Dr. Davide Astolfi

Full-scale wind turbines are mostly a mature technology: this means, on one side, that public subsidies are declining and, on the other side, that the target of 100% technical availability is becoming realistic. This motivates further investment and research for improving the efficiency of wind kinetic energy conversion: therefore, wind turbine retrofitting has become a widespread activity in the recent years. Basically, wind turbine upgrades belong to two categories: aerodynamic improvement (as, for example, vortex generator installation) or control system improvement (as, for example, pitch angle optimization or soft cut-out strategies for extending the operation at very high wind speed). This kind of interventions has material and labor cost and producible energy is lost during the installations. Further, an a priori estimate of the production upgrade is typically provided on the grounds of numerical simulations or field tests under controlled, ideal conditions, that are commonly very different with respect to the ones to which wind turbines are subjected in real environments. Therefore, a realistic estimate of the production improvement is fundamental, in order to inquire if a certain wind turbine upgrade has an advantageous return of investment. Due to the non-stationary operation conditions of wind turbines, it makes little sense to compare the production before and after an upgrade: it makes much more sense to compare the post-upgrade production against a model of the pre-upgrade production under the same conditions. Since the power upgrades are of the order of the percent, a theoretical simplified model of wind turbine operation often lacks the required precision: models should therefore be data-driven. The availability of Supervisory Control And Data Acquisition (SCADA) data is therefore fundamental for this kind of problems, but their fruitful use is non-trivial. This stimulates the cooperation between academia and industry; in particular, this work reports results from the collaboration between Renvico (www.renvicoenergy.com) and the Department of Engineering of the University of Perugia. Renvico owns and manages 335 MW of wind turbines in Italy and France and has established a good practice with the academic partner: wind turbine upgrades are tested on some pilot operating wind turbines and specific models are formulated for computing the production improvement. On these grounds, the decision of extending or not the upgrade to other wind turbines in the Renvico wind farms is based. In this work, several test cases of wind turbine upgrades are discussed: pitch angle optimization near the cut-in, vortex generators and passive flow control devices installation, soft cut-out strategies for power curve extension above the nominal cut-out. Appropriate models are formulated for the study of each test case and estimates of the production improvement are provided. Possible critical issues are discussed, as for example mismatches with respect to the estimates provided under hypothesis of ideal conditions, exacerbation of stressing mechanical conditions possibly affecting wind turbine residue lifetime, and so on. The vastness of the analyzed test cases provides a versatile tool box of methods for a critical study of basically every kind of wind turbine upgrade. 

Davide Astolfi is a post-doc at the Department of Engineering of University of Perugia, Italy. His research activities deal mainly with modelling and control of mechanical systems, condition monitoring and fault diagnosis of wind turbines through data mining and numerical simulation of wind flow and wakes. He also works on wind tunnel testing of micro wind turbines and vehicles for aerodynamic and performance analysis.