Search

The most commonly employed tool for wind turbine performance analysis is the power curve, which is the relation between wind intensity and power. The diffusion of SCADA systems has boosted the adoption of data-driven approaches to power curves. In particular, a recent research line involves multivariate methods, employing further input variables in addition to the wind speed. In this work, an innovative contribution is investigated, which is the inclusion of thirteen sub-component temperatures as possible covariates. This is discussed through a real-world test case, based on data provided by ENGIE Italia. Two models are analyzed: support vector regression with Gaussian kernel and Gaussian process regression. The input variables are individuated through a sequential feature selection algorithm. The sub-component temperatures are abundantly selected as input variables, proving the validity of the idea proposed in this work. The obtained error metrics are lower with respect to benchmark models employing more typical input variables: the resulting mean absolute error is 1.35% of the rated power. The results of the two types of selected regressions are not remarkably different. This supports that the qualifying points are, rather than the model type, the use and the selection of a potentially vast number of input variables. [ABSTRACT FROM AUTHOR]

An increasing amount of wind turbines, especially in Europe, are reaching the end of their expected lifetimes; therefore, long data sets describing their operation are available for scholars to analyze the performance trends. On these grounds, the present work is devoted to test case studies for the evaluation and the interpretation of wind turbine performance decline with age. Two wind farms were studied, featuring widely employed wind turbine models: the former is composed of 6 Senvion MM92 and the latter of 11 Vestas V52 wind turbines, owned by the ENGIE Italia company. SCADA data spanning, respectively, 10 and 7 years were analyzed for the two test cases. The effect of aging on the performance of the test case wind turbines was studied by constructing a data-driven model of appropriate operation curves, selected depending on the working region. For the Senvion MM92, we found that it is questionable to talk about performance aging because there is no evident trend in time: the performance variation year by year is in the order of a few kW and is therefore irrelevant for practical applications. For the Vestas V52 wind turbines, a much wider variability is observed: two wind turbines are affected by a remarkable performance drop, after which the behavior is stable and under-performing with respect to the rest of the wind farm. Particular attention is devoted to the interpretation of the results: the comparative discussion of the two test cases indicates that the observed operation curves are compatible with the hypothesis that the worsening with age of the two under-performing Vestas V52 can be ascribed to the behavior of the hydraulic blade pitch. Furthermore, for both test cases, it is estimated that the gearbox-aging contributes negligibly to the performance decline in time. [ABSTRACT FROM AUTHOR]

In this work, the combination effects on wind turbine performances of wakes and terrain-driven flow are investigated. The test case is a subcluster of four turbines from a wind farm sited in southern Italy in a very complex terrain. The layout, the inter-turbine distance and the wind rose result in a challenging performance scenery. The subcluster is analyzed, when the wind blows from the west, through computational fluid dynamics numerical simulations and experimental supervisory control and data acquisition data mining. Two wind intensity regimes and several simulation setups are employed. It is shown that the main effect of the terrain is the northward distortion of the wake of the upstream turbine. This explains the non-trivial yawing patterns of the cluster and the fact that the wake line affects the overall performances of the subcluster less than it would do in flat terrain. It is further shown that the presence of the rest of the subcluster in operation southward deviates the wake line of the upstream turbine. The dependency on wind intensity of these directional distortions allows to estimate the relative importance of wakes and terrain-driven flow. A bijective feedback between models and data is established and a convincing framework is constructed, for separating and assessing the effect of the terrain and of the single and multiple wake. Copyright © 2017 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Optimal alignment of wind turbines to the wind direction is a crucial condition for the quality of power output and for the health of the turbines. Actually, bad alignment can cause degraded performances and dangerous loads that can affect, on the long run, the mechanical safety of the wind turbine. Supervisory Control And Data Acquisition (SCADA) systems are becoming widespread in modern wind energy technology because of the appreciable costs - benefits ratio. The common time scale of SCADA, yet, usually is not effective for misalignment diagnosis because the wind varies too rapidly. For this reason, misalignment is often diagnosed using ad hoc techniques as LIDAR-based or spinner anemometers. In the present work, it is shown that very useful indications for the diagnosis of misalignment can be obtained also from the SCADA data, without invoking expensive supplementary control techniques. The method is validated on the data set of a wind farm sited in Italy. [ABSTRACT FROM AUTHOR]

Wind turbines undergo dynamic loads along all the phases of transformation of wind kinetic energy into power output to be fed into the grid. Gearbox breakdowns are one of the most common and most severe causes of energy losses and it is therefore crucial to prevent and forecast them. Straightforward vibration analysis is very demanding by the point of view of technology, costs and complexity of signal denoising. A considerable keystone in fault diagnosis is the analysis of Supervisory Control And Data Acquisition (SCADA) systems. In particular, thermal behaviour of wind turbines fits well with the common time scale of SCADA data; heating trends are fairly responsive as a consequence of rotor vibration. Machine learning techniques applied to SCADA data are very powerful in reconstructing inputs - output dependency. On these grounds, in this work an Artificial Neural Network approach is proposed for early diagnosis of gearbox faults. The method is validated on the data of a wind farm operating in Italy. It is shown that the method is capable in recognizing incoming faults with a very manageable advance also with data on short time scales. [ABSTRACT FROM AUTHOR]

Supervisory control and data acquisition (SCADA) systems have become widely diffuse in modern wind energy technology. The slowdown of new installations and the increasing percentage of energy entering the grid from renewable stochastic sources has diverted attention to the careful optimization of operating farms. Elaborating the complex data stream from SCADA systems into knowledge poses technological and scientific challenges. SCADA data analysis therefore lies at the crossroads of mechanical engineering, applied mathematics, statistics and physics. In the present work, mathematical methods are proposed for tackling the complexity of SCADA data. This idea is to elaborate simplified and more powerful data sets through one action: discretization of continuous quantities. The approach is employed for two very different issues: performance evaluation and wake effects analysis, which is investigated from the point of view of power losses, due to the difficulties associated with optimal turbine alignment with the wind. Two indexes for performance evaluation are formulated. Recurrent non-trivial orientation patterns of clusters of turbines are individuated, and the efficiency associated to them is analyzed. The methods are tested on two wind farms situated in southern Italy. [ABSTRACT FROM AUTHOR]

Due to the stochastic nature of the source, wind turbines operate under non-stationary conditions and the extracted power depends non-trivially on ambient conditions and working parameters. It is therefore difficult to establish a normal behavior model for monitoring the performance of a wind turbine and the most employed approach is to be driven by data. The power curve of a wind turbine is the relation between the wind intensity and the extracted power and is widely employed for monitoring wind turbine performance. On the grounds of the above considerations, a recent trend regarding wind turbine power curve analysis consists of the incorporation of the main working parameters (as, for example, the rotor speed or the blade pitch) as input variables of a multivariate regression whose target is the power. In this study, a method for multivariate wind turbine power curve analysis is proposed: it is based on sequential features selection, which employs Support Vector Regression with Gaussian Kernel. One of the most innovative aspects of this study is that the set of possible covariates includes also minimum, maximum and standard deviation of the most important environmental and operational variables. Three test cases of practical interest are contemplated: a Senvion MM92, a Vestas V90 and a Vestas V117 wind turbines owned by the ENGIE Italia company. It is shown that the selection of the covariates depends remarkably on the wind turbine model and this aspect should therefore be taken in consideration in order to customize the data-driven monitoring of the power curve. The obtained error metrics are competitive and in general lower with respect to the state of the art in the literature. Furthermore, minimum, maximum and standard deviation of the main environmental and operation variables are abundantly selected by the feature selection algorithm: this result indicates that the richness of the measurement channels contained in wind turbine Supervisory Control And Data Acquisition (SCADA) data sets should be exploited for monitoring the performance as reliably as possible. [ABSTRACT FROM AUTHOR]

The widespread availability of wind turbine operation data has considerably boosted the research and the applications for wind turbine monitoring. It is well established that a systematic misalignment of the wind turbine nacelle with respect to the wind direction has a remarkable impact in terms of down-performance, because the extracted power is in first approximation proportional to the cosine cube of the yaw angle. Nevertheless, due to the fact that in the wind farm practice the wind field facing the rotor is estimated through anemometers placed behind the rotor, it is challenging to robustly detect systematic yaw errors without the use of additional upwind sensory systems. Nevertheless, this objective is valuable because it involves the use of data that are available to wind farm practitioners at zero cost. On these grounds, the present work is a two-steps test case discussion. At first, a new method for systematic yaw error detection through operation data analysis is presented and is applied for individuating a misaligned multi-MW wind turbine. After the yaw error correction on the test case wind turbine, operation data of the whole wind farm are employed for an innovative assessment method of the performance improvement at the target wind turbine. The other wind turbines in the farm are employed as references and their operation data are used as input for a multivariate Kernel regression whose target is the power of the wind turbine of interest. Training the model with pre-correction data and validating on post-correction data, it is estimated that a systematic yaw error of 4 ∘ affects the performance up to the order of the 1.5% of the Annual Energy Production. [ABSTRACT FROM AUTHOR]

Full-scale wind turbine is a mature technology and therefore several retrofitting techniques have recently been spreading in the industry to further improve the efficiency of wind kinetic energy conversion. This kind of interventions is costly and, furthermore, the energy improvement is commonly estimated under the hypothesis of ideal wind conditions, but real ones can be very different because of wake interactions and/or wind shear induced by the terrain. A precise quantification of the energy gained in real environment is therefore precious. Wind turbines are subjected to non-stationary conditions and therefore it makes little sense to compare energy production before and after an upgrade: the post-upgrade production should rather be compared to a model of the pre-upgrade production under the same conditions. Since the energy improvement is typically of the order of few percents, a very precise model of wind turbine power output is needed and therefore it should be data-driven. Furthermore, the formulation of the model is heavily affected by the features of the available data set and by the nature of the problem. The objective of this work is the discussion of some wind turbine power curve upgrades on the grounds of operational data analysis. The selected test cases are: improved start-up through pitch angle adjustment near the cut-in, aerodynamic blade retrofitting by means of vortex generators and passive flow control devices, and extension of the power curve through a soft cut-out strategy for very high wind speed. The criticality of each test case is discussed and appropriate data-driven models are formulated. These are employed to estimate the energy improvement from each of the upgrades under investigation. The general outcome of this work is a catalog of generalizable methods for studying wind turbine power curve upgrades. In particular, from the study of the selected test cases, it arises that complex wind conditions might affect wind turbine operation such that the production improvement is non-negligibly different from what can be estimated under the hypothesis of ideal wind conditions. A complex wind flow might actually impact on the efficiency of vortex generators and the soft cut-out strategies at high wind speeds. The general lesson is therefore that it is very important to estimate wind turbine upgrades on real environments through operational data. [ABSTRACT FROM AUTHOR]

• Performance decline with age of 2 MW wind turbines is analyzed. • Innovative data-driven methods based on operation curves analysis are employed. • The aging is mainly due to a slight diminishing of extracted power for given generator speed. • The estimated rate of performance decline is in the order of − 0.2% per year. The worsening with age of wind turbine performance is an expected phenomenon, which is practically impossible to estimate a priori. The objective of this study is formulating a method for the estimation of wind turbine performance decline with age, basing on long term SCADA data analysis. Two test cases, featuring in total fifteen 2 MW wind turbines, have been selected because at present there are no similar studies devoted to wind turbines of this size. The study is focused on the operation regime characterized by variable rotor speed and practically fixed pitch and it is based on the analysis of the rotor speed - power and generator speed - power curves through the binning method and through a Support Vector Regression with Gaussian Kernel. The main result is that the average rate of performance decline with age for the considered test cases is in the order of − 0.2% per year, which is compatible with the most recent analysis in the literature based on cumulative data. Furthermore, it is estimated that the gearbox aging does not contribute to the performance decline, while instead the generator aging does. [ABSTRACT FROM AUTHOR]