Clark Gellings: The Future of the Power Grid

Electricity expert Clark Gellings envisions an integrated, agile network to tie together millions of electricity producers and consumers

Loading the podcast player...

Jean Kumagai: Hi, I’m Jean Kumagai, and welcome to IEEE Spectrum’s “Techwise Conversations.” What will the power grid look like 50 years from now? More importantly, what do we want it to look like, and how will we supply reliable, affordable, and sustainable electricity to a global population that may reach 10 billion by midcentury? IEEE Spectrum considered those important questions as part of its recent special report “The Future We Deserve.”

Clark Gellings is one of the world’s leading experts on the electricity system. He’s a Fellow of the Electric Power Research Institute in Palo Alto, California, and also a Life Fellow of the IEEE. During the course of his 46-year career, his ideas, his writing, and his testimony have really helped propel the electricity industry toward greater energy efficiency, more widespread adoption of the smart grid, and more integration of renewable energy and other clean technologies. Clark, welcome to the podcast.

Clark Gellings: Thank you.

Jean Kumagai: How should we be thinking about the electricity grid right now, and why is it important to also be thinking about the grid in 50 years’ time?

Clark Gellings: We designed the grid around the need to provide population centers and industry with power. As the years went on, we built power plants near those population centers and near those industrial facilities, and ultimately we got smart and thought, Well, gee, what if we connect all of this together with wires?

Well, things have changed. First, the nature of the energy sources that you use to generate electricity have changed. We’ve got some new ones that are very promising, like wind-power generation and solar-power generation. We’ve got perhaps just the dawn of an era of an understanding of how to store electricity, much like we store electricity in the batteries in our cars.

But looking down the road, we’re going to see an awful lot more different resources that generate power, and we’re going to have to understand how to make all of these, what will be millions of points of generation, and millions of points of storage and use—how are we going to connect them all together in a very efficient way?

Jean Kumagai: Electricity rates in some parts of the world, like Hawaii, are already so high that it now makes economic sense for people to install solar panels on their roof, hook those panels up to a battery, and basically unplug from the grid. So what’s wrong with this picture?

Clark Gellings: Yes, there are a percentage of people, residential consumers, who have disconnected from the grid—divorced themselves, if you will, from the grid—but actually that percentage isn’t that great. In fact, the grid, even with the consumer installing his own photovoltaic generation, provides some resources that you can’t otherwise get. You can’t start a large central air conditioner without the grid behind you. You can’t have the kind of day and night power reliably supplied to your home or office or industrial facility without the grid playing a role.

So what you’re finding in places like Hawaii and California, Arizona, and so on, where people are installing solar systems, they’re still connected to the grid. And that’s fine, except there’s a question of how much they pay. And if they don’t pay their fair share, then others will be subsidizing them.

Jean Kumagai: When you say “paying their fair share,” what exactly do you mean?

Clark Gellings: Well, let me use some U.S. average data to give you an example. On average, the U.S. residential customer consumes about 1000 kilowatt-hours per month, and on average he pays about $110 for those kilowatt-hours. And if you start digging into that, and you say, Well, how much of that is paying for the power system itself? That’s the grid and the fact that there’s generators tied to it to offer additional strength to the grid. That’s about $51 of that $110.

Now, most of the consumers who are separating themselves from the grid are trying hard to substitute for as many of those 1000 kilowatt-hours as they can. And as they do, they pay less and less to the power company. In fact, depending on the metering scheme, they may even get some kind of rebate, if you will, for the energy sold to the power company. Meanwhile, the utility still has to bear that $51 a month.

We need the grid. We absolutely, all of us, need the grid. But we don’t necessarily need it in the way we’ve needed it in the past. So we’ve got a bit of a problem to sort out here, and it’s not just a technical problem.

Jean Kumagai: Well, you’ve actually come up with a very intriguing answer, this concept of an Internet-like equivalent for the power grid that you’ve called the ElectriNet. What is that exactly, and how would it be better than what we have now?

Clark Gellings: So let me just go back a little bit in history, because all of this kind of started in the late 1970s, and it led to our asking the question, Why do we have to always predict electricity as being a commodity that’s following some kind of a growth curve? Why can’t we think about how we use it differently? So that’s how we began to introduce the idea of actively becoming much more efficient and effective in the way we use electricity.

Well, the next phase of that was this concept that I call the ElectriNet, published first a number of years ago, where I basically suggested that we are going to have the opportunity to generate electricity at our homes and businesses. And why can’t we optimize all of these pieces so that we operate our own building efficiently, we operate the appliances in it efficiently. And then we allow our on-site generation system to optimize itself against all the other generation systems that are connected to that grid. And if we organize this all very effectively, we could end up with this equivalent, if you will, of a fully integrated power system. I sometimes call it the “integrated grid.”

Jean Kumagai: Okay, and that, I guess, would also make it easier for grid operators to have a sense of what’s going on on the grid at any give time, correct?

Clark Gellings: If we think about it for a moment, we’re talking about the electronic age, when all of us, all around us, have all of these devices that are based on the digital revolution, if you will, but yet we’re still controlling the grid mechanically. We’re still relying on pretty slow communications to monitor its condition. We essentially don’t know the condition of the grid at any point in time. So when something goes wrong, it can go terribly wrong very quickly.

If you adopt this idea of the ElectriNet, where essentially you’re overlaying a communication system on top of the power system, you will be installing sensors, communications, and computational ability that would then allow you to make full use of the grid, at the same time balancing it in the way I discussed a few minutes ago, and have much better visibility of the condition of the grid. So it’s to mitigate and prevent outages, and repair or restore from them very quickly.

Jean Kumagai: So what are a few of the things that will have to happen between now and 50 years from now to make your vision of the grid a reality?

Clark Gellings: Well, first, we’re going to need communications standards that allow devices to talk to one another, so that we don’t have the problem we have now. For example, in buildings, the electronics that are being used have as many as 28 different communications architectures. And so one building technology that might control some new thermal storage unit you have may not be able to talk to another device in that building.

Number two, the computer system that would control these millions of nodes in any given region of the United States, they don’t exist. I mean, we can control tens of thousands of nodes, and we do now, but we’re going to need to control millions of nodes. So that’s another area of development.

And thirdly, technology. For example, power electronics to fully be able to control, in a very fluid way, the power systems, even to the point of doing things like having the system self-heal, or taking action so as to mitigate from an outage that it sees, even before necessarily the outage has occurred.

Jean Kumagai: You know, this is also sounding very expensive, so I’m guessing that part of this will also involve policymakers and grid operators and so on being willing to spend the money to make these important investments.

Clark Gellings: You know, it’s not going to come cheap. And I’ve done cost estimates of that, and roughly speaking you’re talking about something on the order of $12 per month for a residential customer, is what it would cost to do the kind of modernization that I’m referring to. Or, as I’ve glibly said, the cost of a pizza pie a month.

And I don’t see a way around it. Without it, you’re going to suffer from decreased reliability, you’re going to suffer from poor power quality, you’re going to suffer from limited resiliency in terms of recovering from major storms and outages, of which we have a lot in the United States. And those costs, although they’re not monetized in power-system terms, they add to the cost of all goods and services in our society. So I think that putting it in the residential equivalent, that $12 a month is well worth it. In fact, overall I believe it has about a 4:1 benefit-to-cost ratio.

Jean Kumagai: Thank you very much for speaking with us, Clark.

Clark Gellings: It’s really been a pleasure. Thank you.

Jean Kumagai: We’ve been speaking with electricity expert Clark Gellings on the future of the power grid. For IEEE Spectrum’s “Techwise Conversations,” I’m Jean Kumagai.

This interview was recorded Tuesday, 6 May 2014.

Audio engineer: Francesco Ferorelli

Images, Left: iStockphoto; Right: EPRI

Read more “Techwise Conversations” or find us on iTunes.

NOTE: Transcripts are created for the convenience of our readers and listeners and may not perfectly match their associated interviews and narratives. The authoritative record of IEEE Spectrum’s audio programming is the audio version.