The Need for Grid-Forming Inverters in the Future Power System
Okay. My name is Ben Kroposki, and I am serving as the organizational director for the Universal Interoperability for Grid-Forming Inverters, or as we like to say, UNIFI Consortium. And I'm going to be talking today about the need for grid-forming inverters in the future power system, and specifically why this is the most critical topic we need to address to achieve 100 percent clean energy goals. So 2020, which was last year, was a seminal year. A lot of important things happened.
And let me just touch on some of these critical things. First, the entire supply in the United States of energy is shifting, and since – you know, over the last decade, we've seen a dramatic decrease in the use of coal, whereas natural gas and renewables have increased, and nuclear and hydro have remained steady. But 2020 was the first year, at least on this graph, and maybe since the use of coal, that renewables surpassed either nuclear or coal in energy generation in the US. So with the 20 percent energy share, that's
a pretty significant achievement, where it finally passed coal and nuclear in energy generation. So let's dig in a little bit more to this renewables. So if we take a look at the renewable portion of that, the 20 percent that was annually produced from renewable sources, you see 8.5 from wind, 7.3 percent from hydro, 2.3 solar, and a little
bit from biomass and geothermal. Again, 2020 was another watershed year in that wind actually produced more energy than hydropower. And you can see over the last several years that wind and solar are the two forms of energy that are the most rapidly increasing on the grid today.
So if we take a look at that and we say why is this really happening, that renewables are starting to take off, we've been working on renewables for 30 plus years, trying to make them cost competitive, but really, finally, if you look down in the 2019, 2020 area, this is when the utility scale wind and utility scale solar have become the most cost competitive forms of new energy. Now this is across the entire United States, and depending on what region you're in, it's going to vary, but in general, we've seen these tremendous decreases, 70 percent in wind, 90 percent in solar, over the last decade. And although this is for the price – levelized cost of electricity for new plants, and so that's why a lot of new plants that are going in are either wind, solar, or natural gas, these prices are so low that they are actually rivaling the cost of electricity from existing nuclear and gas plants, and in a lot of cases around the country are below actual operating costs for existing coal plants. And this is just a huge change in the industry that is happening,
and we are going to see this continue for the next decade plus, where wind and solar are going to be increasingly used, because they are the lowest form cost of new electricity going in. So what's really driving this besides the economics? The economics is always going to be the main driver of practically everything that we do. But we're also seeing states and utilities across the United States put together clean energy targets. So you can see from this graph here there's quite a number of states that have state clean energy mandates over 50 percent, several that are 100 percent clean energy.
And these green-shaded areas you can see in here, now those are utility service territories. So you have places like Xcel Energy here in the Midwest. You also have places like Duke Energy here in the Carolinas, where the individual utilities are setting these 100 percent decarbonization goals. And so a combination of these has really led to a push in clean energy technologies. Now in addition to that, we're – we have finally seen the new administration come up with these clean energy targets. And if you take a look at what we're talking about, 100 percent clean electricity by 2035, and 100 percent clean energy economy by 2050. And the engineer at least in
me is sitting there thinking, hmm, are – what do these goals really mean? How aggressive are they? If you take kind of a little thought experiment and you say, hmm, 100 percent clean electricity by 2035, and we have 20 percent nuclear, 7 percent hydro, maybe you've got 10 percent clean hydrogen fuels of some sort in the future, but that still leaves 70 percent that is going to have to be made up from wind and solar technologies. And this is a really big number, so let's remember that 50 to 70 percent of the entire system has to be made of wind and solar, so what do we need to do to get there? So let's start taking a look at what we know about running grids with high levels of renewable systems. First off, we'll start on the very small end of the scale here. If you look at the graph here, along the X axis you can see system size in gigawatts, and it ranges from maybe 100 kilowatt type system to megawatts, all the way up to 1 gigawatt here in the middle, and out to 1,000 gigawatts, which is the installed capacity of the continental United States. And on the Y axis we're looking at percent wind and solar. So first off, we're going to look at small island power systems, and two very important numbers to set some context. First, this red square, that is the amount of wind and solar at any particular time. So we call that the instantaneous level. Whereas
these gray circles, those are annual energy, and so that's going to be how much energy is produced from wind and solar over an entire year. But if you take a look at these very small island systems that range from around 150 kilowatts in size up to around 10 megawatts in size, you can see, right, we know how to make systems that run at really high levels of renewables. So how are we able to do this? So these systems, number one, they have very good wind and solar resources. Number two, they also typically have a lot of energy storage connected to them. As an example, you can see the El Hierro
system in the Canary Islands has a pumped hydro system that's associated with the wind plant. But in general, these systems really oversize their wind and solar capacity. For example, this is – on Ta'u Island, this is over a megawatt of solar PV for around 150 kilowatt size load. But we do know how to make these types of systems work. Typically, they only have one device inside of them, being the grid-forming device, and everything else just synchronizes to that.
As we go to larger-scale systems, you can see here on large islands, and here, some good examples would be Kaua'i in the Hawaiian islands, Maui in Hawaii, Crete in Greece, and the country of Ireland, and the entire Irish power system. As you can see here, we get fairly high annual levels, somewhere in the 30 to 40 percent range, and we're incrementally increasing the instantaneous level of these grid-forming sources. You can see 90 percent to 60 percent. In fact, this data is from 2019, and even since then we've see increases in these levels. As you can see here, Maui was just as of this year able to get up to closer to 90 percent, and Ireland also got up around 70 percent on an instantaneous level.
Bus you can see we're still, even as we get larger and larger systems, we're not getting really to 100 percent, and there's a variety of technical reasons for this that we'll be going through in this presentation. Now as we go to larger interconnects, I picked two examples here, ERCOT, which is the Texas power system, and the Western Interconnect of the Untied States. As you can see here, these levels range from 10 to 26 percent annually and then 24 to 58 percent on an instantaneous level, although this is older data, too. And since even this year, ERCOT was up around 66 percent instantaneously. But at the end of the day, what we need to be able to do is look at these large-scale systems, figure out the technical constraints, and get them to operate as if they were some of these smaller-scale island power systems. And this is a non-trivial problem to solve,
but what we want to be able to do is come up with technical solutions to get us to 100 percent inverter based on an instantaneous level with really large power systems. And to do this, we need to really unify the operations of inverter based systems and synchronous generators. So let's do a quick analysis, just kind of a thought study here, back of the envelope calculation, just to see why is getting to 100 percent so important. So let's do a little quick case study. This is using some data from EirGrid, which is the utility operator on the
Irish power system. Again, if you look at their 2018 data, which is what we're going to look at, they were running around 65 percent instantaneously, which, again, this is actually limited by the system operator, which got them to around 30 percent on the annual energy basis. But by 2030, they want to be getting up to 70 percent on an annual energy basis, and 100 percent instantaneously. And why 70 percent? Well, that's on your way to
100 percent clean energy system, and if you remember from an earlier slide, for the United States, probably up where we need to be, in the 2035 time period. So let's just take a quick look at this example to understand why getting to 100 percent is important. What you see here is the wind data from 2018 in the Irish power system, and you can see they're limiting this level of wind on the particular system to 65 percent. This is actually limited by the system operator themselves. But you can see that for the most part the wind is going up and down, crazy, throughout the year, going from January to December.
Now what happens when we start adding in enough wind to get us to 70 percent on an annual energy basis? Remember, this was around 30 percent in the red line here. The blue line gives you around 70 percent on an annual energy basis, and back of the envelope, we just quickly multiplied these numbers by around 2.5 – or 250 percent, or 2.5 times the amount, and that would give you enough energy to provide 70 percent of the annual energy needs by wind. But you can see all of a sudden we are getting way over 100 percent at any particular time. Now this graph is pretty crazy to look at, so let's reorganize this data a little bit to make it a little simpler. What we're going to do is create something similar to what especially people in the power system know as a load duration curve, but we're going to create a variable renewable duration curve. So we're going to resort this data from high to low.
As such, we're going to slowly move all the data around. And then we create this curve, right? We've resorted all these blue data points into this curve, which is called a variable renewable duration curve. But it really quickly helps you understand some important parts of the system. So now what I've done is I've transposed that 70 percent renewable duration curve on here, and then the 30 percent or 29 percent duration curve. And if you take a look at this, some important things come across. First off,
okay, so this amount of energy is provided by wind, which is 30 percent of the area under this flat line. And the rest of this has got to be provided by other sources, whether it's coal or natural gas or whatever resources you have available. But at the same time, you really – you can't go over 100 percent, so you're having to basically curtail all of this electricity here.
Now if all of a sudden we were able to install up to 70 percent wind, we wouldn't magically get that much energy, because if, for example, we're still limiting the instantaneous level to 65 percent, we would only pick up this amount of the curve here. And so now instead of 30 percent, we actually get up to around 45 percent on an annual basis. And still, you need other sources to provide the rest of that energy. And we had to curtail all of this wind here.
Now if we think about what we really want to be able to do, what we want to be able to do is come up with some type of technical solution to get us to 100 percent, because all of a sudden that frees up a really big chunk of this annual energy that we're currently curtailing. So if all of a sudden we were able to go to 100 percent, instead of 45 percent, we're now up to 58 percent, and that's getting us a lot closer to this 50 to 70 percent overall need on the system. Now of course, I'm just doing it for Ireland here, but if you do this for any particular system, you will get something similar to this, where going to 100 percent instantaneously will allow you to pick up the most cost effective portion of electricity. Now you can see we're still curtailing a lot of electricity in this scenario, but at the end of the day, that has to be solved with things like energy storage, increased transmission, or demand response or moving load. But the biggest and cheapest and quickest way to access the most renewable energy will be through getting this to 100 percent instantaneous levels of inverter based resources. And that's really why I'm saying that this is the most critical problem to work on in the short term.
Longer term, we're going to have to solve all these additional problems to get and move this excess energy over into this portion here and reduce our dependency on other sources. Okay, so that kind of explains why we want to try to get to 100 percent and why that's so critically important. If we take a look at inverters and machines together, let's understand why this is a challenge and what we need to address. So if we take a look at currently operating large-scale power systems, for the most part, you have – they're based on synchronous generator technology, so whether it's hydropower, coal, nuclear, natural gas. All of these have a synchronous generator inside of them. And when we electrically connect them together, so you can imagine this being a transmission line connecting two generators together, we create a large virtual power machine that is all synchronized together and outputting an AC waveform, whether that's at 50 Hz like in Europe, or 60 Hz in the United States.
When you connect induction machines or inverter based technology, they don't necessarily have a physical connection that is created by these synchronous generators. So you have a somewhat looser connection, if you will. But what we need to do in the future is as we develop PV and wind technologies and inverters, that they sort of replicate what's going on in synchronous generators, although they don't have all the physical characteristics. But if we can come up with sort of rules of the road on how to unify the operations of synchronous generators and other synchronous machines and inverter based resources at any scale, that will enable us to either run at 100 percent inverters, 100 percent synchronous generators, or everywhere in between, which is what we're going to need in future power systems, with wind and solar resources varying throughout the day.
There are a variety of technical challenges with inverter based technologies, and this particular consortium is going to focus on trying to address a lot of these. I'll highlight them here, even though I won't go into massive amounts of detail. We'll be doing those in future parts of this seminar series. But if you think about the system, we also – we always need to be able to maintain frequency stability and voltage stability. So those will be critical aspects. Now although we can do that actually currently with grid following inverter technology by adjusting the control algorithms in there to actually what we call respond fast, or have fast frequency response characteristics that would emulate some of the inertial characteristics you would see in synchronous generators. But there's additional considerations that need to be taken into account, things like system protection. So I often use this graph as a way to just quickly understand why system
protection is so important, and what inverters bring into a system and why it's a challenge. If you take a look at this graph here, what we're assuming here is that a fault happens on this system here at basically time zero, and when a fault happens on a grid with a synchronous generator, you get some output characteristic of the current that looks like this. So it's maybe six times rated output current for several cycles as it decays. But this huge spike in current very quickly allows protective relaying devices or fuses to recognize a fault has occurred and operate properly in order to isolate that part of the system. When you deal with inverter technologies, as you can see down here, you can have several response characteristics. So one response characteristic may be this red line, where the inverter sees that spike happen, and actually reacts extremely quickly, and turns itself off in a quarter cycle, and then just goes flat in terms of current output.
Another thing, it could see that short circuit and actually keep providing current into that fault, but roughly at around its rated power level, for seven cycles, and then drop off. But the challenge here is that neither of those curves looks like this blue curve. And so it's much more difficult to detect if you have a system with high levels of inverters what the actual fault currents are. So we need to look at a variety of different ways to protect the grid in the future when it's running with high levels of inverter based resources. Another important thing that we need to work on is this grid forming capability, which is the whole reason for this consortium that we're talking about. But in addition to those grid forming capabilities, the ability to black start a system, so bringing it up from scratch, from a blackout condition. Historically, we use either hydropower
plants or natural gas plants or something that we're starting to help restore the grid power, and then we bring it all back online. But with inverter based grids, we need to look at having that capability of black starting the entire system, including providing things like transformer inrush currents when you're restarting those, or line-charging currents. There's additional issues around the various control systems that are implemented in inverters, and potentially interactions with those, as well as cyber security concerns when you have so many distributed technology devices. Let me just real quickly go back to this black start capability, because I wanted to highlight an example of grid forming technologies. This is actually an example of not even a project. This was actually a necessity that we had an NREL, where we were able to run our building loads on a large PV plant plus a wind turbine plus a battery. For over 72 hours we provided these building loads
in this 100 percent wind-PV-battery grid during a site outage that we had at our Flatirons campus. But we were able to actually connect all of these together and operate this type of system, and were able to demonstrate black start capabilities. One of the unique things in this particular system was the way that we were developing the controls so that they would actually maintain stable operations, even though you had pretty wildly fluctuating wind and solar output. You can see down here this orange line is the solar output, and then the green is the wind output from the turbine, and then the battery system was either absorbing power or producing power to maintain this nominal building load that you see as this steady state condition in gray.
But even though we were able to demonstrate this, one of the things that we would like to eventually do is have all of the inverters that were associated with these different components actually all be independently grid forming and compatible with each other. In this particular case, we were able to run this system because we had one device, the battery system, that acted as the grid forming and grid reference signal in the system. So I've explained a little bit about what's going on in the power system industry currently right now with renewables. I've explained what we know about how to operate grids.
And I've explained why getting to 100 percent inverter based grids is critically important. But let's talk about what we really are starting here, which is the UNIFI Consortium, a way of bringing the industry together to solve all of these challenges and enable these types of systems. So our vision for the future is future power systems that have any mix of machines and inverter based resources at any scale that enable this affordable, secure, reliable, clean, and resilient power grid. If you take a look over here, you can see a recreation of those graphs I was showing you earlier, but what we need to be able to do is address these fundamental challenges that are limiting these types of large scale grids for getting to a near 100 percent system, and basically create a way to seamlessly integrate grid forming technologies with the power system of the future, and get these systems operating at as close to 100 percent instantaneously, which will enable the annual averages in terms of renewable technologies to get above 50 percent.
So inside the UNIFI Consortium, we plan to conduct research and development, demonstrate these grid forming concepts at scale, develop best practices and standards, and train the next generation workforce for how the future power grid will operate. So let's just think back in history a little bit, and we're really not doing something entirely new here. If you think and go back all the way to the early 1900s, when the power system was just forming, we're really standing on the shoulders of giants. You take a look here at the precursor of what became the IEEE, and people got together and really started putting together presentation material about how to start to parallel all of these synchronous generations that were popping up all over the country. This interoperability between all the synchronous generators really led to the interconnections that we know today, and that operate and provide power across the United States and around the world.
And it's really the sustained engagement of researchers and practitioners – it was super critical for how we can interconnect generators to realize this future power grid. And UNIFI is going to provide this platform for engagement to realize the grid of the future, where we're integrating inverter based resources and synchronous generators completely together. So what do we plan to do inside this consortium? If we think about it, we're going to unify the old technology, the synchronous machines, and the new technologies, with grid forming and grid following inverters. We're going to unify local controls and global controls across the system. We need
to unify how these things operate at both slow and fast time scales to be compatible with each other. We also need to unify how large scale inverters and aggregations of these technologies integrate with much smaller, because inverters are unique. They can be all the way from 200 watts up to several megawatts. And finally, we're going to unify how solar, wind, and storage technologies integrate with the electric power system to create this clean grid of the future. I want to highlight the partners that we have. A lot of you are on the phone here today. but
as you can see, we have a huge collection of people spanning the national laboratories, the universities there in the middle, and a huge number of our industrial partners, both on the inverter manufacturing sector as well as utilities and system operators, as well as the industry labs and real time simulation and software vendors. But we need this huge cross-section of the industry to actually make this a reality, because what we're trying to do in this entire consortium is change the fundamental way that the entire power system operates. So I'll talk at a high level about what's going on in this particular consortium. We have it divided into three main thrusts: research and development, demonstration and commercialization, and outreach and training. And inside each of these thrusts there are key research
areas – let's just look at the research and development one. Modeling and simulation, controls, hardware, integration and validation. And we have several leads across this space. These are not the only people working in here. We're actually tying in all of the partners that you saw on the previous page to contribute into each of these things as they get developed. In addition to the project partners, which are partners that are either receiving DOE funds or contributing cost share, we have an additional set of industry partners that we're reaching out to across the entire scope. So additional sets of utilities, additional vendors, additional software
providers and system operators, as well as other consortia that are in this space, in power systems or in power electronics. But we are looking at a kind of connection between the power system world and the power electronic world in a way that allows for this entire grid to operate smoothly. The other thing I want to mention is the geographic diversity of the partners that we have on this project. All across the United States, you can see little dots here that represent different
locations, but it goes all the way from Alaska to Puerto Rico, from the Hawaiian Islands to New York and the New England States. But in addition to that, we have strong ties to international organizations around the world, including the Global Power System Transformation Initiative, as well as several of the research institutes around the world, because everybody is kind of going through this transformation simultaneously, and we want to make sure that we are making best use of those interconnections, both internally in the United States and externally around the world, to make sure that we as an entire world can make this transition together. I will highlight a couple of the things that we plan on doing in this consortium, and a lot of this will be, again, further talked about in future presentations in this particular seminar series. But we're establishing the US as the leader in this space across the PV, wind, and storage, for cutting edge research in grid forming systems. We want to foster this ecosystem that ties together
the R&D, the commercialization and demonstration, the outreach and training in this space. The other thing that's most important and a piece here that I'll talk about on the next slide, also, is around the idea of developing these interoperability guidelines at a system level, along with the functional requirements at the individual inverter level. And these will be critical for how we actually maintain this operability across the entire system. In addition, we want to cultivate this inclusive culture. As we've opened up this particular seminar series here, we want to invite more and more people into this sort of family, to listen in around grid forming technologies, to create this cooperation around sustained innovation.
And we want to convene this continuous collaboration between the inverter manufacturers on one end and systems operators and utilities on the other, to bridge the gap between power system and power electronic industries. So let me just talk a little bit about these interoperability guidelines and functional requirements, because this is kind of the key piece of this consortium that will be developed over the next several years. Even this year, we'll start at this, and then through research, demonstration, and collaboration, we'll be able to eventually translate a lot of these type of breakthroughs into more standardized products and standards. So if we take a look at interoperability guidelines, what we really are focused there on is how do we promote the coordination and seamless operation of many different grid forming technologies from multiple vendors while ensuring stability and reliability? So this is the key thing here, right? From a systems level perspective, they just want all of these different technologies to connect, make sure the system is stable and reliable, but from many, many different vendors that are going to be providing inverter based technologies. So we'll be looking at how do you develop scalable secondary controls, how do you maintain system level stability, how do you look at voltage and frequency regulation, black start capabilities, and ensure cyber secure communications across these areas? When you go down to the functional requirements from individual grid forming plants or individual inverters themselves, how do we make sure that we're specifying how these requirements operate so that they are developed in a vendor agnostic fashion and satisfy all the system level requirements coming from the top level system? And inside there, you're talking about real time control with dynamic protection, autonomous primary control signals, how do you manage the input and output from communications signals, how do you look at aggregations to maintain stability, and what type of power quality and protection requirements are needed? So this is really the heart of the consortium itself, and really, this is going to be – to develop this, we're going to need the entire industry to get on board and move this forward. So just to sort of start wrapping this up a little bit, as you saw here, our UNIFI Consortium, starting off with 4 national labs, 12 universities, 6 inverter manufacturers, 8 utilities, 2 system operators, 3 real time simulation and software vendors, but the reality is we are trying to get the entire industry connected to this project. So we have over 80 letters of support on this particular
project, and we'll be reaching out to many more parts of the industry as we move this forward. This particular group of people was the initial development around droop control and virtual oscillator control that you may have heard so much about, if you're in this grid forming space. We've got lots of experience in the past doing large demonstrations of inverter based resources, providing grid services. The team has extensive experience in the standards organizations and leadership roles in IEEE 1547 and the new P2800 standards. And we've
really tried to put together a broad coalition of the world-renowned experts in this space. The key goal here is to unify the integration of machine and inverter technologies, and as we've seen in the past, interoperability drove the interconnections that we have today, and interoperability will drive the innovation of the grid of the future. And we think that UNIFI can be a key driving force in the US power sector and renewable energy industries to achieve this 100 percent clean power by 2035. So our first step in this entire journey has been with you here today. We've just kicked off
our seminar series. As you can see here, more information will be coming out about all of the follow-on talks in this particular series. And then we plan on continuing this past the fall into next year and beyond as a way to get people to talk about grid forming technologies and learn from each other in this space. So I want to thank all of the additional speakers that have signed up for this fall, and look forward to hearing all of these future presentations. And with that, I'll say thank you, and we'll take any questions. Thank you.
2021-09-14 16:51
