Skip to main content

Grand Challenges in the Science of Wind Energy (Text Version)

Following up on a 2019 article in Science, researchers from NREL and Denmark Technical University held a webinar to dive deeper into the subject of tackling the three grand challenges in wind energy science.

More than 90 participants attended the webinar, which provided an opportunity to discuss solutions to drive the innovation needed for wind to become one of the world's primary sources of low-cost electricity generation. The following is a transcript of that discussion.

Video begins with opening remarks and overview graphic of the grand challenges in wind energy.

Alexsandra Lemke: Hello, everyone, and welcome to today's webinar, "Grand Challenges in the Science of Wind Energy." My name is Alex Lemke, and I will be your moderator today.

Today's webinar is the first of a series of webinars were experts are given the opportunity to present some of NREL's core wind energy research capabilities, and we begin with a panel of wind energy science and technology leaders for a webinar on the challenges, opportunities, and interdisciplinary collaboration necessary to drive the innovation needed for continued wind energy expansion.

Back in 2017, a group of researchers and scientists from around the world documented the remaining grand challenges in wind energy research, highlighting the need to better understand atmospheric physics where taller turbines will operate, and materials and manufacturing constraints associated with the scale-up and the systems approach needed for plant optimization and grid integration.

Transition to slide showing speakers of the webinar.

So without further ado, I'd like to kick things off by welcoming out first speaker, Paul Veers, chief engineer, senior research fellow, and author of The Grand Challenges at the National Renewable Energy Laboratory. Over to you, Paul.

Paul Veers: I hope that by the time we're finished with this hour, you get a little bit of a sense of the challenge and of the interesting problems that are yet to solve in wind energy, and some of the promise that there is if we can do that. I'm going to share the time today with four other speakers. Julie Lundquist is going to talk about the atmospheric science and the challenges there. Amy Robertson is going to talk a little bit about the machines themselves, how we're trying to grow them to amazing sizes and scales and put them offshore in difficult to access locations.

Katherine Dykes will then talk about aggregating these machines into systems of generation plants and the problem of optimizing that plant and of using it in a way that is appropriate to the needs of the grid that it serves. And finally, at the end, Eric Lantz, also here at NREL, is going to talk about going beyond the physical sciences and looking at the difficulties and challenges that are out there for the social and environmental sciences, and how we actually get large deployments of wind energy within a social and environmental setting.

Slide transitions to graphic showing installed capacity and cost history of wind energy.

So with that, I'll look a little bit backwards. Many of you know where wind has come from. It's been a long process, and over the last 40 years, it's gone from a very expensive and almost undeployed energy technology to where it is today, where the costs have come down to less than $0.04 per kilowatt hour in a good site in the US, and the deployments are at large scales. We now in the US have over 100 gigawatts of wind energy deployed.

And the question that often comes up is that since we've made these great leaps forward, and there is deployment now at scales that are sizeable, essentially, are we done? Have we kind of solved all the difficult problems, and is there nothing left really to do other than keep pushing forward? It's been noted that the machines themselves have rested on a certain configuration of three blades. It operates upwind. They all tend to sit on a tubular tower. And that really, there's nothing left to do.

Transition to slide showing global wind energy capacity forecasts.

Well, that may be true if we want to stay where we are now, but I think the opportunity that presents itself for wind is not to be a small supplier. Right now, wind supplies five percent of the global electricity that we're using, but the International Renewable Energy Agency and other groups that look at such things have projected that if we can push the technology forward and continue to make progress, wind energy could be the world's top energy source by the middle of the century. Thirty-five percent is the goal that IRENA suggested. We think it could even go beyond that. But what that requires is that we get from where we are at about .6 terawatts of installed wind up to between 2 or maybe 6 terawatts in order to get to that position of being sort of a basis for our energy supply can come from wind.

Transition to a slide showing members of IEA Wind Grand Vision for wind energy meeting.

And if we want to do that, we're going to be walking into a whole new world. So to do the assessment of what it's going to take to get to ten times the wind penetration that we have now, as Alex said at the introduction, we convened the leading experts from around the world, hosted them at NREL, and asked that question. If we're going to achieve say half of our electricity supply from wind energy, what scientific issues do we have to solve, what scientific unknowns do we have to actually engage, in order to push the technology to that level?

Transition to slide showing explaining power electronic connections to the grid that differ than they do now.

The reality of getting to that level also has to deal with the fact that wind is not the only player that's going to be changing the way the energy infrastructure of the future operates. Solar energy is making similar gains, and is at this point accelerating even faster than wind. So the grid of the future is not going to be a large grid supplied by large rotating machinery from fossil plants. It's going to be a grid dominated by renewables that have power electronics connections to the grid that are different than we do now. And the systems are going to rely on wind energy and solar energy to be their basis. So wind will not be expected to be a small supplier of just bulk energy into a large grid that already exists. The energy component today is the major thing that wind energy supplies. But in the future, it's also going to provide capacity credit. It's also got to provide those systems services that allow the grid to operate in a reliable and resilient way. And that is an entire different reality that we now live in.

Transition to slide explaining steps towards realizing a future powered by wind and other renewable energies.

So if that's the reality that we have to deal with, the question is, how do we improve the wind systems that we have now, which means we can walk backwards from if we need to have system improvements, what are the innovations that are going to have to be generated to provide that systems improvement, and what research is going to have to be done at a fundamental level to enable that innovation?

The same is true of cost of energy. If the levelized cost of energy or LCOE is going to continue to come down, and it has to in order for wind to continue to grow, what are the innovations needed to do that, and what research has to be done to make that happen? Similarly, also, then, for the increase in value.

And these were the question that were laid to the experts' panel. The workshop was then summarized in a major report, which you can download. Katherine Dykes is the lead author on this report, and it details all of the issues that were brought up by the workshop. It's a comprehensive document, and it's quite long. It's 160 pages. I would recommend that you take a look at this if you're interested in these challenges.

But then we felt there was also a need to take this word outside the wind community and present it to the larger scientific community and make the point that wind is not done, there are really interesting and challenging scientific issues to be researched in order to make this go forward.

Transition to slide discussing science article published October, 2019.

So we took a look at just the research portion of this and condensed that into a paper that with – many of the participants of the workshop contributed as authors, and we published this paper in Science last October. There was a long negotiation with the editors at Science about what this could contain, and we really pressed on them the fact that although they would like to see a very simple and focused thing about one scientific issue, that the wind energy enterprise is one that is so integrated that you can't really make progress in one area without impacting all the others. And in order to actually express the need for scientific progress, we have to actually talk about all these areas.

Transition to slide asking the question, “What are the science challenges for researchers that enable wind energy to achieve 50% or more of the global electricity supply?”

So when we ask the question what are the challenges and really set up this goal of something on the order of 50 percent or more of global electricity supply, we came up with an approach that said, we can condense these areas into three basic areas of unknowns, three areas where we're pushing the limits of what's already known in the scientific community with our wind energy technology.

Transition to slide discussing three challenges in wind energy science.

The first of these is the physics of the atmospheric flow in the critical zone where wind plants operate. The sizes and scales of the machines now being developed, and the sizes and scales of plants we're trying to optimize and control are in a zone of atmospheric science that has not been a particular target of study. Although a lot is known about it, there are many unknowns we need to solve.

The machines themselves have grown to the scale where assumptions we used about their design and how we actually designed these machines 20 years ago or 30 years ago are questionable at this point. We're pushing into zones of the unknown simply because of the scale and location of where we put these machines, and Amy is going to talk to us a little bit more about that, especially when we go offshore.

And finally, an area of unknown is what this grid is going to look like in the future and how it is we're going to be able to take hundreds of independent generating agents and organize them and control them in a way that both optimizes their performance and supports the grid. And Katherine is going to talk to us a little bit more about that. But I'll just go over these at a high level and give you a little background before we turn it over to Julie and Amy and Katherine.

Graphic displaying the large spatiotemporal scales that wind energy covers from global and regional scales, all the way down to nanometer measurements of how air interacts with turbine blades, all taking into account varying time frames.

One of the things that is also fascinating about the wind energy problem is that the scales that it spans are just enormous. The energy source of the fuel for this system that we're building comes from the global atmosphere and its global circulations. Moving air is generated by uneven heating, by Coriolis forces, the rotation of the Earth, and so on. And these scales then cascade through the wind plant all the way down to the tiny turbulence that intersects with the actual surface of the blade, and to do the aerodynamics right, we have to get down to the millimeter, and less thickness of those boundary layers on the airfoil. And within the machine, we're dealing with the mechanics of materials that have to be manufactured on the 100 meter scale, and yet their properties are defined at the nanometer.

At the same time, we also have temporal scales where the weather systems are seasonal. The planning that we do for what resource is going to be available and how much we need to build and how much energy we're going to get is decades in length. So the long scales are critically important to making this an economic enterprise, and yet we have to deal with those tiny time scales of minute by minute balancing of the grid and microsecond kinds of faults and disturbances that require the grid to ride through and be stable and continue to operate.

So it is really kind of an amazing both temporal and spatial range of scales that we have to deal with.

Graphic displaying flows of wind across varying geospatial scales.

Now the global system are those things driven by these circulations that flow. This is a map of water vapor over about a week timeframe, and you can kind of see, although this is not wind itself, but it does show that the flows that we have that drive our wind systems are global in nature. In order to actually do weather modeling, generally, a smaller scale will be fleshed out, and you'll look at what are called the mesoscale models, where we can actually map individual weather systems, and we can look at the kinds of dynamics of the flow at a local level.

But the wind plants then have to actually take this mesoscale modeling and look at how the flows are going to actually impact with all those individual turbines, and how those individual turbines interact with each other. And if you get down and you drill down even lower, the scales of turbulence in the wind flow are going to impact the individual machine and determine its loads and the dynamics of its generation.

Now interestingly, each of these areas of modeling are fairly well developed and understood. One of the biggest challenges we have is making the transition from one scale to the other and getting it a consistent kind of energy flow that really represents the physics of what's going on in the atmosphere.

The other challenge is that not only do these cascade downward, but the individual machine impacts the plant, the plant impacts the regional flows, and those regional flows, again, can potentially have some sort of impact on the global flows.

Graphic displaying the first grand challenge, how to master the physics of the wind resource from the atmosphere.

So we're trying to package all of this, and that's where we came up with this first grand challenge, which Julie will flesh out in more detail. But really, we have these different scales. One of the elements that has been described as the terra incognita is the fact that there is a scale level that in the mesoscale models where we look at weather, it's a little bit smaller than we would ever resolve the individual dynamics of that flow, so we resolve it on average. But in the microscale flow where the wind plants operate, that is the scale that is important and critical to the flow around the machine, and making that represent the appropriate mesoscale flow has been one of the challenges among many that we deal with in the atmospheric science, and Julie will talk about that more.

Video of wind flows at an offshore wind plant demonstrating wind turbine wakes and how they impact power production of wind turbines downstream.

When we try to get from the atmospheric science down to the individual machines, it's useful to look at this video, which shows us coming down from a very high level to an offshore wind plant. This is off the coast – a proposed wind plant off the coast of New Jersey. And it is – shows us mapped in color, if we look at a horizontal plane and look at the wind speed, where red is high speed and blue is low speed. As the flow, which is now very irregular on a microscale, impacts each machine, the machine extracts energy from that flow, and you see behind it a low speed zone of wind, which is the wake of the machine.

In certain orientations and occasionally one machine's wake will impinge on the machine downstream, and that changes the flow characteristics completely in terms of the dynamics of production and also the loads in that machine. So to understand this well and optimize it, we have to be able to understand these wakes, which are meandering left to right, and as this shows, it's a three-dimensional problem as well. Those wakes can go up in the air as well as coming down.

Transition to a slide showing the structural challenges that need to be resolved as a part of the second grand challenge in wind energy science regarding scales and aerodynamics.

So that is going to impact how we deal with the machines themselves. And the second grand challenge really is about the size and scale of these machines, and the fact that the structural dynamics and the aerodynamics of the machines of the past are becoming more and more challenging. The machines that we built 30 years ago, if they were simply scales up to what we have now, those blades and those rotors would be ten times heavier than what we have now.

So over time, we've managed to remove 90 percent of the weight from it, but that means that it's more flexible, it's lighter, and it's more susceptible to instabilities than it would have been, and we need to actually deal with them.

Transition to slide with an animated turbine talking about exascale computational efforts to understand physics at play.

A lot of the challenges that we've looked at in terms of the aerodynamics also are saying that the scale of the blades that we now have are interacting with the turbulence of the atmosphere in ways that change the actual lift and drag in ways that are not necessarily easy to predict, so we're creating exascale type computational models that will allow us to actually solve for the detailed flows right near the blade in realistic conditions where we might have multiple machines.

Transition to slide on the third grand challenge in wind energy science, discussing how each of the wind turbines interact appropriately with the power grid.

And then finally that leads us to our third grand challenge, which is really about taking all of these individual machines and packaging them and controlling them in such a way that we can interact with the grid appropriately.

Transition to an animation showing wind plant simulation with realistic atmospheric flow.

So this is a movie that we've made some years ago, but it does represent this fusion of information from the atmospheric science area, and creating a turbulence resolved flow over natural wind plant. This is a land-based wind plant in Northern Colorado. The red and blue are, again, the higher and lower speed wind speeds. And we're looking at control of that plant as this flows comes over it in a way that we can optimize that plant with actual realistic weather conditions coming across it machine by machine.

Graphic showing power grid simulations across the United States with different amounts of renewable penetration on the grid.

And where we need to get to is from individual plants then to the national system, and NREL has done some amazing work in simulating the actual grid with high penetration wind and solar. This is an example. We've taken a week in history where we have the actual weather conditions for that time period. We actually have the grid loads during that same time period. And we can look at day to night variations. The line coming across here, this now is sunrise coming in, and you can see all of the solar plants come on. And when those solar plants come on, that causes the need to push that energy from where it's being generated to where the demand is, and the flows that have to match that. And then when sunset comes back in, you see the solar goes away. A lot of the dark blue are the wind plants, which tend to pick up a little more at night, and those flows then have to deal with that.

We'd like to get from this example, which was about a 35 percent penetration, up to these very high 75, 80, 90 percent renewable penetrations, and look at how the grid is going to operate in that kind of a stress situation.

Transition to slide that states thank you! A series of visuals show on screen while Julie takes over the presentation.

So with that, I'd like to turn my talk over to our first speaker to get into the atmospheric science grand challenge. Julie Lundquist is going to talk about that. And Julie, I'll let you introduce yourself.

Transition to title image of Grand Challenges in Wind Energy Science.

Julie Lundquist: Okay. Thank you very much, Paul, and I will try to take control of the screen and have it project. So please let me know if you guys can see my screen. Whoops, that is not the slide that should be showing. So Paul, can you see my screen?

Paul Veers: Yes, I can.

Julie Lundquist: Okay. Great. So let me come back here, and is it in presentation mode or is it in presenter view?

Paul Veers: It is not in presentation – there you go. You've got it now.

Julie Lundquist: Excellent. Okay. So it's interesting that some of the challenges of – grand challenges in wind energy seem to often be related to our collaboration software. But I'm very happy to have the opportunity to talk about atmospheric science today. Those of you who know me know that I'm a professor of atmospheric science at the University of Colorado Boulder. I'm actually in a department with atmospheric sciences and oceanographers, so there's a lot of potential collaboration between different parts of NREL and different parts of CU. And I've been in my role as a joint appointee at NREL since 2010, and I feel like half of my job is often translating what atmospheric sciences think about the world into what wind turbine design engineers think about the world, and so I'm really happy to be in that role.

A couple of older photographs of meteorologists is displayed.

So when Paul asked me to put this presentation together, he had started with a question of if you guys have been working on weather forecasting models for so long, why don't we know everything perfectly about how we need to predict wind as it goes into wind plant? And so I just wanted to point out that weather forecasting has evolved considerably over the last 100 years. If we look even 70 years ago, in 1950, a lot of weather forecasting work was done on large tables of printed out maps, and people drawing lines by hand to try to understand the movements of weather systems, and then move from that larger idea of weather systems into thinking about specifically how those systems would affect winds and temperature.

Transition to slide displaying computers in a lab.

And so now we obviously have a lot of computational tools that allow us to assimilate data collected on the ground in situ as well as data collected from satellites above, and all of that data gets assimilated into refined forecasting models that still have a human component to them, but involve a lot of machine learning, data simulation, and those types of things.

Transition to a visual displaying global wind circulation.

But in order to assimilate those measurements and observations, we have to think about the larger physical system, and I was very happy to hear that Paul had mentioned the Coriolis effect and things like that. This is an image that we like to talk about in our introductory weather classes, where we think about the large scale weather patterns on our planet Earth, and we think about on an averaged annual cycle.

And so you can see that if you care about winds, then you tend to find westerly winds in the mid latitudes, between 30 degrees north and 60 degrees north. Similarly, between 30 degrees south and 60 degrees south you also have westerlies. But what we need for wind energy purposes are much more refined forecasts than just this global picture.

Transition to a slide displaying equations and functions to aid with forecasting on a global grid.

And so instead, we take equations of motion, and we take these equations, some of which are presented here – there are many other equations that go into these models. We put these models onto a grid. So we slice up the atmosphere in the horizontal, we slice up the atmosphere in the vertical, and then we make physical parameterizations for the processes that cannot be represented explicitly on that grid.

So embedded within each grid cell, we probably have parameterizations for cloud development, for how an individual beam of solar radiation will be interacting with the aerosol particles in the atmosphere, with cloud particles in the atmosphere, with different chemical constituents in the atmosphere, and so on.

Transition to a slide displaying a wind turbine alongside images of global wind energy systems.

So weather forecasting is very sophisticated, but the problem for wind energy and for other renewables, for that matter, is that we need very specific information. We care a lot in the wind energy world about winds and turbulence between bottom of the turbine rotor disc and the top of the rotor disc. So right now, that tends to be around 3,200 meters above the surface, and that's because turbines of course are integrating winds and turbulence at those altitudes.

However, forecast models were actually designed over the last 70, 120 years or so to focus on predicting surface temperature, precipitation at the surface, and then the movements of those large-scale weather systems. So we think about that in terms of low pressure systems, high pressure systems, and the typical quantitative we look at is the height of the 500 millibar surface. So these are very different goals, and I'm sure that you guys all know from your work that if you design a system for one goal and then need it to do something else, then new changes and new research is required in order to make that work.

Transition to a slide displaying video of wind-plant scale phenomena than is traditionally measured.

And so now we're trying to couple the weather scale phenomena, the mesoscale phenomena, the movements of frontal passages, and that type of stuff, we're trying to couple that with the wind plant scale phenomena that we know a lot about, but tends to happen at much shorter and faster timescales, shorter spatial scales and faster timescales, than what we see.

So I think that Paul showed this animation before, and each of these individual vectors represents the direction of the wind as it flows through this particular wind plant. And I like this image because it shows that at the initial time, we have southeasterly flow, but then at frontal passage comes through from the north, and it changes the directions of the wakes, it changes the power produced by each of those individual turbines, and it changes the loads on those individual turbines.

So if we want to successfully model wind plants, simulate wind plants, predict wind plants, and then control wind plants, we have to be able to capture that larger scale forcing, which is not stationary, which is not averaged out, it includes time varying phenomena, like frontal passages and thunderstorm outflows and things like that, so that we can make these predictions accurately. Okay, now I'm going to try to move on to the next slide.

Transition to a slide displaying wind plant impacts on one another.

So similarly, we can move upscale and think about the interactions of individual wind plants with each other, because if we're going to be integrating large parts of – or large amounts of renewably generated electricity into the grid, we need to think about the effects of wind plants on each other.

So in this plot here on the left, we actually have three wind plants, one down here that I'm circling right now, and then there's another wind plant on the northeast side, and another wind plant on the southwest side. And what we were trying to study in this particular case was to look at how the upwind wind plant and how the wake that that generates affects the downwind wind plant.

And so if you looked at the grand challenges paper in Science, you'll notice that we have a couple of extensive paragraphs talking about inter-array effects of one wind plant on another. And so here, the wind barbs are just showing the wind direction. The colors are representing the wind speed deficit introduced by the presence of this upwind wind plant on the southwest side. And when we go – we can also look at this at the same time in terms of the power production at the downwind wind plant. And the darker, angrier colors in the panel on the right show the wind speed deficit experienced at the downwind wind plant because of the wake from the upwind wind plant.

And this is basically one month of simulations for one particular year that we chose, not because it was the absolute worst case scenario, but because it was a challenging scenario. And we can quantify the impact on the downwind wind plant to see that it lost a significant percentage of its capacity factor due to the waking effect of the upwind wind plant.

So this sounds like a problem, but if we have a good understanding of the atmospheric science behind this, and we understand how wind speed, wind direction, and the ability of the atmosphere or the stratification of the atmosphere actually effects this, then we can understand when we can predict those strongest power losses, those strongest wake effect losses.

Transition to a slide showing power loss events and how they are predictable.

And so what we have here is a plot where we're looking at on the X axis the stratification of the atmosphere. On the right side of zero we basically have daytime situations. On the left side of zero we have nighttime situations. And then, of course, the Y axis is the inflow wind speed.

And I want to draw your attention to when and where we see these large dots. So each dot represents the total amount of power lost at the downwind wind plant for any individual hour over the course of the month. And you can see that if we pay attention to the large dots, those large dots are mostly occurring at night and mostly occurring at wind speeds just less than rated for this particular set of turbines.

So atmospheric effects are very important. They can be extremely consequential. But if we understand how the atmosphere works, we can actually predict those events, and if we can predict them, then we can manage them, from the grid perspective or from property owner perspective, and so on.

Okay, so if we want to think about future progress, there's a lot that we still don't know about the atmosphere, and this animation here on the left is basically going to show you an image of what we know about oceans, because if we think about the marine boundary layer, which Amy will be talking about, or she'll be talking about offshore technology, but that is embedded in the marine boundary layer, there's a lot that we need to learn about how that behaves and how that interacts with the atmosphere.

Transition to a slide with a video and a static image of offshore wind turbines showing demands for more atmospheric research.

So if we look at the ocean and we see how the surface temperature of the ocean varies, so here we can see the gulfstream, this nice red river, essentially, coming up off the East coast of the United States, those dramatic changes in sea surface temperature have a strong effect on the winds and the marine boundary layer, where a lot of offshore wind will hopefully be deployed soon.

So I think that I should wrap up here and hand it over to Amy Robertson. And just I would like to remind you that more research is required. There are a lot of fantastic open questions in offshore wind energy as well as onshore wind energy, specifically in atmosphere interactions. All right, Amy.

Amy Robertson: Okay. Thanks, Julie. Can everyone – can you see my slide, Julie?

Paul Veers: You're good.

Julie Lundquist: Yes, I can see your slide beautifully.

Transition to overview slide on grand challenges.

Amy Robertson: Okay. Thank you. So as Julie said, I'm Amy Robertson, and I work at the National Renewable Energy Laboratory, and I'm going to be focusing on the technology grand challenge, and specifically related to my experience with floating offshore wind.

Transition to a slide showing characterization of offshore wind resource in the United States.

So before we start talking about the challenges associated with offshore wind, I thought it'd be useful to first talk about some of the reasons we want to pursue this technology. So offshore wind provides a significant contribution to the resource potential of the US, as you can see in this picture here of the resource characterization for the US at 100 meter hub height.

The winds offshore also tend to be in general stronger. They're more consistent, and usually have less turbulence associated with them, due to the smooth sea surface that they're flowing over. As with land-based wind, significant cost savings can come from increasing the size of wind turbines, and offshore, we actually have some unique advantages that land-based wind doesn't have that is related to transportation constraints. With offshore wind, you can actually build a wind turbine just at a port just offshore, and then be able to tow out your system to its location. So no transportation necessarily needed over land, which constrains land-based size.

In addition, offshore wind is located closer to population centers on the coastlines, where most of the population lives, and where the energy is needed.

Transition to a slide displaying fixed and offshore wind systems.

So offshore wind is already being done to a large scale in the North Sea and in Europe, and they're using what we call fixed bottom offshore wind systems. These systems mainly look like what we think of as land-based systems. They're monopiles where we have the tower of the wind turbine just being extended into the sea floor to attach it to the sea floor.

So that's the majority of what we're seeing today. As we start to move into larger or deeper water depths, we started seeing larger support structure, like these four-legged jackets, as we need to have a stronger support structure for that deeper water depth. But really, to get the contribution from offshore wind resource that we want to have, we need to move into even deeper water depths, and as we move deeper, it becomes not economical anymore to try to fix a offshore wind turbine to the sea floor. The structure just gets too large and gets too long and flexible.

So instead, we move to what we call these flowing wind systems, where the support structure floats on the water, and is kept in place through mooring lines that are attached via anchors to the sea floor instead.

So floating wind right now is an emerging market. We only have a few pilot wind farms in this area, and there's several challenges that we need to overcome in order to get the cost to the level to make it really a commercial reality. And so here, we're going to be talking about some of the scientific challenges that we need to address.

Transition to a slide displaying video of validation of offshore wind technologies, explaining how these structures differ from other systems because of coupled problem because of complex dynamic characteristics of floating wind systems.

So what we're doing with floating wind is really bringing together two different markets. We're bringing together offshore structures and wind turbines. So we have had modeling approaches to design each of these systems, but now we're when bringing them together, the same theories or same approaches we used to design and model these systems no longer apply, and that's because we're really getting into these unique dynamic characteristics that these systems have that really no other systems in the world presently have.

So if we look at this video – hopefully it'll run – of this scale model test of a floating wind turbine, you can see that the waves acting on the structure are going to excite the system, as is the wind. And since you have this structure that's being excited both by the wind and the waves, those two excitations become coupled to one another. So the way the wind is interacting with the structure is being affected by how the wave is interacting with the structure, and vice versa.

So we have this coupled problem where the dynamic characteristics of the system are different from what we've traditional seen, and the models that we need to use to design and innovate these systems needs to be modified based on these dynamics. So to really design these innovative floating wind systems that we need to achieve the cost levels for commercialization, we need to fundamentally understand the dynamic characteristics of these systems and be able to model them accurately.

And I do have a video, a short video in here also of a full-scale wind turbine. I just wanted to – that's my own video, which is why it's not so good. I just wanted to emphasize that this dynamic characteristic we're looking at here for a scale turbine is sped up and for an extreme condition, so it may look a little bit larger than what you see in a real system, because of that speed up and looking at extreme conditions. But it helps to show those dynamic characteristics that we're interested in.

Transition to a slide displaying aerodynamic challenges of floating wind turbines.

Okay, so digging into some of the specific aspects, when we look at the aerodynamic challenges of these systems, as you saw, we have this increased motion. So therefore, for the design of the systems offshore, we need to make sure that we have a turbine that can work with this large motion. So if you think of historical land-based systems, for instance, that use a gear type drive train, that is no longer viable offshore, when you start having a lot of motion. So that's one area.

We also need to understand and develop a control system that does not destabilize the system. So if we have a system moving around, and you can think as you start pitching your blades, that's going to actually create interaction with the dynamics into the system. And when we first start putting wind turbines on floating support conditions, that's one of the first issues that we encountered, is if you simply put a land-based control system on a floating wind turbine, it creates a destabilizing or negative damping type issue where your system will essentially flip over. So we need to make sure we can design control systems that work with those dynamics.

And in fact, we actually have an opportunity there, not just a negative, but an opportunity. Since those control systems can interact with the dynamics, it can actually help to better stabilize the system than without it. And that could actually lead us to even more optimized designs that use less support structure steel and can lower costs.

As we said before, one of the ways we will lower costs is to have these ever-increasing size of these machines, and as you can see in this upper right picture here, we're moving to even larger systems offshore. Again, because of the lack of constraints with transportation, we see these now proposed 50 megawatt machines offshore, so really large.

But as Paul was alluding to, when you start moving to these larger machines, the way they're interacting with the air is very different than these smaller machines. And historically, the models that we've been using to design these systems are based on the aerodynamics of these smaller machines. So we need to start adapting those for these new large designs that we want to build.

In addition, the added motion that we've been looking at will mean that the turbine is now moving in and out of its own wake. That means that the traditional theories such as BEM are no longer valid, and it also means that we're going to have increased interaction with the blades with its own vorticity.

So we do have projects that are trying to address these issues. And so you can see on the lower right here the IEA Wind OC6 project, phase III, we're looking at using measurements from a wind tunnel with a robotic excitation system that's going to move that wind turbine like it's moving in the waves to look closely at the aerodynamic models and whether we can accurately predict how that wind turbine responds when it's on a floating foundation.

Transition to a slide discussing hydrodynamic challenges of floating offshore wind with a semisubmersible image and chart displaying wind/wave load cases.

Next, looking at the hydrodynamic side, we see that the hydrodynamics differ from oil and gas in several ways. First, we have a wind turbine now on top of a floating structure. That turbine is going to create this high center of gravity, and there's going to be a large thrust force trying to push that wind turbine over. Therefore, we're going to have to design a fundamentally different support structure that can resist that overturning movement while being cost effective.

With offshore wind, we're also moving into shallower water depths, which mean we need to have novel mooring systems that can be designed for those shallow depths.

We also have different design drivers. For oil and gas, it's all about designing to the extreme conditions. For offshore wind, we have some of our largest loads coming when the turbine's operating. So we have this mix of design drivers from operational and extreme conditions that we need to consider.

The risks are also severely different for offshore wind that oil and gas. There's no real risk – high risk when it comes to wind turbines. If we lose a wind turbine, there's no risk of loss of life. There's no large environmental risk of creating large barrels of oil going into the water. We just simply lose the system. So the constraint for offshore wind design really isn't that high risk scenario, but rather trying to find a cost effective solution. So we really need to adapt the methods that we use to design these systems to address the right level of risk that we have in this industry, and not simply keep using the oil and gas approach to designing these systems.

On the modeling side, for these floating structure, the size of these structure are very different than the large oil and gas structure you see. We have different dynamic characteristics. And we find that we sit in actually this kind of no man's land of appropriate hydrodynamic theory. In some of our past work, we found that our hydrodynamic models aren't predicting the loads on these systems because of these issues to the accuracy level we'd like. We've seen that in OC6 phase I we have about a 20 percent under-prediction of the loads across different conditions of these floating wind systems. And for some industries, say again oil and gas, you may be able to live with that 20 percent under-prediction because you have this large safety margin due to the associated risk, but for floating wind, we need to live to these more optimized cost optimal solutions, and therefore we need to eliminate this level of uncertainty and get to more accurate solutions.

There's another – a number of hydrodynamic components that we need to consider, including combined nonlinear and irregularity of sea states, breaking waves, viscous effects, etcetera. And also, as we go to these more optimized designs, we're going to be having more flexible structure, and we need to get the right hydrodynamic theory for those as well.

Transition to a slide showing some land-based wind turbines and a satellite view of a hurricane.

Finally, we also have challenges regarding the metocean representation, which is the wind/wave climate for these offshore structure. As Julie was mentioning, when we go offshore, the wind no longer has the same characteristics as on land, and we have limited measurements right now for what that behavior looks like.

We do have wave measurements, but we don't have that simultaneous wind/wave behavior in combination, and it would be great – that's an outstanding need, to better understand what those characteristic are offshore.

And with that lack of information and these different dynamics, the wakes themselves that Julie was also showing is going to behave differently for these floating wind systems than they would on land or even for a fixed bottom system. And we also need to consider things like these extreme conditions, such as hurricanes and breaking waves in the US, sea ice floating, and start understanding how the wind and wave interact together and change the loading behavior.

Transition to thank you slide.

So I think that's it. That's where I'll end. Thank you. I'm sorry. I think I went a little bit over. But I do want to introduce the next speaker, which is Katherine Dykes. And she's going to be talking about the optimization and control of fleets of wind plants working together individually and synergistically as a whole system. Thank you.

Katherine Dykes: Thank you, Amy. I'm going to try to now put my presentation in full screen mode. Is the right screen showing, hopefully?

Amy Robertson: Yep, it's good.

Paul Veers: You're good.

Transition to title slide on third grand challenge of wind energy.

Katherine Dykes: Thank you. Okay. So yeah, I see a lot of familiar names on the participant list. Nice to virtually meet up with you guys and be part of this webinar on the Science article on grand challenges. I'm certainly not the best person to present this grand challenge three, but I'm very happy to be part of it.

So for those of you who don't know me, I was at NREL for about nine years, and last year moved to DTU Wind Energy, where I lead the section on loads and control that does air elasticity and controls research, but also turbine and plant optimization research.

So grand challenge three is all about everything from kind of looking at plant level and control all the way to how do we actually support the electricity grid with the wind power plants of the future, and what are the research needs there.

Transition to a slide with charts showing possible changes to the grid system and markets.

So probably many of you in the audience are actually much more knowledgeable on the larger system integration and grid technology and science than myself, but for context, really, as Paul spoke to in his presentation, we had kind of two things that were really motivating thinking about the research needs in wind energy, and one was this traditional LCOE driven perspective, which aligns very well with the historical space for wind energy, which largely had power purchase agreements or feed-in tariffs, and so really just the goal was to generate as many electrons as possible at the lowest possible cost.

But as wind starts to participate in these other markets, whether it's markets for energy or whether it's capacity markets, how much wind is in that system actually has an inverse correlation with the types of revenue that a wind power project might see. And so this is kind of the future outlook, so nice graphic from MISO from 2015 showing how the capacity credit and potentially then the capacity value or what wind could achieve in revenue in a forward capacity market decreases with the amount of wind in the system. A snapshot of PGM just through 2017 showing that the trends with more renewables, variable renewables in the system, especially wind, is that the revenue share from energy or capacity markets are starting to come closer together.

And so Paul spoke to this. In the future, there's speculation about how will the future markets look. Will we have an energy-dominant market like we do today, or will it be capacity-denominated? Even so, even in an energy market, will a lot of the revenue come from times of the year when – or the day or the hours where wind is not producing as much? So there's a lot of concern, industry is quite concerned about how do we make sure that wind power plants are profitable as we look to the future.

Transition to a slide showing conceptual graphic with wind turbines more closely tied into grid system.

And then, of course, not just from a profitability standpoint, but of course, there are bigger issues that a lot of the grid integration community is thinking about, is as we reduce physical inertia in the system, as we have more and more wind turbine brains in the system, can we maintain the reliability and the stability of the grid that we require? And there's a lot of speculation and a lot of ongoing discussion about may in some ways such a system, kind of based on brains and power electronics, may even be better. But these are of course outstanding research topics.

Transition to a slide of chart showing how wind needs to support the grid at different time scales.

And the other point then is if we're going to have this system where wind is supporting the overall grid, it needs to be supporting it in all these different timescales, so going from very short term stability to operations and planning.

Transition to a slide that spells out the third grand challenge of wind energy.

So that kind of motivates the grand challenge three, which is all about how in this large variable renewable grid system, based grid system, how do we actually orchestrate all of these wind plants to perform in the best possible way to provide low-cost energy, stability, resiliency, reliability, and affordability in a future power system.

Transition to a slide showing the key issues related to the third grand challenge in wind energy.

So that was the background, and then we broke things down into a set of more specific issues. So there's a whole lot of research, and NREL has been very active in this space, around advanced plant control, so moving beyond every turbine focusing on its own control schemes, but looking at coordinating controls at a plant level, and doing that not just to maximize power, but also do other things as well, which I'll talk about.

Also, looking at the whole shift from this traditional fossil fuel-based system with lots of massive inertia to this system of the future, again, with a lot of these brains and power electronics kind of providing the backbone of the grid system.

And yeah, even going to the level of forming the grid with a largely inverter-dominated system. Yeah, and following from that, wind can provide these services, but to what extent? And how far can we push things?

And underlying all of this, one key thing that came up across all the challenges, but was really prominent when we talk about grand challenge three, is the importance of data analytics, and how we can – in order to do what we need to do, we really need to be smart about how we're using the data and all the data science activities.

Transition to a slide discussing wind farm controls and how to improve wind power energy through coordinated efforts and reduce loading on turbines.

So, a little bit on wind farm control. So this is a very, very active area of research which NREL is very involved with, and part of a world-leading group of researchers that are really pushing the envelope on this, and looking at how we can improve power and energy production from wind farms through coordinated wind farm control, how we can reduce the loading across the turbines and distribute the loading to improve overall reliability, and more and more, how can we actually provide grid services through these coordinated wind farm control activities.

Transition to a slide explaining wake steering with a flow dynamic graphic.

So one example that's gotten a lot of press and a lot of traction in particular at NREL, Paul Fleming, Jennifer King, and others are still doing a lot of work in this space, looking at wake steering. How can we increase power production by deflecting the low energy wakes away from downstream turbines? And that work has been going on for some time, and we're starting to see field demonstrations of that in the commercial scale, but there's still a huge amount of research still to do on that topic and related wind plant controls topic.

And now we're trying to think about, okay, that's the flow dynamics and the flow control side of things. What about the grid control, and how can we marry these things? And I don't have, unfortunately, the video working for this, but some very nice work actually done already some years ago now at NREL with Paul Fleming and _____ where they were actually coupling the flow modeling side with the dynamic grid simulation side to look at how you can actually do active power control at a farm level. And so another area where there's some things going on, but there's a whole lot more we need to do.

Transition to a slide discussing how to move to 100% renewables with variable renewable inverter based technologies.

And then the kind of broader wind grid integration community aligned with ESIG as well as IEA Wind Task 25, has really been looking at pushing the envelope at how can we move towards 100 percent renewable systems with variable inverter-based renewable energy sources as the basis. So looking across different cases and what are the challenges there, and this area has been really developing quite a bit, and there's going to be a whole session on this at the upcoming wind grid integration workshop in October that Hannele is coordinating. And I think there's also some papers that just came out or are about to come out also led by – in collaboration with colleagues at NREL as well.

Transition to a slide showing Cerberus dog of three challenges to be tackled.

But looking at how to challenge some big – some big remaining challenges with this variable inverter renewable energy based system, stability, balancing, flexibility, and adequacy, and market operation. So how do we address all of these challenges as we move to more and more variable renewable energy generation in the grid system?

Transition to a slide discussing digitization and data mining needed to inform future developments.

And then, yeah, coming back to finally the data, digitalization, data-driven modeling, massive amounts of data of course involved just at the wind farm or even wind turbine level. Now we're talking about expanding that to the grid level. How can digitalization really help us realize the full potential of these high renewable systems? So there's a lot of, of course, work, which I feel like we're really just scratching the surface of on this front.

Transition to a slide explaining next steps in this arena.

So in closing, future grid with large amounts of variable renewable energy generation faces new challenges. I'm sure that's not a surprise to this group. There's a lot we're already doing with advanced controls, but there's a lot more that we can do and need to do on this space, especially when it comes to the grid side of things. And we really need new tools, models, technology development, and more research, and far greater interaction than we've ever seen before between the wind – traditional wind community and the grid community. And I think NREL is really at the forefront of that, and that's great, and hopefully, we'll help to solve some of these critical challenges we face.

So I think that's my last slide, so I, again, I think we're running super late on time, so I'll quickly turn it over to Eric so he can queue up his presentation.

Transition to thank you slide.

Transition to Eric Lantz’s slide deck and title slide.

Eric Lantz: Great. Thanks, Katherine. I'm assuming folks can see my slides, but someone should let me know if that's not the case.

Paul Veers: It looks good, Eric.

Eric Lantz: Excellent. Thank you, Paul. Yeah, so it's my pleasure to kind of round out this panel by extending the grand challenges that we wrote about in the Science article beyond those disciplines that were the focus there.

Transition to a slide from WindToolKit showing 24 hour wind flow in North America.

My talk is really going to look at how the wind energy community also needs to devote time and attention and mental cycles to engaging with social scientists and biological scientists. I like to start off with this particular graphic which shows sort of the immense wind resource that we have that we're really blessed with here in North America and in the United States in particular. It is a variable resource, but it's really quite continuous. The wind truly is blowing somewhere.

And it was interesting when we first published the Science article, and Paul and others did some briefings for the press community, one of the main questions we got was do we actually have enough energy to do this from the wind. And it's kind of funny, because those of us from within the wind energy space, we know that's actually a really easy question to answer from a physical science perspective. The quantity of resource is really, truly tremendous, and there's plenty of it there. In the US alone, the simple math is something on the order of 20 terawatts of wind resource potential in terms of power. Today, we have a little over a terawatt of installed and operating capacity to serve our current electricity system needs. Capacity factors are a little bit – they factor into that, too, but the short answer is we have plenty of resource.

But from my view, at least, you know you've got somebody that's paying attention when they pose the question of can we actually place enough wind turbines either on land or offshore in order to get to the 50 percent penetration levels that are referenced in the Science article? Then you know you've got somebody that's really paying attention, and that's what I'll focus on for the rest of my talk here this morning.

Transition to a slide discussing key biological and social science themes for wider wind energy deployment.

So with that in mind, there's really four key themes to the presentation at the moment. One, the cost and benefits of the deployment of wind energy are not evenly distributed. Quite understandably, this can become a bit of a source of friction among many stakeholders that are impacted by wind energy.

Second, to achieve a high penetration wind future, the interface between wind energy, people, and wildlife will increase, perhaps dramatically.

Third, the physical scientists and engineers that are designing wind turbines and wind plants need to extend their thinking beyond cost and value strictly in the monetary sense. We really need to think more broadly about people, wildlife, and land use, as well as the traditional metrics of cost and value.

And finally, I'll spend a few minutes just talking about the value of the information exchange and communications among the many disciplines. One of the things that I really find exciting about wind energy is how interdisciplinary it truly is, and I think as we think about these really high penetration renewable energy futures with very high contributions from wind energy, the need for that interdisciplinary exchange and engagement just really compounds upon itself, so we'll talk a little bit about that at the very end.

Transition to a slide showing lifecycle air pollution emissions data and wind energy’s effectiveness in enhancing air quality.

All right, so in terms of the distribution of costs and benefits, this slide here presents life cycle air pollution emissions for a variety of electricity generation technologies, focusing on some of the traditional regulated air pollutants. And obviously, no surprise to this group, onshore and offshore wind both do very well from a life cycle air pollution perspective.

However, air pollution is a relatively regional issue, and it's largely concentrated in urban centers. There are of course a few exceptions to this. However, it really is in general, on average, a bit of an urban problem, but we don't site wind plants in urban locations. We site wind plants in rural areas. So that air quality and air pollution mitigation impact that you might achieve from wind doesn't benefit the communities that are actually hosting wind plants.

Transition to a slide showing chart with lifecycle greenhouse gas emissions data.

Not surprisingly, similarly, from a life cycle greenhouse gas emissions perspective, again, looking across a variety of electricity generation technologies, wind energy does very well. But we all know that within the policy debate and within the general conversation around climate and climate change mitigation, the issue of free ridership is very real. Not only do – are countries hesitant to sign on to agreements if they feel that others might not be adhering to the same standards, or if they feel like others might have a particular advantage, that plays out at the community and local level, too.

So when you think about the concept of free ridership, it's not hard for somebody to say, well, yeah, I think climate change is a problem, but these turbines are spoiling my view, and I'm not really convinced that they're going to have an impact on the climate because of the actions of other countries or other sectors of the economy.

Transition to a slide showing other impacts that may be significant to social acceptance of wind energy, but are often more subjective.

Going beyond those sort of traditional benefits of wind energy, we think about things like energy diversity, jobs, land lease payments, and then issues like public acceptance and wildlife. These are results from our 2015 Wind Vision study. The results themselves are less relevant to the current conversation, but these are benefits that we often think about with respect to wind, or impacts that are relevant for wind.

And unfortunately, in my mind, they actually complicate the conversation a little bit. Things like energy diversity is often touted. There's this sort of long term hedge value of wind energy that gets brought up, hedging against future fossil fuel price volatility. But that's a benefit that is kind of abstract, and it extends across all consumers and different power producers, owner/operators of plants. And so it's hard for that to be a tangible benefit for a community that might be hosting a wind project.

Jobs are an interesting one, too. That's probably one of the most compelling local benefits of a wind energy deployment. But even that, it's not uniform, how those impacts play out. In some cases, a project might bring in a construction company from some other region to build the plant. They might be importing their O&M services also from another region or from an urban center close by. And with respect to land lease payments, you have absentee landowners, and sometimes you have landowners who actually live in a separate state, but they're seeing sort of the benefits of that local land lease payment as well.

And so there's always complications and nuances, and this distribution of cost and benefits just is not a super straightforward conversation.

Transition to a slide showing U.S. Wind Fleet map in 2014 and then subsequent map of U.S. Wind Fleet in 2018 and then what a 2050 scenario with 22% of U.S. energy needs being met by wind.

Shifting gears a bit to talk about the interface between wind energy, wildlife, land use, and people, this slide here – I'm going to go through a couple of these slides that show how the footprint of wind energy could change as we go forward in time across a handful of modeled scenarios.

This particular plot shows wind plants installed in 2014. This was about 65 gigawatts serving about 5 percent of our electricity supply. Do note that the size of the dots are scaled to the actual area of the wind plant, so some of the smaller installations in the northeast in particular and in other regions of the country are a bit hard to distinguish at this resolution. But this was the picture in 2014.

Zoom ahead to 2018. Year end, we were approaching 100 gigawatts at that time, serving about 6.5 percent of our national electricity needs. We're now beyond the 100 gigawatt mark, but this is where we were at the end of 2018.

Zooming further into the future, we have a 2050 scenario here. This was a business as usual scenario from our NREL standard scenarios effort. It gets us up to about 260 gigawatts of wind energy, 22 percent of our electricity supply in 2050, with installations really expanding across the country. You see a substantial buildout in the East, in the Midwest, in different parts of the interior region, and increased deployment even in the more Pacific regions, the sort of Western, mountain, and Pacific regions, so really expansion of the footprint across the country.

If we think about more aggressive or more – higher penetration renewable scenarios, here's one where wind gets to about 500 gigawatts, 39 percent of our electricity supply in 2050. And again, you can see that footprint just kind of really take hold in many parts of the country, including even significant portions of the Southeast in these higher penetration scenarios.

The last one I'll show here, which is the highest one from our standards. I believe it was a low cost wind and solar scenario, and there may have even been a carbon constraint within the scenario. I don't recall the specifics. But it got us up to 630 gigawatts of wind, nearly 50 percent of our electricity supply, and virtually every portion of the country is impacted by operating wind facilities in some form or another.

So clearly, ignoring the expanding footprint of wind energy just isn't going to work. We really have to think about these impacts.

Transition to a slide displaying two maps showing habitat distribution of wildlife species of interest and competing uses of land.

Just amplifying that theme even further, these two maps show competing uses of land. On the left hand side, you have the habitat distribution for particular wildlife species of interest. It's not comprehensive, just a handful of species of interest. You can see them listed in the fine print down _____ the slide. Very few areas that are impacted by at least some form of wildlife.

On the right hand side, you see the same wildlife data overlaid with public acceptance as well as radar consideration, and the darker colors there reflect areas where you have multiple competing uses for a particular piece of land. Obviously, the circular areas kind of amplify that. Those are the radar installations. But even the darker shaded regions are areas where you're going to be relatively close to residences or people.

Transition to a slide showing coverage of bats in the U.S. in relation to wind energy.

Just zooming in on the bat problem in particular, the gray shaded areas here reflect areas where bats are relatively absent, so any area that's not shaded gray is going to have to think about bats. There are significant swaths of Texas and the sort of interior wind belt that are unaffected, but if you think about those images I showed a few slides ago, large portions of the Eastern half of the country and really just that buildout that affects the whole country, bats are really a significant consideration that we as a wind energy research community are going to have to grapple with if we want to achieve the levels of deployment and penetration that are envisioned as a basis for the Science article that we wrote.

Transition to a slide showing estimated wind plant power density by region as informed by existing facilities.

So how do we start to extent our optimizations and our analyses to go beyond cost and value? This is actually something that's already happening to some degree today. What this map shows you, and this is a work in progress, but what it shows you is when you take the plant densities of the existing fleet and do some robust statistical work to try and extend that across areas that are similar, you see variants in the relative power density of existing facilities and how that might play out on broader geographic scales.

In the plot on the right hand side, you sort of see the regional distribution of plant densities when applied across the entire region. And just for reference, the mean there is basically between two and three megawatts per square kilometer.

In the image, though, the lighter areas show relatively higher power densities. The darker areas show relatively lower power densities. And what this shows us is that plant designers are already taking into account local topographic considerations, property boundaries, population densities, sort of other factors than the wind energy resource alone or the construction challenges, to think about how to design our wind facilities.

And my contention would be that we have to increase our focus and our ability and our sophistication of doing these types of things by an order of magnitude, or perhaps more.

Transition to a slide showing sections of a map explaining wind curtailment regimes across time and space.

Another example of things that are getting a lot of attention today and that I think need to be increased and really built upon is thinking about how to mitigate wildlife impacts. In this case, we've looked at land area that would go from a positive to a negative NPV. If you had an existing wind facility in that, that's reflected by the red areas in the blow-up of the map here. And if you had a bat issue in those localities and you just installed sort of a blanket curtailment regime so you had _____ curtail below five meters per second for six months out of the year, something relatively simplistic, you take out basically about a terawatt or 1,000 gigawatts of wind energy potential from a positive NPV and put it into the negative NPV range.

If you instead though employ a smart curtailment strategy that focuses on the higher periods of risk or really tries to minimize the impact on plant operations while still capturing the same benefits in terms of reduced mortality, you could potentially recover about 90 percent of that impact, put those areas back into the black from an NPV perspective, make them economically viable once again.

So these are the types of solutions that need to be considered and given a great deal more attention. And again, extend this beyond the sort of traditional cost and value optimization framework.

Transition to a slide with text and graphic explaining interdisciplinary nature of work for a future with a great degree of wind energy penetration.

So just to wrap up here, the – not only do we need sort of the knowledge from the social and biological scientific communities, we need the information exchange. We really need strong interactions between a variety of disciplines, including, as Katherine mentioned, some of the data science community, in order to come up with the innovative solutions that are going to be really essential to getting us to the high penetration futures that we hope to see out there in the 2050 timeframe, with wind energy playing a big role.

So with that, I'll conclude, and pass it back to Paul. Thanks so much.

Paul Veers: Thank you, Eric. I think if you'll give up control, I'll turn on the camera again, and we'll – I don't know if Alex was going to moderate questions, but we're open to take questions, if you want to type them in the meeting chat. We can certainly respond to them. I think we have lost Katherine by this time, most likely. She had to move on. In Denmark, it's quite late. We'll take some questions if there are any.

Audience: Paul, this is Jason Fields. I had a question for you.

Paul Veers: Sure. Thank you.

Audience: So this – the grand vision process was really intended to be kind of a call to arms to the academic community and other disciplines which may not be directly engaged with wind energy yet. With the publication of the – the journal publication and the report, have you seen any uptick? Have we been effective in achieving that goal of outreaching to other areas?

Remaining time in webinar shows faces and icons of different people on the webinar speaking during a webinar Q&A session.

Paul Veers: That's a good question, and I'll leave it open for my other authors as well. A little bit hard to get some metrics on that. I'm not quite sure where we'd see that. Certainly, there's been growing interest in that area.

I think what we have to do, though, is move from a very high level and somewhat squeezed publication which tried to make the case simply that we're not done with this, that there are going to be some critical progress yet to be made from a scientific perspective, and flesh out those details somewhat.

So we're working on a series of follow-up articles that can actually drill down into the nature of those challenges, somewhat like the presenters did today, so they presented a lot of material that we simply did not have room for in that Science article, and get that out to the wind community first, and to the broader community after that. So I think there's still work to be done in that area, Jason. I think simply publishing the Science article was a first step, but then this is going to allow us to kind of take that further and say, okay, now what's the nature of these scientific uncertainties?

We know there are uncertainties, but what do they actually look like, and how do they impede our ability to do innovation? I think that's really one of the key elements, is that it's not just that there's something unknown there, but the fact that it's unknown is keeping us from doing things that might be an improvement, that might be a new and innovative solution to an existing problem.

So I think those are the kinds of things we're trying to flesh out now. And Julie is involved in an innovative to look at the atmospheric grand challenge. I'm working with Amy on one to push forward on the machine and those grand challenges. Katherine is working with people from the controls community and from the grid community, engaging perhaps with Paul Fleming and Jennifer King here at NREL, as well as those on the outside, to flesh out that third grand challenge.

So we have much left to do. But thanks. Good question, Jason.

Audience: Thank you.

Paul Veers: If my other speakers have something to add to that specific topic? Eric, I think maybe you could talk a little bit about the need to follow up. I think one of the biggest criticisms of the grand challenge effort was that we couldn't talk about everything, and that we did limit ourselves essentially to the physical sciences, and did not reach out into biological and environmental and social sciences. So that initiative is still yet to be completed. So could you say a couple of words about that, Eric?

Eric Lantz: Yeah. Absolutely. No, that comment is spot on about sort of the gap or the criticism that we received. There are actually a handful of efforts underway that are potentially going to result in papers that will address this concern or issue in some form. We have some of those actually within our modeling and analysis portfolio, really looking at this question of land use and how different types of siting considerations might impact the opportunities to deploy wind both from a technical potential perspective, but also a market potential perspective.

There is also activities within IEA Wind that other colleagues are really driving, and I expect that within the coming months we'll have another handful of papers within that topic that will make that extension and connection.

To the broader question that you raised, Jason, in my mind, it's a conversation, and I do think that we've opened the door to a lot more conversations about the research needs within wind energy. Certainly, some of those are within the wind research community, and where there may be kind of speaking to the choir, so to speak. But I think that there are a lot more – there's a lot more dialogue from the wind energy community out to other parts and portions of society and the decision makers that exist outside of wind energy.

So I'm pretty optimistic that the long term effect will be quite positive. Of course, as Paul mentioned, it's really hard to measure that, but I think if you just think about the conversations that we're experiencing in engaging around – from the foundation of the grand challenges paper, I think there's good reason to be optimistic.

Paul Veers: And if I could just follow up with that, give a pitch, we're right now organizing the next North American Wind Energy and WindTech Conference, which will be hosted at the University of Delaware in 2021, so it's a whole year away. But the intent there is really twofold, and it's to bridge two of the gaps that were brought up today. Katherine talked about the increased need to connect between the wind technology suppliers and the grid integration, and that conference is meant to be built around a theme of interface between the traditional wind energy research and grid research.

And it's also going to host an increased look at the social sciences and how they are integrated to how we push forward with the technology. So keep that on your calendar. It's a dialogue, it's a discussion, and there's going to be an attempt to bring these groups together and enhance that dialogue at the NAWEA/WindTech Conference in 2021.

Alexsandra Lemke: So Paul, we have one more question that came in from Julian. Julian, do you want to unmute yourself and ask your question?

Audience: Sure. I hope you can all hear me.

Paul Veers: Loud and clear.

Audience: Great. So Amy discussed this issue of safety factors and how since human life isn't at stake and it's more about costs, we should be able to have lower safety factors, where it may be economical to do so. And I'm just reading Julie's response to me now. There's this huge variety of types of inflows a plant could see. I mean, I think to simulate all of those would just be an ungodly amount of computational time. So I guess for me, it's kind of hard to envision a pathway towards having more trust in loads, because there's just so much variability that there could be, and there must be some _____ or _____ into our predictions.

So I'm just wondering, sort of looking ahead, next 10 or 20 or 30 years, like what pathways could the research community have towards a better trust in loading predictions between stakeholders and modelers? I really hope that makes sense.

Paul Veers: Well, you open a large can of worms on that one, but it's right in target with what Amy was talking about in the sense of the fact that when we look at the loads and turbines, and we have some simplified design rules that we use in order to make sure that they're safe, we have then built in some sort of safety factors, we also have built in some sort of a sense of a characteristic load.

So – and those things were calibrated with respect to the turbines that were designed 20 and 30 years ago, which were much, much smaller. So the nature of the atmospheric dynamics and those factors and those approaches that we built in place were built around assumptions of a smaller machine and smaller scales within the atmosphere.

So the experience since then has been that if we use these approaches and look at the track record of the industry, of the machines that have been built, and whether they're failing or surviving, they're doing quite well. So there is a strong need or a strong desire by the industry to not mess with something that's working well.

On the other hand, when you go offshore, we know that as Julie described some of these flow conditions, and Amy talked about some of the conditions, we know that these are not only a larger machine, but it's also a significantly different atmosphere. And there is a need for those offshore systems to come down in cost substantially, and there is a need to keep them as safe as they need to be.

So if we're not going to continue to leave a lot of cost on the table, and we're going to try to optimize these systems to the actual conditions they need to survive, we're going to have to reexamine that set of load conditions that we have so that we can take and remove unneeded excess material and excess cost in order to optimize these things down, which means we have to characterize that flow field.

In terms of the second question you asked is really about the safety factors, and the fact that these are not human life critical, so that we can then – that puts these structures in a different category of structures, and we can apply different safety factors. The discussion goes on among the community about what an appropriate safety factor might be, and the difference between ensuring that an offshore plant survives no matter what versus the realization that we may want to for cost purposes install a plant that survives up to a certain level of safety or a level of storm, and then we buy insurance essentially to say that we know that over a 100 year time period, storms of a high intensity are going to hit at some place or another. They're not going to hit everywhere, so I don't have to design all of my wind plants to do this, but I have to buy insurance for those places where the worst and most intense storms will hit, and we can then replace what gets destroyed by the most high intensity storms.

Others within the community would say that this is not all cost, that there is a public relations issue here, and if we have a storm early in the deployment of offshore and the hurricane happens to hit our wind plant, that there will be a penalty to pay for the industry as a whole. So I think those discussions continue to go on, and everything that comes up in some sense, as Eric said, begins to look like a social issue as well as a technical issue. So these things are never quite completely separable.

And I would just – I guess we have other questions we should get to. But Amy, if you had other comments to make, just jump in.

Amy Robertson: Yeah, I think you covered most of it. I think it might be good to go to the next question, because it's also related to offshore wind technology.

Alexsandra Lemke: Yes, Christian, do you want to go –

[Crosstalk happening]

Alexsandra Lemke: Christian, do you want to ask your question?

Amy Robertson: Or it was – yeah.

Alexsandra Lemke: I think he might be muted, but Amy, you can read the question. You want to go ahead and answer it, about the feasibility of fixed bottom in larger, deeper water?

Amy Robertson: Christian, are we hearing you now, or should I read it?

Audience: Yes. Oh, whatever, which one you prefer.

Amy Robertson: Go ahead.

Audience: All right. Yeah. So I was thinking about these offshore – floating offshore technology, and even that the availability of space and – when talking about offshore fixed bottom, is reaching its maximum capacity. Do you think that the floating technology will be ready in the near future, so that it can overtake the existing technology?

Amy Robertson: Yeah, that's a great question. I think – we've done multiple studies in this area. I think we see that floating wind technology is still more costly than fixed bottom, and I think it's just due to, as you said, the technology readiness level that we're at for floating wind. I mean, there's a lot more issues that we need to overcome with floating wind, as I talked about the complexity of the dynamics of that system. We're adding a lot more things we need to consider.

But our models do show that we are progressing. We're trending downward for cost. We do have some pilot plants in that area. And our expectations are kind of in line with what you're saying. Around 2030 we're expecting to kind of be on the same cost level as fixed bottom.

There's actually some opportunities we have with floating wind that you don't have with fixed bottom, that if you could design your system fully cay-side at a port and just tow it out, you have to do a lot less work at sea than you have to do for fixed bottom. You don't have to transport these large cranes offshore, things like that. So if we can figure out these challenges we need to overcome, we can actually get to maybe even a lower cost point than fixed bottom for the shallow water locations.

And in the US, yeah, this is really important. The US is one of the unique areas in the world, Japan is another one, that has a lot of that deep water offshore resource. The whole West coast will probably be 90 percent floating wind, and it's a challenge that we need to overcome, but I think, yeah, around that timeframe we'll get there. We have some pilot projects now, and we're on our way.

Audience: Yes. Thank you. One quick thing. So you're saying that by 2030, it will – I mean, the floating offshore technology will reach an LCOE, same level as fixed bottom or even lower, but do you think that people will be – or governments will be eager to invest within this technology? I understand that the LCOE will go down, but do you think that people are keen on looking forward to implementing this kind of technology? 

Amy Robertson: Yeah, I mean, the advantages with – I mean, there's some advantages and disadvantages with floating. Environmental challenges will be different for floating wind systems. As you go out in deeper waters and have the mooring lines, they can create challenges. I think for the social acceptance, having these systems further offshore where the waters are deeper are going to be advantageous to social acceptance, because they won't be as readily seen from the shore.

We do see that a lot of local governments are very interested in offshore wind in general, because there's a lot of ports that have been shut down over the years. The shipping industry has become more large scale. A lot of the smaller ports no longer can be useful. And this can help revitalize I think a lot of local economies with their port infrastructure.

Audience: I see. Yes. Thank you for your time and your answer. Yeah.

Alexsandra Lemke: Okay. I wanted to express my gratitude not only for all of the attendees but for our esteemed speakers today. Unfortunately we did run long, but ultimately, I would like to encourage anybody, if you have specific questions, to email me, and I'll vet those with the appropriate individuals, and we'll get right back to you.

So without dragging it on any longer, we really appreciate your time, and thank you for joining us today.

Paul Veers: All right. Thank you very much, Alex, for moderating and for organizing this. Look forward to your next webinar.

[End of Audio]

Webinar ends with an animation of the NREL logo.