Skip to main content

The Power System of the Future – Innovations, Trends, and Signposts Text Version

This is the text version of the video "The Power System of the Future: Innovations, Trends, and Signposts," which is a seminar held at NREL about innovations, trends, and signposts along the evolution path of the power system of the future.

Good afternoon, everybody. Thanks for taking your time out of your day, your busy day, to come here and share a little knowledge and insight for the next hour or so on where we at EPRI see the future of the power system evolving to. As Ben mentioned, my name's Bryan Hannegan. So lots of different things to talk about. We're gonna talk about politics. We can do that.

Want to talk about the weather, we can do that. Want to talk about the environment; we can do that as well. But what I'll do in the next 20, 25 minutes or so is walk you quickly through how the viewpoint from the C suite within the utility industry is really changing and changing rapidly. I would say that what I'm gonna portray here is very different than what a lot of the leaders of the power sector were thinking 18 months, 24 months ago. And so you're getting kind of a real-time snapshot of what's inside their heads.

I'm also gonna give you a few things to watch, a few signposts, a few trends to watch because as we move from the power system of today to this power system of the future that I'll describe in a moment, one of the real questions is how fast are we gonna move that ‘cause that matters for investments, that matters for R and D, that matters for business model. So how fast are we gonna move there? Which pieces are gonna move first? What's gonna be driving it?

And how are people gonna continue to make money in this brave new world? Because ultimately, that's what the businesses of the future are gonna have to develop a lot of that. So hopefully by the end of the hour, you'll walk away maybe looking a few things differently, but certainly with a better sense of what the utility folks are thinking these days. First, let me say a bit about EPRI. For those of you who don't know who we are, founded in 1972. We're past our 40th year now. A non-profit research collaborative organization.

Think of EPRI as the pool that all of the R and D funds from the individual utility companies are put into so that instead of everybody paying 100 percent of the cost to do the same project 50 times, they pay 2 percent of the cost to do this large project once. Our annual budget is on the order of about $380 million. A lot of that is in our nuclear area; about a half of that is in our nuclear area. We've got a strong sector in power delivery and utilization, which is where a lot of the smart grid, energy efficiency, electric vehicles, et cetera, is from.

We have a fossil generation division that focuses on coal and gas, but also includes carbon capture and storage. And then we've got my area, which includes basically everything else that doesn't fit into one of those three buckets. We have over 450 participants. These could be power companies. These could be universities. They could be NGOs.

They are located in over 30 different countries on nearly every continent. I don't think we've gone to Antarctica yet, but we're working on it. And here in the United States, our members represent about 90 percent of the electrodes flowing on the system. And as you see on the last one, $1.00 in every $4.00 is actually coming in from outside the U.S. So as you think of working perhaps with EPRI in the future, we're a great partner to help you gain international perspective that you may not get directly sitting here at a DOE national lab. So I mentioned the three key aspects of EPRI, what makes us different from a contractor or a consultancy.

One is that we're independent. I like to joke with our advisors that you get what you get and you don't get upset. For those of you with little kids, you may recognize where that – we can talk later. I've got a seven-year-old girl and a four-year-old girl and they're always throwing a fit about something. We're independent, objective, scientifically-based research. I think we enjoy a really unique position in that we can provide a lot of unbiased and actual truth to power. When it comes to environmental data that goes into the EPA rulemaking docket, when it's testimony in front of Congress, whether it's interacting with the headquarter staff of DOE or the folks in the various other agencies or even the White House itself, you know that when you're interacting with EPRI, you're getting solid, clear, transparent work.

It's also non-profit, so everything that we do has to have a public focus to it and we try to put as much of our research into the public domain as possible. And then we really focus on collaboration. We don't do one-offs. Everything that we do has got a group of individuals from across the industry, various and different types of individuals, and if we don't have the best available knowledge in-house, we're more than willing to go out and find it. In some cases, we're coming here, particularly on renewable energy, and hopefully soon on energy integration as well.

I mentioned our four areas of research programs: the power delivery and utilization, and you can see some of the topics that we focus. Of interest to ESIF is grid operations and planning, how are we gonna operate this power system in the future with the reliability that our customers have come to expect. We've got a big area of work in nuclear. Nearly every sort of civilized world nuclear operator is in this program and it sets a lot of the R and D basis for the NRC, and by extension, all throughout the world, so it's a great collaborative there. I mentioned the fossil generation group. Increasing work on combustion turbines and materials and chemistry as we're looking at moving fossil units from base load operation into more recycling or load-following mode.

So that puts stresses and strains on the materials. And a big focus there also on water, which I know is a critical issue in the minds of those of you here at NREL. And then my area, environment and renewables. I'm gonna focus today on the renewable energy piece. But I'm certainly happy to answer questions about any of the other stuff. So all of this work arises and sort of bubbles up to the three parts that challenge us that going all the way back to the early stirrings under Edison. This was the goal of the electricity energy was to provide reliable service at an affordable price, and increasingly now as we go forward in time, it's gotta be environmentally-sustainable or environmentally-responsible.

And keeping in mind that we're doing it not just for ourselves, but for a wide variety of stakeholders throughout the world with that. And we have to take this challenge on, preserve that reliability, preserve that affordability, extend the sustainability of our work, even as we're moving a power system that looked somewhat like this, a sort of a hub-and-spoke model where you had large generations, a one-way transmission and distribution system, and relatively passive customers in the way that you had a very simple equation for keeping the lights on. You had base load generation. You had generation that was dedicated and load following and intermediate or peaking.

You had some bulk energy storage, bulk hydro, you had compressed air in some places, and all of that was against a pretty predictable customer demand. And largely at the commercial and industrial level, some interruptible that you could use for peak shaving and you can sort of prearrange the deal with your large consumers in the area. Very simple, very straight-forward. Not too hard to operate. We kind of new what was coming. And so it was a pretty linear system from an engineering standpoint.

Well, what's happening? We're now seeing the onset of centralized wind and solar, largely on the great technology work that's been done here in your partners around the world that have been driving down the costs of wind and solar. We've got favorable policies now that are helping deploy those. So we're bringing on those resources and the variability that they come with. We're also seeing a lot of interest in distributed generation. I'll come back to that in a moment.

Not just solar PV. I live in Northern California. My average rate is $0.18 per kilowatt hour. Most everybody I know is seeing solar coming in well below that unsubsidized. And then you add in all the state and federal supports and it's a great deal. I don't know why anybody wouldn't do it. We've got lots of distributed generation coming in, gas micro turbines at $3.00 or $4.00 per MMBtu, starting to look like a great deal, particularly if you're concerned around your own energy security. Let's say you've got a high technology fab line and you want seven-nines of reliability, this is a pretty easy way to guarantee that and a lot of innovation in these spaces.

So you've got variability on the upstream side, on the generation side. You've got variability coming on now on the downstream end from the customer side, much of which the utility is blind to, so you've got a whole different equation for your generation. And then you add all the changes that we're seeing in the customer base, whether it's new pricing systems that allow for consumers to respond to time-of-use pricing or the tiered pricing structures peak, super peak, and off-peak demand response. You've got electric vehicles starting to proliferate, 100,000 vehicles so far, on its way to 1 million and more in the next few years at current gasoline prices.

And then new customer-side demands, folks that have been using coal or natural gas facing maybe emissions limits in the Regin region or in California starting to think about how do I source my energy from cleaner sources of electricity and can I start to use my industrial processes on electricity bringing spiky megawatt type loads onto the grid where they weren't there before. So there's a whole set of variabilities coming in on the customer side as well, and that's leading to a much more variable system and one that frankly looks more like an optimized delivery network rather than a simple linear pipe. In fact, when you look at what the power system of the future graphically starts to behave and appear on paper, you've now got your large wind and solar coming in on the generation side.

You've got all sorts of data and analytics to come along your various generation controls. You've got your coal and perhaps even your nuclear assets starting to run to follow the load, as opposed to being base-loaded with everything on top of it. On the customer side, you now see two-way power flow, both from the point of creation back up the grid, as well as down the grid in the traditional sense. You've got consumers with PV on their rooftops. You've got community and residential-scale power generation, energy storage, both in the bulk and in the distributed lens.

You've got new uses from electric vehicles. And sort of in the middle of this, you've got a grid operator that's now dealing with trillions of data points of variability on seconds, minutes, hours of time scale because solar, as you know, when the clouds move over, it's a very different time mode of variability than a cold front moving through and shutting down some of the wind. So you've got a much more dynamic distributed and decentralized power system, and a lot of these technologies by themselves are here today. You hear about the Nest Thermostat.

You hear about solar city and the success that they're having coming in behind the meter. And you hear about the utility smart meter installations and all of the building energy management analytics that folks like SAP and Schneider Electric are starting to put forth. The bits and pieces are coming together. But how fast are we going to be moving towards this power system of the future, this optimized delivery network, this highly-variable generation mix, this highly-variable customer demand, how soon as we gonna move there? How fast?

Could it be that we get there first on the basis of customers and the generation is slow to respond? Or is it the other way around? And are both of them gonna get there before our grid capability is ready to absorb? And what does that mean for the technology and the rates of penetration of that technology? And then at the end of the day, people are gonna be in this business to make money.

Let's be honest about that. So what are the business models that are gonna accompany this transformation, the power system of the future? The operation of the grid system may continue on the old monopoly franchise model because nobody's gonna go right out and build another incremental mile of transmission to serve given the costs of the permitting itself. They're gonna want a predictable rate of return. So maybe the traditional utility compact à la Con Edison in New York is we become a wires company.

We've got to invest all of our generation and our customer assets. We become a wires company. We're highly regulated. We get a nice steady rate of return. The banks love us. That's our niche of the business. I can also imagine a future cleantility, to coin a phrase, that specializes in helping the customer bring all this stuff into their homes and into their businesses without the customer having to go to Home Depot, install it themselves, put widget A with widget B, and actually create a value proposition where not unlike what we do with our cell phones today, we're buying energy not as a commodity, but we're buying it as a service.

We're asking who can provide me with more comfort at a lower cost or more mobility at a lower cost. Already we see folks like NRG packaging their electric vehicle charging infrastructure with the sales of the electricity, and it's not long before they'll probably do the car as well. Why not do it on the basis of total cost of ownership when in fact some of our own research shows that for an EV, the total cost of ownership is about 25 percent less than a convention gasoline vehicle. Why not realize that value proposition and make some of that profit?

So there's a whole lot of business models that are out there. Someone's gonna specialize in microgrids, going from a large commercial and industrials and helping them do things in a turnkey. Someone's gonna reach out to Marin County and say, "Fine, if you guys don't want to be with PG&E anymore, here's how we can do it." And perhaps that somebody might well be PG&E. There might be cobranded products. There might be new divisions set up.

Already people are starting to think, "Gosh, if we see our customers leaving because of aggregation and disintermediation, which is a word that I despise. But if I see these new business models coming along, maybe I should work with my regulators to get a piece of that market so I don't lose my customer. And the regulators are sitting over here, as they were at our meetings earlier this week, and thinking, "Wow, to get all that technology innovation, we're gonna need some regulatory innovation as well. We've got to look at our pricing structures."

"We've got to look at who can be in our markets and whether competition helps or hinders." So there's a lot of moving pieces going on, and it's not straight-forward how soon, how fast, and what business models will arise. What I want to do is give you a few things to watch to give you some guidance into how you might assess each of these questions. And I want to start with the customer side. These are, in my view, five things to keep an eye on.

I'm not gonna cover 'em all. But I'm gonna give you some examples of signposts that we're watching at EPRI to see how fast the markets and the technologies are moving and to see where the emerging research needs are, both from a technology, as well as a policy and a regulatory sense. There is a variety of smart appliances, hyper-efficient appliances. We have tested quite a few of them here. Some of them have deployed in Japan and in Europe and elsewhere.

You bring these into the United States, you start to get these hyper-efficient devices to communicate with one another, and you will begin to realize a variety of efficiency gains and also the ability to optimize. The growth in demand for datacenters, for example, is a huge load on the system. But we also are seeing tremendous improvements in the efficiency of those datacenters and the fact that we're now starting to see their demand flattening out, even as the total demand for service increases, and that's kind of a theme. You'll hear me come back to it.

We've got heat pump water heaters, variable refrigerants, low air conditioning, LED street lighting. One of the nice things about LED street lighting is because of the reduced load that the LED provides, you can actually imagine a traffic signal or a streetlight that has a solar panel over the top of it, has enough battery storage to keep going during a blackout. How many times have you been in a neighborhood where the power's been out and the worst problem was not necessarily that you couldn't get phone service, but because nobody knew how to drive? And so if we could take some of the public safety issues out of the way through modern technology, keep the lights on for that very important purpose.

So watch these items as they evolve and the rates of deployment as one indicator of how far we're moving. Electric vehicles; today, we're looking at silicone anode technologies that get us these advanced lithium ion batteries in the 150 to 200 watt hour per kilogram range. With advanced technologies, we're looking at upwards of maybe 400 watt hours per kilogram, twice the effective range for the same size vehicle. So we would take the range, the all-electric range of the Leaf from 50 to 100, and if we take that to 300, our studies of the consumers in the auto market suggest that would be the point at which range anxiety would all but disappear and you can start to see consumers setting that aside as a purchasing consideration for tomorrow's vehicle purchase.

And that then would lead to proliferation of these batteries on wheels, which have a tremendous impact on what the distribution architecture's gonna look like. So we're working on lithium sulfur batteries, lithium air chemistries, a variety of things through our energy storage program, intending to kind of keep this curve going up and hopefully move this number somewhere this side of 2030. So keep an eye on battery range as these new vehicles come out; what's the range equivalent? Once we get close to gasoline, that's a good sign maybe we're gonna see more and more of these in the marketplace.

I mentioned distributed energy resources. Folks, this is the levelized cost of electricity for distributed generation of various types in commercial applications today. Here we are at a $4.00 natural gas price. This is sort of the current market spot, if you will, for where we think natural gas prices will go. In places like Northern California where my marginal rate is actually up here, my average rate is down here, something like a solid oxide fuel cell or I maybe the industrial turbine or a micro turbine start to make a lot of sense, particularly if I'm a big box manufacturer or I'm a market chain like Whole Foods, this is one of the Connecticut fuel cell energy fuel cells that's been put in there.

We're starting to see more and more of these installed, and the key is even without combining heat and power, we're seeing good economics. Now, also use that for space heating, for waste heat process, and you've got a better value proposition. So why do utility folks live in fear of never building a large capital asset ever again? Because between this and distributed solar, it's not unreasonable to think that in some parts of the country, you might be able to cover all your incremental growth behind the substation, which is a very different business model than what we usually come to expect.

Let me talk a little bit about the networks that are gonna go along. So we're moving from the upstream, sorry, I should say the midstream. Smart energy systems go beyond the smart grid; go beyond the smart meter first; go to the smarter grid and then go through the rest of the system. Will there come a day where we're thinking more about systems rather than individual components? That's I think a big hurdle that we have to challenge. But I think here at ESIF, you guys are in a great spot to accelerate that by in effect allowing for plug-and-play of different components in a system that you can change on the fly and test and experiment.

I think you'd see a really massive improvement in our society's capability to do this. It's not just about the technology, but it's about the data. You build supercomputing for a reason. Massive dataflow. Gotta sift through all of that to find intelligence. It's not enough to have the data and the information. We've got to figure out what to do with it so that it can help us optimize how we run this network and how we deliver those services. And that data has got to be secure from cyber and physical attack because if I'm a consumer and I'm working with you as a third party and you're saying, "Well, we're gonna collect all this data and we're gonna optimize your building structure and you don't have to worry about a thing."

Say, "Well, wait a second. You're gonna know when I'm home. You're gonna know when my lights are on or when my car's out of the garage or when my lights have gone out for the night and I'm going to bed. You're gonna know a lot of things about me as an individual." You may already know that on Facebook. In fact, if you've looked me up, you've learned more than you ever think you'd want to know. And that may be okay for some parts of society.

But I can imagine other parts that are saying, "Hey, if it comes with infringement on my privacy, then I don't want to have anything to do with it, even if I'm the greenest person on the planet." So smart energy systems, you hear this phrase, the virtual power plant, CPS Energy in San Antonio likes to talk about how they're able to peak shave upwards of 10 megawatts on their super peak days in Texas where the demand is hitting the all-time high and they talk about going further and doing 20 and 100 megawatts over time; what you want to see is groups and blocks of different loads sort of acting as one and actually serving as a biddable product so that the California ISO can call up the University of California Davis and say, "Gang, we need about five megawatts of demand response and we need it two hours from now. Can you guys do that?"

And the campus building energy manager says, "Absolutely." Hits a button on the campus manager system and the lights are dimmed, the AC starts to cycle, and a few other processes go offline and go to variance. And to the grid, it's a single prompt. It doesn't matter whether it came from A or B or C, it's just five megawatts of DR. It's a virtual power plant. It's as though you had a five megawatt turbine kick in. That's what we're looking for and we haven't yet seen that in the real world. But we'll keep an eye out on whether this concept of the virtual power plant can deliver, basically aggregating all of these things together and not having to manage them individually, but instead managing them as a block.

I mentioned data. Each one of these 1.0s on the axis here is 250 million DVDs worth of data and you can see just over the last 10 years, we're starting to really peak the curve and move up in our big data challenge when you bring in all of these sensors and new meters and building energy management systems and even things on the TMD system itself to help boost reliability. One of the challenges with feeding power from a customer back up through the distribution system is that in many cases, our distribution feeders were never built for that, number one. And number two, we have no understanding of what that does to the feeder itself from a physical standpoint because we don't have any visibility into its performance.

We've gone out at EPRI and done some studies where we've looked at high penetration of PV on the rooftops and we've gone out and manually measured the voltage of a function of a 24-hour day and the folks in the distribution company were absolutely shocked. He said, "There's no way that distribution feeder should've continued to operate because you were three times out of speck with what we expected it to be because of all of the voltage fluctuations coming on and back and forth from the PV." Now, we've since followed behind that and said, "What if we paired the PV with a smart inverter that can provide voltage support when you need it and power when you don't and sort of moderate things?"

And surprisingly, we're now seeing better voltage profiles. By better, I mean more stable over the 24-hour day, better with 20 percent PV in the smart inverters that we had, even without the PV to begin. So in effect, by putting PV on the distribution feeder, we made it more reliable. Think about that for a moment ‘cause all you hear about PV is, well, it's gonna mess things up. We've actually gone the other way and showed that that's actually the converse, you actually do improve things. But you could only realize that if you know what to do with the data that you're collecting.

So the big data challenge is a huge one. Watch for the Oracles, the SAPs, the IBMs to get into this business more than they already are and once you see them providing turnkey solutions to utilities, then you know we've crafted that. I mentioned cyber and physical attack. Just a few examples recently, things like sleeper agents and impacting remote operations, mass disconnects. You know, somebody gets on the computer, a 17-year-old teenager gets on the computer, probably my daughter in 10 years time, gets on the computer and decides to shut off the lights at her friend's house just for fun and in so doing manages to take out the entire subdivision.

It's possible. Might not be malicious, but it could happen if we don't have our cyber defenses up and in place and continuously improving as well. Even things like disabling the billing system. You want to hurt a utility? Hit ‘em in the wallet. Take out the billing system. Doesn't affect the consumer; sure creates havoc when it comes time to figuring out how to pay the bills. So lots of opportunities I think for improvements in the cyber security area. Lots of folks having clearances much higher than mine working on this, and I think this is an issue that as we make the transition, there may be some hiccups, but I think over time, we've been able to make the Internet secure, even as each of us now have 20 or 30 devices in our house connecting to it.

And I think we can do the same with the electric power system. So keep an eye on the signposts. If you start to see interruptions from whatever cause, and I would say physical attack; it might not be human-induced. It might also be Mother Nature-induced. So if you start to see impacts to a more smarter energy system that has more points of entry, if it turns out that those points of entry actually make it more fragile rather than less, then you're gonna see a slowing down perhaps in the move towards this power system of the future. On the other hand, if we're able to show that a more distributed energy system would've done better during Hurricane Sandy, or should I say Superstorm Sandy.

As a meteorologist, it kind of hurts me because that's not really even a defined term. But we'll talk about that later. If you were able to demonstrate that a different kind of power system would've had greater resilience, then you might see more power companies and more regulators and more government officials moving towards that future. Finally, let me wind up with the story of the upstream side and the power generation system of the future. We tend to overuse this phrase, "We've got to fix the airplane while we're flying it." But in every respect, that's what we need to do in order to keep the lights on.

We've got to change the generation fleet even as we continue to operate that generation fleet going forward. And so right at the top of the list is the long-term operations of the existing fleet. And one of the things that it's a real concern; it's probably overblown by half and under-blown by half; but when you start to run something that was engineered to run at a consistent temperature and pressure, when you start to move that up and down on a regular basis or even on an irregular basis, you start to stress the metal. You start to accelerate the corrosion. You start to have more fatigue.

You start to bring your major components to the end of life a lot quicker. This is some thermal-induced cracking in some piping. We've got it under a microscope here, a remote sensor that allows us to go in and do some non-destructive evaluation of different components. You can see the printout here on the computer screen that tells you sort of what the depth of metal that's available to it so you could do predictive maintenance. You can go into an outage. You can schedule it during a regular outage.

All of the things that you would have to do to develop a more flexibly-operating fossil and nuclear fleet. And so keep an eye on nuclear unit cycling; yes, nuclear. They do it in France a little bit. They're thinking about doing it a lot more here as those units come to the end of their current commissioning cycle and then look at the outage rates because we've seen instances where not having that – enough capacity is just as damaging as having too much demand when it comes to the reliability. So lots of interest in long-term operations.

We're also gonna have to continue to operate in particular the coal fleet, but to the lesser extent, all these natural gas assets that are coming on in a world of increasingly tighter environmental statements. Mercury, air toxics, conventional air pollutants, sulfur and nitrogen, particulate matter, these are a variety of technologies that we have under tests in one form or another either at EPRI or one of our facilities. This one in particular, this carbon activation process is kind of interesting. We're actually generating the activated carbon that we use for mercury control out of the coal itself, whereas instead, we have to go and buy it, ship it in, and there's logistics and even though UPS says they know logistics, power plants, they've got their own troubles with that, why not generate it onsite and in fact generate it onsite at a lower overall cost to meet the new mercury controls that are out there?

The difference in these environmental control challenges is that most of these technologies do extremely well within kind of a narrow window, a narrow parameter space, a certain flow rate, a certain temperature. But again, if they're load-following and not base load, you're doing this. You are all over the map. And if EPA says, "Well, you've got to meet your emission standards on a continuous basis" now we've got a technology challenge if that continuous basis is moving all over the place. And sometimes we're in the sweet spot and sometimes we're not.

So there's a lot of innovation there as well. And I mentioned natural gas deliberately because our friends in the natural gas community, they might think, "Well, you know, we can meet the standards now." Well, what about five years from now when they come up again in the context of the Clean Air Act? What about all the new science that we're looking at at EPRI that says it's actually the ultra-fine particulate matter that goes way into the respiratory system that causes the most cardiovascular damage and the pulmonary diseases that EPA is trying to minimize in the first place under the Clean Air Act?

So looking at particulate matter, control technologies, understanding what we're gonna see in a more natural gas-driven world, that's one of those signposts to watch, the timing and the magnitude of future regulations. On natural gas, I have never been either a bull or a bear on natural gas. I've had a gut feeling for the last five years that we're gonna be in a $4.00 range for the near future going up to $6.00 in 2015, 2020. Don't tell me why. Don't ask me why. But it's sort of an accumulated reading sense, looking at the marketplace, reading the 10-Ks of the firms that are out there producing in the shales right now and seeing the difficult economics that they're facing here in $3.00 to $3.00 range, thinking about as you move up this supply curve and you employ some of the best practices for handling things like produced water, induced seismicity and so on, as you start to move into more and more of these new finds, you see the gas prices starting to trickle up.

And also thinking about in an energy system when I have a low commodity and it is the one that is the most favored, I'm gonna have people rush to that. Someone once said to me that made me laugh, "The surest way to get $8.00 gas is to act like it's gonna be $4.00 forever." You get all this demand, the onshoring, the Dow Chemicals, the DuPonts, they're all bringing their operations back, the ones that were leaving when I was in the Senate. They're not coming back because natural gas costs are affordable here.

What does that do? Increases the demand, puts pressure on these resources, moves the market up. At some point, there's a correcting factor; it's no longer cheaper for a utility to run their natural gas units over their coal, so now the dispatch goes the other way. Demand comes back down, everything settles in a nice equilibrium. $4.00, my guess in the long-term within the next 5 to 10 years, we'll see that come up $5.00 to $6.00, and that's gonna bring on a lot of technology. It's gonna bring on wind and the best resource areas without subsidy.

It's definitely gonna bring on solar in the Southwest and other areas. So you're seeing a world where yes, we're in kind of a period now for natural gas that it seems to be that's the most favored and it's the easiest thing to do from a utility standpoint. Well, watch the geology. Do we get enough recovery? Do we keep making the new finds? Watch the oil prices.

Why? Because all of these red bars are selling their natural gas liquids at the oil price and not at the gas price. So they're making most of their money on anything but natural gas. This is just a byproduct. And they're flooding the market with it at a very low price. You want to watch these guys, the Pinedales, the Jonahs, the what's so-called dry plays without natural gas associated with them. When you see them producing, then you know the price supports are gonna be there, and that's something that utility planners can bank on. And then I mentioned regulation and demand as well.

Keep an eye on people and their new uses for gas. Do we have more LNG? Do we start to export it? And if so, does that mean we start to link with the global gas price rather than our own here? And keep in mind that in Europe, they're paying 13. In Japan, they're paying 18. So that brings price pressure up. I live in California near some of the most fertile farmland in the country, if not the world, yet it costs me more to go two miles to my grocery store to get that strawberry that was produced five miles down the road than if I went to Washington, D.C. and bought it off the shelf at the Whole Foods.

We've created these global networks that don't necessarily keep oversupply in one place. And if we go to a global market with natural gas, the same is likely to be true for us. And that matters when it comes to where the electricity comes from. Finally, renewable energy, you guys know this as well or better than I do. We've got most PV costs situations now beginning to cross over with the range of the utility electric retail price; retail price, so that as I mentioned, the solar cities of the world are starting to see a viable business opportunity.

And that's likely only going to continue over time when you think about the utility having operations and maintenance expenses that go up, having infrastructure to support and to repair, having labor costs that are increasing over time, and if you look at the learning rates, let's say that we go back to the standard learning rate for PV and the balance of system, that'll still continue to go down. So you've got more and more of an intersection over time. And as we see those perhaps accelerate, if SunShot is successful bringing down the balance of system costs to the level that we've set as a goal, this becomes a huge crossover, again, to the point where a utility might have to honestly start thinking about how do I engage the customer and become a retailer, like in Australia where they're already focused on this as a result of their carbon cries, the renewable energy standard, and the structure of their market.

Last thing, and I think this is actually the sleeper issue of ‘em all, we spend a lot of time in our industry thinking about carbon and that's well and good. Carbon, climate change, huge concern. We have to get moving on that. But in some respects, we set the carbon cap. We make the decision on how much as a society we want to emit in the environment, what constraints we want to put on the system. When it comes to water, the constraints set for us, there's only so much water in a watershed and it's gotta be parceled out between people, food, ecosystem function, and then everybody else.

And more people using more water for municipal and industrial needs, eating more food that's gotta be irrigated with agriculture, ecosystems probably using more and more water in a world where it's hotter or humid; you've got some impacts of climate variability and change coming. Who's at the end of the line? Users like the power plants. And so we've been actively encouraging our utility members who think about a world in which there is reduced water availability to the energy sector to the point where we're looking at perhaps zero liquid discharge requirements, where we're looking at desalination or wastewater treatment and reuse.

We've just opened up a multi-million dollar water research center down in the Southeast in Georgia where we're working with the Department of Energy to really explore this energy and water access because when you think about all of the advantages that our energy system has, water efficiency isn't one of them. And oh, by the way, let's not rule the possibility that a water utility might start to recover the embedded energy in their waste flows. They might recover the bio solids that are coming in the wastewater streams, use that to drive the energy of their pumps, the clarifiers, their other treatment requirements and actually start to wean themselves off the electric utility system as a customer.

And perhaps as in the case of the East Bay Oakland Municipal Utility District, sell the excess power back into the grid, as they're doing from their seven megawatts of anaerobic digesters where their load is only two megawatts. That other five goes back on the grid and PG&E has to actually pay them for it. So there's a crossover there between energy, waste, water, and if you broaden your concept of the smart energy system to smart energy water and waste system, then the ability that you have to optimize becomes a manifold grid.

So I'll wind up by saying I hope that at least I've convinced you that there's a lot of change out there on the horizon and the power system as we know it is likely to see much more change in the next decade than we've seen in its entire 100-year plus history today. The real question that I find myself asking in boardrooms, in public meetings, with our advisors on research projects, is how are we going to respond. How are we going to respond as consumers? How are we going to respond as technologists? How are we gonna respond as those who talk with policymakers?

And how are we gonna respond as leaders? I was an early champion from my EPRI perch of the investment that you've all made in ESIP here because I think that is one of the most important things we're gonna have to do is start to actually do these smart energy systems, put them in practice, monitor the heck out of them, learn from those experiments, and then get better and better over time so that when we make this move to a power system of the future like the one that I've described, we do it in a way which maintains the reliability, it maintains the affordability, and it gives us the environmental virtues that we're after.

I thank you for your 45 minutes or so and hope that I've given you a little something to think about as you head home today. Thank you. [Audience claps]