Dynamic wind load effects in a photovoltaic single-axis tracker mounting rail.

dc.contributor.advisorBredell, JRen_ZA
dc.contributor.advisorVenter, Gen_ZA
dc.contributor.authorKoekemoer, JHen_ZA
dc.contributor.otherStellenbosch University. Faculty of Engineering. Dept. of Mechanical and Mechatronic Engineering.en_ZA
dc.date.accessioned2024-02-26T12:16:21Zen_ZA
dc.date.accessioned2024-04-26T17:06:22Zen_ZA
dc.date.available2024-02-26T12:16:21Zen_ZA
dc.date.available2024-04-26T17:06:22Zen_ZA
dc.date.issued2024-02en_ZA
dc.descriptionThesis (MEng)--Stellenbosch University, 2024.en_ZA
dc.description.abstractENGLISH ABSTRACT: Single-axis trackers are actuated structures often used in utility-scale photovoltaic (PV) installations. These installations are sensitive to dynamic wind load effects due to their lightweight, flexible support structures and large PV module area. Significant damage to single-axis trackers have been reported in literature, despite the use of modern design methods. Design codes prescribe wind loads for a representative geometry but exclude potentially aeroelastic sensitive structures. Additionally, boundary layer wind tunnel testing and computational fluid dynamics studies are often associated with significant uncertainties. This study aimed to determine wind load effects on an existing 32 m × 25 m single-axis tracking PV array using field measurements. In this way, the uncertainties associated with wind tunnel testing and computational studies could be avoided. The project focuses on a critical component of the support structure, namely the mounting rail that attaches the PV modules to the rotating torque tube. Representative mounting rails were instrumented with strain gauges to capture dynamic wind load effects over periods of up to 109 days. The strain gauge locations were determined using results from calibrated finite element analyses of the structure. Equivalent static normal forces and moments acting on the mounting rail could be calculated using strain data and load calibration. These equivalent static loads would produce similar deformations and stresses compared to the dynamic wind loads while assuming simplified load distributions. The experimentally determined wind loads were correlated with wind speed, wind direction, and tracker tilt angle. The most critical wind directions (east and west) have a larger PV module area projected normal to the oncoming wind and subsequently showed high wind loads. Higher equivalent wind loads were also observed when the PV modules were more inclined relative to the oncoming wind. An interior located rail showed lower peak loads compared to an exterior rail, likely due to shielding from the surrounding structure. Wind load coefficients were lower compared to design codes for the range of wind conditions and tracker positions seen during the measurement period. This was expected, since design codes represent a critical combination of geometrical parameters to provide conservative estimates of wind loads. Analysis of dynamic load effects revealed contribution of torsional and bending modes of the torque tube to the normal forces and moments experienced by the mounting rails. A fatigue assessment found an insignificant fatigue damage for wind speeds below 21 m/s. Measurements suggest wind speeds above 21 m/s may be expected to cause fatigue damage. Design trends in utility-scale trackers show a decrease in mounting rail length to reduce the capital cost of new installations. The calibrated finite element model was expanded to assess the impact of a reduction in mounting rail length. Significant increases in the stress in the mounting rail and PV module were seen with a reduction in rail length, assuming all other parameters remain unchanged. Increases in stress may be non-linear, depending on the position and component.en_ZA
dc.description.abstractAFRIKAANSE OPSOMMING: Enkel-as volgerstrukture is geaktueerde strukture wat dikwels gebruik word in nutskaal fotovoltaïese (PV) installasies. Hierdie installasies word blootgestel aan dinamiese windlaseffekte as gevolg van hul ligte, buigbare ondersteuningstrukture en groot PV-module oppervlaktes. Beduidende skade aan enkel-as volgerstrukture is gerapporteer in die literatuur, ten spyte van die gebruik van modern ontwerpsmetodes. Ontwerpkodes skryf windbelastings voor vir ’n verteenwoordigende geometrie, maar sluit moontlik aëro-elasties sensitiewe strukture uit. Daarbenewens word grenslaag-windtonneltoetse en numeriese vloeidinamika studies dikwels geassosieer met aansienlike onsekerhede. Hierdie studie het ten doel gehad om windlaseffekte op ’n bestaande 32 m × 25 m enkel-as volgerstruktuur te bepaal deur middel van veldmetings. Op hierdie wyse kon die onsekerhede geassosieer met windtonneltoetse en vloeidinamika studies vermy word. Die projek fokus op ’n kritieke komponent van die ondersteuningstruktuur, naamlik die monteerbalk wat die PV-modules aan die roterende torsiebuis heg. Verteenwoordigende monteerbalke is geïnstrumenteer met rekstrokies om dinamiese windlaseffekte oor tydperke tot en met 109 dae te meet. Die rekstrokie posisies is bepaal met behulp van resultate van gekalibreerde eindige-element analises van die struktuur. Ekwivalente statiese normaalkragte en momente wat op die monteerbalk inwerk kon bereken word met behulp van vervormingsdata en laskalibrasie. Hierdie ekwivalente statiese laste veroorsaak soortgelyke deformasies en spannings as die dinamiese windlaste terwyl vereenvoudigde lasverdelings aanvaar word. Die eksperimenteel bepaalde windbelasting is gekorreleer met windspoed, windrigting, en die volgerstruktuur kantelhoek. Die mees kritieke windrigtings (oos en wes) het ’n groter PV-module oppervlakte wat loodreg op die aankomende wind geprojekteer word en het dus hoë windbelasting getoon. Hoër ekwivalente windbelasting is ook waargeneem wanneer die PV-modules teen ’n hoër kantelhoek was relatief tot die aankomende wind. ’n Binnegeleë monteerbalk het laer piek belasting getoon in vergelyking met ’n buitegeleë monteerbalk, waarskynlik as gevolg van beskerming deur die omringende struktuur. Windbelasting koëffisiënte was laer in vergelyking met ontwerpkodes vir die reeks windtoestande en volgerstruktuur posisies wat gedurende die meetperiode waargeneem is. Dit was verwag, aangesien ontwerpkodes ’n kritieke kombinasie van geometriese parameters verteenwoordig om konserwatiewe beramings van windbelastings te voorsien. Analise van dinamiese laseffekte het gewys die torsionele en buig modes van die torsiebuis dra by tot die normaalkragte en momente wat deur die monteerbalke ervaar word. ’n Vermoeidheidsanalise het getoon dat daar onbeduidende vermoeidheid-skade is vir windsnelhede onder 21 m/s. Metings suggereer dat windsnelhede bo 21 m/s verwag kan word om vermoeidheid-skade te veroorsaak. Ontwerptendense in nutskaal volgerstrukture dui op ’n afname in monteerbalk lengte om die kapitaalkoste van nuwe installasies te verminder. Die gekalibreerde eindige-element model is uitgebrei om die impak van ’n afname in monteerbalk lengte te beoordeel. Beduidende toenames in die spanning in die monteerbalk en PV-module is waargeneem met ’n afname in monteerbalk lengte, indien alle ander parameters onveranderd bly. Toenames in spanning kan nielineêr wees, afhangend van die posisie en komponent.af_ZA
dc.description.versionMastersen_ZA
dc.format.extentxxi, 128 pages : illustrations.en_ZA
dc.identifier.urihttps://scholar.sun.ac.za/handle/10019.1/130424en_ZA
dc.language.isoen_ZAen_ZA
dc.language.isoen_ZAen_ZA
dc.publisherStellenbosch : Stellenbosch Universityen_ZA
dc.rights.holderStellenbosch Universityen_ZA
dc.subject.lcshWind-pressureen_ZA
dc.subject.lcshPhotovoltaicsen_ZA
dc.subject.lcshField measurementsen_ZA
dc.subject.lcshFinite element analysisen_ZA
dc.subject.lcshUCTDen_ZA
dc.titleDynamic wind load effects in a photovoltaic single-axis tracker mounting rail.en_ZA
dc.typeThesis en_ZA
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
koekemoer_dynamic_2024.pdf
Size:
13.07 MB
Format:
Adobe Portable Document Format
Description: