We're a digital substation team looking at, how do we increase the ability to use data and communicate between equipment to do things more efficiently, as well as support the increased amount of renewables and other changes in our energy landscape, which are happening quicker than they have in a long, long time.
Mark Thompson: I’m Mark Thompson, I am the director of the digital delivery group within electric system engineering at National Grid.
Robert Stoner: Welcome to the podcast, Mark. It's great to have you here. We're all laboring away at MITEI in the energy transition, the transition away from fossil fuels that the grid was constructed to support, to one that uses renewables. Of course, that's enormously attractive and we're making progress, but it creates challenges. I'm looking forward today to talking with you about what those challenges are and how they can be approached using digitization and other techniques. Maybe before we get going, could you orient us a little bit in how the grid works and where National fits into it?
MT: Sure thing. National Grid is an electric transportation and delivery company in the Northeast part of the U.S. We're responsible for the transmission of the wholesale power. These are the big power lines that you see on the side of the road as you drive down highways. They transport power at a bulk level, traditionally from the large generation plants, whether they're coal, nuclear, or gas to local cities, municipalities.
The other part of National Grid is the distribution companies. Distribution companies transform the power from the bulk system to the smaller poles and wires that you see within neighborhoods. These connections in the distribution system are made to individuals' homes as well as industries and businesses. Within National Grid, we have our transmission organization for the bulk power. There's a substation group, which is actually the node between the transmission system down to the distribution system, which delivers power to the end users.
RS: You're involved in transporting energy from power plants to us through this double layer system. If we talk about renewables, especially solar and wind, which are so-called non-dispatchable resources, they come on when they come on, they go off when they go off. You can switch them off, but you can't make them come on when the sun's not shining or the wind isn't blowing. How do you manage that on a grid? That's basically built to be one-way carrying power from a power plant to somebody's home and where you have a lot of control over what's generated historically throughout the day as demand fluctuates. What do you do when you can't control it?
MT: As you mentioned, conventionally, over the last hundred plus years, there's been a one-way flow of power. That one-way flow of power has been built into our systems. There's engineering involved in these systems, which make sure that we can provide safe and reliable power. In days like today, when there's a snowstorm, there's always the opportunity for something to go wrong, a pole to fall down or not. When that happens, how do we make sure that this entire grid, which is one big balancing act of the power being created versus the power being used, doesn't swing off balance and fall over? We have intelligent systems within the grid that protect us. Sorry, this is a lot of background here.
RS: We need it, keep going.
MT: These intelligent systems continuously monitor the power grid. Monitor how much power is being used, how it's flowing, making sure that the equipment isn't over capacity. Similar to someone's house, where you have a breaker panel. If you plug-in too many applications, too many devices into an outlet, you could pop a circuit breaker. Same thing happens at a bulk power system and at the distribution system. What gets complicated now, is if you have the power strip plugged into your wall, traditionally you're just receiving power, but what happens if you start adding generation to that mix as well? That breaker in your house is sized. It doesn't matter which way power is flowing. It will operate once it hits 15 amps, 20 amps, or whatever it is. You can offload some of that, the device power usage by having generation. Same thing happens in the distribution system. If you have a PV on your roof, you see a large-scale wind farm someplace else, that could offset load.
What gets challenging though, is that there's intelligence within those production systems. It's a little more complicated than that breaker in your house where it doesn't matter which way power is flowing. At large power amounts, the engineering is more complex and there's a lot of directional components there. As we increase the amount of distributed power, whether that distributed power is from PV, solar, biomass, whatever it is, the complexities in how the power flows in the grid increases. It becomes much more challenging to manage this in order to provide a safe, reliable power system.
RS: Some of that must be happening because the renewable sources are closer to the consumer than the substation. In other words, they're in the distribution system, so that you've got reversals going on there that you can't control.
MT: That's a great point. The transmission system, honestly, the bulk power system has been designed for two-way power flows since the beginning. You have these large plants connected at these precise locations with load centers all over as a plant goes up, it goes down. It's been designed by two-way power flow. When you get down to the distribution system, we have not conventionally had generation out of the distribution system, or very little. It's really been like the home analogy with the circuit breaker panel there. As we're adding more renewable generation, it's changing how we look at these systems, how we protect these systems.
We don't want to have a backflow onto a power system, which could cause an injury or damage equipment, which will take out customers. Still, there's controls that need to be placed into service at these substations, which are really the nexus of these links, as well as on our distribution system.
A term called smart grid is something you folks may have heard about, and this is the intelligence controls and communication between these devices. That's really where my group comes into play. We're a digital substation team looking at, how do we increase the ability to use data and communicate between equipment to do things more efficiently, as well as support the increased amount of renewables and other changes in our energy landscape, which are happening quicker than they have in a long, long time.
RS: What sorts of things do you have to change because, say, a solar array comes on in the middle of the day when sun is shining in the sky? What do you have to do differently than you do in the fossil or one-way grid that occasions the use of these digital controls?
MT: Yes. Another component with the renewable energy, or distributed generation generally, is the intermittency of the supply. The power is not constant. Even though it's flowing out two directions, it’s also flowing intermittently. It could be flowing at a hundred percent one minute, and then an hour and a half later, it says zero. These load flows are changing throughout the day. That's completely new for us. The systems within the substations, which are acting like, I mentioned, the breakers in someone's home panel, were originally designed for static conditions.
These variable conditions now are requiring some flexibility and automation that was never required in the past. Right now we're really bumping into the limits in how much renewables we can put on some parts of the grid because of our requirements of these static conditions. There's some variability everywhere. There always is, but if we want to go above a certain percentage of renewables on our power system, we need much more flexibility. Systems like ours, where there's this digital substation system, we can automatically, we're setting up the stage and we're not doing it right away, but we're setting the stage to automatically analyze the generation, the load components, and then make intelligent decisions in how to protect and monitor the grid in real time.
RS: Where are we in converting our grid from conventional resources to renewables on a percent basis in New England in your service territory?
MT: I think in doing that, we're actually ahead of the curve a little bit in some regards. We have some very ambitious state target. National Grid itself is targeted to have net-zero carbon emissions by 2050. We're well along the way to achieving that and there's many components to that. The influx of renewable generation is going to be key for that as well as other alternative energy sources.
One thing to take into consideration when we're looking at the power grid is it's a regulated entity. In the past, it's been considered a natural monopoly. In such that there's a lot of physical barriers to entry. You need to construct these large power lines, in the past large power plants as well. So therefore it's been regulated. That regulation is done at multiple levels. There's the federal regulation, which some of the, like the FBI, want to take the law enforcement analogy, which is monitoring interstate commerce. FERC, exactly, the Federal Energy Regulatory Commission.
Then the state level, there’s a state regulators. The New England PUC, the New York state public service commission. They are responsible for the distribution systems within those States. Where I'm going with this is, in New England, we have some of the most ambitious renewable goals at the state level. It's really an exciting time to be in the industry. New York, for example, is pushing up something called REV, Reforming the Energy Vision, really trying to put the consumer in the middle of the future of the energy industry. How do you enable these distribution systems that can be interacted with by the consumer and also encourage the increased amount of renewable energy? That's a long-winded response to your question.
RS: Are we in this 10%-15% range now, on your system of renewable? Not non-hydro renewables, by the way, because you could control hydro.
MT: Yes, actually, that's one thing. In New York, there's a lot of hydro renewables right now. Niagara Falls, of course, is one of the first ones, as well, some of the ones in the sea way. I'll be honest, Rob, I don't know that percentage. That sounds about right. What's interesting is, as we're increasing this, we are getting to the point now where, 10 years ago, we did not really have these grid stability challenges. We're starting to see that now in certain regions.
This is where digitalization is helping us, especially on specific feeders. A feeder is one circuit coming from a substation supplying several streets or a section of a city. What we're finding is that we're reaching penetrations of renewables on that feeder, which are above what the current systems can handle. We're unfortunately having to discourage, we're not able to connect as much renewables on those circuits as maybe we would like to ideally, especially in rural areas where there's land. Land is available for, say, large-scale PV installations. We're having issues there. That started coming to the play on the transmission level as well. In Northern New York and Western Mass, there's areas which could handle our prime areas for large-scale renewables. If we want to meet some of the state's targets, whether it's 50% renewables or more in the next couple of decades, the infrastructure definitely needs to be increased in order to support that.
RS: Let's just unpack that a little bit because I'm imagining that I'm living in a rural area. In fact, I am living in a rural area right now. I've got neighbors. I might myself have solar panels on my roof. I might have a wind station down the street and suddenly there's a sunny windy day. I and my neighbors aren't consuming all the electricity that's being produced on that feeder. That means that you've got to deal with it. It's coming back to you at your substation and you've got to try to redeploy that throughout the system. Is that the fundamental situation here? The reverse situation as well, where you're counting on them to be producing a certain amount of wind or solar on a given day and it doesn't happen and you've got to find the power to get to them, but how does digitalization help you solve that problem?
MT: Both of those are two instances that we have to accommodate. We have some very smart engineers and planners looking at those scenarios. Before we interconnect any sort of large scale renewables, those are the considerations we have to take into place. Today based on certain percentages and the existing equipment and service, there are some limits onto the capacity that can be installed there in order to, one, just really allow a more reliable power system.
What digitalization allows us to do is... taking a step back here. At the substation, there are these intelligence systems, which in the past, for the last 100 years, were originally magnets and coils and things like that who could monitor the power that the electrical wave forms. Then make a predetermined decisions based on that. They could detect, did a line fall over? Is there a short? Let’s isolate that short so that it's not a wide-scale blackout. Over the last 100 years, we've had progressively more stringent reliability standards and requirements after several of our large-scale blackouts have required that. Those engineering practices and those devices have protected the power system, but that protection is based on the generation and the consumption of that power. There's certain parameters it has to meet. If it's outside of the parameters, it's because it was assumed something had happened that was not good. That was how it was in the past.
Where digitalization comes into play is replacing these devices that have communicated in a very analog needs. They've had miles upon miles of copper cables, tying them together to communicate whether it's power system, measurement values, or reactions to that. By enabling more of a network-based communication between these devices, and also having much smarter devices there, we're enabling a much more flexible power system. What that means now is that, at that generation changes by 50%, 100%, but it changes and we can still see it's because of generation coming out of load online rather than losing load customers, now we know that it's a normal event, that an unexpected event, and therefore we're not going to trip the circuit. Therefore we can allow greater numbers of generation, or distributed generation, onto these circuits as well. It's really the flexibility. It's the optionality, which it gives us which we had had in the past.
RS: You've gone away from a fault-based operating scheme to one where you have a continuum of false, if you like, or conditions that you just have to deal with.
MT: It’s still fault-based but now we’re smarter. We can know if a fault is not a fault. Is it really just a normal operating condition, which is outside the initial parameters?
RS: How computerized is the process? Digitization implies that there's a big computer out there someplace making decisions. Is there still a bunch of people sitting in front of computer terminals, deciding things on the fly?
MT: I'm going to start, first of all, I'm very proud of where we are right now. National Grid just installed one of the first fully-networked systems protecting live line power systems in the U.S. This is the first of its kind. But that said, the industry itself has slowly transitioned from these electric mechanical devices. Really that started at, say, the 70s and 80s. Started using these microprocessor devices. These microprocessor devices operate similar to the older devices, but now instead of having 20 unique devices, you’d have one purpose-built computer, if you will, or even before that transistor digital machine. They were still very much interconnected by dozens upon dozens of cables. There really was no network technology, and they were also set by engineers. There's these specialist engineers sitting within an office who are planning what the power system will do in different scenarios.
Based on those circumstances, they input settings into these microprocessor now based machines. As we move forward, what we're enabling is these machines, these purpose-built computers, are now, we're going from this physical architecture to interconnect to them to more of a digital network-based architecture.
It was an MIT connection with Robert Metcalf, where he said that the value of a network increases exponentially based on the number of nodes and that's true here as well. The more devices we have talking to each other, the more power we have in the future to, one, protect the system in regards to these now, what once was abnormal power conditions, these renewable generations, distributed generation, to making that more of the normal so we have flexibility with a dynamic power grid. We can also make much smarter asset health decisions. We can know proactively if a transformer potentially could fail and as well as make it monitor the grid in a much smarter way than we have in the past where we had to use rule of thumbs or had physical maintenance going to the site, taking samples. Now we're continuously monitoring its performance.
RS: Do you control me, Mark? I’ve got a solar array on my roof and I'm generating power and the sun is shining and I'm happy. Do you have any say over whether that power flows out onto the grid or not?
MT: No, we really don't today. I'm pretty optimistic and passionate about where we're coming from a technology point of view. But if you compare it to other industries, the utility industry has a long ways to go. Within the utility industry, we're digitizing our internal systems. We're trying to enable these future connections as well. When it comes to the customer connections, we have limited control there.
I mentioned New York REV earlier. One of the goals is to make a transactional marketplace where your solar inverter, which is responsible for converting the DC to AC, could be connected to this marketplace. Depending on real-time conditions or marketplace conditions parameters, it could make a decision to charge up your local battery or power directly to the grid or disconnect entirely. We're not there yet. I think there's a lot of great thoughts right now in how do we do that. There are also some pilots in service where we're starting to do this a little bit. Generally, we do have a little bit of a ways to go until we have the full connection. There's a lot of other concerns that we have to think about before we get there as well.
RS: One of those concerns has got to be storage. It's one thing to have a computer that's flipping switches and trying to keep the system balanced all the time and from overloading, but sometimes we just generate more power than we're consuming when we have a lot of wind and solar connected to the system. People talk about adding storage devices and, certainly, there's a lot of activity at MIT to develop new storage technologies, a lot of activity elsewhere as well, of course. Do you have a sense of how far we can get though, without storage? Or to what degree, perhaps, digitization can mitigate the need for storage on the system? Are we anywhere near a limit?
MT: No, we nowhere near a limit. I think energy storage has been considered the holy grail of the power industry for a little while now. It would allow a lot of our aspirations around renewables, a lot of our challenges we have. Take a step back, a minute ago I talked about the balancing act of the power grid. It's not like a water system where we can easily build a tank or natural gas, where we can fill up old caverns. The power grid is really about creating the generation to match the load and balancing this. If that balance goes, if it gets off balance, the whole system can fall down. Energy storage gives us a whole lot of flexibility there.
That said, the technologies around energy storage, as you and I'm sure your listeners know, have not advanced nearly as much as, say, some of the technological microprocessor-based devices have. Within National Grid, we do have several pilots, and we're actually interconnecting these large-scale battery storage pilots to our digital substations and trying to enable this, a much smarter approach it. What do I mean by that? There's a couple of large-scale batteries in our system which are acting as pilots which are communicating back and we're trying right now to gather data about how well are they responding to grid conditions, working with some key vendors around that as well.
To your original point, where are we on the transition? I think we're very early right now. I think there's some parts of our system that could use it sooner than later, but we definitely have a long ways to go.
RS: Sometimes people talk about using storage devices, and other techniques including digitization, as non-wires approaches to upgrading the network. In other words, you install a storage device or a switch or some power electronic device that enables you to adjust power flows in a way that keeps the load on a given line within its existing range so that you don't have to add to it. We've been looking at how do you understand those investment decisions in my group. Are you looking at trying to minimize network investments with these techniques? Is that a big part of what you're doing? Are you just trying to keep the system from catching fire?
MT: I guess goal number one is to keep it from catching fire and make sure no one gets hurt. Goal number two is definitely to drive efficiencies. I talked a little bit about the reduction in the physical infrastructure when it comes from communication infrastructure with the number of copper cables connecting devices together to communicate. The number of devices that we need using these digital microprocessor-based computers, we could have multiple functions in a single box, have redundant boxes, and still have less of them.
Going to what you're saying for the non-wires alternative, that's a really interesting approach as well. I've seen digital substations being used. Actually, I think there's a pilot in Germany right now. They've installed large-scale solar array, and because of the intermittency of that generation, the load on that line was going... it shifted drastically from very high levels to near capacity to very low levels. The digital substation was used to analyze what that looked like. It actually changed these protection systems so that we didn't damage any of the equipment, did not have anything burn down, as you mentioned. It also helps by reducing the amount of added infrastructure there if we can.
Another component of this is monitoring what is actually on the line. Real-time line monitoring can allow us to maximize how much power is flowing across that transmission line. In the past, we made rules of thumbs, we assumed the worst case was a 30°C day. There's so much, as temperature increased, the line would dip. As the load on line increased, the line would sag as well. We really limited how much capacity we could have on that.
Now we can actually monitor, what's the temperature of that line? How much power is really flowing across there? Having more dynamic ratings for that line. That could help save us from installing new infrastructure to support larger renewables or load shifting. These non-wire alternatives I'm seeing being investigated for a variety of installations.
RS: If I were a Russian hacker, I'd be listening to this program thinking, "Well, this is great. I can get involved in manipulating lines and blowing the system up." Cyber security has got to be a big concern for you and a risk to the system. How do you handle that? Are you going through the internet?
MT: First, no, we're not going through the internet. I can say that with some confidence here. But that said, you're right. The first thing that people think about, or one of the first things they think about, is what are the potential nefarious uses for all this interconnectivity? There were two principles the legacy power grid used for security. One was “security by obscurity”, I'm sure your listeners may have heard about that. The other was air gapping. Air gapping is maybe having an intelligent system but not letting it talk to anything else.
I'll address each of those individually. “Security by obscurity” is no longer valid. These systems are becoming way too prevalent. Also, one of the things driving the use of these systems driving the efficiencies and the ability, capabilities are standardization. Industry standardization. One example, if your listeners want to look it up, is IEC 61850. It's a purpose-built suite of standards for communication within the power grid, focused particularly on power grid performance and safety.
That said, this is an industry-standard. You can download the mechanics of how these devices will communicate. There's been instances for example, in 2015 and 2016 there were hacking incidents in Ukraine which brought down part of the distribution system. Tracing that back, the use of IEC 61850 was a contributing factor, as well as some of the proprietary system data being published on the internet. Which really set the blueprints for anyone to see that and do what they did, was take down part of that grid.
When we're designing our digital substation, these are some of the considerations we're taking in from day one. Security is no longer an add-on component. Security is part of the design from the beginning. I mentioned a while ago FERC, the Federal Energy Regulatory Commission. To complicate the regulatory scene, there's a subgroup called NERC, North American Energy Reliability Council, whose responsible for security reliability standards. They set general compliance guidelines, which in the past we've typically tried to meet, or we've not tried, we've had to meet.
We're no longer just meeting the compliant guidelines. Now, how can we proactively design these systems to be best in practice? To be honest, this is one of the challenges I face that, as a regulated industry, most of our compliance and regulatory aspects are based on previous events. When it comes to security, we can't be looking at what happened in the past, we've got to look at what could happen in the future. That's something every single day now, as we're designing these systems, we're taking into account.
RS: Let's talk about standards in 61850 and what that means. Does it simply ask you to operate in a particular way or design in a particular way to ensure compliance so that vendors, for example, can sell their equipment to you and invest in engineering and manufacturing with some confidence that it will be useful? Or is this about protecting you from invasion from the outside? Or what's its purpose?
MT: The original purpose of 61850 was not around protection or security. It was really around interoperability and defining the rules of engagement for how network-connected power system equipment will operate. The power industry is lagged a lot of other industries when it comes to the use of intelligent electrical devices and networks. We're in 2021 right now and we're just now implementing some of the first in the world network-connected devices. Because of the need for reliability and safety, this industry is a little more hesitant to invest in some emerging technologies.
I think, as we talked in the early part, some emerging market conditions such as renewable penetration and whatnot, require us to be more innovative but 61850 allows that the use of network-connected intelligent devices to interop between vendors also defines how do we engineer these systems, test these systems so that whether it's a utility in the US or someone overseas, we can get some efficiencies of scale, we can do it more effectively, more efficiently.
There's a certain level of confidence that it's going to operate the way we expect it to, just like your Bluetooth mouse there's a sub-standard behind that so that you can plug it into any USB computer, same thing when it comes to 61850, it allows any 61850 enabled device if the vendors truly follow it to be interoperable. Security that said, is something we have to be very cognizant of, because the standard itself is not secure. There's security aspects of a standard, but generally it requires additional thought and engineering practice to make it secure.
Another concern of mine when it comes to security is in regards to the disparity between the state and federal regulators. There's some very ambitious and really, really great targets for the increased use of renewables, and some net zero goals at the state level. But these state goals are often focused so much on the renewables that I, personally, am a little concerned about a disconnect when it comes to things like cyber security. Entities like NERC, the North American Energy Reliability Council, have cyber security targets, but they're not effective at the state level. It's only the wholesale bulk power system that needs to be compliant with those regulations. When it comes to the state and some of these really cool, innovative power systems that they're trying to push, there may be a gap when it comes to security.
RS: This is really recognizing this transition of ours is a global phenomenon. There are economies of scale to be realized by enabling a marketplace and devices and stability and practice and training that are enabled by having common standards amongst different utilities in different parts of the world. You guys operate outside New England as well. I know you have some firms in the UK as well. Can you tell us, are we lagging as a nation? They're very aggressive about wind, especially in the UK. Are we on the tail of that or at the forefront?
MT: It's kind of interesting. A little bit about the digitalization journey at National Grid, at least from my point of view, in the digital substation world. Some of this technology, the 61850 standard, has been out since the mid 2000s. We've investigated it in the early, around 2010, 2011, decided not to pursue it, we weren't quite ready for it. 2016 is when we really started down this journey in earnest. At that point in time, I got involved and I started reaching out to some peers.
It's interesting, because across New England, other utilities in the Northeast have actually been using 61850 now for several years, for the better part of a decade. I started also reaching out outside of just the northeast, or even the U.S. There are some very ambitious goals and in some of the Canadian utilities where they're using these standards. Our counterparts, half the company, is also in the UK, as you mentioned. They're responsible for the electric and gas transmission networks there. They've piloted some 61850 systems for better part of 10 years as well.
In mainland Europe, I’ve worked with some of the utilities there, where they're using this as their standard. They've seen the efficiencies in deploying this type of technology. The same is true elsewhere in the world. Where we're not seeing as much of it, although a digital interconnected substation is cheaper than a conventional substation with the same functionality, if you're not as concerned about reliability and keeping the power grid online for 99.99% of the times, you could probably skimp on some of these protections.
In some of the developing world, where they have nothing today, and they're trying to get something there as quick as possible, they're not using the digital technologies as much. Other parts of the world, for example, a counterpart of mine from another utility has been helping Puerto Rico get back on the feet after the hurricane a couple years ago. As they redesigned the grid, digitalization is one of the key tenets. The reason for that is this reliability that we talked about, but the interconnection of renewables there. They want the ability to leverage multiple energy sources. The flexibility that digitalization allows is really necessary for some of their goals they're looking at.
RS: I’m remembering what a mess that was in Puerto Rico. We had an event at MITEI to try to get our heads around it and look for opportunities. My memory was that the major power generating stations were on all on one side of the island, the south side of the island, whereas the industry is mostly on the north side of the island. What happened during the hurricane was that the lines carrying power over the mountains to the load centers on the north were disrupted. There was a lot of interest in trying to enable the system to stay up, almost entirely with distributed resources, and then no access to those centralized power plants. There were a lot of ambitious discussions, I don't remember exactly how far that went. Islands essentially, is my point, are a really interesting case, because unlike the power system that we live in here, where we have all kinds of sources of electricity and a diverse range of assets—including hydro from Canada, solar from southern Massachusetts, and nuclear—you don't have that in an island economy, especially the lack of a large hydro supply. Balancing becomes very difficult when you have a high penetration of renewables. I just wonder how far you went in your work in Puerto Rico. Are you involved in other islands, trying to deal with this very high penetration challenge?
MT: To be honest, I'm not. That was part of actually a working group which we formed. Shortly after we started our program, as I was reaching out to counterparts across the industry, we formed a group of about two dozen of us utility partners across North America, so U.S., Canada, the EU, even South Asia. We get together on a quarterly basis. One of those members is the one that's leading that effort in Puerto Rico. I'm not personally involved there, I have talked to them, and we've shared some lessons learned as far as what we're doing and how they could maybe learn from that as well. I don't really have any experience with the island architecture itself.
RS: We may find ourselves learning from all kinds of island systems in the future as we get further and further into this. Last question, what about China? Are they adhering to the same sorts of standards in China as we are? Are they encountering the same problems finding ways to solve them that we could also reflect on?
MT: Given the current, I guess, geopolitical considerations, our interactions with China are definitely limited. We do see some Chinese products from time to time coming in. We're seeing them using these standards as well. This is a set of standards, they work, so they're using them. That said, there was an executive order that was under our previous administration that was pushed, currently been put on hold, that looked at supply chain, and also investigating, really understanding where it came from. It did not name names, but it did allude to several nation states where we wanted to be very conscious of before we accepted any products. This is fresh off the presses. I do know there was an incident where supply chain, where it was a subcomponent within a subcomponent, there was some security compromises in there. That's something we're realizing now, that it's not just the overall computer you're buying. It could be a chip on that computer that was bought by someone three levels down in the supply chain. We're responsible for understanding what that supply chain looks like and knowing the risks that are associated with that as well.
RS: This isn't your grandfather's grid. Breaking new territory. Mark, it's been great talking with you. I've enjoyed this. I hope our listeners have learned something about how the system operates. I look forward to speaking with you again about opportunities. Thank you.
MT: Rob, thank you very much. It's been a pleasure.