ZX Complex Flow Solver

Introduction

Since October 2012 the DNV GL ‘Stage 3’ statement in the Roadmap for Remote Sensing Devices (RSDs) has allowed for a standalone ZX 300 wind Lidar (i.e. no need for an onsite met mast) to be used for finance-grade (commonly referred as ‘bankable’) Energy Yield Analysis in the support of onshore wind farm development in simple, non-complex terrain.

Announced in January 2021 based on a broad evidence base, an industry collaboration has developed a ‘Complex Flow Solver’ that combines 50-point Continuous Wave Lidar measurements, high resolution open source Computational Fluid Dynamic modelling and powerful cloud computing to deliver ‘the best available solution for bankable wind resource and energy yield assessment based on standalone Lidar’ in complex terrain and unlocks the use of ZX 300 without the need for a met mast in any type of terrain onshore. The use of ZX CFS is approved by Deutsche WindGuard.

The details

All current commercial RSDs assume homogenous flow across the volume from which the measurement samples are taken. For the majority of Use Cases this assumption is valid and a wide body of evidence supports this assumption as is evidenced by more than 500 Performance Verifications of ZX 300 units at the UK Remote Sensing Test Site:

Table 1: Regression analysis of wind speed from > 500 ZX300 performance verifications from the UK RSTS

In complex terrain depending on the location of the Lidar this assumption can either still be valid or become invalid in the same way that if were possible to move a met mast across a complex site it would provide different wind measurements with only very small movements particularly if you were, for example, on the side of a hill. In cases where the assumption of homogenous flow is no longer valid, any Lidar measurement will differ when compared to a met mast that is in the same location and to date, this difference has resulted in the use of standalone Lidar to be limited as the onsite validation and therefore traceability of measurements is not available. The benefits of Lidar in complex terrain – portability, cost, planning, H&S – can be lost and it is this that drives the need for change and a new solution to address the challenge.

Figure 2: Left to right: homogenous flow in flat terrain; another example of homogenous flow in complex terrain, and; non-homogenous flow in complex terrain

Traditional cup anemometry itself also shows different responses to conditions such as inflow angles and turbulence (Albers et al., DEWI-Magazine 18, February 2001). This adds to the challenges of comparing RSDs with their own dependencies to these cups but as noted above the industry still demands this traceability, for now.

A white box approach born from industry collaboration

Full transparency and the understanding of any post-processing of data is essential in delivering any solution to using Lidar standalone in complex terrain and the challenge is well documented in IEC 61400-12-1 Ed 2 L.4.5 ‘Uncertainty due to non-homogenous flow within the measurement volume’ which states that the combination of a suitable flow model and a model reproducing the remote sensing device horizontal wind speed extraction algorithm might be used to assess the uncertainty due to non-homogenous flow.

Consideration of non-homogenous flow can legitimately be undertaken by detailed CFD analysis and results in the derivation of ‘conversion factors’ for Lidar measurements. Comprehensive analysis and documentation of the situation provides all info relevant for a bankable wind assessment. The uncertainty associated with the impact of complex flow on the Lidar measurement is reduced and replaces the otherwise more conservative Mann-Bingöl assessment of uncertainty and provides the input necessary for third-party bankable assessments.

Deutsche WindGuard, ZephyScience and ZX Lidars have collaborated to deliver the ‘ZX Complex Flow Solver’ (ZX CFS) that operates in this way; converting ZX Lidar data to be representative of a point / cup-equivalent measurement and in doing so providing a clear reduction of measurement uncertainty and data that can be used for bankable energy assessments in the development of wind farms.

High resolution CFD modelling

ZX CFS solves the three-dimensional flows using OpenFOAM[1], which is a flexible CFD toolbox and is applied world-wide for various purposes, including automotive and wind energy applications. The fully automated flow model provides a numerical solution of the steady-state Reynolds-Averaged Navier-Stokes Equations, modelling turbulence using the RNG-k-ε turbulence closure model, which is a widely applied and validated model describing turbulent transport and dissipation. OpenFOAM has the particular advantage of offering a parallelized unstructured calculation solver, which allows complex topographies to be resolved with more efficiency and with higher resolution. In order to detect small scale effects, a horizontal resolution of the calculation mesh of just 10 metres is applied with a synoptical resolution of 10°.

During calculations for the Bolund Case[2], which is the most comprehensive systematical flow model comparison for wind energy purposes performed to date, three out of the top five results (from all 50 participating models) were calculated by the model OpenFOAM.

This high resolution CFD model is made accessible to users through the ZephyFarm platform[3] – a front-end web-browser client application, powered by cloud computing to allow conversion factors to be produced for each Lidar deployment in ~1hr.

ZX Complex Flow Solver approach

The basic steps for running ZX CFS are defined as:

1. User uploads to ZephyFarm portal:
   1. Lidar wind data measurement file; or
   2. Lidar location & measurement heights (if a. is considered too sensitive or for pre-deployment site assessment)
2. Outputs include:
   1. A Conversion Intensity Map for the optimisation of Lidar siting before deployment which minimises project uncertainty as much as possible and is a useful, perhaps essential, element of considering project value before any actual measurement campaign commences.
   2. Conversion factors to be applied to 10-minute wind speed data (user then simply multiplies wind speed by the conversion factor, which is a function measurement height and direction)
   3. Converted data (if original Lidar data is uploaded)
3. Deutsche WindGuard provides approval (with quantified uncertainties) for use of standalone Lidar data in complex terrain.

Validation and Impact

Over the past 12 months, well-instrumented sites have been assessed using ZX CFS to convert the installed ZX 300 wind data and compare the obtained results to that of the collocated cup-instrumented meteorological mast. ZX CFS Validation results demonstrate an improvement in agreement between the ZX 300 Lidar and masts in complex terrain. There is a clear removal of the bias caused by non-homogenous flow. In addition, no boundaries in the use of ZX CFS have been found to date. The correlation coefficient (the statistical relationship between the two) can be seen to improve towards unity in the vast majority of cases. An in-depth review of the validations focussed on various sites where all cup calibration documentation was available and in addition to the correlation improvement, an improvement in regression was also achieved.

Conclusions

Through careful validation by independent wind engineering company Deutsche WindGuard, ZX Complex Flow Solver (CFS) appropriately addresses the challenges of using Lidar in complex terrain and provides a solution for bankable wind resource and EYA based on a standalone ZX Lidar. The use of ZX CFS is approved by Deutsche WindGuard.

This study builds on previous work in this area including:

• Post conversion of Lidar data on complex terrains, Sanquer, Woodward – Meteodyn, ZephIR Lidar
• Uncertainty of CFD-adjusted lidar measurements, Marmander, Abiven, Juguet – Natural Power
• Validated adjustment of remote sensing bias in complex terrain using CFD – Harris, Locker, Douglas, Girault, Abiven, Brady – Natural Power

[1] The OpenFOAM Foundation: OpenFOAM 8.0; 2020; www.openfoam.org

[2] CFD+engineering: The Bolund Case – Comparison of Simulation Results with Measured Data, Version 1.0.0 , 08/2010.

[3] ZephyFarm – Wind Farm Engineering Solutions; zephyfarm.io

The solution is described in detail in the following 1hr webinar: