Today the very big technical news for me is dual-doppler meteorology radar applied to create very high precision wind maps of operational wind farms. Brian Hirth of the Texas Tech University and his team have applied technology typically used for high-accuracy real time imaging of storm fronts to wind farms. Before I arrived in Vegas, I didn't expect that radar would turn out to be the most exciting story of the week, but at least from a technical perspective it is.
Yesterday was the excellent and under-reported news on the Lockheed Martin TPS-77 radar which through pencil-beam radar and smart digital signal processing can eliminate wind turbine farm clutter for military or civilian radars.
Doppler radar can detect moving air. Dual triangulated Doppler radar can detect moving air around a wind farm of 50 or more devices with an accuracy of nine to fifteen meters everywhere in the wind farm and in the airspace leading up to it in real time.
That degree of precision and real time data allows a number of use cases which previously were considered separately, were done badly or were impossible.
The big one is that this will allow smart automation to slightly sub-optimize upwind turbines so that their wakes miss downwind turbines completely. This is done by yawing the turbine out of square with the wind and pitching the blades so that the wake is nudged off to the side a bit and misses turbines completely, instead of slowing them and inducing mechanical stress. In the picture of the imaging, you can see the turbine in the upper left is in a bright wake band from an upwind turbine. A bit of tweaking and it won't be. This has the potential to raise existing wind farm output 5% to 10% by eliminating downstream wake interactions with upstream tweaking all day every day. That's huge. A modern 50 unit wind farm could increase revenue $2 - $4 million dollars annually just by tying these devices plus some sophisticated software to their existing pitch and yaw supervisory control and data acquisition (SCADA) system.
That's astounding but another use case is wind regime automation. In the Great Plains right now, there are winds called low-level jets. These occur mostly at night above thermal inversions. The wind above the inversion is often two to three times the velocity of the wind below the inversion. Because accessible wind energy is cubed as speed increases, that means that the energy in the wind above the inversion is eight to twenty-seven times greater.
Because wind turbines are now taller, they are starting to push the tips of the blades above the inversion layer into radically different wind conditions. If the stronger wind is from the same direction, the wind turbine gets a little boost and a little strain as the tips pass into the stronger wind. If the wind is from a different direction, interesting things start to happen. One of them is that instead of increasing output, the blade can stall instead not only reducing output but making a louder noise. Anecdotally I have heard of wind turbines being shut down due to this. According to the best study on the subject from the UK, blade tips above the inversion layer is the most likely cause of occasional thumps from wind turbines.
But if you know the actual direction and speed above the inversion and the speed and direction below the inversion, you can optimize the yaw and pitch of the turbine so that it gets the maximum from the blend of the two wind speeds. People are already doing this in a basic way by figuring out the roughly prevalent seasonal and hour-of-day inversions and directions, and then tuning the wind turbine a bit to improve outputs. It's guesswork because the wind meter is below the inversion 99% of the time. With real time monitoring of all wind speeds, tuning yaw and pitch to the actual wind conditions in real time is possible.
A third use case is blade stress under turbulence and specific conditions. Right now, blade pitch and yaw are estimations based on averaged wind speeds, directions, and turbulences. There are products that promise to improve this by putting LIDAR on blades then controlling pitch and yaw to the best compromise for the actual winds and turbulence hitting the blade. But if you have a painted picture of actual wind speed and turbulence for the entire wind farm already, you don't need to install devices on the turbines to figure that out. The value proposition is reduced blade and hub strain due to gusts and turbulence, so it's a cost avoidance measure, but that just means maximization of profit which is nothing to sneeze at. Existing products in this space will be unsaleable wherever Doppler is deployed; that destroys a niche market.
The fourth use case is noise management through precise tuning in specific conditions where noise occurs. Right now, there are occasional situations where neighbors are exposed to the thump instead of the usual swish and some get annoyed. However, the thump is maximized under inversion conditions in a plane from the turbine. Houses to the side will hear it, but if they are a little more or a little less to the side they won't. This allows precise direction, pitch and yaw control to optimize generation and minimize thump. It's a minor case, but it's big enough that a report of 800 or so pages was put together by the UK government after much study due to it.
The final use case is about complaints. When neighbors complain about wind turbine noise, consulting acousticians get engaged by someone to determine what is really going on. Right now the level of data is relatively low, but with this device in place very specific edge conditions can be discovered and adjusted for.
And all of this can be delivered incrementally through software upgrades after implementation. The first use case will deliver all of the payback and then incremental refinement will deliver another 1%-5% benefits through improvements over time. Early adopters will win, then win bigger, and late adopters will get a complex suite of performance tuners.
Of course, that's the adaptation model. But what about how wind farms are planned today? Current micro-wind maps are coarse-grained with models and met tower or two. But what if you parked a couple of these devices for a month and used the insights to nudge wind turbines around? And what if you could put wind turbines a lot closer together, as much as 1-2 rotor widths closer, because you knew you could use the doppler radar and software to fix wake interactions? And what if noise planning accounted for this technology allowing you to detune specifically loud and annoying conditions and as a result allowed devices to be closer to homes with a higher assurance that less annoying noise would reach the home? Combined with so-called low wind turbines which spread the potential economically viable generation area much more broadly, this would reduce siting concerns nearer to homes while substantially increasing generation.
The downside is that this sub-optimization might increase mechanical stress in some situations. As a result, the overall benefits will be slightly lower, but that can be modeled and figured out.
The potential is amazing. Brian Hirth would neither confirm nor deny when asked whether conversations with very major players in the industry who thought data and analytics were key drivers were under way.
The above is really, really good news. But there is bad news. Brace yourself: cave bats in the eastern United States will probably be mostly extinct in a few decades.
This has nothing to do with wind turbines and everything to do with white nose syndrome. It's a fungus that kills bats. Cave bats have central sleeping caves and distribute in a star-shape to places where they feed, so every cave has members that run into members of nearby caves with white nose syndrome. That's a vector of contagion that has a communicable chain. That communication chain has resulted in white nose syndrome spreading from a single cave in 2006 almost half-way across the USA by 2014. That's an epidemic that's going to peak in fewer rather than more years.
My background in public health tells me that in human populations facing disease we can vaccinate, isolate, quarantine and treat. In animal populations we can also cull, as with avian flu and mad cow disease. Most of the interventions aren't possible with hundreds of millions of bats living in caves. Culling could create firebreaks between caves but no one has been willing to be that draconian, and my moderately educated eye on the outbreak map today suggested that it is too late. And no vaccination or prophylactic has been figured out for several bat species dealing with a fungal infection.
It's unclear what this probable cave bat species die off means, except that cave bats are the most common type of bats so other bat species likely won't be able to pick up the slack in the food chain, so the ecosystem is going to be out of alignment somehow, either with too many crop destroying insects or with an unexpected predator proliferating.
The proviso of eastern USA is that there aren't many caves in central USA so there is a sort of natural firebreak that hopefully white nose syndrome won't jump. Culling of border caves might occur, but even there it's a fungus and they tend to survive better than bacteria and viruses, so accidental long-distance transmittal is more likely.
What does this have to do with wind turbines? They don't cause white nose syndrome, and as yesterday's update pointed out, they don't impact many cave bats. My inexpert judgment suggests that wind farms will have about a zero impact on the survival or non-survival of cave bat species given the overwhelming magnitude of the threat and lack of any interventions.
This will not prevent -- nor should it -- wind farms being subjected to strict environmental assessments and conditions related to cave bats. More money and time will be spent on this subject so that a space will be be opened up to allow smart people to find a reasonable solution that has a reasonable chance of success. And if it's just a delaying action that slows eventual extinction down, that is what it is.
To end on a high note, there is other good news from today. A major US offshore wind announcement was made by the DOE. Three firms -- Fishermen’s Energy, Principle Power and Dominion Virginia Power --, will receive about $47 million each to implement test offshore wind farms spread along the US eastern seaboard.
They will employ different innovative approaches to floating or embedding the wind turbine mast. It's a great step forward. Unlike the first two points, the pain and gain will be more delayed.
Net-net: expect a lot of bat-grief around much more productive wind farms in the next five years. Wind turbines are an ever optimizing peripheral player in a species-level extinction event.