Evaluating The Reliability And Recognizing The Weak Parts Of Wind Energy Conversion Systems

Under the dual pressures from energy crisis and environmental protection,large-scale wind power development and utilization have become important measures tochange the energy composition of many countries in the world. In recent years, theinstalled capacity of wind power and rated capacity of individual wind energyconversion systems (WECS) have been increasing, especially in China, for many yearsto maintain a high growth rate of installed capacity. Currently, China has the largesttotal installed capacity of wind power in the world. The WECS reliability has significanteffects on the WECS design, development, maintenance, repair, wind farm planning,power system operation and management. Therefore, it is important to study the relativeissues of WECS reliability.Supported in part by the National Natural Science Foundation of China(“Probabilistic model and application of wind farm reliability evaluation”, No.51077135), this thesis conducted the researches and explorations of WECS reliability.The studies focus on the impacts of wind speed and other factors on the reliability ofwind turbine power converter system, impacts of wind speed and temperature on theinsulation life and capacity determination of wind power transformer, impacts ofparallel structure of generators and converter modules on the WECS reliability and theidentification method of WECS weak links.A multi-state probability analysis method is proposed considering the operatingcharacteristic of WECS power converter varying with wind speeds. Wind speed effectmodels of failure rate for power converter subsystems (Generator-side inverter, DC-linkand Grid-side inverter, etc.) and their power electronic components in WECS are built,and then wind speed effect model of WECS power converter reliability is built. Inaddition, the impacts of wind speed and wind turbine parameters (Cut-in, Rated andCut-out wind speeds) on the reliability of power converter in a WECS are analyzed. Theresults of case studies indicate that the reliability of wind turbine power converter isaffected significantly by the variations of wind speed and rated wind speed of windturbine. The relationship between wind speed and the converter failure rate is similar tothat between the output power of WECS and wind speed. In addition, the powerconverter reliability changes with seasons.Similarly, considering the characteristic that the operating state of transformer in a wind farm is affected by wind speeds and load factor of transformer changes with windspeeds, an evaluation model of wind farm transformer insulation aging is proposedbased on the IEEE transformer aging model. On basis of this, the capacity optimizationmodel of wind farm transformer is proposed to minimize the life cycle cost (LCC) oftransformer and meet its technique constraints (oil temperature, winding temperatureand maximal load factor). In addition, the impacts of energy storage system on thetransformer insulation life and capacity determination are analyzed. Case studies showthat the level of wind farm transformer insulation aging is related to wind speeds, andthe rated capacity of wind farm transformer can be reduced greatly using the proposedoptimization model. In addition, a positive correlativity between the rated capacity ofwind farm transformer and its LCC is verified. That is, the transformer with minimalcapacity to meet the technique constraints is economically optimal.According to the new structure of existing multiple generators or convertermodules in parallel in WECS, a new reliability index, annual expected energy loss(AEEL), is proposed and an analytical reliability model for this WECS is builtconsidering the effect of wind speeds to analyze the effects of different structures on thereliability performance of WECS. The results of case studies show that there is a largegap in essence of improving WECS reliability for two parallel structures.Based on the principle that the responsibility of system unreliability is assigned tofailed components and proportional-sharing-principle, a reliability-tracing model ofWECS is proposed to quantify the contribution of each subassembly to the WECSunreliability indices. According to the identification and analysis of WECS weak partsbased on different reliability indices, the annual losses due to forced outage arerecommended to be the most reasonable index for weak part identification because itcan comprehensively reflect the WECS reliability and economic performance. At thesame time, a flow chart for tracing the WECS unreliability and identifying weak parts ispresented. Finally, the impacts of improving the subassembly reliability performance onthe losses are analyzed. The proposed method can provide useful information foroptimization design and theoretical foundation for decision-making of operation andmaintenance strategy. The case studies show that the gearbox, blade/pitch and generatorare weaker parts of a double-fed WECS, which should be given great attention in theprocesses of system design, operation and maintenance.