Viewpoint: How blockage effects skew energy-yield forecasts

In most global markets, pre-construction wind-energy assessments, and the energy yield estimates they contain, play a central role in the financing of utility-scale wind farms. 

Ørsted's announcement specifically mentioned unaccounted-for blockage effects as part of the overestimation and cited a 2018 paper from DNV GL on the subject as evidence that this likely represents an industry-wide issue.

This article aims to explain some of the issues behind these developments in simpler terms.  

Blockage effects

As wind approaches an obstacle — the blockage — its flow slows down and diverts around it ­— the blockage effect.

A similar slowdown occurs upstream of a wind turbine.

While this is also a blockage effect, it is more often referred to as turbine induction and has been well-known since the start of the wind industry.

There are many blockage-related effects that occur at a wind farm. Some are addressed in longstanding wind energy practices, but others are not.

It is the unaccounted-for blockage effects that have led to bias in commonly used energy-prediction methods.

For a given wind resource, accurate prediction of a turbine’s energy output requires a power curve that faithfully represents the turbine’s production when operating in isolation and an accurate estimate of how that production changes when the wind farm’s other machines are present.

Unaccounted-for blockage effects probably cause prediction bias in both areas.

Traditionally, when calculating turbine-interaction losses, the industry has used what we call the "wakes-only" approach.

In this approach, a turbine can only affect machines located downstream — any influence on turbines upstream or to the side is almost always ignored.

In other words, the turbines in the front row of a wind farm, collectively and individually, are assumed to produce the same amount of energy as they would operating in isolation.

Measurements and high-fidelity simulations, however, indicate that front-row turbines are very likely to produce less energy than they each would in isolation, mainly due to wind-farm-scale blockage effects — and the related prediction bias extends to the entire project.

This is because "wakes-only" turbine-interaction models are tuned to predict the production of downstream turbines accurately relative to the front row.

Power curve curveballs

Turbine-power performance curves are another potential source of bias. The standards for these are designed to mitigate the undue influence of blockage on measured power curves.

Recent studies, however, indicate that these effects still have a material impact. The induction from the test turbine alone can affect the wind speed measured upstream.

Blockage effects from the surrounding turbines can skew the wind-speed relationship between the met mast and test turbine.

The result is a power curve that is different, and typically more energetic, than the power curve we need for a wind-energy assessment.

Therefore, to the extent that power curves used in these calculations are consistent with measured power curves, the bias in the latter will generally affect wind-farm energy predictions.  

Remedies

These biases pertain to energy-estimation methods used throughout the wind industry. They are probably material and apply to most wind farms, offshore and onshore.

Their magnitudes can vary significantly and depend on site-specific details, such as the turbine layout and local meteorological conditions.

Reliable corrections should be based on a wind-farm flow model capable of resolving the important physical drivers all the way down to turbine scale; the model should also be validated with relevant field observations.

Research continues into the characterisation of these unaccounted-for blockage effects, and pre-construction energy-assessment methodologies are being refined.

The wind industry is focused on quantifying blockage effects in the development phase, and on balancing the impact of all losses in the operational stage through wind-farm control methods, such as wake steering, and smarter operations.

DNV GL encourages partners from the industry to collaborate to better understand these effects throughout all phases of the project lifecycle.

Stefanie Bourne is business director for renewable energy project development at DNV GL - Energy. Ioannis Papadopoulos, Christiane Montavon and James Bleeg contributed to this report