Lidar and the integration of wind data into digital workflows

Lidar is a versatile wind measurement instrument that can both emulate the capabilities of conventional met mast mounted anemometry and go beyond those capabilities to acquire richer data sets that would not otherwise be available. These additional data may not be useful for the purposes of the strict implementation of established wind resource assessment procedures based on limited met mast capabilities. However, they give us a pathway to full digitisation of wind data during the earliest stages wind power projects, such as resource assessment, in a manner that is consistent with measurements that can be acquired post-construction. This supports fuller integration of these wind data into the digital workflows required to achieve a lifetime of successful project performance, and the bench-marking and monitoring of that performance against real-world wind conditions. 

The limitations of met mast methodologies are now more apparent than ever. For example, the wind power industry is currently confronting issues such as global blockage effects, which can be seen as artefacts of methodologies based on the limited capabilities of met masts. The real implications of this are that we can no longer apply "simple terrain" approximations or distinguish between pre- and post-construction phases of project delivery. The circumstances that determine the wind conditions on which project performance is based only arise once the wind farm is constructed. The wind farm itself modifies wind conditions in a similar way to the terrain. It is part of the terrain, and in "simple terrain" the wind farm is the terrain. That is, there is no such thing as "simple terrain". We only ever imagined there was such a thing as a way of describing a situation in which met masts perform in a conveniently consistent manner in different locations, obviating the need for site calibration. It transpires that that situation was a fantasy. We need new measurement and analysis methodologies that can be applied consistently both pre- and post-construction to accommodate reality. 

When considering wind turbine design load cases, met mast measurements are used to initiate computational fluid dynamics (CFD) models which describe wind conditions. These are then coupled to aeroelastic models (AEM) which describe the aerodynamic forces that arise as a consequence of the wind turbine encountering the wind. These are then coupled to wind turbine models using finite element methods (FEM) to propagate these forces through the structure to predict, for example, fatigue loads such as blade root bending moments, tower bending moments, torque variance in the drive train, and so on. These loads can then be accommodated in the design of the wind turbine.  

Where lidars replace met masts they must emulate their functionality. In the language of lidar use cases adopted by IEA Wind Task 32, this means using measurement methods that fulfil the same data requirements as met masts, under the same circumstances or operational conditions they fulfil them in. However, lidars operate in a fundamentally different manner. They directly measure so-called intermediate variables such as radial wind speed, the time-of-flight of the emitted pulse (and hence the distance along the line-of-sight to the probe volume where the radial wind speed measurement is acquired), beam azimuth and elevation angles, time, and so on. These must be processed using a wind field reconstruction (WFR) algorithm based on a model of the relationship between wind conditions and intermediate variables, to produce final variables that fulfil the same data requirements as are fulfilled by met masts, such as horizontal wind speed and wind direction. The WFR model describes what a met mast would have measured under the same conditions in which the lidar recorded its intermediate values.  

There are two problems. 

• Firstly, the WFR model is typically inadequate under all but the simplest circumstances, such as horizontally homogenous flow. This can result in measurement ambiguity and uncertainty. You may be familiar with "complex terrain bias," where variation in wind conditions within the volume surveyed by the lidar violate the assumptions of the WFR model.
• Secondly, the met masts themselves, or as emulated by the lidar, do not uniquely determine CFD predictions, introducing uncertainty in any analysis that relies on them. Met masts acquire relative sparse data, compared to lidar. There are a range of different simulated conditions which are consistent with the same met mast measurements. This range represents an uncertainty in any predictions based on these measurements.  

The contribution to uncertainty by WFR is normally overcome by brute force, in the sense of acquiring more data using lidar than could be obtained using met masts, so that the reduction in uncertainty due to volume of data more than compensates for the increase due to WFR. 

But the solution lies in going beyond the capabilities of met masts and properly exploiting the capabilities of the lidar. There is no need to replicate the limitations of met masts with WFR. Minimum uncertainty is achieved by directly validating CFD with intermediate variables. This eliminates the two major sources of uncertainty: the WFR algorithm and CFD initiation and/or validation using met masts. The richer, more detailed lidar data sets can validate the CFD predictions more precisely. At the same time, if the CFD is used to predict intermediate, rather than final values, such as radial wind speed relative to the lidar position, in multiple locations throughout the volume being modelled, then there is no need to employ WFR. Ideally, to close the loop, we would use CFD codes that can be initiated by intermediate rather than final values, and cut the met mast out of the loop entirely.  

This is compatible with the procedures that cope with our inability to neatly distinguish between pre- and post-construction scenarios. The same instruments and methods can be applied in a consistent manner during any phase of project delivery. The same measurements can be made pre-construction as post-construction, the only difference being that the wind turbines pre-construction only exist in silico. However, it is possible to have greater confidence in earlier benchmarks due to the consistency of the methods. So one consequence is that the data requirements arising at every phase of project delivery becomes aligned. Lidar, its digitisation of wind data, and their integration into a coherent digital workflow, makes this possible. This means the development of realistic uncertainty budgets and the adoption of better informed measures to reduce them, as described in my article "one step backwards, two steps forwards: digitalisation and uncertainty in wind energy assessments," can be achieved.