Only one thing is worse than the lights not coming on when the switch is flicked—and that's the lights going out right afterward. The fact that the problem is most often a burned-out lightbulb is testimony to the reliability of what's sometimes called the world's largest machine—the U.S. transmission and distribution grid for electricity.

But that reliability is tenuous at best and perhaps temporary: the machine needs an update to meet increasing demands for more electricity and to deliver it reliably and safely, according to the Obama administration and others. "If Alexander Graham Bell returned to Earth today, the progress in telecommunications over the last 125 years would be mystifying," said Robert Catell, chairman of the New York State Smart Grid Consortium, at a smart grid event in New York City at New York University (NYU) in February. "If Thomas Edison came back today, not only would he recognize our electricity system, he could probably fix it" when problems arise.

That's no surprise: Today's grid was largely finished by the 1970s and contains mostly the same system of devices in use since the 1920s. And, after the wholesale power market was deregulated in 1992, many utility companies stopped investing in the grid—leaving it in a perilous state of disrepair today. An update could cost, according to some estimates, as much as $1 trillion over the next several decades—the stimulus plan alone provided $11 billion for a smart grid, including 32 demonstration projects in 21 states administered by the U.S. Department of Energy.

What exactly a smart grid is depends on who you ask. "What really makes a grid smart?" asked electrical engineer Farshad Khorrami of Polytechnic Institute of New York University at the February event. His answer: "the control system in that grid." In essence, it's the telecommunications and information technology industries applying their innovations to the infrastructure that made computers possible, in large part, or overlaying the utility infrastructure with communications and control systems that will allow energy technology to be more productive.

Grid full of detectors
According to Khorrami, that means a wide array of new two-way sensors throughout all parts of the grid, detecting voltage, current, power, temperature, pressure, wind, sunlight, anomalies, stress, failures, hacking and more. "We want to be as comfortable with the grid running itself as we are with pilots walking out of the cabin and knowing the plane is flying itself," Khorrami said. The smart grid also needs to be "like the Internet in that adding another computer doesn't stress the rest of the system, for when people add a new generator," such as solar panels on their roof, he noted.

In fact, the most important part of the smart grid will likely be in the home. "The electric meter hasn't changed a lot since it was invented," said Thomas Kuhn, president of industry group Edison Electric Institute (EEI), based in Washington, D.C., at an EEI event last October. In fact, the 130 million or so electromechanical meters in homes, businesses and industries across the country "are born obsolete," says electrical engineer Reza Ghafurian, a technical leader at the utility company Consolidated Edison in New York City. That's because they are largely the same technology that Elihu Thomson invented in 1888, and they passively record the electricity used by a given household in kilowatt-hours.

Of course, customers aren't buying kilowatt-hours or even electricity per se. Rather, "it's the comforts they're buying," Kuhn said. "The residential consumer is really only interested in making sure the lights are on, the TV's working and the beer is cold," Catell noted. "An educated and informed consumer is the best weapon in the war against energy demand, and the smart grid is the best way to educate the consumer."

Better displays, smarter consumers
More specifically, the best teachers are smart meters and energy-use displays that reveal everything from how your electricity consumption compares with your neighbors' to the best time of day to run your dishwasher on cheap electricity. The initial prognosis from smart grid pilots, such as the Olympic Peninsula Project led by the Pacific Northwest National Laboratory, based in Richland, Wash., is that such information can save consumers 10 percent on electricity bills on average; that program is now being expanded from 112 homes to 60,000 throughout the northwestern U.S. over the next five years. "The technology is here to do this today," said IBM's Jane Snowdon, senior manager for smarter building research, who participated in the pilot program.

The smart grid is particularly good for utilities: the Olympic Peninsula Project dropped the peak power usage by 15 percent, and a similar project from Constellation Energy in Baltimore, Md., cut peak power demand by at least 22 percent—and as much as 37 percent—just by alerting customers to the price of power (and hence demand for it) through an orb that changes color, according to Constellation CEO Mayo Shattuck. Hence, it can eliminate the need for some of the redundancy in the system, such as extra distribution cables or substations, which are built-in to deal with a peak in electricity demand that lasts for two to four hours once a year—a major expense for utility companies. "It helps you not to have to build back-up infrastructure," says Con Ed's Ghafurian.

The smart grid may also forestall some of the growth in electricity consumption—forecast to rise by at least 1 percent per year between now and 2035, according to the U.S. Department of Energy: usage will jump from 3,873 billion kilowatt-hours in 2008 to 5,021 billion kilowatt-hours in 2035. Globally, electricity growth will nearly double from 17.3 trillion kilowatt-hours to more than 33 trillion kilowatt-hours. That demand may jump even faster if electric cars begin charging from the grid in the near future. "They're coming," said Andrew Tang, senior director for the smart grid at Pacific Gas & Electric, a utility in northern California, at the October EEI event, noting that car companies have unveiled at least 35 models of plug-in hybrid or electric vehicles.

Bumps in the grid
But the smart grid has already run into resistance. Class-action lawsuits have been filed in California, alleging inaccuracy in the 5.5 million smart meters installed by PG&E, resulting in electricity bills as much as 300 percent higher. The company contends that such increases are a result of already approved rate hikes as well as a hotter than average summer in 2008—but also admits that thousands of meters were improperly installed and have manifested various problems, including communication malfunctions. Such growing pains can literally turn off customers: a smart homes pilot program in Westchester County, N.Y., lowered bills for almost all participants, but still 30 percent quit the program entirely.

Ultimately, the customer pays for all the fixes through rate increases. Whether a consumer with the smart grid saves money or, at least, breaks even "remains to be proven," admitted Aubrey Braz, Con Ed's corporate vice president in charge of smart grid technology.

Balance of power
Another key challenge for a smart grid is the fact that electricity is an instantaneous commodity—it is consumed at the exact same moment that it is produced. Running an electric grid is "harder than rocket science," says Stephen Wright, president of the Bonneville Power Authority based in Portland, Ore., simply because supply and demand must be so closely matched and both vary throughout the day.

Utility companies generally cope by keeping some electricity generation in standby mode—so-called spinning reserve—to meet spikes in demand. By more closely monitoring changes in supply, demand or both, a smart grid could help reduce the burning of fossil fuels, for example, for nothing. "We run about 16 percent spinning reserve. I can't use it, I can't store it, it's just there as a safety net," says Mike Carlson, former chief information officer at Xcel Energy in Minneapolis, Minn., which is running a smart grid city in Boulder, Colo. The smart grid "can take half of those out" by matching "supply and demand instead of being always on."

Installing sensors, smart meters and other smart devices throughout the hundreds of thousands of miles that constitute the nation's electric grid will be the work of decades. "If you do all the smart grid, what is the efficiency improvement we'll gain?" asked Stephen Hammer, executive director of the energy smart cities initiative at the Joint U.S.-China Collaboration on Clean Energy at the NYU event. "How much will it cost, and is that the best way to spend our money rather than building or vehicle efficiency?"

For example, the smart grid city initiative in Boulder alone has cost Xcel and its partners $100 million, or $2,000 per customer. "It's unsustainable and nondeployable at that cost," says Carlson, noting that cost would have to drop to $500 per customer to be viable. "We know these things will be effective in delivering something. The question is: will they justify their cost?"

And, ultimately, it comes down to this equation: one watt of light generally requires burning 400 watts' worth of coal, according to mechanical engineer Saul Griffith of Other Lab in San Francisco, Calif. That equation will need to get smarter, whether it be through the smart grid enabling cleaner, more variable resources like wind generation to be more widely deployed or the smart grid reducing the amount of that essential inefficiency that must be tolerated.

13 Comments

1. 1. Power Hungry 11:12 AM 5/10/10

   The last time I checked, my outdated meter as was stated does not know when peak times are and when energy is cheaper. All it does is record how much power is consumed. So the only one benefiting from cheap power is the power company. Me as a consumer do not get a break for say drying cloths at night when demand is low. The reason is so the power company can keep an even load on the grid at all times. I see a class action for the continuing lies about how consumers save money by using power at low peek times. My dumb meter does not know that the power is cheaper at night then during the day, all it knows is how much power was consumed.

   Reply | Report Abuse | Link to this

2. 2. jtdwyer 11:44 AM 5/10/10

   The only capability that can justify the capital expenditure to replace the current power transmission grid is a significant improvement in long distance transmission efficiency. This requirement is seldom even mentioned as part of the 'Smart Grid' proposal.
   
   The enormous amount of power dissipated into the environment by current long distance transmission facilities requires the generation of enormous amounts of unusable power, consuming enormous resources and tremendously increasing the environmental impact of the power actually consumed.
   
   Increasing the separation distances between power generation facilities and consumers, as proposed by many non-solar, renewable energy generation options, only increases the transmission power losses.
   
   The main justification for smart meters is to provide a market for electronic meter manufacturers (prominently IBM) and to shift the power usage management responsibilities to the consumer.

   Reply | Report Abuse | Link to this

3. 3. MisterA 01:26 PM 5/10/10

   Dream on!

   Reply | Report Abuse | Link to this

4. 4. SpoonmanWoS in reply to Power Hungry 03:43 PM 5/10/10

   @Power Hungry: the problem is you don't live in an area with variable rates. There are a lot of places in the country where you are charged lower rates at off-peak hours, so you would benefit in such a place. Building the smart grid encompasses providing this level of variability to everyone so the benefits available to some are available to all.

   Reply | Report Abuse | Link to this

5. 5. Insania 05:02 PM 5/10/10

   I see smart grid as expanding the role the electric service provider has in providing that service. Smart grid will never result in lowered cost to consumers, and indeed it never has been about that, nor should it be.
   
   Look to the California ISO, MISO, and PJM -- entities whose purpose is to extract market efficiencies, and tell me that they've resulted in lowered costs to consumers. Smart grid is about extracting grid efficiencies, but the cost of extracting this 10% to 15% through smart grid technologies will be greater than the savings in the same way the ISOs have spent more to develop and operate energy markets than they've saved. In my opinion, smart grid does not benefit consumers.
   
   Again, it's about expanding the role your provider has in providing you electric service. Utilities will hire legions of smart grid managers, integration middleware engineers, automation technologists...and consumers still might choose to turn on their AC at 4:30 PM when it's hot -- because it's hot. Regardless of price.

   Reply | Report Abuse | Link to this

6. 6. Macrocompassion 03:40 AM 5/11/10

   Whats the true situation about supply and demand of electrical power? When demand is high there is a small reduction in voltage over the line, this should immediately tell the power supply source to increase the flow. Similary when demand is low, the voltage rises lightly and the supplier should know to provide less juice. This simple idea is known in the electronics business as "negative feed-back" which is about as smart as any power supplier needs to be. So what is going on? Can our suppliers be so dim as not to be using this device?
   
   And what about line losses? Surely the optimum distribution of parallel power supply from a number of generators into a grid can and is being continuously monitored for minimum power loss? I imagine that this is being done and as a result the article's concern for the way that the consumer behaves and what his meter shows (or does not) is the most vital and "smart" link to concentrate on. Using thes same power lines it is easy to send every consumer a signal telling him/her when the power supply is at lower cost etc. Any consumer who wants to save can surely afford the cost of a suitable receiver (plugable to any outlet in his house) to show when the flow is at a lower cost. Is this information too much for the writer of the article and the electrical companys to explain and provide?

   Reply | Report Abuse | Link to this

7. 7. Mark Mastrocinque in reply to Power Hungry 07:18 AM 5/11/10

   If you talk to your power company, they must tell you when power is cheaper. You can then adjust your schedule to run large appliances when power is cheaper and save yourself some money. Your meter may not be smart but I'm sure you are.

   Reply | Report Abuse | Link to this

8. 8. Power Hungry 07:34 AM 5/11/10

   Once again you miss the point. The meter does not care what time of day it is or if the power is cheaper. It is a meter it records how much energy is used, end of story. The consumer is charged x amount for the power used. Off peak or on peak, there is no difference to the meter.

   Reply | Report Abuse | Link to this

9. 9. sethdayal 01:10 PM 5/11/10

   If we assumed $2000 a pop average for the entire smart meter gig, we'd need $300B to do Americas 160 Million meters
   
   Money would be much better spent on nuclear power
   
   Seems with the many new nuclear builds near completion in Asia, some of them the American designed AP-1000 reactor, real nuclear construction costs are headed shortly to $1B/Gw leaving plenty of cash to convert the US off fossil fuel. To do it, we'd need 2500 gigawatts of nukes at $2500B financed by ending our payments to our friends at Big Oil/Coal for $600B in crude, $100B in natural gas and $75B in coal. Could be all done ten years from now.
   
   If we convert all our natural gas power plant and heating applications first we will free that $100B in current gas usage, exactly enough to replace all our gas and diesel use with CNG auto fuel, and NG derived methanol and dimethyl ether. The last two would replace ethanol in current E85 flex fuel vehicles and diesel in trucks and locomotives.
   
   As we switch to electric vehicles, and enough nuclear capacity is available off peak nuclear produced methane or methanol would fuel the remaining non electric vehicles.
   
   Mass produced conversion kits and home fill stations would quickly become available.
   
   The $300B required for smart meters would instead buy 300 new mass produced nuclear plants which could be distributed around the country close to load centers virtually eliminating the need for long distance transmission grids.
   
   Offpeak nukes would generate power for liquid fuel production, desalination, heating and cooling storage systems, vehicle charging and other time insensitive apps instead sending more moolah into the greedy pockets of power companies executives for high daytime rates.

   Reply | Report Abuse | Link to this

10. 10. BobLento 04:56 PM 5/11/10

    Its a little known secret that, in the telecommunications industry, free nights and weekends is a successful strategy to move leisurely, non-business calls off peak hours and thus reduce infrastructure costs for the carrier.
    
    Now with the smart grid, utilities can use similar strategies to financially motivate consumers to avoid peak-use and save a lot of money as well as the environment. Theres a lot energy companies can learn from the telecommunications industry, which has had a real-time information network for years and operates a similar infrastructure-based business.
    
    -Bob Lento, Convergys

    Reply | Report Abuse | Link to this

11. 11. BobLento 11:55 AM 5/12/10

    The comparison of telecommunications to the electrical grid is particularly compelling. Both are infrastructure-based businesses where ‘the network’ is a substantial cost burden that is multiplied by peak use. Our company has been providing billing and customer care technology and services for telecom for over 20 years. Telcom companies have had a 2-way communications network all along, and the concepts telecom have learned over the years are very applicable to utilities.
    
    For example, people with poor credit and no deposit could now pre-pay for electricity. Of course one of the best examples is “free nights and weekends” which was designed to reduce peak wireless network use lowering network costs for the provider and creating a valuable cost savings opportunity for the consumer. The Smart Grid creates a similar opportunity for the utility industry.
    
    -Bob Lento, Convergys
    

    Reply | Report Abuse | Link to this

12. 12. Wayne Williamson 05:57 PM 5/13/10

    the most i can positively say about the article is wiring the major parts of the infrastructure makes sense...knowing when such and such is down and rerouting power...
    
    showing the user power consumption by hour of day and the cost incurred might help...
    
    i used teco's(tampa electrics) version of the household limiter for years....it allowed them to shut off the water heater and the air conditioner when demand was high...for an 18 dollar discount on your bill every month...after about 8 years on it, they dropped the discount to i believe 10 dollars and started to shut off my air conditioner on the hottest days for hours at a time....after 1 or 2 years of that i canceled it....
    
    

    Reply | Report Abuse | Link to this

13. 13. ultimobo 06:37 PM 6/4/12

    good point Bob - providing for peak capacity is expensive and wasteful if only needed for a minority of the time, so shifting peak demand to off-peak times is a sensible strategy to avoid the need to build more peak capacity - reminded me of a TV program about London electricity central controller, they know that peak demand occurs at the end of the most popular TV program, when millions get up and turn on the electric kettle - so the controller gets on the phone to France and buys peak electricity from France (nuclear plants? cough cough ...)
    
    here in Oz a lot of supposed investment for renewable/green energy has instead been going into 'goldplating' the poles and wires infrastructure rather than seeking out new generation capacity, so I guess those industry folk know which side of their bread is buttered ... they're rubbing their hands for smart meters to bring them rivers of gold - oh gee, sorry it made a mistake in your billing rate !(not really - hee hee!)

    Reply | Report Abuse | Link to this