TY - JOUR
T1 - Value of load monitoring in lifetime extension decision making for wind turbines
AU - Nielsen, Jannie Sønderkær
AU - Dimitrov, Nikolay Krasimirov
PY - 2020
Y1 - 2020
N2 - For wind turbines close to the end of the certified lifetime, life extension is in many cases an economically viable option. Although the design life is expired, a new assessment of fatigue life can be made considering updated information about the actual load exposure of the turbines. Information such as more accurate wind speed distributions, turbulence distributions and time in various operational conditions is generally available for larger offshore wind farms, whereas less information may be available in other cases. Load monitoring can be applied to reduce model uncertainties on the fatigue load, which can result in better estimates of the remaining fatigue capacity, asserting the availability of additional fatigue life. In this paper, the decision problem for load monitoring is set up, and a procedure for estimation of the value of load monitoring is proposed based on Bayesian decision analysis. The additional fatigue life arising from the site conditions being more benign than the IEC design conditions is assessed using a surrogate model trained on a generic aeroelastic model, and with probabilistic methods applied for uncertainty propagation. The value of load monitoring is found to mostly depend on the uncertainty structure of the fatigue load estimation, and on the fatigue life estimated using already available data.
AB - For wind turbines close to the end of the certified lifetime, life extension is in many cases an economically viable option. Although the design life is expired, a new assessment of fatigue life can be made considering updated information about the actual load exposure of the turbines. Information such as more accurate wind speed distributions, turbulence distributions and time in various operational conditions is generally available for larger offshore wind farms, whereas less information may be available in other cases. Load monitoring can be applied to reduce model uncertainties on the fatigue load, which can result in better estimates of the remaining fatigue capacity, asserting the availability of additional fatigue life. In this paper, the decision problem for load monitoring is set up, and a procedure for estimation of the value of load monitoring is proposed based on Bayesian decision analysis. The additional fatigue life arising from the site conditions being more benign than the IEC design conditions is assessed using a surrogate model trained on a generic aeroelastic model, and with probabilistic methods applied for uncertainty propagation. The value of load monitoring is found to mostly depend on the uncertainty structure of the fatigue load estimation, and on the fatigue life estimated using already available data.
M3 - Journal article
SN - 1475-9217
JO - Structural Health Monitoring
JF - Structural Health Monitoring
ER -