 Now, our speaker today is Bruce Lohman. He is a research scientist in the eutronics lighting and network groups at LPRNL with a focus on the intersection of network technology with energy systems. His work includes early work on energy efficiency of eutronic devices and outlining the role that IT plays in the electricity. His degree is from UC Berkeley with a better degree in architecture and a master degree in energy resources. He has been working with the lab since 1916. Without further delay, I will let Moose share his slides and begin his presentation. Yeah, thank you. I agree with the slide there. I absolutely welcome the clarification during it, but for longer discussion items, try to hold them at the end so we can first get through the slide and then go back one slide. Yeah, thank you. It's a pleasure to be here. As you noted, I've been at LBL for a long time. Starting in the early 90s, I worked on electronics energy use, computer, televisions, network equipment, and towards the end of that time realized that being connected to a network was a really important part in driving energy use of electronics. Like, particularly when we had desktop PCs, commonly they would be left on 24-7. So I realized I needed to get into network technology to understand the reporting. I started getting to looking at network technology to understand how we could adapt and change and use network technology to reduce energy use. And that got me involved with organizations like IEEE and the IETF, the people who brought you the internet protocol. And while some of my colleagues were doing things like embedding better window coatings or better ways to construct fluorescent light bulbs or working on appliance standards to make it illegal to sell appliances that use more than a certain amount, for electronics, I realized that technology standards kind of played the same role that the law of physics played for a lot of other end users. As a public sector, if I was going to try to influence the market since we don't make or sell anything, getting involved in technology standards was one of the most effective things I could do because technology standards can either prohibit or require energy-saving features, so they're extremely powerful. If you ever want to find me, click on my name to find my page. I'm easy to find. And I welcome many questions or comments after this. So this slide, I actually used several years ago for a different talk to help clarify that this notion of how you see the world really depends on where you're starting from. And in my area of working with electricity, oftentimes most people in the conversation take perspective of the utility grid, which is an important and valuable perspective to take. But I'm a buildings person, so when I think of what's behind the meter, for me, I'm sitting in my house or office, there's the meter and beyond it is the grid, whereas grid people think they're reversed. So it really, people thinking about what's possible and what's desirable gets shaped by their context. And I thought this was kind of funny because I googled around and found other examples of this New York cover and there was one from Stanford. So that was a nice tie-in with today. So for myself, I was born in Southern California in the summer of 1979, my family was born from Southern California up to Palo Alto, where I grew up. And about three months later, there was the first computer communication and distance from UCLA to Menlo Park. So almost the same path at almost the same time. So I kind of feel like I have some visceral connection to IT network technology from that way. I went to high school, Palo Alto High School on land that is entangled with Stanford. When I was, I don't know, maybe second or third grade, I got a new name around CRT, it was pretty cool, particularly back in the early 70s. When I was in eighth grade, I decided that we did time wrong. I liked to question fundamental principles. And, you know, this clock, the digital vegetable clock that had a LED display that showed time and decimal time in fractions of the day. But I also realized it was feasible to change the second as the fundamental unit of time in the metric system. So I moved on. But then I got hired right after that, and when I was in high school, I worked in the building, the style of the press headquarters. Now, at that time, it was Hewlett Packard's research lab. So I did programming when I was in high school. So I have a bunch of connections to here. I then went to Berkeley and never came back. So I'm now pleased to actually be back here sort of returning closer to where I started from. One thing I keep coming back to is that in the terms of the technologies we have, the best things in life are simply universal. Every email address on the planet since about 1992 is exactly the same format as a simple user at domain. And of course, the domain name system was a technology that already existed for reaching the Peter. So it was one simple technology built on another symbol technology, same with web browsing, USD, the power symbol. If you're going to turn something on and off, you're going to probably look for a power symbol, aren't you? And it turns out, fun fact, I am the world's expert on the power symbol. And if you don't believe me, go to Wikipedia, look up power symbol. You'll see that I wrote all the important references. So anyways, and money. Money makes the world around and money is pretty universal. So I mean, money isn't always the best thing in life, but it's really important to have. So an awful lot of problems that I come up with with people is people who insist on wanting to build really complicated systems and complexity as costs. It impedes some privacy. It reduces interoperability. It's just complexity is our enemy, what we don't need it. Now, occasionally we need it, but we should try to avoid it if we can. And something that's occurred to me that hopefully this is an example of where we are, is that in biology, people realize that in the fossil record, you have these long periods of time where the species will be relatively static, slowly evolving, and then you have some events which cause a dramatic change in the genome pool of the animals there, or plants too, presumably. And you get this punctuated equilibrium, and then you have a lot of change over a short transition period of time, and then another static thing. So with communication, we clearly have that right nearby. We invented internet technology, and that took a few decades. And I would say we're now past it. We've literally thrown away the old phone system, not only thrown away the hardware, but also a lot of the ideas that understand the old phone system. And while things do evolve with internet technology, the fundamentals really have not particularly changed. Yeah, we went from IPv4 to IPv6. It didn't really change anybody's experience of it. And other four technologies, TCP and such, are all quite stable. So hopefully with electricity, we are at that cusp of a quick trans... Okay, it was better. To a much better future. So we look at our electrical technology from 1882 when the electricity grid was invented in New York City. The way we generate power has dramatically changed. It would be unrecognizable to Edison. The way we use electricity has dramatically changed. And power distribution inside the building, because again, I'm a building's person, has also hardly changed. Sure, we've got a few buildings with batteries. But the vast majority of the infrastructure in buildings would be very recognizable to Edison if he somehow returned to life today. I just added this other column which I hadn't previously used for sort of the wide area distribution of electricity, which has also not changed that much. And a lot of this is just to the inherent nature of AC power systems. Because the way we construct our systems, it's really hard. And it's really unlike internet technology, where internet technology, because of loose coupling and layered architectures, it's much easier to evolve. So I would say that it's highly time to rethink how we do electricity to move to a much better technology, not just a little better technology, because if we're going to address climate change meaningfully, we have that much better electrical technology that's lower cost, more flexible, easier microgrids, that kind of thing. So the electricity grid today is what I call a unitary system, just like the old phone system was one big system. The phone in my house when I was a child was black, had a dial. It could only make phone calls and there was no local intelligence. It was just a phone. And of course, that's a completely different model from how we do things with internet technology, wherein we have rich infrastructure inside the buildings. And the network in a house or dorm room or office building is loosely connected to the wider internet. It's not tightly coupled. But with the electricity grid, everything is tightly coupled on a 60-hearth frequency. And so it ends up being one single big system, which again is part of why it makes it difficult to evolve. And it's also fundamentally an analog system. So there are a lot of similarities between the old phone system and the electricity grid we have today. They will hold a long, obviously, because they're very similar. And all these characteristics, fundamental characteristics, which then change when we went to internet technology. And so there's this obvious question of what is the electricity equivalent of the internet? Now some people will refer to a phone called Internet of Energy. And if somebody uses that term, you know they either do not understand how the internet works. They do not understand how energy works or both, because Internet of Energy makes no sense. And I'll get to that in the next slide. But clearly there is something which I call a network model of power that we need to do. And the way I've explored this is only inside a building. And whether it applies to the utility grid, I don't know. I don't think it does in the near term. It may in the long term. But I'm only trying to make claims about how things should operate inside a building. But the way things operate inside a building relates to the grid, because we need to coordinate the grid, particularly for demand flexibility, more and more. So if we look at these two systems, we need technology to move a resource from where it's available to where it's wanted. So if the information is on the web server, you want it to move it to your web browser. If the electrons are in a power plant and you want to move to this computer so we can use it, you need to have technology to move things from point A to point B. All data packages are different source, different destination, different content. The electrons are all the same. And I know that because I work for a physics laboratory and the people down in the hill for me tell me that electrons are all the same. Which in a sense actually makes it a much simpler problem than the communications problem where all the packets are different, which we have already solved. So we've done the hard thing. We just need to do the relatively simple thing of managing flows of electricity. So the internet goes over the internet protocol. No matter what your application is, it all goes through the protocol, no matter what your physics layer. So what we really need is we need a fundamental mechanism which can play the same role that the internet protocol does for electricity. And in thinking about that, I only come up with price and quantity as that possible mechanism. It's not like I came with five mechanisms and price and quantity seem to be the best. I actually just can't even think of another mechanism that could possibly be a mechanism. And if somebody else has some nominations, please tell me because in the last 15 years I've been thinking about this, I haven't come up with another. And in a sense that's not surprising because we use price and quantity every day in our lives. It's a basic way that our economy and our infrastructure work. So that's not a surprise. So the internet protocol, that's fundamental mechanism, routes data over multiple hops, data point P, point B because the packets are different. Since the electrons are all the same, routing an individual electron or a group of electrons over multiple hops in a network doesn't make sense because the electrons are all the same. Now you do need to manage the location, the quantity and the climbing of the electrons. So you're managing flows, but they're flows of identical things. They're not routing. It doesn't make sense. So that's why I said concept of internet of energy doesn't actually make any sense. Now there's this additional concept here of capacity. Any type, whether it's a data communication path or an electrical connection has a maximum power capacity it can do. So on the internet, when there's a capacity constraint, things slow down and just take longer to get there. And we have TCP to manage the data rates of packet flows. So that's how we manage capacity constraints in IT networks and electric circuits that we don't overload circuits because you can introduce a safety problem or trip circuit breakers or such. What we typically do with electricity systems is we weigh over build capacity to make sure we're not going to run out and then have extremely conservative consumptions about how we use it because our electricity system is basically built on not knowing what's going on. And it's amazing how little utility grids know about what's going on at the edge of their system make conservative assumptions to make sure we don't exceed it, which means that you're actually wasting a lot of capacity that might be available. So if I plug something into this USB port, it doesn't just get power, it has to ask for it. So there's actually digital management of capacity in devices that we all own today. So we just need mechanisms to scale up capacity management to entire customers. So in terms of how the internet works, it is a collection of both mesh networks and tree structure networks. And the wonderful thing about internet technology is that it doesn't really matter. It's flexible and operates both ways. And that's true on the wider network and it's also true inside the buildings where you can have a tree structured IT network or you can have a mesh network of switches and routers collectively. So it's a flexible technology, which is nice, which means that you can make or break a connection at any point and the system will automatically adjust. Has somewhat similar characteristics, but because it's driven by the laws of physics on the wide area network, has flowing over the transmission system is a complicated topic that I know nothing about where that comes into play. Inside the buildings, you just tend to have tree structures of electricity because with an AC power system, having a mesh network doesn't make a lot of sense. But if you use DC distribution, it actually can. It's the internet protocol that's operating across all of these lanes no matter whether you're sending a photo from your phone to your TV or there's the emailing somebody in Africa. It's the same mechanism in all countries and all building types for all application office. Even if you're not connected to the wider area network, you still use the internet protocol to move data. It's a universal fundamental mechanism and price can have to be that. So there's the OSI model, the seven layers that you can simply simplify to just five. The internet protocol sits in the middle to isolate your physical layers from the application layers and doing so is really essential to making the internet work at all because if we had things that were more complicated or more entangled, it would get so difficult that it would make it much harder to manage it. But it's the simplicity of internet architecture which allows it to be flexible to scale, to drive down costs, make products interoperable. These system architecture foundations are really essential to making a highly functional system. So there's a key notion that you want to isolate the complexity of a land from the WAN and vice versa so that the wider area network doesn't need to know what's going on inside your building and vice versa. You want to separate, isolate the complexity of applications from the complexity of physical layers and that allows you to then evolve systems in a very healthy way. But like how our electricity system works today where we have entangled technologies and that's why we haven't really evolved it very much since 1882. So there's other things and this is certainly not a complete list where on the phone system you had the actual data plane of a switching circuit and you had a control plane that was separate which orchestrated how those circuits were set up. With the internet protocol, those were unified. There's also where the control signals go and there's a whole lot of advantages to that so that when you break a connection on the data you also break connection on the control and so you really know what's going on. With our electricity systems with these things separate from each other you don't necessarily actually know what's going on. This smart grid diagram that somebody created about 10 or 15 years ago I have never seen such a diagram in my work with internet people. Nobody who works with never technology could ever create such a structure like this because it's just not workable. But in the electricity system people are so used to having these things that are organized like that because they've been sort of patched together with duct tape and bailing wire that they think it's okay. It's not okay because it's going to be expensive. It's going to be brittle. It's not going to be highly functional. You need simple systems in order for them to be highly functioned. Another important thing is that you have to throw away ideas when their time has passed and there's a whole lot of ideas that underpin our electricity system which are simply obsolete. They're not okay for the future. Like demand charges. Demand charges and billing are evil and I think that's the only flight term for it. We need to just get rid of it. It's like the electoral college or any number of things like that. We just need to get rid of them. Analog communication is very 19th and 20th century. Everything's, you know, future is digital. These separate and patrolled data planes, unitary systems versus network systems, heavy covers with loosey couples. Deterministic operation is a really critical one. So, you know, the old phone system was completely deterministic. You placed a phone call. It set up a circuit. You were guaranteed this much data capacity whether you used it or not. The internet was completely non-deterministic. You send out a packet. You just hope it gets there. It almost always does if it doesn't be broadcasted. But it's completely non-deterministic. In the 1950s, multiple people went to Bell Labs and they said, there's this thing called packet switching. We think it's the future of communications. And the clever phone company engineer said, we were aware of that. We know it's impossible. Therefore, we do not need to study it. So, they just couldn't imagine because of the way they thought about the world because they only believed in deterministic system that a non-deterministic system could actually work. Not only that, but the deterministic alternative. Now, the weird thing about the electricity system is the demand side has always been non-deterministic. Now, I as a customer, I never tell the utility what it is that I plan to do as far as electricity. The utility just has to forecast things based on past experience. And luckily for the utility, they don't care what I use. They only care what everybody collectively uses within some region. So, since, you know, variation out, then it's all fine. But it's completely non-deterministic. On the supply side, things are very deterministic where you orchestrate with wholesale markets which power plants come on. But a lot of people in the electricity system believe that the only way to have a better future is to make the demand side deterministic. And they are just wrong, just like the phone company engineers who didn't believe that internet technology was possible. And that's that mental problem of people thinking having misconceptions about how we are, I guess, barriers. So, throwing technology away is a critical thing to do. So, I sort of started collecting some things that are these myths that sort of impair people having clear thinking. So, we used to have applications specific physical layer, so communication. TV broadcast signals, you could only send out how TV signals. FM radio or AM radio, they only send out audio over those things. Telegraphy. All these technologies were efficient technologies for a particular individual application. And what we replaced it with is technologies that are useful for all applications, generic technologies. And that's really critical. So, like when electric vehicles come along, people think, oh, we need to have a specific solution for this one particular end use. No, you need to have a solution for customers in general who may have bursty loads, but you don't want to make it specific to an individual device. Mechanisms are always better than the application. There's this notion that electrons aren't all the same, which as we know is non-thin. And so then some people would say, oh, I want to charge one tariff for your electric vehicle charging and a different set of prices for your other devices. That doesn't actually make any sense with respect to this. It's because again, the electrons are the same or that they want to arrange flexibility differently for different devices. What your thermos, what your furnace is using as controlled by your thermostat is different from what your lights and other devices are. It's nonsense. The electrons are all the same. We need universal mechanisms. There's, I already covered this on non-determinant systems. We rely on non-deterministic systems every day for a whole lot of things and that's the future. And then some people for engaging flexibility want to sell electricity at one price through one mechanism and then engage in flexibility like through aggregators or such to make the advanced side operate in a way that's better for the supply side. But they're only doing that because they charge the right through on price in the first place. If you charge the right price in the first place, you wouldn't need to have this separate mechanism to account for the charge the wrong price. When I was in my elementary school, I started learning how to program computers and garbage out. If you have a great algorithm but you get a bad data, you'll have a bad result. If you have bad prices, you'll have a bad result and that should be not a surprise. So, again, as I said, I haven't thought of any other mechanisms that could possibly serve this role as a fundamental mechanism. And while you can do other things in some context, they always break down when you move to other contexts. Well, Dan, and you want to operate off-grid, if you've got some cloud-based energy or trading your devices in your buildings and you don't have an internet connection, you're screwed. But if you're using pricing, you can generate the price locally and your water heater won't even know or care that the grid went down. If you're using prices that's not new, there's this paper from 82, another every paper that used the phrase prices to devices from 2006. And there's this thing I can actually read either to do that when I find time. So, I only came to this in 2010. I came to this much later than other people did. So, the whole notion of using quality retail prices has been around for a long time. I got my start in thinking about this from this photograph, which I know exactly this is very convenient. And later, when you move electricity around in an off-grid context, either within an individual household or between a village, this off-grid, that maybe actually in five or 10 years gets a grid connection. But maybe it's an unreliable grid connection. So, you want to have a ways to move around electricity over space and over time, over time through storage in ways that can be completely self-organizing, can be inherently safe, and can really scale from those small systems of the largest system. And I came up with price. So, that's how I came to it from thinking about what's good for the customer whereas the other people were thinking, let's use price because that's good for the grid. Now, it turns out the price is actually better for both the customer and the grid. But if you think of it from a customer-centric point of view, it gives you a different sense of how you think about it from if you have a grid-centric point of view. So, another critical thing we want to use is we want to manage all power flows digitally just like we manage all data flows digitally. So, USB does this. It has for a few decades now, you own devices that do this. On the same table, you have general and data communication power and critically, communications about the power to manage the flow of power. And that allows you to do things like change the voltage on it which you can do with both USB and with Ethernet. It allows you to do this permission-based capacity management and other things. The key thing is that it's all on the same cable. Well, but they're all free on different wires. With Ethernet, it's the same cable. They're all on the same wires. That's an implementation detail that's confined to the link technology. It doesn't really matter beyond that. The important thing is the same cable. This is, again, where the control and the data plan are unified. They're not separate. So, this is much easier to do with DC power than with AC. You know, basically, it's just not worth doing with AC. With AC, but with DC it is. And so, that's one reason we ended up using all AC because we didn't have power electronics back in the 1880s. But now we do. So now we can actually make much more use of DC than we used to. And so, that's part of my work, is to work on technologies to more easily use DC distribution inside a building. So, we all use digitally managed power today. We just need to scale it up to a higher level. On the utility grid, what a lot of people do is they take utility grid and say, oh, we want to make your microgrid. Let's take our expensive brittle utility technology and make it smaller, which makes it even more expensive and more brittle. But if you take simple technology and you scale it up, just like the first airplane was one person 12 seconds, 80 feet. Now we have jumbo jets. Most technologies, they scale up. You get to a different place, depending on where you start. What I came up with, with somebody else, it wasn't just me, was the cost of power integration where technologies were moving the electricity around inside of a building. Technologies for how devices know what to do, sort of like the on or off, dim, the blue flashing. The functionality is from the application protocols, the same application protocols we already used, so nothing changed there. We just need to have these power distribution technologies, which we have in the form of things like USB and Ethernet, but they're only isolated link technologies. You know, Ethernet or Wi-Fi, you know, they're also lists by themselves because you almost never want to talk to the device at the other end of the link. You want to talk to the device multiple hops away. It's the Internet protocol which allows you to go from a link context to a network context, and that's what price and quantity does is it allows you to go from a link context to a network context. That division between those allows each of them to evolve separately. That's just how things operate on the Internet and that's what allows systems to evolve much better. So one thing that comes out of this is you can network electricity itself. So this could be your, you know, my house or this building or a car. It could be any power distribution system. You could, you can arrange what are called nanogrids in a mesh network or in tree structures. Nanogrids has a memory controller and infrastructure device that's analogous to an Ethernet switch. It's not analogous to an IC router because again, we're not routing Pakistan, analogous to an Ethernet switch. And then you, just to be part of the load and you may have, so you have, you're downstream, you know, local connections and then you upstream connections to other nanogrids and then you have a microgrid controller which then is actually the same as the other ones, but it's just the one that interfaces with the encode example. And some things could be AC, some could be DC, there could be different voltages and you can make or break the power connections whenever you want to and the system will automatically adjust. The grid has its own local price. So how, how long you get a net flow power from here over there? Well, this is a nanogrid controller not only never talks about it does not even know that it exists. In this model, all the communications is only over a, over a direct electrical connection to the watt or the device at the other end of the link. But the value of this might be zero cents, maybe it's 10 cents here, 12 cents, 60 cents. You can have a net flow of electricity over multiple links of communication. So it's a model which is both simple and completely flexible also to when, when a power connection fails or a piece of equipment fails, the system is automatically rebounds the power around it. So in 1969, they built this device here about the size of a refrigerator. I'm sure it would cost millions of dollars in today's money. It's basically a four port ethernet switch. Okay? Now, you can buy them the technology scale because it has these inherent characteristics. You don't find that kind of scaling with traditional electrical technology. So one of my goals has been to build the world's first fully functional nanoroute controller and hopefully early next year we actually actually have that. So, um, how does this actually all work? So in that model there are only four messages for this physical layer distribution between the two things. It turns out the first three messages already exist in USB and an ethernet in devices that you already own. The only thing that gets added is a price and a non-binding forecast of future price. This is really simple and this simplicity makes it allow us to be interoperable and low cost and cyber secure because if you're not sharing information you can't get hacked or you can't violate. So this, there's a huge number of advantages to this model. It could extend anywhere from this building again to an off-grid village on another cost. So key to this then is to separate as much as possible the customer domain from the grid domain. I used to only ever talk about buildings but as I talked to more grid people I realized I should be talking about customers. So they realized the only reason we have the utility riddance is to serve the needs customer. The other way around I thought that they're just wrong. So from the local area network from the building the grid should be just a black box. You shouldn't know or care what's on the other side of the meter and the reverse should be true. The utility should have no knowledge of what devices you own for their status and patrol it should just be about price and quality. And actually another critical thing here is that a lot of people even today will say that you have to exactly which is not true because now we have battery now we have storage. And in order to create internet technology we of course had to have visual communication but we also had to have data storage. The storage was essential to have more loosely coupled systems instead of tightly coupled. And so now that we have electricity storage we can be more loose about how important. We have we can afford use prices at the meter always use prices usually terrible price. Now we can have better prices there and then we can have the screech infrastructure inside the buildings of prices for different power to name one mechanism at all scales. So there's I like to use this term coordination architecture which is ultimately gets down to this is usually between the grid and the customer or the grid and customer device who talks to whom about what. So in a particular way of doing things this and what are they communicating about either providing information or exercising control. So 50 years ago or so there was the right low control where utilities would send out signals over radio or power lines to cycle off water heaters or air conditioners that works. They don't really get any good feedbacks. They don't like it. Then we about 20 years ago 30 years ago a lot of event-based demand response where a dozen days a year you have an event and you took something specific. The other 353 days are quite regular days so you do nothing at all. It works but it's an extremely limited tool. There's prices which is what I'm advocating here. And then there are complicated things with bidding options which you could make work. I think they would end up working about the same as prices except that they add this enormous amount of complexity to the system and cost which could never be justified by over how pricing works. So those systems are just never going to happen. People love to talk about them because they're really cool constructs but they're never going to happen in the real world. You know just like you know flying cars. I drove down here. I didn't say a single flying car around on the freeway even though people are talking about it 16 years. So ultimately the ideal number of ways of doing this globally is one. On the internet engineering task force where I've worked collaborating for several years if there's a technology solution like a communication protocol for a problem that works you do not develop a second solution for the same thing. You just use the one. The key thing is simple because they understand the value of simplicity. So a lot of times people talk about grid services and talk about the various ways that grids and customers might interact and they're always unsatisfying conversations I've been in hundreds of these because they're mixing things together that actually don't really have much energy. We can deal with through pricing. There's coordinating with inverters about power issues power quality reactive power frequencies and such. I'm not an electrical engineer. The inverter people seem to have that well in hand. It's a separate problem and needs a separate solution. And there are solutions so that we don't have to complicate our discussions about energy just by polluting with these power discussions. And then this last part about managing capacity we typically don't do but we absolutely need to for vehicle charging in Australia they are using it today to manage excess solar from feeders where you have a whole lot of customers with PV panels. Sometimes it was having somebody was that I want to put PV on my house and the grid would say you cannot because a few hours a year there'll be too much power coming back from all these panels. So now what they're doing is they're broadcasting out an interval limit each customer can export to the grid to make sure we don't overload the feeder. It's not a limit on what the PV produces because it's the PV minus whatever the loads are. So if you're in danger of exporting too much you could charge your PV or you could just increase your load you could charge a car you need some water things like that. So capacity management is something that absolute utility needs to be working on. And but again it's it's a separate problem and there might be prices associated with this but it's separate. So what might these prices look like? These are some real prices that people were charged on those like about eight nine years ago. In Illinois now they don't vary that much. So in this case it's on your customers. So not much dynamic range so it doesn't give you a lot of opportunity to save money. The more the prices range and why they're talking about maybe 40 cents difference between the high price time in the day and the low price time make it really worthwhile for customers to buy technologies to shift load. My heat pump controller for my heat pump takes in prices today and optimizes to them. Now I only pay to you prices because I can't pay an hourly price here but once that gets charged my the device I already own my house can can operate do that directly. That's what we need. We need to start distributing technology that's ready for the future. So I never use the term real-time price in this concept. Real-time price is a wholesale market construct. It's about wholesale markets it's not about retail markets. If you go to the store to buy tomatoes or bananas that's a retail transaction. You don't know what the wholesale market for bananas or bananas are. That's a separate thing. Now there's some loose relationship between two but they're only loosely coupled or not probably coupled. And if we went from flat prices to a retail price that was calculated from a wholesale price we'd be better off but that would not be the right price. It's a which is the one that's actually best for the grid and best for the customers than something that's calculated directly from a wholesale price. So high-economic prices the time interval of the price is between hourly and five minutes. I've worked on this area for the last three and a half years. Zero people have a time that they think we need prices that are fashionable inside them. Nobody thinks that. And it doesn't make sense to have intervals less than ever. So that brackets that. You want to set it no further than every day because if you don't they're not responsible for the actual grid conditions and then they're not the right price. Again it's garbage and garbage out. Now this doesn't guarantee it's the right price but you have to have these if you have any. And we will no doubt start with prices that are announced today in advance and then over time as people come more comfortable with it have prices that are day of. You know like weather forecasts these days are pretty good. You know you can look at the temperature for the little temperatures will track it pretty well and the same will be true of electricity prices. So people will relax about the distinction between prices that are guaranteed versus prices that are just forecast. You also want to set it to health. Well it's probably one of the different for what the utility sell versus what they buy back. Maybe they sell like 20 cents buy a back of 10 cents. That's done in some places that's being proposed for fuller customers in California. There are even score against it but it's legitimate full box. You also want to merge on greenhouse gas emissions so that customers who want to use that can. What you do is you pick your dollar per ton figure. $50, $100, $1,000 a ton multiply it by the kilogram per kilowatt hour figure this broadcast. You add that to the retail price and then you optimize to this hybrid combined price because since climate change is real the retail price by itself is actually the wrong price because it doesn't take into account the real cost of carbon. And every so often you have financial alert, if we're going to see power shots wildfire but that's it. There's no other information you need to do for energy. And this is my model for how things to work in buildings from a building person. You have a price server that's like a web server. It doesn't compute any information that's distributed. The California Energy Commission has one operating today. You can go query to get prices. Today it's only things like time. But when we start getting hourly prices it'll do that too. We have a research price server for a project that we're going to work on. And then we read some of the algorithms for the price for the DHB and the emergency alerts. Some device down here translates from price to functional control to turn a compressor on and off or change the level of light. So the functional control signals of the orange one. So anything for devices to provide yourself or some central devices in the building or some cloud-based entity can do this. You might wonder why would you send the price from the building out to the cloud instead of sending ground from here because this is a local price. The value of electricity in the building may diverge from what it is on the grid. So if you're not optimizing for the local price you're optimizing for the wrong price. So this I drew this about three years ago and it has one bit since I originally drew it which makes me pretty confident that it's some good I don't know I'll try to stand for a while here Let's see. So these heavy-damaged prices have a lot of benefits. It's simple improves the privacy because the utility doesn't know what devices you own it only observes. And you can actually layer other mechanisms on top of the price but if those other mechanisms fail or are not present the system will always continue to keep operating. And a lot of cases we have aggregator models where utilities will pay like Google or Ecobee to manage thermostat they typically take half of the value of the flexibility that gets engaged. Now I spent thousands of dollars at a time to put a heat pump and a heat pump controller in my house and if I lost half the value of flexibility to somebody else and didn't get it we'd make a far less cost effective to invest in these electrification and advanced flexible resources. So I think we simply cannot afford a society to have this model where somebody else takes away a half the value. Now customers should be empowered to hire somebody to optimize a system. So it paid me 50 cents a month to optimize your thermostat a lot of people might be very happy with that. But then the customers in charge instead of the aggregator being charged because if there's an aggregator model that prices the customer has no choice. They can only do it through the aggregator. But if you have pricing the customer is empowered to have choice and that's what we need. So ultimately everybody in terms of the aggregators better off with prices and so I think that's absolutely the way to go. So these are the other mission models that people talk about and as I say I think they're all of higher cost and they all perform less well than pricing. And so it's really I don't I don't think there's any contest as to what the right way to go is. So again one mechanism on all scale it turns out today according to the Ethernet standard you can put a price for the value of electricity on that Ethernet cable not according to an application protocol but according to the Ethernet standard itself. I don't think there's any products in the market that use this but it's there on the standard. I know it's there because I put it there and one of my thoughts you know that's how the internet works that's what pricing can do. So people talk about virtual power plants but for the today's rebranding of aggregators but customers are not power plants they have different characteristics and if people think of customers as power plants they are going to think about the problem wrong. We need to have aggregators and virtual power plants in the near term but we need to make sure that pricing can compete on an equal footing so that customers who want to take advantage of flexibility themselves so when devices they buy can do it directly. And so yeah actually I was supposed to read that quite so yeah we need to have a way where these both coexist we need to start offering people good prices as soon as possible so that manufacturers have an incentive can take in these prices directly and everything needs to be globally standard we don't want things to be different in different countries because we're all trying to solve the same problem that's what we learned with communication everybody has the same problem but now there's a communication fair in openavr 3.0 to very simply and easily communicate these kinds of safety rules and it was actually only completed about a month ago I led the process to create it so I think it's the best thing for my friend and critically energy access where I started offering people with almost no money it supports that so there are hundreds of millions of people on the planet today they own one electrical device it's a mobile phone and they charge with usd just like we all do today how can those people who have almost no money have access to this wonderful technology it's because we all use the same technology we need to make sure that whatever technology we use on the standard campus and wherever you live wherever I live that it's can scale up and down to be suitable for everybody on the planet that's the only way to get it into a little future that's it thank you questions answers so it's technically possible to have a system where every customer created this big curve of supply demand and saying depending on the prices I would consume different amounts in the next hour there are no buildings at all on the planet today that I know of or maybe there's a you know it doesn't in research projects would do this how my house wouldn't necessarily do a good job at forecasting its demand and forecasting this flexibility when it doesn't know when I'm going to be turning on the oven to make number of things it's a mind boggling the complex problem that basically nothing does today but if you did it you would probably end up with the same result as if you just use prices the key thing is that with prices you rely on the large value of large numbers of things evening out sort of the randomness of different customer behavior averaging out and that's how the grid works today the reason people want to go to those systems because they think the only way the system can work is if it's but it would be so expensive to do this you would not get to a better result than just using prices do and so it's never going to happen because nobody is in the right line is going to say as a siding week should invest tens of billions of dollars to create this whole new system to get to a result that we can get to far more cheap so in theory you could do it but we never will like I don't I need to buy an airplane ticket this morning I'm going to go to a meeting in Europe which was wonderful news I don't put in a bid for which different flights I would do and then once a week the airlines like do all these things they just post the price and I look at the prices available and I select the way the world works is with these non-determinist systems sure there are these niche cases like wholesale markets where there's determinist assistance but that's not how the world works in generally the people who want those systems believe don't have a non-determinist system and they're just wrong okay but does that get any sense yeah so those nano rate controllers I had within the building each free controller whether it's a micro controller or a demand controller sets its own local price it has to add the algorithm do that and how those work can be a subject of innovation so the algorithms for both the grid controllers and the devices don't have to be standard the communication has to be standard so with HTTP and HTML we have a structure where we have interoperability because things are standard but that doesn't mean that all web pages look the same so with others we can have a system which does both innovation and interoperability so it's the cost a set of price and I have some ideas but I know that we need a whole lot of other smart people maybe some of you in this room to work on that but you learn from what happened yesterday you see about supply patterns or patterns of different prices you look at demand patterns from yesterday look at your battery state of charge use that to how you manage the battery and how you set the price or not two problems but there are two aspects of the same problem but the price and if it seems to be like it's too low and your battery gets depleted too much so we just raise the price so you can always it's like flying an airplane you can always just adjust to keep things balanced yes soon for them to become mandatory I think people always should have the option of a less than the price that they want although I don't think it's up to the utility necessarily to be the one that does that basically if we had dynamic prices today and somebody proposed oh we should have flat prices and that's kind of like saying that you should be able to buy any seat a seat on any flight from here to New York any hour of the day any day of the year for the same amount airlines it was almost for $2,000 you could have whatever thing you want but if you want to save money you would be on a dynamic price so whether it's through a third party that provides insurance insurance always is a premium or by the utility directly sure you can have this flat price but you'll be probably paying a lot more money at the end of the day than a grand dynamic price but as we introduce it initially we want to just make it an attractive offer so that people who want to try it out do they will save money they'll tell their neighbors and their you know friends and family hey by saving money by doing this and I'm also helping the environment and getting more renewables on the grid there's no downside you should do the same thing and then it'll spread and also all of their price response needs to be automated we do not want humans in the loop people you should express preferences but you shouldn't be actually doing it like I can express preferences like flash or not on my phone phone it does all the complex stuff I'm not in the loop setting the focus exposure time because the automation is far better than I could be and I really don't have the interest to do it there's one percent of population that wants to do those things and they should totally have that option but you know your water heater needs to do things while you're asleep people don't want to spend time doing this I don't it has to be automated and it can be like my key control is automated I know is that other comments questions is there anything from online okay so yeah so I've been sort of trying to follow things the trees and there's in a lot of places notion to create flexibility markets that are somehow separate from the wholesale supply markets and also separate from the retail markets which and these proposals like looks particularly like the one in the UK are incredibly complicated lots of different kinds of entities doing lots of different kinds of transactions all because they're refused to charge the right price for retail now if you go to the grocery store you don't next to the tomato saying I'll pay you a dollar not to buy tomatoes this week and there's not somebody asking you are you using this tomato for pasta or for salad they're just tomato right they're just electrons so flexibility is not hard to engage in a retail market if you charge the right price and then of course that people have devices that can do that if you charge the wrong price you will have to go through all kinds of calisthenics to account for the fact that you're charging anyone it'll be a definitely a substandard solution in California our energy commission has some a process called the load management standards to require utilities to offer highly dynamic prices by I think 2027 the public utilities commission also has a process called demand flex to do the same thing which is good because otherwise because it's usually the regulator that does this not the energy commission the policy organization you screw things up for in principle within you know three or four years you should all have the opportunity to pay these prices save money and help the environment that's that's the goal on the plan and then there's a few things in other countries where such prices are being introduced you can do a half hour price in the UK and some other ones but I think we have the most coherent vision and we're getting into the technology details of how more than people are in other places so yeah so the local price is the price on the customer side of the meter so I may have gone through a few of them so one is if there's a difference in the buyer sell price utilities sell the 20 cents buys back a 10 the local price is going to be somewhere between 10 and 20 and it could be in between if you want to take into account carbon emissions that's another one you add the carbon burden to the price to optimize to that if you have a DC power domain or maybe you have an AC power domain that has a DC power domain between it and the grid domain so when the grid goes down you have some reliable DC and some reliable AC each of those domains of power will have a different price because if they're ever going to move power between them there has to be sufficient price differential the microgrid case when the grid goes down there is no retail price to generally get the price locally to do this and there's probably a few more but those are certainly the major ones where the availability is different from the prices to it no no no it's all automatic the vacuum knows that unless it's really expensive vacuuming is you want to do it right now so it's just going to pay for the prices unless it's completely extreme you know it's yeah it's usually when you use the vacuum it's really important yes yeah so if you had a device that was more discretionary let's say a sauna or something where if the price is above certain amounts okay it's just not worth taking a sauna it would look at the price and if the price went above that threshold it would just stop eating and if you had a medical device it would never look at the price and if you had a light as the price does that maybe the light above some high levels starts to dim but the way the pricing operates is not to reduce the service out but most of the way the price operates is to maintain the service but shift the time of the consumption so I have 120 gallon water tank in my house which provides both my space and my water heat it can heat that tank during the during the course of the day soak up the excess PV from my solar panels through the evening coast room and even into the night because I've got a whole lot of hot water there to either use as hot water or as space so I get the same heat I get the same hot water but the time of the consumption of it so shifting load is the vast majority of what occurs there's a there's a company eat blocks from my house that makes an induction stove with a battery inside of it so that it can be 120 volt appliance not a 241 so you don't have to run a new wire to it if you're replacing gas stove and it just charges up the battery of the force of the day so if you have those inexpensive electrons from the PV we have and then when you want to turn the oven on or the burners it draws some power but most of it the energy comes from the battery so you're shifting load you're still getting the same cooking thing so you only shed load you only reduce services on extreme days you know one to three days a year hopefully the other 362 days you're only shifting load and that's a problem because a whole lot of people get introduced to demand response for these emergency conditions so they only think about shedding load but you know van flexibility is 98% about shifting load so it's a it's a misperception of what's really going on I think there was another I'm sorry oh well I mean the utility rate today uses price in wholesale markets and you know PGD has a big battery down in the Monterey area and I assume they charge that battery when the prices are low and they discharge the price when the batteries are high so prices are already used on the grid side because companies that you know have hundreds of millions of dollars in the state they want to use the right mechanism so they use prices right if they didn't use prices the system wouldn't work nearly as well so the price they're already used over there it's using the prices at the meter and using the prices inside the grid is that I'm a building person I'm kind of a person and stuff I don't know how the grid can be operated differently if I thought I did I would probably avoid saying it because I don't have expertise in that area so I don't know how the grid should be work I just know how things should work inside the grid but for Stanford where you do have this you know rich infrastructure of chilled water and hot water presumably you could have a price for the value of the the hot water and the chilled water and you use the same mechanism for for the water energy as you do for the electric energy so again we want is all applications it does generalize in that way it all depends on how the price is so a few years ago the winter time of use price from PG&E which then translated to the CCA but I was do was one cent per kilowatt hour difference between the high price and the low price was like 30 cents was the low price that's really an insanely stupid difference I think now it's at least four cents and the summer was 10 cents and why they're talking about 40 cents so how much you to save depends on how big that dynamic ranges and that varies enormously from place to place so it all depends on the tariffs that are chosen to be charged and we should expect different you know variations in different days on some days when there's less need to shift load there'll be less difference and other there'll be a very large difference when there's more need to shift load so you will engage more flexibility when there's a bigger price and less when there's less because things have different efficiencies you know my water tank's pretty efficient because whatever I lose through the tank heats the house though in the winter it's not really a loss even batteries always there's a round trip efficiency loss in the capital cost so there's you if you have one bigger price difference okay and I'll be hanging around for anyone who wants to keep keep chatting thank you