One-Year Monitoring PV Power Plant Installed on Rooftop of Mineirão Fifa World Cup/Olympics Football Stadium

energies

Article One-Year Monitoring PV Power Plant Installed on Rooftop of Mineirão Fifa World Cup/Olympics Football Stadium

Luís G. Monteiro 1,*, Wilson N. Macedo 2, Pedro F. Torres 2, Márcio M. Silva 3, Guilherme Amaral 1, Alexandre S. Piterman 1, Bruno M. Lopes 4, Juliano M. Fraga 4 and Wallace C. Boaventura 1

1 Graduate Program in Electrical Engineering—Federal University of Minas Gerais (PPGEE/UFMG), Belo Horizonte-MG 31270-901, Brazil; [email protected] (G.A.); [email protected] (A.S.P.); [email protected] (W.C.B.) 2 Group of Studies and Development of Alternatives Energy—Federal University of Pará (GEDAE/UFPA), Belém-PA 6625-972, Brazil; [email protected] (W.N.M.); [email protected] (P.F.T.) 3 Federal Center of Technological Education of Minas Gerais (CEFET-MG), Belo Horizonte-MG 30421-169, Brazil; [email protected] 4 Companhia Energética de Minas Gerais (CEMIG), Belo Horizonte-MG 30190-131, Brazil; [email protected] (B.M.L.); [email protected] (J.M.F.) * Correspondence: [email protected]; Tel.: +55-31-99764-6268

Academic Editor: Senthilarasu Sundaram Received: 6 December 2016; Accepted: 4 February 2017; Published: 14 February 2017

Abstract: This paper presents results of one-year monitoring of AC side electrical parameters and the characterization of local solar radiation at the biggest rooftop PV Power Plant, with an installed capacity of 1.42 MWp, mounted at Mineirão Football Stadium in Brazil. This stadium is one of the sport facilities that hosted 2014 FIFA World Cup and Rio 2016 Summer Olympics Games in the country. Results showed how it is important to study and characterize the solar resource in the region of interest, based on historic data, to provide the understanding of solar radiation and thus project PV power plants with better performance. Furthermore, AC electrical data show the behavior of active, reactive and apparent powers and the influence of the PV system on the power factor at the local grid utility connection point. Finally, PV power plant performance data (as annual final yield, performance ratio and capacity factor) are also presented and compared with data from PVsyst software simulations. The results over the monitoring period were good considering the specificities of the stadium.

Keywords: PV power plants; monitoring; rooftop; sport stadiums; power factor; performance metrics

1. Introduction At the end of 2015, world cumulative installed solar PV power capacity reached 229.3 GWp. Only in this year, 50.6 GWp was installed and a growth of 29% was registered [1]. Thus, year after year PV systems become increasingly popular. Around the world, there are many different applications, ranging from stand-alone systems, in rural and isolated areas, to grid-connected systems, usually huge power plants and/or distributed systems, inside or close to urban areas. One of these applications (and distinguished application) is to install PV systems on sports stadiums (football, basketball, baseball, cricket, etc.), as they promote the visibility of such structures. Nowadays, since International Olympic Committee (IOC) ﬁrst included environmental protection (Green Programs—“the third pillar of Olympics games”) as one of the requirements for a successful bid

Energies 2017, 10, 225; doi:10.3390/en10020225 www.mdpi.com/journal/energies Energies 2017, 10, 225 2 of 23

to hostEnergies the Games,2017, 10, 225 PV systems and solar heat collectors have been promoted as sources of renewable2 of 23 energy to be used in stadiums, multi sports arenas and athletes Olympic villages. Insuccessful particular, bid theto host 2000 the Summer Games, PV Olympic systems Games and solar in heat Sydney collectors (ﬁrst have IOC been greening promoted program, as sources inspired by theof commitment renewable energy due toto conservationbe used in stadiums, measures mult takeni sports during arenas1994 and athletes Winter Olympic Olympics villages. in Lillehammer, Norway),In as particular, well as Winter the 2000 Olympics Summer GamesOlympic in Games Turin in (2006), Sydney Beijing (first IOC (2008), greening Vancouver program, (2010), inspired London (2012)by and the Rio commitment (2016) all included due to substantialconservation greening measures programs. taken during 1994 Winter Olympics in Lillehammer, Norway), as well as Winter Olympics Games in Turin (2006), Beijing (2008), Vancouver Green programs have ﬁve key performance areas: energy conservation, water conservation, (2010), London (2012) and Rio (2016) all included substantial greening programs. reductionGreen of waste programs generation, have five pollution key performance avoidance areas: and energy protection conservation, of natural water environment conservation, [2,3]. Today,reduction sustainability of waste is angeneration, essential pollution part of any avoidanc moderne and Olympic protection Games’ of natural project environment and a central [2,3]. concept in theToday, Olympic sustainability Agenda 2020 is an initiative.essential part Thus, of an solary modern energy Olympic in sports Games’ is becoming project and usual a central as professionalconcept sportsin seek the Olympic a competitive Agenda edge 2020 ininitiative. sustainability. Thus, solar energy in sports is becoming usual as professional Manysports areseek initiatives a competitive around edgeworld in sustainability. that use solar power in sports stadiums to save and/or export energy toMany a local are utility initiatives grid. around According world to that Solar use Energysolar power Industries in sports Association stadiums to [save4] in and/or United export States of Americaenergy (U.S.), to a local the total utility accumulative grid. According solar to Solar power Energy capacity Industries in professional Association [4] sports in United facilities States reached of 25.4 MWpAmerica in 2015(U.S.), (a the growth total accumulative of 13% by year).solar power Figure capacity1 shows in solarprofessional installations sports facilities at professional reached U.S. 25.4 MWp in 2015 (a growth of 13% by year). Figure 1 shows solar installations at professional U.S. sports facilities, ranked by the total cumulative capacity in kilowatts (kWp). sports facilities, ranked by the total cumulative capacity in kilowatts (kWp).

Figure 1. U.S. solar stadiums: (a) Lincoln Financial Field; (b) Staples Center; and (c) Metlife. Figure 1. U.S. solar stadiums: (a) Lincoln Financial Field; (b) Staples Center; and (c) Metlife. In Asia, the National Stadium in Kaohsiung, Taiwan has a 1 MWp PV system installed on the Inrooftop. Asia, Built the National for the 2009 Stadium World inGames Kaohsiung, and used Taiwan in the 2014 has aAsian 1 MWp Games, PV systemit is the world’s installed first on the rooftop.stadium Built to for rely the solely 2009 on World solar power. Games Thyagraj and used Stadium in the in 2014India’s Asian capital Games, Delhi City, it is which the world’s hosted ﬁrst stadiumthe to2010 rely Commonwealth solely on solar Games, power. installed Thyagraj India’s Stadium first in1 India’sMWp rooftop capital solar Delhi plant City, [5]. which In 2015, hosted in the 2010 CommonwealthBangalore, India, Games,the M. Chinnaswammy installed India’s stadium ﬁrst 1 MWpreceived rooftop a 400 solarkWp PV plant system [5]. Inon 2015, its rooftop. in Bangalore, In 2008, Summer Olympics Games in Beijing, China the Olympic Stadium “Bird’s Nest” was partially India, the M. Chinnaswammy stadium received a 400 kWp PV system on its rooftop. In 2008, Summer supplied by a 130 kWp PV system [6]. Olympics Games in Beijing, China the Olympic Stadium “Bird’s Nest” was partially supplied by a 130 kWp PV system [6].

Energies 2017, 10, 225 3 of 23

Energies 2017, 10, 225 3 of 23 In Australia, in 2000 Summer Olympics Games hosted in Sydney, the basketball Super Dome and OlympicIn Australia, Park received in 2000 rooftopSummer PVOlympics systems Games of 70 hosted kWp in and Sydney, 64 kWp, the basketball respectively. Super In Dome Townsville City, aand 348 Olympic kWp PV Park power received plant rooftop was installedPV systems on of the 70 kWp Townsville and 64 kWp, Basketball respectively. Stadium In Townsville in 2012 [7 ] and City,Energies a 3482017 ,kWp 10, 225 PV power plant was installed on the Townsville Basketball Stadium in 2012 [7]3 andof 23 in Carrara football stadium (220 kWp/BIPV) at Gold Coast City. Figure2 illustrates examples of PV in Carrara football stadium (220 kWp/BIPV) at Gold Coast City. Figure 2 illustrates examples of PV systemssystems installedIn Australia,installed at rooftops at in rooftops 2000 inSummer in Australia, Australia, Olympics India India Games and Malaysia. Malaysia. hosted in Sydney, the basketball Super Dome and Olympic Park received rooftop PV systems of 70 kWp and 64 kWp, respectively. In Townsville City, a 348 kWp PV power plant was installed on the Townsville Basketball Stadium in 2012 [7] and in Carrara football stadium (220 kWp/BIPV) at Gold Coast City. Figure 2 illustrates examples of PV systems installed at rooftops in Australia, India and Malaysia.

Figure 2. PV systems installed in: (a) M. Chinnaswammy; (b) Townsville; and (c) Kaohsiung. Figure 2. PV systems installed in: (a) M. Chinnaswammy; (b) Townsville; and (c) Kaohsiung. In Europe, some countries have adopted a method for space usage on rooftops of stadiums to Ingenerate Europe, electric some energy countries from solar have panels adopted denominat a methoded as for“Solar space Stadia”. usage Therefore, on rooftops during of the stadiums 2006 to generateFIFA electric (FédérationFigure energy 2. PV Internationale systems from installed solar de panels Footballin: (a) M. denominated Chinnaswammy;Association) World as (b) “Solar Townsville; Cup in Stadia”. Germany and (c) Kaohsiung. Therefore, and 2008 UEFA during the 2006 FIFA(Union (Fédération of European Football Internationale Associations) de Football Cup in Switzerland Association) this method World was Cup put in into Germany practice and[8]. 2008 In Europe, some countries have adopted a method for space usage on rooftops of stadiums to UEFAThus, (Union for ofthe Europeanfirst time in Football FIFA history, Associations) the World Cup Cup in Championships Switzerland thisspecifically method addressed was put into environmentalgenerate electric concerns energy fromby creating solar panels the so-called denominat “Greened as “SolarGoal” Stadia”.program Therefore, (created induring 2003 theand, 2006 in practice [8]. Thus, for the first time in FIFA history, the World Cup Championships specifically Germany,FIFA (Fédération 2006 was Internationale the first greening de Football program Association) for a football World competition) Cup in Germany which has and the 2008 following UEFA addressedmain(Union areas: environmental of European water, waste, Football concerns energy, Associations) and by creatingtransportation Cup the in Switzerland so-called [9]. By doing “Green this so, method FIFA Goal” did was programnot put have into the practice (created purpose [8]. in 2003 and, inofThus, Germany,creating for the a short-term 2006first time was invision, the FIFA first but history, greening rather the maki programWorldng a Cuplong-term for Championships a football and lasting competition) specifically contribution addressed which to the has the followingimprovementenvironmental main areas: of concerns environmental water, by waste, creating protection energy, the so-called andin hosting transportation “Green a mega-sporting Goal” [9 ].program By event doing (created[10]. so, FIFA in 2003 did and, not havein the purposeGermany, ofBadenova-Stadio creating 2006 awas short-term the (Mage first greeningSolar vision, Stadio) butprogram rather located for making ain football Freiburg, a long-term competition) Germany, and whichwas lasting the has first contributionthe Europeanfollowing to the main areas: water, waste, energy, and transportation [9]. By doing so, FIFA did not have the purpose improvementfootball stadium of environmental to receive a PV protection system, a in 259 hosting kWp power a mega-sporting capacity PV gene eventrators [10 ].was installed on of creating a short-term vision, but rather making a long-term and lasting contribution to the Badenova-Stadioits rooftop, in 1995 (Mage [11,12]. Solar Others Stadio) examples located are: inFritz-Walter-Stadion Freiburg, Germany, at Kaiserslautern was the first (1.35 European MWp/rooftop)improvement of and environmental Bremer Weser-St protectionadion (1.27 in hosting MWp/rooftop) a mega-sporting in Bremen, event both [10]. in Germany; Football football stadium to receive a PV system, a 259 kWp power capacity PV generators was installed StadiumBadenova-Stadio Wankdorf/Stade (Mage de Suisse Solar inStadio) Bern (1.35located MWp/rooftop) in Freiburg, andGermany, in Biel wascity, theTissot first Arena European (2.11 on its rooftop, in 1995 [11,12]. Others examples are: Fritz-Walter-Stadion at Kaiserslautern MWp/rooftop)football stadium in to Switzerland receive a PV [13]; system, and athe 259 Stadio kWp powerMarc’Antonio capacity Bentegodi PV generators stadium, was installedin Verona, on (1.35 MWp/rooftop)Italiy,its rooftop, which becamein 1995 and renewable[11,12]. Bremer Others Weser-Stadionwith the examples commissioni (1.27are: ngFritz-Walter-Stadion MWp/rooftop) of a 1 MWp solar in photovoltaicat Bremen, Kaiserslautern both installation in (1.35 Germany; FootballoverMWp/rooftop) Stadium its dome Wankdorf/Stade andin 2009 Bremer [5]. Weser-StFigure de Suisse3adion illustrates (1.27 in Bern MWp/rooftop) examples (1.35 MWp/rooftop) of PVin Bremen, systems both andinstalled in in Germany; Biel at city,rooftops Football Tissot in Arena (2.11 MWp/rooftop)Germany,Stadium Wankdorf/Stade Italy and in Switzerland. Switzerland de Suisse [13 in]; Bern and (1.35 the Stadio MWp/rooftop) Marc’Antonio and in Biel Bentegodi city, Tissot stadium, Arena (2.11 in Verona, Italiy,MWp/rooftop) which became in renewable Switzerland with [13]; the and commissioning the Stadio Marc’Antonio of a 1 MWp Bentegodi solar photovoltaicstadium, in Verona, installation Italiy, which became renewable with the commissioning of a 1 MWp solar photovoltaic installation over its dome in 2009 [5]. Figure3 illustrates examples of PV systems installed at rooftops in Germany, over its dome in 2009 [5]. Figure 3 illustrates examples of PV systems installed at rooftops in Italy andGermany, Switzerland. Italy and Switzerland.

Figure 3. Some Europe PV systems installed at rooftops in: (a) Badenova-Stadio; (b) Tissot Arena; (c) Fritz-Walter-Stadion; (d) Stadio Marc’Antonio Bentegodi; and (e) Bremer Weser-Stadion.

Figure 3. Some Europe PV systems installed at rooftops in: (a) Badenova-Stadio; (b) Tissot Arena; Figure 3. Some Europe PV systems installed at rooftops in: (a) Badenova-Stadio; (b) Tissot Arena; (c) Fritz-Walter-Stadion; (d) Stadio Marc’Antonio Bentegodi; and (e) Bremer Weser-Stadion. (c) Fritz-Walter-Stadion; (d) Stadio Marc’Antonio Bentegodi; and (e) Bremer Weser-Stadion.

Energies 2017, 10, 225 4 of 23 Energies 2017, 10, 225 4 of 23

Recently,Recently, inin SouthSouth America,America, BrazilBrazil hostedhosted fourfour importantimportant internationalinternational sportsport events:events: 2013 FIFAFIFA ConfederationsConfederations Cup,Cup, 2014 2014 FIFA FIFA World World Cup, Cup, Rio Rio 2016 2016 Summer Summer Olympics Olympics and Paralympicand Paralympic Games, Games, thus, anthus, opportunity an opportunity emerged emerged to solarize to solarize some stadiumssome stadiums across across the country. the country. Therefore,Therefore, thethe so-calledso-called programprogram “Estádios“Estádios Solares”Solares” (stadia(stadia solar)solar) [[14],14], anan initiativeinitiative takentaken byby thethe IdealIdeal InstituteInstitute andand FederalFederal UniversityUniversity ofof SantaSanta CatarinaCatarina (UFSC),(UFSC), waswas meantmeant toto installinstall PVPV powerpower plantsplants inin thethe 20142014 FIFAFIFA WorldWorld Cup Cup stadiums.stadiums. ItIt waswas thethe chancechance toto retrofitretrofit solarsolar PVPV ontoonto rooftopsrooftops ofof existingexisting stadia,stadia, andand atat thethe samesame time,time, deploydeploy aa fewfew systemssystems ontoonto newlynewly constructedconstructed ones.ones. InIn Brazilian’sBrazilian’s SoutheastSoutheast region,region, twotwo stadiumsstadiums hadhad PVPV systemssystems installed.installed. OneOne ofof them,them, JornalistaJornalista MárioMário FilhoFilho StadiumStadium (Maracanã),(Maracanã), locatedlocated inin RioRio dede Janeiro city,city, receivedreceived aa 390390 kWpkWp PVPV powerpower systemsystem atat itsits rooftop.rooftop. InIn BeloBelo HorizonteHorizonte city,city, capitalcapital ofof MinasMinas GeraisGerais State,State, aa 1.421.42 MWpMWp PVPV powerpower plantplant waswas installedinstalled atat thethe MineirãoMineirão Stadium,Stadium, alsoalso onon thethe rooftop.rooftop. InIn NortheastNortheast regionregion ofof Brazil,Brazil, PernambucoPernambuco ArenaArena inin Recife/PernambucoRecife/Pernambuco State has a 11 MWpMWp groundground mountedmounted PVPV systemsystem installed.installed. BrazilianBrazilian 2014 World CupCup diddid notnot utilizeutilize PituaçuPituaçu StadiumStadium locatedlocated atat SalvadorSalvador citycity inin BahiaBahia State,State, althoughalthough thisthis stadiumstadium waswas thethe firstfirst stadiumstadium inin LatinLatin AmericaAmerica toto receivereceive aa 403403 kWpkWp rooftoprooftop PVPV system.system. DuringDuring thisthis internationalinternational sportsport event,event, thethe PituaçuPituaçu StadiumStadium servedserved onlyonly asas anan officialofficial trainingtraining field.field. FigureFigure4 4 shows shows the the four four stadiums stadiums that that received received PV PV systems systems inin Brazil.Brazil.

Figure 4. PV systems installed in stadiums in Brazil: (a) Maracanã Stadium; (b) Pituaçu Stadium; (c) Figure 4. PV systems installed in stadiums in Brazil: (a) Maracanã Stadium; (b) Pituaçu Stadium; (Pernambucoc) Pernambuco Arena; Arena; and and (d) ( dMineirão) Mineirão Stadium. Stadium.

Despite these facilities, there are not many publications that disseminate the operating results Despite these facilities, there are not many publications that disseminate the operating results associated with Brazilian solar stadiums, and, to help change this, this paper presents a one-year associated with Brazilian solar stadiums, and, to help change this, this paper presents a one-year (October of 2014 to September of 2015) monitoring of electrical parameters (at AC side) and the (October of 2014 to September of 2015) monitoring of electrical parameters (at AC side) and the characterization of local solar radiation for a PV Power Plant installed at the rooftop of Mineirão characterization of local solar radiation for a PV Power Plant installed at the rooftop of Mineirão Football Stadium. First, Mineirão Football Stadium and its rooftop PV solar power plant are Football Stadium. First, Mineirão Football Stadium and its rooftop PV solar power plant are presented. presented. In sequence, the used methods are presented and the main results of the solar resource In sequence, the used methods are presented and the main results of the solar resource characterization characterization and monitoring of PV power plant are analyzed and discussed. Finally, final and monitoring of PV power plant are analyzed and discussed. Finally, ﬁnal comments are presented. comments are presented.

Energies 2017, 10, 225 5 of 23

2. Mineirão Football Stadium Energies 2017, 10, 225 5 of 23 Mineirão stadium, officially named Estádio Governador Magalhães Pinto, was opened in 1965. The football2. Mineirão arena Football is part Stadium of Pampulha architectural complex (UNESCO’s World Cultural Heritage), designedMineirão by Oscar stadium, Niemeyer officially (Brazilian named architectEstádio Governador who is considered Magalhães onePinto, of was the opened key figures in 1965. in the ◦ 0 ◦ 0 developmentThe football of modernarena is part architecture) of Pampulha and architectural located (geographic complex (UNESCO’s coordinates: World 19 Cultural51 57” S Heritage), 43 58 15” W) in thedesigned North regionby Oscar of Niemeyer capital Belo (Brazilian Horizonte architect (Minas who Gerais is considered State). one The of stadium the key isfigures also ain Culturalthe Heritagedevelopment of Belo Horizonte of modern city.architecture) Therefore, and Mineirão located (geographic is one of Brazil’scoordinates: most 19°51 famous′57″ S football 43°58′15 stadiums.″ W) Thein the stadium North region had capacityof capital forBelo 130,000 Horizonte spectators (Minas Gerais and used State). it The until stadium the 1990s is also when a Cultural maximum capacityHeritage decreased of Belo Horizonte to 75,000 city. spectators. Therefore, When Mineirão Brazil is one won of Brazil’s their bidmost to famous host thefootball 2014 stadiums. World Cup, MineirãoThe had stadium a large redevelopment.had capacity for 130,000 The project spectato includedrs and used a complete it until the reconstruction 1990s when maximum of bottom tier, capacity decreased to 75,000 spectators. When Brazil won their bid to host the 2014 World Cup, an extension of roof, and further refurbishments to upgrade the stadium to FIFA standards and achieve Mineirão had a large redevelopment. The project included a complete reconstruction of bottom tier, the green goals commitments. Therefore, capacity decreased to 58,259 spectators for comfort and an extension of roof, and further refurbishments to upgrade the stadium to FIFA standards and safetyachieve reasons. the green goals commitments. Therefore, capacity decreased to 58,259 spectators for comfort Afterand safety redevelopment, reasons. in 2014, Mineirão became the first stadium in Brazil, second in the World, to receiveAfter the LEED redevelopment, Platinum in Sustainability 2014, Mineirão Certification became the first (Leadership stadium in Brazil, in Energy second and in Environmentalthe World, Design,to receive the highest the LEED level Platinum of the Sustainability seal) granted Certi by thefication U.S. (Leadership Green Building in Energy Council and Environmental (USGBC) due to managementDesign, the of resourceshighest level (energy, of the water,seal) granted etc.) performed by the U.S. byGreen the building.Building Council (USGBC) due to Inmanagement 2013 Confederations of resources Cup, (energy, the stadiumwater, etc.) hosted performed three by matches the building. (including one semi-final); it hosted six matchesIn (including2013 Confederations one semi-final) Cup, the in 2014stadium FIFA hosted World three Cup; matches and during (including Rio 2016 one Summer semi-final); Olympics it hosted six matches (including one semi-final) in 2014 FIFA World Cup; and during Rio 2016 Summer Games, it hosted six group games, two quarter-finals, a women’s semi-final and the men’s third-place Olympics Games, it hosted six group games, two quarter-finals, a women’s semi-final and the men’s play-off.third-place Figure 5play-off. illustrates Figure an aerial 5 illustrates view of thean stadiumaerial view (Figure of the5a) stadium before its (Figure modification 5a) before (Figure its 5b) after itsmodification modification (Figure for 5b) the after World its modification Cup. for the World Cup.

Figure 5. Aerial view of the Mineirão Stadium: (a) before its alteration works; and (b) after its Figure 5. Aerial view of the Mineirão Stadium: (a) before its alteration works; and (b) after its alteration works. alteration works. As one can see, Figure 5b shows a PV system on the rooftop of the stadium received as part of Asits onemodernization can see, Figure efforts.5b showsThe solar a PV power system plan ont theis a rooftop partnership of the between stadium German received Technical as part of its modernizationCooperation efforts. Agency The (GTZ), solar powerthe local plant utility is aco partnershipmpany (Companhia between Energética German Technical de Minas Cooperation Gerais Agency(CEMIG)) (GTZ), and the the local dealership utility companyMinas Arena. (Companhia It was partially Energética financed de by Minas KfW Development Gerais (CEMIG)) Bank, and the dealershipfrom Germany, Minas and Arena. local state It was government partially through ﬁnanced CEMIG. by KfW This Development PV system is the Bank, biggest from rooftop Germany, and localPV installation state government in football through stadium CEMIG.in Brazil and This presents PV system challenging is the biggest conditions rooftop in order PV to installation achieve in an optimal system performance, as described in next sections. football stadium in Brazil and presents challenging conditions in order to achieve an optimal system performance,Photovoltaic as System described at Mineirão in next Football sections. Stadium

PhotovoltaicThe System photovoltaic at Mineirão plant Footballinstalled Stadium at rooftop of Mineirão Football Stadium is composed of 5910- polycrystalline silicon PV modules Mprime (Martifer) model M 240P. The 360° rooftop is divided Theinto 88 photovoltaic physical segments plant with installed 42 segments at rooftop (east and of west Mineirão sectors) Football containing Stadium 75 PV modules is composed each of ◦ 5910-polycrystalline(18 kWp/segment) silicon and PV 46 modules segments Mprime (north (Martifer)and south model sectors) M 240P.with The60 360PV modulesrooftop iseach divided into 88(14.4 physical kWp/segment). segments with 42 segments (east and west sectors) containing 75 PV modules each (18 kWp/segment) and 46 segments (north and south sectors) with 60 PV modules each (14.4 kWp/segment). Energies 2017, 10, 225 6 of 23 Energies 2017, 10, 225 6 of 23

◦ All PV modules that compose each physical segment have 88° of tilt tilt following following the roof roof inclination inclination angle. The The modules may may not appear due to restrict restrictionsions imposed by agency that gives buildings the title ofof CulturalCultural Heritage Heritage of Municipality.of Municipality. Thus, Thus some, some structures structures of the stadiumof the stadium could not could be modiﬁed, not be modified,and problems and withproblems shadowing with shadowing over the panels over duringthe panels day isduring constant day over is cons thetant year. over The the system year. has The a 2 systemtotal installed has a total power installed capacity power of 1.42 capacity MWp of and 1.42 occupies MWp and an areaoccupies of 11,530 an area m (70%of 11,530 of total m2 (70% rooﬁng of 2 totalarea—16,424 roofing marea—16,424)[15,16]. Figure m2) 6[15,16]. shows anFigure overview 6 shows scheme an ofoverview segments scheme where photovoltaicof segments powerwhere photovoltaicplant is installed. power plant is installed.

Figure 6. Schematic overview of each segment that make up the Mineirão Football Stadium’s PV Figure 6. Schematic overview of each segment that make up the Mineirão Football Stadium’s PV power plant.

Each segment in Figure 6 is composed of 6–7 strings (9–12 PV modules each) that are connected Each segment in Figure6 is composed of 6–7 strings (9–12 PV modules each) that are connected to one inverter from Ingeteam, model Ingecon Sun Smart TL, with nominal power rated at 15 kW and to one inverter from Ingeteam, model Ingecon Sun Smart TL, with nominal power rated at 15 kW and includes three maximum power point trackers (MPPT) (2–3 strings for each MPPT; 1 MPPT per includes three maximum power point trackers (MPPT) (2–3 strings for each MPPT; 1 MPPT per phase). phase). Therefore, 88 inverters compose the whole system and were installed in eight rooms (11 Therefore, 88 inverters compose the whole system and were installed in eight rooms (11 inverters per inverters per room) in the stadium. room)The in theoutput stadium. of each inverter provides 380 AC industrial voltage (phase-phase) and a total nominalThe power output of of 1.32 each MW inverter to two step-up provides electrical 380 AC substations, industrial voltagelocated at (phase-phase) north and south and sectors. a total Eachnominal substation power ofis 1.32rated MW at 750 to twokW step-upand connecting electrical tosubstations, four inverter’s located rooms. at northFinally, and the south substations sectors. step-upEach substation the voltage is rated to a at13.8 750 kV kW bus and for connecting distribution to foursystem inverter’s [16]. Figure rooms. 7 shows: Finally, (a) the an substations overview schemestep-up theof inverters voltage torooms’ a 13.8 kVand bus substations; for distribution (b) PV system module [16 string;]. Figure (c)7 shows:inverters (a) room; an overview (d) PV plant/utilityscheme of inverters grid connection rooms’ substation; and substations; (e) step-up (b) PV substation; module and string; (f) switchgear (c) inverters for room; photovoltaic (d) PV powerplant/utility plant grid(interface connection with distribution substation; (e)grid). step-up Figure substation; 8 shows andan electrical (f) switchgear line diagram for photovoltaic for one segmentpower plant that (interfacecomposes withthe PV distribution system. grid). Figure8 shows an electrical line diagram for one segment that composes the PV system.

Energies 2017, 10, 225 7 of 23 Energies 2017, 10, 225 7 of 23 Energies 2017, 10, 225 7 of 23

Figure 7. (a) Scheme of eight inverters room’s and substations sectors A and B; (b) PV module string FigureFigure 7. ( 7.a) ( Schemea) Scheme of of eight eight inverters inverters room’sroom’s andand substations sectors sectors A Aand and B; B;(b) ( bPV) PV module module string string detail; (c) inverter’s room; (d) PV plant/utility grid connection substation; (e) step-up substation detail;detail; (c) inverter’s(c) inverter’s room; room; (d) PV(d) plant/utilityPV plant/utility grid grid connection connection substation; substation; (e) step-up(e) step-up substation substation (380 V (380 V to 13.8 kV); and (f) PV plant/utility grid main switchgear. to 13.8(380 kV); V to and13.8 (kV);f) PV and plant/utility (f) PV plant/utility grid main grid switchgear. main switchgear.

Figure 8. Electrical line diagram for a photovoltaic power plant segment. Figure 8. Electrical line diagram for a photovoltaic power plant segment. Figure 8. Electrical line diagram for a photovoltaic power plant segment. 3. Solar Resource Characterization—Method, Results and Discussion 3. Solar Resource Characterization—Method, Results and Discussion 3. SolarPV Resource systems Characterization—Method, performance is strongly dependent Results on weather and Discussion conditions [17–20]. Therefore, for the PV systems performance is strongly dependent on weather conditions [17–20]. Therefore, for the PV system at rooftop of Mineirão Football Stadium, four meteorological/solarimetric stations that PVPV system systems at rooftop performance of Mineirão is strongly Football dependent Stadium, on four weather meteorological/solarimetric conditions [17–20]. Therefore, stations that for the measure global irradiance in horizontal plane (GHI) using a Kipp & Zonen pyranometers model CMP PVmeasure system atglobal rooftop irradiance of Mineirão in horizontal Football plane Stadium, (GHI) using four a meteorological/solarimetricKipp & Zonen pyranometers model stations CMP that 21; ambient temperature/humidity using a PT-100 and humidity sensors; and wind speed/direction 21; ambient temperature/humidity using a PT-100 and humidity sensors; and wind speed/direction measureusing globalultrasonic irradiance anemometers in horizontal Lufft, model plane 200A (GHI) were using installed. a Kipp Furthermore, & Zonen pyranometers four crystalline model silicon CMP using ultrasonic anemometers Lufft, model 200A were installed. Furthermore, four crystalline silicon 21; ambientreference temperature/humiditysolar cells are also installed using in the a plane PT-100 of andarrays humidity (POA). Stations sensors; and and sensors wind (denominated speed/direction reference solar cells are also installed in the plane of arrays (POA). Stations and sensors (denominated usingas DL01-DL04) ultrasonic anemometers are located on Lufft, the north, model south, 200A east were and installed. west of the Furthermore, stadium as shown four crystalline in Figure 9. silicon referenceas DL01-DL04) solar cells are are located also installedon the north, in the south, plane east of and arrays west (POA). of the Stationsstadium andas shown sensors in Figure (denominated 9. as DL01-DL04) are located on the north, south, east and west of the stadium as shown in Figure9.

Energies 2017, 10, 225 8 of 23 Energies 2017, 10, 225 8 of 23 Energies 2017, 10, 225 8 of 23

Figure 9. Location of four meteorological/solarimetric stations and four crystalline silicon references Figure 9. Location of four meteorological/solarimetric stations and four crystalline silicon references cellFigure (DL01–DL04) 9. Location at of Mine fourirão meteorological/solarimetric Football Stadium. stations and four crystalline silicon references cell (DL01–DL04) at Mineirão Football Stadium. cell (DL01–DL04) at Mineirão Football Stadium. All data from sensor are collected by a programmable logic controller (PLC) and are supervised by a SCADAAll data systemfrom sensor that scans/integratesare collected by aparameters programmable every logic 15 min. controller (PLC) and are supervised by a ToSCADA validate system irradiation that scans/integrates data from Mineirão’s parameters parameters stations every every and15 15 min. min. sensors, nine years of historic GHI hourlyTo data validatevalidate (2007–2015) irradiation were data used. from Hourly Mineirão’s data were stations measured and sensors, by a Kipp nine & years Zonen of pyranometer historic GHI modelhourly data CMP (2007–2015)21 from meteorologic were used. Hourlyal/solarimetric Hourly data data were station measured of Nati by onal a Kipp Institute & Zonen of pyranometerMeteorology (INMET).model CMPCMP This 21 21 fromstation from meteorological/solarimetric is meteorologic located at theal/solarimetric Federal University station station of Nationalof Minasof Nati InstituteGeraisonal Institute(UFMG) of Meteorology campusof Meteorology (INMET). and it is justThis(INMET). 2 station km Thisaway is locatedstation from atMineirãois located the Federal PVat the Plant. University Federal Figure University of 10 Minas shows Geraisof MinasINMET (UFMG) Gerais meteorological/solar campus (UFMG) and campus it is station just and 2 it km atis UFMGjustaway 2 fromkm campus away Mineirão while from PV FigureMineirão Plant. 11 Figure presentsPV Plant. 10 shows a FigureGHI INMET correlation 10 shows meteorological/solar from INMET a set meteorological/solar of data station (May at to UFMG November station campus ofat 2014)whileUFMG between Figure campus 11 INMET whilepresents Figure station a GHI 11 and presents correlation DL02 stadium a GHI from correlation station. a set of data from (May a set to of November data (May of to 2014) November between of 2014)INMET between station INMET and DL02 station stadium and station.DL02 stadium station.

Figure 10. INMET meteorological/solarimetric station at UFMG campus. Figure 10. INMET meteorological/solarimetricmeteorological/solarimetric station station at at UFMG UFMG campus. campus.

Energies 2017, 10, 225 9 of 23 Energies 2017, 10, 225 9 of 23

Figure 11. GHI correlation (between INMET station and DL02 station). Figure 11. GHI correlation (between INMET station and DL02 station).

Figure 11 showsFigure that 11. GHIGHI correlation data from (between INMET INMET station station have and good DL02 correlationstation). for the entire Figureirradiance 11 showsrange to thatFigure data GHI 11.from dataGHI DL02 correlation from station, INMET (between with station aINMET correlation have station good factorand correlation DL02 of 1.53.station). Since for the the entire data irradiancefrom Figure 11 shows that GHI data from INMET station have good correlation for the entire rangeINMET to data station from DL02have station,a broader with time a correlationrange (historic factor GHI of data), 1.53. Sincethis dataset the data was from chosen INMET for the station Figure 11 shows that GHI data from INMET station have good correlation for the entire irradiancesubsequent range analysis to datato analyze from DL02the variation station, thatwith occurs a correlation over time factor providing of 1.53. a betterSince understandingthe data from haveirradiance a broader range time rangeto data(historic from DL02 GHI station, data), with this a dataset correlation was factor chosen of 1.53. for the Since subsequent the data from analysis INMETon the characteristic station have profile a broader of the timesolar rangeenergy (historic resource GHI in the data), region this where dataset Mineirão was PVchosen power for plant the to analyzeINMET the station variation have that a broader occurs time over range time providing(historic GHI a better data), understanding this dataset was on chosen the characteristic for the subsequentis installed. analysisFigure 12 to showsanalyze monthly the variation average that of occurs daily horizontalover time providingglobal solar a better irradiation understanding (by year) proﬁlesubsequent of the solar analysis energy to resourceanalyze the in thevariation region that where occurs Mineirão over time PV providing power plant a better is installed.understanding Figure 12 onand the Figure characteristic 13 shows profile annual of average the solar of energy daily horizontal resource in global the region solar where irradiation. Mineirão PV power plant showsison installed. monthly the characteristic Figure average 12 profile shows of daily of monthly the horizontal solar average energy global reofsource daily solar inhorizontal the irradiation region global where (by solar Mineirão year) irradiation and PV Figure power (by year)13plant shows annualandis installed. average Figure 13 ofFigure shows daily 12 annual horizontal shows average monthly global of averagedaily solar horizontal irradiation. of daily globalhorizontal solar global irradiation. solar irradiation (by year) and Figure 13 shows annual average of daily horizontal global solar irradiation.

Figure 12. Monthly average of daily horizontal global solar irradiation.

Figure 12. Monthly average of daily horizontal global solar irradiation. FigureFigure 12. 12.Monthly Monthly average average of ofdaily daily horizontal global global solar solar irradiation. irradiation.

Figure 13. Annual average of daily horizontal global solar irradiation by year.

Figure 13. Annual average of daily horizontal global solar irradiation by year. Figure 13. Annual average of daily horizontal global solar irradiation by year. Figure 13. Annual average of daily horizontal global solar irradiation by year.

Energies 2017, 10, 225 10 of 23

Energies 2017, 10, 225 10 of 23 EnergiesAs can2017, be10, 225 seen in Figures 12 and 13, daily, monthly and annual GHI varies throughout10 of year23 As can be seen in Figures 12 and 13, daily, monthly and annual GHI varies throughout year demonstrating the stochastic behavior of the solar resource in location. Years such as 2014 and 2015 demonstrating the stochastic behavior of the solar resource in location. Years such as 2014 and 2015 achievedAs thecan highestbe seen valuesin Figures for daily12 and average, 13, daily, 5.57 monthly and 5.28 and kWh/m annual2, respectively,GHI varies throughout standing 7% year and achieved the highest values for daily average, 5.57 and 5.28 kWh/m2, respectively, standing 7% and 2%demonstrating above the average the stochastic presented behavior in Figure of the14. solar resource in location. Years such as 2014 and 2015 achieved2% above the the highest average values presented for daily in Figure average, 14. 5.57 and 5.28 kWh/m2, respectively, standing 7% and 2% above the average presented in Figure 14.

Figure 14. Average deviation of annual global irradiation. Figure 14. Average deviation of annual global irradiation. Figure 14. Average deviation of annual global irradiation. Years such as 2014 and 2015 were particularly very dry, 994 mm and 1245 mm of precipitation, respectively,YearsYears such such compared as as 2014 2014 and and to 2015total 2015 wereannualwere particularlyparticularly precipitatio ve verynsry from dry, dry, 994period 994 mm mm 2007 and and 1245to 1245 2015. mm mm These of of precipitation, precipitation, values are respectively,respectively,lower than compared the compared average to precipitationto total total annual annual precipitationsfor precipitatiothe period ns(1574 from from periodmm). period This 2007 2007fact to 2015.canto 2015.explain These These the values highvalues are values lowerare thanlowerfound the than averagefor global the average precipitation irradiation precipitation for for these the for period two the years. period (1574 Figure mm).(1574 15 Thismm). presents fact This can factannual explain can totalexplain the precipitations high the high values values for found a forfoundseries global forof irradiationnine global years. irradiation for these for two these years. two Figure years. 15Figure presents 15 presents annual annual total precipitations total precipitations for a series for a of nineseries years. of nine years.

Figure 15. Annual total precipitations for period 2007 to 2015. Figure 15. Annual total precipitations for period 2007 to 2015. From the point ofFigure view 15. of Annualthe irradiance’s total precipitations frequency for distribution, period 2007 the to 2015.period from October 2014 to SeptemberFrom the 2015point was of view considered. of the irradiance’s This kind of frequency analysis ranksdistribution, the data the depending period from on theOctober frequency 2014 tothat SeptemberFrom they the are point 2015found ofwas in view considered.a certain of the range. irradiance’s This Forkind irradi of frequency anancealysis data, distribution,ranks the the analysis data the depending periodwas based from on on the October the frequency received 2014 to Septemberthatenergy they and 2015are thefound was occurrence considered. in a certain time. Thisrange. Energy kind For of analysisirradi analysisance in ranks data,dicates thethe the dataanalysis total depending amountwas based onof theonsolar frequencythe irradiance received that theyenergyavailable are foundand in theeach in aoccurrence certainrange, while range. time. time For Energy irradianceanalysis analysis indicates data, indicates the how analysis longthe total waseach basedamountirradiance on of the rangesolar received irradianceprevailed. energy andavailableFigure the occurrence 16 inshows each frequencyrange, time. Energywhile distribution time analysis analysis for indicates ener indicatesgy (irradiation) the totalhow amountlong as aeach fu ofnction irradiance solar of irradiance irradiance. range prevailed. available in eachFigure range, 16 shows while timefrequency analysis distribution indicates for how ener longgy (irradiation) each irradiance as a rangefunction prevailed. of irradiance. Figure 16 shows frequency distribution for energy (irradiation) as a function of irradiance.

Energies 2017, 10, 225 11 of 23 Energies 2017, 10, 225 11 of 23

Energies 2017, 10, 225 11 of 23

Figure 16. Energy frequency distribution (October 2014 to September 2015). Figure 16. Energy frequency distribution (October 2014 to September 2015).

In Figure 16, the bar value corresponds to the left vertical axis and it indicates the annual amount Figure 16. Energy frequency distribution (October 2014 to September 2015). Inof Figuresolar energy 16, the received bar value per square corresponds meter for to theeach left irradiance vertical range. axis and Cumulati it indicatesve percentage the annual value amount is 2 of solarshown energyIn Figurein the received solid 16, the line, bar per right value square vertical corresponds meter axis. for Total to each the amount left irradiance vertical of annual axis range. and energy it Cumulative indicates available the percentageis annual 1985 kWh/mamount value is shown(appearsof insolar the energy solidon the line,received label right of perthis vertical square axis) and axis.meter this Totalfor value each amount isirradiance 5% ofabove annual range. the total energyCumulati average availableve percentage amount is 1985of valueannual kWh/m is 2 energy for the period of 2007 to 2015 (1885 kWh/m2). (appearsshown on in the the label solid of line, this right axis) vertical and this axis. value Total is 5%amount above of theannual total energy average available amount is 1985 of annual kWh/m energy2 2 2 (appearsIn Figure on the 16, label the irradianceof this axis) ranges and thisthat2 valuecontribute is 5% more above are the 650–700 total average W/m (148 amount kWh/m of annual), 700– for the period2 of 2007 to 20152 (1885 kWh/m ).2 2 750 W/m (184 kWh/m ) and 750–800 W/m (1452 kWh/m ) as highlighted on black bars. Energy Inenergy Figure for the16, period the irradiance of 2007 to 2015 ranges (1885 that kWh/m contribute). more are 650–700 W/m2 (148 kWh/m2), contributionIn Figure of 16, irradiances the irradiance up to rangesthese ranges that contribute accounts morefor 69% are of 650–700 the energy W/m measured2 (148 kWh/m by sensors,2), 700– 700–750 W/m2 (184 kWh/m2) and 750–800 W/m2 (145 kWh/m2) as highlighted on black bars. Energy and750 W/mthese2 three(184 kWh/mranges 2alone) and contribute750–800 W/m about2 (145 23% kWh/m (7%, 9%2) andas highlighted 7%, respectively) on black of totalbars. incident Energy contribution of irradiances up to these ranges accounts for 69% of the energy measured by sensors, irradiationcontribution on of PV irradiances plant over up the to period. these ranges accounts for 69% of the energy measured by sensors, and theseand theseAnother three three ranges analysis ranges alone is alone irradiance contribute contribute frequency about about distri 23%23%bution (7%, 9% 9%against and and 7%,occurrence 7%, respectively) respectively) time (Figure of total of total17). incident incident irradiationirradiation on PV on plantPV plant over over the the period. period. AnotherAnother analysis analysis is irradiance is irradiance frequency frequency distribution distribution against against occurrence occurrence time time (Figure (Figure 17). 17).

Figure 17. Irradiance frequency distribution versus occurrence time (October 2014 to September 2015).

Similar to energy distribution in Figure 16, the time analysis in Figure 17 presents time intervals Figure 17. Irradiance frequency distribution versus occurrence time (October 2014 to September 2015). Figurein which 17. solarIrradiance irradiance frequency was in distribution certain ranges, versus considering occurrence daily time (Octobermeasurements 2014 to from September 6:00 a.m. 2015). to 7:00 p.m.Similar to energy distribution in Figure 16, the time analysis in Figure 17 presents time intervals Similarin whichIn Figure to solar energy 17, irradiance considering distribution was the in in certainperiod Figure fromranges, 16 ,2014 the considering timeto 2015, analysis the daily ranges in measurements Figure of irradiance 17 presents from most 6:00 frequent time a.m. intervals into the region of Mineirão Stadium are 50 to 200 W/m2 (17% or 930 total hours) and from 400 to 750 W/m2 in which7:00 p.m. solar irradiance was in certain ranges, considering daily measurements from 6:00 a.m. to (66% or 1544 total hours). Range of 700–750 W/m2 must be highlighted, with 254 h, or 5.5% of total. 7:00 p.m. In Figure 17, considering the period from 2014 to 2015, the ranges of irradiance most frequent in Furthermore,the region of Mineirão the analysis Stadium shows are that, 50 tofor 200 almost W/m 281% (17% of or the 930 time total (cumulative hours) and hours),from 400 the to irradiance 750 W/m2 In Figure 17, considering the2 period from 2014 to 2015, the ranges of irradiance most frequent in is(66% lower or 1544or equal total to hours). 750 W/m Range. of 700–750 W/m2 must be highlighted, with 254 h, or 5.5% of total. the region of Mineirão Stadium are 50 to 200 W/m2 (17% or 930 total hours) and from 400 to 750 W/m2 Furthermore, the analysis shows that, for almost 81% of the time (cumulative hours), the irradiance 2 (66%is or lower 1544 or total equal hours). to 750 RangeW/m2. of 700–750 W/m must be highlighted, with 254 h, or 5.5% of total.

Energies 2017, 10, 225 12 of 23

Furthermore, the analysis shows that, for almost 81% of the time (cumulative hours), the irradiance is lower or equal to 750 W/m2. Energies 2017, 10, 225 12 of 23 It is important to mention that the ﬁrst range of irradiance (0–50 W/m2) has values out of range, but theyIt can is important be read (733to mention h) at thethat topthe first of bar.range Even of irradiance though (0–50 this W/m range2) has has values a substantial out of range, time of occurrence,but they its can energy be read contribution (733 h) at the is quite top of poor. bar. Even though this range has a substantial time of Consideringoccurrence, its both energy Figures contribution 16 and is17 quite, despite poor. the fact that irradiance prevailed at lower levels (0 to 350 W/m2Considering), energy generation both Figures is concentrated16 and 17, despite primarily the fact on that medium irradiance levels prevailed of irradiance at lower (400 levels W/m (0 2 to 800 W/mto 3502 ).W/m On2), these energy ranges, generation even is the concentrated frequency primarily distribution on medium for time levels being of average,irradiance energy (400 W/m transfer2 is highto 800 (for W/m the2 whole). On these range), ranges, causing even the these frequency levels distribu to presenttion for a time higher being energy average, contribution. energy transfer Finally, to completeis high (for the the frequency whole range), distribution causing these analysis levels for to solar present radiation a higher (fromenergy October contribution. 2014 Finally, to September to complete the frequency distribution analysis for solar radiation (from October 2014 to September 2015), in Mineirão Stadium Power Plant, the daily irradiation values were classiﬁed into energy ranges, 2015), in Mineirão Stadium Power Plant, the daily irradiation values were classified into energy as shownranges, in as Figure shown 18 in. Figure 18.

Figure 18. Daily irradiation classification of values in energy ranges. Figure 18. Daily irradiation classiﬁcation of values in energy ranges. Figure 18 shows number of days that irradiance occurred for each irradiation range. The bar Figurelengths 18correspond shows number to the amount of days of thatsolar irradianceenergy per square occurred meter for th eachat is irradiationavailable by range.those days, The bar lengthsshown correspond on the left to vertical the amount axis. The of solid solar line energy represents per square the total meter cumulati thatve is energy, available shown by thoseon the days, shownright on vertical the left axis, vertical labeled axis. total The amount solid of line energy represents (1985 kWh/m the total2). cumulative energy, shown on the right verticalFurthermore, axis, labeled Figure total 18 illustrates amount ofa tendency energy (1985of values kWh/m concentratin2). g on ranges starting from 4.4 2 2 Furthermore,kWh/m up to 5.8 Figure kWh/m 18 illustrates(35% of total a energy). tendency Since of 49% values of the concentrating total accumulated on ranges energy startingoccurred from over ranges up to 5.6 to 5.8 kWh/m2 (contributes with 6% of total energy on this range), it 4.4 kWh/m2 up to 5.8 kWh/m2 (35% of total energy). Since 49% of the total accumulated energy demonstrates the good solar resource availability in the region. occurred over ranges up to 5.6 to 5.8 kWh/m2 (contributes with 6% of total energy on this range), it demonstrates4. Monitoring the of good PV Power solar Plant—Method, resource availability Results in and the Discussion region.

4. MonitoringMonitoring of PV a Power PV system Plant—Method, is a well-known Results and andwidespread Discussion practice around world [21–32]. Monitoring all operational parameters of a PV solar plant (e.g., DC-AC voltages, currents, power, Monitoringsolar radiation, a etc.) PV systemis crucial is for a fault well-known detection and energy widespread production/yield practice aroundevaluation. world The AC [21 –32]. Monitoringside of allMineirão operational PV power parameters plant was of monitored a PV solar for plant the period (e.g., DC-AC of October voltages, 2014 to currents, September power, 2015, solar radiation,by Fluke etc.) 435 is series crucial II Power for fault and detection Energy Quality and energy Digital production/yield Analyzer. evaluation. The AC side of MineirãoSuch PV power equipment plant is was used monitored for commissioning for the period and monitoring of October power 2014 plants to September [22,29] and 2015, also by to Fluke 435 seriesmeasure II Power many electrical and Energy parameters Quality as Digital voltage Analyzer. (up to 1000 V—phase to neutral—±0.1% Vnominal), Suchcurrent equipment (5 A to 6000 is A, used CC + forAC—±0.5%), commissioning power (W, and VA, monitoring VAr, W up powerto 6000 MW—±1%), plants [22,29 harmonics] and also to (THD’s voltage/current, 1 to 50th) and others. measure many electrical parameters as voltage (up to 1000 V—phase to neutral—±0.1% Vnominal), In this study, average values of voltage, current, active (P), reactive (Q) and apparent (S) powers ± ± currentas well (5 A as to THDs 6000 A,and CC other + AC— variables0.5%), of interest power were (W, sampled VA, VAr, per W phase up to and 6000 stored MW— at intervals1%), harmonics of 5 (THD’s voltage/current, 1 to 50th) and others.

Energies 2017, 10, 225 13 of 23

In this study, average values of voltage, current, active (P), reactive (Q) and apparent (S) powers as wellEnergiesEnergies asTHDs 20172017,, 1010,, and225225 other variables of interest were sampled per phase and stored at intervals1313 ofof 2323 of 5 min. The data acquisition device (Fluke 435) was connected to the point of common coupling (PCC) min. The data acquisition device (Fluke 435) was connected to the point of common coupling (PCC) between utility grid and PV Power plant of Mineirão Football Stadium as illustrated in Figure 19. between utility grid and PV Power plant of Mineirão Football Stadium as illustrated in Figure 19. Figure 20 shows daily-accumulated energy (in MWh, MVArh and MVAh) measured for the period Figure 20 shows daily-accumulated energy (in MWh, MVArh and MVAh) measured for the period fromfromfrom October OctoberOctober 2014 20142014 to September toto SeptemberSeptember 2015. 2015.2015.

Figure 19. InterfaceInterface diagramdiagram ofof GridGrid UtilityUtility withwith PVPV Power Plant indicates the installation of Figure 19. Interface diagram of Grid Utility with PV Power Plant indicates the installation of Fluke 435. Fluke 435.

Figure 20. Cont.

Energies 2017, 10, 225 14 of 23 Energies 2017, 10, 225 14 of 23

Figure 20. Cont.

Energies 2017, 10, 225 15 of 23 Energies 2017, 10, 225 15 of 23 Energies 2017, 10, 225 15 of 23

Figure 20. (a) Daily energy generation from October 2014 to December 2014; (b) Daily energy Figure 20. (a) Daily energy generation from October 2014 to December 2014; (b) Daily energy generation generation from January 2015 to March 2015; (c) Daily energy generation from April 2015 to June from January 2015 to March 2015; (c) Daily energy generation from April 2015 to June 2015; (d) Daily Figure2015; 20.(d) Daily(a) Daily energy energy generation generation from July from 2015 October to September 2014 2015.to December 2014; (b) Daily energy energygeneration generation from fromJanuary July 2015 2015 to to March September 2015; ( 2015.c) Daily energy generation from April 2015 to June 2015;As ( illustratedd) Daily energy in Figure generation 20, black from dashed July 2015 circles to September indicate days 2015. of data lost due to maintenance (5– As8, 21 illustrated and 22 December in Figure 2014,20, black just dashed1.6% of circlesthe observed indicate period) days ofand/or data lostgeneration due to far maintenance below the (5–8, 21 andexpectedAs 22 Decemberillustrated (26 October in 2014, Figure 2014 just and20, 1.6% black1 April of thedashed 2015, observed 1% circles of the period)indicate observed and/or days period). of generationdata For lost the due one-year far to below maintenance monitoring, the expected (5– 8, 21 and 22 December 2014, just 1.6% of the observed period) and/or generation far below the (26 Octoberthe PV 2014system and presented 1 April a 2015, generation 1% of theof 4.83 observed MWh/day, period). totalizing For the 1761 one-year MWh over monitoring, the period, the PV expectedresulting (26 in October an annual 2014 average and 1 Aprilproductivity 2015, 1% (annual of the yieldobserved (YF)) period).of 1240 kWh/kWp For the one-year (with daily monitoring, mean system presented a generation of 4.83 MWh/day, totalizing 1761 MWh over the period, resulting theY PVF of system3.40 kWh/kWp), presented an a annualgeneration average of 4.83capacity MWh/day, factor (CF) totalizing about 14%.1761 FigureMWh 21over presents the period, the in an annual average productivity (annual yield (YF)) of 1240 kWh/kWp (with daily mean YF of resultingaverage in performance an annual average ratio (PR) productivity for all the monitored (annual yieldmonths. (YF )) of 1240 kWh/kWp (with daily mean 3.40 kWh/kWp), an annual average capacity factor (CF) about 14%. Figure 21 presents the average YF of 3.40 kWh/kWp), an annual average capacity factor (CF) about 14%. Figure 21 presents the performanceaverage performance ratio (PR) ratio for all (PR) the for monitored all the monitored months. months.

Figure 21. PV system performance ratio averaged for all monitored months.

Figure 21. PV system performance ratio averaged for all monitored months. Figure 21. PV system performance ratio averaged for all monitored months.

Energies 2017, 10, 225 16 of 23

Figure 21 shows that the lowest values of PR happen in months with less solar radiation, hence low PVEnergies productivity, 2017, 10, 225 while reaches higher values in sunnier months. To calculate this16 of parameter, 23 the global irradiationFigure 21 shows data that from the the lowest INMET values meteorological of PR happen in station,months with described less solar in radiation, the previous hence section, and PRlow equation PV productivity, deﬁned while in [24 reaches] were higher used. values It is possiblein sunnierto months. notice To that calculate system’s this parameter, PR varies the between 0.71 andglobal 0.62 irradiation (blue circles) data from over the the INMET year, presentingmeteorological an station, annual described average in ofthe 0.68 previous (red section, dash line) and for the 2014–2015PR equation period. defined in [24] were used. It is possible to notice that system’s PR varies between 0.71 In orderand 0.62 to (blue compare circles) the over energy the year, results presenting found an by annual monitoring average atof PV0.68 power(red dash plant, line) afor set the of 2014– simulations 2015 period. using PVsyst software was performed [33]. All 88 segments were 3D modeled in software and for In order to compare the energy results found by monitoring at PV power plant, a set of each one,simulations losses due using to PVsyst shading software were was taken performed into account[33]. All 88 for segments energy were production 3D modeled estimation. in software As an example,and Figure for each 22 one,a presents losses due data to shading input/3D were drawing taken into in account PVsyst for for energy PV generators productionat estimation. segment As 67, which is orientatedan example, to the Figure north, 22a as presents well asdata a photo.input/3D Figure drawing 22 inb PVsyst presents for PV the generators software at generated segment 67, shading diagramwhich for thisis orientated segment. to the north, as well as a photo. Figure 22b presents the software generated shading diagram for this segment.

Figure 22. (a) Data input/3D drawing in PVsyst software/photo with shading early in the morning Figure 22.over( aPV) Data generators input/3D at segment drawing 67; and in (b PVsyst) shading software/photo diagram for this segment. with shading early in the morning over PV generators at segment 67; and (b) shading diagram for this segment. As Figure 22a shows, shading over PV modules is causing by concrete arms of the segment. Shadows are representative in the morning (beginning around 8:00 a.m., with 40% shading loss, and As Figure 22a shows, shading over PV modules is causing by concrete arms of the segment.

Shadows are representative in the morning (beginning around 8:00 a.m., with 40% shading loss, and ending about 9:30 a.m., with 1% shading loss) and late afternoon (beginning approximately 3:00 p.m., with 5% shading loss, and ending about 4:00 p.m., with 40% shading loss) throughout the Energies 2017, 10, 225 17 of 23 year, as demonstrated by the shading diagram in Figure 22b. For segment 67, which has 60 PV modules (14.4 kWp), the energy simulations results, made by PVsyst for PV generators, were: 16,641 kWh/year, annual YF = 1156 kWh/kWp, CF = 13.4% and PR = 0.71. Considering all 88 segments of the PVEnergies power 2017, plant,10, 225 the overall energy simulations results, according to [33], were: 161017 MWh/year;of 23 YF = 1134 kWh/kWp, CF = 12.9% and PR = 0.72. Table1 summarizes the overall energy results for monitoringending period about 9:30 and a.m., PVsyst with simulations1% shading loss) for and Mineirão’s late afternoon PV system.(beginning approximately 3:00 p.m., with 5% shading loss, and ending about 4:00 p.m., with 40% shading loss) throughout the year, as demonstrated by the shading diagram in Figure 22b. For segment 67, which has 60 PV modules (14.4 Table 1. Summary of overall energy results for monitoring period and PVsyst simulations for Mineirão’s kWp), the energy simulations results, made by PVsyst for PV generators, were: 16,641 kWh/year, PV system. annual YF = 1156 kWh/kWp, CF = 13.4% and PR = 0.71. Considering all 88 segments of the PV power plant, the overall energy simulations results, according to [33], were: 1610 MWh/year; YF = 1134 kWh/kWp,Figures of CF Merit = 12.9% and PR = 0.72. Table 1 summarizes Mineirão PVthe Poweroverall Plantenergy Metrics results for monitoring

period and PVsyst simulationsEnergy for Mineirão’s Produced PV system. Annual YF CF PR Reference Irradiation Energy Parameters 2 Table 1. Summary of overall(MWh/Year) energy results for (kWh/kWp) monitoring period (%) and (%)PVsyst simulations(kWh/m for ·day) MonitoringMineirão’s PV system. 1761 1240 14 68 5.18 (Figure 13) PVsystFigures [33] of Merit 1610 Mineirão 1134 PV Power Plant 12.9 Metrics 72 4.32 Relative difference (%) 8.57 10 7.86 5.56 16.61 Energy Produced Annual YF CF PR Reference Irradiation Energy Parameters (MWh/Year)(kWh/kWp) (%) (%) (kWh/m2·day) Monitoring 1761 1240 14 68 5.18 (Figure 13) As illustratedPVsyst in Table [33] 1, the monitoring1610 results1134 were better12.9 compared72 to the4.32 PVsyst simulations. The results forRelative annual difference energy (%) produced, 8.57 Y F and CF with10 a relative7.86 5.56 difference for16.61 these energy metrics were about 8.6%, 10% and 8%, respectively. For performanceAs illustrated ratio in Table (PR) 1, parameter, the monitoring PVsyst results presented were better a bettercompared value to the (72%) PVsyst compared simulations. to PR from The results for annual energy produced, YF and CF with a relative difference for these energy metrics monitoring, i.e., 68%. However, if an annual reference irradiation of 5.18 kWh/m2 per day (the same were about 8.6%, 10% and 8%, respectively. utilized inFor monitoring performance PR) ratio is considered(PR) parameter, for PVsyst calculating presented PVsyst a better performance value (72%) compared ratio, the to newPR result for PVsystfrom PRmonitoring, will be i.e., 67% 68%. and However, the relative if an annual difference reference to irradiation monitoring of 5.18 PR kWh/m will2be per just day (the 1% greater. This demonstratessame utilized that in monitoring it is important PR) isto considered use the same for calculating reference PVsyst parameters performance when ratio, comparing the new reality to simulationresult to for calculate PVsyst PR energy will be metrics67% and inthe a relative PV system. difference to monitoring PR will be just 1% greater. AccordingThis demonstrates to Figure that 20 it, is the important best generation to use the same period reference occurs parameters from September when comparing to February, reality months to simulation to calculate energy metrics in a PV system. with higher incidence of solar radiation, as illustrated before in Figure 12. For the entire monitored According to Figure 20, the best generation period occurs from September to February, months period,with reactive higher power incidence (Q) of issolar always radiation, negative, as illustrated due to before plant in inductiveFigure 12. For power the entire factor monitored and is always relatedperiod, to active reactive power power (P) produced(Q) is always throughout negative, due day. to As plant an inductive example, power Figure factor 23a and shows is always this behavior for activerelated (green to active diamond, power P_Total),(P) produced reactive throughout (yellow day. square,As an example, Q_Total), Figure apparent 23a shows power this behavior (grey triangle, S_Total)for and active line (green voltage diamond, (blue P_Total), line) over reactive 1–10 (ye Januaryllow square, 2015 Q_Total), (powers apparent are expressed power (grey in triangle, kW, kVAr and S_Total) and line voltage (blue line) over 1–10 January 2015 (powers are expressed in kW, kVAr and kVA and voltage in kV, as presented in the secondary axis (right)). Figure 23b shows the irradiance kVA and voltage in kV, as presented in the secondary axis (right)). Figure 23b shows the irradiance proﬁle (orangeprofile (orange line), line), measured measured by by INMET INMET station station for the the same same period. period.

(a)

Figure 23. Cont. Energies 2017, 10, 225 18 of 23

Energies 2017, 10, 225 18 of 23 Energies 2017, 10, 225 18 of 23

(b))

Figure Figure23. (a) 23.Power (a) Power between between PV PV system system and and gridgrid overover ten ten days days in Januaryin January 2015; 2015; and ( band) irradiance (b) irradiance Figure 23. (a) Power between PV system and grid over ten days in January 2015; and (b) irradiance profile forprofile 1–10 for January 1–10 January 2015. 2015. profile for 1–10 January 2015. As illustrated in Figure 23a, active power (P) is close to apparent power (S) indicating a low As illustrated in Figure 23a, active power (P) is close to apparent power (S) indicating a low Asreactive illustrated demand in Figureat the measurement 23a, active point. power In (P)Figu isre close23a, the to increase apparent of photovoltaic power (S) indicatinggeneration a low reactivethroughout demand theat theday measurementincreases consumption point. ofIn reactive Figure power. 23a, theAs theincrease measurement of photovoltaic point is located generation reactivethroughoutafter demand transformers,the day at theincreases measurementone factor consumption can be point.raised of to In reactiveex Figureplain the 23power. behaviora, the As increase of the reac measurementtive of power photovoltaic consumption: point generation is located throughoutafter transformers,as the the transformer day one increases loadfactor increases, consumptioncan be there raised is a oftonatural reactiveexplain increased the power. behavior consumption As the of measurement reac of reactivetive power power, point consumption: which is located afteras the transformers, transformerleads to a more load one intense factorincreases, inductive can bethere behavior raised is a tonaturalof the explain transformer. increased the behavior consumption of reactive of reactive power consumption:power, which asleads the to transformer a moreFigure intense 23b load presents increases,inductive the very behavior there good is irradiance a of natural the transformer. pr increasedofile for the consumption first 10 days of ofJanuary reactive 2015, power, which which are correlated directly with the solar PV plant generation and power. On these days, irradiance leadsFigure to a more 23b intensepresents inductive the very behaviorgood irradiance of the transformer. profile for the first 10 days of January 2015, which reached values between 866 W/m2 (lowest) to 1151 W/m2 (highest) with average of 1033 W/m2 and Figure 23b presents the very good irradiance profile for the first 10 days of January 2015, which are correlatedthe PV system directly had withan average the solar energy PV production plant ge ofneration 6.55 MWh/day and power. with a daily On finalthese yield days, of 4.61 irradiance arereached correlatedkWh/kWp values directly between and a withPR =866 0.70 the W/m (with solar2 average PV(lowest) plant nominal generationto 1151 powe W/mr of and 21.14 (highest) power. MW). Figures On with these 24–26average days, present irradianceof 1033the same W/m reached 2 and 2 2 2 valuesthe PV between parameterssystem had 866 and W/man analysis average(lowest) done energy for todata 1151 production from W/m Figure (highest)23a,of 6.55 but withMWh/day with better average detail with for of a days 1033daily with W/m final different yieldand theof 4.61 PV systemkWh/kWp hadprofiles. and an average a PR = 0.70 energy (with production average nominal of 6.55 MWh/day power of 1.14 with MW). a daily Figures final yield 24–26 of present 4.61 kWh/kWp the same andparameters a PR = 0.70 and (with analysis average done nominal for data power from Figure of 1.14 23a, MW). but Figures with better 24–26 detailpresent for the days same with parameters different andprofiles. analysis done for data from Figure 23a, but with better detail for days with different profiles.

Figure 24. Power profile for a partly cloudy day.

Figure 24. Power profile for a partly cloudy day. Figure 24. Power proﬁle for a partly cloudy day.

Energies 2017, 10, 225 19 of 23 Energies 2017, 10, 225 19 of 23 Energies 2017, 10, 225 19 of 23

Figure 25. Power profile for a cloudy day. Figure 25. Power profile proﬁle for a cloudy day.

Figure 26. Power profile for a sunny day with high photovoltaic generation. Figure 26. Power profile for a sunny day with high photovoltaic generation. Figure 26. Power profile for a sunny day with high photovoltaic generation. As can be seen in Figures 24–26, the voltage (phase to phase) oscillates within the range from As can be seen in Figures 24–26, the voltage (phase to phase) oscillates within the range from 14.0 to 14.25 kV, which is above nominal voltage of 13.8 kV (1.4% and 3.2% above nominal value, 14.0 Asto 14.25 can be kV, seen which in Figures is above 24 –nominal26, the voltagevoltage (phaseof 13.8 tokV phase) (1.4% oscillatesand 3.2% withinabove thenominal range value, from respectively). This is normal since the PV power plant is connected directly to the substation bus, by 14.0respectively). to 14.25 kV, This which is normal is above since nominal the PV power voltage plant of 13.8 is connected kV (1.4% directly and 3.2% to abovethe substation nominal bus, value, by means of a dedicated feeder. As already described, for Figures 24–26, active power (P) is following respectively).means of a dedicated This is normal feeder. since As already the PV described, power plant for is Figures connected 24–26, directly active to power the substation (P) is following bus, by apparent power (S), indicating a low reactive demand at the measurement point and the increase of meansapparent of apower dedicated (S), indicating feeder. As a alreadylow reactive described, demand for at Figures the measurement 24–26, active point power and (P) the is increase following of photovoltaic generation throughout the day increases consumption of reactive power. apparentphotovoltaic power generation (S), indicating throughout a low the reactive day increases demand consumption at the measurement of reactive point power. and the increase of Figure 24 shows a power profile for a partly cloudy day. For this day, the irradiance achieved photovoltaicFigure 24 generation shows a power throughout profile the for day a partly increases cloudy consumption day. For this of reactive day, the power. irradiance achieved 865 W/m2 at solar noon and had a 5.84 kWh/m2·day of irradiation. For this meteorological condition, 865 W/mFigure2 at 24 solar shows noon a powerand had profile a 5.84 for kWh/m a partly2·day cloudy of irradiation. day. For For this this day, meteorological the irradiance condition, achieved the PV power2 plant reached an average energy2 production of 5.03 MWh/day (with 866 kW for 865the W/mPV powerat solar plant noon reached and had an a average 5.84 kWh/m energy·day production of irradiation. of 5.03 For MWh/day this meteorological (with 866 condition, kW for nominal power PV power plant) and a daily final yield of 3.60 kWh/kWp and PR = 0.62. In Figure 25, thenominal PV power power plant PV reachedpower plant) an average and a energydaily final production yield of of3.60 5.03 kWh/kWp MWh/day and (with PR 866= 0.62. kW In for Figure nominal 25, cloudy day, the solar radiation values were 797 W/m2 (at solar noon) and irradiation was 4.26 power PV power plant) and a daily final yield of 3.60 kWh/kWp2 and PR = 0.62. In Figure 25, cloudy cloudy 2day, the solar radiation values were 797 W/m (at solar noon) and irradiation was 4.26 kWh/mday, the·day. solar On radiation this day, values PV system were 797 generated W/m2 (at 3.55 solar MWh/day noon) and (YF irradiation = 2.54 kWh/kWp/PR was 4.26 kWh/m = 0.60)2 with·day. kWh/m2·day. On this day, PV system generated 3.55 MWh/day (YF = 2.54 kWh/kWp/PR = 0.60) with a nominal PV power of 1.02 MW at solar noon. Ona nominal this day, PV PV power system of generated 1.02 MW 3.55at solar MWh/day noon. (YF = 2.54 kWh/kWp/PR = 0.60) with a nominal PV Figure 26 shows a day profile with the highest generation for the PV system among all monitored powerFigure of 1.02 26 MW shows at solara day noon.profile with the highest generation for the PV system among all monitored days, with nominal power of 1.15 MW at solar noon (1162 W/m2, maximum recorded irradiance), days,Figure with nominal26 shows power a day profileof 1.15 with MW the at highestsolar noon generation (1162 W/m for the2, maximum PV system recorded among all irradiance), monitored generation of 8.66 MWh over the day (8.63 kWh/m2.daily recorded2 irradiation), final daily yield of days,generation with nominalof 8.66 MWh power over of 1.15the day MW (8.63 at solar kWh/m noon2.daily (1162 recorded W/m , maximum irradiation), recorded final daily irradiance), yield of 6.20 kWh/kWp and PR = 0.72. Figures 27 and 28 show2 active versus reactive power and power factors generation6.20 kWh/kWp of 8.66 and MWh PR = 0.72. over Figures the day 27 (8.63 and kWh/m 28 show.daily active recorded versus reactive irradiation), power final and power daily yield factors of as a function of active power for different generation profiles. 6.20as a kWh/kWpfunction of andactive PR power = 0.72. fo Figuresr different 27 and generation 28 show profiles. active versus reactive power and power factors as a function of active power for different generation profiles.

Energies 2017, 10, 225 20 of 23 Energies 2017, 10, 225 20 of 23 Energies 2017, 10, 225 20 of 23

Figure 27. Active/reactive powers and power factor as a function of active power for first 10 days of Figure 27.27. Active/reactive powers powers and and power power factor factor as as a function of active power for ﬁrstfirst 10 days of January 2015. JanuaryJanuary 2015.2015.

Figure 28. Active and reactive powers and power factor as a function of active power for a good day Figure 28. Active and reactive powers and power factor as a function of active power for a good day (19Figure November 28. Active 2014, and sunny reactive day) powers and bad and day power (24 January factoras 2015, a function cloudy ofday). active power for a good day (19 November 2014, sunny day) and bad day (24 January 2015, cloudy day). (19 November 2014, sunny day) and bad day (24 January 2015, cloudy day). Finally, Figures 27 and 28 demonstrate that the PV power plant power factors are close to unity Finally, Figures 27 and 28 demonstrate that the PV power plant power factors are close to unity when it operates at nominal power. It is possible to notice a tendency to lower power factor values whenFinally, it operates Figures at nominal27 and 28 power. demonstrate It is possible that the to PVnotice power a tendency plant power to lower factors power are closefactor to values unity when submitted to partial loads. The decrease of power factor is important for the description of whenwhen itsubmitted operates atto nominalpartial loads. power. The It isdecrease possible of to power notice factor a tendency is important to lower for power the description factor values of power quality [28]. PV power plants operating with low power factor may pose problems to voltage whenpower submitted quality [28]. to PV partial power loads. plants The operating decrease with of power low power factor factor is important may pose for problems the description to voltage of regulation, the reason for the desired unity power factor. According to IEEE [34], PF (power factor) powerregulation, quality the [ 28reason]. PV for power the desired plants operating unity power with factor. low power According factor to may IEEE pose [34], problems PF (power to voltage factor) of an inverter must always operate above 0.85 when its output exceeds 10% of nominal power. regulation,of an inverter the reasonmust always for the desiredoperate unityabove power 0.85 when factor. its According output exceeds to IEEE [10%34], PFof (powernominal factor) power. of However, with the consent of utility grid, power factor can operate with values below 0.85 for anHowever, inverter with must the always consent operate of aboveutility 0.85grid, when power its outputfactor exceedscan operate 10% ofwith nominal values power. below However, 0.85 for compensation or correction. Nowadays, inverters have control mechanisms that can adjust reactive withcompensation the consent or ofcorrection. utility grid, Nowadays, power factor inverters can operatehave control with valuesmechanisms below that 0.85 can for adjust compensation reactive power flow in accordance with the needs of the utility grid. orpower correction. flow in Nowadays,accordance inverterswith the needs have controlof the utility mechanisms grid. that can adjust reactive power ﬂow in accordance with the needs of the utility grid. 5. Conclusions 5. Conclusions 5. Conclusions In this paper, it was demonstrated one of applications (and distinguished application), which is In this paper, it was demonstrated one of applications (and distinguished application), which is increasingIn this nowadays, paper, it was that demonstrated is, the use of one PV of gr applicationsid connected (and systems distinguished on sports application), stadiums. whichMany increasing nowadays, that is, the use of PV grid connected systems on sports stadiums. Many countriesis increasing have nowadays, already installed that is, solar the use PV of power PV grid plant connected on rooftop, systems facades on and sports ground stadiums. mounted Many at countries have already installed solar PV power plant on rooftop, facades and ground mounted at sportcountries stadiums have already(arena, installedfacilities) solar to PVsave power and/or plant export on rooftop, energy facadesto a local and groundutility grid. mounted Great at sport stadiums (arena, facilities) to save and/or export energy to a local utility grid. Great internationalsport stadiums sport (arena, events facilities) such toas save the and/orOlympics export Games energy and to Football a local utility World grid. Cup Great are important international to international sport events such as the Olympics Games and Football World Cup are important to disseminate and encourage the use of solar power technologies as PV systems and solar heat disseminate and encourage the use of solar power technologies as PV systems and solar heat collectors, thus they can accomplish their green goal agenda and provide a sustainable legacy. collectors, thus they can accomplish their green goal agenda and provide a sustainable legacy.

Energies 2017, 10, 225 21 of 23 sport events such as the Olympics Games and Football World Cup are important to disseminate and encourage the use of solar power technologies as PV systems and solar heat collectors, thus they can accomplish their green goal agenda and provide a sustainable legacy. In this direction, for the 2014 FIFA World Cup, Mineirão Stadium, located in Belo Horizonte City (Brazil), through the program “Estádios Solares”, installed a 1.42 MWp PV power plant on its rooftop in 2013. This paper presented results and analysis of one-year (October 2014 to September 2015) monitoring of electrical parameters (at AC side) and characterization of local solar radiation. It was shown that the solar resource in the stadium region varies stochastically over the year but has a good annual average of daily horizontal global solar irradiation (5.18 kWh/m2·day). Besides that, 2014–2015 achieved high values of GHI, above average, those years being drier than usual. For the period considered in this analysis, the total amount of annual energy available was found to be 1985 kWh/m2 (5% above average) and the solar irradiance ranges with higher energy contribution (23% of total energy) in the region are from 650 W/m2 to 800 W/m2. When considering the time length, ranges from 400 to 750 W/m2 have relevant contribution (1544 h, 66% of total time). Furthermore, irradiation is concentrated in the range from 4.4 kWh/m2 to 5.8 kWh/m2 (35% of total energy) demonstrating a good availability of solar resource in the region and the importance of analyzing such availability over time, thus understanding the characteristic profile of the solar energy resource for a better gain of energy yield at the PV power plant. Good results were found on monitoring the electrical data (AC side) of PV system, considering the significant influence of partial shading at certain periods and the days that the system was completely disconnected from the grid. Energy Generation results presented an average of 4.83 MWh/day, totalizing 1761 MWh in the period, with an annual YF of 1260 kWh/kWp, CF = 14% and annual PR = 68% (which varies from 62% to 71% over the year). These results were compared with PVsyst annual simulations (1610 MWh/year; YF = 1134 kWh/kWp, CF = 12.9% and PR = 0.72) and the monitoring results for energy produced, YF and CF were better than PVsyst simulations. However, PVsyst simulations presented a better value for PR (for average irradiation of 4.32 kWh /m2·day). When a mean irradiation of 5.18 kWh/m2·day as a parameter in simulations (the same used in monitoring period) was considered, the new result for PVsyst PR was 67% and the relative difference to monitoring PR was just 1% greater. This demonstrates that it is important to use the same reference parameters when comparing reality to simulation to calculate energy metrics in a PV system. Figure 26 shows the highest PV system generation, of all monitored days, that reached the mark of 1.15 MW of nominal power at solar noon (1162 W/m2) and generated 8.53 MWh for a particular day 2 with high solar irradiation 8.63 kWh/m . Therefore, the PV system reached a daily YF = 6.10 kWh/kWp and a PR about 71%. Reactive power (Q) is always negative, characterized as inductive and is always related to active power (P) produced throughout the day. As the photovoltaic generation increases, consumption of reactive power increases. However, at night, system displays an approximate self-consumption of 10 kW associated with losses in transformers and supervision and control system. On analysis, it was possible to observe that voltage between phases oscillates and it is above the nominal voltage of 13.8 kV. This fact is usual since the PV power plant is connected directly to the substation bus, by means of dedicated feeder. The power factor control of inverters installed in plant can be used to absorb reactive power in order to avoid an increasing level of voltage at its output, which does not happen in this particular system where the power factor tends to one when power generation grows. In addition, power factor of the inverters, which works close to unity when operating at nominal power, have a tendency to lower value when submitted to a partial loads, which is important for the description of power quality. Finally, this paper presented the beginning of a research work with some general results found about PV power plant installed at the rooftop of Mineirão football stadium. During monitoring period, many challenges arose (shading in segments, degradation of PV modules over years, power Energies 2017, 10, 225 22 of 23 quality analysis and much more). Therefore, this kind of sport facilities is a great opportunity for the development of research with real problems that will contribute to the PV area.

Acknowledgments: “Call No. 13” of R&D technology program of electricity sector regulated by Electrical Energy National Agency (ANEEL), Companhia Energética de Minas Gerais (CEMIG), National Institute of Meteorology (INMET) and Group of Studies and Development of Alternatives Energy (GEDAE) from Federal University of Pará (UFPA). Author Contributions: Luís G. Monteiro, Márcio M. Silva and Wallace C. Boaventura conceived and designed the experiments; Luís G. Monteiro, Márcio M. Silva and Juliano M. Fraga performed the experiments; Luís G. Monteiro, Wilson N. Macedo, Pedro F. Torres, Alexandre S. Piterman and Guilherme Amaral analyzed the data; Bruno M. Lopes contributed materials/analysis tools; and Luís G. Monteiro wrote the paper. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Solar Power Europe. Global Market Outlook for Solar Power/2016–2020; Solar Power Europe: Bruxelles, Belgium, 2016; p. 32. 2. The Environmental Games. Environmental achievements of Sydney 2000 Olympic Games; Olympic Games 2000; National Library of Australia: Canberra, Australia, 2000; p. 26. 3. FIFA. Green Goal 2010—Legacy Report; FIFA: Cape Town, South Africa, 2011; p. 137. 4. Solar Energy Industries Association (SEIA). Solar for the Win: A Study on Solar in U.S. Professional Sports. 2016. Available online: http://www.seia.org/research-resources/solar-win-study-solar-us-professional- sports (accessed on 10 July 2016). 5. Natural Resources Defense Council (NRDC). Demand Latest Green Building Leaders in Professional Sports. In Game Changer: How the Sports Industry Is Saving the Environment; NRDC: New York, NY, USA, 2012; pp. 107–111. 6. United Nation Environment Programm (UNEP). Independent Environmental Assessment—2008 Beijing Olympics Games; UNEP: Nairobi, Kenya, 2008; p. 139. 7. International Energy Agency (IEA). Implementing Agreement on Photovoltaic Power Systems; PVPS Annual Report (348 kWp Townsville Sports Stadium); IEA: Paris, France, 2012; p. 118. 8. Ruther, R. Feasibility Study—Project: Solar World Cup 2014—Minas Gerais, Utilization of the Mineirão-Mineirinho Complex Rooftops for the Generation of Electric Energy with Photovoltaic Panels; Final Report; Universidade Federal de Santa Catarina (UFSC): Florianópolis, Brazil, 2010; p. 153. 9. FIFA. Green Goal 2006—Legacy Report; FIFA: Frankfurt, Germany, 2006; p. 122. 10. Dolles, H.; Södermana, S. Addressing ecology and sustainability in mega-sporting events: The 2006 football World Cup in Germany. J. Manag. Organ. 2010, 16, 587–600. [CrossRef] 11. World Heritage Encyclopedia. Badenova-Stadio—List of Association Football Stadiums by Country, 2005–06 DFB-Pokal, 2007–08 2. Bundesliga, 2009–10 Bundesliga; National Public Library: Freiburg, Germany, 2011. 12. Lisboa, A. Estádios Solares—Minas Solar 2014, Inova FV (Presentation); Unicamp: Campinas, Brazil, 2011. 13. Swissinfo. Swiss Champion Solar Power but Struggle with Targets. 2015. Available online: http://www. swissinfo.ch/eng/energy_swiss-champion-solar-power-but-struggle-with-targets/41603882 (accessed on 20 June 2016). 14. Ruther, R.; Montenegro, A.; Zomer, C.; Santos, I.; Nascimento, L.; Grabolle, P. Estádios Solares. Opção Sustentável para a Copa de 2014 no Brasil; Relatório Final; Instituto Ideal/Universidade Federal de Santa Catarina (UFSC): Maio, Brazil, 2010; p. 52. 15. Companhia Energética de Minas Gerais (CEMIG). Avaliação do Desempenho da Usina Fotovoltaica do Mineirão—Maio a Novembro (Presentation); CEMIG: Belo Horizonte, Brazil, 2014. 16. Martifier Solar. Central Fotovoltaica do Estádio do Mineirão (1.42 MWp)—Projeto de Execução, Infraestruturas elétricas; Cálculos Justificativos; Martifier Solar: Oliveira de Frades, Portugal, 2013; p. 49. 17. Ramli, M.A.M.; Prasetyono, E.; Wicaksana, R.W.; Windarko, N.A.; Sedraoui, K.; Al-Turki, Y.A. On the investigation of photovoltaic output power reduction due to dust accumulation and weather conditions. Renew. Energy 2016, 99, 836–844. [CrossRef] 18. Ghazi, S.; Ip, K. The effect of weather conditions on the efficiency of PV panels in the Southeast of UK. Renew. Energy 2014, 69, 50–59. [CrossRef] Energies 2017, 10, 225 23 of 23

19. Congedo, P.M.; Malvoni, M.; Mele, M.; De Giorgi, M.G. Performance measurements of monocrystalline silicon PV modules in South-eastern Italy. Energy Convers. Manag. 2013, 68, 1–10. [CrossRef] 20. Correa-Betanzo, C.; Calleja, H.; Lizarraga, A. Photovoltaic system assessment considering temperature and overcast conditions: Light load efficiency enhancement technique. Sol. Energy 2016, 137, 148–157. [CrossRef] 21. Müller, B.; Hardt, L.; Armbruster, A.; Kiefer, K.; Reise, C. Yield Predictions for Photovoltaic Power Plants: Empirical Validation, Recent Advances and Remaining Uncertainties. In Proceedings of the 29th European PV Solar Energy Conference and Exhibition, Amsterdam, The Netherlands, 22–26 September 2014; pp. 1–9. 22. Woyte, A.; Richter, M.; Moser, D.; Mau, S.; Reich, N.; Jahn, U. Analytical Monitoring of Grid-Connected Photovoltaic Systems—Good Practices for Monitoring and Performance Analysis; IEA PVPS Task 13, Subtask 2. Report IEA-PVPS T13-03; International Energy Agency (IEA): Paris, France, 2014; p. 90. 23. Moreno-Garcia, I.; Palacios-Garcia, E.; Pallares-Lopez, V.; Santiago, I.; Gonzalez-Redondo, M.; Varo-Martinez, M.; Real-Calvo, J. Real-time monitoring system for a utility-scale photovoltaic power plant. Sensors 2016, 16, 770. [CrossRef][PubMed] 24. Photovoltaics System Performance Monitoring-Guidelines for Measurement Data Exchange and Analysis; IEC Standard 61724; IEC: Geneva, Switzerland, 1998. 25. Carigiet, F.; Baumgartner, F.; Sutterlueti, J.; Allet, N.; Pezzotti, M.; Haller, J. Verification of Measured PV Energy Yield versus Forecast and Loss Analysis. In Proceedings of the 28th European PV Solar Energy Conference and Exhibition, Paris, France, 30 September–4 October 2013; pp. 1–6. 26. Spertino, F.; Corona, F. Monitoring and checking of performance in photovoltaic plants: A tool for design, installation and maintenance of grid-connected systems. Renew. Energy 2013, 60, 722–732. [CrossRef] 27. Carrillo, J.M.; Martínez-Moreno, F. Use of PV plants monitoring to characterize PV arrays power. In Proceedings of the 31th European PV Solar Energy Conference and Exhibition, Hamburg, Germany, 14–18 September 2015; pp. 1–5. 28. Macêdo, W.N.; Zilles, R. Operational results of grid-connected photovoltaic system with different inverter’s sizing factors (ISF). Prog. Photovolt. 2007, 15, 337–352. [CrossRef] 29. Martínez-Moreno, F.; Lorenzo, E.; Moretón, R.; Narvarte, L. Bankable Procedures for the Technical Quality Assurance of Large Scale PV Plants. In Proceedings of the 29th European Photovoltaic Solar Energy Conference, Amsterdam, The Netherlands, 22–26 September 2014; pp. 2864–2869. 30. Bizzarri, F.; Brambilla, A.; Caretta, L. Monitoring performance and efficiency of photovoltaic parks. Renew. Energy 2015, 78, 314–321. [CrossRef] 31. Firman, A.; Toranzos, V.; Vera, L.; Casa, J. Passive Monitoring of the Power Generated in Grid Connected PV Systems. Energy Procedia 2014, 57, 235–244. [CrossRef] 32. Drews, A.; Keizer, A.; Beyer, G.; Lorenzo, E.; Bercke, J.; Van Sark, W.; Heydenreich, W.; Wiemken, E.; Steller, S.; Toggweiler, P. Monitoring and remote failure detection of grid-connected PV systems based on satellite observations. Sol. Energy 2007, 81, 548–564. [CrossRef] 33. Martifier Solar. Central Fotovoltaica do Estádio do Mineirão (1.42 MWp)—Estudo Energético; Martifier Solar: Oliveira de Frades, Portugal, 2013; p. 406. 34. IEEE Recommended Practice for Utility Interface of Photovoltaic Systems; IEEE Standard 929; IEC: Washington, DC, USA, 2000.

© 2017 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).