling effect will be discussed in acquisition (SCADA) systems developed for hydro- chapters 2 and 5. power plants. More details on this will be provided in chapter 2. Synergy with existing electrical 1.3.2.  infrastructure 1.3.4.  Easier installation and deployment Many inland freshwater bodies, especially the res- In cases where complicated anchoring and mooring ervoirs of hydropower plants, have nearby grid con- are not required (see also section 1.4), the process nections. As a result, the length of medium-voltage of installing FPV is in many cases simpler than for lines required to connect FPV to the grid is likely to be ground-mounted PV. No civil work is needed to pre- short. This can reduce investment in electrical infra- pare the site, since typical floating platforms in the structure. In the case of large irrigation reservoirs, market are modular, made of small individual floats per water treatment plants, cooling ponds for industri- module and interconnecting units. They do not require al use, or other energy-intensive infrastructure, the heavy equipment during construction. The platforms on-site self-consumption of the electricity produced are assembled on land and get pushed into the water by the installed FPV plants would further decrease as the number of rows increases (figure 1.8 top). costs and energy losses. Thereafter they get towed to an exact location on the reservoir (figure 1.8 bottom). In sum, deployment times The existence of electrical infrastructure is location are shorter, and costs lower, than for ground-mount- and project specific. Depending on the situation, FPV ed PV. For example, a major FPV developer from Chi- project developers need to make proper arrangements na recently reported that a 1 MWp system can be for the metering and integration of electrical systems. installed by 50 people in one day, provided that a sup- They also need to check relevant regulatory provisions, ply chain is in place. Installation and deployment will e.g., for self-consumption or net metering. be discussed in greater detail in the next publication of Where Sun Meets Water series. 22 •  FLOATING SOLAR MARKET REPORT FIGURE 1.8. Deployment ramp (top) and towing of FPV platform into exact location (bottom) Source: © Lightsource BP Floating Solar Array, London. Source: © Pixbee/EDP S.A. 1.4. Challenges ground-mounted PV, owing chiefly to the expenses for the floats, mooring and anchoring, and more stringent At this moment, FPV still comes with challenges that requirements for electrical components. The cost of will require further research and learning to facilitate floats is expected to drop over time, but economies of wider adoption. scale today remain constrained by a relatively small installed capacity. 1.4.1.  Capital expenses The capital costs of FPV are currently still slight- Optimizing the floating platform design by reducing ly higher than or at best comparable to those of unnecessary buoyancy and cutting some mainte- CHAPTER 1:  WHY FLOATING SOLAR? • 23 nance pathways (e.g., by employing dual-pitch struc- on water than on land, FPV islands have been seen to tures) may also help to save some costs. attract birds (and their droppings) in the United King- dom and Singapore. Protecting installations from birds 1.4.2.  Anchoring and mooring is possible but would increase O&M expenses. Anchoring and mooring fixes a platform, and keeps PV Electrical safety and long-term reliability 1.4.4.  panels correctly oriented toward the sun. The anchor- of system components ing has to withstand wind load, waves, and potential currents. In some cases, the system needs to accom- When electrical systems are constantly exposed to modate large fluctuations in water levels (e.g., in coun- humidity and possibly also salinity (in offshore or near- tries with dry and wet seasons). In some reservoirs, shore installations), this poses risks to their operation, water depth and the terrain of the water body’s bed can especially over the long run. Also, floating structures pose challenges to the installation and maintenance of are in constant motion. Degradation and corrosion the anchoring. Here, more complicated solutions may occur more quickly than on land, and bio-fouling is be required, adding to the cost of the project. an additional challenge not faced by land systems. System components may need to be periodically rein- 1.4.3.  Operation and maintenance forced or replaced to ensure systems’ long-term reli- ability and safety. Operation and maintenance activities are generally more difficult to perform on water than on land. Boats Temperature fluctuations may cause floats to bloat are usually required to access PV arrays, even for and shrink, which can cause cracks. Freezing may installations with maintenance pathways. Anchoring stress system components, particularly joints. Howev- and mooring cables must be regularly inspected, an er, experience from the past few years (in Japan and activity that may require divers. Replacing parts is also China, among others) suggests that floating platforms more complex, and workers’ safety must be adequate- can well survive ice and snow (figure 1.9). ly protected. While dust collection is less of an issue FIGURE 1.9. Deployment of FPV in freezing conditions in Japan (left) and China (right) Source: © Sungrow. Source: © Sungrow. 24 •  FLOATING SOLAR MARKET REPORT FIGURE 1.10. Stackable floats for efficient transport Source: © ISIGENERE. Source: © ISIGENERE. 1.4.5.  Transportation of floats floats such that they can be more easily transported as illustrated in figure 1.10. Most floats are bulky and have a very low weight-to- volume ratio, making them difficult to ship. The cost of transporting them to remote locations may be prohibi- 1.5. Comparison with tively high. The manufacturing of floats for many large ground-mounted systems FPV projects has been done locally to avoid this prob- As has been noted in this report, FPV installations lem. In the future, mobile manufacturing equipment offer several benefits over land-based systems, even may offer a solution. Else, float suppliers try to collabo- as they pose additional challenges and costs. Table rate with local plastic molding manufacturers to reduce 1.2 provides a comprehensive look at both types of cost of transport. Some suppliers are also designing systems. CHAPTER 1:  WHY FLOATING SOLAR? • 25 TABLE 1.2. Floating and land-based photovoltaic systems: A comparison Parameter Floating PV Land-based PV Land/water surface •  Suitable/affordable land may be far Does not compete for land with agricultural, industrial, •  use or residential projects away from load centers, thus requiring • Often easier to find sites near densely populated areas costly transmission infrastructure • Since water bodies often have a single owner, the Requires change in land use, which •  permitting process is often less complicated can be time consuming •  Expected lower leasing cost Competes for land with city dwellings, •  •  Potential integration with aquaculture industrial development, and agriculture • May save water resources by reducing water evapo- ration Plant design •  Modular design on flat surface • Design must accommodate terrain • Limited tilt due to wind load considerations imply a and area constraints lower energy yield in high-latitude regions •  Easier to implement tracking • Anchoring cables require periodic inspection and •  Yield prediction is better established maintenance Performance/ • Lower module temperatures (effect is Can benefit from tracking, bifacial, •  energy yield dependent on climate) and optimum tilt angle •  Nearly no shading More temperature losses in hot •  •  Lower soiling from dust climates • Overall 5–10 percent higher initial performance ratio (climate specific) • Long-term degradation (e.g., potential induced degradation) is still uncertain Installation and • In general, easy assembly, but highly variable • Efficiency varies depending on loca- deployment depending on location and workforce availability tion and workforce availability • Transportation of floats to site is difficult; favors local • Needs heavy equipment and land production preparation •  Needs suitable launching area •  Depends on soil quality Power system Synergy with existing electrical •  Costs of grid interconnection are •  benefits infrastructure often borne by project developer •  Possible hybrid operation with hydropower and can be prohibitively high Environmental • Long-term effects on water quality are not well • Some adverse impacts during established construction •  Potential to reduce algae growth •  Potential habitat loss or fragmentation •  Potential to reduce water evaporation •  Potential impact on aquatic ecosystems Investment Slightly higher costs on average due to floats, •  • Huge installed capacity and hence anchoring, mooring, and plant design very established investment and •  Cost of floats may drop as scale of deployment financing sector increases •  Costs continue to drop Higher perceived risk due to lower level of maturity •  Operation and •  Harder to access and replace parts •  Easy to access maintenance • Biofouling •  More affected by vegetation growth •  Animal visits and bird droppings •  Easier to deploy cleaning routines •  Harder to maintain anchoring •  Easy access to water for cleaning •  Lower risk of theft/vandalism continued 26 •  FLOATING SOLAR MARKET REPORT TABLE 1.2. continued Parameter Floating PV Land-based PV Durability Normally 5 to 10 years of warranty on floats •  Key system components durable for •  >20 years Safety • Close to water, tend to have lower •  Generally safe insulation resistance to ground • Constant movement poses challenge for equipment grounding •  Risk of personnel falling into water Regulation and • More difficult for natural lakes and easier for artificial • More established permitting process permits ponds •  Clearer regulations •  Lack of specific regulations Experience/level of Cumulative capacity as of end of 2018: >1.3 GWp •  • Cumulative capacity as of end of maturity •  4 years of experience with large-scale projects 2018: >500 GWp •  Thousands of projects built •  10–30 years of experience Source: SERIS. Note: GWp = gigawatt-peak. CHAPTER 1:  WHY FLOATING SOLAR? • 27 References BNEF (Bloomberg New Energy Finance). 2017. “China’s Renewables Curtailment and Coal Assets Risk Map.” https://data. bloomberglp.com/bnef/sites/14/2017/10/Chinas-Renewable-Curtailment-and-Coal-Assets-Risk-Map-FINAL_2.pdf. ———. 2019. “Clean Energy Investment Exceeded $300 Billion Once Again in 2018.” https://about.bnef.com/blog/clean- energy-investment-exceeded-300-billion-2018/. Choi, Y.-K., N.-H. Lee, and K.-J. Kim. 2013. “Empirical Research on the Efficiency of Floating PV Systems Compared with Overland PV Systems.” CES-CUBE. https://www.semanticscholar.org/paper/Empirical-Research-on-the-efficiency- of-Floating-PV-Choi-Lee/a38910b49973d099603974c24cd55f43a8b0f64c. Connor, P.M. 2009. Performance and prospects of a lightweight water-borne PV concentrator, including virtual storage via hydroelectric-dams. Proceedings of the ISES Solar World Congress 2009: Renewable Energy Shaping Our Future. Helfer, F., C. Lemckert, and H. Zhang. 2012. “Impacts of Climate Change on Temperature and Evaporation from a Large Reservoir in Australia.” Journal of Hydrology 475: 365–78. IEA (International Energy Agency). 2018. Renewables 2018. OECD/IEA, Paris, France. IFC (International Finance Corporation). 2015. “Utility-Scale Solar Photovoltaic Power Plants: A Project Developer’s Guide.” Working paper, International Finance Corporation, Washington, DC. https://openknowledge.worldbank.org/han- dle/10986/22797. IRENA (International Renewable Energy Agency). 2018. “Renewable Power Generation Costs in 2017”, International Renewable Energy Agency, Abu Dhabi. https://www.irena.org/publications/2018/Jan/Renewable-power-genera- tion-costs-in-2017. Kenning, T. 2017. “India’s SECI Invites Expression of Interest for 10GW of Floating Solar.” PV Tech, December 19. https:// www.pv-tech.org/news/indias-seci-invites-expression-of-interest-for-10gw-of-floating-solar. ———. 2018. “Maharashtra Discom Consults on 1GW of Floating PV, Retenders 1GW of Solar.” PV Tech, April 11. https:// www.pv-tech.org/news/maharashtra-disocm-consults-on-1gw-of-floating-pv-retenders-1gw-of-solar. Le Nouvelliste. 2019. “Le projet de parc flottant sur le lac des Toules entrera en fonction cet automne.” https://www.lenouvel- liste.ch/articles/valais/martigny-region/le-projet-de-parc-solaire-flottant-sur-les-lac-des-toules-entrera-en-fonction-cet-au- tomne-827633. Liu, H., V. Krishna, J. L. Leung, T. Reindl, and L. Zhao. 2018. “Field Experience and Performance Analysis of Floating PV Technologies in the Tropics.” Progress in Photovoltaics: Research and Applications 26 (12): 955–1012. Maisch, M. 2017. “40 MW Floating PV Plant in China Connected with Sungrow’s Inverters.” PV Magazine, May 19. https:// www.pv-magazine.com/2017/05/19/floating-pv-plant-in-china-connected-with-sungrows-inverters/. National Energy Administration. 2018. “ 国家能源局新闻发布会介绍2017年度相关能源情况等 [National Energy Administration Press Conference on the Energy Situation for the Year 2017].” http://www.nea.gov.cn/2018-01/24/c_136921015.htm. Patil, S. S., M. M. Wagh, and N. N. Shinde. 2017. “A Review on Floating Solar Photovoltaic Power Plants.” International Journal of Scientific & Engineering Research 8 (6). Planair and PITCO. 2017. Assessment of Floating Solar PV Potential for Pakistan; Task 1: Commercial Readiness of FSPV—Global Market and Performance Analysis. Unpublished report prepared for International Finance Corporation, Washington, DC. PV-Tech. 2017a. “Masdar and Indonesian Power Giant to Build World’s Largest Floating Solar Plant.” PV-Tech, November 28. https://www.pv-tech.org/news/masdar-and-indonesian-power-giant-to-build-worlds-largest-floating-solar-pl. ———. 2017b. “Sungrow Building Another Record 150MW Floating Solar Project in China.” PV-Tech, June 1. https://www. pv-tech.org/news/sungrow-building-another-record-150mw-floating-solar-project-in-china. 28 •  FLOATING SOLAR MARKET REPORT Qin, B., G. Zhu, G. Gao, Y. Zhang, W. Li, H. W. Paerl, and W. W. Carmichael. 2010. “A Drinking Water Crisis in Lake Taihu, China: Linkage to Climatic Variability and Lake Management.” Environmental Management 45 (1): 105–12. Sahu, A., N. Yadav, and K. Sudhakar. 2016. “Floating Photovoltaic Power Plant: A Review.” Renewable and Sustainable Energy Reviews 66 (Supplement C): 815–24. Santafé, M. R., P. S. F. Gisbert, F. J. S. Romero, J. B. T. Soler, J. J. F. Gozálvez, and C. M. F. Gisbert. 2014. “Implementation of a Photovoltaic Floating Cover for Irrigation Reservoirs.” Journal of Cleaner Production 66: 568–70. Saurabh. 2016. “Indian Hydro Power Company Plans 600 MW Floating Solar Power Project.” Cleantechies, June 26. http:// cleantechies.com/2016/06/27/indian-hydro-power-company-plans-600-mw-floating-solar-power-project/. Shiklomanov, Igor A. 1993. “World Fresh Water Resources.” In Water in Crisis, edited by P. H. Gleick. Oxford: Oxford University Press. Sivakumar, B. 2019. “Tangedco aims for 250MW solar floating projects.” The Economic Times, January 25. https://energy. economictimes.indiatimes.com/news/renewable/tangedco-aims-for-250mw-solar-floating-projects/67682064. Trapani, K., and M. Redón Santafé. 2015. “A Review of Floating Photovoltaic Installations: 2007–2013.” Progress in Photovoltaics: Research and Applications 23 (4): 524–32. WEC (World Energy Council). 2013. “World Energy Resources: Solar.” https://www.worldenergy.org/wp-content/ uploads/2013/10/WER_2013_8_Solar_revised.pdf. World Bank Group, ESMAP and SERIS. 2019. “Where Sun Meets Water: Floating Solar Handbook for Practitioners.” Forthcoming. Washington DC: World Bank. CHAPTER 1:  WHY FLOATING SOLAR? • 29 PORTUGAL © PIXBEE/EDP S.A. 2 TECHNOLOGY OVERVIEW The electrical configuration of a floating PV (FPV) aspects of FPV systems are covered in section 2.2.2 system is similar to that of a land-based PV system, Section 2.3 deals with the operation of an FPV system except the PV arrays and often the inverters float on in tandem with a hydropower station, a hybridization water. Figure 2.1 shows the typical configuration of a that opens a huge potential market, given the vast large-scale FPV power plant using a central inverter. amount of hydropower installed capacity worldwide. Electricity generated by PV modules is gathered by combiner boxes and converted to AC power by invert- FPV systems can be installed on a wide variety of ers. The floating platform, together with its anchoring water bodies such as industrial ponds, hydropower and mooring system, is an essential part of any FPV reservoirs, agricultural ponds as well as other types of installation. In this chapter, we offer an overview of the man-made water bodies like flood control reservoirs. components and technologies of FPV installations. All these applications are mainly inland freshwater bodies. However, FPV systems can also be built off- Section 2.1 describes mainstream FPV platforms and shore or near-shore. Figure 2.2 illustrates these various solutions, including anchoring and mooring. Novel applications.  FIGURE 2.1. Schematic of a typical large-scale FPV system, showing key components Transmission Central Lightning protection inverter (from other arrays) PV modules system (connected to metal frames supporting modules and grounded) Floats/pontoons Transformer Combiner box Mooring lines Anchoring Source: SERIS. 2. Except in section 2.2, less commonly used technologies are not covered in this report. 31 CHAPTER 2:  TECHNOLOGY OVERVIEW  • 31 FIGURE 2.2. Typical FPV applications M n-m d w t r bodi s Industri l ponds • Reservoirs for flood control • Cooling ponds • Water catchment areas • Wastewater-treatment ponds • Hydropower reservoirs • Mining and quarries water bodies A ricultur ponds Offshor nvironm nt • Irrigation ponds • Deployment near shore Source: Authors. Key components and system 2.1.  ble. Details on the various technologies and designs designs currently available in the market are presented in the subsections that follow. Most large-scale FPV plants have pontoon-type- floats, upon which PV panels are mounted at a fixed 2.1.1.  Floating platforms tilt angle. The floating structure can consist of floats alone (called pure floats), floats with metal trusses, Pure-floats design or special membranes or mats. The platform is held Pure-float configurations use specially designed buoy- in place by the anchoring and mooring system, the ant bodies to support PV panels directly. Table 2.1 design of which depends on factors such as wind summarizes the pros and cons of this platform type. load, float type, water depth, and variation in water level. The layout of the PV plant is generally simi- As an example, the Hydrelio floats from Ciel & Terre lar to that of land-based installations, except in the International are illustrated in figure 2.3. The float sys- case of smaller floating plants located close to shore, tem is modular and consists of two types of floats. which offer the option of placing the inverters on “Main floats” support the PV modules and provide land, i.e., separated from the PV array. Both central an optimum tilt to the module (different tilt angles are and so-called string inverter configurations are possi- possible, depending on the model used). “Secondary 32 •  FLOATING SOLAR MARKET REPORT TABLE 2.1. Advantages and disadvantages of pure-float design Advantages Disadvantages •  Systems are easy to assemble and install Modules are mounted very close to water. This reduces air circulation •  • Systems can be scaled without major and cooling effect from evaporation. changes in design. It also generates a high-humidity environment for both PV modules and cables. • Few metal parts are required, minimizing corrosion. It is not cost-effective to transport pure floats over long distances, •  so they may need to be molded in nearby facilities • Platform adapts to wave motion and relieves stress. Constant movement may cause stress and fatigue to joints and •  connectors. Source: SERIS. FIGURE 2.3. Components of floats from Ciel & Terre International Main float supporting the PV module Standard framed Secondary float for 60 cells PV module maintenance/buoyancy Connection pin Rail to fix the PV module Source: © Ciel & Terre International. on the floats floats” ensure connection with the main floats, provide Pure-float designs from other suppliers such as Sun- sufficient spacing to limit the shading of PV modules, grow Floating (Sungrow) are conceptually similar, with and are used as maintenance walkways while lending their own features. Figure 2.4 shows Sungrow’s floating additional buoyancy. The floats are connected with pins platform design. or bolts to form a large platform. The material used is UV- and corrosion-resistant high-density polyethylene A further example, shown in figure 2.5, comes from (HDPE) that is manufactured through a blow-molding Sumitomo Mitsui Construction Co., Ltd (SMCC). It process. It is compatible with drinking water.3 This type features a more regularly shaped float for denser of floating structure has established itself as the most packing and easier transportation. Filled with polysty- common solution, with several suppliers in the market rene foam, the float will not sink even if damaged. In and an installed capacity worldwide of several hun- addition, the connecting parts are banded together, dred megawatt-peak (MWp). which, according to the manufacturer, reduces the risk of structural failure. To be compatible with drinking water, the material must pass certain 3.  A last example, shown in figure 2.6, comes from standards. More details are provided in World Bank Group, ESMAP and SERIS, 2019. ISIGENERE S.L. who developed the ISIFLOATING CHAPTER 2:  TECHNOLOGY OVERVIEW  • 33 FIGURE 2.4. Sungrow floating platform design (top) and floats (bottom) Main floating body Connecting floating body Aisle floating body Multi-floating body Source: © Sungrow. FIGURE 2.5. Illustration of Sumitomo’s floating platform design Solar panels Upright stands Floats Solar panel brackets Anchor bolts Aluminum plate Binding bands Polystyrene foam Source: © SMCC. design, which was one of the pioneering floating Pontoons + metal frames solar systems since 2008. Their solution is charac- Another common design is to use metal structures terized by using a HDPE pure bi-float design, which (frames or trusses) to support PV panels as with is compact, nestable and stackable (thereby easy to land-based systems, but to affix the structures to pon- transport) and forms a closed volume when the PV toons, which serve only to provide buoyancy. In this panel is fixed on the top side. case, there is no need for specially designed floats. 34 •  FLOATING SOLAR MARKET REPORT Often used are capped pipes having technical spec- Alternatively, the metal trusses can be built on floats of ifications similar to those of pure floats with respect other shapes, as in the examples from NRG Energia, to strength, non-toxicity, and durability. Pipes may be Takiron Engineering and Scotra shown in figure 2.8. easier to obtain locally than pure floats. Such platforms are offered by companies such as 4C Solar and Koinè In another design from Solaris Synergy, the metal frame Multimedia (figure 2.7). stands on four specially designed floats (figure 2.9a). FIGURE 2.6. ISIFLOATING platform design Maintenance platform Same modular float covered with plastic top Photovoltaic panel Floats connection Quick clip fixing Modular float Source: © ISIGENERE. FIGURE 2.7. Various designs using metal frames and pipes to support PV panels, 4C Solar (top) and Koine Multimedia (bottom) Source: © SERIS. Source: © SERIS. CHAPTER 2:  TECHNOLOGY OVERVIEW  • 35 FIGURE 2.8. Various designs using floats and metal frames to support PV panels, NRG Energia (top), Takiron Engineering (middle), Scotra (bottom) Source: © SERIS. Source: © SERIS. Source: © Scotra. 36 •  FLOATING SOLAR MARKET REPORT FIGURE 2.9. Solaris Synergy design: Floats (a), outer ring (b) and an illustration of automatic wind adaptation (c). a b Source: © SERIS. Source: © SERIS. c wind NoNo wind Wind Wind Source: Authors based on Solaris Synergy. Each floating assembly supports several PV panels Membranes and mats to form a unit. Multiple units are then held together by Another type of platform is created by simply cover- cables and encircled by an outer ring, as shown in fig- ing the entire water surface with rubber mats to cre- ure 2.9b. Within a single ring, a maximum of 2MWp can ate a base for PV installation (figure 2.10). Although be installed; rings can be connected in a honeycomb much less common than the previous two types of pattern to achieve any desired total capacity. One inter- platforms, this option is being explored by Continental esting feature of this design is that panels can auto- Corporation and other companies. Covering the entire adapt to reduce wind load, because wind produces water surface is particularly suitable for desert areas torque that flattens the tilt of PV panels, which subse- (e.g., parts of Israel) to prevent evaporation losses quently relieves the drag forces produced by wind (fig- and save scarce water for irrigation or drinking. The ure 2.9c). design is conceptually simple and provides an easy base for installation and maintenance. In figure 2.10, The chief advantages and disadvantages of this type the membrane is fastened to a circular concrete rim of platforms are listed in table 2.3. and equipped with weights and floats to preserve its shape and to form trenches of varying depth that CHAPTER 2:  TECHNOLOGY OVERVIEW  • 37 TABLE 2.3. Advantages and disadvantages of pontoon + metal structures. Advantages Disadvantages •  The concept is simple. With more rigid structures, waves cause stress to •  • Floats are easy to make and therefore can be easily concentrate at certain points. sourced locally. •  Structures are more difficult to assemble. • Wave movement between PV modules is less variable, • Access for maintenance can be difficult in certain thus reducing wear and tear on module connection designs. components and wires. Source: SERIS. FIGURE 2.10. Floating solar membrane cover concept (left) and installation (right) Source: © Continental Corporation. Source: © Continental Corporation. accommodate changes in water level and hold rain- 2.1.2. Anchoring and mooring systems water. This technology may not be easily scalable. At An appropriate anchoring and mooring system is a the moment, it is more suitable for smaller-scale sys- critical part of an FPV plant. There are three basic tems on reservoirs or irrigation ponds up to around ways to hold a floating platform in place: bank anchor- 100,000 to 200,000 m2 in size. ing, bottom anchoring, or piles (figure 2.11). Devel- opers choose the design that best suits the platform Similarly, Ocean Sun uses large, round membranes location, bathymetry (water profile and depth), soil fixed to a floating ring up to 72 meters in diameter. The conditions, and variation in water level. More informa- system was adopted from fish farming in Norway and tion on how to choose and design an anchoring and was initially used for offshore applications (and thus is mooring system is provided in the next publication of discussed in section 2.2.5) but is starting to be also Where Sun Meets Water series (World Bank Group, deployed on inland reservoirs. ESMAP and SERIS, 2019). Membrane-based systems have the advantage of Bottom anchoring being in direct contact with water; heat from sunlight Bottom anchoring is used in the vast majority of exist- is discharged into the water, thus lowering the oper- ing FPV plants. The anchor must keep the FPV arrays ating temperature of the PV modules and increasing in place for 25 years or more, unlike the kedge anchors energy yield. used on ships, which need only resist lateral movement over a limited period of time. Many mature anchoring It is also possible to float specially designed PV pan- solutions exist in marine and ocean engineering, as els directly on water or in a semi-submerged manner. well as in watercraft industries, solutions that can be However, so-called submerged FPV (discussed fur- easily transferred and adapted to the FPV context. ther in section 2.2.3) is not yet a mainstream solution and is not widely deployed. 38 •  FLOATING SOLAR MARKET REPORT FIGURE 2.11. Schematics of bottom anchoring (here using so-called concrete sinkers), bank anchoring, and anchoring on piles Source: Authors. There are two broad types of permanent bottom Where the terrain and soil conditions are more com- anchors—self-seating anchors and installed anchors. plex, or where loads are large, installed anchors may One self-seating anchor commonly used for FPV con- be needed to provide a stronger hold to the bottom. sists of a dead weight, usually a large concrete block A helical anchor is a shaft equipped with wide spiral (called a “concrete sinker”) that resists movement blades that allow it to be screwed into the substrate. by its sheer weight and, to a lesser degree, by set- Installed anchors are generally more expensive than tling into the substrate. This cheap and simple option the self-seating variety; specialized boats and divers is effective in many cases. Other common types of are often required. self-seating anchor include mushroom anchors and pyramid anchors. CHAPTER 2:  TECHNOLOGY OVERVIEW  • 39 With any anchor, mooring lines must be selected figuration is particularly useful for installations with and deployed. Enough slack should be present to special features such as tracking and concentration accommodate stress levels and variations in water (see section 2.2.1). In this case, the pile provides a level, but not so much as to permit excessive move- central pole around which the platform revolves. In ment of the platforms. More details about anchor- response to variations in water level, the platform can ing and mooring are offered in World Bank Group, (in principle) slide up and down the piles. However, ESMAP and SERIS, 2019. pile drilling usually involves specialized equipment and civil works; hence it is much more costly than Bank anchoring anchoring. Bank anchoring is particularly suitable for small, shal- low ponds, where the FPV plant is close to shore (fig- Electrical configuration (central vs. 2.1.3.  ure 2.12). Bank anchoring may also be used when string inverters) other options are not available—for example, when the Like ground-mounted PV plants, FPVs use either bottom of the basin is lined with plastic and cannot central inverters or string inverters for their electrical accommodate an anchor. layout. For large FPV plants, it is beneficial to install the inverters on water instead of on shore so as to Whenever possible, bank anchoring should be con- avoid excessive resistive losses. Special floats made sidered, as it is often the most cost-effective option. of stainless steel or concrete are used to support It allows easy access to anchoring points, both for containerized central inverters. Depending on the deployment and for periodical inspection during supplier, the inverters may be integrated with trans- O&M. But feasibility of bank anchoring may also formers, with medium-voltage cables connecting the depend on shore conditions and on permission from transformers to the transmission grid. This is the con- the pond owner. figuration used by Sungrow at its large FPV farms in China (figure 2.13). Piles For some (typically shallow) water bodies, it may be For use on water, some engineers advocate string possible to drill or ram piles into the basin floor. The inverters. They are lighter than large central invert- float platform is then moored to the piles. This con- ers and can be placed on regular floats (figure 2.14). Although string inverter solutions tend to be more FIGURE 2.12. Bank anchoring example expensive than central inverters in large FPV plants, they offer higher granularity, so that in the event of a failure only small sections of the PV plant are affect- ed. The failed inverter can be quickly switched out if a few replacement units are kept on site. In FPV plants, easy access to the string inverters should be part of the layout. For example, rather than being placed in the center of a large array, the inverters ought to be on the periphery, accessible by boat. Source: © ISIGENERE. 40 •  FLOATING SOLAR MARKET REPORT FIGURE 2.13. Sungrow FPV farm with a central inverter on a floating island (left), detailed view of the floating island for the central inverter (right) Source: © Sungrow. Source: © Sungrow. 2.2.  Novel FPV concepts FIGURE 2.14. String inverters placed on the floats together with PV arrays FPV plant designs are not limited to the standard com- ponents or options discussed above. In this section, we describe some additional features that could be implemented in FPV installations. These include track- ing, concentration, and active cooling, all of which are still in an early stage of development. Most pilot plants are small in scale, and have been deployed for research and testing purposes. 2.2.1. Tracking Tracking can be achieved by rotating the entire float- ing platform to follow the sun from east to west. This Source: © SERIS. type of vertical-axis azimuth tracking is particularly relevant for FPV, since it is relatively simple to move FIGURE 2.15. Illustration of azimuth tracking for an an array on water (with its lower resistance) than on entire platform around a central pile land. In addition, because alignment with the sun’s position need not be completely accurate, the dis- turbances caused by wave movements are of minor consequence. Platforms can be moored around a central pile or surrounded by a fixed outer ring (or polygon), as illustrated in figure 2.15. The platform is usually circular, with a diameter of up to 100 meters. The platform is rotated by motors. Some engineers have proposed the use of bow thrusters to complete the rotating motion. Pilot tracking systems have been installed at the Lotus project in Suvereto, Italy, and in Source: © Koine Multimedia. CHAPTER 2:  TECHNOLOGY OVERVIEW  • 41 FIGURE 2.16. The 100 kW tracking FPV plant at Hapcheon Dam, Korea Source: © K-Water. Source: © K-Water. Navacchio, Pisa, also in Italy (Patil Sujay, Wagh, and tures. On water, high concentration is not possible Shinde 2017; Rosa-Clot and Tina 2018a). A larger because constant movement of the platforms impede plant of 100 kilowatt (kW) capacity, installed at Hap- precise position control. However, a certain degree of cheon Dam in the Republic of Korea, consists of four concentration can be achieved using mirrors or Fres- rotating structures (figure 2.16). A few tracking FPV nel lenses. For example, light can be concentrated to systems have been deployed in combination with a horizontal PV panel using V-shaped mirrors (Rosa- concentration (see section 2.2.2). Clot and Tina 2018b; Tina, Rosa-Clot, and Rosa-Clot 2011), as illustrated in figure 2.17. Calculations show Cost is the biggest challenge for tracking systems. that the concentration factor can be as great as three. Both the initial capital investment and the maintenance costs are rather high. In addition, the platform size is Concentration in FPV systems pairs naturally with limited, so scaling up is more challenging. In general, tracking, as indicated by the so-called floating tracking single-axis trackers improve the energy output of a cooling concentrator system. Here, mirrors are placed solar farm by about 20 to 30 percent (Mousazadeh in front of each PV panel, and the entire platform and others 2009). The gain is less for azimuth track- rotates in a circle to track the sun using azimuth track- ing in low-latitude regions, since the sun position is ing, as described in the previous section. One such at high angles at midday. Moreover, to reduce wind system was built in Australia (figure 2.18). loads, FPV systems are typically installed at a maxi- mum tilt angle of 10–15 degrees, which might further In principle, dual-axis tracking is also possible. A reduce the effect of azimuth tracking. In these cas- system in India has implemented dual-axis tracking es, horizontal tracking is needed. This may be more on water at a very small scale, as shown in figure difficult to realize for FPV systems, but novel tracking 2.19. This technology, called the Liquid Solar Array, mechanisms are being explored in this area. uses plastic concentrators that float on water and are mounted on anchored rafts. A thin focusing Fres- nel lens rotates to track the sun. Silicon PV cells are 2.2.2.  Concentrated FPV housed in a PV container that floats on water where Concentrated PV could be a relevant and attractive the cells are cooled thanks to the surrounding water, option on water, since the ambient temperature tends while allowing the concentrated light to enter through to be lower and water, as a coolant, is readily avail- a glass window. In bad weather the lens is protected able. Both help to alleviate the common problem in by rotating under the water surface to avoid damage concentrated PV systems of high operating tempera- in high winds. Water therefore becomes an essential 42 •  FLOATING SOLAR MARKET REPORT FIGURE 2.17. Low concentration with V-shaped mirrors for FPV 2 V-trou h Conc ntr tion Source: © Koine Multimedia. Source: Authors based on Rosa-Clot and Tina 2018b. FIGURE 2.18. FPV with azimuth tracking (1-axis, vertical) in a wastewater facility, Jamestown, Australia Source: © Infratech Industries. component of the design, both for cooling and pro- 2.2.3. Submerged FPV tecting (Connor 2009). Putting PV panels in direct contact with water to exploit its cooling properties can significantly lower As with general FPV tracking systems, concentration operating temperature, thereby increasing power out- systems suffer from the drawbacks of high cost and put. This is a major benefit of the membrane-based less scalability owing to the large number of accesso- FPV systems described earlier, but it also accounts for ry components and structures. the appeal of submerging PV modules just beneath CHAPTER 2:  TECHNOLOGY OVERVIEW  • 43 FIGURE 2.19. Sunengy’s Liquid Solar Array with dual-axis tracking and concentrators (left), detail of the collector with lens concentrator (right), Whalvan Hydroelectric Dam, Lonavala, Maharastra, India Source: © Sunengy. Source: © Sunengy. the water’s surface or floating flexible modules direct- Submerged FPV still has a long way to go to prove ly on top of the water’s surface (figure 2.20). Aside its industrial relevance. Among the challenges to be from lower module temperatures, submerged FPV overcome are those related to long-term reliability and systems offer the advantage of reducing mechani- electrical safety. The electrical components are often cal load, especially from wind or currents, as well as in contact with water, so they must be able to resist internal stress from wave motions, thus simplifying moisture and corrosion. Another problem is adhesion mooring. Buoyancy for thin, flexible films (made of or accumulation of dirt or sediment on the panel sur- crystalline silicon or other thin-film materials) can be face, especially in the case of semi-submerged thin- supplied largely by the PV panels themselves, which film panels. When water dries in the sun, dirt can be typically are laminated with materials that encapsu- left on the surface of flat-lying panels. Bio-fouling is late air while resisting moisture. They can be installed another potential concern. Perhaps most importantly, easily from large rolls, which are also easy to trans- the actual performance gains of submerged FPV have port to the site. The total material usage for deploying yet to be demonstrated. self-buoyant PV modules is dramatically less than for conventional PV panels. 2.2.4.  Active cooling Some float suppliers also integrate pumping systems Submerged FPV has been explored and discussed into the floating platform to spray water onto the PV by several authors (Rosa-Clot and Tina 2018c; Tra- modules to cool them (figure 2.22). Sprinklers are pani and Millar 2014; Trapani and Redón Santafé triggered when the module temperature (as detected 2015). The first test system, with a 0.57 kilowatt-peak by sensors) reaches a certain threshold. After spray- (kWp) capacity, was deployed in 2010 in Sudbury, ing, the module temperature drops quickly, improving Canada, by MIRARCO Mining Innovation (Trapa- performance. ni and Redón Santafé 2015), shown in figure 2.21. Since then, several companies have tested systems This solution seems quite natural and sensible for FPV, using floating thin films. since water is readily available. In hot climates, this can indeed reduce temperature-related reductions in power output, which is a major loss factor. How- ever, since the pumping also consumes energy, the 44 •  FLOATING SOLAR MARKET REPORT FIGURE 2.20. Semi-submerged floating thin-film module and the forces to which it is exposed Buoyancy Wind lift Wind e Wav Tension Surface tension Tidal current Weight Source: Authors based on Trapani and Redón Santafé 2015. FIGURE 2.21. The 0.57 kWp MIRARCO Mining Innovation semi-submerged floating thin-film system in Sudbury, Canada Source: © MIRARCO Mining Innovation. CHAPTER 2:  TECHNOLOGY OVERVIEW  • 45 operation algorithm needs to be carefully optimized may necessitate a different platform design or the use to ensure a net energy gain. In addition, the improved of different technologies. However, the rich experience performance must be large enough to make the of marine and offshore industries should make it possi- investment worthwhile. Another problem, depending ble to meet the challenges. Compared to the open sea, on the water quality of the reservoir, is that soiling may areas such as lagoons and bays are relatively calm— occur over repeated cycles of spraying and drying. and thus more suitable for FPV installations. There are Currently, active cooling has been employed for only many such areas along the world’s coastlines, offering a few FPV systems, and no rigorous assessment of its a large potential market for FPV. Moreover, offshore or overall benefit is yet available. near-shore FPV may be the only way for small islands such as the Maldives to “go green” without having to 2.2.5.  Offshore/near-shore FPV clear scarce land to make room for ground-mounted PV installations. Offshore or near-shore FPV is conceptually similar to FPV on inland water bodies. However, offshore or near- Several companies are researching offshore and near- shore environments present some additional challeng- shore FPV solutions. Austria’s Swimsol has launched a es and difficulties: pilot plant in the Maldives (figure 2.23). Its system con- • Water surface conditions are much rougher because sists of 25 kWp modular platforms, each supported by waves and winds are higher. floating buoys and moored by helical anchors. Accord- ing to Swimsol, the platform can withstand waves two • Mooring and anchoring becomes even more critical meters high and winds of 120 km/h. due to tidal movements and currents. • The salinity of seawater is tougher on components. For its offshore FPV platforms, Norway’s Ocean Sun • Bio-fouling is much more likely. borrows the floating technology from offshore fishing farms, as shown in figure 2.24, using round floaters The more stringent requirements for floats, anchors, that can stretch a membrane over diameters up to and components imposed by the harsher environment 72 meters. The membranes, which are in permanent FIGURE 2.22. Active cooling solution Source: © Ciel & Terre International. 46 •  FLOATING SOLAR MARKET REPORT FIGURE 2.23. Swimsol’s pilot offshore FPV plant on a resort island in the Maldives Source: © Swimsol. FIGURE 2.24. Ocean Sun’s offshore floating platform in Norway, with membrane to hold PV panels Source: © Ocean Sun. contact with water, provide good thermal conduction Offshore or near-shore FPV is still in a nascent stage, of heat from the PV panels, effectively reducing the and practical experience is limited. The biggest uncer- operating temperature of the modules. In addition, the tainties are long-term reliability and costs. In general, membranes are strong enough to walk on (for installa- marine grade materials and electrical components tion and maintenance) while also being flexible enough are needed, and the structural design must withstand to accommodate waves. extreme weather. It is likely that PV modules, as well, would have to be reinforced for offshore conditions. CHAPTER 2:  TECHNOLOGY OVERVIEW  • 47 Operation and maintenance costs may also be higher Establishing synergy between hydroelectric dams than for inland FPV installations. and FPV plants to generate more electricity is becom- ing an attractive option for the operators of existing Hybrid operation with 2.3.  hydropower plants. In every case, total output from the hydropower plants hybrid system must meet grid dispatch demand. This can be achieved through the following adjustments: PV power generation is inherently variable owing to cloud cover and the diurnal cycle. Research is being • Electricity generated by the PV system is transmit- conducted to reduce ramping rates and smooth the ted to the hydropower substation. The PV system output from PV systems using energy storage sys- is treated as a nondispatchable virtual unit of the tems or other solutions. However, while the cost of hydropower plant. From the perspective of the utility-scale energy storage systems is dropping, it power grid, the hybrid system constitutes a sin- remains high. Maintenance and disposal of the stor- gle dispatchable source of power, analogous to age system after its shelf life also needs to be con- a conventional power plant (An and others 2015; sidered. One storage solution in common use today is Fang and others 2017; Gebretsadik 2016). pumped-storage hydropower, in which reservoirs are • The automatic generation control (AGC) system used as a storage system. When energy demand is of the hybrid system monitors the real-time out- low (e.g., overnight), water is pumped from a down- put power from the PV source, receives set points stream reservoir into an upstream reservoir behind a from the grid-dispatch center, and calculates the hydroelectric dam and then released through the dam total power set point for hydropower (Gong and to generate electricity at times of peak demand. others 2014). The AGC then determines the active power set points for each hydro unit. In the same spirit, the rise of FPV offers a new and promising alternative: combining FPV with hydropow- • In the short term, the hydropower plant can count- er stations. Being a kind of instantly adjustable ener- er-adjust its output through a small movement of gy source, hydropower has the potential to become guide vanes, smoothing the variable output curve a real-time compensator for variable PV power. The of the FPV system (An and others 2015). reservoirs behind hydroelectric dams can store water • In daily operation, the hydropower plant can adjust during periods of high irradiance and release it at the water level in the reservoir to compensate for cloudier times or when demand spikes and it there- the randomness of PV output. At times of high PV fore serves as a storage system of the hybrid solar and output and low system demand, the hydropower hydropower operation. units can reduce their output and store water in Apart from utilizing the reservoir surface, the combina- the reservoir. At times of low PV output and high tion also allows easy grid connection through the infra- system demand, the hydropower plant releases structure of the hydropower plant. This option is best water and increases its output. To meet the water conceived as a way to maximize the utility of existing requirements of other reservoir functions—such hydropower stations rather than as a way to justify the as irrigation, downstream environmental flows, building of new dams. Where rainfall patterns are highly and flood control—the daily water balance of the seasonal, as in monsoon areas, there is an additional reservoir should be maintained at its level before advantage of complementarity over the course of the the installation of the hybrid system. year: More solar power is generated during the dry sea- • During the rainy season, when water run-off is high, son (when water levels and hydropower output are low); hydropower plants with limited reservoir capacity the reverse is true for the rainy season. must operate at maximum output, and therefore will not be able to compensate for PV variations. Hybrid 48 •  FLOATING SOLAR MARKET REPORT operation will be ineffective under such conditions. (using solar energy during the day and hydropower Water will have to be spilled or PV power curtailed. at night), but also across the seasons. During the dry season, for example, when there is low water Hybrid hydropower and FPV systems can offer great storage and low hydropower output, the bright, sun- advantages in terms of grid integration, equipment ny weather allows for higher PV generation. PV thus utilization, and cost: makes up for the hydropower deficiency. Supported by the PV output, hydropower can dispatch electric- • Hybridization with hydropower improves the qual- ity in a more flexible manner. ity of PV power. As variable, nondispatchable PV power is at least partly converted to stable and dis- • Deploying PV systems on reservoir surfaces can patchable electricity, the consumption of PV power save on the cost of land. The existing road access rises, as do the profits of the developer of the PV to the hydropower plant likely reduces construc- power plant. From the point of view of the power tion and transportation costs, as well. system, stable and dispatchable PV power means The development of grid-connected hybrid hydro- lower requirements for spinning reserves and ener- power and FPV projects is still in the early stages. A gy storage, thus reducing the overall operational small FPV system of 220 kWp has been deployed on cost of the power system (An and others 2015). a hydropower dam in Portugal. Many other such proj- Another great advantage of hybrid operation is the ects, some large in scale, are being discussed or are benefit of making maximal use of existing electrical under development. infrastructure, including high voltage grid access and transformers, which can lower capital costs Currently, the world’s largest hybrid hydropower and and accelerate project implementation. solar PV project is one where the PV component is • Water resources and solar energy can compen- ground-mounted. The Longyangxia hydro/PV power sate for each other when operated together as a plant project in Qinghai, China, is striking in its size hybrid. This is true not only over the diurnal cycle and hence is a role model for such conjoint operation (Qi 2014; Zhang and Yang 2015). FIGURE 2.25. Satellite image of Longyangxia hybrid hydro/PV power plant Source: © Google Earth. CHAPTER 2:  TECHNOLOGY OVERVIEW  • 49 The Longyangxia hydropower plant was commis- 2.25). The first phase was built and commissioned in sioned in 1989 with four 320MW Francis turbine-gen- 2013 with a nameplate capacity of 320 MWp and aver- erator sets that generate 5.942 GWh of electricity age annual energy generation of 0.498 GWh. An addi- each year. The plant serves as the major load-peak- tional 530 MWp (Phase II) was completed in 2015. It ing and frequency-regulation power plant on Chi- is one of the largest solar PV installations in the world. na’s northwest power grid. The dam is located at the entrance of the Longyangxia canyon on the Yellow The PV power plant is directly connected to the River in Gonghe County, Qinghai Province. It provides reserved line inside the Longyangxia hydropower carryover storage and excellent multi-year regulation substation by a 330 kV transmission line. capability. The designed normal storage water level is 2,600 meters; the dead water level, 2,530 meters; the The hybrid system is operated in a complementary regulating storage, 193.5 × 108 cubic meters. manner (Zhang and Yang 2015). Figure 2.26 com- pares the total system output and the hydro output The associated Gonghe solar plant is located 30 kilo- before and after hybrid in a relatively dry year (Qi meters from the Longyangxia hydropower plant (figure 2014). After the PV plant was added, the grid opera- FIGURE 2.26. Before and after hybridization operation on a day in December in a dry year: hydropower output (top) and total system output (bottom). Power output 1,200 1,000 800 MW 600 400 200 0 1 2 3 3 5 4 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hours Power output 1,200 1,000 800 MW 600 400 200 0 1 2 3 3 5 4 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hours Hydro before hybridization Hydro after hybridization Hydro+PV after hybridization Source: SERIS based on Qi 2014. 50 •  FLOATING SOLAR MARKET REPORT tor began to issue a higher power dispatch set point reservoirs. All power generated by the hybrid system during daylight hours. The increased portion is shown is absorbed by the grid without curtailment. in the right side of the figure. Despite this, output from the hydro facility is lower than it was before the PV The hybrid operation closely follows power dispatch plant came on stream, especially from 11 am to 4 pm, set point on sunny days. On cloudy days the hybrid when PV generation is highest. The saved energy is operation compensates the variability of solar output then used during the early morning and late-night by using flexibility of hydropower production with the hours. Although the daily generation pattern of the maximum deviation within the limits required by the hydropower plant has changed, the daily water bal- dispatch operator. The deviation, together with other ance in the reservoir has been kept as it was in order variables depicting the hybrid operation, can be seen to meet the water requirements of other downstream in figure 2.27. FIGURE 2.27. Hybrid operation on a sunny day (top) and a cloudy day (bottom) during daylight hours. Hybrid system output on a sunny day 1,400 1,200 1,000 800 MW 600 400 200 0 9 10 11 12 13 14 15 16 Hour Hybrid system output on a cloudy day 1,400 1,200 1,000 800 MW 600 400 200 0 9 10 11 12 13 14 15 16 Hour Solar power Hydro power Grid set point Total power Source: SERIS based on Qi 2014. Note: Total power plot is not visible on the top graph since Grid set point plot mostly corresponds to the same values as Total power plot. CHAPTER 2:  TECHNOLOGY OVERVIEW  • 51 Field operations have shown that the hydropower tur- of four turbines exceeds 150 MW per minute, and the bines can provide adequate response to variations in maximum output deviation is 60 MW. These parame- demand and PV output. The active power ramp rate ters meet grid-dispatch requirements. References An, Y., and others. 2015. Theories and methodology of complementary hydro/photovoltaic operation: Applications to short- term scheduling. Journal of Renewable and Sustainable Energy 7(6): 063133. Connor, P.M. 2009. Performance and prospects of a lightweight water-borne PV concentrator, including virtual storage via hydroelectric-dams. Proceedings of the ISES Solar World Congress 2009: Renewable Energy Shaping Our Future. Fang, W., and others. 2017. Optimal sizing of utility-scale photovoltaic power generation complementarily operating with hydropower: A case study of the world’s largest hydro-photovoltaic plant. Energy Conversion and Management 136: 161–172. Gebretsadik, Y., and others. 2016. Optimized reservoir operation model of regional wind and hydro power integration case study: Zambezi basin and South Africa. Applied Energy 161: 574–582. Gong, C., and others. 2014. AGC control strategy and application in Longyangxia Hydro/PV complementary operation. Mechanical and Electrical Technique of Hydropower Stations 37(3): 63–64. Kim, S.-H., and others. 2016. Application of floating photovoltaic energy generation systems in South Korea. Sustainability 8(12): 1333. Mousazadeh, H., and others. 2009. A review of principle and sun-tracking methods for maximizing solar systems output. Renewable and Sustainable Energy Reviews 13(8): 1800–1818. Patil Sujay, S., M.M. Wagh, and N.N. Shinde. 2017. A review on floating solar photovoltaic power plants. International Journal of Scientific & Engineering Research 8(6). Planair and PITCO. 2017. Assessment of Floating Solar PV Potential for Pakistan; Task 1: Commercial Readiness of FSPV— Global Market and Performance Analysis. Unpublished report prepared for International Finance Corporation, Washing- ton, DC. Qi, S. 2014. The analysis of complementation in PV grid-connected plant of Longyangxia 320MWp. Engineering Depart- ment, Xi’an University of Technology. Rosa-Clot, M. 2018. Floating Tracking Cooling Concentrator (FTCC) System. http://www.scintec.it/ricerca/energia/ftcE.html. Rosa-Clot, M. and G.M. Tina. 2018a. Integration of water-based PV systems. In Submerged and Floating Photovoltaic Sys- tems (chapter 9). Academic Press. Rosa-Clot, M. and G.M. Tina. 2018b. Concentration systems and floating plants. In Submerged and Floating Photovoltaic Systems (chapter 6). Academic Press. Rosa-Clot, M. and G.M. Tina. 2018c. Submerged PV systems. In Submerged and Floating Photovoltaic Systems (chapter 4). Academic Press. Tina, G.M., M. Rosa-Clot, and P . Rosa-Clot. 2011. Electrical behavior and optimization of panels and reflector of a photovoltaic floating plant. Proceedings of 26th European Photovoltaic Solar Energy Conference and Exhibition, Hamburg, Germany. Trapani, K. and D.L. Millar. 2014. The thin film flexible floating PV (T3F-PV) array: The concept and development of the prototype. Renewable Energy 71: 43–50. 52 •  FLOATING SOLAR MARKET REPORT Trapani, K. and M. Redón Santafé. 2015. A review of floating photovoltaic installations: 2007–2013. Progress in Photovoltaics: Research and Applications 23(4): 524–532. World Bank Group, ESMAP and SERIS. 2019. “Where Sun Meets Water: Floating Solar Handbook for Practitioners.” Forthcoming. Washington DC: World Bank. Zhang, P. and T. Yang. 2015. The study of operation strategy of Longyangxia Hydro/PV complementary operation. Journal of North China University of Water Resources and Electric Power (Natural Science Edition) 36(3): 76–81. CHAPTER 2:  TECHNOLOGY OVERVIEW  • 53 HONG KONG SAR, CHINA © Water Supplies Department (WSD) of Hong Kong SAR, China 3 GLOBAL MARKET AND POTENTIAL Availability of floating solar 3.1.  3.1.1.  Global irradiation resource Global horizontal irradiation (GHI) generally decreas- Where does it make the most sense to harness the es as one moves away from the equator to the north sun’s energy using photovoltaic (PV) panels floating on and south. Looking at a color-coded map of long-term water? In this section, global hotspots for floating pho- average yearly GHI, this is indicated by a shift from tovoltaic (FPV) installations are assessed by looking at warm tones (pink and red) to cool ones (green and (i) global irradiation data and (ii) the locations of water blue) (see figure 3.1). The irradiation data used for bodies such as lakes, reservoirs, dams, and ponds. A this study were obtained from the Global Solar Atlas third factor, also key, is the availability of nearby electric provided by the World Bank Group and funded by power lines but such evaluation would require local- the Energy Sector Management Assistance Program ized prefeasibility studies. (ESMAP). Irradiation rates were calculated using atmo- spheric and satellite data, and considering the effects of terrain, with a spatial resolution of 1 km. Note that uncertainty ranges from about 3 percent to 10 percent, depending on the location. FIGURE 3.1. Average GHI levels around the world Source: Global Solar Atlas (https://globalsolaratlas.info), © World Bank Group (2019). Note: kWh/m2 = kilowatt-hour per square meter. 55 CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 55 3.1.2.  Availability of water bodies ment limit, but it is far from realistic due to feasibility and environmental concerns. Water bodies are of two main types: natural and man- made. This study considers only man-made reservoirs • A range of 1–10 percent of the total surface area and dams, using data from the Global Reservoir and is defined as “useable” for the purposes of this Dam Database (GRanD) compiled by Lehner and analysis. It is assumed that covering this much of others (2011b) and distributed by the Global Water the surface area would not have significant adverse System Project (GWSP) and the Columbia Universi- impacts on the environment (although in practice ty Center for International Earth Science Information this should be investigated specifically for each res- Network (CIESIN). Natural water bodies are not con- ervoir). This does not mean that a higher surface sidered here, for a couple of reasons: (i) to compile a coverage ratio could not be considered, as seen complete global list of natural water bodies (approx- in various realized projects worldwide, but this will imately 177 million) would be a cumbersome task, depend on the specificities of each reservoir. and, also (ii) environmental considerations that apply • This useable area was then multiplied by an area to these are different than for man-made water bodies. factor of 100 watt-peak per square meter (Wp/ m2), which is within the range reported by existing There are several databases of man-made water FPV projects.4 This results in a total installed peak bodies. For example, the Food and Agriculture Orga- capacity in gigawatt-peak (GWp). nization of the United Nations developed AQUASTAT, • In order to derive the potential electricity genera- a geo-referenced database of dams and associated tion, the installed capacity was multiplied by the reservoirs, which was used as an input when GRanD energy yield, using the local irradiation with a was set up. standardized assumption of an 80 percent perfor- mance ratio (PR). This analysis utilizes GRanD instead of the AQUASTAT database for the following reasons: The potential capacity and energy generation of FPV • Geo-referencing. Although the number of total reser- projects are summarized by continent in table 3.1. The voirs in the AQUASTAT database is higher, the num- continent of Antarctica is omitted because of its rela- ber with geo-coordinates is lower than in GRanD. tively low irradiation and low power demand. • Greater detail. GRanD offers more details on its If just 1 percent of man-made reservoir surfaces were data sources, selection criteria, and methods used used, FPV capacity could quickly reach 400 GWp, to compile and document data (FAO 2016). which is the total installed capacity of all conventional solar PV systems combined at the end of 2017. Even The following steps were carried out during the if only 10 percent of the surface area of every third assessment: man-made reservoir in the world were covered, the • Using geographic information system (GIS) data, all FPV market would represent a terawatt (1,000 GW) the selected water body vectors were charted onto scale market opportunity. That is before even tapping a global solar irradiation map, and the surface area the resource potential of the world’s natural landlocked of each water body was calculated. This resulted in water bodies or its oceans—which receive the majority a detailed list of the average irradiation potential of of the solar energy received on earth. the world’s man-made water bodies. • This average potential would only be realized if 100 The projects investigated include (i) Yamakura Dam Reservoir, 4.  percent of these water bodies’ surface were to be Japan; (ii) Umenoki, Japan; (iii) Agongdian Reservoir, Taiwan, China; (iv) Godley Reservoir, United Kingdom; and (v) Queen utilized. This is the theoretical maximum deploy- Elizabeth II, United Kingdom. 56 •  FLOATING SOLAR MARKET REPORT TABLE 3.1. Floating photovoltaic potential, capacity and energy generation by continent (man-made reservoirs and dams only) Total FPV capacity Total annual FPV energy output potential [GWp] potential [GWh/y] Total (% of water surface used (% of water surface used for surface area for PV installation) PV installation) available No. of water Continent [km2] bodies assessed 1% 5% 10% 1% 5% 10% Africa 101,130 724 101 506 1,011 167,165 835,824 1,671,648 Asia* 115,621 2,041 116 578 1,156 128,691 643,456 1,286,911 Europe 20,424 1,082 20 102 204 19,574 97,868 195,736 N. America 126,017 2,248 126 630 1,260 140,815 704,076 1,408,153 Oceania 4,991 254 5 25 50 6,713 33,565 67,131 S. America 36,271 299 36 181 363 58,151 290,753 581,507 Total 404,454 6,648 404 2,022 4,044 521,109 2,605,542 5,211,086 Source: SERIS calculations based on the Global Solar Atlas, © World Bank Group (2019) and the GRanD database, © Global Water System Project (2011). Notes: *Middle East is included in Asia. FPV = floating photovoltaic; GWh/y = gigawatt-hour per year; GWp = gigawatt-peak; km2 = square kilometer; PV = photovoltaic. FIGURE 3.2. FPV capacity potential worldwide based on total surface area available Source: SERIS based on the Global Solar Atlas, © World Bank Group (2019) and the GRanD database, © Global Water System Project (2011). Note: GWp = gigawatt-peak; kWh/m2/y = kilowatt-hour per square meter per year; kWp/m2 = kilowatt-peak per square meter; PV = photovoltaic. CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 57 The global potential is outlined in the global map pre- of more than 950 MWp, representing about 73 per- sented in figure 3.2, under the assumption that up to 10 cent of the world’s total. The remainder of the installed percent of man-made water surfaces are covered. The capacity is mainly spread between Japan (about 16 size of the circles indicates the size of the considered percent), the Republic of Korea (about 6 percent),Tai- reservoirs’ FPV potential. wan, China (about 2 percent), the United Kingdom (about 1 percent) whilst the rest of the world accounts 3.2.  Current market status for only 2 percent. FPV plants totaling more than 180 MWp have been installed to date in Japan; most of The world market for FPV has been surging over the them are below 3 MWp. past few years (as outlined in chapter 1), and the installed capacities of individual projects are increas- Figure 3.3 ranks the FPV projects listed in Table 3.5 ing year on year. The largest FPV systems in operation as well as projects smaller than 5 MWp based on their are in China, where two projects with capacities of installed capacity. Plants were divided into five cate- 150 megawatt-peak (MWp) each were developed by gories: (i) smaller than 2 MWp, (ii) between 2 and 3 Sungrow Group and China Three Gorges New Energy MWp, (iii) between 3 and 5 MWp, (iv) between 5 and 15 Co., Ltd. The global installed FPV capacity exceeded MWp, and (v) larger than 15 MWp. Most of the installa- 1.3 GWp as of December 2018 and has been grow- tions to date are small systems with capacities below ing exponentially since 2017. Table 3.2 lists the world’s 3 MWp. However, the number of large systems has largest FPV projects (with capacities of at least 5 MWp) been increasing significantly since 2017 and this trend completed as of December 2018. is set to continue, with many FPV projects larger than 10 MWp under development. The world’s 13 largest Market data suggests that with the installation of a plants (>15 MWp) account for more than 70 percent of few large FPV systems in the last two years China has all FPV installed capacity. become the FPV market leader with installed capacity FIGURE 3.3. Distribution of FPV plants according to their size, as of December 2018 1,400 1,200 13 projects UK 1% Others 2% Taiwan, China 2% 1000 934 Korea 6% 800 MWp Japan16% 600 400 China 73% 63 projects >200 projects 9 projects 200 127 152 6 projects 78 23 0 TOTAL 1,314 MWp <2 MWp 2–3 MWp 3–5 MWp 5–15 MWp >15 MWp >300 projects Cumulative installed capacity (MWp) Source: Authors’ compilation based on various external sources (public media releases and direct insights from industry representatives). Note: MWp = megawatt-peak. List of projects attempts to be exhaustive, but omissions might have occurred. 58 •  FLOATING SOLAR MARKET REPORT TABLE 3.3. Overview of largest (5 MWp and above) FPV installations in the world, ranked by size, as of December 2018 Floating system supplier(s) Com- Size (and subcontractor, pletion (kWp) Water body and nearest city Country City/Province if applicable) year 150,000 Coal mining subsidence area, China Anhui Province Beijing NorthMan, Zhongya, 2018 Huainan City (Panji—China Hefei Jintech New Energy Co. Three Gorges New Energy) Ltd., Anhui ZNZC New Energy Co. Ltd., CJ Institute China 150,000 Coal mining subsidence area, China Anhui Province Sungrow Floating (Anhui ZNZC 2018 Huainan City (Fengtai Guqiao— New Energy Technology Co. Ltd.) Sungrow) 130,000 Yingshang coal mining China Anhui Province Anhui ZNZC New Energy Tech- 2018 subsidence area (Liuzhuang nology Co. Ltd., Shanghai Qihua mine—Trina Solar) Wharf Engineering Co. Ltd, etc. 102,000 Coal mining subsidence area, China Anhui Province Sungrow Floating (Anhui ZNZC 2017 Huainan City (Fengtai Xinji) New Energy Technology Co. Ltd.) 100,000 Coal mining subsidence area, China Shandong Sungrow Floating 2018 Jining City Province 70,005 Mine lake, near Huaibei (China China Anhui Province Ciel & Terre International 2018 Energy Conservation and Environmental Protection (CECEP)) 50,000 Coal mining subsidence area, China Shandong Sungrow Floating 2017 Jining City (Shandong Weishan) Province 40,000 Renlou coal mine in Huaibei City China Anhui Province Shanghai Qihua Wharf 2017 (Trina Solar) Engineering Co. Ltd., etc. 40,000 Coal mining subsidence area, China Anhui Province Sungrow Floating 2017 Huainan City (20+20 Panji) 32,686 Mine lake (Golden Concord Ltd China Anhui Province Ciel & Terre International 2018 (GCL)) 31,000 Coal mining subsidence area, China Shandong Sungrow Floating 2017 Jining City (Shandong Weishan) Province 20,000 Coal mining subsidence area, China Anhui Province N/A 2016 Huainan City (Xinyi) 18,700 Gunsan Retarding Basin Korea, North Jeolla Scotra Co. Ltd. 2018 Rep. 13,744 Yamakura Dam reservoir Japan Chiba Ciel & Terre International 2018 10,982 Xuzhou Pei County China Jiangsu Province Ciel & Terre International 2017 9,087 Urayasu Ike Japan Chiba Ciel & Terre International 2018 8,500 Wuhu, Sanshan China Anhui Province N/A 2015 8,000 Lake in Xingtai, Linxi County China Hebei Province N/A 2015 7,550 Umenoki Irrigation Reservoir Japan Saitama Ciel & Terre International 2015 6,800 Hirotani Ike Japan Hyogo Takiron Engineering Co. Ltd. 2018 6,776 Amine Lake, Jining City China Shandong Ciel & Terre International 2018 Province 6,338 Queen Elizabeth II Drinking Water United London Ciel & Terre International 2016 Reservoir Kingdom Source: Authors’ compilation based on various external sources (public media releases and direct insights from industry representatives). Notes: N/A = not available; kWp = kilowatt-peak; MWp = megawatt-peak; PV = photovoltaic. List of projects attempts to be exhaustive, but omissions might have occurred. CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 59 FPV offers significant advantages in countries where 3.2.2. Bangladesh land is scarce or expensive, and suitable water bodies As of early 2019, two FPV systems are planned in are present. Some economies, such as Taiwan, China, the country, including one of 50 MW, which should offer financial incentives for the use of water bodies receive the support from the Asian Development Bank for PV deployment. Several large FPV installations are and will be built on Kaptai Lake in the Chittagong dis- integrated with hydropower plants. These arrange- trict (Islam 2019). ments increase the overall efficiency of both solar and hydropower production and allow the sharing of exist- 3.2.3. Belgium ing transmission infrastructure. The first 998 kilowatt-peak (kWp) FPV system was The following subsections consider the present and commissioned in early 2018 at Hesbaye Frost in Geer. future FPV capacity of selected countries (presented Other pilot FPV installations, including a 5 MWp system in alphabetical order), based on available literature, to be owned by Sibelco in Dessel,5 are being devel- including various online sources. The countries listed oped and supported at the subnational level, by the in this chapter have large installed FPV capacity, size- Flemish government (Bellini 2018d). Meanwhile, the able planned or tendered future FPV capacity, or are Belgian government is looking into the possibility of considering developing their near-shore and offshore building offshore FPV plants in the North Sea. potential. 3.2.4. Brazil 3.2.1. Albania The first FPV system completed in Brazil in September Statkraft is planning to build a 2 MW FPV system at its 2017 has a capacity of 305 kWp (figure 3.4). The sys- 72 MW Banja hydropower dam. This project might be tem was developed by Ciel & Terre International and is eligible for the feed-in tariff applicable in the country located on a rainwater accumulation pond in the state (Bellini 2019). The largest Albanian power producer, Korporata Elektroenergjitike Shqiptare (KESH), is plan- 5. https://www.sibelco.com/media/sibelco-is-supporting-sustainable- ning to develop a 12.9 MW FPV system (Jonuzaj 2018). energy/. FIGURE 3.4. FPV system (of 305 kWp capacity) in Goias, Brazil Source: © Ciel & Terre International. Note: kWp = kilowatt-peak. 60 •  FLOATING SOLAR MARKET REPORT of Goias. Ciel & Terre International is also involved in developed by Xinyi, Trina Solar and Sungrow) are the development of two other FPV plants of 4.99 MWp unique in the way that they used central inverters and each, in Balbina (Amazon region) and in Sobradinho transformers on dedicated floating pontoons, allow- (Bahia region) (Kenning 2017a). The first phase of ing for shorter direct current (DC) cabling (Planair these two projects started in early 2016, when 1 MWp and PITCO 2017). each was installed near the Balbina and Sobradinho hydroelectric power plants. The aim is to evaluate the Currently, the largest FPV plants in operation are two performance of two similar systems in different climat- 150 MWp projects, one completed by Three Gorg- ic conditions (Zaripova 2016). es New Energy Company and the other by Sungrow. Both are located in Anhui Province. Additional projects 3.2.5. Cambodia under China’s “Top Runner” program were disclosed in 2017, but how many of these will be realized remains A 2.8 MWp FPV system developed by Cleantech Solar to be seen (Bin 2018). with floats from Ciel & Terre International was com- pleted end of 2018 and commissioned early 2019 3.2.7. Colombia at Chip Mong Insee Cement Corporation industrial pond. The Natural Heritage Institute (NHI) evaluated A 99 kWp FPV system, consisting of two units, was the feasibility of installing a utility-scale FPV system recently completed in 2018. The system was deployed on the recently built 400 MW Lower Se San 2 Hydro- on the water reservoir of Peñol-Guatapé, owned and power Dam, as an alternative to the Sambor Dam and operated by Empresas Públicas de Medellín (EPM), hydroelectric power plant project (National Heritage the local energy and telecommunications utility of Institute 2017). Medellín (Bellini 2018a; Ciel & Terre International 2018). New large-scale projects are in the pipeline. 3.2.6. China 3.2.8. France Multiple pilot and small-scale projects were devel- oped in China before early 2016, when large projects In France, a flagship project, O’MEGA 1, is being started to take off. Since then, the country has seen developed in Piolenc in the department of Vaucluse. astounding growth: total installed capacity was more The 17 MWp project, developed by Akuo Energy, is than 950 MWp as of December 2018, surpassing under construction and expected to be completed in by far the combined capacity of all other countries March 2019. It is built on a former quarry lake and will in the world. The large majority of China’s FPV proj- be financed through nonrecourse project financing ects are located in Anhui Province and utilize lakes from Natixis Energeco (Kenning 2018a). Other large- formed when irregular depressions in the terrain scale FPV projects are currently under development caused by the collapse of mines flooded with rain- in Hautes Alpes and Bouches-du-Rhône regions. water. A further 400 MW was tendered in Shandong Province; this combines FPV with “PV over water,” 3.2.9. Ghana that is, installed on piles in shallow water. The winning In February 2018, Eni Ghana and Eni Energy Solutions bidders (among them Sungrow, Trina, GCL, Xinyi, signed two separate memorandums of understand- CECEP, and China Three Gorges New Energy) sell ing (MOUs) with Bui Power Authority, the company the generated electricity to the State Grid Corpora- responsible for the management of the 400 MW Bui tion of China at rates ranging from yuan (Y) 0.71 to Y Hydroelectric Power Project in Ghana, and Volta River 0.81 per kilowatt-hour (kWh) ($0.11–0.12/kWh). Many Authority, respectively. Both relate to the joint develop- of these large-scale FPV systems (including systems ment of power generation from renewable sources, including FPV systems (Eni.com 2018). CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 61 3.2.10. India ning 2017c). In 2018, SECI launched a tender for three FPV projects with 50 MWp capacity each at the India’s numerous hydropower plants have a total Rihand Dam, located in the Sonbhadra district of Uttar capacity of 44 GW (equivalent to 13.6 percent of India’s Pradesh. An upper ceiling tariff of Rs. 3.32 (~$0.0476)/ total energy output), offering tremendous potential for kWh has been fixed for this tender. Shapoorji Pallon- the integration of FPV and hydropower. India’s largest ji won package B (50 MWp) with a tariff of Rs. 3.29/ FPV plant to date is a 2 MWp FPV system, installed on kWh (Kabeer 2018). Additional projects are foreseen the Mudasarlova reservoir in Visakhapatnam (Andhra in Tamil Nadu, Jharkhand, and Uttarakhand. Pradesh) and developed by Greater Visakhapatnam Smart City Corporation Limited, a company created in In November 2017, the Lakshadweep Energy Devel- 2016 to implement various smart city projects (Prateek opment Agency invited developers to submit an 2018b). Another 3 MWp FPV project on the Meghad- expression of interest in FPV projects of 10 MWp in rigedda reservoir in Visakhapatnam, tendered in 2017 the Lakshadweep islands. In June 2018, the National by the Greater Visakhapatnam Municipal Corpora- Thermal Power Corporation launched a tender for a 22 tion, was recently awarded to ReNew Power Limited MW FPV system to be developed at the Rajiv Gandhi (ReNew Power 2018). The corporation is planning to Combined Cycle Power Plant in Kayamkulam in Ker- build another 15 MWp FPV system on the Meghad- ala. The project would be financed by the same cor- rigedda reservoir (Rao 2018). poration that launched it (Prateek 2018c). In the same month, the Irrigation and Water Resource Department Given the relative shortage of inexpensive land and of Uttar Pradesh issued a tender to develop 100 MW of very ambitious solar targets in the country, India’s FPV grid-connected canal-top solar PV (ground-mounted, project pipeline is growing fast, with many large-scale not FPV) projects under a public-private partnership projects under study. A 5 MWp plant is currently under model (Prateek 2018d). The Maharashtra State Elec- construction in the district of Murshidabad in West tricity Distribution Company is also looking at develop- Bengal following a turnkey engineering, procurement, ing 1 GWp of FPV at the Ujani Dam in Solapur District and construction tender won by International Coil Ltd. (Kenning 2018b). at a price of Rs. 269.12 million (about $4.1 million, or $0.83 per Wp) (Prateek 2018a). Many other new FPV tenders have been launched end of 2018 and early 2019, some of which by NTPC National Hydroelectric Power Corporation has announc- Limited, one of the largest power utilities in the coun- ed plans to set up a 600 MWp FPV at the 1,960 MW try, with the aim to build FPV systems at its existing Koyna Hydropower project, with an estimated capital power plants. Most recently, TANGEDCO from Tamil cost of $1,350–$1,500/MWp (Saurabh 2016). In addi- Nadu announced a plan to open tenders for 250 MW tion, an FPV project of 5 MWp has been planned by the of FPV systems on three dam reservoirs in the state mining and power firm NLC India, in the Andaman and (Sivakumar 2019). Nicobar Islands (Planair and PITCO 2017). Two projects totaling 40 MWp in Maharashtra and Kerala, funded by 3.2.11. Indonesia KfW Development Bank, were announced in 2016, with an investment of $44 million (Planair and PITCO 2017). Indonesia has significant FPV potential, and one of the major developers in the country, the Abu Dhabi– In December 2017, the Solar Energy Corporation of based Masdar Clean Energy, is looking ahead in that India (SECI) announced an expression of interest in direction. Many forested areas across the islands of 10 GWp of FPV on artificial bodies of water across Indonesia are not suitable for solar deployment, and the country, with the aim of gathering information on their feasibility through market consultations (Ken- 6. Exchange rate as of August 31, 2018. 62 •  FLOATING SOLAR MARKET REPORT FIGURE 3.5. FPV installation (with a capacity of 13.7 MWp) at the Yamakura Dam in Japan Source: © Kyocera TCL Solar LLC. Note: MWp = megawatt-peak. land prices are high. The company has identified more kWp Pontecorvo system located on an irrigation pond than 60 reservoirs that could host FPV plants. Recent- in Savona Province, completed by Ciel & Terre Inter- ly, Masdar signed a project development agreement national.7 with the local power utility PT Pembangkitan Jawa-Bali to build a 200 MWp FPV plant covering 225 hectares 3.2.13. Japan of the surface area of the Cirata Hydroelectric Plant Japan is the country with the longest history of Reservoir in West Java Province (Rambu Energy 2017; MW-scale floating PV installations. The first 20 kWp Publicover 2017b). The Asian Development Bank also project was completed in 2007 in Aichi, Japan, as performed a preliminary opportunity assessment of a research prototype by the National Institute of FPV in Sulawesi and Kalimantan by identifying six sites Advanced Industrial Science and Technology. It is esti- with a cumulative capacity of 975 MWp. It estimates mated that more than 180 MWp of FPV systems were FPV potential to be in the range of several gigawatts deployed in Japan by end 2018. FPV offers many ben- at similar sites (e.g., hydropower reservoirs, estuaries, efits to Japan given its mountainous and heavily for- bays) across the country. ested terrain (over 70 percent of its land is unsuitable for ground-mounted PV). Japan also has abundant 3.2.12. Italy water surfaces; over 200,000 agricultural reservoirs Several companies in Italy are pioneering the develop- are used for irrigation or rainwater retention, among its ment of FPV systems, such as Koine Multimedia, with its floating tracking cooling concentrator, and NRG 7. https://www.ciel-et-terre.net/essential_grid/floating-solar-system- Energia. One of the largest systems to date is the 343 pontecorvo-34320-kWp/. CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 63 FIGURE 3.6. Offshore FPV system in the Maldives Source: © Swimsol. many lakes, dams, and reservoirs (Planair and PITCO power reservoir in the Xe Kong Basin, the 290 MW Xe 2017). Japan used to benefit from a generous feed- Kaman 1 hydropower plant, as an additional source in tariff (FiT) scheme for both small- and large-scale of electricity generation. The plan was endorsed by solar PV systems (FiTs for FPV were the same as for the prime minister, who issued two directives (dated ground-mounted PV). However, the subsidy scheme February 16 and August 13, 2018) to the relevant line was revised in 2017, when FiTs were removed for solar ministries to adopt and implement its findings and systems larger than 2 MWp.8 recommendations to further the nation’s renewable energy development. The world’s second-largest FPV installation outside China is on the Yamakura water retention dam in Chi- 3.2.15. Malaysia ba Prefecture, with a capacity of 13.7 MWp (MI News About 78 lakes in peninsular Malaysia have been iden- Network 2017) (figure 3.5). tified as suitable for developing FPV systems (Reve 2015). Malaysia’s largest FPV project to date was built 3.2.14.  Lao People’s Democratic Republic by Cypark Renewable Energy, a subsidiary of Cypark A Japanese company, TSB Co. Ltd., plans to build Resources Berhad (CRB), on Ulu Sepri Dam, a reser- a 14 MWp FPV plant at Nongheo and Nahai water voir for drinking water. CRB partnered with Ciel & Terre ponds in Hadxaifong district, near Vientiane (Vien- International in this 270 kWp FPV installation, which tiane Times 2018). The Natural Heritage Institute was connected to the grid in November 2016 and developed a master plan (National Heritage Institute which benefits from the feed-in tariff that was in place 2018) that was submitted to the government of Lao at the time (PV Tech 2018b; Ciel & Terre Internation- PDR in February 2018. The master plan examines al 2018). CRB plans to build two more projects, each the deployment of FPV at the largest existing hydro- 30 MW, one at Terip Dam and one at Kelinchi Dam in Negeri Sembilan under a contract with Cove Suria (PV Tech 2018b). The company from Taiwan, China Tien 8. https://www.iea.org/policiesandmeasures/pams/japan/name- 30660-en.php. Ching Energy along with UMILE signed a memoran- 64 •  FLOATING SOLAR MARKET REPORT dum of understanding for a 48 MWp FPV project with 3.2.17.  The Netherlands an estimated investment of up to $90 million (Planair The Netherlands has 52,000 hectares of shallow inland and PITCO 2017). In December 2017, the winners waters that could potentially be used for FPV installa- of the country’s second large-scale solar PV auction tions. In 2017, a national consortium called “Zon op were announced; four of the winning projects are FPV, Water”, Sun on Water, was created, initiated by the Min- including a 9.99 MW to be developed by Coral Power istry of Infrastructure and Water management, and led Sdn Bhd in Manjung, Perak (Bernama 2018). These by the Solar Energy Application Center (SEAC). The projects are expected to start commercial operation consortium comprises more than 40 companies with between 2019 and 2020. In April 2018, Sarawak’s the aim to promote the development and installation of chief minister made a proposal to LONGi to explore 2 GWp of FPV in the Netherlands by 2023. The con- the possibility of developing FPV systems at dams and sortium is developing a series of projects, including rivers in the state (The Sun Daily 2018). one on the permitting regulatory framework as well as multiple testbeds in various environments (e.g., inland, 3.2.16.  Republic of Maldives sea, with varying levels of wave and wind exposure). In the Maldives, Swimsol developed the first offshore FPV systems in 2014. These offshore projects are of a One of these testbeds was initiated by Waterschap small scale and meant to be complementary to rooftop Rivierenland, the Dutch Water Authority partnering PV installations as a means to reduce island resorts’ with Dutch companies Blue21 BV, Hakkers BV, and reliance on expensive diesel generators (figure 3.6). the Photovoltaic Materials and Devices unit (PVMD) of In 2018, eight different platforms with a total capacity Netherlands’ Delft University of Technology (TU Delft). of 200 kWp have been installed at various locations by The consortium is named INNOZOWA (INNOvatieve the same company. New projects are currently being ZOn-pv op Water) and is financially supported by developed. Early 2019, the government published a the government-run Netherlands Enterprise Agency tender to install 5 MW of grid-tied solar PV systems in (Bellini 2018e). the Greater Male region. Even though floating PV is not being specified in the tender documents, such tech- In addition, the “Zon op Water” consortium installed four nology could be considered for future tenders. FPV systems of 50 kWp each on De Slufter, a contami- FIGURE 3.7. FPV installation (with a capacity of 1.85 MW) in Azalealaan, Netherlands Source: © Ciel & Terre International. Note: MWp = Megawatt-peak. CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 65 nated dredging depot in the Port of Rotterdam, to func- on a local reservoir near Lingewaard in the eastern tion as a pilot testbed. The companies involved include Netherlands (Doo-soon and Ha-yeon 2018) (figure 3.7). Wattco, Texel4Trading, Sunprojects, and Sunfloat. 9 If this pilot proves successful, as much as 100 MWp In September 2018, the water utility NV PWN Waterleid- of capacity could be developed at this site (Osborne ingbedrijf Noord-Holland launched a tender to select 2017b). The project aims to demonstrate the feasibility a contractor to design, supply, and install FPV systems of FPV installations on rough waters. The pilot is locat- with a combined capacity of 7 MWp at two different ed in a tough environment where the water is brack- locations (Andijk and Hoofddorp). Projects could be ish, contains many contaminants, and where wind and expanded in the future to 16.7 MWp (Tsanova 2018). waves with heights up to 1 meter are common. 3.2.18. Norway Also under the aegis of “Zon op Water,” a consortium Ocean Sun has successfully tested in Norway two off- formed by the Energy Research Centre of the Neth- shore floating PV systems, based on their hydroelastic erlands, the Netherlands Organization for Applied membrane concept. A third system of 100 kWp sup- Scientific Research, the Maritime Research Institute plies off-grid power (with back-up diesel generators) Netherlands, the Abu Dhabi National Energy Compa- since April 2017 to a large fish farm on the western ny PJSC, and the Dutch startup Oceans of Energy, coast of Norway. More projects are in the pipeline to announced in February 2018 that it would develop and power other fish farms in Norway, as well as install deploy an offshore FPV project (“Solar@Sea”) over the such system on hydropower dams. next three years. The panels’ performance will be test- ed in salt water and inclement weather conditions (Bel- 3.2.19. Panama lini 2018b). The testbed will be financially supported by the Netherlands Enterprise Agency. In February 2017, a 24 kWp project was completed and connected to the grid by Ciel & Terre Internation- A 780 MWp FPV system at De Krim Resort, Texel al10 on a water retention pond (figure 3.8). It consists Islands, has been tendered by Texel4Trading on a rain- water reservoir currently used to irrigate golf courses. 9. https://www.zonopwater.nl/ Additionaly, a 1.85 MWp FPV system was recently built 10. https://www.ciel-et-terre.net/essential_grid/fl/. FIGURE 3.8. FPV system (with 24 kWp capacity) at Miraflores near the Panama Canal Source: © Ciel & Terre International. Note: kWp = kilowatt-peak. 66 •  FLOATING SOLAR MARKET REPORT FIGURE 3.9. FPV system in Alto Rabagão in Portugal Source: © Pixbee/EDP S.A. of 96 solar panels located in a semi-closed recess shallow body of salt water separated from the sea by of the great Gatun Lake and close to the Miraflores an industrial estate. locks, on the Pacific side of the Panama Canal (Pan- ama Today 2017). 3.2.22. Singapore Launched in October 2016, Singapore operates the 3.2.20. Portugal world’s largest FPV testbed, with a total installed The first FPV project to be built at an existing hydro- capacity of 1 MWp (figure 3.10), located on Tengeh electric power station was at a dam at the mouth of Reservoir. The project was a collaborative initiative the Rabagão River in Montalegre, Portugal. The 220 by PUB, Singapore’s National Water Agency, and the kWp system occupies an area of around 2,500 m . 2 Singapore Economic Development Board (EDB). It The pilot project was initiated by Energias de Portu- was designed, implemented, and is operated by the gal (EDP) in 2015 and has been operational since the Solar Energy Research Institute of Singapore (SERIS) end of November 2016 (figure 3.9). The mooring of at the National University of Singapore. It enables this floating power plant was very challenging, as the observers to compare the performance of 10 FPV bottom of the reservoir is more than 60 meters deep, installations (100 kWp each) to each other and to a and the system must deal with a regular fluctuation in reference 20 kWp rooftop PV system that is mount- water level of up to 30 meters (Osborne 2017a). ed on top of an inverter room located on the bank of the reservoir. The testbed also allows to study its 3.2.21. Seychelles own environmental impacts on the reservoir, such as reduced evaporation as well as effects on water The first African utility-scale FPV tender was announced quality and biodiversity. A comprehensive monitor- in April 2018 for a 4 MW system on Mahé Island, in the ing system tracks more than 500 parameters in real Seychelles (Beetz 2018). The tender is organized by time, ranging from electrical to meteorological and the Seychelles Energy Commission with support from module-related factors. Inertia sensors track move- the African Legal Support Facility of the African Devel- ments of the floating structures along six degrees opment Bank and the Clinton Foundation. The project of freedom. The 10 subsystems (see table 3.3 for an will be one of the first in the world to be installed on a CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 67 FIGURE 3.10. SERIS FPV testbed (with a 1 MWp capacity) at the Tengeh Reservoir in Singapore 1b 5 2 3 4 1a 6 7 8 10 9 Source: © SERIS. Note: MWp = megawatt-peak. TABLE 3.3. Key elements of SERIS FPV testbed at the Tengeh Reservoir System Floating platform Modules* Tilt Other features 1a Floats and stainless steel Glass-glass, frameless 7° east Small water footprint 1b Pipes and aluminum Framed 7° east Small water footprint 2 Pure floats Framed 12° east and west Dual-pitch design 3 Pure floats Framed 12° east — 4 Pure floats Framed 12° east Active cooling 5 Individual float modules Framed and frameless, 10° east Good ventilation, wind glass-glass load adaptation 6 Pipes and metal structure Framed 5° east — 7 Floats and aluminum Framed multi-Si, half-cut 7° east Rigid structure multi-Si, bifacial mono-Si 8 Pure floats Framed multi-Si, bifacial 10° east — mono-Si 9 Pure floats Frameless, glass-glass 10° east and west — 10 Pure floats Framed multi-Si 15° east — Rooftop reference — Half-cut, bifacial mono-Si 7° east Free standing on system rooftop Source: SERIS. Note: *All systems use multi-Si modules unless otherwise stated. 68 •  FLOATING SOLAR MARKET REPORT overview) use largely different types of PV modules 3.2.23.  The Republic of Korea (including some bifacial and frameless glass-glass Together with Japan, Korea was one of the first coun- modules), inverters, and floating structures. One of tries to adopt FPV. As of December 2018, FPV proj- the systems includes an “active cooling” feature: ects with a combined capacity of more than 75 MWp water is sprayed on to the solar modules to cool them had been installed with many projects in the pipeline. down and thus improve their performance. About 90 percent of the country’s mines have been closed or abandoned, and where collapsed mines Singapore has great FPV potential, given its limited land have flooded, this means a huge FPV potential. Var- mass and the fact that water bodies make up about 8 ious floating technologies have been developed and percent of the surface area (quarries and reservoirs are tested by Korean companies such as the state-owned mainly used to capture rain water to generate potable water management company Korea Water Resources water). Ongoing studies are considering how much of Corporation (K-Water), the Korea Rural Community those water bodies can be sustainably utilized for FPV, Corporation, and the Korea East-West Power Corpora- without compromising natural habitats or the intended tion. One tracking design features a floating structure use of the reservoirs. Estimations of future FPV poten- that rotates on the water’s surface to cope with freezing tial are in the hundreds of megawatt-peak, to be grad- conditions and ice formation on the lake. ually developed over the years to come. In 2013, Korea introduced its first megawatt-scale FPV In September 2017, PUB launched tenders for engi- power plant, at the cooling water intake channel of the neering and environmental studies to be conducted for thermoelectric power plant in Dangjin-si (Planair and a potential 50 MWp FPV system in Tengeh Reservoir PITCO 2017). The world’s largest (18.7 MWp) FPV proj- and a 6.7 MWp FPV system in Upper Peirce Reservoir ect outside China was completed in June 2018 and is (PUB 2017). In December 2017, Linyang Energy and located on the Gunsan Retarding Basin in North Jeolla Sunseap signed a memorandum of understanding to (Scotra 2018). Another notable FPV installation has a collaborate on various renewable energy and energy capacity of 0.465 MWp and was developed by Solkiss efficiency projects, including FPV, in Singapore (Ken- in 2014. Called “the Sunflower,” its modules follow the ning 2017f). sun using patented rotating motors. According to one estimate, this technology enables a 16 percent increase In 2018, the Housing Development Board announc- in energy yield over static FPV modules (Quirke 2017). ed it would initiate a research program focused on developing FPV systems in coastal marine conditions K-Water is actively looking at using its reservoirs to (Tan 2018). In October 2018, EDB issued a request build FPV systems and envisions installing more than for information to explore the feasibility of building a 1 GWp by 2022. In early 2016, K-Water signed an commercial 100 MWp FPV project on the Kranji Res- agreement with LG Electronics to build FPV projects ervoir, where a private company (or consortium) would on ponds and reservoirs throughout Korea (Publicover be first chosen as the offtaker, and in a second stage 2017a). an independent power producer would be selected to build and own the system (Economic Development Also, the only state-owned Korean energy firm, Korea Board 2018). In November 2018, Sunseap announced Hydro & Nuclear Power, signed a memorandum of they will develop a 5 MWp near-shore FPV system, to understanding in February 2018 with renewable ener- be located along the Straits of Johor, with the support gy company Hwaseong Solar Energy for a 100 MWp of EDB. In April 2019, PUB called for an EPC tender for FPV plant on Hwaseong Lake, a man-made body of FPV deployment in the Bedok (1.5 MWp) and Lower water on Korea’s western shoreline. The state-owned Seletar Reservoirs (1.5 MWp) (PUB 2019). firm is reportedly investing $202 million in the project. CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 69 The PV panels would cover 8.3 percent of the lake’s of the reservoir. This has been the first step in a wider surface (Clover 2018). plan to set up FPV plants on various dams and res- ervoirs, governed by the Sri Lanka Mahaweli Authori- The Korea Rural Community Corp. reported plans to ty (Office of the Cabinet of Ministers 2016). The Asia install 280 MWp of FPV capacity over three sites by Power Management Group, jointly with the Ministry of 2019 (Publicover 2017a). In August 2018, Hyundai Mahaweli Development and Environment, is expect- Heavy Industries Green Energy Co. announced that ed to develop a series of FPV projects that could total it had signed a memorandum of understanding with about 2 GWp. KEPCO Plant Service & Engineering Co. to cooper- ate in establishing FPV power plants with a combined 3.2.25.  Taiwan, China capacity of 170 MWp (Ji-woong and Mira 2018). In Taiwan, China, stands to benefit a great deal from FPV. September 2018, Korea Western Power Co. signed Available land for ground-mounted PV is limited in this a memorandum of understanding with Ansan City to economy, where much of the land is used for agricul- build a 102.5 MW FPV system on Sihwa Lake in Ansan ture or is mountainous and forested. High demand by 2020. This lake also accommodates the world’s for available land, meanwhile, is pushing up rent and largest tidal power installation, totaling 254 MW (Yon- purchase prices. Taiwan, China, has implemented a hap News Agency 2018). FiT regime, updated in January 2019, that favors FPV installations (NT$4.5016, or about 14.6 cents per kWh) 3.2.24.  Sri Lanka over ground-mounted PV (NT$ 4.1094, or about $13.3 In March 2017, an international tender was announced cents per kWh) to promote the uptake of FPV (Ministry for a 100 MW FPV plant to be located on the Maduru of Economic Affairs 2019). A slightly higher FiT is avail- Oya Reservoir in the eastern part of the country (Ken- able for FPV projects connected to the high voltage ning 2017e). The plant would cover around 4 percent transmission grid.11 FIGURE 3.11. FPV installation (with a capacity of 3 MWp) Cheongpung Lake, Chungju Dam in Korea Source: © LSIS. 70 •  FLOATING SOLAR MARKET REPORT FIGURE 3.12. Floating solar installation in Taiwan, China (a typhoon-prone area) Source: © Sungrow. However, harsh weather conditions, such as Solar Energy Corp were awarded the right to build the typhoons, pose technical challenges for the imple- special zone (Chia-erh 2018). mentation of FPV in the country, hence increasing system costs. Indeed, high wind speeds can desta- 3.2.26. Thailand bilize and damage FPV systems, calling for addition- The Thai solar company SPCG announced that it al stress testing of structural components (Kenning would work with a Japanese renewable energy com- 2016a) (figure 3.12). pany, InterAct, to implement FPV in Thailand to power shrimp farms (Nikkei 2014). Also, Ciel & Terre Inter- Many developers are looking to develop FPV projects national recently opened a new float manufacturing in Taiwan, China. In partnership, Taiwan Power Co. facility in Thailand with a maximum annual production and Taiwan Water Co. plan to install FPV systems on of 50 MWp (Kenning 2017b). SCG Chemicals signed eight reservoirs in Chiayi County (Planair and PITCO a memorandum of understanding with the main utility, 2017). New Green Power and J&V Holding have also Electricity Generating Authority of Thailand (EGAT), to announced a joint 20 MWp FPV project on an irrigation work collaboratively in the research and development pond in Taoyuan County (Kenning 2017d). of a mooring system for an FPV farm on the utility’s reservoirs and dams (SCG Chemicals 2018). SCG In July 2018, the Ministry of Economic Affairs announc- Chemicals has become Thailand’s first company to ed the development of a 320 MW special zone for successfully design and manufacture an FPV sys- solar power at the Changhua Coastal Industrial Park tem; this has a 979 kWp capacity (All Around Plastics that would also feature FPV systems. Chenya Energy 2018) and is situated on an industrial pond at Ray- Co., Yeheng Energy, and a major subsidiary of Taiwan ong’s Map Ta Phut Industrial Estate. EGAT recently announced it will facilitate the development of 2.7 11. Tariffs valid for the first half of 2019 and FX exchange rate as of January 31, 2019. GWp of hybrid floating solar-hydro projects across CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 71 FIGURE 3.13. FPV project on the Queen Elizabeth II Reservoir, United Kingdom Source: © Lightsource BP Floating Solar Array, London. nine dams throughout the country. Two projects, of 45 Europe’s largest FPV project is located on the Queen and 24 MW respectively, are already in the develop- Elizabeth II Reservoir run by Thames Water and fund- ment phase (Kenning 2018d). ed, built, and operated by Lightsource BP and Ennovi- ga Solar (figure 3.13). It has a capacity of 6.3 MWp 3.2.27. Ukraine and was in 2016 one of the first FPV projects to be installed on a deep-water (18.4 meters) reservoir (Ciel The UK-based asset manager Touchstone Capital & Terre International 2018). Ciel & Terre International Group Holdings Ltd. is looking at developing a hybrid designed the system and supplied the floating pon- 1.3 GWp wind-solar power project at the Kakhovka toons. Power generated by the site covers 20 percent water reservoir, located on the Dnieper River, alongside of the water treatment facility’s total energy demand the Kakhovka Hydroelectric Power Plant. It would com- (PV Tech 2018a). prise 1 GWp of wind and up to 300 MWp of FPV, with installations located between the wind turbines and The United Kingdom’s second-largest FPV project is anchored to their foundations (Bellini 2018c). located at the Godley Reservoir in Hyde (Hill 2015). This 2.99 MWp project (a £3.5 million investment) 3.2.28.  United Kingdom developed by United Utilities was a bid to hedge a Limited space available for land-based PV and high water treatment facility against increased power pric- on-site energy demand from water treatment plants es. It is able to provide 33 percent of the facility’s elec- are two key reasons for developing FPV in the United tric energy requirements (Energy Matters 2015). Kingdom. Several 100–200 kW FPV power plants have also been built on farms’ irrigation reservoirs. A start-up, AqvaFloat, is also launching a pontoon manufacturing facility in the United Kingdom with a capacity equivalent of 12 MWp (Parnell 2018). 72 •  FLOATING SOLAR MARKET REPORT FIGURE 3.14. FPV project in Orlando, Florida, United States Source: © Ciel & Terre International. 3.2.29.  United States Agency (Sonoma County Gazette 2015). Pristine Sun is leasing the ponds for about $30,000 per year The world’s first commercial FPV system of 175 kWp (Brown 2015). is since 2008 on an irrigation pond at the Far Niente Winery in Napa Valley, California. The pond and an Ciel & Terre International is also building four different adjacent land system are integrated, with 994 panels FPV systems in the country, totaling 5.3 MW. One of on the pond and 1,302 on land, covering 2.5 acres of them, a 252-kWp FPV system on a wastewater treat- space in total and producing more energy than the ment pond in Kelseyville (California) was completed in winery needs (Margaronis 2013). September 2018 (Osborne 2018). A 75-kWp FPV sys- tem was recently installed by GRID Alternatives Colo- A 4.4 MWp FPV system was completed in 2016 in rado at the drinking water treatment facility of the town Sayreville, New Jersey, and produces electricity for the of Walden (Jackson County, Colorado) (Grid Alterna- Bordentown Avenue Water Treatment Plant. The Orlan- tives 2018). do Utilities Commission in Florida has also a strong interest in FPV. A 31.5 kW system was built in 2017 on 3.2.30. Vietnam one of its storm water storage reservoirs (figure 3.14) with the support of D3Energy and Ciel & Terre Interna- In a bid to encourage the large-scale implementation tional (Pickerel 2017). of renewable energy technologies, Vietnam’s Minis- try of Industry and Trade established a FiT scheme In the coming years, Sonoma Clean Power in Cal- (Decision No. 11/2017/QD-TTg of the prime minister, ifornia is building a 12.5 MWp project, contracting passed on April 11, 2017 and taking effect from June with San Francisco–based Pristine Sun to build solar 1, 2017) for utility-scale solar installations, that also systems to be mounted on docks across six waste- applies to FPV projects. It was set to expire on June water ponds operated by the Sonoma County Water 30, 2019, but was extended in Ninh Thuan Province by CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 73 another 12 months (with a cap of 2 GWp). Grid-con- two FPV projects in Vietnam, each with a capacity nected power plants are granted a FiT equivalent to of 40–50 MWp (Kenning 2018c). In addition, Ciel about $0.0935/kWh, not counting value added tax 12 & Terre International has opened a manufacturing (Baker McKenzie 2018). An updated policy on FiT is facility in Vietnam (Kenning 2018c). A 47.5 MWp FPV being drafted and should be finalized by June 2019 project initiated by the Da Nhim-Ham Thuan-Da Mi when the current FiT will expire. Hydropower Joint Stock Company is under construc- tion on the reservoir of its 175 MW Da Mi hydro pow- Given the availability of freshwater bodies in Vietnam, er plant in Binh Thuan province, with the financial and land constraints, there is ample room for the support of the Asian Development Bank (ADB 2018). implementation of FPV systems. Many companies have announced plans to expand the country’s FPV In 2017 the Korean company Solkiss announced its potential. For example, Vasari Energy, a California- intention to develop a 500 MWp FPV plant in Yen Bai based green tech company, is planning to develop Province in southern Vietnam (Clover 2017a; 2017b). 12. /https://www.lexology.com/library/detail.aspx?g=530843be-3857- 4c97-a07d-e59db9a8a3a7 74 •  FLOATING SOLAR MARKET REPORT References ADB (Asian Development Bank). 2018. “Viet Nam: Floating Solar Energy Project.” https://www.adb.org/projects/51327-001/ main#project-pds. All Around Plastics. 2018. “Illuminating the Future—Sunlight and Innovation for Business Sustainability.” Innovation, All Around Plastics, May 8. http://www.allaroundplastics.com/en/article/innovation-en/1818. Baker McKenzie. 2018. “Vietnam’s Solar Feed-in-Tariff for Ninh Thuan Province Officially Extended for a Total Capacity of 2,000 MW.” Lexology, September 12. https://www.lexology.com/library/detail.aspx?g=530843be-3857-4c97-a07d-e59d- b9a8a3a7 Beetz, B. 2018. “Africa Announces Utility-Scale Floating Solar Tender.” PV Magazine, April 9. https://www.pv-magazine. com/2018/04/09/africa-announces-utility-scale-floating-solar-tender/. Bellini, E. 2018a. “Colombia to Host Its First Floating PV Project.” PV Magazine, April 19. https://www.pv-magazine. com/2018/04/19/colombia-to-host-its-first-floating-pv-project/. ———. 2018b. “Dutch Consortium Plans World’s First ‘Off-shore’ Floating PV Plant for the North Sea.” PV Magazine, Febru- ary 7. https://www.pv-magazine.com/2018/02/07/dutch-consortium-plans-worlds-first-off-shore-floating-pv-plant-for-the- north-sea/. ———. 2018c. “Floating PV Goes Everywhere.” PV Magazine, October 3. https://www.pv-magazine.com/2018/10/03/float- ing-pv-goes-everywhere/. ———. 2018d. “Belgium: Flemish Government to Support Pilot Floating Solar Projects with €6 Million.” PV Magazine, April 3. https://www.pv-magazine.com/2018/04/03/belgium-flemish-government-to-support-pilot-floating-solar-projects-with-e6- million/. ———. 2018e. “Netherlands: Newly Created Consortium to Develop Floating PV.” PV Magazine, January 11. https://www. pv-magazine.com/2018/01/11/netherlands-newly-created-consortium-to-develop-floating-pv/. ———. 2019. “Statkraft plans its first floating solar plant in Albania” PV Magazine, January 15. https://www.pv-magazine. com/2019/01/15/statskraft-plans-its-first-floating-solar-plant-in-albania/. Bernama. 2018. “TNB to Buy Solar-Generated Power from Two Perak-Based Facilities.” New Straits Times, March 30. https://www.nst.com.my/news/nation/2018/03/351151/tnb-buy-solar-generated-power-two-perak-based-facilities. Bin, L. 2018. “China’s Solar Industry Is at a Crossroads.” Chinadialogue, August 13. https://www.chinadialogue.net/article/ show/single/en/10775-China-s-solar-industry-is-at-a-crossroads. Brown, M. 2015. “Sonoma Clean Power Inks Deal for Floating Solar Panel Project.” The Press Democrat, February 26. Accessed on March 9, 2018. http://www.pressdemocrat.com/news/3580954-181/sonoma-clean-power-inks-deal. Chi Lieu, D. 2018. “Vietnam Plans to Extend FiT Scheme for Solar Power Projects in Ninh Thuan Province until the End of 2020.” Lexology, July 6. https://www.lexology.com/library/detail.aspx?g=d2006854-8459-4a43-a2d1-34beb01e6e56. Chia-erh, K. 2018. “Special Zone for Solar Power Set Up in Changhua.” Taipei Times, July 18. http://www.taipeitimes.com/ News/biz/archives/2018/07/18/2003696863. Ciel and Terre. 2017. “Goias Farm—305 kWp—Floating Solar Power Plant in Brazil.” https://www.cieletterre.us/project/ goias-farm-305-kwp-floating-solar-power-plant-brazil/ ———. 2018. “Our References: The Floating Solar Expert with Hydrelio Technology.” https://www.ciel-et-terre.net/wp-content/ uploads/2018/07/CT-References-May-Jun.-2018.pdf. Clover, I. 2017a. “South Korean Company Builds $1.1b Floating Solar Plant in Vietnam.” Pakistan Defence, March 27. Accessed March 9, 2018. https://defence.pk/pdf/threads/south-korean-company-builds-1-1b-floating-solar-plant-in- vietnam.485941/. CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 75 ———. 2017b. “Vietnam Province Taps Korean Government for Assistance on 500 MW Solar Project.” PV Magazine, March 27. https://www.pv-magazine.com/2017/03/27/vietnam-province-taps-korean-government-for-assistance-on-500-mw- solar-project/. ———. 2018. “South Korean State Energy Firm Plans 100 MW Floating PV Project.” PV Magazine, February 12. https://www. pv-magazine.com/2018/02/12/south-korean-state-energy-firm-plans-100-mw-floating-pv-project/. Doo-soon, K., and L. Ha-yeon. 2018. “Hanwha Q Cells to Build Floating Solar Power Plant in Netherlands.” Pulse, March 10. https://pulsenews.co.kr/view.php?year=2018&no=228807. Economic Development Board. 2018. “Request for Information: Large-Scale Floating Solar Photovoltaic System.” Economic Development Board, Singapore, October 31. https://www.edb.gov.sg/content/dam/edbsite/news-and-resources/news/ solar-pv-system/request-for-information-large-scale-floating-solar-pv-system-for-private-sector-consumption.pdf. Energy Matters. 2015. “Huge Floating Solar Power System on UK Reservoir.” Energy Matters, October 29. Accessed on March 9, 2018. https://www.energymatters.com.au/renewable-news/floating-solar-uk-em5156/. Eni.com. 2018. “Eni to Step up Cooperation in Renewable Energy in Ghana.” Eni.com, February 13. https://www.eni.com/ en_IT/media/news/2018/02/eni-to-step-up-cooperation-in-renewable-energy-in-ghana#. FAO (Food and Agriculture Organization). 2016. “AQUASTAT website.” http://www.fao.org/nr/water/aquastat/dams/index.stm. Global Solar Atlas: https://globalsolaratlas.info/, World Bank Group (2019). Grid Alternatives. 2018. “Unveiling Colorado’s First Floating Solar Array.” Grid Alternatives, October 23. https:// gridalternatives.org/regions/colorado/news/unveiling-colorado%E2%80%99s-first-floating-solar-array. Hill, J. S. 2015. “Construction Begins on Europe’s Largest Floating Solar Plant.” Clean Technica, 0ctober 28. Accessed on March 9, 2018. https://cleantechnica.com/2015/10/28/construction-begins-europes-largest-floating-solar-plant/. Islam, S. 2019. “Bangladesh prepares plans for two floating solar plants.” PV Magazine, January 17. https://www.pv- magazine.com/2019/01/17/bangladesh-prepares-plans-for-two-floating-solar-plants/. Ji-woong, M., and C. Mira. 2018. “HHI Green Energy and KEPCO KPS Join Hands in Floating Solar Projects.” Pulse, August 1. https://pulsenews.co.kr/view.php?year=2018&no=483381. Jonuzaj, K. 2018. “Albania’s KESH seeks building permit for 12.9 MWp floating solar park.” Renewables now, December 4. https://renewablesnow.com/news/albanias-kesh-seeks-building-permit-for-129-mwp-floating-solar-park-635433. Kabeer, N. 2018. “Shapoorji Pallonji Wins SECI’s 50 MW Floating Solar Auction Quoting ₹3.29/kWh.” Mercom India, Novem- ber 27. https://mercomindia.com/shapoorji-pallonji-wins-seci-50-mw-floating-solar-auction/. Kenning, T. 2016a. “20GW by 2025: Behind Taiwan’s Big Solar Numbers.” PV Tech, November 2. Accessed on March 9, 2018. https://www.pv-tech.org/features/20gw-by-2025-behind-taiwans-big-solar-numbers. ———. 2017a. “Brazil’s First Floating Solar Plant Completed by Ciel & Terre.” PV Tech, September 6. https://www.pv-tech. org/news/brazils-first-floating-solar-plant-completed-by-ciel-terre. ———. 2017b. “Ciel & Terre to Launch Floating Solar Manufacturing in Thailand Next Year, Vietnam to Follow.” PV Tech, November, 16. https://www.pv-tech.org/news/ciel-terre-to-launch-floating-solar-manufacturing-in-thailand-next-year-vie. ———. 2017c. “India’s SECI Invites Expression of Interest for 10GW of Floating Solar.” PV Tech, December 19. https://www. pv-tech.org/news/indias-seci-invites-expression-of-interest-for-10gw-of-floating-solar. ———. 2017d. “NGP to Build 14MW Floating Solar Plant in Taiwan.” PV Tech, October 20. Accessed on March 9, 2018. https://www.pv-tech.org/news/ngp-to-build-14mw-floating-solar-plant-in-taiwan. ———. 2017e. “Sri Lanka to Tender 100MW Floating Solar Plant, Funds Module R&D.” PV Tech, March 3. https://www. pv-tech.org/news/sri-lanka-to-tender-100mw-floating-solar-plant-funds-module-rd. ———. 2017f. “Sunseap and Linyang Energy Partner for 500MW Rooftop and Floating Solar in Singapore.” PV Tech, Decem- ber 12. https://www.pv-tech.org/news/sunseap-and-linyang-energy-partner-for-500mw-rooftop-and-floating-solar-in. 76 •  FLOATING SOLAR MARKET REPORT ———. 2018a. “Akuo Energy Starts Construction on 17MW Floating Solar Plant in France.” PV Tech, September 25. https:// www.pv-tech.org/news/akuo-energy-starts-construction-on-17mw-floating-solar-plant-in-france. ———. 2018b. “Maharashtra Discom Consults on 1GW of Floating PV, Retenders 1GW of Solar.” PV Tech, April 11. https:// www.pv-tech.org/news/maharashtra-disocm-consults-on-1gw-of-floating-pv-retenders-1gw-of-solar. ———. 2018c. “Vasari Plans 180MW of Ground-Mount and Floating Solar in Vietnam.” PV Tech, January 22. Accessed on March 9, 2018. https://www.pv-tech.org/news/vasari-plans-180mw-of-ground-mount-and-floating-solar-in-vietnam. ———. 2018d. “Thailand utility eyes 1GW of floating solar on hydro dams, pilots energy storage” PV Tech, November 14. https://www.pv-tech.org/news/thailand-utility-eyes-1gw-of-floating-solar-on-hydro-dams-pilots-energy-sto. Lehner, B., C. Reidy Liermann, C. Revenga, C. Vörösmarty, B. Fekete, P . Crouzet, P. Döll, M. Endejan, K. Frenken, J. Magome, C. Nilsson, J.C. Robertson, R. Rodel, N. Sindorf, and D. Wisser. 2011a. “High-Resolution Mapping of the World’s Reser- voirs and Dams for Sustainable River-Flow Management.” Frontiers in Ecology and the Environment 9 (9): 494–502. ———. 2011b. “Global Reservoir and Dam Database, Version 1 (GRanDv1): Dams, Revision 01.” NASA Socioeconomic Data and Applications Center (SEDAC), Palisades, NY. http://sedac.ciesin.columbia.edu/data/collection/grand-v1. Margaronis, S. 2013. “Far Niente Winery’s Floating Solar Experiment.” RBTUS, October 19. MI News Network. 2017. “World’s Largest Floating Solar Power Plant to Be Built in Japan.” Marine Insight, January 19. Accessed on March 9, 2018. https://www.marineinsight.com/shipping-news/worlds-largest-floating-solar-power-plant-to- be-built-in-japan/. Ministry of Economic Affairs. 2018. “2018 Renewable Energy FIT Rates Are Announced.” Bureau of Energy, Ministry of Economic Affairs, Taiwan, January 25. https://www.moea.gov.tw/MNS/english/news/News.aspx?kind=6&menu_ id=176&news_id=76457. National, The. 2018. “Japan’s biggest floating solar plant sparks into life.” April 3. https://www.thenational.ae/business/energy/japan-s-biggest-floating-solar-plant-sparks-into-life-1.718330. National Heritage Institute. 2017. “Sambor Hydropower Dam Alternatives Assessment.” https://n-h-i.org/programs/restor- ing-natural-functions-in-developed-river-basins/mekong-river-basin/cambodia-sambor/. National Heritage Institute. 2018. “Sustainable Hydropower Master Plan for Xe Kong Basin in Lao PDR.” https://n-h-i.org/ programs/restoring-natural-functions-in-developed-river-basins/mekong-river-basin/laos-xe-kong-2/. Nikkei. 2014. “Floating Solar Panels to Cut CO2 at Thai Shrimp Farms.” Asian Review-Nikkei, November 15. Office of the Cabinet of Ministers, Sri Lanka. 2016. “Making Use of the Mahaweli Economic Zones for the Development of Renewable Energy.” Press briefing of Cabinet Decision taken on November 15, 2016. http://www.cabinetoffice.gov.lk/cab/ index.php?option=com_content&view=article&id=16&Itemid=49&lang=en&dID=7238. Osborne, M. 2017a. “First Ever Hydro-Electric and Floating Solar Project Operating in Portugal.” PV Tech, July 27. https:// www.pv-tech.org/news/first-ever-hydro-electric-and-floating-solar-project-operating-in-portugal. ———. 2017b. “Floating Solar Pilot Projects in the Netherlands Set Sail.” PV Tech, July 17. https://www.pv-tech.org/news/ floating-solar-pilot-projects-in-the-netherlands-disembark. ———. 2018. “Ciel & Terre Building Four Floating Solar Projects in US.” PV Tech, September 18. https://www.pv-tech.org/ news/ciel-terre-building-four-floating-solar-projects-in-us. Panama Today. 2017. “Floating Solar Panels: A Panama Canal Green Project.” Panama Today, November 25. https://www. panamatoday.com/panama/floating-solar-panels-panama-canal-green-project-5836. Parnell, J. 2018. “UK Start-Up Eyes Global Floating Solar Market.” PV Tech, April 13. https://www.pv-tech.org/news/uk- start-up-eyes-global-floating-solar-market. Pickerel, K. 2017. “Trial floating solar installation in Orlando is first of hopefully many for Florida utility” Solar Power World, May 2. https://www.solarpowerworldonline.com/2017/05/trial-floating-solar-installation-orlando-first-hopefully-many- florida-utility/. Planair and PITCO. 2017. Assessment of Floating Solar PV Potential for Pakistan; Task 1: Commercial Readiness of FSPV— CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 77 Global Market and Performance Analysis. Unpublished report prepared for International Finance Corporation, Washington, DC.. Prateek, S. 2018a. “Breaking: West Bengal Auctions 5 MW of Floating Solar PV.” Mercom India, March 16. https:// mercomindia.com/west-bengal-auctions-5-mw-floating-solar-pv/. ———. 2018b. “India’s Largest 2 MW Floating Solar Project Commissioned in Greater Visakhapatnam.” Mercom India, August 28. https://mercomindia.com/india-largest-2mw-floating-solar-vishakhapatnam/. ———. 2018c. “NTPC Tenders 22 MW of Grid-Connected Floating Solar Project in Kerala.” Mercom India, June 25. https://mercomindia.com/ntpc-tenders-22mw-floating-solar-kerala/. ———. 2018d. “Uttar Pradesh Issues Tender for Development of 100 MW of Canal Top Solar PV Projects.” Mercom India, June 7. https://mercomindia.com/uttar-pradesh-tender-100mw-canal-top-solar-projects/. PUB. 2017. “PUB Studying Clean Energy Solutions from Blue Spaces.” Press release PUB, September 28. https://www.pub. gov.sg/news/pressreleases/pubstudyingcleanenergysolutionsfrombluespaces. ———. 2018. “PUB Looks to Bedok and Lower Seletar Reservoirs for More Solar Power.” Press release, PUB, April 29. https://www.pub.gov.sg/news/pressreleases/publookstobedokandlowerseletarreservoirsformoresolarpower. Publicover, B. 2017a. “Korean Firm to Build 280 MW of Floating PV Capacity.” PV Magazine, July 10. https://www.pv- magazine.com/2017/07/10/korean-firm-to-build-280-mw-of-floating-pv-capacity/. ———. 2017b. “Masdar to Build 200 MW Floating PV Array in Indonesia.” PV Magazine, November 29. Accessed March 9, 2018. https://www.pv-magazine.com/2017/11/29/masdar-to-build-200-mw-floating-pv-array-in-indonesia/. PV Tech. 2018a. “Europe’s Floating PV Pathfinder.” PV Tech Power. PV Tech. p. 13. ———. 2018b. “Malaysia’s First Grid-Connect Floating PV Project: A Matter of National Security.” PV Tech Power. PV Tech. p. 12, 13. ———. 2018c. “ROUND-UP: Colombia’s ‘First’ Floating Plant, Iran’s Khorasan Gets 10MW, Clenergy East Asia Orders.” https://www.pv-tech.org/news/round-up-colombias-first-floating-plant-irans-khorasan-gets-10mw-clenergy-e. Quirke, J. 2017. “South Korea to Build World’s Largest Rotating Floating Solar Plant.” Global Construction Review, July 12. http://www.globalconstructionreview.com/news/south-korea-build-worlds-largest-rotating-floating/. Rambu Energy. 2017. “Indonesia to Have Its First and Largest Floating Solar Power by 2019.” Rambu Energy, December 2. https://www.rambuenergy.com/2017/12/indonesia-to-have-its-first-and-largest-floating-solar-power-by-2019/. Rao, U. 2018. “GVMC to Set up 15 MW Floating Solar Project on Meghadrigedda.” Times of India, October 24. https:// timesofindia.indiatimes.com/city/visakhapatnam/gvmc-to-set-up-15-mw-floating-solar-project-on-meghadrigedda/article- show/66354338.cms. ReNew Power. 2018. “ReNew Power Wins 3 MW Floating Solar PV Project at Visakhapatnam.” ReNew Power, October 18. https://renewpower.in/wp-content/uploads/2018/10/ReNew-Power-Wins-3-MW-Floating-Solar-PV-Project-at- Visakhapatnam.pdf. Reve. 2015. “Potential for Floating Photovoltaic Solar Energy in Malaysia.” Reve, June 4. Accessed on March 6, 2018. https://www.evwind.es/2015/06/04/potential-for-floating-photovoltaic-solar-energy-in-malaysia/52537. Saurabh. 2016. “Indian Hydro Power Company Plans 600 MW Floating Solar Power Project.” SCG Chemicals. 2018. “SCG Joins Hands with EGAT to Develop Floating Solar Farm’s Mooring System for Power Genera- tion.” Press release, July 25. https://www.scgchemicals.com/en/news-media/news-events/press-release/detail/431. Scotra. 2018. “It Is the Case of a SCOTRA Floating PV System Installed.” Scotra, June 22. http://scotra.co.kr/bbs/board. php?bo_table=en_case01&wr_id=63. Sivakumar, B. 2019. “Tangedco aims for 250MW solar floating projects.” The Economic Times, January 25. https://energy. economictimes.indiatimes.com/news/renewable/tangedco-aims-for-250mw-solar-floating-projects/67682064. 78 •  FLOATING SOLAR MARKET REPORT Sonoma County Gazette. 2015. “S Sonoma Clean Power Contracts to Build Largest US Floating Solar Project.” Sonoma County Gazette, February 27. https://www.sonomacountygazette.com/cms/pages/sonoma-county-news-article-3597.html. Sun Daily, The. 2018. “Sarawak CM Suggests LONGi Group to Develop Floating Solar Park.” The Sun Daily, April 16. http:// www.thesundaily.my/news/2018/04/16/sarawak-cm-suggests-longi-group-develop-floating-solar-park. Sungrow. N.d. “1.2+1.9MW Floating PV Power Plant Project in Tainan City, Taiwan.” https://en.sungrowpower.com/refer- ence?id=22&ref_cate_id=30. Tan, A. 2018. “HDB Exploring Floating Solar Panels in Open Sea.” The Straits Times, July 9. https://www.straitstimes.com/ singapore/environment/hdb-exploring-floating-solar-panels-in-open-sea. Tsanova, T. 2018. “Dutch PWN Launches 7-MWp Floating Solar Tender.” Renewables Now, September 17. https:// renewablesnow.com/news/dutch-pwn-launches-7-mwp-floating-solar-tender-626921/. Vientiane Times. 2018. “Japanese Firm to Build 14MW Floating Solar Power Farm in Hadxaifong.” Vientiane Times, Asia News Network, June 21. http://asianews.eu/content/japanese-firm-build-14mw-floating-solar-power-farm-hadxai- fong-75282. Yonhap News Agency. 2018. “S. Korea to Create Floating Solar Farm on Sihwa Lake.” Yonhap News Agency, September 4. http://english.yonhapnews.co.kr/news/2018/09/04/0200000000AEN20180904001600320.html. Zaripova, A. 2016. “Brazil Launches First Stage of 10 MW Floating PV System.” PV Magazine, March 7. https://www.pv- magazine.com/2016/03/07/brazil-launches-first-stage-of-10-mw-floating-pv-system_100023576/. CHAPTER 3:  GLOBAL MARKET AND POTENTIAL  • 79 NORWAY © Ocean Sun 4 POLICY CONSIDERATIONS AND PROJECT STRUCTURING As highlighted in this report, deployment of floating ity ceiling.14 Grid-connected power plants are granted solar photovoltaic (FPV) power is cost-competitive a FiT equivalent to $0.0935/kWh, before value added under many circumstances and therefore should not tax. On 29 January 2019, MOIT released a first draft require financial support. Nevertheless, initial projects update of the country’s current feed-in tariff structure. may require some form of support to overcome bar- The draft FiTs will vary based on (i) the completion date, riers associated with the industry’s relatively limited (ii) location, and (iii) type of solar projects (i.e., floating, experience with FPV technology. ground-mounted, integrated storage system or rooftop solar). In the latest draft, released on 12 April 2019, the Countries have taken various approaches to FPV pow- new FiT for floating solar projects will be 8.5% high- er. Many of the policies supporting FPV installations fall er than for ground-mounted PV, but is still subject to into one of two categories: (i) financial incentives or (ii) potential changes.15 supportive governmental policies. These are discussed in the sections that follow. In Malaysia, a FiT in force since 2011 has pre-set capacity ceilings for each technology. The so-called Financial incentives and 4.1  “RE quota” administered by the Sustainable Energy Development Authority16 is set every six months and support mechanisms in covers a period of three years. However, there is cur- selected countries rently no quota available for FPV projects. Large-scale Few countries have provided financial incentives spe- solar PV (including FPV) projects are implemented via cifically for FPV systems. However, most countries that auctions, as discussed in section 4.2.3. are still implementing preferential feed-in tariffs (FiTs) for solar PV typically also include FPV. This is the case In Japan, large-scale PV solar systems were eligible for in Japan, Malaysia, and Vietnam, among others. a FiT until 2017, when systems of 2 MWp and above (including FPV) became ineligible. Offtake prices are In a bid to encourage the large-scale implementation now determined through a competitive auction system. of renewable energy technologies, Vietnam’s Ministry of Industry and Trade (MOIT) established a FiT scheme Some economies have specific FPV support mecha- for all on-grid utility-scale solar installations in 2017; it nisms. Examples are Taiwan, China; the state of Mas- also applies to FPV projects.13 Set to expire on June sachusetts in the United States; and the Republic of 30, 2019, the scheme was extended for another 12 Korea (table 4.1). months in Ninh Thuan province, but with a 2 GW capac- 13. Prime Minister’s Decision No. 11/2017/QD-TTg, passed on April 11, 2017, and taking effect in June 2017. 14. Resolution No. 115, dated August 8, 2018; https://www.lexology.com/library/detail.aspx?g=530843be-3857-4c97-a07d-e59db9a8a3a7 15. https://www.bakermckenzie.com/en/insight/publications/2019/04/updated-draft-policy-on-feed 16. http://www.seda.gov.my/ CHAPTER 4:  POLICY CONSIDERATIONS AND PROJECT STRUCTURING  • 81 TABLE 4.1. Examples of financial support mechanisms for FPV systems, 2018 Economy Support mechanism Taiwan, China In Taiwan, China, a specific feed-in tariff applies to floating solar photovoltaic (FPV) power; it is higher than the FiT for ground-mounted photovoltaic systems. In the second half of 2018, the FiT for ground-mounted systems was NT$4.2943/kWh as opposed to NT$4.6901/kWh for FPV systems. In early 2019, the Ministry of Economic Affairs announced that the FiTs for FPV would be reduced in the first half of 2019 to NT$4.5016/kWh (for projects not connected to the high-voltage transmission grid) or NT$4.9345/kWh (for connected projects). These tariffs are about 10 percent higher than the FiTs for similar-size (≥1 MW) ground-mounted PV projects. FPV FiTs will drop by another 1.5 percent in the second half of 2019a (Ministry of Economic Affairs 2018–2019). Massachusetts, The Solar Massachusetts Renewable Target (SMART) Program was implemented in 2018. It United States offers a location-based compensation rate add-on of $0.03/kWh (Tranche I—80 MW) for FPV under certain conditions (Mass.gov n.d.). Korea, Rep. As part of the Renewable Portfolio Standards, power producers with installed generation capac- ity greater than 500 MW must produce a minimum proportion of their power using new and renewable energy sources. The 2018 obligatory renewable service supply ratio is 6 percent of total power generation (excluding new and renewable energy generation). A weighting scheme is applied to various renewable technologies for purposes of computing the ratio. A weighting of 1.5 applies to FPV installations, as opposed to 0.7 for land-based systems larger than 3 MW (Korea Energy Agency n.d.). Source: Authors’ compilation based on sources mentioned in table. Note: FPV = floating photovoltaic; kWh = kilowatt-hour; MW = megawatt; NT$ = New Taiwan dollar. a. https://www.moea.gov.tw/MNS/populace/news/News.aspx?kind=1&menu_id=40&news_id=82734. 4.2 Supportive governmental The first pilot project in India, funded by the Ministry policies of New and Renewable Energy, became operational in 2014. Following its success, other institutions and Governmental policies favoring clean energy—such as government bodies began considering installing small ambitious renewable energy targets, construction of demonstration projects across the country. Numerous pilot FPV plants, and solar PV tenders and auctions— FPV tenders were launched in 2018, and more than 1.8 have helped FPV projects come to fruition in certain GWp of capacity is either planned, tendered, or under countries. construction. 4.2.1 Renewable energy targets After more than two years of analyzing the operating results of the world’s largest FPV testbed in Singapore, Most of the world’s countries have renewable energy the country’s Economic Development Board (EDB), in targets, some of which are very ambitious (table 4.2). late 2018, commenced the first phase of a commercial In locations where population density is high, and land tender to build 100 MWp of FPV atop a reservoir, with is scarce or has a high opportunity cost, FPV may have more projects in the pipeline on other reservoirs. a key role to play in reaching these ambitious targets. In the Netherlands, the national consortium Zon op 4.2.2 Pilot plants Water was created in 2017 at the initiative of the Minis- Dedicated agencies of some governments have sup- try of Infrastructure and Water Management. Its ambi- ported the set-up of demonstration or pilot plants. tious aim is to promote the development and installation India, the Netherlands, and Singapore are examples. of 2 GWp of FPV in the Netherlands by 2023. Multiple demonstration plants are being implemented to test FPV in various environments. 82 •  FLOATING SOLAR MARKET REPORT TABLE 4.2. Selected ambitious solar PV targets Country Target By Year Source China 210-270 GWp solar PV 2020 https://www.pv-magazine.com/2018/11/05/china-may-raise- 2020-solar-target-to-over-200-gw/ India 100 GWp solar PV 2022 https://www.greentechmedia.com/articles/read/woodmac- expects-india-to-miss-2022-renewables-target#gs.NOv- VS8ym Japan 64 GWp solar PV 2030 https://www.pv-magazine.com/2017/12/12/japan-may- surpass-2030-pv-target-of-64-gw-within-two-years-rts/ Korea, Rep. 30 GWp solar PV 2030 https://www.pv-magazine.com/2017/12/20/solar-installations- to-soar-under-new-south-korean-energy-plan/ Taiwan, China 20 GWp solar PV 2025 https://www.pv-magazine.com/2018/12/12/obscured- policies-in-taiwans-fit-scheme-to-impact-on-sustainable- development-of-local-solar-supply-chain/ Source: Authors’ compilation based on sources mentioned in table. Other pilot plants (led by either the private or public as solar panel assemblers and maintenance person- sector) are being considered in Afghanistan, Albania, nel. They are earning better wages and are no longer Azerbaijan, Kyrgyz Republic, Liberia, the Philippines, exposed to harmful mine conditions known to cause Thailand, and the United States, to name just a few lung disease. countries. As shown in figure 4.1, the 150 MWp FPV project built 4.2.3 Tenders and auctions by Sungrow in Anhui Province is located at a subsid- ence area in Guqiao town, which covers an area of In 2016, as part of its so-called Top Runner program, 422 hectares. It is estimated that the project will reduce China’s National Energy Agency issued a tender for the annual average standard coal consumption by 62,900 installation of 1 GWp of FPV in coal-mine subsidence tons and reduce annual carbon dioxide emission by areas, mainly in Anhui Province. An additional 400 MWp 150,000 tons (Sungrow 2019). was tendered in Shandong Province. As reported in chapter 3, the winning bidders (among them Sungrow, In India, many solar PV tenders are organized by the Trina, GCL, Xinyi, CECEP, and China Three Gorges New Solar Energy Corporation of India or other utilities and Energy) sell the generated electricity to the State Grid distribution companies facing stringent renewable Corporation of China at rates ranging from yuan (Y) 0.71 purchase obligations mandated by central and state to Y 0.81 per kilowatt-hour (kWh) ($0.11–0.12/kWh). governments. These tenders are for specific FPV proj- ects to be built on pre-determined reservoirs in the Producing solar power in mining regions while scal- states of Andhra Pradesh, Himachal Pradesh, Kerala, ing back coal-based power production is one way to Maharashtra, Rajasthan, Telangana, Uttar Pradesh, address local air pollution in several regions of China and West Bengal, among others. More specifically, fol- (Mason and BBC, 2018). There are dozens of flood- lowing new regulations from the central government, ed coal mines in China. Spurred by the so-called Top the thermal power generation company, NTPC Ltd, has Runner program, solar developers are turning these launched several tenders for FPV projects as compo- environmental and social challenges into an opportu- nents of a series of renewable power plants built at nity. Anhui Province is home to the world’s largest FPV existing power stations to meet renewable generation installations to date, ranging from 20 MWp to 150 MWp obligations. Building FPV projects at existing conven- per site. Local people who just a few years ago worked tional power plants brings the additional advantage of underground as coal miners are now being retrained allowing distribution companies to meet their renew- CHAPTER 4:  POLICY CONSIDERATIONS AND PROJECT STRUCTURING  • 83 FIGURE 4.1. Coal mine subsidence area in Anhui Province, China, rehabilitated with Sungrow Guqiao 150 MWp FPV system. Left: after construction of FPV system; right: local people employed by Sungrow Source: © Sungrow. Source: © Sungrow. able purchase obligations through existing power pur- is to be unlocked. Positive examples include tenders chase agreements. 17 for water-lease contracts in Korea with K-Water, in Sin- gapore with PUB, and in the Netherlands with NV PWN Under France’s large-scale solar PV auction scheme, Waterleidingbedrijf Noord-Holland. the Ministry of Ecological and Solidarity Transition in 2017 tendered 70 MWp of innovative solar capacity. Other policy and regulatory 4.3  Among the winning bidders were several small-scale considerations FPV projects. Even countries in which FPV power has undergone In Malaysia, two Large-Scale Solar tenders (LSS1 significant development lack clear, specific regulations and LSS2) have been completed with a total award- on permitting and licensing of such plants. To a great ed installed capacity of 958 MWp. Four projects from extent, regulatory processes can be based on those LSS2, totaling about 80 MWp have been attributed to employed for ground-mounted PV, but legal interpre- FPV projects. The LSS is an initiative from the gov- tation is still needed. In some countries, reservoirs for ernment to achieve Malaysia’s national renewable drinking water and hydropower plants are considered energy roadmap. Past tenders included an additional national-security sites, making permitting more com- merit point in the comparative price of bid calculation plex and potentially protracted. to encourage use of former mine lands, and which For most countries hoping to develop a well-func- have benefited FPV projects foreseen on flooded col- tioning FPV segment as part of their solar PV market lapsed mines, like in China. A new 500 MWp Large development, key policy and regulatory considerations Scale Solar 3 (LSS3) scheme was tendered in early remain to be addressed. These include: 2019. An open tender, it is expected to include FPV projects, as it includes the same merit point mecha- • Unique aspects of permitting and licensing that nism as for previous tenders. necessitate interagency cooperation between ener- gy and water authorities. This also includes environ- Receptivity of the entities responsible for managing mental impact assessments for FPV installations. water bodies will be essential if FPV’s broader potential 17.h ttps://www.financialexpress.com/industry/ntpc-invites-1000- mw-renewable-energy-tenders/1396219/ 84 •  FLOATING SOLAR MARKET REPORT • Water rights and permits to install and operate and whether there is an energy-intensive user close an FPV plant on the surface of a water body and to the FPV system (e.g., a water treatment facility).18 anchor it in or next to the reservoir. In Japan and China, most FPV-generated electricity is sold to the grid, whereas in the United Kingdom most • Tariff setting for FPV installations, which could be FPV plants produce for self-consumption, and only done as for land-based PV, for example, through surplus is injected into the grid. Some examples are FiTs for small installations and tenders or auctions given below: for large projects. • Japan. Because the FiT for solar energy is high, Other questions pertain to access to existing transmis- FPV plants usually sell their generated solar elec- sion infrastructure. How will this be managed? Who will tricity to the grid. However, because systems larg- be responsible? What permits and agreements will be er than 2 MWp no longer benefit from a FiT (as required? of 2017), a shift toward self-consumption could become more common. Hydro-connected plants call for special consideration. Will the owner/operator of the hydropower plant be • United Kingdom (Queen Elizabeth II and God- allowed to add an FPV installation? Will it be permit- ley). The two largest FPV plants in the United King- ted to offer a concession to a third party to build, own, dom both sell electricity (behind the meter) to a and operate such an installation? If so, rules must be local water treatment facility. The surplus is then devised to coordinate dispatch of the solar and hydro- injected into the grid. Both plants were realized power plants’ output. Finally, risks and liabilities relat- under the United Kingdom’s Renewable Obliga- ed to the hydropower plant may also affect connected tion scheme. solar facilities. • China (Anhui Province). Most of China’s mega- watt-scale FPV plants are being built under the Business models and project 4.4  so-called Top Runner program. All generated elec- structuring tricity is sold to local electricity companies at a tariff As of end 2018, the largest completed FPV projects determined through competitive bidding.19 are located in China (up to 150 MWp). Other large proj- To understand how FPV projects are typically financed, ects are located in Korea (up to 18.7 MWp) and Japan it is useful to divide them into two groups: those whose (up to 13.7 MWp). Most other FPV systems are much installed capacity is less than or equal to 5 MWp, and smaller in scale (i.e., typically around 5 MWp or less), those whose installed capacity is greater than 5 MWp. though this is expected to change soon, as many util- Table 4.3 summarizes typical financial structures for ity-scale projects are under development around the these two classes, which are similar to those for land- world. based PV deployment. To build trust in the technology, public grants are often provided to finance research The business model and type of financing of an FPV and development and pilot projects (<1 MWp) that project will depend on its size and offtake arrange- could be run by companies (as in the case of the Top ment. Many models are appropriate; most are simi- Runner program projects in China) or by universities lar to those used for ground-mounted and rooftop PV and public research institutions (as in the case of Sin- installations. gapore’s testbed). Either the electricity produced can be self-consumed or it can be sold to a local or national power utility. 18. “Task 1: Commercial Readiness of FSPV—Global Market and Which of the options is chosen depends on national Performance Analysis,” in Planair and PITCO (2017). regulations (e.g., net metering), existing FiTs for PV, 19.  Ibid. CHAPTER 4:  POLICY CONSIDERATIONS AND PROJECT STRUCTURING  • 85 TABLE 4.3. Typical financing structures of FPV systems System Business model Ownership Financing structure size (MWp) ≤5 Self-consumption Commercial Pure equity or mix of equity and corporate financing (or “bal- (with excess sold and industrial ance sheet” financing). Owner would typically be an energy- to the grid where companies intensive commercial or industrial company with ponds, lakes, possible) or reservoirs on its premises and willing to install a floating solar system for its own use. In this case, the project owner (a developer; engineering, procurement, and construction contractor; or a corporate consumer) funds the project by borrowing against the company’s balance sheet. Vendor financing is also possible in cases where one of the equip- ment manufacturers (e.g., the module or float structure suppli- er) is an established company with a strong balance sheet. >5 Power sold to the Independent Mix of debt and equity (typically 80:20); on balance sheet or grid through a power pro- nonrecourse project finance. Projects larger than 10 MWp power purchase ducers and will likely use project finance structures similar to those of agreement public utilities utility-scale ground-mounted photovoltaic projects. Source: Authors’ compilation. Note: MWp = megawatt-peak. Except in China, most FPV projects are small and as a part of the water surface owned by another party financed in local currencies by local or regional banks. via the Dutch legal concept of accession (natrekking), In Japan and Taiwan, China, local commercial banks the owner of the FPV system must receive a right of have taken advantage of the favorable long-term FiTs superficies (opstalrecht). According to the Dutch Civ- available for FPV. Large international commercial banks il Code (Article 5:101 [1] DCC), a right of superficies as well as multilateral development finance institutions enables its proprietor—the “superficiary”—to have or are expected to get involved as larger projects start to acquire for himself buildings, constructions, or plants be developed in low-income countries. (vegetation) in, on, or above an immovable thing owned by someone else. This means that under Dutch civil Given their many advantages, projects that combine law, the owner of an FPV system (e.g., the asset owner) FPV with hydropower are likely to proliferate. New could be different from the owner of the water surface financing structures could enable FPV systems to be (e.g., the public water utility). By obtaining a right of built on the reservoirs of hydropower plants by offer- superficies, the developer of the FPV system can avoid ing the lenders financing the FPV system recourse to a the risk of accession. The bank will usually require a share of the cash flows of the hydropower plant. mortgage on this right of superficies. It is therefore important to understand the property rights and rights Table 4.4 outlines a few examples of financing struc- of ownership of movable and immovable assets appli- tures and business models used in FPV systems. In cable in the jurisdiction where an FPV system is being the Netherlands, a bank’s ability to identify appropriate built, as these may affect the lenders’ options to request security is a major challenge to implementing FPV proj- or enforce security interests in the project. ects. To ensure that the FPV system is not considered 86 •  FLOATING SOLAR MARKET REPORT TABLE 4.4. Selected business models and project finance structures used for FPV structures Country Project Status Observations United 6.3 MWp Queen Operational The London’s Queen Elizabeth II FPV project, which cost Kingdom Elizabeth II floating about £6.5 million, was funded, installed, and operated by photovoltaic (FPV) Lightsource BP . The floating array covers less than 10 per- system cent of the reservoir’s surface. The project generates about 5,750 megawatt-hours (MWh) of power per year and sells it to Thames Water, the United Kingdom’s largest water and waste- water company, via a private-wire power purchase agreement with Lightsource BP . The FPV system satisfies around 20 per- cent of Thames Water’s energy needs, as part of the utility’s ambitious bid to self-generate a third of its own energy by 2020 (Lightsourcebp 2016). Netherlands 1.8 megawatt-peak Operational Tenten Solar Zonnepanelen B.V. has developed the project for (MWp) Lingewaard Drijvend Zonnepark Lingewaard B.V. under the SDE+ scheme FPV system of governmental subsidies in the Netherlands (Netherlands (Gelderland) Enterprise Agency n.d.). The project was financed through a nonrecourse project finance loan from ING. France 17 MWp O’MEGA 1 Under The project is located in Piolenc, Vaucluse, and developed FPV system construction by Akuo Energy. It is the first in France to use nonrecourse financing, with a loan of €12.8 million from Natixis Energeco. The project has a mixed ownership structure with capital from the local municipality, Akuo Solar, and residents (via crowdfunding) (Kenning 2018). Akuo Solar holds 60 percent of the shares while the municipality and residents hold 40 percent. The debt-equity ratio is 73:27, and the loan struc- ture is similar to that of ground-mounted PV projects. Vietnam 47.5 MWp DHD Under DHD is expected to enter a 20-year PPA with EVN under FPV System at construction the current feed-in tariff regime of $0.0935/kilowatt-hour Da Mi equivalent, paid in Vietnamese dong but indexed to the U.S. dollar. The Asian Development Bank has proposed to provide a direct loan of up to $20 million as well as concessionary loans up to $22 million (from various sources). All loans will have a tenure of up to 15 years, including a 1-year grace period on the principal repayment. The bank can also rely on DHD’s 722 MW of existing hydropower as a financial back- stop (ADB 2018). Source: Authors’ compilation. CHAPTER 4:  POLICY CONSIDERATIONS AND PROJECT STRUCTURING  • 87 References ADB (Asian Development Bank). 2018. “Project Number: 51327-001: Report and Recommendation of the President to the Board of Directors.” Manila. September. https://www.adb.org/sites/default/files/project-documents/51327/51327-001- rrp-en.pdf Baker McKenzie, September 12 2018: “Vietnam’s solar feed-in-tariff for Ninh Thuan Province officially extended for a total capacity of 2,000 MW.” https://www.lexology.com/library/detail.aspx?g=530843be-3857-4c97-a07d-e59db9a8a3a7 BBC (British Broadcasting Corporation). 2018. “Solar Farm Means ‘I Can Breathe More Easily.’” Video story, BBC News, April 24. https://www.bbc.co.uk/news/av/business-43881280/solar-farm-means-i-can-breathe-more-easily. Decision No. 11/2017/QD-TTg of the Prime Minister dated 11 April 2017, Vietnam. http://www.lse.ac.uk/GranthamInstitute/ wp-content/uploads/laws/8277.pdf Duane Morris LLP, February 1 2019 : “Vietnam’s draft new solar tariffs—more sun, less cents, more sense.” https://www. lexology.com/library/detail.aspx?g=93915ecf-4410-4246-8260-ef7d81d0a0f6&utm_source=lexology+daily+news- feed&utm_medium=html+email+-+body+-+general+section&utm_campaign=lexology+subscriber+daily+feed&utm_ content=lexology+daily+newsfeed+2019-02-04&utm_term= Dutch Civil Code, Article 5:101 [1]. Korea Energy Agency. N.d. “Renewable Portfolio Standards.” http://www.energy.or.kr/renew_eng/new/standards.aspx. Mason, Pauline. 2018. “Meet the Ex-Miners Who Are Now Walking on Water.” BBC News, April 27. https://www.bbc.co.uk/ news/business-43864665. Mass.gov. N.d. “Solar Massachusetts Renewable Target (SMART) Program: General Information.” https://www.mass.gov/ info-details/solar-massachusetts-renewable-target-smart-program# general-information-. Ministry of Economic Affairs. 2018. “2018 Renewable Energy FIT Rates Are Announced.” Bureau of Energy, Ministry of Economic Affairs, January 25. https://www.moea.gov.tw/MNS/english/news/News.aspx?kind=6&menu_id=176&news_ id=76457. Ministry of Economic Affairs. 2019. “附表—、108 年度太陽光電公告費率.” Bureau of Energy, Ministry of Economic Affairs, January 30. https://www.moea.gov.tw/MNS/populace/news/News.aspx?kind=1&menu_id=40&news_id=82734. Planair and PITCO. 2017. Assessment of Floating Solar PV Potential for Pakistan; Task 1: Commercial Readiness of FSPV—Global Market and Performance Analysis. Unpublished report prepared for International Finance Corporation, Washington, DC. Government of Vietnam. 2018. “Resolution No. 115/NQ-CP.” August 31. Hanoi. http://vepg.vn/wp-content/ uploads/2018/10/115_NQ-CP_310818_CP_Dev-of-Ninh-Thuan_EN_GIZ- unofficial-transl.pdf Press articles and other sources Press articles https://www.pv-magazine.com/2018/12/12/obscured-policies-in-taiwans-fit-scheme-to-impact-on-sustainable-development- of-local-solar-supply-chain/ https://www.pv-magazine.com/2018/11/05/china-may-raise-2020-solar-target-to-over-200-gw/ https://www.greentechmedia.com/articles/read/woodmac-expects-india-to-miss-2022-renewables-target#gs. NOvVS8ym https://www.pv-magazine.com/2017/12/12/japan-may-surpass-2030-pv-target-of-64-gw-within-two-years-rts/ https://www.pv-magazine.com/2017/12/20/solar-installations-to-soar-under-new-south-korean-energy-plan/ https://www.pv-magazine.com/2018/12/12/obscured-policies-in-taiwans-fit-scheme-to-impact-on-sustainable- development-of-local-solar-supply-chain/ 88 •  FLOATING SOLAR MARKET REPORT https://www.financialexpress.com/industry/ntpc-invites-1000-mw-renewable-energy-tenders/1396219/ https://www.pv-tech.org/news/akuo-energy-starts-construction-on-17mw-floating-solar-plant-in-france Other sources https://www.lightsourcebp.com/uk/stories/qe2/ https://english.rvo.nl/subsidies-programmes/sde http://www.seda.gov.my/ CHAPTER 4:  POLICY CONSIDERATIONS AND PROJECT STRUCTURING  • 89 MALDIVES © Swimsol 5 COSTS OF FLOATING SOLAR In this chapter, the theoretical costs of floating and a float structure made of high-density polyethylene, ground-mounted photovoltaic (PV) systems are com- HDPE). Total project costs are not always accurately pared, using average figures based on industry feed- disclosed in the public domain and should be taken as back and publicly available data. Since floating PV indicative, as it is difficult to independently verify and (FPV) systems are not as common or widespread as compare them. Some could contain grid connection ground-mounted systems, it remains difficult to have costs, water surface rental/lease costs, and import tax- data about their capital and operating costs that could es and duties on PV modules and other components. be generalized. The analysis presented here uses rea- In some cases, costs may be affected by stringent sonable assumptions based on information available in local content rules, making them less comparable. the public domain and best practices from the indus- Some projects benefit from grants for feasibility and try. A more detailed analysis would need to be per- engineering studies, thereby lowering development formed at a country level, and of course on a project costs. Finally, from an engineering point of view, some basis, for a complete picture of how FPV compares to projects are easier to implement than others (e.g., ground-mounted in given circumstances. where the water depth is low and water levels vary lit- tle), considerably reducing project design, anchoring, It should be clarified here, at the outset, that even and mooring costs. though the two are being compared, FPV is not being put forward as a competitor to ground-mounted proj- To date, most of the projects operational outside Chi- ects. Floating installations are complementary to na are small ones of around 5 megawatt-peak (MWp) ground-mounted systems and serve different needs or less, with the exception of a few large installations and purposes. For example, when integrated with in Japan and the Republic of Korea. But the FPV hydropower plants, they can help reduce the seasonal market is burgeoning and many large-scale projects variability of hydropower generation. Or they can be (ranging between 20 and 150 MWp) are currently used to harness the sun’s energy even where land is under development or construction in various coun- expensive or scarce, or to help commercial and indus- tries in the world. It will be interesting to watch the trial companies garner profits from large, unused water evolution of capital expenditure (CAPEX) in the mar- bodies on their premises—and benefit from supply of ket to see how economies of scale affect investment additional electricity. Many business models exist; the costs. Obviously, the costs of small systems can vary best choice depends on the context, including the pol- significantly by location. icy and regulatory framework. On a per watt-peak basis, the CAPEX of FPV projects is still slightly higher than of ground-mounted PV, main- 5.1  Recent disclosed FPV costs ly due to the expenses of the floating structure (the number of floats required depends on the design), To start, we will consider the publicly announced turn- the inverter floating platform (where relevant), and the key engineering, procurement, and construction (EPC) anchoring and mooring system. It is fair to expect that costs of projects using similar types of technology (i.e., the costs of floats will drop over time. The FPV mar- CHAPTER 5:  COSTS OF FLOATING SOLAR  • 91 ket is still at a nascent stage, and cumulative installed In March 2018, the India-based West Bengal Pow- capacity is only about 1.3 gigawatt-peak (GWp) (the er Development Corporation Limited announced the total global installed PV capacity was about 500 GWp results of an EPC tender for a 5 MWp FPV system to at the end of 2018). However, the extent to which costs be developed in the district of Murshidabad, on a raw will drop is difficult to predict, especially since HDPE water pond of the Sagardighi Thermal Power Project. floats remain dependent on crude oil prices. If nothing International Coil Limited, with the support of Ciel & else, economies of scale should help to reduce costs Terre International, won the turnkey EPC bid with the as the scale of float production increases and experi- lowest quote of Rs. 269.12 million (no grants provid- ence from past production and deployment is applied. ed) (Prateek 2018), which corresponds to about $4.13 Nevertheless, the design of the floating structure and million or $0.83/Wp (using the 2017 average annual its anchoring and mooring system will always remain exchange rate). The average price of the five bidders site sensitive, and costs will vary depending on the (i.e., International Coil Ltd, Adani Infra, Vikram Solar, complexity of the engineering challenges involved. Sterling and Wilson, and Giriraj Renewables) was sub- However, unlike ground-mounted PV, generally no stantially higher at $1.14/Wp. heavy civil engineering work is required to set up an FPV installation. On this basis, a 2017 International Future capital costs will depend on the costs of solar Finance Corporation (IFC) report 20 estimates that for modules as well as the development of new floating a “cost per watt-peak” installed, FPV should not devi- technologies, beyond HDPE plastic floats that are the ate significantly from ground-mounted PV installations. most common floating structures on the market today. This has been confirmed by recent FPV tenders. 5.1.1  Capital expenditure Information on FPV investment costs, mainly retrieved The main difference between the cost of investing in from public press releases, are summarized in fig- ground-mounted or FPV resides in the floating struc- ure 5.1. Projects have been sorted by their month ture and the related anchoring and mooring system, of commissioning. Total CAPEX for FPV systems in which are highly site specific. There are too few data 2018 ranged between $0.8 and $1.2 per watt-peak, points available in the nascent market to provide an depending on the location, water body depth and vari- “average” cost figure with a high level of confidence. ation, and system size. Large projects on deep-water Another issue that affects costs is the use of direct cur- reservoirs are likely to be the most complex, pushing rent (DC) (in some cases submarine) electric cables up project development and capital costs. That said, with additional insulation and shielding properties to based on the interviews with industry experts, the protect them from moisture degradation, thereby add- scale matters for projects up to about 30 MWp, after ing a premium to the CAPEX of FPV when compared which economies of scale become less significant. to ground-mounted PV. As can be seen in figure 5.1, system prices remain rel- The following subsections outline reasonable assump- atively high in Japan. China and India have achieved tions regarding the average cost per component of much lower FPV costs than the global average, a trend a hypothetical 50 MWp FPV system on a freshwater, also observed for ground-mounted and rooftop solar inland reservoir (with a maximum depth of 10 meters systems. A 500 kilowatt-peak (kWp) FPV system in and minimal water level variation). Kerala, India, is an exception: here only high-quality components were used without any attempt to realize cost-benefit efficiencies. “Task 1: Commercial Readiness of FSPV—Global Market and 20.  Performance Analysis,” in Planair and PITCO (2017). 92 •  FLOATING SOLAR MARKET REPORT FIGURE 5.1. FPV investment costs, 2014–18 (realized and auction results) UK—0.2 MWp Sheeplands (2014) 1.14 Japan—2 MWp Shiroishi Saga (2015) 3.12 Portugal—0.2 MWp EDP Hydro (2016) 2.31 UK—6.3 MWp Queen Elizabeth II (2016) 1.22 China—20 MWp Anhui Xinyi (2016) 1.48 Japan—2.4 MWp Noma Ike (2017) 2.93 China—40 MWp Anhui Sungrow (2017) 1.13 India—0.5 MWp Kerala (2017) 2.84 Japan—1.5 MWp Mita Kannabe (2017) 2.93 Japan—13.7 MWp Yamakura Dam (2018) 0.97 India—2 MWp Andhra Pradesh (2018) 0.92 China—150 MWp Three Gorges (2018) 0.99 India—5 MWp West Bengal Auction Lowest Price (2018) 0.83 India—5 MWp West Bengal Auction Avg Price (2018) 1.14 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.8 1.2 $/Wp Source: Authors’ compilation based on media releases. Notes: Using 2017 U.S. dollar annual exchange rates, as released by the Organisation for Economic Co-operation and Development. FPV = floating photovoltaic; MW = megawatt; $/Wp = U.S. dollars per watt-peak. The cost component assumptions used in this chapter This design is relatively resistant to moisture. Frameless are based on the experience and investigations of the modules have been used in some projects, especial- Solar Energy Research Institute of Singapore (SERIS), ly for floats using membranes, as they allow for direct and guidance from equipment suppliers, EPC con- contact with the surface and eventually also reduce tractors, and developers. It is important to reiterate the risk of potential induced degradation (which rises that these figures are estimates that will need to be when humidity rises). It is also too early to confirm that adjusted once more data become available after the glass-glass modules perform better than glass-back- completion of more large-scale FPV systems across sheet modules. PID-free or glass-glass modules could the world. Also, the cost of a specific FPV project will be advantageous where humidity is high, but generally depend on its design and location. come at a slight price premium. Solar PV module In this analysis, a standard PV module price of $0.25/ There are no particular standards for FPV modules. The Wp is used to calculate the levelized cost of electric- most commonly used are framed glass-glass mono- or ity (LCOE). This is considered representative of the polycrystalline silicon modules with 60 cells or 72 cells. average price of polycrystalline silicon (poly-Si) mod- CHAPTER 5:  COSTS OF FLOATING SOLAR  • 93 ules (with efficiencies typically in the range of 17–19 The estimated average price of central inverters for a percent)21 and of mono-Si high-efficiency/passivated solar PV system of about 50 MWp is about $0.06/Wp; emitter rear cell (PERC) modules from Chinese man- this is the price used in the LCOE calculations. This ufacturers (with a typical efficiency greater than 20 figure is in line with the inverter cost estimates for utility- percent) 22 in the third and fourth quarters of 2018. scale PV systems cited in NREL (2017). As reported by EnergyTrend and as a direct conse- quence of China’s “531” policy, the average prices of Floating structure, anchoring and mooring system mono- and polycrystalline silicon modules fell by 19.8 HDPE floats are the most common, cost-competi- percent and 25.5 percent, respectively, in the first three tive structure used for FPV plants. The quality of the quarters of 2018 (Bellini 2018). HDPE material, including additives for long-term dura- bility, is important to consider. Potential investors and No import or safeguard duties were assumed in the developers should ensure that floats are sourced from estimates of PV module prices. Currently, most large- high-quality manufacturers with a strong track record. scale FPV plants are deployed using pontoon-type Floats should also be recyclable, nontoxic, resistant to floats, with PV panels mounted at a fixed tilt angle. ultraviolet radiation, salt corrosion, water, alkalis, and acids, and have a lifetime of over 20 years. Experience A fixed array installation is simple to install in different from the maritime industry has shown that a lifetime of types of reservoirs, and the space needed between PV 20-25 years (and even longer) is possible. As a safety panels is relatively small. Furthermore, its complexity measure, particularly when being installed on a drink- and thus its cost is low, and the system does not occupy ing water reservoir, floats should be food grade and a large surface area. Since this type of installation does compliant with strict drinking water tests. not require any moving parts, it is relatively resilient and needs little maintenance (ERM—ADB/ Da Nhim–Ham Because they are the most common, costs of high- Thuan–Da Mi Hydro Power Joint Stock Company 2018). quality HDPE floats were used to calculate the LCOE in this analysis. Anchoring and mooring costs are Inverter included in the total price of floats. Their costs vary Unlike solar PV modules, inverter prices are negotiat- according to site conditions, such as local wind load ed at a regional level; hence, no exchange price data (more anchoring points are needed where winds tend are available for estimating a global benchmark price. to be strong) and maximum depth and water level However, inverter prices have come under similar pres- variation (where the level fluctuates widely, more com- sure as panel prices; it is likely that they will continue to plex mooring is required). A system in calm and shal- fall, gradually, levelling off in the medium term. low waters, for example, could simply be anchored to a bank. The design of the floating structure and Both string and central inverters have been used in the anchoring and mooring system, and to a certain FPV installations around the world. Generally, central extent cabling costs, depends on the following input inverters are used for large-scale FPV systems and parameters: string inverters for smaller systems. Inverters can either be installed on the surface of a water body (on • Bathymetry (including subsurface soil conditions) a floating pontoon) or on land (typical for smaller sys- • Water-level variation tems). If inverters are mounted on floats, they should • Wind and wave characteristics have an ingress protection rating of at least IP67 to • Type of banks (for launching) withstand the high moisture. • Water quality and level of salinity Another important cost relates to logistics and trans- port; HDPE floats have a high volume-to-weight ratio. 21.  EnergyTrend, 2018/11/07 update. 22.  PVinsights, 2018/11/04 update. For larger systems, local manufacturing processes 94 •  FLOATING SOLAR MARKET REPORT using redeployable equipment may be worth con- required, which would have a direct impact on costs. sidering. Electrical cable routing and the slack needed for the constant movements of the floating installation also Recent (2018) cost estimates of pure HDPE float- affect the balance-of-system costs. ing structures (including anchoring and mooring) range between $0.14/Wp and $0.22/Wp. An estimate The rest of the equipment required, such as the com- of $0.15/Wp is used in the LCOE calculations. This biner boxes, a switchboard, transformer, and a proper would be for a standard, large-scale FPV project of monitoring system are not different from those needed 50 MWp that does not require a complex anchoring for ground-mounted PV projects. and mooring system, and whose floats can be pro- duced locally. An estimate of $0.13/Wp for the balance-of-system components is used in the LCOE calculations. Even though straightforward HDPE float islands offer an ideal solution in many cases, structures with frames or Design, installation, civil works, testing, and various mooring and anchoring systems might be better commissioning costs suited to certain environments and climates. According Typically, HDPE float islands are easy to install and can to the Solar Energy Application Centre in the Nether- be quickly mounted on the banks of a water body or on lands and findings from its pilot test “Zon op Water,” a platform. Certain civil works and site preparation ele- HDPE floats may not offer the most durable solution ments may need to be constructed such as an inverter under certain conditions (Hutchins 2018). A researcher housing structure (floating or on land) or a dedicated at the Solar Energy Application Center (SEAC) in Eind- launching platform (dependent upon accessibility to hoven stated, “We are also testing steel systems, where the water surface). However, heavy civil and founda- you build up mounting structures from a steel pipe, and tional works are in most cases not required for FPV another which is based on a floating cement used in projects, as it is for ground-mounted PV projects. the marine industry—it is pretty solid and easy to walk on, and has a type of foam on the underside.”23 Such With regards to the speed of FPV installation, leading alternatives to HDPE floats are typically more expen- float manufacturers report that a team of 50 trained sive, and their feasibility requires further research (and installers can deploy between 500 kWp and 1 MWp also depends on prevailing steel prices). per day, provided that a supply chain is in place. Balance-of-system components: Cabling, combiner Even though the costs outlined in this subsection can box, switchboard, transformer, and others vary substantially across projects; an estimate of It was observed in some sections of the SERIS test- $0.14/Wp is used in the LCOE calculations. bed in Singapore that the insulation resistance of the systems dropped in certain instances. This in turn Grid interconnection costs caused the inverters to temporarily shut down for safe- Another factor relevant to costs is the availability of ty reasons, that is, because a current leakage was existing grid interconnection infrastructure. Grid con- suspected. This sequence of events might have been nection, upgrades, or additional substations might be prompted by the high moisture content around the required where transmission and distribution infra- insulation of the cable. At this stage, no specific stan- structure are not present or are inadequate. Where an dards have been developed for floating PV cables, but FPV system is located close to a load center or hydro- in some cases enhanced cabling insulation might be power plant, the costs of grid interconnection will be 23. https://www.pv-magazine.com/2018/11/03/staying-afloat-whatever- the-weather/ CHAPTER 5:  COSTS OF FLOATING SOLAR  • 95 much lower since the system can benefit from existing ground-mounted PV. On a per-watt-peak basis, industry electrical infrastructure. experience indicates that the CAPEX for FPV projects tends to be $10 cents higher than for ground-mounted To simplify the LCOE calculations, it is assumed that PV projects under similar conditions. there are no grid interconnection costs. With increased competition and higher economies of Summary scale, the future cost of float structures is expected to The average total investment cost of an FPV system drop further. But, it is to be hoped that quality will not in 2018 varied between $0.8/Wp and 1.2/Wp, depend- be compromised. ing on the system’s size and location. The West Ben- gal EPC auction prices (unsubsidized) are from March TABLE 5.1. A comparison of capital investments: 2018; other listed projects were completed in the first Floating vs. ground-mounted photovoltaic systems half of 2018, and would have included higher PV mod- Ground-mounted ule prices. The CAPEX of large-scale but relatively CAPEX FPV 50 MWp PV 50 MWp component ($/Wp) ($/Wp) uncomplicated FPV projects (around 50 MWp) was in the range of $0.7–$0.8/Wp in the third and fourth Modules 0.25 0.25 quarters of 2018, depending, of course, on the loca- Inverters 0.06 0.06 tion and the type of modules involved. Mounting system 0.15 0.10 (racking)* The CAPEX of a hypothetical 50 MWp FPV installation is BOS** 0.13 0.08 laid out in figure 5.2 and table 5.1, by component, and Design, 0.14 0.13 also compared with a ground-mounted system (both construction, T&C fixed tilt) at the same location. The module and inverter Total CAPEX 0.73 0.62 costs of both types of systems are assumed to be iden- Source: Authors’ compilation based on 2018 industry data. tical. The costs of the mounting structure (including, Note: *For FPV, the mounting system includes a floating structure, in the case of the FPV system, a floating structure as and anchoring and mooring system. **Including monitoring system. BOS = balance of system; CAPEX = capital expenditure; MWp = well as anchoring and mooring) and balance-of-system megawatt-peak; PV = photovoltaic; T&C = testing and commission- costs are significantly higher for FPV projects than for ing; $/Wp = U.S. dollar per watt peak. FIGURE 5.2. Investment costs of floating vs. ground-mounted photovoltaic systems, by component Floating PV Ground-mounted PV 19% 21% Modules 34% Inverters 40% Mounting System 18% 13% BOS Design, Construction, T&C 8% 21% 16% 10% Source: Authors’ compilation based on 2018 industry data. Notes: Numbers are indicative only, for a hypothetical 50 megawatt-peak system. BOS = balance of system; MWp = megawatt-peak; PV = photovoltaic; T&C = testing and commissioning. 96 •  FLOATING SOLAR MARKET REPORT 5.1.2  Operating expenditures and an underwater robot to inspect the mooring may also be viable options. The main operating costs of an FPV system are iden- tical to those of ground-mounted PV: the leasing or As with CAPEX, the O&M costs of an FPV system will rental of the space where the system will be installed vary depending on the site’s conditions. Depending on (but in the case of FPV it’s a water body, not land), the wind forces present on the site, annual inspection operation and maintenance (O&M), insurance, and of the mooring cables and sporadic inspection of the inverter replacement costs. A lease or rental fee will not anchoring system should be performed. The need for be considered in the LCOE calculations here, for FPV maintenance also strongly depends on the variation of or ground-mounted PV, since this cost varies so widely water level that the plant undergoes. Likewise, replace- across locations and projects. Nonetheless, it is likely ment of parts of the equipment is more complicated that the cost of leasing a water body is cheaper than and time intensive. Since operating on a more or less leasing land, since the water body is not in competition deep-water body, worker safety is another aspect that with agriculture or real estate development. needs to be considered and potentially adds to the maintenance cost.24 Operation and maintenance Based on industry experience, O&M costs can vary One FPV developer, Ciel & Terre International, lists typi- a lot across jurisdictions and according to the invest- cal annual O&M efforts on its website as two man-days ment strategy of the developer/owner. Even though the per MWp to check the floating structure, plus three use of boats, or in some circumstances even divers, man-days per MWp every three years to check the might increase O&M costs at certain times, industry mooring and anchoring. representatives indicate that these costs are generally comparable to those of ground-mounted PV over the But O&M costs are difficult to estimate. For example, lifetime of a project. When placed on water bodies, leading renewable energy institutions and developers solar panels will generally incur less soiling from dust, of utility-scale ground-mounted PV projects use very and the water needed to clean them is directly avail- different assumptions (see table 5.2). able. However, corrosive bird droppings have been reported in Singapore and the United Kingdom. Their Thus, O&M costs will vary significantly depending on effects should not be underestimated since they could the project’s environment, the investor’s strategy, and negatively affect the energy yield if panels are not also the related labor costs. In the LCOE calculations cleaned regularly (thus driving up maintenance costs), used for this analysis, a general assumption of $0.011/ particularly in areas where avian life is abundant. Using Wp is used for the first year. In the real world, this figure remotely operated robotic systems to clean the panels, would vary significantly depending on where the proj- TABLE 5.2. Estimated operating and maintenance costs of ground-mounted photovoltaic systems (fixed tilt), various sources Utility-scale fixed tilt O&M ($/Wp/year) Geographic focus NREL (September 2017) 0.0154 United States Lazard v12.0 (November 2018)* 0.009 United States Fraunhofer ISE (March 2018) 2.5% of CAPEX Germany Source: Lazard 2018; NREL 2017; Fraunhofer ISE 2018. Note: *Same figure as reported in Lazard LCOE Analysis v11.0 dating from November 2017. CAPEX = capital expenditure; NREL = National Renewable Energy Laboratory; $/Wp = U.S. dollar per watt-peak. 24. “Task 1: Commercial Readiness of FSPV—Global Market and Performance Analysis,” in Planair and PITCO (2017: 30). CHAPTER 5:  COSTS OF FLOATING SOLAR  • 97 ect is located (Europe, the United States, India, or Chi- enues for a period of up to 10 or 12 years. This type na, for example) and the general climate conditions. It of coverage is optional and will depend upon the risk is therefore important to include relevant sensitivities in appetite of the sponsors/owners of the system and/ this particular cost item. Industry experience with O&M or the lenders. According to 2018 data, the estimat- costs over the lifetime of an FPV project is nascent, ed cost of insurance for both irradiance and entire and the assumptions here may be found to be overly PR risks for a 50 MWp FPV system are about 0.8–1.2 conservative. percent of insured revenues, on average (this would typically cover 85–90 percent of the P50 output). A Insurance one-off insurance cost premium of about $1.1 million Similarly to ground-mounted PV, different types of would be paid up front and cover irradiance and PR insurance coverage exist, including policies covering risks for 10 years (to match with the debt tenure). Using physical and/or nonphysical damage risks. Premiums a 50 MWp FPV system cost of $0.73/Wp, equivalent vary widely depending upon the location of the proj- to a total system cost of $36.5 million, this insurance ect, system design and quality, and climatic condi- premium would be equivalent to 3 percent of the total tions. According to a report from SolarBankability on system cost or to 0.3 percent of the system cost on general PV investments (2017: 26): an annual basis (for 10 years). Yet these numbers are solely indicative, and additional research and compar- Insurance coverage for technical risks is available ison should be performed on a project basis. both during the project’s construction and opera- tional phase. The former phase can be covered by a general liability and a construction insurance. The According to Speer, Mendelsohn, and Cory (2010), the latter phase can be covered by a general liability, annual cost of insurance can range from 0.25 percent a property damage, a business interruption, and to 0.5 percent of total CAPEX (and it is highest in areas optionally by a performance guarantee insurance. where extreme weather events are likely). Premiums The coverage is offered for technical risks caused by will vary over time. external root causes such as storm, external surg- es, fire, theft, etc. Usually, the insurance includes a deductible which the PV system owner has to cover The insurance cost used for the LCOE calculations himself. The business interruption insurance covers is 0.3 percent of the system price, paid annually and revenues lost on power feed-in for the duration of adjusted to the inflation rate. This assumption is simi- a breakdown of up to 12 months. In recent years, lar to the one used for large-scale ground-mounted PV insurers started to differentiate insurance premiums projects due to a lack of empirical data received from between new and used PV systems, with significant- the industry. More data needs to be collected from the ly higher premiums for aged PV systems. In case of an insurance claim, the insurer usually reserves the implementation and realization of FPV projects across right to cancel the insurance. the world to better understand what potential distinc- tive factors from ground-mounted PV projects could For performance guarantee insurance—such as a sys- be. An FPV insurance premium could be applied in tem output guarantee protecting against a reduced certain instances, especially when projects are built in system performance ratio (PR) or reduced solar irradi- environments that are more complex. ation—the cost will vary depending on the percentage of the revenues insured (e.g., 90 percent, 85 percent, Inverter replacement or 80 percent of P5025 output), the project’s materials Similar to ground-mounted PV plants, certain plant and design, among other variables. The insurance pre- components will need to be replaced over an FPV mium can be paid up front or in installments, subject system’s operating lifetime even though most should to the project size, and will protect the project’s rev- be operational for at least 20 years. The highest risk comes from the inverters. Experience from the field 25. https://solargis.com/blog/best-practices/how-to-calculate-p90- shows that a “mean time between failures” of 1-16 or-other-pxx-pv-energy-yield-estimates/ 98 •  FLOATING SOLAR MARKET REPORT years can be observed, and inverter manufacturers 5.1.3  Residual value/decommissioning typically offer warranties over a 5-12 year period. In this example, it is assumed that the residual value of Therefore, with an offtake contract tenure of 20 years, the floats (recycled plastic), the module frames (alumi- the replacement cost of inverters needs to be taken num), and the cables (copper) would be used to cover into account at least once during the operation of the decommissioning costs. This assumption will need to PV assets. Apart from accounting for the replacement be further verified with additional experience from the cost of inverters at the time of failure, the inverter industry and as FPV projects reach the end of their supplier usually offers an option of buying a warranty operating lifetime. extension for another five years at about 20 percent of the prevailing inverter cost. A detailed cost-ben- efit analysis needs to be carried out to compare the Calculating the levelized cost 5.2  expected operating lifetime of the inverters against of electricity the cost of warranty extension. 5.2.1  Financial assumptions For this present analysis, it is assumed that the war- Financial assumptions vary substantially from one ranty will be extended in five-year intervals. The war- country to another, and largely depend on which risk ranty extension cost is assumed to increase with the mitigation mechanisms have been put into place to age of the inverter portfolio. An inverter manufacturer ensure the reliable operations of an FPV system over might be less willing to extend a 10-year-old inverter 20 years. Experience from sponsors, developers, and portfolio (when some but probably not all inverters EPC and O&M contractors are paramount to build were replaced in the previous five-year period) than a trust among investors. Given that the deployment 5-year-old one. For the base case, it is assumed that of large-scale FPV systems remains limited to date, the warranty extension cost will be 20 percent of the lenders and potential institutional investors might prevailing inverter price in year 5, 45 percent in year require a higher cost of capital to compensate for 10, and 60 percent in year 15. The increase of the the lack of experience in this market segment. Three premium reflects the inverter supplier’s reluctance to WACC scenarios are therefore considered. The same extend the warranty in line with the increasing age financial assumptions have been used to calculate of the inverter fleet. Inverter prices are assumed to the LCOE of hypothetical 50 MWp ground-mounted continue a slow declining trend, leveling out at about and FPV systems, as detailed in table 5.3. $0.05/Wp in year 10. The nominal amount of all invert- er warranty expenses over the project’s 20 years of TABLE 5.3. Financial assumptions used to calculate operation would be calculated on an annual basis the levelized cost of electricity for 50 MWp ground-mounted and FPV projects (not discounted) at about $200,000 (equivalent to $0.004/Wp). Based on this methodology, the inverters Assumption are assumed to be replaced about 1.33 times in the Debt equity ratio 80:20 20-year period. WACC Scenario A: 6% Scenario B: 8% Some inverter manufacturers also offer the option of Scenario C: 10% paying a one-off premium to extend the 5-year inverter Debt premium 4% warranty into a 20-year warranty, at a cost equivalent Maturity of loan 10 years to about 60-70 percent of the initial inverter purchasing Inflation rate 2% price. Based on an initial price of $0.06/Wp, this would Economic system life 20 years add a cost of $0.039/Wp to the initial investment costs, Source: Authors’ compilation. which is quite significant. This option has therefore not Note: LCOE = levelized cost of electricity; MWp = megawatt-peak; been modelled. WACC = weighted average cost of capital. CHAPTER 5:  COSTS OF FLOATING SOLAR  • 99 BOX 5.1 Methodology The LCOE is calculated by dividing the entire lifecycle cost of an FPV system by its cumulative solar electricity generation. It is presented in net present value terms, with each year’s cost discounted by the investor’s hurdle rate. For this particular generic analysis, some simplifications have been used: • No interest during construction, as lenders often offer a grace period • No residual value/decommissioning cost • No taxes, as these vary significantly across jurisdictions The LCOE (before tax) formula used in this analysis is shown below: *Inflation adjusted Where: EPCI =  Equity project cost investment IEI =  Inverter warranty extension investment IC =  Insurance cost LP =  Loan payment N =  Number of years in the system’s service life IRD = Irradiance OM =  Operation and maintenance PR =  Performance ratio DR =  Nominal discount rate SDR =  System degradation rate The numerator sums up all the possible cost items over interests and amortizations. The denominator includes the system’s entire lifetime. The investment cost com- the system’s lifetime electricity generation. The specific prises the equity project cost investment (EPCI). The yield is the energy yield of the system in the first year, annual operating cost is split in two parts, namely the which is calculated by the product of the available irra- operating and maintenance cost (OM) and the insur- diance (IRD) and the performance ratio (PR). After the ance cost (IC). The inverter warranty extension invest- first year, the generation output is annually adjusted ment (IEI) represents the warranty extension cost for according to the system degradation rate (SDR). Both the systems’ entire operating life. The year in which the values are discounted by the nominal discount rate warranty is extended depends on inverter suppliers. (DR) for net present value calculations, which is based The model assumes a warranty extension at years 5, on the weighted average cost of capital (WACC) con- 10, and 15. In case a part of the up-front CAPEX is cept. OM, IC, and IEI are adjusted with the inflation rate debt financed, the loan payments (LP) include annual after the first year. 5.2.2  Energy yield ditions. Therefore, the irradiation level and ambient temperatures where the project is located are key vari- The key difference between FPV and ground-mount- ables that will influence the energy yield and thus the ed PV projects is the modelling of the cooling effect LCOE of projects. due to water evaporation. It has been reported across the world that FPV systems have a higher energy yield Preliminary results show that in hotter climates, the than ground-mounted PV systems under similar con- energy yield gain of an FPV plant over a ground-mount- 100 •  FLOATING SOLAR MARKET REPORT ed one is higher than in temperate climates, since The representative “average” P50 global horizontal irra- the cooling effect of water makes a great difference diance and performance ratio for ground-mounted PV to their relative efficiency. This means that in certain figures has been estimated for each climate zone (table regions of the world, the energy yield gain could be 5.4). The performance ratio of FPV systems under sim- around 10 percent (typically in warmer regions with ilar conditions is estimated to increase by 5 percent in a global horizontal irradiation higher than 1,600 kilo- the conservative scenario and 10 percent in the opti- watt-hour per square meters per year [kWh/m /year]) 2 mistic scenario. The bold underlined PR values are the while in other regions it would be only about 5 percent “likely” cases per climate zone. (typically in colder regions or where irradiation is low- er than 1,600 kWh/m2/year). However, more studies Table 5.5 shows the energy output of hypothetical 50 are needed to verify this assertion and to more accu- MWp ground-mounted and FPV plants in their first rately quantify the correlation between energy yield year, across the three climates. gains and various climates. Because the FPV market is nascent and lacks empirical data, the analysis here 5.2.3  System degradation rate uses preliminary estimates of the energy yield gain As of the end of 2018, there are no sufficient records of FPV projects in three climates. These estimates are yet for the degradation rates of FPV systems. Gener- based on assumptions, and require verification when ally, crystalline silicon modules degrade at a rate of data become available. no greater than 0.8 percent to 1.0 percent per year, respectively. It is assumed here that the annual system Three types of climates are considered in the LCOE degradation rate is 1 percent (Ye et al. 2014) in a tropi- calculations: temperate, tropical, and arid/desert. Cold cal climate, 0.7 percent in an arid/desert climate (Cop- and polar climates have been excluded from the anal- per, Jongjenkit, and Bruce 2016), and 0.5 percent in a ysis as building large-scale solar PV plants in these temperate climate (Jordan and Kurtz 2013). regions is less likely. TABLE 5.4. Representative average global horizontal irradiance and performance ratio, by climate zone Floating PR (%) GHI Ground-mounted PR Conservative Optimistic (kWh/m2/year) (%) (+5%) (+10%) Tropical 1,700 75.0 78.8 82.5 Arid/desert 2,300 75.0 78.8 82.5 Temperate 1,300 85.0 89.3 93.5 Source: SERIS estimations based on: Baker et al. 2015; and Reich et al. 2012. Note: GHI = global horizontal irradiance; kWh/m2/year = kilowatt-hours per square meter per year; PR = performance ratio. TABLE 5.5. First year’s energy output, by climate Floating PV (GWh) Ground-mounted PV (GWh) Conservative (+5%) Optimistic (+10%) Tropical 63.8 66.9 70.1 Arid/desert 86.3 90.6 94.9 Temperate 55.3 58.0 60.8 Source: SERIS calculations based on estimated data. Note: GWh = gigawatt-hour; FPV = floating photovoltaic; PR = performance ratio. CHAPTER 5:  COSTS OF FLOATING SOLAR  • 101 5.2.4  LCOE calculation results • Same insurance cost The following assumptions were made for both ground- • Same O&M costs: even though this assumption mounted and FPV technologies: can be argued; therefore a sensitivity analysis on this variable (+15 percent for FPV) will be provided • No lease cost, since this varies widely across proj- in the next section ects and regions • Same system degradation rate • No contingency costs (typically at 3 percent of • Calculated on a pretax basis EPC costs [NREL 2017]) • Same inverter replacement methodology Ideally, to fine-tune this analysis, system prices, O&M costs, insurance, and inverter warranty extension costs TABLE 5.6. Summary of assumptions used in calculations General assumptions Ground-mounted Floating System size (MWp) 50 50 System price ($/Wp) 0.62 0.73 O&M costs ($/Wp/year) 0.011 0.011 Yearly insurance (in % of 0.3% 0.3% system price) Inverter warranty extension Year 5: 20% of prevalent price Year 5: 20% of prevalent price Year 10: 45% of prevalent price Year 10: 45% of prevalent price Year 15: 60% of prevalent price Year 15: 60% of prevalent price ~$0.004/Wp ~$0.004/Wp Debt equity ratio 80:20 80:20 WACC 6% / 8% / 10% 6% / 8% / 10% Debt premium (%) 4% 4% Maturity of loan (years) 10 10 Surface lease cost ($/year) — — Inflation (%) 2% 2% Years of operation 20 20 Floating PR (%) Climate-related GHI System degradation Ground-mounted Conservative Optimistic assumptions (kWh/m2/year) rate (%) PR (%) (+5%) (+10%) Tropical 1,700 1.0 75.0 78.8 82.5 Arid/desert 2,300 0.7 75.0 78.8 82.5 Temperate 1,300 0.5 85.0 89.3 93.5 Source: SERIS. Note: GHI = global horizontal irradiance; kWh/m2/year = kilowatt-hour per square meter per year; MWp = megawatt-peak; O&M = operation and maintenance; PR = performance ratio; $/Wp = U.S. dollar per watt-peak; WACC = weighted average cost of capital. 102 •  FLOATING SOLAR MARKET REPORT TABLE 5.7. Results of (before tax) calculations LCOE ($cents/kWh) Ground-mounted PV 50 MWp Floating PV 50 MWp Conservative Optimistic (+5% PR) (+10% PR) Tropical WACC 6% 6.25 6.77 6.47 8% 6.85 7.45 7.11 base case 10% 7.59 8.28 7.91 Arid/desert WACC 6% 4.52 4.90 4.68 8% 4.96 5.39 5.15 10% 5.51 6.01 5.74 Temperate WACC 6% 6.95 7.53 7.19 8% 7.64 8.30 7.93 10% 8.49 9.26 8.85 Source: SERIS calculations. Notes: kWh = kilowatt-hour; LCOE = levelized cost of electricity; MWp = megawatt-peak; PV = photovoltaic; WACC = weighted average cost of capital. The bold LCOE values are the “more likely” cases per type of climate. should also be varying by location/climate. Without 5.3  Sensitivity analysis empirical data on these particular variables, the analy- The following scenario was chosen as the base case to sis considers their costs to be similar across the three perform a sensitivity analysis: climate zones. Base case = 50 MWp FPV system in a tropical In the conservative scenario (+5 percent PR), the climate with a WACC of 8% and an optimistic LCOE of the FPV system is between 8 and 9 percent PR (+10%) higher than the LCOE of the ground-mounted PV LCOE ($cents/kWh) = 7.11 system, while in the optimistic scenario (+10 percent PR), the FPV LCOE is only 3-4 percent higher than Reduced CAPEX (-15 percent) and a higher perfor- the ground-mounted LCOE (table 5.7). This difference mance ratio (88 percent) will have the highest positive is likely to reduce, become zero, or even reverse as impact on LCOE, while higher CAPEX (+15 percent) FPV volumes grow and anticipated cost reductions and a reduction of 5 years of the operational lifetime are realized (installed capacity today is still very small will have the highest negative impact on the LCOE, compared to ground-mounted PV systems around as depicted in figure 5.3. A 2 percent change in the the world). WACC, even though not reflected in the figure but cal- culated in table 5.7, will also have a significant impact The LCOE calculation represents only a “break-even” on the LCOE, almost as important as a 15 percent analysis—that is, if the tariff were set at the LCOE, the change in CAPEX. This highlights the fact that conces- net present value of the project would be zero. Equity sionary financing from multilateral and bilateral lenders investors would presumably require a higher tariff from could boost FPV adoption. the offtaker to make the project economically viable for them, assuming debt financing was accessible. CHAPTER 5:  COSTS OF FLOATING SOLAR  • 103 FIGURE 5.3. Levelized cost of electricity sensitivities vs. base case Base Case CAPEX ($/Wp) –15% 0.73 +15% Years of operation +5 years 20 –5 years System degradation rate 0.5% 1% 2% Irradiance P50 P90 Yearly insurance 0.1% 0.3% 1% Performance Ratio 88% 82.5% 78% Financial leverage 90% 80% 70% O&M ($/Wp) –15% 0.011 +15% –0.8 –0.6 –0.4 –0.2 7.1 HIH 0.2 0.4 0.6 0.8 $ cents/kWh (FPV LCOE) Source: SERIS calculations. Note: CAPEX = capital expenditure; FPV = floating photovoltaic; LCOE = levelized cost of electricity; O&M = operation and maintenance; $/Wp = U.S. dollar per watt-peak; $ cents/kWh = U.S. dollar cents per kilowatt-hour. 5.4  Risk assessment FPV deployment opportunities will be mainly driven by (i) jurisdictions where permitting favors them and From a financing perspective, risks associated with where (ii) access to land and the scarcity/price thereof new technologies like FPV are critical for the premium are major issues. on interest rates compared to more established forms of PV deployment like ground-mounted systems. Five Compared to rooftop and ground-mounted PV instal- main risk categories are outlined in table 5.8. lations, MW-scale FPV is brand new. This technology is at the earlier stages of its learning curve, and great- 5.5 Conclusion er cost reductions are to be expected. This is not only true for the cost of the floating system itself, but also for There is no significant difference in the LCOE of engineering and project development costs. As will be ground-mounted, fixed-tilt systems and FPV installa- shown in the following chapter, only a few EPCs have tions. The higher initial capital costs of FPV systems realized a sizeable number of FPV plants. are mostly balanced out by their higher energy output. Meanwhile, other considerations might favor FPV, such Finally, it is important to differentiate between risks and as the opportunity costs of using agricultural land. FPV unknowns. Increased transparency and knowledge costs are approaching those of ground-mounted sys- sharing with regards to the capital costs, environmen- tems and may converge in time, eventually leading to tal impact, and performance of FPV systems will help an equal or lower LCOE. build trust among international investors and lenders, which will in turn help reduce financing capital costs. 104 •  FLOATING SOLAR MARKET REPORT TABLE 5.8. Overall FPV risk assessment Risk category Comment Technology/capital • Even though deployment of floating photovoltaic (FPV) systems remains limited to expenditure (CAPEX) risk  date, the technology risk on inland freshwater reservoirs is considered low given the fact that developers apply experiences from (i) the established forms of PV deployment, especially ground-mounted PV systems; and (ii) the offshore and maritime industry where floating structures made of high-density polyethylene (HDPE) and mooring and anchoring has been applied for decades. Nevertheless, quality matters, and especially floats (which have the shortest track record) need to undergo thorough stress testing and certifications of their long-term durability and reliability. • Mooring complexity, corrosion and aging, equipment fatigue, and the impact of waves and wind must be carefully analyzed to find the appropriate FPV system design. All of these elements will influence a project’s structural and mooring costs. • “The floating dynamics of the FPV system may lead to fatigue-based micro-cracking in the panels over time. This will lead to reductions in the performance of the panels but the magnitude of this deterioration is not yet fully understood” (Leybourne 2017). Operation and • In principle, the O&M costs of FPV systems should not be higher than for ground- maintenance (O&M) risk  mounted PV, although accessibility may play a role, as in most cases boats need to be used and the replacement of components may be more complex. • O&M requirements depend on the context. For example, severe soiling due to bird drop- pings will either increase O&M costs or reduce energy yield if not dealt with properly. Financial risk • FPV is still at a nascent stage, hence the long-term data needed for statistical analysis and to assess the performance of loans do not exist. Most projects are still being financed on balance sheets. • Technical due diligence takes longer where nonrecourse (or limited recourse) project financing is involved due to FPV systems’ lack of a track record. This can lengthen the process of reaching financial close. However, this risk is expected to reduce in time as more and more FPV projects are built around the world and their track record information becomes available to developers and financial institutions. • From an economic perspective, FPV projects will generally have a high share of domes- tic content, thereby having a positive impact on the local economy, as it is much more economical to manufacture HDPE floats locally. This in turn will have a positive impact on job creation, local-currency funding, and the development of a local commercial financial sector (reducing foreign exchange risks, unlike when most equipment is imported). • Interestingly, Sungrow reported the following on the bankability of FPV systems: “The banks are willing to provide us financial support because even though the ROI of these floating plants can be a little bit lower than the other ground-mounted PV plants, this kind of plant does not have a real estate problem” (PV Tech 2017). • It is expected that investors and financing organizations would collaborate with public or semi-public utilities that own a series of water bodies (e.g., water, rural/agricultural author- ities) to adopt a portfolio approach with a series of projects in different geographic loca- tions to diversify risks. In rural areas, it is important to involve local communities, especially if they are depending on the reservoir for other activities such as fishing. Regulatory risk • The ownership of the water body and/or water surface, as well as the contractual setup of FPV projects, is an important point to consider. This will vary by jurisdiction. In view of the lack of specific regulatory frameworks, diligent legal advice will be required to ensure that the right business model and project structure are chosen. • Enforcement of typical lenders’ securities must also be analyzed as challenges may arise when the owner of the asset (FPV system) is not the same as the owner of the reservoir (surface). • “There are projects in which the opportunity for mitigating risks is not undertaken due to lack of clarity as to who bears the overall responsibility” (IEA 2017). CHAPTER 5:  COSTS OF FLOATING SOLAR  • 105 TABLE 5.8 continued Risk category Comment Environmental risk • There is still a lack of empirical data on the long-term environmental impact of FPV installations on reservoirs. Developers tend to cover only small fractions of reservoirs and regularly monitor whether there are any adverse impacts on the water quality and flora/ fauna. Installed projects to date (with a 2–3 year track record) have shown that they are not creating any negative environmental impacts. • This topic is of great importance to the FPV community and therefore will be addressed in detail in the next next publication of Where Sun Meets Water1 series, to follow shortly after this publication. Source: Authors’ compilation. Note: ROI = return on investment. World Bank Group, ESMAP and SERIS. 2019. “Where Sun Meets Water: Floating Solar Handbook for Practitioners.” 1.  Forthcoming. Washington DC: World Bank. References Baker, R. S., M. Bieri, W. K. Cher, K. Zhang, R. Rüther, T. Reindl, and A.M. Nobre. 2015. “National Solar Repository—5-Year Country-Wide Assessment of PV Systems’ Performance in Singapore.” 31st European Photovoltaic Solar Energy Conference and Exhibition, Hamburg, Germany. Bellini, E. 2018. “Module Prices Have Decreased by up to 25% so Far This Year, TrendForce Says.” PV Magazine, October 9. https://www.pv-magazine.com/2018/10/09/module-prices-have-decreased-by-up-to-25-so-far-this-year-trendforce-says/. Copper, J., K. Jongjenkit, and A. Bruce. 2016. “Calculation of PV System Degradation Rates in a Hot Dry Climate.” Confer- ence paper at Asia Pacific Solar Research Conference, Canberra, Australia, December 2016. Energy Trend. https://www.energytrend.com/pricequotes.html. 2018/11/07 update. Fraunhofer ISE. 2018. “Levelized cost of electricity—Renewable energy technologies”, Freiburg, Germany, March 2018. Authors: Christoph Kost, Shivenes Shammugam, Verena Jülch, Huyen-Tran Nguyen, Thomas Schlegl. Hutchins, M. 2018. “The Weekend Read: Staying Afloat, Whatever the Weather.” PV Magazine, November 3. https://www. pv-magazine.com/2018/11/03/staying-afloat-whatever-the-weather/. IEA (International Energy Agency). 2017. Technical Assumptions Used in PV Financial Model: Review of Current Practices and Recommendations. Report IEA PVPS T13 08-2017. Switzerland: International Energy Agency Photovoltaic Power Systems Programme. Jordan, D. C., and S. R. Kurtz. 2013. “Photovoltaic Degradation Rates—An Analytical Review.” Progress in Photovoltaics: Research and Applications 21 (1): 12–29. Kenning, T. 2018. “Akuo Energy Starts Construction on 17MW Floating Solar Plant in France.” PV Tech, September 25. https:// www.pv-tech.org/news/akuo-energy-starts-construction-on-17mw-floating-solar-plant-in-france. Korea Energy Agency. N.d. “Renewable Portfolio Standards(RPS).” http://www.energy.or.kr/renew_eng/new/standards.aspx. Lazard. 2018. “Lazard’s levelized cost of energy analysis—version 12.0”, November 2018. Leybourne, M. 2017. “Floatovoltaics—Thinking Beyond the Cost of Floating Solar PV.” Linked in, July 4. https://www.linkedin. com/pulse/floatovoltaics-thinking-beyond-cost-floating-solar-pv-mark-leybourne/. Lightsourcebp. 2016. “Reservoir Floating Solar.” Lightsourcebp, May 2. https://www.lightsourcebp.com/uk/stories/qe2/. Mass.gov. N.d. “Solar Massachusetts Renewable Target (SMART) Program: General Information.” https://www.mass.gov/ info-details/solar-massachusetts-renewable-target-smart-program#general-information-. Ministry of Economic Affairs. 2018. “2018 Renewable Energy FIT Rates Are Announced.” Bureau of Energy, Ministry of Economic Affairs, January 25. https://www.moea.gov.tw/MNS/english/news/News.aspx?kind=6&menu_id=176&news_ id=76457. 106 •  FLOATING SOLAR MARKET REPORT Netherlands Enterprise Agency. N.d. “Stimulation of Sustainable Energy Production (SDE+).” https://english.rvo.nl/subsidies- programmes/sde. NREL. 2017. “U.S. Solar Photovoltaic System Cost Benchmark: Q1 2017”, September 2017. Authors: Ran Fu, David Feldman, Robert Margolis, Mike Woodhouse, and Kristen Ardani. Planair and PITCO. 2017. Assessment of Floating Solar PV Potential for Pakistan; Task 1: Commercial Readiness of FSPV— Global Market and Performance Analysis. Unpublished report prepared for International Finance Corporation, Washing- ton, DC. Prateek, S. 2018. “Breaking: West Bengal Auctions 5 MW of Floating Solar PV.” Mercom India, March 16. https://mercomindia. com/west-bengal-auctions-5-mw-floating-solar-pv/. PV Tech. 2017. “World’s Largest Floating Solar Project: Making Use of the Unusable.” PV Tech Power Volume 12, September 2017. https://www.pv-tech.org/technical-papers/worlds-largest-floating-solar-project-making-use-of-the-unusable. PVinsights. http://pvinsights.com/index.php. 2018/11/04 update. Reich, N. H., B. Mueller, A. Armbruster, W. G. J. H. M. van Sark, K. Kiefer, and C. Reise. 2012. “Performance Ratio Revisited: Is PR > 90% Realistic?” Progress in Photovoltaics: Research and Applications 20 (6): 717–26. SolarBankability. 2017. “Technical Bankability Guidelines—Recommendations to Enhance Technical Quality of Existing and New PV Investments.” SolarBankability, February 15. Speer, B., M. Mendelsohn, and K. Cory. 2010. Insuring Solar Photovoltaics: Challenges and Possible Solutions. Technical Report NREL/TP-6A2-46932 Revised February 2010. Golden, Colarado: National Renewable Energy Laboratory. Ye, J. Y., T. Reindl, A. G. Aberle, and T. M. Walsh. 2014. “Performance Degradation of Various PV Module Technologies in Tropical Singapore.” IEEE Journal of Photovoltaics 4 (5): 1288–94. World Bank Group, ESMAP and SERIS. 2019. “Where Sun Meets Water: Floating Solar Handbook for Practitioners.” Forthcoming. Washington DC: World Bank. CHAPTER 5:  COSTS OF FLOATING SOLAR  • 107 UNITED KINGDOM © Lightsource BP Floating Solar Array, London. 6 SUPPLIERS OF FLOATING PV SYSTEMS 6.1.  General overview Among the float suppliers entering the market are inverter manufacturers, developers of large solar PV The ecosystem of floating photovoltaic (FPV) power is projects, suppliers of mounting structures, plastic similar to that of other PV applications, with the addi- manufacturers, and firms active in related engineering tion of suppliers of float systems. The main industry fields, such as offshore and marine industries. Some players are investors, sponsors, developers, contrac- suppliers are start-ups; others are subsidiaries of play- tors (for services including engineering, construc- ers established in their own industry. tion, operation, and maintenance), and suppliers (of PV modules, float systems, and other equipment and Most suppliers of FPV systems have their own propri- components). Many of these players are active in the etary design of floats and floating systems. Some man- ground-mounted and rooftop PV sector, and in other ufacture their own floats, whereas others procure them renewable energy systems. from third parties. In addition, an increasing number of FPV system suppliers also offer engineering, pro- A growing number of developers is expanding their curement, and construction (EPC) and operations and portfolios to include FPV by integrating floating plat- maintenance (O&M) services, with a few being able to form technologies. They often start by partnering with offer even full turnkey solutions (figure 6.1). float suppliers. The new and unique element applica- ble to FPV is thus the design and supply of the floating Float suppliers may cooperate with EPC contractors platform as a key structural component, which often or developers to deliver complete FPV system solu- also includes the design and supply of the anchoring tions. For example, Ciel & Terre International collabo- and mooring solution. rates with Innova Capital Partners in Colombia,26 Akuo FIGURE 6.1. FPV ecosystem (simplified) O&M contractor Developer / Owner OEMs: PV modules, floats, inverters, cables EPC contractor Floating PV system integrator Contract Support Source: Authors. Note: O&M = operations and maintenance; EPC = engineering, procurement, and construction; OEM = original equipment manufacturer; PV = photovoltaic. CHAPTER 6:  SUPPLIERS OF FLOATING PV SYSTEMS  • 109 Energy in France, and D3Energy in the United States.27 supplies the FPV system. These collaborations mean Some suppliers have even started certifying local EPC that certain projects appearing in the table may have contractors. As a result, more and more EPC contrac- been double-counted. tors with experience in ground-mounted or rooftop PV technologies are being trained to install FPV plants. As of early 2019, the FPV market is currently dominat- With regard to the anchoring and mooring system, they ed by two main system suppliers: Sungrow Floating generally borrow expertise from the marine industry, and Ciel & Terre International. Both might be more or even develop in-house expertise through their own accurately described as system integrators. research and development. For example, Sweden’s Seaflex provides anchoring and mooring solutions for FPV in the Republic of Korea, India, and the United Providers of floating 6.2.  States, in partnership with FPV system suppliers. As a technology solutions for inland result of such collaboration, many float suppliers can freshwater applications provide design services, or at least advice, related to anchoring and mooring. This section describes in alphabetical order suppliers of FPV technology for inland freshwater applications Some FPV system suppliers are also exploring near- that could claim an FPV installed capacity of at least shore or offshore floating applications, where different 5 MWp in December 2018. The section 6.3 describes technologies and more robust designs are required. two suppliers of systems for offshore or near-shore Section 6.3 offers a detailed look at several of these applications of FPV. Some suppliers develop FPV sys- suppliers. tems for both inland and offshore applications. Apart from float suppliers, manufacturers of other 6.2.1.  Ciel & Terre International equipment are exploring the development of products Ciel & Terre International (C&T) has been developing and solutions specifically for FPV applications. These FPV plants for commercial, industrial and government include polymer resins, PV modules, inverters, DC applications since 2011 (figure 6.2). By the end of cables, and other mechanical support components. 2018, the company had installed in excess of 300 MWp However, this trend is still in its infancy and will not be of FPV systems with more than 130 installations in 30 covered in this chapter. countries, including in Europe, the United States, India, Australia as well as in several Asian countries. More Table 6.1 presents a nonexhaustive list of FPV system than 80 projects are in Japan, but the two largest—with suppliers as of end 2018. The first part lists major capacities of 70 MWp and 32 MWp— are in China. The FPV system suppliers—i.e., those with a cumula- company’s project portfolio shows experience in man- tive installed capacity in excess of 5 megawatt-peak aging all aspects of a project’s development by supply- (MWp); the second lists other FPV system suppliers. ing design, financing, EPC, and O&M services. In both parts, the companies appear in alphabetical order. Many companies are entering the FPV market, C&T’s brand Hydrelio, a modular system consisting of so the table below is not exhaustive. two types of HDPE floats, is a well-recognized brand in the industry. Three different models are currently avail- Some of the suppliers mentioned in table 6.1 are col- able: laborating on FPV projects, such as LSIS and Scotra in Korea, where LSIS generally performs EPC and Scotra • The Classic (12° tilt) • The Equatorial (5° tilt designed for lower-latitude 26. https://www.pv-tech.org/news/ciel-terre-and-innova-capital-to- countries) develop-floating-solar-in-colombia • The aiR (adaptable angle up to 15°). 27. http://www.d3energy.com/about-us.html 110 •  FLOATING SOLAR MARKET REPORT TABLE 6.1. Nonexhaustive list of inland freshwater and offshore FPV system suppliers as of December 2018 Services offered Total Total FPV FPV capacity under Turn- capacity development/ Country (Co-) key Location of installed construction Company name of origin Owner EPC O&M Others completed FPV projects FPV technology (MWp) (MWp) Website MAJOR FPV SYSTEM SUPPLIERS (INSTALLED CAPACITY ≥ 5 MWp) Ciel & Terre France Floating system design Worldwide Specialized pure 319 330 https://www.ciel-et-terre. 4 6 4 International and procurement HDPE floats net/ Jintech New China Floating system design China Specialized pure 150 80 http://www.jnnewenergy. 4 4 0 Energy and procurement HDPE floats com Kyoraku Co. Japan Floating system design Japan, Taiwan, China, Specialized pure 51 N/A http://www.krk.co.jp/ 0 6 0 and procurement, Thailand HDPE floats tracking LG CNS Korea, Floating system design Korea, Rep. Floating island + 6 80 http://lgcns.co.kr/ 4 4 4 Rep. and procurement racks LS Industrial Korea, Floating system design Korea, Rep., Japan Floating island + 30 250 http://www.lsis.com/ko/ N/A 4 4 Systems Co. Rep. and procurement racks NorthMan China Floating system design China Specialized pure 230 N/A https://netsolar.solarbe. Energy 4 4 0 and procurement HDPE floats com/ Technology SCG Thailand Floating system design Thailand, Specialized pure 5 N/A https://www.scgchemi- 4 4 4 Chemicals and procurement Singapore HDPE floats cals.com/en Scotra Co. Korea, Floating system design Korea, Rep., Japan, Floating island + 40.3 19.3** http://www.scotra.co.kr/ Rep. N/A 4 4 and procurement Taiwan, China, racks en/ Philippines Sumitomo Japan Floating system design Japan, Singapore, Specialized pure 9.7 100 https://pv-float.com/ Mitsui Con- 4 4 4 and procurement Thailand, Taiwan, HDPE floats english/ struction Co. China Sungrow China Floating system design China, Germany, Israel, Specialized pure 500 600 https://en. and procurement, Japan, Philippines, HDPE floats sungrowpower.com/ 4 4 4 specialized platform Singapore, Thailand, product_category?id= for central inverters. Taiwan, China 22 Full turnkey Xiamen Mibet China Floating system design Brazil, China, Ger- Specialized pure 30 120 https://www.mbt-energy. New Energy and procurement, float many, Israel ,Japan, HDPE floats com/ 4 6 4 Co. supplier, tracking Southeast Asia, Spain, Taiwan, China, 0 = No    6 = Limited support    4 = Yes    N/A = Not available CHAPTER 6:  SUPPLIERS OF FLOATING PV SYSTEMS  • 111 TABLE 6.1. continued Services offered Total FPV Total FPV capacity under Turn- Location of capacity development/ Country (Co-) key completed FPV installed construction Company name of origin Owner EPC O&M Others projects FPV technology (MWp) (MWp) Website OTHER FPV SYSTEM SUPPLIERS (INSTALLED CAPACITY < 5 MWp) Floating Solar Netherlands Floating system design N/A Floating island N/A N/A https://floatingsolar.nl/en and procurement, (with pipes) + N/A 4 4 tracking frames, optimal 112 •  FLOATING SOLAR MARKET REPORT solar tracking ISIGENERE Spain Floating system design Spain, Chile Specialized pure 1.9 50 https://isifloatingcom. 0 6 6 and procurement HDPE floats wordpress.com/ Koiné Multime- Italy Floating system design Singapore, Italy, Floating island 0.4 N/A http://www. dia (Upsolar 0 4 0 and procurement, Korea, Rep. (with pipes) + koinemultimedia.eu/wp/ Floating) tracking, concentration frames NRG Energia Italy Floating system design Italy, Iran, France, Floating island + 1 10 http://www.nrg-energia. 0 4 4 and procurement India, Canaries racks it/index-en.html Island Oceans of Netherlands Floating system design Netherlands Offshore floating N/A N/A https://oceansofenergy. Energy 0 0 0 and procurement, systems *** blue/ mooring systems Ocean Sun Norway Floating system design Norway, Singa- Floating island + 0.1 2.2 http://oceansun.no/ 0 4 0 and procurement pore membrane *** ProFloating Netherlands Floating system design N/A Specialized pure N/A N/A https://profloating.eu/ 0 6 0 and procurement HDPE floats en/ 4C Solar USA Floating system design Singapore, Chile, Floating island N/A N/A https://www.4csolar. and procurement, Maldives (with pipes) + com/ N/A N/A N/A tracking frames, one-ax- is tracking + offshore SolarisFloat Portugal Floating system design N/A Modular floating N/A 20 https://www.solarisfloat. and procurement, platforms with com/ N/A N/A N/A tracking two-axis tracking *** 0 = No    6 = Limited support    4 = Yes    N/A = Not available TABLE 6.1. continued Services offered Total FPV Total FPV capacity under Turn- Location of capacity development/ Country (Co-) key completed FPV installed construction Company name of origin Owner EPC O&M Others projects FPV technology (MWp) (MWp) Website OTHER FPV SYSTEM SUPPLIERS (INSTALLED CAPACITY < 5 MWp) Solaris Israel Floating system design Israel, Singapore, Special island 1 50 http://www.solaris- Synergy 0 4 0 and procurement, USA design with HDPE synergy.com/ tracking floats + frames Sunengy Australia Floating system design India Plastic concentra- N/A N/A http://sunengy.com/ and procurement tors with tracking, 0 0 0 mounted on rafts (Liquid Solar Array) *** Sunfloat Netherlands Floating system design Netherlands Floating island N/A N/A http://www.sunfloat. N/A N/A N/A and procurement, (with pipes) + com/ tracking (bifacial) aluminum frames Sun Rise E&T Taiwan, Floating system design Japan Floating island N/A N/A http://www.srise.com. Corporation China 0 0 0 and procurement (with pipes) + tw/v2/ frames Swimsol Austria Floating system design Maldives Offshore modular 0.2 0.4 https://swimsol.com/ and procurement, floating platforms 4 4 4 floating substructure *** supplier Takiron Japan Floating system design Japan Floating island + N/A N/A https://www.takiron. N/A N/A N/A Engineering and procurement racks co.jp/english/ 0 = No    6 = Limited support    4 = Yes    N/A = Not available Source: Authors’ compilation based on information received from suppliers and/or their websites. Notes: HPDE = high-density polyethylene. *O&M for own projects only. ** Under construction, excluding bidding projects. *** R&D or early stage of commercialization. CHAPTER 6:  SUPPLIERS OF FLOATING PV SYSTEMS  • 113 FIGURE 6.2. Examples of C&T FPV projects in Japan (left) and Brazil (right) Source: © Ciel & Terre International. Source: © Ciel & Terre International. The C&T system has been designed to be assem- tial adverse effects on the environment. This includes bled like Lego blocks with screws and nuts made of irrigation ponds, mining lakes, water retention ponds, polypropylene and fiberglass. The modular approach waste water treatment ponds, industrial reservoirs, and allows for installations from the kilowatt-peak (kWp) to hydroelectric dams. the gigawatt-peak range. Hydrelio also offers ways to enhance the final design by adjusting the buoyancy, The system is designed to be safe to install, as no energy yield, Wp/m², ease of access (for O&M), and heavy tools or machinery are required. Furthermore, footprint of the system layout. during project development and construction, C&T offers system integrators assembly instructions and C&T claims that its technology is tested to ensure risk assessments such as development support, engi- endurance against severe tension loads over the life- neering expertise, EPC support, O&M services and time of the plant. Being a modular technology, it allows financing solutions. for many different configurations and layouts, depend- ing on the degree of buoyancy and stability required. 6.2.2. Kyoraku To offer a bankable and long-lasting solution, C&T’s Kyoraku Co., Ltd. was established in 1917 as a real FPV projects undergo testing for reliability, quality, and estate development company initially. The company safety related to compliance with drinking water stan- started manufacturing and selling plastics in 1947 to dards; floatability; and resistance to wind, waves, cur- become one of the leading suppliers of plastics in rent, ultraviolet rays, and temperature, among others. Japan today. The company provides a wide range of plastic products used in various industries, including A key requirement for floats is not to affect water qual- the food and beverage industry. ity, especially when deployed in drinking water reser- voirs. Hydrelio floats have been certified as “drinking Based on their long-standing experience as a blow water compliant” pursuant to tests performed by the molder, Kyoraku has developed a floating structure English Independent Water Quality Control Center, specifically for floating solar systems, called the “Min- attesting that the installation is safe on water intend- amo Solar System”. As of January 2019, the compa- ed for human consumption. C&T is also focusing on ny has provided float structures for 33 different FPV man-made reservoirs in which wildlife is limited or even projects totaling 51 MW. Most of the FPV systems are entirely absent, with an eye to minimizing any poten- located in Japan (figure 6.3). 114 •  FLOATING SOLAR MARKET REPORT FIGURE 6.3: Examples of Kyoraku’s FPV systems in Japan Source: © Kyoraku. Source: © Kyoraku. Leveraging on their 60 years of experience in produc- ing types. The company handles the logistics of all ing plastic containers for the high-standard food and components—including panels, structures, floating medical industries, the company only uses food-grade objects, and wiring. material resin to develop their floats. They also have developed specific expertise in developing outdoor Floating structures of LG CNS include multiple marine buoys. design configurations depending on the application and environment (e.g. wind load and water surface The structure of their floats is relatively flat with few motion). They can include a frame (array), pure connections points. Their float system can resist 65 HDPE float matrix, as well as mats or membranes. m/s wind speed and have undergone real scale wind Most of their larger FPV projects (more than 1 MWp) tunnel testing. use a frame system whilst smaller projects (less than 1 MWp) typically use HDPE float matrix. According 6.2.3.  LG CNS to the company, projects have witnessed increased energy yield ranging between 7 to 13 percent, and Founded in 1987 in Korea, LG CNS is the first informa- have been able to cope with humidity, rust and saline tion and communications technology company in the environments. country to make its way into the smart energy indus- try. The company applied its ICT capabilities to clean Ease of deployment with lightweight materials is a energy sources such as solar, wind, energy storage key feature of LG CNS systems. Design and systems systems, and hydrogen to create integrated energy have been tested to ensure structural safety including management solutions. fatigue test under two million cycles of dynamic load, Within their solar segment, LG CNS is a developer and wind tunnel test, and other tests of load resistance a turnkey provider of EPC plus financing. It is one of and performance. Furthermore, environmental impact the few EPC companies and FPV solution providers assessments have been conducted by various agen- to possess both technical and financial capability, cies to study the effects on water quality, sediment, having completed 6 MWp of large-scale floating proj- aquatic life, and birds. These tests involve electromag- ects (figure 6.4). LG CNS provides detailed and varied netism, temperature, humidity, light reflection, noise designs attuned to wind speed, water surface fluc- and odor tests, among others. tuation rates, water depth and conditions, and moor- CHAPTER 6:  SUPPLIERS OF FLOATING PV SYSTEMS  • 115 FIGURE 6.4. Sangju FPV systems built by LG CNS in Korea Source: © LG CNS. Source: © LG CNS. FIGURE 6.5. LSIS FPV installations in Korea Source: © LSIS. Source: © LSIS. 6.2.4.  LS Industrial Systems (LSIS) LSIS’s floating structure design is based on rigorous studies of stability under conditions of wind velocity In 2011, Korea’s LSIS Co., Ltd. built the country’s first (35 m/s), dead load, snow load, wave and flow veloc- FPV power plant at Hapcheon Dam, following research ity, and others, based on the Korean Building Code and development carried out in collaboration with and the Harbor and Fishery Design Code. For use in K-Water. LSIS has experience with various aquatic aquatic environments, LSIS has developed an exclusive environments, including dams, reservoirs, and run- eco-friendly PV module that are completely lead-free; it off ponds (figure 6.5). LSIS and K-Water are currently is the first company authorized by the Korean govern- researching FPV systems in marine environments. ment to use its PV modules as water supply equipment. 116 •  FLOATING SOLAR MARKET REPORT 6.2.5. Scotra Scotra has made substantial efforts to make its plat- forms eco-friendly. All three dams on which Scotra has Scotra is a leading supplier of FPV systems in Korea. built FPV platforms provide drinking water to the sur- It has constructed over 40 MWp of FPV systems there, rounding population. Minimal contact with the water including the country’s largest, 18.7MWp at Gunsan surface is a feature of Scotra’s eco-friendly design Retarding Basin (figure 6.6). It also exports its FPV (only 10.6 percent of total FPV area), allowing sub- solutions to Japan, the Philippines, and Taiwan, Chi- stantially more sunlight to reach the water. To minimize na, among others. Scotra is the lead institution on a effects on underwater ecosystems, including benthic Korean government research project on FPV systems organisms, the Scotra system does not block the nat- in marine environments. In that role, the company is ural flow of water. Ample free passages through the focusing on structural stability and strength to with- structure reduce O&M costs considerably. The minimal stand typhoons, and on eco-friendliness by minimizing use of floaters, combined with solar panels’ maximum the surface area of FPV installations. exposure to open air, leads to more-efficient genera- tion of electricity owing to the cooling effects of pass- The company claims to have built more than 1,200 ing water and air. The company generally provides a floating structures since its establishment in 2004. three-year warranty on the float system, although this Most have been for recreational purposes, such as is negotiable. marinas, water parks, buildings, bridges, stages, and mooring facilities. Scotra’s FPV platform business Salt water is one of the great challenges for FPV sys- began in 2011 with a partnership with K-Water, a Kore- tems, as it is for solar panel manufacturers. Scotra an public corporation in charge of managing water has built an FPV system on salt water in Korea and resources, including dams. Because the difference has been monitoring it for the past five years to ascer- in level between high and low water in some dams is tain environmental effects and the impact of tides more than 35 meters, Scotra tested a variety of moor- and salinity. Based on the confidence obtained from ing methods before arriving at an optimum system, this experience, Scotra has organized a research the 360 degree multi-point catenary mooring method consortium of 15 institutions and is leading a gov- with patented elastic devices. Scotra is now applying ernment research project on FPV systems in the sea to reservoirs the knowledge and expertise it gained in environment. its dam projects. FIGURE 6.6. Scotra’s FPV installations in Korea (left is Korea’s largest FPV system) Source: © Scotra. Source: © Scotra. CHAPTER 6:  SUPPLIERS OF FLOATING PV SYSTEMS  • 117 6.2.6.  Sumitomo Mitsui Construction (SMCC) In August 2018, SMCC announced the development of a new float supporting 72-cell solar panels (instead of Sumitomo Mitsui Construction Co., Ltd. (SMCC) is a 60-cell panels).29 At present, one solar panel is mount- large Japanese general construction company. Since ed on each float at a tilt of 10°. As with other suppliers’ 2015, SMCC has ventured into supplying float systems systems, float assembly is simple and quick. Moreover, for FPV installations. It sells floats under the brand the compact and regular shape of floats increases name “PuKaTTo” that have been deployed on various packing density and reduces transportation costs. types of water bodies such as lakes, water reservoirs, industrial water retaining ponds and flood control res- 6.2.7. Sungrow ervoirs (figure 6.7). Established in 2016, Huainan Sungrow Floating Mod- SMCC manufactures floats with a proprietary design ule Scientific and Technical Co., Ltd (Sungrow Floating) that is slightly different from other mainstream suppli- is a subsidiary of Sungrow Group, which has 21 years ers. Like for other suppliers, floats are made from stur- of solar power research and production experience, dy, UV-resistant HDPE, but its uniqueness lies in the predominantly in the area of PV inverters. The compa- fact that the floats are filled with polystyrene foam. ny has been devoting research and development on Consequently, floats will not sink even when dam- FPV systems for the past three years. More than 30 aged. SMCC further claims that these floats are three patents cover aspects of the HDPE pure-float, matrix- to five times more rigid than hollow products, thereby type floating platforms supplied by the company. minimizing the risk of plastic expansion. Another specificity of SMCC float system is the use of flexible Sungrow Floating has supplied many projects in China binding bands to connect the floats, allowing the (figure 6.8), including very large “Top Runner” program floats to move along with the waves, thereby minimiz- projects in coal-subsidence areas at the 100+ MW ing impact on the fixing parts and the modules. Also, scale, and smaller projects in lakes, agricultural ponds, to enhance cooling from water, central part of the and water-treatment reservoirs. The company has also floats contains a large aperture. installed test systems in extremely cold and typhoon-af- fected regions. In addition, Sungrow is taking the lead SMCC can also provide mooring design services to establish floating technology standards in China. building on the group’s experience in offshore wind. SMCC also has its own wind tunnel testing facility. 29. https://tech.nikkeibp.co.jp/dm/atclen/news_en/15mk/081902303 /?ST=msbe FIGURE 6.7. Examples of SMCC’s FPV systems in Japan Source: © SMCC. Source: © SMCC. 118 •  FLOATING SOLAR MARKET REPORT FIGURE 6.8. Examples of Sungrow’s projects in China Source: © Sungrow. Source: © Sungrow. Sungrow’s floating structure consists in pure HDPE Sungrow can also provide additional services such float matrix where floats can accommodate both alu- as designing the anchoring system and turnkey EPC minum frame panels and glass-glass panels in various design and construction via its parent company, Sun- layouts. Panels can be tilted at 5° or 12°. The maxi- grow Power Supply Co., Ltd. In general, Sungrow mum buoyancy of the floating matrix is 200 kg/m2. Floating warranties its products for five years, extend- Inner stress is effectively neutralized through the flex- able depending on contracts. ible ear connection of the floating matrix. The stability of the system is significantly enhanced by Sungrow’s 6.2.8.  Xiamen Mibet New Energy anchoring solutions, which are based on experience Xiamen Mibet New Energy Co., Ltd (Mibet Energy) with ocean engineering. specializes in researching, developing, manufactur- ing, and selling PV-related products, mainly mounting The HDPE material has passed more than 20 tests structures and trackers. With its independent intellec- (including photoxy-aging and environmental stress tual property, Mibet Energy offers first-class mounting crack resistance). Float products have passed more solutions around the world. Its ground-based, roof- than 10 tests (e.g., of watertightness and wind resis- based, and floating MRac PV mounting systems, as tance) and earned certification from TÜV SÜD (water well as its MRac tracker mounting system, are sold in quality detection, damp-heat aging, oxidation induc- more than 100 countries. They have received interna- tion time, impact brittle temperature, strain relief test of tional certifications such as AS/NZS1170, TÜV SÜD, opposite side angle, UV-irradiation aging, bend fatigue MCS, UL, and SGS. test, restriction of hazardous substances, environmen- tal stress crack resistance). All materials are food- G4S is the latest version of the MRac FPV mounting grade and meet environmental protection standards system. Its HDPE floats have passed the Hunt Water for drinking water. Absorption Test, Anti-Aging Test, and Anti-UV Test, among others. Mibet Energy claims a product lifetime Efficient cooling can be achieved using the flat surface of more than 25 years, increased volume of floats to of the main floating body and aluminum brackets. This improve buoyancy (which can reach 150 kg/m2), mod- combination not only ensures proper panel tilt, but ularity with various array designs that are easily com- also maintains enough space between the panels and bined to form complex islands and product durability the main floating body to facilitate ventilation and heat established by extensive tests and certifications such dissipation. as (i) aging test by TÜV SÜD, (ii) wind-load resistance CHAPTER 6:  SUPPLIERS OF FLOATING PV SYSTEMS  • 119 FIGURE 6.9. Examples of Mibet Energy’s projects in China Source: © Mibet Energy. Source: © Mibet Energy. tested by TÜV SÜD, and (iii) quality-of-water test by significantly more energy than regular FPV systems NSF (United Kingdom), ensuring environmental friend- under good insolation. By minimizing the materials liness and compatibility with drinking water. Examples needed, the design also reduces costs and logistics of the company’s projects are shown in figure 6.9. challenges. The hydroelastic membrane is attached to an outer Providers of floating 6.3.  perimeter of moored buoys so that the floater is not dragged under the mooring, even in strong currents. technology solutions for Elements of the mooring technique are derived from offshore or near-shore industrial fish farming in rough waters in Norway. Accu- applications mulated rainwater is diverted over the freeboard using This section describes in alphabetical order two sup- bilge pumps. The circular geometry is beneficial with pliers of FPV technology for offshore or near-shore respect to external forces under harsh conditions. A applications. Because these applications are still limit- rectangular shape can be used in more benign waters ed to date, suppliers in this segment have not reached or inland reservoirs. the threshold of 5 MWp of installed capacity. However, this could change in the near future. Thanks to the hydroelastic properties and dampening effect of the membrane, the system can cope with 6.3.1.  Ocean Sun relatively large waves. Watertight, it offers a protective barrier against saltwater. In 1.5 years of marine test- Ocean Sun was founded in 2016 to develop and com- ing, the system has performed well. The concept has mercialize a novel floating solar concept based on the also been modelled to scale in basin laboratory and installation of PV modules onto a large, free-floating, behaves well in waves up to three meters (Hs29<1.5 hydroelastic membrane. The method differs funda- m). If necessary, the torus rim and freeboard can be mentally from existing FPV systems since the mod- optimized further to cope with the slamming force of ules are cooled by direct contact with the hydroelastic higher waves. membrane as opposed to conventional air-cooling. The operating temperature of PV cells is held close to Hs=significant wave height, defined as the mean wave height 30.  the water temperature, enabling the system to produce (trough to crest) of the highest third of the waves. 120 •  FLOATING SOLAR MARKET REPORT The membrane is made of a strong, polymer-coated ly bankable. Early adopters are strong independent textile, dimensioned to withstand the tensile forces power producers interested in a new, low-cost float- exerted by waves, current, and wind. The buoyant er technology with high yield, able to carry risk, and double torus is constructed from HDPE piping. All the possessing the engineering resources to do in house floater materials are approved for drinking water and assessments. Certification work has been initiated with are carefully selected with respect to UV and hydro- a major third-party classification company. lysis resistance. Mathematical modelling using both analytical and the finite element method, as well as Ocean Sun operates two smaller test systems in Nor- instrumented tests in a basin laboratory, shows that the way and in Singapore. A third 100 kWp off-grid sys- PV modules are subjected to low stress and deflec- tem supplies power to a large fish farm on the western tions. For modules with adequate stiffness (such as the coast of Norway (figure 6.10). Ocean Sun is currently typical 60-cell glass-glass module), mechanical stress building 2.2 MWp on two hydroelectric power dams in is significantly lower than the stress occurring in the Southeast Asia and South Europe, respectively. wind-load test with four-point clamping described in the IEC61215 standard. Modules also have more stable 6.3.2. Swimsol thermal contact with the water body, and the thermal- Swimsol was founded in 2012 and has become the ly induced stresses acting on the metallic conductors major solar PV company in the Maldives, with an from temperature fluctuations between day and night installed capacity of 2.5 MW (rooftop PV) through end is eliminated. Because the modules are horizontal, the 2018. system performs best in the lower latitudes; the sleek design offers excellent wind resistance. Construction Swimsol´s first pilot SolarSea system was implemented has been modelled with good results using computa- in 2014 (figure 6.11). By 2018, eight platforms with a tional fluid dynamics for wind speeds up to 275 km/h. total capacity of 200 kWp had been installed at three This velocity is equivalent to typhoon category 4. different locations (two island resorts, one local island). By the second quarter of 2019 another 2 MWp rooftop Ocean Sun offers design specifications, EPC for 1 MW and 400 kWp FPV system will be installed. demonstration installations, consultancy, and follow-up. The unconventional floater design is still subject to Swimsol´s SolarSea solution has been developed and development and qualification and is not yet direct- continuously optimized over more than eight years. FIGURE 6.10. Offshore floaters in Norway, 50 meters in diameter (left) and 20 meters in diameter (right) Source: © Ocean Sun. Source: © Ocean Sun. CHAPTER 6:  SUPPLIERS OF FLOATING PV SYSTEMS  • 121 FIGURE 6.11. Swimsol’s FPV systems in the Maldives Source: © Swimsol. Source: © Swimsol. The system has proven since 2014 to withstand waves, wave tanks and under actual conditions. Its low-vol- wind, and harsh conditions at sea, and is built to last ume, truss-like floating structure, with a patented float 30 years. It is designed in such a way as to be easily distribution, creates an elevated surface area that iso- assembled on site (i.e., on a beach) and is commer- lates solar panels from the effects of waves. Several cially competitive with diesel generators. versions of SolarSea are available for different wave conditions and can produce electricity at costs as low Swimsol systems are designed and dimensioned to suit as US$0.12 per kWh. specific requirements. To this end, Swimsol provides related services such as site selection and preparation Working with electronics in a tropical marine envi- and analysis of the existing electrical grid. Typically, ronment is always a challenge. For each installation, FPV system components are preassembled at Swim- Swimsol selects the most appropriate stress-tested sol’s plant in Austria. Swimsol installs the systems on components of the highest quality, including heavy- site, including mooring and anchoring, to ensure the duty, high-performance panels developed specifically quality of the entire system. The company also applies for tropical marine environments. Systems have a life- hybrid solutions to integrate the solar power generated time of around 30 years. by its systems into the existing power grid. This is par- ticularly beneficial for users of diesel generators, who SolarSea’s effects on marine flora and fauna have are able to reduce fuel costs by not running certain been found by Swimsol to be negligible. Platforms are generators during sunshine hours. Swimsol systems installed only above sandy sea beds and coral patches include equipment that monitors the system via live are strictly avoided. A detailed environmental study of Internet feed. On request, Swimsol can also propose Swimsol’s longest-serving floating systems has shown financing for floating projects. significant positive effects, such as new coral growth, whereby the platforms have become a habitat for fish Swimsol’s SolarSea product is the result of five years and crustaceans. of modelling, computer simulations, and testing in 122 •  FLOATING SOLAR MARKET REPORT Funding gratefully acknowledged from