Estimating the wind resource available around the world is a major component of the development of wind energy technology. Not only do we need to know how much wind there is to meet the energy demand, but we also need to know what specific locations have the most wind when compared to others. So what if I told you that the way many scientists and engineers do this estimation is inherently flawed?
It may surprise you to know that most wind resource estimations aren’t based on direct measurements, because we don’t have many. Most of our wind data come from surface measurements – airports, ocean weather buoys, even most satellite wind measurements are surface based. The problem is, wind turbines don’t depend on wind at the surface. They can be anywhere from 20 to 100 meters high, and the wind speed up there can be radically different than whatever we measure at the surface. Historically, we have assumed this change in wind speed with height to be based on a logarithmic boundary layer. That just means that the wind speed increases exponentially with height based on a “power law” equation
Where the U’s are wind speeds at height, Z2, and at reference height, Z1. If you don’t know anything about math, this is just indicating that the wind speed will increase the higher you go. We also have a generally accepted value for alpha, gained from many different experiments that is used almost universally in these types of estimations. In order to estimate the available wind resource, you would estimate the wind speed at the height of your turbine using a surface measurement, and use that to calculate the average amount of power at that location. It should stand to reason that an accurate estimate of this power depends greatly on that estimated wind speed being close to what’s actually going on at 80 meters above the ground.
The problem is that our atmosphere is not as simple as fluid flowing against a wall, and the wind speed profile is not accurately estimated by the power law as we typically use it. One of the main culprits is atmospheric stability.
You probably learned in middle school earth science that the air temperature changes as you move up in our atmosphere. In fact, in the lower parts of our atmosphere that we’re interested in, there can be distinct and quick changes in the atmospheric structure that create layers of air at different temperatures. After a hot day, the ground might release heat to the air, resulting in a layer of warmer air under a layer of colder air. You know that heat rises, so it makes sense that if you have a layer of colder air over a layer of warmer air, then there’s something funny going on. We call this type of configuration an “unstably stratified” atmosphere. If the reverse happens, and there’s a layer of warmer air over a layer of colder air, this is what’s known as a “stably stratified” atmosphere. But how does this affect wind speed?
In an unstable configuration, most of the movement in the wind is going to be upwards – that warm air wants to rise. In the stable configuration, where you have defined layers that are okay with where they’re at, most of the wind movement will be along those layers in a horizontal direction. The wind profiles tend to reflect that motion, see figure for clarification. A power law estimation neglects this difference and assumes the neutral case. You can see how this may lead to problems with extrapolating a wind speed up to height. In an unstably stratified atmosphere you would be overestimating your resource, and in a stably stratified atmosphere you would be underestimating your resource.
It’s clear that atmospheric stability is an influential part of extrapolating a wind speed at turbine height. Air pollution studies have utilized this concept for decades, but wind energy researchers seem reluctant to address it. It could mean that the potential energy available from wind has been underrated, especially over the ocean. A more accurate estimate might give offshore wind energy developers the financial incentive they need to push forward with this technology.