The Smart Grid - Transcript

On April 10, 2008, Popular Mechanics and the National Science Foundation presented "Bridges to the Future: A vision for infrastructure in the 21st century," a webcast discussion exploring the best ideas for improving American infrastructure and building a better, safer future. This is a transcript of one of the panels. For links to the other panels' transcripts or for more information about the webcast, see the Bridges Webcast Conference page.

>> Jerry Beilinson: Hi. This is Jerry Beilinson for Popular Mechanics magazine. We're here in Arlington, Virginia, with the National Science Foundation, our partner in today's conference, which we're calling "Bridges to the Future." What we want to talk about today are challenges facing infrastructure in the United States in the 21st century and just as importantly, of course, we want to talk about solutions and smart technologies to help meet those challenges. We'll be having three sessions today. We're starting with the conversation on the electrical grid, and specifically on Smart Grid concepts. And we'll be starting here in a couple of minutes, and we will be having questions in a few minutes, both from our live audience here in the studio. We'll also have questions from those of you joining us online. There's a phone number that you can call and we'll be taking questions by email as well. First, I'd like to introduce our panelists. Joining me in the studio is Dr. James Momoh, Director of the Center for Energy Systems and Control at Howard University in Washington, D.C. Also here in the studio is Roger Anderson, Director of the Edison program at the Center for Computational Learning Systems at the Columbia University in New York City. Joining us remotely is Dr. Anjan Bose, distinguished Professor of Electrical Power Engineering, Washington State University in Pullman, Washington. Thank you all for being here. Dr. Momoh, the United States suffered in 2003 a significant blackout on the East Coast that shut down a big portion of the country. There was another blackout in 2006 in California that was blamed for 141 deaths. And there have been smaller power outages throughout the country. So I guess the first question is are we in trouble in the United States with our electrical system, and are we in fact facing a crisis at some time in the future?

>> Dr. Momoh: Thank you, Jerry. It is my pleasure to be a part of this panel. Indeed, the electrical power grid as we know it today is the hub of our economy. It generates all the resources for all the stakeholders, whether it be manufacturing, whether it be industrial, whether it be commercial. It's a fabulous challenge and it continues to be. First of all, let's look at the grid coming from the generation side of the business all the way from end to end, the transmission system to the distribution networks. This networks of networks as we know them is aged, has not been maintained for a long time, and in fact the investment continues to dwindle. So, in other words, when we look at the demand that is placed on the grid as we know it today, it's so much that we are not sure we'll be able to sustain the capability for this network to handle the demands at all levels. And by that I mean that the demands in the generation continue to increase and is not able to meet the absolute need of today's need in terms of the different drivers, be it communication, be it all the new technologies that rely on the power source systems. The network that is transmission system equally is overloaded with lots of (inaudible). I can play. I can add my power generation to the networks. In doing so, the question is therefore how do we essentially determine what is the capability, can sustain the new players. The distribution subsystem, as you know, it is where the customers are. We need to turn the dishwasher on quickly. We consume energy in the home. The places are not smart. They consume a lot of energy. Yet the cost of living continues to increase. So the big question facing all of us is what can we do to make the grid of the future smart, to have accessibility, to anticipate the demands that will be coming from different participants; in other words, different players. Overall, there are no revelations. To open the market, to make sure we have new technologies in this system.

>> Jerry Beilinson: There are several different points you raise. Certainly the Smart Grid is exactly what you were just talking about. Dr. Bose, I wanted to ask you, in a Smart Grid, it's something that we've been interested in addressing today, but it's a pretty broad term. So what are people talking about when they talk about the Smart Grid? What does it actually mean?

>> Dr. Bose: Well, I think the Smart Grid usually refers to the fact that we can put in a lot more computers and communications and control on top of the grid to make it work better and more efficiently and so on. Once you start getting into the details, of course, you have to look at different parts of that, how do you make the transmission much smarter. How do you make the distribution part of it much smarter? How do you make the consumer end of it much smarter? (Webcast experiencing technical issues with Dr. Bose's audio feed.)

>> Jerry Beilinson: Dr. Bose, I'm sorry. We're having a little bit of a problem that I'm sure our technicians will get it fixed quickly. Dr. Anderson, turning to you, I know in New York you're working with ConEd, a local utility there, and some of the Smart Grid technologies are already being put in place.

>> Dr. Anderson: We're trying to give it capability for the city to do twice as much work in transportation and enjoyment of life with the same amount of power. We're trying to create more of an internet-like system where we can route power from sources to sinks, what we call load pockets, places where we have difficulty getting power to and from and consequently we have to overpower the entire city and we have excess power. So if we could make it more efficient using computer-aided systems, much like Google makes search engines more efficient. Same technology. We have high hopes of being able to accomplish that task.

>> Jerry Beilinson: So the analogy is you send an email out and it may take different routes, but eventually the packet of information gets where it needs to go. How does that work on the energy level? What's the--how does the analogy play out?

>> Dr. Anderson: Well, in the electricity world in the distribution side the electrons slosh around to wherever the low spots--think of it as the ocean sloshing around. We distribute it into neighborhoods in New York that we call networks. Those networks are highly populated sometimes and during the day, for instance, Manhattan has eight million people in it during the day and then at 5:00 p.m. they all distribute out into the Bronx, Brooklyn and Queens. And the distribution of that power--right now we have to keep maximum power for both locations even though we only use it at 2:00 p.m. in Manhattan and at 8:00 p.m. in Brooklyn.

>> Jerry Beilinson: So you want to be able to move that power where it's needed.

>> Dr. Anderson: Just like the emails. Shift the power to where it's needed and therefore not have to have such an excess capacity in the system.

>> Jerry Beilinson: Right. That's part of the power optimization, Dr. Momoh, that you talk about, I assume.

>> Dr. Momoh: Yes. (Inaudible) from the generation point of view, (inaudible). Generation has to suspect, anticipate the demands on it. You have to make sure that the new generations are so optimized. Beyond that, you now have the transmission system, where you are trying to make sure you push power efficiently at minimal cost at favorable pricing. To do that, you need a global optimization that anticipates problems and provide solutions, at the same time optimize control systems that will allow you to push optimal power through. Given that, then you go to optimizing the distribution subsystem. There again, you are trying to make sure the customized (inaudible). To make sure it's paying affordable prices, you have to ensure that energy loss is minimized. You also have to make sure the price signal is appropriate, in which case you price incentive, to want to use less, is part of the equation in optimizing. So there are market issues. There are also tax credits you have to give to people. There are barriers you have to overcome if you trying to optimize the enterprise. And those barriers, some are regulatory, some are lack of investment. Some are also lack of research, R and D support, to make this happen. Again, the customer has a choice. It's part of the equation that you must put in.

>> Jerry Beilinson: I want to see if we have the sound back for Dr. Bose. Is that working now? Dr. Bose, are you up? Do we have you online?

>> Dr. Bose: I can hear you perfectly. I don't know if you can hear me.

>> Jerry Beilinson: Yeah. We do have you back now. Thank you. A little earlier we were talking about the blackout of 2003, and I thought that would be a good example, because this can be fairly abstract, this discussion of electricity and the grid. Can you take us through how that actually took place? I know that it was essentially a transmission level problem. And what the Smart Grid--if the Smart Grid were in place the way you want it to be, how that would have been prevented.

>> Dr. Bose: Well, the main thing that happened for the 2003 blackout was that the problems in Cleveland were not being detected by the operators in Cleveland, who were looking at the computers and not seeing any of the problems happening because of computer malfunction. And because they didn't know that things were going wrong, the disturbance cascaded right through Pennsylvania, New York and Ontario to put 50 million people without lights. So the main thing, of course, it started with the computer malfunction and that's where the Smart Grid comes in, where you're not dependent on one or one area's computers to manage the whole grid. And the Smart Grid would make it more available to other operators, ways to correct the situation and take corrective action. So I think the Smart Grid at the transmission level would come in very much, very much into play if we did that. Now, there were other ways of, of course, saving some of the distribution areas that would have also come into play if we had a lot more sensors and computer control on the distribution side as well.

>> Jerry Beilinson: Mm-hmm.

>> Dr. Anderson: On that distribution side, in New York we were back in power in a couple of days, but one thing that it pointed out, it took us--after we got power established to the neighborhoods in the city, it took another 24 hours after that before the subways got up and running. And we realized that the electric economy of the future, where we're all headed, in which we're all going to have electric cars and subways and trains and we're going to have excess demand on the grid, we're going to have an integrated system that has these other infrastructures in it. For example, water. Three days of water supply in all of the vertical towers that people live in in New York City, all that water is pumped up with electric pumps. So after three days we're out of lights, we're out of subway and water. The sewers work, but the sewage doesn't get clean. And the natural gas is pumped also. So the infrastructure of the whole system needs to be understood by these computer systems of the future.

>> Jerry Beilinson: Mm-hmm.

>> Dr. Momoh: The distribution subsystem seems to be (inaudible) incorporate some of the solutions that we have in building the future intelligent breed. At the distribution level, typically it's radio. Being radio, there are problems of restoration and so on, so forth, to restore the system when power fails. What you do, therefore, is to come up with new and better systems which may well--communications subsystems (inaudible). An example therefore would be do we have an automated function that will be able to reconfigure the system. So we are looking for distribution functions that will reconfigure the reconfiguration system, put the customer in the right loop or feeder so they can get power at the right time and on time. At the same time, you have to restart the system quickly, and to do that you need an intelligent system that will allow you, give you the human ability to think, to solve the problem (inaudible) rather than doing manual, in the feeder, for example, (inaudible) reestablish customers at the appropriate time. So you need a distribution which involves automatic billing, automatic metering, smart meters that will be able to tell when (inaudible), so on, so forth. So it is indeed true that distribution subsystem there are some quick fix we could do that could make this network very smart.

>> Dr. Anderson: Beyond this quick fix, really quickly, though, you'd have to be able to turn it into a smart machine. And a smart machine, more like a plane, is going to fix itself, or more like your car, is going to have several computers that are predicting what's going to fail, not fixing it after it's failed, but predicting it before it fails so that you could keep the system up. And there we have peers from New York City standpoint, for instance. Shanghai, London, Paris are headed to New York for a meeting next week where we talk about our integrated problems. And we find the same problem in all vertical integrated cities around the world.

>> Jerry Beilinson: Sure. I'd like to talk more about that and get to the customer level of the smart metering and what the future would feel like to somebody on the grid. But I just want to check with the operator to see if we have any questions online right now. Hello? I'll take that--

>> Male Speaker: As no.

>> Jerry Beilinson: I'll take that as no. I don't think we have questions online yet. So the question that you just brought up brought up with smart metering and how the individual household can adjust to the amount of electricity they use I think would be of a great deal of interest to investigate. We have a piece of art, if we could pull up, the connected home of the future. Perhaps, Dr. Anderson, you can actually take us through what we're looking at. This obviously is a home that has advanced meters in place. How does that work?

>> Dr. Anderson: Well, everybody's familiar with Energy Star. Your appliances have a little star on them, they're very efficient. We're starting to talk about within the electric economy of the future what we're calling with ConEd the energy double star world, where the appliances are smart, they negotiate with each other without any human interaction. In a larger apartment building, for instance, the refrigerators will turn themselves off for an hour a day in sequence as they rotate around the apartment building. The other big problem we have is if we go to electric cars and everybody drives to work with electric car and then plugs it in, we even have designs for plugging into your parking meter in New York City. You suddenly load up the grid with massive amounts of additional draw that we'd like to dampen out. In order to do that, you have to be able to introduce storage as well as generation. Alternative energy, such as solar, wind, anything you can think of, needs to be stored so that it can efficiently be delivered, not just delivered when the peak demand is. That's the idea of this cartoon, is that you have everything communicating with each other without humans, negotiating, using perhaps price signals to decide when--one refrigerator makes a bid to another refrigerator, I really need to keep my beer cold, so I'm going to give you an extra $2 off your bill because I'm going to pay it for you. The other refrigerator says fine.

>> Jerry Beilinson: Right. And that's in response to the larger grid sending a signal saying we are getting close to peak here, the situation is getting a little (inaudible), who is willing to reduce their demand?

>> Dr. Momoh: Yeah. I think that makes sense. That is where we are looking at--when I said the lower hand inputs, when I mean we have quick fix, what I really mean is we have quick opportunity to make impact, to demonstrate the capability of these so-called Smart Grid. These are systems, you now have an opportunity to say, okay, if I put a cheap (inaudible) in homes (inaudible) a dishwasher, for example, how much you need to have it on or off. By turning it off when you don't need it, you are saving energy. So the equivalent of that is that you are allowing us to balance the energy requirement from end to end. If I limit the generation at this point, then it is wise that a customer has participated in minimizing the energy required at home. So by doing that, you helping to stabilize the network or the system itself. And that is why I said you can now provide a price incentive for those customers that are already cooperating so much to reduce the amount in their different homes. If we're able to do that, we might be saving a lot of money, whereas generating company will have been trying to minimize loss, customize (inaudible) is helping me to reduce the demand. It's a win/win in this case. It is that kind of home that we call the intelligent homes of the future. They are smart. They know the energy required by all the appliances are controlled and communicated through a central box, like you have your internet web page, where you can see how much energy is used by each of the devices in the home. Remotely, therefore, you can shut down or turn on as you need. I'm saying the price savings from that should be part of the equation that we use to have your public perception of good public policy in this case with our customers.

>> Jerry Beilinson: Right. One of the things that that will do, of course, is that if we can bring down peak demand at certain times, we can get a lot more stability in the grid without actually raising the generating capacity by a tremendous amount. Dr. Bose, I thought I'd come back and ask you about that. I know that you focused on transmission of electricity. How does that same kind of concept work on the transmission level, which is going from the power plants down to the level of the local neighborhood or home?

>> Dr. Bose: Well, I think it's somewhat similar. That is, anytime you are doing intelligent grid, you have two aspects of it. One is to sort of match the load to the generation and also make sure that you're doing it at the cheapest possible way and saving money at it. But the other important thing that you cannot do today at all is the speed at which you can take corrective action. That is, when you have a--at the transmission level, when you have cascading events, that could happen in milliseconds, and there is no way that the operator can react to do something and take corrective action after things have started cascading, whereas computers can, computers and controls and so on. And so that's one of the major things that the transmission system has to focus on to avoid these cascading events that have occurred. And the cascading could happen because of just bad luck or a storm going through.

>> Jerry Beilinson: Just to be really clear, what kind of corrective action would be taken? What does that mean? What would the grid do?

>> Dr. Bose: Well, for example, the simplest one to think of is that you could isolate parts of the grid that are in trouble to make sure that the problem doesn't cascade to the other parts. For example, the 2003 problem, it could have been contained into the Cleveland region alone, which would have been only a couple of million people rather than 50 million people in Canada and the U.S., which is what it did. And these new systems we're talking about can certainly do that.

>> Jerry Beilinson: Okay. Yes.

>> Dr. Momoh: I would like to join him in the transmission subsystems. With respect to what he is saying, it makes sense. But we are still assuming that the transmission of the future is going to be deterministic. I'm not thinking so. I'm assuming that there is going to be a lot of anticipation in terms of the demands. There is going to be therefore be a need for predictivity capability of the transmission system of the future to predict when storm is coming, to predict when there is overload, to predict temperature change, to predict when the demands are going to be more than the designed limits. Core engineers currently have this idea of the limit can be plus or minus 10%. I say no. That is deterministic. But when you are looking at randomness in terms of what was possible to happen. It's true that is needed. Therefore, it's what we are currently being sponsored by the National Science Foundation and (inaudible), to look at this so-called dynamic optimization of the grid. This idea means the grid, the optimization strategy will predict, will anticipate. What if a (inaudible), will predict for 90%. But is already at 10. Can you do it? Can you handle it? So this kind of enabler, this software embedded in smart devices that are capable, that are controllable with communication strategy, of course, is the real way to go in designing and deploying and demonstrating the future intelligent grid and transmission system. That means you must have new paradigm of specification of intelligent softwares that are going to be anticipatory, predictive and adapt to new situations.

>> Dr. Anderson: I've got a great example of that. A great example of that was in Texas a month ago.

>> Jerry Beilinson: I'm sorry, but I want to hear that and then afterwards we'll pause. We're starting to have a couple of questions. Please go ahead.

>> Dr. Anderson: I just wanted to say a great example a month ago in Texas where we now have a really significant portion of the grid supplied by wind power. The winds stopped blowing after several days in which it had blown hugely. Because the wind was blowing so much, power plants came off-line to do maintenance. And then the wind stopped blowing and they couldn't get the maintenance back on in time. That shows you the integration of a smart system that has to go beyond just the electric grid. We call it lean systems in other industries. The electric utility industry is not lean. It doesn't know what's happening at the transmission side, at the power plant maintenance level, when wind or solar farms are producing. That is very doable. It's being done in virtually every other major industry in the world right now.

>> Jerry Beilinson: Okay. If I could just take a question, this is from Martin Schroeder, the American Public Transportation Association, and he's asking what steps are being taken to organize and inform major users, like public transportation authorities running electric train systems and subways in how to prepare for the coming Smart Grid?

>> Dr. Anderson: That's exactly what we're trying to do in New York, is get the NTA and the New York power authority and the ConEdison together in terms of computer systems so that one knows the other. Each sees something different. The drawdown from the electric grid is seen as power on the subway trains, and all of that system--those two systems were built completely independently. And the subway system in New York used to have its own power plants. Now they're using the integrated power system of New York, New York state, all the way into Canada. And the visibility that's required from a computer-sensing standpoint has to cross the boundaries that have been artificially set, actually, between metropolitan transfer authority, metropolitan power authority, blah, blah, blah. It's a very interesting problem in terms of breaking down silos.

>> Dr. Momoh: In addition, I think they should prepare for environmental problems, global climate change concern that we all have. When you are solving this issue, we must not forget the need for environmental impact. So as we encourage new participants, like the transportation authority, of course we will enjoy using them as immobile power, possible power because they already have it, so we might say, well, they might be connected to the grid when it's not being used, when the trains or buses are not being used. That's one idea.

>> Dr. Anderson: Another one is to give them carbon credits to the transportation authority for taking all those automobiles off the road. Taking money from transportation and moving it over into public transportation is a radical thought. You all know what Mayor Bloomberg is trying to do and what the City of Glendon has already done. Those kind of dramatic changes in transportation patterns have direct effect not only on the environment, but also on the electric grid.

>> Jerry Beilinson: Correct. Do we have any questions on the line now? I don't know if we have our operator on the line.

>> Operator: We have one question.

>> Jerry Beilinson: Go ahead.

>> Operator: Your line is open. Go ahead, sir.

>> Audience: I previously asked the question if we could see the graphic. This is Bob Wilson in Lakewood, Colorado, where we're building a big transit system. The question that came up is the design of the Smart Grid, from my background as a former field engineer and holder of power system clearances, we got to think about the safety aspects for the general public, and especially the electrical workers. How is that going to be incorporated in this very noble project?

>> Dr. Anderson: Again, this is an industry where we still electrocute several workers per year in New York alone. You have to build the system so that you--the computer system so that the humans who work in and around it know what's energized and what's not.

>> Jerry Beilinson: So you're saying there's an increased--I think the caller is asking whether there is an increased danger because the grid is actually by itself. In effect what you're saying, sending power to certain lines or reducing power to other areas.

>> Dr. Anderson: You have to be able to make sure that the workers who are out there. That's the people that I'm mostly worried about. The question was probably related to the public as well. We don't want to kill any subway riders.

>> Audience: Some of the increase in danger is actually being put into the system because we are allowing generation to be put into the distribution system which allows power to flow in the other direction than it normally does. And that has put in a lot of new ideas now into how to protect the system and the people who work on that system. This is a big question because what it means is that the technicians and so on that work on the systems have to be trained differently than they used to be in the past. So this is partly because of the intelligence being put into the system and partly because we are putting more distributor generation on the system at the other end of the line.

>> Jerry Beilinson: If I just ask you, Dr. Bose, you're talking about distributed generation, which is actually the kind of idea that at Popular Mechanics we've written about several times and just sounds like a great idea. Could you explain to people who might not know just what that is and what the benefits are, what sort of the goal of distributed generation is?

>> Dr. Bose: Well, the distributed generation would mean that if you have a large number of wind turbines and so forth concentrated on a wind farm, is to spread that around in a larger area connected to different points, okay? A simpler example would be if a lot of us bought electric plug-in cars, we could use those to actually supply the grid during the time when there's a shortage of power from the generating stations. So that now you're plugging it into your home, whereas instead of the electricity coming from the electric company into your home, now you are supplying power into the grid. And that's what I meant by distributed generation and the electricity flowing in the other direction than it normally does. And that means that the whole design pattern of the distribution system--and this is what I think Roger was referring to, has to be now much more protected by computers and other devices from the system itself as well as endangering any of the people on the other end.

>> Jerry Beilinson: We can also take questions from the live audience. We just need to get a microphone or else it won't be picked up for people online. But we do have a question here. Go ahead, sir.

>> Audience: Hi, there. Is it on? This is (inaudible), Department of Energy. We have, of course, a very vigorous research program in Smart Grid technology and we found a lot of work. But we realize that you need an analog counterpart to the digital part. And what we need is not only distributed intelligence, but, as was mentioned, distributed generation and distributed storage.

>> Dr. Anderson: Storage, yes.

>> Audience: In particular the storage that provides the buffer that will make all of this work much better. It's trying to say the electronics will respond instantaneously, but with a buffer, it will work even better.

>> Dr. Anderson: That's absolutely correct. Every big computer server farm, or even small ones these days, has a universal power storage to accommodate the blinking. In Paris, for instance, they allow blinking, they have so much support of their computer systems. It has to be extended to the house, to the home, the basement of apartment buildings, subway system having batteries and storage. The storage is really critical to pulling off the distributed generation.

>> Dr. Momoh: With that generation, the question original was (inaudible), and I thought I should coin one more S, which is the security of the system itself. Whereas we develop a new technology such as (inaudible) generation sources, there are issues with (inaudible) with regard to the (inaudible). Therefore, I think technology that will be needed will be control technology, just to be able to detect the malfunction of these devices and provide corrective measures in a timely fashion. Security, of course, the overall security of this new system we are talking about, security against enemy, against attack. Because there are going to be a lot of communications systems embedded in the electricity system. So what does that mean with respect to (inaudible) cyber attack? We got to be safe, safety in terms of electrical, safety in terms of you put this out there, you have not educated customers properly. So my third point is customer satisfaction in the process of all this.

>> Jerry Beilinson: Right.

>> Dr. Momoh: How much will it cost to put all this together. So, yes, we need to include storage. We need to include distribution generation, that's all four types. I give you one example just to close this gap. The wind energy type. Sure, it's one of the DG options, called distributed generation options. But then it's built on this speed. What if you designed this wind energy source and the speed go out of order? Before you know it, it rotates at an enormous speed. What you find very quickly is it would just fly out of the region, the region of interest, of safety, and it could easily hurt a lot of people. So we need to design a control strategy that will monitor the speed of your distributed generation to be sure it's within safe limit. And I can give you an example of all the DG options, how you have to make sure you protect them and anticipate problems. All this anticipated, not after the fact.

>> Jerry Beilinson: Thank you. One of the distributed--I'm sorry. Go ahead.

>> Audience: My name is Dan (inaudible). I'm with Tetrologics. I have a question at a larger scale. Given that we have facilities, infrastructure in place that we can use more efficiently, prolong its life, get more benefits from it, we still have to invest in enough infrastructure to provide reliability at the catastrophic stand; that is, extraordinarily strong power demands for short periods of time, storms that take things out. It's generally not in the economic interest of any particular utility or any particular individual to provide that kind of reliability. Doesn't make economic sense. But we as a society need that reliability. What kind of a regulatory framework would you guys propose to take care of that problem?

>> Dr. Momoh: If I were to ask that question, what do we do first in building the Smart Grid? You have to get your regulations right. You have to get your standards right. It will be a waste of time and effort if we don't agree on what the standards should be. Because I will be building my own intelligent systems, technology for intelligent grid or nobody use it at the end of the day because you building yours, I building mine, and they're not compatible. So the first step is to make sure we have regulation, public policy, public awareness, public education to make sure that we all agree in terms of the framework in which we're going to build the different technologies that is going to be embedded or enhance the existing system to make them smart. We're not going to throw what we have away. We going to embed software into intelligent software schemes, put in some intelligent devices, we're going to add some added value of communication systems. We're going to use human capacity. By law, you going to regulate. But you got to also ask legislation pass, to support research that will make this happen. Otherwise, there will be no one to back it up. We need human intelligence as well as the grid intelligence to converge at one time to solve this problem. It doesn't start at the same time. There are short-term goals. There are mid-term goals. There are long-term goals that we have in building this. We not trying to build an intelligent grid by tomorrow. We're saying that in the next five years what are we going to do? What can be reachable? Then we look another 10-year span, then another 15, 20-year span. I'm suggesting that by 2020 what is our vision? Vision in technology development, in support of generation, transmission, distribution, subsystems, vision for the operational support for the system, the planning over 100 years, of course. Then you have to look at the regulation. And the people, the culture also has to come in. We would like to sustain the quality of life as we know it. That is why we are supporting the idea of resiliency of the network, sustainability of the network, because we feel that our lifestyle has come to be so good and we depend on these resources, sorry, on the energy and the resource constraints that goes with it, be it telecommunications, be it water, be it transportation, be it people, and we want to sustain it. Energy will be one of the elements to make it happen. A smart home will be one of the elements, one of the subsystems that we must put together to make it happen. Of course I'm thinking that tomorrow I can walk to a Best Buy and pick up something to make my home smart, save on my energy bill. Maybe she will want to redo the bill tomorrow.

>> Jerry Beilinson: Dr. Bose, it looks like you had something to add.

>> Dr. Bose: Could I jump in here for a minute?

>> Jerry Beilinson: We'd love for you to.

>> Dr. Bose: I think the reliability of the grid being a public good is an interesting question, because we in the U.S. have a slightly different structure of the power grid than they do in most other countries, which is government-owned. And they can actually--they actually in other countries can determine the reliability for which we are willing to pay. Here, because private companies have to come in, the way we have to make sure that the reliability is there for the public good is by standardization and by regulatory moves. And that makes it much more interesting because the way utilities and power companies recover their investment is through rate regulation by states. So we have a round robin problem in this country which requires sort of a high-level look at how to make sure that the right amount of reliability is built into the grid for the public good that we think is necessary.

>> Jerry Beilinson: I have a question. Maybe, Dr. Bose, you can help us get at that. Reliability of the grid would have to do with a number of different elements, but certainly as demand is likely to rise, how much can be done in terms of a Smart Grid technology and to what extent do we just need to build more transmission lines and more power plants? I think the country now uses about four trillion kilowatt hours of electricity each year and that number is likely to rise. So how do you balance the various ways we have of increasing reliability?

>> Dr. Bose: Yeah. Another way to look at that statistics is 40 percent of all energy consumed in the United States is actually turned into electricity. So we consume it as electricity. The interesting thing that we have to point out is we can put a tremendous amount of intelligence under the system which consists of computers, control and communications. None of that produces any energy or transmit any energy. So ultimately the reliability of the grid fundamentally has to be met by building more generation and transmission and distribution to meet the power demand or the energy demand of the country. What we do by increasing the intelligence of the grid is better optimization, better economic benefits, better reliability, so we are taking what we have as an infrastructure, which has to be the basis of the reliability, and then trying to squeeze the last element out of that by doing it intelligently.

>> Jerry Beilinson: All right. I know, Dr. Anderson, you had some ideas about how much we can increase the transmission capacity in places like New York.

>> Dr. Anderson: Well, think of the 40 percent number for now. What will it be in 20 years? 30 years? We believe it will be 80 percent is turned into electricity. That's what we talk about when we say the electric economy. And you just cannot put that much wire in the ground. All of ours in New York City is underground, and we have the largest copper mine in the world already in our streets. We're running out of room. So hopefully we can figure out how to transmit it more efficiently and do more work with less electricity. That means efficiency. That's what efficiency is. That's why we keep casting our glances at leaner industries that do just that. The airplane, the 787 is going to fly on half the fuel that the 777 does.

>> Jerry Beilinson: Assuming they get it built.

>> Dr. Anderson: Assuming they ever get it in the air. They will. They will.

>> Jerry Beilinson: We have another question from the audience here.

>> Audience: There are a lot of national issues at stake here. As a consumer I have to confess the rising cost of electricity and the environmental effects with the CO2 are the things that really worry me. The other things seem pretty small compared to that. I look at the way our natural gas bills have been going up, and my electricity bills for the daytime are going up because we have to import natural gas more and more because of these little natural gas generators they're building in the big cities. That takes us back what you guys were saying earlier, that renewable energy is an essential part of what we need to do. And then, as you said, we need storage for renewable energy or for other reasons because if the wind blows sometimes and not others, we better be able to store it if we want to be able to use it. But there's a big hole in the discussion. The old way of running the power grid that 90 percent of the experts still do today is some combination of local little algorithms that don't know the big picture, human operators that get mixed up and lead us to blackouts sometimes because they don't know how to cope with the information and automated systems. The only automated systems I know of today to do this optimization of power flow is the optimal power flow kind of algorithm that Jim Momoh here has developed here. He's made it real. But the problem is what we have today is static. It gives you the best power flow today. But if you want to use storage, you can't use storage efficiently unless you have a new kind of intelligence. It's not enough to have a lot of computers. It's not enough to give contracts to people who do it the old way. You need to extend these kinds of algorithms to be able to handle the problem of anticipation and time shifting. And that's where there's some new mathematics coming online, but it's going to take a major paradigm shift because most of the world isn't doing it yet, and that's a big part of the struggle, to get beyond words about intelligence to making it really real.

>> Jerry Beilinson: Dr. Anderson, maybe you can discuss that.

>> Dr. Anderson: Yeah. Even in the energy business, refineries and liquid natural gas facilities have to see the market in the future. They build their economic optimization based on profitability. And they have to see the future. The future also includes weather and all of these uncertainties. There are total plant software systems right now running virtually all of the high-tech refineries and petrochemical plants in the world that would be the envy of the electricity industry. It's not that the optimization problem is not seeing the future. It's that the electric optimization problem is not seeing the future yet.

>> Jerry Beilinson: Now, you do have some predictive software in ConEd you can talk about.

>> Dr. Anderson: That's right. Where did it come from? It came from developments in other industries. I mean--

>> Jerry Beilinson: And the idea there, just to elaborate on what I saw Dr. Anderson was trying to show me around the grid control center in New York, you were there to predict what might fail so that repairs could be made in advance. Maybe you could describe how that works.

>> Dr. Anderson: Yeah. So every system in mechanical factories and machinery in general know about sensing vibration and wear and tear. Jet engines go in for repair basically by calling home themselves and saying I'm not quite feeling right, I need to have somebody look at me. A car does that. The electric grid has a lot of the sensors, has a lot of the autopsies done on the failed systems so that you can develop failure models of the components, like a transformer or a feeder or a--or a--anything. What it doesn't have is the system integration that understands the plane crash, as opposed to the equipment failure. So a plane doesn't crash anymore because of one system failure or two system failures. It takes three, three or more. Three will take a plane down, even though you have redundant systems in each of those three. In order to build that integration of those odds of failure, you have to have a really good, solid model of how the system works. Toyota built a better Lexus by building a model of the Lexus. Only after they had the car model in the computer could they build the optimal car for Toyota. We are struggling with building that optimal model. For New York City, for example, there are five million nodes, five million connection points to the electric grid just inside city limits alone, and we don't even have it connected to the transmission grid. Those are massive computational problems, but they're not insurmountable. They're simply not.

>> Jerry Beilinson: Let's see if there's any other questions on the line. I want to not ignore our phone call participants. Is there anybody on the line? No? Okay. Go ahead.

>> Dr. Momoh: I'd like to just add a little comment on the issue of the need for advanced strategies, which we are referring to as optimization. In fact, in the energy bill signed by the President in December, (inaudible) There is a need for that optimization. And if you pay attention to what one of the audience said, that we do need a global, adaptive optimization of the grid. It has to be funded. There are only a few people who wants to do this kind of work, because some of our colleagues are used to the old, traditional optimization, which I myself participated in. They want to keep this. They don't want to take new risk, don't want to fund new ideas. This is the first step we must take. Because in doing that, these new technique will be able to look into the future, will have anticipation for what if the change in the flows in the network suddenly increase. Are we ready for it? What if the population demand for power changes radically. We don't even know what the customers have out there. We know they are using all kinds of gadgets now. Then we also have to ask what if the source of energy itself dwindles or suddenly changes its pattern? Imagine that. We don't even know the color of power we have. The electricity has no color. It's just color. We don't even know whose power we are using in the old system. So if you are developing a tool, you have to think carefully will my tool be able to have this predictability. And the ability to do that is for us education also. You need to fund research of students of the future, whose mindset is about new things, just like you said in the auto industry. If I go to see my doctor, he has a history of my--he has a small history of my family history, of my health history. He can predict how long I will live, maybe, what type of medicine I should take, maybe, to sustain my life. This is the type of remedy we don't have. I call it remedy for detection. I call it remedy for prevention. So in a different subsystem of power, whether that be in the generational system, this new thinking has to be there. In fact, the word smart means taking my human ability to predict, to think, and put it in system that don't talk before, that looks just silent (inaudible). Degeneration only hum, make noise. That's all they do. They don't think. But can we make them think, to anticipate what is going to happen? I think I have a picture like that in one of my (inaudible).

>> Jerry Beilinson: Is that the generation transmission distribution--

>> Dr. Momoh: Yes.

>> Jerry Beilinson: I think we do have that.

>> Dr. Momoh: You might. (Inaudible) I have this in the (inaudible) posted by my students. We think when we grow up, when we finally get there, we'll have this network all talking to each other intelligently. It is at that point that degeneration mix, before you were asking, one of the audience, we have to integrate with renewable resources in a smart way, minimizing all the pollution of the world at the same time, optimizing the output, either through materials that you are to come up with, smart materials, superconductors, nanotechnology. All those things will be embedded in that. I'm not saying it will be done tomorrow. I have a time frame for all of you. Then you see that generational mix will predict the demand from the intelligent networks, the transmission system. The transmission system again would distribute the controls that are also located intelligently. One of my colleagues, (inaudible) talks about it, how to use this intelligent system to schedule flexible transmission devices that can grow the network without getting right away, without having to wait to build a new network, okay? A new (inaudible) conductor will be built (inaudible) that will enhance the capacity of the transmission system intelligently, software-wise. Thirdly is what we've been saying all day, the smart consumption.

>> Jerry Beilinson: Right.

>> Dr. Momoh: You can now also bring in an optimizer that optimizes the energy usage and the reliability you are looking for. You were talking about reliability earlier. By the way, there is I call society acceptable reliability. What is that definition? In other parts of the world, they took it for granted that, well, you can have power failure every day on a given day of the week. But that is because society allows it. In the U.S. we are defining reliability, it's not just a number. It has so many parameters with it. It has a GDP issue with it. It tells the population the region where you live in defining what I call your reliability. So by definition we are coining at Howard University a new definition of reliability. It's not just (inaudible) on soft energy anymore. It's societal expected on soft energy.

>> Jerry Beilinson: Thank you, Dr. Momoh. I just want to see if there's one more question in the audience. We'll just need to get you a microphone. Oh, you have one. Great.

>> Audience: I'm a professor of power at Virginia Tech. I would like just to argue a little bit about the capability to predict failures in a system like the power system, which is extremely large scale, continental scale, which make it different from let's say an airplane or a car. We don't have hundred of thousands of components here. We have millions of components distributed over large-scale areas. To make the system resilient--and I'm going to make a difference between resiliency and reliability. They are very different. Reliability, you build reliability through redundancy. I totally agree with the panel on that. Resiliency is very different, to make the system be able to respond extremely fast in a time constant, in few milliseconds, to eventually unpredicted failure. That means that no matter what your capability of intelligence and predictability, you cannot predict all failures. There are some failures that are unpredictable and you have to face that possibility. There is a risk there. The risk cannot be zero. Okay? There is a risk. So you need to make the system agile. And for that you need to have distributed, coordinated control. What is missing today in power system, we have distributed protection system that are not so well coordinated. You need to coordinate those actions, and in addition you need to put the load in the picture. The load can contribute in a very effective manner to make the system resilient to cascading failure. And we have a very good example. The question is resiliency. So the distributed generation, the load can contribute to resiliency of the system.

>> Jerry Beilinson: All right. I'm sorry actually to cut you off. I'm being told that I have about two minutes left. I'd like to give people a chance to respond to that and also to sort of bring up anything that we haven't covered yet. And, Dr. Bose, maybe we can turn to you. You might have a response to that. And also we want to make sure if you have any final comments that you think we should walk away from here with.

>> Dr. Bose: Well, I thought that it was interesting, the last two comments from Professor Mealy (phonetic) and Professor Momoh were pointing to the fact that there are questions that needed to be resolved by doing a lot more research and development. And one of the issues I think we should leave with is that research and development in the power grid has been tremendously lacking in the United States in terms of the amount of money you would expect that would be spent on something that is a $100 billion industry that we are looking at. So that's one thing. The added--because there hasn't been much research and development, there has been also an issue about where the new people are going to come from into this industry who are going to solve these problems, who are going to build the intelligence in this. We're not training enough engineers and scientists and people and even technicians who can do this work. And these are major issues that we have to face. We can talk forever about how intelligent this system ought to be and how reliable it ought to be, but it's not going to happen if we don't solve some of the other infrastructural problems which has to do with people and education and research.

>> Jerry Beilinson: Thank you very much, Dr. Bose. I want to thank you before we end for participating very much. That's actually almost a perfect comment for us to close with because part of the mission, of course, at Popular Mechanics and certainly National Science Foundation is to encourage the best ideas and the smartest people to go into solving the country's problems. That's something we're all engaged with here, and we have on the panel with us educators who are dedicated to doing exactly that. So I am being told we do have to close right now. I want to thank you also both very much for being here and to encourage anyone watching from home to rejoin us when we look at water infrastructure issues and issues related to bridge building later this afternoon. So thank you very much. Once again, this is Jerry Beilinson for Popular Mechanics and the National Science Foundation. Thank you.

>> Male Speaker: Thank you.

(Applause)

Return to BridgesWebcast Conference page.