Elsevier

Renewable Energy

Volume 121, June 2018, Pages 123-132
Renewable Energy

Small wind turbines: A numerical study for aerodynamic performance assessment under gust conditions

https://doi.org/10.1016/j.renene.2017.12.086Get rights and content

Highlights

  • An unsteady CFD model of small-sized wind turbine has been developed.

  • The EOG gust model presented in IEC-61400 has been used as reference.

  • The NREL Phase VI HAWT has been considered as case study.

  • Both aerodynamic and structural analyses have been performed.

Abstract

In the present work, a numerical study aimed to analyse the effect of an extreme loading event on a Horizontal Axis Wind Turbine (HAWT) is performed. A 3D unsteady CFD model of the NREL Phase VI small-sized wind turbine is validated against experimental data, with the incompressible solver of ANSYS Fluent and an unstructured moving mesh strategy. Then the Extreme Operating Gust (EOG) model from IEC 61400-2 is considered as the inlet condition. The results of the aerodynamic response and of the structural ultimate check, based on the IEC guidelines, are presented for both the operating and the parked turbine, in order to underline the benefit of the safety position in terms of lower stress transferred to the critical root section.

Introduction

Wind turbines are a clean alternative to conventional fossil fuel-based power production plants [1,2]. However, despite technological advancements occurred in the last decades, aerodynamic design and assessment are still challenging steps because these machines have to hold up strong and variable operating conditions during their lifetime [3,4]. Along with a higher efficiency [5,6], also a more reliable and consequently safer design needs to be considered [7,8]. Therefore a deep knowledge of the environmental conditions in which the turbine operates is fundamental. For this purpose, several methods and models, able to simulate wind gusts, have been developed in order to predict structural loads caused by unpredictable extreme events [9]. The one-minus-cosine profile is the simplest model and it is currently adopted for load assessment in aerospace industry [10]. Larsen et al. [11] discussed and validated a model to extract mean gust shapes based on the peak over threshold detection procedure applied to a suitably large number of measured gusts, but the availability of useful data is also limited in space and time. Seregina et al. [12] proposed the extrapolation of wind gust velocities from common hourly measurements, with the hypothesis that both wind and gust velocities follow a Weibull distribution. They also provided a validation using meteorological data. Mann [13] built an algorithm to simulate a 3D turbulent wind field with realistic atmospheric boundary layer behaviour, so including even gust events. The algorithm has been validated using two atmospheric experiments, both running for one year.

On the other hand, the IEC guidelines [14] are less accurate. Wright and Wood [15] pointed out an actual wind direction change that exceeds the IEC maximum, whereas Knigge and Raasch [16] used a virtual 3D atmospheric dataset from a LES simulation and obtained a steeper increase and decrease of the wind speed, if compared to the IEC profile. However, despite the accuracy presented by the above mentioned methods, the IEC gust models are simple and suitable for numerical analyses of wind turbines. As an example, Gillebaart et al. [17] simulated the active flap control of a wind turbine airfoil operating under an Extreme Operating Gust (EOG) model from IEC. They performed 2D Unsteady Reynolds Averaged Navier-Stokes (URANS) calculations and studied the aeroelastic response. Other works concern a whole turbine, and used the datasets of the NREL 5 MW offshore wind turbine for the validation. For instance, Tran et al. [18] analysed the response of the NREL machine operating under Extreme Coherent Gust (ECG) and EOG. They compared results of the unsteady CFD and the unsteady Blade Element Momentum (BEM) method. The same turbine was considered by Kim et al. [19], who investigated the structural behaviour of the machine for offshore application during an EOG event. Norris and Carter [20] adopted the ECG model with direction change and performed a LES analysis on a wind farm. The actuator disk method was applied, and it was combined to the FAST aeroelastic code by NREL. Gutpa [21] studied several IEC gust models using both the BEM theory and the lifting line theory.

NREL published also the dataset of the Unsteady Aerodynamics Experiment [22], performed in the NASA's Ames Research Center, was devoted to give aerodynamic and structural results of the Phase VI HAWT in controlled operating conditions. Wind tunnel testing facilities can provide reliable data, which can be used as input for numerical analyses. The development of accurate CFD simulations allows to improve the knowledge on the 3D unsteady effects, such as the radial development of stall and the vortex generation, characteristics of the field operation. After the experiment, NREL revealed the measured data to verify predictions from computational BEM and CFD codes developed by international experts [23].

Several HAWT models have been validated using the Phase VI dataset. Steady conditions and symmetry hypotheses have been adopted by Mo and Lee [24] and Sørensen et al. [25], using incompressible CFD solvers, whereas Le Pape and Lecanu [26], Yelmule and Anjuri [27] and Huang et al. [28] performed compressible validations. Sezer-Uzol and Long [29] and Song and Perot [30] considered the whole turbine rotor instead, and adopted an unsteady approach and moving grids. Also alternatives to full CFD analyses have been proposed. Wang et al. [31] predicted the blade root loads using an aeroelastic code. Hsu et al. [32] coupled an iso-geometric analysis with low-order FEM calculations to study the fluid-structure interaction. Lanzafame and Messina [33] implemented and improved a model based on the BEM method. Breton et al. [34] applied a lifting line code to test the accuracy of different stall delay models, used to correct 2D coefficients employed in BEM methods. The models showed a general over prediction of the performance of the turbine in stalled conditions. In fact the present study proposes an unsteady CFD methodology with high fidelity in modelling the 3D aerodynamic effects.

In the present work, the EOG model from Ref. [14] has been adopted to study the response of a small HAWT in different operating conditions. In Section 2 the gust model is explained, while in Section 3 the turbine model is presented and validated against wind tunnel measurements. The selected test case for the gust analysis is the small-sized NREL Phase VI, unlike the previously cited works, which focused only on the offshore multi-megawatt NREL machine. The geometry of the Phase VI is not typical for a small wind turbine, as explained by Wood [35], because the blades employed in the experiment are a scaled-down version of large wind turbine blades. Nevertheless the machine is considered a small-sized wind turbine for its dimensions and for the Reynolds number range, that is 0.4÷1.1106 in design conditions, so the work is presented taking in account the peculiarities of the machine. Indeed, the Reynolds effects are not completely understood for the small-sized machines, which are continuously spreading in southern Europe [36,37]. The aero-structural optimization also represents one of the main topics for the analysis of small-sized turbines [38,39]. However, the structural response of the blades to an impulsive load, such as a gust condition, has never been deeply investigated. Few authors developed an URANS model of a HAWT under gust conditions; specifically, the present paper performs the calculation using an hybrid structured-unstructured grid and the moving mesh method. Both the aerodynamic and structural aspects of the gust analysis are discussed in Section 4, while concluding remarks are given in Section 5. The turbine model has been studied in both operating and parked conditions.

Section snippets

Gust model

A wind gust is defined as a short term speed variation within a turbulent wind field. The typical shape can be characterized by the following parameters: the gust relative amplitude, the gust rise time, the maximum gust variation and the lapse time [40]. In IEC 61400-2 a series of severe transient events are defined as the variation of a wind profile superimposed over a mean flow, along with a set of extreme loading cases which the wind turbine should withstand during its lifetime [14]. The

Solver and validation

The adopted numerical method is based on the URANS incompressible pressure-based implicit solver of ANSYS Fluent 16.1. The sliding interfaces have been created by coupling the intersection surfaces initialized in the previous phase. The solver setup used to validate the model is presented in Table 3, while the velocity and density conditions are the same used in Ref. [27]. I=0.5% simulates the controlled inflow conditions of the wind tunnel [23]. The pressure measurements are referred to the

Results and discussion

The numerical CFD results of the NREL Phase VI wind turbine operating under gust conditions are organized as follows: the evolutions in time of the aerodynamic torque are presented in Section 4.1, bending moments and axial loads are discussed in Section 4.2, whereas the pressure distributions of 3 representative instants are analysed in Section 4.3. In Table 5 the main values are summarized.

Conclusions

A numerical model of a HAWT under gust conditions has been built and successfully validated, against the NREL Phase VI measurements, using unsteady CFD simulations of an unstructured moving mesh. The behaviour of the turbine operating under an EOG from IEC 61400-2 standard has been presented for the undisturbed operation and the parked position. An overall coherent trend is noticed in the evolution in time of bending moments, axial load and consequent stresses at the root section. The benefit

References (51)

  • WWEA Half Year Report

    (2016)
  • Key World Energy Statistics

    (2016)
  • M.O.L. Hansen

    Aerodynamics of Wind Turbines

    (2015)
  • T. Burton et al.

    Wind Energy Handbook

    (2001)
  • V. Vandecauter

    CFD Simulation of Atmospheric Wind Gusts

    (2011)
  • R.E. Bartels

    Development, Verification and Use of Gust Modeling in the NASA Computational Fluid Dynamics Code FUN3D

    (2012)
  • G.C. Larsen et al.

    Mean Gust Shapes, Risø-R-1133(EN)

    (2003)
  • L.S. Seregina et al.

    Development of a wind gust model to estimate gust speeds and their return period

    Tellus Dyn. Meteorol. Oceanogr.

    (2014)
  • J. Mann

    The spatial structure of neutral atmospheric surface-layer turbulence

    J. Fluid Mech.

    (1994)
  • Design requirements for small wind turbines

    IEC 61400-2:2013 International Standard

    (2013)
  • A.K. Wright et al.

    Rotor speed and gyroscopic loads on a small horizontal Axis wind turbine

    Wind Eng.

    (2007)
  • T. Gillebaart et al.

    Active flap control on an aeroelastic wind turbine airfoil in gust conditions using both a CFD and an engineering model

    J. Phys. Conf. Series

    (2014)
  • T. Tran et al.

    Extreme load estimation for a large wind turbine using CFD and unsteady BEM

  • B. Kim et al.

    Ultimate load characteristics of NREL 5-MW offshore wind turbines with different substructures

    Int. J. Energy Res.

    (2015)
  • R.C. Storey et al.

    Modelling turbine loads during an extreme coherent gust using large eddy simulation

    J. Phys. Conf. Series

    (2014)
  • Cited by (29)

    • Peak shaving strategy for load reduction of wind turbines based on model predictive control

      2023, Energy Reports
      Citation Excerpt :

      The peak thrust may destroy wind turbines potentially, which must be within the permissible range. In Ref. Menegozzo et al. (2018), the force and load characteristics of a small wind turbine under gust are studied by means of computational fluid dynamics. Adjusting the pitch angle is recommended as a good way to optimize loads.

    • Mechanism of mooring line breakage and shutdown opportunity analysis of a semi-submersible offshore wind turbine in extreme operating gust

      2023, Ocean Engineering
      Citation Excerpt :

      They found that the sudden increase in wind speed could lead to a nearly 25% and 20% gradual increment of rotor thrust and blade tip deformations. Menegozzo et al. (2017) studied the wake characteristics and blade pressure responses of a small-sized onshore wind turbine in EOG based on CFD numerical simulation. Storey et al. (2014) modeled turbine loads by large eddy simulation and found that ECG-induced wake turbulence can significantly affect turbine fatigue loading in transient atmospheric conditions.

    • Preset angle, aspect ratio, and stall representations effects on the aeroelastic responses of small-scale wind turbine blades

      2022, International Journal of Non-Linear Mechanics
      Citation Excerpt :

      In addition to low aspect ratios, the preset angle of attack characterizes most of the SSWT designs [43–47]. In fact, many studies showed how the aerodynamic performance of the latter wind energy harvesters can be remarkably enhanced by simply identifying the most suitable initial preset angle and overall twist distribution [45,48–50]. As a consequence, the effect of the latter parameter on the system’s stability and flutter boundary should also be assessed.

    • Modeling of near wake characteristics in floating offshore wind turbines using an actuator line method

      2022, Renewable Energy
      Citation Excerpt :

      The key advantage of ALM is its comparatively low computational cost. Henceforth, the computational cost for the current ALM code, a Blade-resolved RANS simulation performed by writers [48], and a Blade-resolved LES simulation [49] are compared in Table 4. As it can be seen, even using a shorter time step, the ALM is by far less computationally expensive.

    • 2.06 - Energy Yield of Contemporary Wind Turbines

      2022, Comprehensive Renewable Energy, Second Edition: Volume 1-9
    View all citing articles on Scopus
    View full text