Advantages of Distributed Wind Generation and How It Plays an Important Role in the Energy Mix
>>Alex: Alright. Well, I think this is a good time for us to get started while everybody's kind of funneling in. Welcome. Good morning, good evening, good afternoon, wherever you're calling in from today. Welcome to the Distributed Wind Generation and How it Plays an Important Role in Energy Mix.
My name is Alex and I'd like to say "Congratulations – you're almost toward the end of the week and thank you for being a part of NREL's Wind Energy Science Leadership series." These webinars have had great success addressing the challenge facing wind energy and the pathways forward for making wind one of the most prevalent energy sources of the future. In today's webinar, you will learn about NREL's distributed wind research capabilities and how they span the innovation pipeline, including design, modeling, simulation, resource characterization, analysis, and manufacturing.
Before we begin, today's webinar will be recorded and available on demand in a few weeks and will be posted on NREL's website. Please, be sure to mute your phone lines and turn off your cameras if you're not speaking today, and next – we encourage you to use the chat feature or raise your hand to ask questions, and we will be answering those questions at the end of the webinar today. And finally, I would like to introduce program manager, Ian Baring-Gould and lead for our Distributed Wind portfolio and moderator of today's webinar.
Ian? >>Ian: Hello, everybody, and thank you, again. Thank you, Alex, for the introduction and thank you, everybody, for joining us on our session today. Very much looking forward to talking a little bit about distributed generation and distributed wind in specific and how that can play a role in the energy future going forward.
So, thank you all for your attendance. As Alex said, please, don't hesitate to type stuff into the chat – questions and things of that nature. We'll be taking questions at the end of the presentations today.
Quickly, just want to talk – kind of set the stage of what we're gonna be talking about today. It's a very exciting time in the energy space and kind of the challenges that are in front of us are pretty mind-boggling if we're looking at meeting what the administration is looking at trying to do. So, we have these kind of national goals that are before us that I list here. Primarily, the 2035 Clean Energy goal, which is monumentous by itself, but then, we add to that a goal to reduce the carbon content of buildings by 50 percent in that same time frame. So, not only are we planning on – or are we trying to or are we going to try to do or what we will succeed, I guess, in reaching a zero carbon from the energy sector, we also have a huge effort in front of us on the building sector. In addition to this, the goal of the administration to really tackle this question of economic development opportunities in communities – both disadvantaged, but then, also ones undergoing energy transition.
So, all of these communities – the energy veterans that have made America what it is today in all of these communities across the country focused on fossil fuel technologies, making sure that as we do this transition that the administration has put forth, we do it in a way that understands and recognizes and supports the development of those communities in this energy transition. And then, the last one is critically important as well. To be able to achieve these goals, we have to put in a lot of new hardware. And the whole manufacturing and development supply chain for this transition is gonna be monumentous. And so, these challenges are facing us as we look at the market going forward, but I think there's a lot of hope and a lot of interest in the distributed generation space, because we see DG as being able to play a key role in looking at communities and helping communities develop going forward. So, generating local energy and improving the energy system, supporting community resilience in the face of climate change and the other changes within the energy specter, expanding the ability to provide local jobs and energy services within the communities, keeping the electricity and the money that is generated from local energy generation, within the communities.
And then, as I said, the ability of distributed generation to provide disadvantaged and transition communities with a way to move in this kind of new energy space with distributed manufacturing, both supporting domestic and international markets. So, that's the kind of context for the work that DOE is doing and NREL and the laboratories are doing to support that. So, we've got three great speakers today. I probably shouldn't say that.
We have two great speakers and then myself today to talk through the activities that NREL – but also the larger laboratory community, our sister laboratories at Sandia National Laboratory, Idaho National Laboratory, PNNL, Argonne, and Los Alamos, are all working to support the Department of Energy in moving the distributed wind energy space forward. And so, three speakers talking about three specific topics. I'll kind of give an introduction, look at the market, and talk about some of the technology innovations we see followed by Jim Riley, who's gonna talk about the MIRACL project, which is one of our large kind of efforts at integration, and then, followed by Heidi Tinnesand, who is gonna talk about performance assessment and how do we understand what the long-term performance of distributed wind technologies are gonna be, which is a key kind of challenge within the market space. So, some good speakers here that will hopefully set the stage for some good Q&A at the end of the session. So, I want to talk a little bit and give a kind of introduction to wind and distributed wind in specific. I always like leading off with a slide talking about the wind resource potential within the United States, because it is phenomenal, and we need to keep that in mind.
It's always good to remind everybody that the capability of wind to provide energy services across this country is astounding, and so, we're not working from a place of deficit; we're working from a place of strength. And we have kind of Alaska and Hawaii here – not colored in blue – but there are tremendous wind resources there. They just are not at the same level on the maps. And we already know that this transition to wind and solar is happening, and so, the question is, how can distributed technologies – primarily wind and solar – play a role in in supporting this wider energy transition? We all are pretty familiar with the large wind turbines that we see in media – so, large wind turbines, both land-based and offshore – but wind turbines come in a whole range of sizes, and so, we have small wind turbines like the one up here in the upper left that is primarily focused at homes, markets, mid-scale turbines that can be deployed in isolated communities, behind the meter applications – so, there's a whole range of wind turbine technology that's out there. It's not just the big wind farms or the big wind turbines that go into the big wind farms. We also have a number of different applications that we look at specifically in the distributed wind space, and those range from isolated communities like the community of Tutka Bay up in Alaska.
Behind the meter applications – and this is a picture of a facility in Ohio where they've installed, obviously, utility scale wind technology behind the meter of an industrial plant. We have microgrids which operate – or can operate – independently of the power sector, and then, we have utility-owned wind farms that operate on distribution voltage. And so, each of these different applications can use any of the technology that I talked about on the last slide. It can range from very small turbines up to very large turbines, and they all have kind of different economic considerations when you're looking at these different cases. And so, we have these – lots of ranges of turbine technologies that we can use, and a lot of ranges of applications, all of which have different kind of technical requirements, but also, different economic requirements.
And even if we take that one example of the behind the meter installation, there's lots of ways that can look. It can be a small turbine on a farm. It can be turbines as part of a community energy project. It can be at a local manufacturing facility behind the school, or – as we see up here on the upper right – connected to the distribution voltage – or distribution system – just across the community.
So, within each one of these different applications, there are lots of ways that distributed wind technology can be used. So, with that, as a little primer, I want to talk a little bit about the distributed wind market first to give you the context of how distributed wind plays. So, distributed wind has been around for a long time, and, to a degree, all of the turbines – even utility-scale turbines that were deployed kind of in the '90s and 2000s – were all distributed wind scale project in wind farms or what have you.
So, the market was pretty developed and moving along pretty well. We did run into the – the industry, I should say, ran into a roadblock due to policy changes back in 2012 when the investment tax credit was adjusted to only cover solar technologies. And so, as you can see there, after that change in policy, the distributed wind industry hit a real kind of complicating point, which resulted in kind of wide-scale shutdown of that industry, and it hasn't really recovered since then. This chart, which is part of the market report – it's worth noting that the 2020 market report that is being done by Pacific Northwest National Laboratories for DOE is gonna be released really soon, so, we're looking forward to updating both this slide, but also, having that report come out. But what it shows is that a lot of the turbines that are being deployed right now are actually large turbines that are deployed in distributed applications – primarily behind the meter or in front of the meter deployments. But it's large wind turbines, and that's because those can take advantage of some of the tax benefits – PTC and otherwise – that the very small turbines can't take advantage of.
So, policy plays a huge role in both what the market looks like today, but also, what the market could look like in the future. We've done a bunch of analysis here at NREL, supported by DOE – the Wind Energy Technology Office – that looks at where wind plays – or distributed wind plays – and this is looking at behind the meter applications. So, if you kind of take the country and kind of break it up into where could you deploy wind, there's a lot of places where wind can be deployed. And so, just under kind of 50 million residentials, commercial or industrial sites that have kind of the land space available for distributed wind, you can see here – 3 terawatt hours – 3,000 gigawatts of potential capacity for turbines that can be deployed.
So, it's pretty clear that there is very large market across the United States for distributed wind if we can get the kind of pieces in place to allow us to take advantage of it. Another study that was done here at NREL – again, supported by DOE – looks at the types of areas where distributed wind makes the most sense. And so, these four quad charts give you an indication.
The one up in the upper left show where there's kind of wide-scale deployment opportunity for wind, and the darker the green, the more deployment. The upper right shows you wind resource – and these are maps of Colorado, by the way – and this report looks at this for a couple of additional states. But what you see here is if you look at where the great wind resource is, you can see, in Colorado, it's out in Eastern Colorado – those dark blue spaces in the kind of B figure up there, and that's where we see all the big wind development.
But where you want to put distributed wind is not necessarily in areas where there's great wind resource, because there's not a lot of load out there, and there's not a lot of people. And so, the figure in the lower left shows you load and so, these are areas – again, in Colorado – where, on average, the consumer isn't consuming a lot of energy. So, you can see here that this kind of corridor north and south around Denver are areas where you have a lot of load, and then, the lower right shows you land areas. So, these are taking all of this land and breaking it up to where could you deploy a wind turbine – a distributed wind turbine – which could be a 5 kilowatt all the way up to megawatt scale turbines and behind the meter applications.
And so, when you add all these – the kind of B, C, and D together to get A, and what it shows you is where you have lots of capacity for distributed wind is not in the areas that have really high wind resource, but in areas where you have big enough chunks of land where you can deploy turbines where the resource is reasonably good and where you have high load. And that, to a degree, makes a fair amount of sense, but it's not where you expect a lot of kind of large-scale wind development to happen. We've also looked at the United States and what happens when you start lowering the cost of power. Right now, depending on the size of the wind turbines – and I'll get to this in a second – we're talking about $4,000.00 a kilowatt hour for kind of your typical residential
turbines, all the way down to about $2,000.00 for kilowatt hour for large kind of 300–400-kilowatt turbine sizes. But, if we could get all of the technology down to this kind of $2,000.00 per kilowatt and you overlay the wind resource and where people have loads and of these kind of things, you can see that there's a huge amount of potential for distributed wind development – about 5 gigawatts – 4 and a half gigawatts – at the residential scale and about 15 gigawatts at the commercial scale based on kind of current technology and current development pathways, current incentives, and things like that. So, clearly, there's a lot of potential out there, and distributed wind can play a key role in going those forwards.
I want to take – we focus primarily on the US market, but I want to take a quick second to talk about how distributed wind can play internationally, and there are a number of reports that have come out in the last year that really look at the transition that we're seeing here in the United States, but that the rest of the world is looking at. So, the World Energy Outlook shows that renewables will meet 80 percent of the future growth by 2030. Now, in a lot of cases, this is gonna be large-scale solar, wind, hydro as examples, but we have to recall and remember that a lot of the power systems in the rest of the world are not as strong as what we see in the US – kind of North America and Northern Europe – or in Europe – and so, the role that distributed technologies are gonna play in supporting this 80 percent of the growth in global electricity is gonna have to be distributed, because we're not gonna be putting 10 megawatt offshore wind farms off the coast of Kenya anytime soon because the grid infrastructure just can't handle it. And so, how distributed technologies and distributed wind play is gonna be critical.
And then, the World Bank and Sustainable Energy for All have also identified very strong needs for kind of microgrid power systems of energy access to the tunes of hundreds of billions of dollars that will be invested in this – right now, mainly focusing on solar technologies, but distributed wind can play a key role in those market segments as well. So, why should we be thinking about wind? I mean, if you talk about and you read the media and all of that in regards to distributed technologies, it's always solar and storage. That's what you hear all the time, and that's certainly what the developed market is.
But why should we be thinking about wind in these contexts as well? One is certainly cost. And the Department of Energy has done a lot of work recently to lower – to work with manufacturers to lower the cost of energy. And I would say this is a very successful effort that Department of Energy has undertaken and is showing really good results with manufacturers in the distributed wind space, being able to come in with costs of certified technology – so, reliable technology – that are well within the bounds of solar technologies. So, unsubsidized rooftop solar is about this $0.15 per kilowatt hour, and you can see here
that turbines – kind of 10 kilowatts and above – are really coming in at those price points. And so, lots of capability for the technology – again, certified technology – within the next couple of years to be able to be cost competitive with solar to be at the same kind of busbar cost as solar is really a critical way to allow wind to go forward. We also have a bunch of other kind of benefits to looking at wind, and the speakers will dive into this a little bit more, but certainly, higher energy density or high energy density in areas with good wind resources – that makes perfect sense; a really small footprint allowing increased onsite generation. So, if we're looking at how do you get to 50 percent decarbonization of buildings, onsite generation is gonna be critical. And so, clearly, you're gonna put solar on the roof. That, to a degree now, is becoming a no-brainer.
But then, how do you go beyond that? Because solar on the roof is gonna get you 20-30 percent of your local energy needs depending, obviously, on where you are and all those details, but how do you go above that? And as soon as you kind of use your available real estate, being able to put in larger amounts of distributed generation become increasingly harder. And so, being able to put in large scale wind, distributed wind technology – so, you have solar on the roof, and you have a wind turbine in the parking lot, and so, being able to increase that onsite generation is gonna be really important. We also have – wind turbines have different – and to a degree, improved – grid support capabilities because you have rotating mass within – in the rotor system, and so, you have some additional capabilities in regards to grid support.
Wind is also very complimentary to solar, which I'll show in a second, and so, the ability to have daily and seasonal variation means that you don't have to rely so much on storage to be able to provide high contributions of onsite power generation. And then, certainly from a wind perspective, a whole – a high amount of domestic content really drives to that kind of local economic development that distributed wind can bring to the table. I mentioned the complementarity of wind and solar because that's really critical and ends up being, I think, one of the key drivers of why it's not going to be a question of wind or solar, it's going to be a question of – or it's going to be a requirement – to have wind and solar as we go forward.
So, Jim will talk a little bit about this in more detail, but you can see that from these maps, that if you look at the complementarity – which you see in darker reds – both from a daily and an hourly basis, in huge parts of the country, the idea of installing wind and solar at the same time – it makes all the sense in the world. And so, as we get to high renewable contributions for behind the meter or onsite generation, you really have to have wind and solar – and likely, wind and solar and storage – to be able to get to those high contributions, but it has that economic value. There is certainly ongoing challenges to the deployment of wind that I'll touch on quickly and then, Jim and Heidi will follow on in kind of more detail. Clearly, there needs to be more cost reductions and reliability improvements. That's been a key issue across the industry, but certainly, with the current policy situation that we have, distributed winds have to be 50 percent lower than solar to be able to compete. And so, those cost reductions and reliability improvements are really important.
Interconnection requirements and acceptance for rural utilities is critical, and Jim will talk a little bit about that. Access to financing is a key issue, and Heidi will dive into that a little bit with the performance prediction. Zoning and permitting variance is a very big issue. If you have to go through a zoning process to get a variance for any wind turbine you put up in a community, that adds cost and complexity to the process, which makes it really hard to do. And the, I think one of the biggest ones is the lack of information, knowledge, and understanding in regards to how distributed wind can play a role going to the future, and we see that all the time by kind of the focus on solar plus storage within the distributed generation context – so, a huge thing that has to overcome.
The great thing is – the Department of Energy is funding work at a host of laboratories, as I mentioned at the top of the presentation, to address a bunch of these activities, and Jim and Heidi will dive into a couple of them, but I just want to give a quick sampling of what is out there. So, two ones that I'll talk about in a second – the Competitiveness Improvement project, which has really spurred the lowering and the cost of the technology; a D3T project – which looks at deployable turbines, improvements and standards, additional research platforms that I'll talk about, as well as improving prediction and then, a lot of looking in market dynamics with modeling and analysis, all of which are funded by the Wind Energy Technology Office under the Department of Energy. Two projects that I really wanted to highlight. I talked about how the costs of wind turbines have gone down substantially – in some cases, over 50 percent in the last few years.
That's largely been driven to the Competitiveness Improvement project, which NREL has the pleasure of running, and that's working collaboratively with a whole bunch of companies – 23 here – to really take the technology of 10 years ago and move it into the modern space. And so, that is much larger rotors. So, we're looking at specific powers of down to 150. So, much lower than even the utility-scale turbines that are available today. Lots of work in power electronics to be able to provide the grid support services that Jim is gonna talk about.
Large efficiency gains in regards to manufacturing. So, a lot of work being done by the Department of Energy, but in partnership with these companies to be able to lower the cost of the technology to a point where it's cost competitive, as well as looking at new markets. So, the Sandia-led D3T project is really looking at how do you deploy turbines at low cost to be able to address the kind of – some of the market niches, but then, have those trickle down into the balance of station costs to be able to continue to lower the cost of technology. Another big investment that DOE is doing at NREL but also at other laboratories – Idaho National Laboratory; Sandia most specifically – is investing in the technology and the research infrastructure to be able to support the kind of ongoing research tasks that are gonna be needed for distributed energy future and allow wind to play a role in that. And a huge investment right now, at the NREL Flat Iron's campus under the Aries project – to update the turbines and the research platforms that are available for industry to use in supporting the next generation of distributed generation technologies or distributed energy – distributed wind technologies to support distributed generation.
So, with that, I would like to hand off the presentation to Jim Riley to talk more specifically about the MIRACL project. So, I'll drop out and Jim, if you want to pull up your slides. >>Jim: Thanks, Ian. Pulling 'em up now.
>>Ian: There we go. >>Jim: There's a lot of [Inaudible] through – it's coming through your Inaudible I think. Ah. Turn off my speakers.
So, like Ian said, I'm Jim Riley. I'm an electrical engineer here at NREL. I have the opportunity to lead the MIRACL project. It's Microgrids Infrastructure Resilience and Advanced Controls Launchpad. This is a multi-lab partnership between Pacific Northwest National Lab, Idaho National Lab, NREL, and Sandia National Labs. Quick overview I'll step through today is just an overview of the overall MIRACL project and dive deeper into some of the research areas, including valuation and modeling, resilience and cyber security, advanced controls, and then, hybrids.
So, the main goal of the MIRACL project is to try to advanced distributed wind research and try to demonstrate, basically, that wind can provide additional services to the grid beyond just – to distribution grids – beyond just watt hours or energy to the system. We're doing this in a couple of different ways. The first – this research area's being led by PNNL. They're looking at valuation and modeling. I'll dive a little bit more into each of these on the next slide. Idaho National Labs is leading research on the resilience and cyber security space and then, NREL and Sandia National Labs are collaborating on the integration and advanced controls side.
Ian had sort of introduced the use cases that we're structuring this research into because, as we mentioned, distributed wind is deployed in a ton of different types of scenarios, and each one of those have different opportunities and challenges. So, we've structured into four different use cases, the first being wind and isolated power systems such as islands out on the Caribbean or up in rural Alaska; the second is wind and microgrids; the third is wind and behind the meter deployments; and the fourth is front of the meter deployments. The way we're defining and accepting the US Department of Energy's definition of a microgrid – as, "A power system that can operate both in grid connected" – so, behind the meter or front of the meter configuration – "and then, by opening or closing a disconnect device you can operate in islanded mode as well." So, there's a lot of relationships for a microgrid between isolated power systems and grid-connected deployments. So, diving into the PNNL research a little bit, this is split into three main focus areas. The first is modeling to try to improve how distributed wind is represented in grid modeling tools by developing – what's on the bottom left here is – a PNNL generic power curve where they're trying to get just a more normalized power curve that can then be included in a number of different modeling tools across the industry.
And this is being presented to a number of tools like GridLAB-D, NREL's REopt, OpenDSS – a number of other tools that would represent a wind turbine and just trying to at least make it so that it's an apples-to-apples comparison between them. This is just one of the examples that's coming out of this modeling work at PNNL. The second research area at PNNL is focused on valuation, largely trying to understand what are some of the benefits, and Brian, in the chat window I had seen, put in "Synthetic inertia from small wind turbines". That's a good example of where is there a direct benefit that could be seen – a quantitative or a qualitative benefit – that could be seen from a small wind turbine providing synthetic inertia into the system. What are the systems where that makes sense? Is there an economic market or are there socioeconomic benefits to that? And there's a great publication if you follow this link.
I tried to shrink down the URL to make that easier to pull up, but what are some of the benefits of distributed wind and are they accurately represented? The third area is co-simulation, which ties us into some of the work that we're doing with Idaho National Labs is to develop a co-simulation platform to try to connect and leverage some of the physical assets at the different national laboratories so that we can leverage each other's assets. There's been a significant amount of investment at the individual national laboratories and we're trying to make it so that we can share some of those assets with each other through real-time data connections and things like that. The second research area is led by Idaho National Labs.
They're diving into the resilience and cyber security space to better understand some of the cyber vulnerabilities and improve resilience in microgrids and distribution systems by including wind in the system. So, to do this, NREL – or INL – has established resilience metrics and a resilience framework for distributed wind and is trying to structure it by the MIRACL use cases, again, so that someone that's in an isolated grid can find specific examples that relate to their scenario. So, on the left, this is one of the INL's DIRE curve – Disturbance and Impact Resilience Evaluation curve – where they step through some of the stages of a catastrophic or an event that may take your power system offline. So, there's the Prepare, Detect, Adapt, Recover, and Return to Normalcy. This is the kind of the logical step, and this could be a number of minutes to hours for recovery, or it could be a much longer time frame, and during that Adapt, Recovery, and Normalcy time frame, you're almost full rebuilding your power system, which is – a good example of that is in Puerto Rico right now.
They're trying to understand how to fully rebuild their system from the ground up after the recent hurricane there a few years ago. Second, on the right side, is Idaho National Labs is developing a cyber security guide where that is relevant specifically to distributed wind and other distributed energy resources to try to give an understanding on what are some of the tools and methodologies that are out there to address cyber security concerns, but also, to make wind and distributed wind be a portion of your actual cyber security plan to improve the overall system. With most generation going online now, there's a lot of control capabilities, and that enables you more visibility and control of your system, which can enable you to detect threats sooner. So, the bottom right is just an example of some of the access control limitations that are typically on a power system.
There's a number of different potential threats – whether it's a benign threat or a malicious threat – and then, insider group – how can people get access to a power system or distributed generation source? So, this is just an example of how to categorize people that have access to your different devices. Advanced controls side – so, this is a collaboration between NREL and Sandia National Labs. We structured our work into three different research areas and are partnering on this using some of the hardware and software capabilities that we have.
So, the first research area that we're focusing on is fault-adaptive controller logic. I was relatively surprised, when I first started learning about this, but there are a number of faults that could occur within a wind turbine, and in order to protect the wind turbine, typically, the turbine will just follow this right side here. For example, if there's an oil temperature sensor that causes an issue, the wind turbine would typically just trip offline. And while that may not be a significant impact to a large grid-connected deployment – such as a big wind plant – if that's your only wind turbine – such as Saint Mary's up in rural Alaska – that wind turbine is a significant portion of your generation fleet so, an outage on that one turbine could be catastrophic to your power system and take it offline. So, Ben Anderson is working on developing this fault-adaptive controller logic that it tries to understand is the defective fault critical.
If it's not, there could be an opportunity to reduce power and re-evaluate to see if that mitigated the fault or to initiate a time shutdown – or an instantaneous shutdown – depending on what the fault is and how dangerous it is to the large spinning mass of the wind turbines rotor. The second research area is advanced controls integration. The bottom two graphs are examples of how a wind turbine can support. The middle one is how a wind turbine can support along with other generation sources – how wind can support reactive power on your system by injecting bars into the system, and on the right is how wind can support – this is not reactive power support. That's frequency – how wind might support frequency.
This is past research that we're looking to reproduce on distribution systems – isolated grids and microgrids and distribution system deployments. We're also looking at how IEEE-1547 – how that impacts and enables additional grid services on distribution networks, but also how it may become a little bit of a challenge for some small wind turbine manufacturers with the increased certification requirements. And we have a paper that we're – recently published. That's on the NREL website on that that outlines some of those challenges and how the Department of Energy is actively trying to find ways to support those small wind turbine manufacturers. The third research area is focused on wind, solar – or distributed wind hybrids – how does wind work with solar storage, diesel, or natural gas – any variation of those three – and others. So, one of the opportunities that every time we present at one of the conferences is basically, how does wind work with some of these other generation sources? How can you make it more plug and play? How do you quantify what some of those benefits are? And that's exactly what we're trying to look at under the MIRACL project – is trying to better understand what are the benefits, challenges, and opportunities that hybridizing wind with other generation sources might provide, and also, aligning the resource – the controls research with resource availability.
So, aligning wind and solar or wind and diesel – does that enable sort of increased valuation or other additional benefits from the distributed wind to a power system? So, some research that I'm really excited about is leverage the HOPP project and some existing partnerships that NREL has with other organizations. This is trying to understand where are some locations where it may make sense to deploy wind with solar or, instead of adding additional solar to a system, would there be benefit to adding wind? In some of my other work, we've worked with the Department of Defense to try to help them install large solar assets, and they've done a great job of deploying a significant amount of PV on a number of these sites. Now, they're trying to include that in their energy resilience scenarios and they're trying to build out microgrids using this. And the simplest example I always end up using is – with solar, if you're trying to build a microgrid, you need something at night.
As you can see on the right here, this is sort of a normalized solar PV curve, producing zero at night, as you would expect. What this map is trying to do is try to understand the wind and solar complementarity. How do they complement each other through daily changes, monthly, or even yearly? Over the course of the year, how do they complement each other? And this is a specific example at Naval Support Activity Mid-South where there's a large solar PV deployment – that we curtailed for this scenario – and this is sort of how the wind would produce in that area on average over the course of the year. And we believe that instead of building a large diesel generator or a very significant amount of energy storage, if you build wind and solar at this site, they would complement each other, hopefully, to reduce the amount of capital expenses you would have to provide on other sources of generation – like diesel or storage – but still serves some of those critical loads that the Navy might have. So, this area is a portion of the – so, the MIRACL data hub and the co-simulation platform are intended to enable the launchpad portion of MIRACL. I won't reiterate the entire acronym again as it's a pretty long one, but that L in "MIRACL" stands for "launchpad".
Idaho National Labs is leading development of what we're calling the MIRACL data hub, which is intended to try to make it so that we can share data, model some of the physical assets across the laboratories – for both research demonstration and validation. The launchpad portion is – we're not trying to do that just for the national laboratories. We're trying to establish a platform and an opportunity to better engage with industry so that the distributed wind, microgrids, and isolated grid communities can come and engage with the national labs, have some of their research and validation done on the infrastructure that's listed here – as well as many others – to try to validate some of the concepts that you're trying to deploy at some of these power systems across the country and across the globe.
Just to reiterate the project team. So, myself, as the principle investigator; PNNL is leading modeling and valuation – that's Alice Orrell from PNNL; Jake Gentle from INL is the cyber security and resilience PI, and Brian Naughton is the controls co-PI from Sandia National Labs. And I think Ian, were we gonna pause for questions here or at the very end? >>Ian: No. We'll wait for the very end. So, Heidi's up next.
>>Jim: Okay. Pass it over to Heidi's for TAP. >>Heidi: Thanks, Jim.
>>Ian: And then, people continue to type questions into the comment field or once we get to the end, we'll raise hands and go around the room like that. But we've got over 100 people on the call so, typing them into the chat is probably an easy – may be an easier way to do it if you have questions that you're holding on to. Heidi, go ahead. >>Heidi: Perfect. Thanks, Ian.
Good morning, everyone. So, I'm gonna talk about the project that I've been leading for the last couple of years. It's also a multi-laboratory project. It's a collaboration between NREL, Pacific Northwest National Lab, Los Alamos National Lab, and Argonne National Lab.
I'm a researcher at NREL, and I want to start by mentioning that this name of this project is a little bit of a misnomer, but it really gives a nod to the impact that we want to have. So, the project is laser focused on accurately estimating wind resource, but the ultimate goal is for the wind resource characterization or capabilities to be integrated with a lot of different user-facing tools that _____ been resourced for lots of different applications in addition to estimating project performance. So, a little bit of background of where the distributed wind industry is, for those that aren't familiar. Estimating wind project performance in the distributed wind industry has been really scatter shot, so there are a number of tools I'm kind of highlighting here on the top, and they all have certain limitations. So, for Bergey Wind Power, they're the longest standing manufacturer of small wind turbines in the US.
They have their own proprietary tool. It's wonderful, but it only estimates project performance for Bergey turbines. Then, we have this wonderful tool in New York, but it only estimates for projects in New York, and then, we have this great tool – System Advisor Model – at NREL, but it's using a wind resource data set that is a little incomplete and inaccurate for what we want it to be used for. And then, most of the industry uses some proprietary approach. They use spreadsheet tools, Excel, plus, the nice wind resource maps that NREL has provided. So, what we're seeing from this – this is another graph from the market report that Ian referred to – is that we're getting a real wide variety of project performance, and I want to bring your attention to the plot on the left – the green plot.
These [Break in audio] certified turbines. We sort of can't make too many claims about non-certified turbines there, but this does kind of refer to something that Ian was mentioning before about policy. So, you know, as we have incentive programs, we can incentivize certified rather than non-certified programs, as well as use of certain tools – like the NYSERDA tool in New York. So, if we're looking at this project performance – and this is data that's been collected from USDA grants as well as the NYSERDA grants – and so, we see really wide variety. So, NYSERDA, you have to use the tool, and USDA, there's no real guidance on what you can use. So, we've got kind of a nice average there of estimated versus actual performance, but a really wide variety.
And when I think about buying a turbine and having somebody estimate what I'm gonna get, this plot makes me feel a little bit unsure about what my results might be. So, these are really the core challenges that TAP is trying to address. So, the inability to consistently and accurately predict project performance. And a couple of those key reasons – one I referred to already about these desktop modeling approaches not being accurate enough or not comprehensive enough to estimate all scenarios, but also, there really isn't the sort of project financing available to install a MET tower or wait for a whole year. So, it's really time and cost prohibitive to employ utility scale approaches to distributed wind projects.
And that really has a huge impact on stakeholder confidence, and also, access to financing. And when you think about kind of access to financing, it's a little different than how you might think – typically, in the utility industry –but you can look to the solar industry and the ability to use these lease models. And that has – that's – because a company is able to have a whole portfolio of projects and work with financiers and have that financing ready right as the consumer is going to buy that project, and then, pay sort of overtime rather than small wind, a lot of times. you have to sort of pay the whole upfront cost right at the outset. So, this is our primary mission – is really to enable the widespread adoption of distributed wind technology by improving resource characterization capabilities.
So, as I mentioned, we're really laser focused on estimating the wind resource. That's really one of the biggest things that the industry, with its limited resources, can't do on their own so, we are applying the wisdom and the expertise of four laboratories to really get this part done right. And then, you know, we're gonna share this characterization capability with tools, but there's some unknowns there. So, we're focusing on precision first and then, accuracy later as we really start to understand and characterize more of the project losses. And a big one there is really availability of turbines as we have sort of a wide variety of sort of service in the areas and monitoring, as well, for systems – and those are other key challenges for the industry. So, for the project, we've got these key work products, and it really falls under kind of three main areas and then, grounded in deep stakeholder engagement.
We really like to follow sort of an agile methodology where we're sharing our innovations and getting user feedback as we're going along the way. But the first key product here is the National Wind Resource Data Set – and Ian mentioned that earlier with one of his maps – and then, getting into obstacle flow modeling. So, with small wind – distributed wind industry – the turbines tend to be installed closer to trees and buildings where those are having a much larger impact rather than sort of turbine-to-turbine interactions.
And then, we're developing a framework to really help make the data quick to access and easy, sort of single point of access for all of the inputs that you would need in order to calculate the wind resource. So, in terms of wind resource modeling, this isn't an activity that's exclusively valuable for the distributed wind industry. This is a product that many of you may be familiar with already.
We have a wind toolkit. It has seven years of data at five minutes, but it has some limitations with it. It was originally developed to mimic forecast errors, and it doesn't really give you a lot of indication about the sort of uncertainty in the models. So, we're developing a new data set that we are sort of – have two years on now and are currently turning the crank on getting the rest of the 20 years.
So, we will have 20-year data sets starting from today, going backwards 20 years. It will be five minutes for the first five years and then, to hourly data. We're gonna actually publish what the uncertainty of this model is, and I'll go a little bit more into that in a minute, and then, characterize the bias in this where we can, where we have available sort of validation data sets, and it will be with all the same two-kilometer spatial resolution, but we were actually adding quite a few more heights so, that will be really exciting – everywhere up to sort of several hundred meters to look at air [Inaudible]. So, in terms of uncertainty quantification, there are sort of two key components there – there's internal variability and then, structural uncertainty. So, with internal variability, you know, there's a lot of difference in sort of what the model outputs look like, depending on when you initialize that model. And then, in terms of structural uncertainty, one of the key – or the most important scheme there – is the planetary boundary layers.
So, we have an entire ensemble of simulations that we've been running at different regions in the US at different times to really characterize what the spread of the uncertainty will be relative to the one – the best data set that we've picked that will be the underpinnings of the production data set. And then, we're also looking at characterizing the bias with a couple hundred data sets that we have now to see, as it sort of indicates, you know, how much bias there is in this model relative to actual measurements. And so, we're looking at that with the existing wind toolkit, developing a methodology to then apply to the 20-year data set once that's complete, and so, we're looking at a lot of different elements of the bias. They're regional, diurnal, seasonal – looking how it changes with hub height, wind speed class, stability, and then, ultimately – at least for the distributed wind portfolio – we're going to look at approaches to actually correct the bias, at least in specific sites where we have sufficient, accurate measurements to be able to correct that data.
Now, with obstacle modeling, we're looking at high-fidelity simulations. There's always lots of amazing pictures that come out of that. I'm sure many of you are familiar with these models.
So, the industry has been using – when they use models to actually look at obstacles and the impact of those – it's a simple 2D model or it's a rule of thumb that we just sort of avoid certain areas. But there was a 2D model; it was based on a fence experiment from about 30 years ago. And then, the utility industry researchers at NREL have many different complex computational fluid dynamics models, and what we're really hoping to do is find something in the middle. We need the models to be able to run very quickly, to be able to run on personal computers.
We can't have a whole army of Ph.Ds to model these sites for us so, we're really trying to bring all of the – bring more of the fidelity from these complex models into a sort of mid-fidelity model. And what we're trying to do is really balance time, cost, and accuracy. So, we started, last year, looking at the different models that we could employ – so, looking at RANS and LES models and a mid-fidelity model out of Los Alamos called QUIC, and then, looking at how that compares with the 2D models that the industry has been looking at. So, this yellow line – which is a little bit hard to see – is the 2D model and, as you can see, it's just wildly inaccurate.
And so, you know, can we better this by the high-fidelity approach? Of course we can. And then, how does the mid-fidelity do? It's looking pretty good so far, the existing mid-fidelity, but has known errors that we want to correct. And then, you know, how do these all look against each other? 'Cause we wanted to run a lot of these models in order to then build this new and improved mid-fidelity model. And then, this year, we're looking – so, that previous evaluation was looking at some wind tunnel data and now, we're looking at actually validating our models in real scenarios so that we can have full stability conditions and really look at more complex sites. So, looking at a single cube in a wind tunnel is a pretty simple case. We're looking pretty good, but what does it really look like in the real world? And so, we've got a partnership with a company in the Netherlands.
It's sort of a mid-complexity scenario we have. It's nice, because we've got a bunch of turbines that were deployed in a very short period of time in very flat terrain with very similar configuration of how they installed those turbines. So, we've got all the performance data from that. Not a perfect validation data set because we don't have MET towers outside of one site where they did their IEC testing, so then, we have two additional sites to work through this year.
So, we've got a building at Argonne National Laboratory, where we're deploying, currently, a bunch of instrumentation in METs towers and lidar and SODAR, and then, another building at Texas Tech, which is very cool. It's actually been one of the challenges of this project that there aren't very many cube-like buildings that are sitting off by themselves without anything else around that we can put instrumentation around. So, not only is this in that scenario for us, but it actually is rotatable. And so, we've just finished deploying a bunch of instruments down there and are really excited to just work through some model validation this year from these new mid-fidelity models that the teams have created. And then, finally, we have this computational framework.
So, we want to be able to implement consistently best practices for spatial and vertical interpolations. We can take the two-kilometer grid and then, get to the exact point of where the turbine is going to be stalled at the correct height. Then, we have access to GIS data through this portal as well to be able to characterize obstacles and terrain. We're working on new methodologies to do automatic obstacle detection so we can have the shape and the height that will feed into the obstacle models, then provide really, a suite of different obstacle models. So, as we sort of have more complex sites that we will enable the user to be able to choose more complex models that would take longer to run. And then, building a scalable data framework that will allow users to access this enormous amount of data quickly so they can run these assessments very, very quickly, and then, support the integration of this capability into commercial tools that already exist on the market.
And so, then, ultimately, we want to enable market transformation. So, there's, as I mentioned before, lots of different ways that this capability can be fed into existing tools to really help to – help different stakeholders to understand where distributed wind might be valuable for their stakeholders. So, if it's a county, you know, how can we use this to really look at the magnitude of the opportunity within the region? How can we then kind of feed this into user-facing tools that do preliminary site screening? And are there additional layers that we can bring in that would make these processes easier and reduce the cost to do project performance estimates? You know, can we impact sales better and help to get to those right locations? One of the biggest things for distributed wind is, unlike the utility industry where we really focus on the levelized cost of energy, distributed wind projects really want to minimize capital cost. So, how close can we bring the turbine to the building so we can have small cable runs while still having, you know, the project performance that we want to have, and then, detail siting as we sort of get into the end of the sales process? So, as Ian alluded to before, distributed wind is a viable distributed energy resource on its own or in combination with other technologies – on-site solar – for a range of energy customers. And it really can be a key component to 100 percent clean power system, especially if you want to sort of engage all the different stakeholders from the states that have 100 percent renewable goals down to, you know, companies that are really motivated to participate in the renewable energy future. It also has a really – a wonderful opportunity to – you know, along with solar and other distributed technologies – to provide a path to support communities that are underserved or undergoing energy transitions.
So, we see this quite a lot with community solar where a lot of folks don't own their homes, but can participate in something like community solar, but we have virtually no community wind projects and there are other opportunities there as well. And also, at – you know, we've seen amazing progress in solar over the last 10 years in terms of bringing down costs and really increasing deployment that indicate that distributed wind challenges can be solved with a combination of research, deployment assistance, and supportive stable policy. And with that, we can go to questions. >>Ian: Great. Thank you so much, Heidi.
We certainly – I know it's the top of the hour. We'll kind of tackle a couple of the questions, but Heidi, Jim, and I can stay on longer to kind of answer additional questions. Jim, why don't you pop on, put on your camera so we can see you, but why don't we start off, Heidi, quickly? A couple of questions in regards to kind of what level of accuracy do you think you can get with wind resource assessment and performance prediction? I mean, a lot of the work that you talked about addresses one of the key issues that we have is local obstacles make it really hard to predict resource. Understanding that you're midway through the process, what is your kind of shot from the hip thought about how close the TAP project and the laboratories are gonna be able to get? >>Heidi: Well, from the research we've done with the industry so far – understanding where they're at – they've been able to, fairly consistently, estimate for the experts – you know, like Bergey and NYSERDA. They can kind of get a plus/minus 20 percent on their predictions and so, our aim is to get close to plus/minus 10 percent. >>Ian: And is your sense that's good enough based on the discussions with the stakeholders that you've had? Or does it need to get better than that? >>Heidi: Well, honestly, if we got plus/minus 20 percent, we hear that's good enough so, we like to be not just good enough but better – really rocket this industry into the – into being part of the, you know, solution.
>>Ian: Great. I had one question in regards to kind of how does this look in Europe, and I think the answer to that is – very distributed. So, outside of offshore wind – which is kind of beast of its own – most of the projects in Europe would actually be classified as – under what we're calling here distributed wind.
So, kind of plans under 15 megawatts in scale that are connected to the distribution network. They don't have tons of communities – distributed small scale distributed wind – so, it's more common large wind turbines deployed in small applications – like we commonly see in Iowa as an example. So, a lot of kind of distributed generation, large turbines. Following on that question, Jim, how many, kind of, of the current fleet of turbines really can provide advanced axillary services like spinning inertia and things of that nature at this point? And what do you see as a research path needed to be able to have that happen? >>Jim: Well, I will say we haven't gone turbine by turbine looking at which individual turbines can provide the different grid services. We've tried to structure it by the various turbine types and we're leaning pretty heavily into what looks like the forward-looking number – like, the turbines that'll likely be deployed largely in the future, which is type three and type four turbines, which are typically full conversion turbines with inverters and converters in them. So, using the power electronics, you can provide that synthetically using the electronics of the inverter to provide synthetic inertia and things like that.
The limitation is really do you want to do it and is there a cost benefit for that specific scenario? And then, what are the tradeoffs? So, to provide inertia, do you need to operate sort of permanently curtailed to provide inertia up to inject power into the system so support frequency up? It depends on the scenario. So, it really – backtracking a little bit, in the future, we're expecting most turbines to be capable of doing that; it's just how often do you want to? What's the specific scenario and need? So, I redirected from your question there. >>Ian: No. no. I think that's a perfect response. I would also kind of say that there are a couple of turbine manufacturers that we're working with through CIP that are looking at integrated storage – and so, looking at the question of do you put storage at the base of the turbine, or do you integrate the storage into the turbine itself? And what is the value proposition of that? And the facilities that are being deployed at NREL under the Distributed Integrated Energy Laboratory are there to be able to test that capability and understand what those dynamics are – getting to Jim's point – about the economics. You can certainly do it.
How much value do you get from doing it, and does that warrant investment in that space, I think, is a key question. We have one question about policy, and I kind of mentioned how policy has hurt the industry. Obviously, under the Department of Energy, we don't advocate for policy one way or another. I know the Distributed Wind Energy Association is working in this area and is looking at trying to implement kind of level playing field policy in regards to the ITC and the ability to do cash grants instead of the ITC for those entities that don't have a tax appetite.
They are also looking at increasing the limit to up to 10 megawatts, because a lot of the – both solar and wind facilities – industrial clients and things like that are larger scale projects. So, again – we can't advocate for policy, but if you're interested in more on the policy, I would encourage engaging with the trade associations and I'm sure they would be happy to talk with you. Another question in regards to environmental impacts – mainly avian impacts with distributed wind. There is really very little data out there. The industry – there's not a lot of distributed wind out there and so, not a lot of research has happened in regards to that.
Some pockets of studies have been done that look at distributed wind – especially guyed towers/un-guyed towers, and distributed wind projects that have been installed in sensitive areas, and the kind of anecdotal is the impact is quite small, especially for the smaller turbines, because they operate much faster than the large turbines. Their rotational speed is a lot faster. But I would highlight that as an area with not a lot of research at this point, and something where more research could be likely or could be needed as we go forward. Posed for additional questions. I don't see any questions in the chat.
Do people want to raise hands or un-mute themselves to ask questions? Silence? Any questions out there? Okay. Jim, you talked really briefly about the launchpad concept within MIRACL. Could you talk a little bit about how companies or other stakeholders would engage in MIRACL, kind of under that launchpad concept? >>Jim: So, we're still developing this – sort of the construct around the launchpad. We've proposed a number of different ways to pursue this with the Department of Energy and so, yeah, it's still sort of under development. But right now, what we're trying to do is establish something so that industry can come to one of the four national labs – or additional national labs outside of the MIRACL project – and try to propose some challenge that they're trying to address with their technology or to validate one of the solutions that their technology might provide or a methodology. They would propose that to the national lab.
There's a ton of expertise and hardware that's deployed across the labs and we would try to use that expertise and infrastructure to help industry to validate these things, again, using lab infrastructure, lab expertise, but providing that out to industry. So, what we all, as a community at the labs, have a lot of interest and understanding on what some of the challenges are, but we're trying to better engage with the industry and pull them into the research – try to answer the industry's questions and challenges. So, the structure is not defined yet so, we're working on that. >>Ian: Great. Thank you. Quickly open up for kind of any additional questions from people out there.
If not, I can continue for a couple more minutes. [Crosstalk] >>Ed: This is [Inaudible] here. One thing about distributed wind is that the permit is very, very difficult, right, because of the turning blades. You know, it's different. If you have a PV, nothing will be so dangerous that – permit will be a lot easier. Now, on the other hand, there are a lot of applications where, you know, you've got a lot of space and even if something happens, it will probably be most likely happen during the night when the wind is blowing.
So, is there a way that we somehow can convince them to make it easier to permit, you know, based on that environment? >>Ian: Yeah. Thank you for the question, Ed. First off, I want to be really clear – the kind of the issues around safety – the whole certification process that has been implemented by DOE and the industry itself has really worked to address that problem to the extent humanly possible.
So, I mean, obviously, there are always things that can happen. PV can be blown off the roof of your house in a wind storm. So, you do have to consider that, obviously, but kind of disintegrating wind turbines are a thing of the past, and we need to be very clear about that. In regards to your question about permitting, I think the answer is, "Yes", and again, going back to Heidi's last comment – the kind of the solar industry and the work that the Department of Energy has done to allow solar permitting has been astounding, to the point now where a solar installer can, on their phone, request and instantly get a permit to install solar on someone's roof. And again, that's investment that the Department of Energy has made with the industry with a lots of other groups – Solar Foundation and others – to be able to make that possible. There is nothing that says that that can't be done here, and obviously, depending on how you're gonna deploy the turbine, the permits should be easier or harder.