1. Reliability: The other dimension of quality, W.Q. Meeker, L.A. Escobar, Qual. Tech. & Quant. Management., 1, 1,pp. 1-25, 2004 2. Method for Testing the reliability of complex systems, Ch. Gray, N. Haselgruber, F. Langmayr, Patent application EP12180254.0, 2012 3. Physics of failure approach to wind turbine condition based maintenance, Chr. S. Gray, S.J. Watson, Wind Energy online, DOI: 10.10002/we.360, 2009 4. Reliability of wind turbine blades: An overview of materials testing, J.W. Holmes, B.F. Sörensen, P. Bronsdsted, Wind Power Shanghai 2007, proceedings Perform risk mitigation activities in parallel with product development Combine contributions from OEM and suppliers in a joint process Start component durability testing as soon as possible to prepare system validation Start validation activities with component maturity demonstration Turbine reliability is key to lifetime profitability of wind farms. Standards give a guideline for proper selection of components with respect to expected wind conditions, i.e. primarily to cope with different levels of fatigue load. Failure statistics show a wide range of failure modes in the field, not all of them being well reflected by wind classification. The contribution of turbine test operation to lifetime demonstration is limited due to lack of acceleration potential. However, component testing can play a major role in maturity demonstration if durability tests are performed. Moreover, there is further potential for risk reduction by dedicated material tests. Detailed understanding of potential failure mechanisms is necessary to identify useful tests and quantify their contribution to reliability demonstration. A reliability improvement program for wind turbines is proposed. It starts with the investigation of risks, including technical, quality and organizational aspects. According to the identified risks for each failure mode the corresponding risk mitigation tasks are derived: load case simulation, failure mode investigation, quality system review, supplier assessment, etc. Input data quality is critical for adequate test design: time logs of wind and climate conditions, grid quality, site location, failure statistics, technical specifications of turbine candidates are necessary. Lack of knowledge or low quality data can turn out as the major risk during early project phases. Proper component testing requires the understanding of component failure modes. Therefore, in a consecutive step the potential failure mechanisms of turbine components have to be analyzed. If this bottom-up analysis is done strictly failure mode related, it delivers damaging operation conditions and boundary conditions. These correlations can be used for describing damage kinetics with physics of failure models, based on observable operation and climate data logs. Such models have been successfully used for test development, for adaptation of test procedures to field conditions, for evaluation of test acceleration with respect to certain failure modes. These models are used to evaluate the overall demonstration potential of a turbine validation program. Such an assessment delivers weak-points as well as over-testing. It identifies particularly aggressive conditions for a given location with respect to certain failure modes and checks whether complementary measures have to be taken for risk mitigation. During recent years the process described above has been applied successfully to various industries such as automotive (pass cars, heavy duty trucks, railway) or industrial equipment (e.g., hydraulics). The potential for wind turbines is illustrated by showing results of a particular analysis, with details given for the turbine blades. It shows clearly the potential of dedicated test rigs to cover real lifetime fatigue load while further tests are required for various other failure modes. A corresponding component maturity demonstration plan is presented. Component maturity demonstration is a powerful initial step of turbine reliability validation Front loading of reliability and lifetime demonstration with component tests: accelerated / dedicated to certain failure modes / cost effective / parallel Elimination of risk is possible for certain failure modes by component testing Physics of failure has to be understood for quantitative test acceleration Duty cycle load histories – e.g. from SCADA - are essential for test parameterisation Suppliers are able to perform a high level of component maturity demonstration Subsequent turbine durability tests are necessary Representative turbine operation addresses a-priori unknown failure modes, interactions and interfaces As turbine tests are generally not accelerated reliability growth is a realistic system validation target SCADA systems bear a high potential for detection of failure mode precursors Adaptation of data processing and classification is necessary for monitoring of failure mode evolution Feed-back from service staff and detailed failure analyses are essential Abstract Maturity Check for Superior Component Reliability Franz Langmayr, Christopher Gray, Nikolaus Haselgruber Uptime Engineering GmbH, Graz, Austria, [email protected] PO. ID 141 Component Validation Potential Reliability Improvement Program Conclusions Durability & Reliability Failure Potential References Investigate correlation between failure modes and operation conditions Derive action plan for failure mode related risk reduction Combine simulation and measurement to eliminate component risks Action plan for failure mode investigation Set-up test hierarchy for each failure mode Example: Blade Step-by-step validation Parallel testing Supplier contribution Quantify load conditions – clarify failure physics - measure endurance and load capacity – define test hierarchy – demonstrate component maturity – perform turbine tests Promote understanding of failure modes, develop adequate tests Collect data to characterize duty cycles and durability tests Evaluate Test efficiency for all relevant failure modes Develop program for homogeneous high level of validation Component Maturity Demonstration Subsystem Component Failure mode Failure location Load Case Simulation Load Case Measurement Load Capacity Assessment Internal structure internal laminates debonding adhesive layer joining skin and main spar at pressure side eigenmode analysis to identify critical areas strain measurement for FEA calibration HCF testing of blade section - lifetime curve (Wöhler) internal laminates debonding sandwich panels - face to core: main spar web eigenmode analysis to identify critical areas strain measurement for FEA calibration HCF testing of blade section - lifetime curve (Wöhler) internal laminates fatigue main spar laminates eigenmode analysis to identify critical areas strain measurement for FEA calibration effect of aging (thermal, UV, ozone) on fatigue endurance of laminates internal laminates fatigue main spar - pressure side eigenmode analysis to identify critical areas strain measurement for FEA calibration HCF testing of blade section - lifetime curve (Wöhler) Lightning protection tip cables arc formation around receptor heat transfer from conductor to surroundings electric power of lightning stroke number/energy of lightning strokes to failure slip-rings at bearings wear running surface thermal condition during lightning stroke electric power of lightning stroke, temperature variation of wear rate with current, temperature and humidity, investigation of failure mode C-brushes wear running surface thermal condition during lightning stroke electric power of lightning stroke, temperature variation of wear rate with current, temperature and humidity, investigation of failure mode C-brushes wear running surface thermal condition during lightning stroke electric power of lightning stroke, temperature variation of wear rate with current, temperature and humidity, investigation of failure mode Paint and coatings gel-coat cracking, debonding - blade surface temperature (rotating, stationary) UV intensity, blade surface temperature (rotating, stationary) exposure to UV, temperature, Arrhenius lifetime tests paint aging - blade surface temperature (rotating, stationary) UV intensity, blade surface temperature (rotating, stationary) exposure to UV, temperature, Arrhenius lifetime tests paint erosion leading edge CFD momentum transfer of particle stream particle freight in air (season, location) CFD assessment of erosion on uplift, measurement of erosion effect on power curve Skins - laminates fatigue Skin - sandwich panels - face to core; skin laminates eigenmode analysis to identify critical areas strain measurement for FEA calibration HCF testing of blade section - lifetime curve (Wöhler) fatigue leading and trailing edge adhesive layer, joining the pressure and the suction side eigenmode analysis to identify critical areas strain measurement for FEA calibration HCF testing of blade section - lifetime curve (Wöhler) fatigue skin and main spar at pressure side eigenmode analysis to identify critical areas strain measurement for FEA calibration HCF testing of blade section - lifetime curve (Wöhler) pollution leading edge CFD momentum transfer and flow re-direction, deposition particle freight in air (season, location) wind tunnel, polluted air, deposition erosion leading edge CFD momentum transfer of particle stream particle freight in air (season, location) CFD assessment of erosion on uplift, measurement of erosion effect on power curve T-bolt/root inserts fatigue threat ground, stud basis - pressure side FEA of stress under critical load superposition strain measurement for FEA calibration HCF testing of mounting zone and stud De-icing system local overheating surrounding of heating wires / hot air channels heat transfer from conductor via surroundings to ice layer heating rate cyclic freezing-de-icing lifetime test Risk Mitigation - Action Plan - Blade Preview IEC I, High Wind; v = 10 m/s, winter IEC II, Medium Wind; v = 8,5 m/s, winter Fatigue test pulsating bending Failure Mode Fatigue Test IEC II IEC II IEC I Complementary Test Duration [h] 50 58400 58400 58400 Debonding of sandwich panels (HCF) 18230,0 1,0 1,7 21,2 - erosion and deposition of paint and coating 0,0 1,0 1,0 1,3 wind tunnel test with deposing and abrasive materials on blade segments Debonding of laminates, shear forces (HCF) 28,4 1,0 1,8 28,5 - HCF at suction side 1563292,6 1,0 1,7 1936,8 - Local overheating around de-icing system 0,0 1,0 1,0 1,1 cyclic icing de-icing test on blade segments Thermal aging of coating 0,0 1,0 1,0 1,1 UV and thermal aging of blade segments Thermal aging of laminate 0,0 1,0 1,1 1,1 UV and thermal aging of laminates Equivalent Lifetime (normalized) Certain fatigue failure modes are well represented by a pulsating bending test. Complementary durability tests have to be performed for several other failure modes. Maturity demonstration can be achieved via component testing to a high degree. Full system validation requires subsequent complementary turbine testing. Detailed test design is based on duty cycle data for all future sites and conditions. Lifetime limiting failure mechanisms are reflected in damage driving operation modes and boundaries Physics of failure models quantify the damage effect of load cases and tests Component testing is specified for fast and cost effective risk reduction EWEA 2013, Vienna, Austria: Europe’s Premier Wind Energy Event