Mooring System Design for Floating Wind Turbines - comparison of different methods and concepts
MetadataShow full item record
- Institutt for marin teknikk 
Full text not available
Det overordnede målet med denne masteroppgaven er å designe og analysere forankringsystemer for flytende vindturbiner, sammenlikning av forskjellige forankringsystem, mooring line materialer og analyse metoder. System som skal bli analysert:Fire alternative forankringsystemer for en 10 MW flytende vindturbin ved Donghae utenfor Sør Korea:• System I - Tradisjonelt kjetting system• System II - Stramt polyester system(taut) - konstant aksiell stivhet EA• System III - Stramt polyester system(taut) - fiber tau modell stivhet• System IV - Polyester system med bøyer Oppset av analyser:Oppset av koblet Simo-Riflex modellDefinisjon av load cases: -3 load caser med 50 års retur periode bølger og vind in-line lasting av forankringsystemet - 2 operasjonelle load caser med rated vind og en med shut down vind - 1 50 års retur periode forankringsystem belastet in-betweenBeregning av vind koeffisienter: - 50 års vind - rated vind, rotor trust inkludert som koeffisient - shut down vind, rotor trust antatt konstantSeed convergens test for oppsett av time domain bevegelses analyse - 16 tilfeldige vind og bølge seed Sjekk av system I:ULS sjekk basert på 16 analyser og MPM tensionSjekk av bevegelser og linestrekk i lo og le lineSjekk av kildene til bevegelse, vind, bølger, strøm, resonans Sammenlikning av design metoder:Simo-Riflex inkluderer dynamikk, Simo alene gjør ikke. Sammenlikning av systemer:Samme laster for hvert system. Jag offset er forskjellig for systemene på grunn av forskjellig stivhet i jag. Hiv og stamp er likt last tilfelle for last tilfelle. Line strekk: - Lo line strekk størst for system I etterfulgt av III, II, IV - Mye dynamikk for system I - Polyester systemene har mye mindre dynamikk, følger kvasi statisk oppførsel - System I har størst le line strekk - Ingen slakk i le liner, noe slakk i lo line for load case 1 og 3, men få tilfeller Jag bevegelse: - System I har størst avdrift (jag), etterfulgt av IV, III og II Hiv og stamp bevegelse: - Bølgefrekvent hiv - Generelt lik hiv og stamp bevegelse load case for load case alle system - Værst stamp krengevinkel for load case 4 ULS sammenlikning av alle systemer: - System I har kapasitet for ULS og derfor godkjent i henhold til regelverk DNV-GL-ST0119 - Ikke mulig å gå ned i kjetting dimensjon eller kvalitet - Polyester systemene har høyere kapasitet for ULS enn syestem I, dog ikke mye - Høy kapasitet for ULS i operasjonelle load caserThe overall objective of this thesis has been to design and analyse mooring systems for wind turbinesin intermediately deep waters, comparing different systems, mooring line materials as well as analysismethods. System to be analysed:The study covers four alternative mooring line systems for a CSC 10MW FWT (Floating WindTurbine) for an offshore wind park in Donghae area in South Korea:• System I - Catenary chain - Base case system• System II - Taut polyester system - Constant axial stiffness EA• System III - Taut polyester system - Sima Fibre rope stiffnes model• System IV - Polyester system with bouys Analysis findings:• Coupled Simo-Riflex model of FWT with base case mooring system with 147mm R3 qualitychain has been made and set up for 3 hour time domain analysis• Decay tests in surge, heave, pitch and yaw for both Simo only and coupled Simo-Riflex modelhas been performed resulting in natural periods, frequencies, and damping versus criticaldamping• Slowly ramped pull-out analysis performed resulting in system restoring characteristics withand without current, and line characteristic for windward line and one of the leeward lines• 3 Load cases has been defined for 50 year return period waves and wind in direction causingin-line loading of mooring system, 1 50 year return period load case defined causing in-betweenloading of mooring system, 1 load case defined for rated wind velocity and 1 load case definedfor before shut down wind velocity both for operational turbine – Wave and wind data come from Metocean report ME2019-096 Rev05 supplied by Equinor – Wave data for operational load cases defined by wind probability – Current 2-3 % of wind velocity – Water depth is 150 [m]• Quadratic wind coefficients calculated for: – 50 year return period wind – Rated wind – Shut down wind• For load case 1 a seed convergence test was made with regards of most probable maximumtension (MPM) in windward line resulting in the use of 16 different random wind and waveseeds for use in time domain analysis of the systems• 16 3 hour time domain analyses were performed for each load case for the base case system • These were post processed in both Sima post processor tasks and Microsoft Excel resulting in: – Mean and most probable max values for degree of freedom surge, heave, and pitch – Mean and most probable max tension in windward line – Mean and most probable minimum tension for leeward line, accounting for slack• Mean surge offset was checked by hand calculating mean loads from wind and current andchecking that against the system restoring characteristics• Heave motion result was spot checked for a wave with corresponding heave motion from RAOin kinetics folder in Sima• Mean pitch was checked using hand calculated mean pitch moment from wind and currenttogether with water plane stiffness C55 in order to find hand calculated mean pitch angle• Surge, heave and pitch checks corresponded well with analysis results from Sima. Time series ofsurge, heave and pitch motion as well as windward and leeward line tension for a characteristicsea state with random wind and wave seed equal one was presented• Frequency analysis was performed using an auto spectrum on time series of motion andwindward line tension subtracting mean – The sources of loads were checked from the frequency spectrum and natural frequencyfound from decay tests• Base case system was checked regarding ULS capacity using consequence class 1 and requirements from DNVGL-ST-0119• Dynamic effects for the different load cases was checked by plotting windward line tensionversus offset for 3 hour time domain analysis results on top of windward line characteristic – Dynamic effects were seen for the 50 year load cases while the operational turbine loadcases had a quasi-static behaviour – Furthermore the plots showed that the dynamic part of design tension, TC,Dyn correosponds well with mean and max tension around the line characteristic curve Comparison of design methods:An analysis was made using Simo only for the base case system and compared with coupled Simo-Riflex analysis result. This showed that the dynamic part of line tension is much higher for coupledSimo-Riflex analysis and the Simo result has a quasi-static behavior. System I – Catenary chain - Base case system:The following observations were made about the base case system:• ULS capacity is achieved according to DNVGL-ST-0119 requirement for all load cases for R3quality chain and consequence class 1– Because the worst seastate capacity margin is just above one, the chain quality ordimension cannot be reduced.• For load case 1-3 the system has a dynamic behaviour while for the operational load cases thesystem follows a quasi static behaviour • Some slack is observed from the time series of windward line load case 1 and 3, but it is notmany instances. Most probable minimum line tension for leeward lines show likelihood forslack in the leeward lines for all load cases – Furthermore a check with increased pre-tension for load case 1 showed that increasingpretension to 1800 [kN] increases dynamic effects and doesn’t reduce instances of slack.The few instances are assumed to be acceptable due to statistically 50 years periodbetween each time the instances occur• Surge motion is worst for the in-between load case, most probable max 38 [m] followed byload case 1, 3, 2, 4 and 5. Whether this large offset is ok or not would need to be checked inaccordance with the requirements from the umbilical, power cable.• Offset from hand calculated mean loads were checked against analysis results, the differenceswere small• Pitch has a mean angle of about 2 [deg] for the 50 year return period load cases and is worstfor load case 4 due to trust force from turbine, most probable max 13 [deg] with somewhatincreased mean – For shut down wind the pitch angle is slightly higher than for 50 year return period loadcases, but no where near as high as for load case 4 Comparison of mooring systems :• The loads for each load case are generally the same for all systems, but the mooring systemsare different. This is why surge motion is very different for each system while heave and pitchmotion are very similar load case by load case. This also applies to the frequency spectra foreach load case as compared with the other systems• Mooring line tension: – Windward line tension is largest for system I, followed by III, II and IV. While meantension between the systems are very similar as compared load case by load case – The dynamic portion of windward line tension is large for system I, base case syste – The polyester systems have a much more quasi static relation of tension versus surge offset. Furthermore the polyester systems incur much less loads in the mooring lines – Generally system I has the largest leeward line tension as well – The lowest leeward line tension occurs for load case 6 system II and III – Otherwise all most probable minimum tensions are nonzero. This is confirmed by the time series of leeward line tension, there is no slack in any of the systems• Surge motion: – Surge motions are different for each system – System I has largest surge offset, followed by IV, III and II. Given the stiffness of thesystems this trend is likely correct • Heave and pitch motion: – Heave and pitch motions are very similar load case by load case – The worst pitch angle occurs for load case 4, between 13 and 15 [deg] pitch angle. – Furthermore pitch motion is largely pitch resonant for all load cases. The system with most resonance is system IV followed by II, III, and I• System II and III are very stiff with a natural period of 54 [s] – Because of less steep line characteristics for system III than system II, system III is softer and this is seen from surge offset and line tension comparison• This also applies to the frequency spectra for each load case as compared with the othersystems• Capacity margins regarding ULS were calculated for all the systems and load cases. They areall greater than 1 meaning all the systems have capacity for ULS. In that sense all systemsare approved according to DNVGL-ST-0119. The 50 year return period load cases are mostcritical regarding ULS and the operational load cases have larger capacity, but since the worstload case is just above 1 for system I-III the mooring line dimensions cannot be reduced.System IV has slightly higher capacity in load case 1, the worst load case. Therefore there is aslight possibility to reduce the dimensions of polyester rope, but that would require another 16random seed 3 hour analyses to confirm that the new dimension complies with the capacityrequirement.