Advanced Search

Please click here to take a brief survey

Making the Grid "Smart"
Joel Makower, 24 Oct 05

The realization that America's electricity infrastructure is shakier than a palm tree during a hurricane hits us every few years, when some blackout or rolling brownout reminds us of our electro-vulnerability.

But to truly understand what we're up against, it's important to step back for a moment to see just how vast -- and how vulnerable -- our electricity infrastructure is:

The North American electric power industry comprises more than 3,000 electric utilities, 2,000 independent power producers, and hundreds of related organizations. Together, they serve 120 million residential customers, 16 million commercial customers, and 700,000 industrial customers. [...] The continent has 700,000 miles of high-voltage transmission lines, owned by about 200 different organizations and valued at more than $160 billion. It has about 5 million miles of medium-voltage distribution lines and 22,000 substations, owned by more than 3,200 organizations and valued at $140 billion. The North American electric power industry will purchase more than $20 billion in grid infrastructure equipment in 2005, nearly one quarter of the worldwide total of $81 billion.

That analysis comes from a report released today: "The Emerging Smart Grid" (Download - PDF), produced by the Redmond, Wash.-based Center for Smart Energy. According to the report, as much as $45 billion is up for grabs by new advanced technologies for modernizing the electric power infrastructure.

The notion of a smart grid is familiar to WorldChanging readers (see here and here, for example.) The idea is to make the existing grid work more efficiently -- so much more, in fact, that it could reduce the need for additional power plants, or for costly redundant systems designed to work "just in case" of peak demand. That's the vision of a growing corps of researchers and companies working on grid optimization, a term that describes a wide range of information technologies that better understand and analyze exactly what's going on in a complex energy system on a minute-by-minute basis, then optimize the system in a way that's cost-effective.

This isn't entirely news. Wired magazine published a seminal piece on The Energy Web in 2001. My company, Clean Edge, suggested in its Clean Energy Trends 2003 report that "optimizing the grid" would soon propel both investors and innovators to grab onto a multi-billion-dollar opportunity. But the CSE report takes that view to a much deeper level. For starters, it offers the seven key characteristics of a modern, optimized grid:

  • Self-healing. A grid able to rapidly detect, analyze, respond and restore from perturbations.

  • Empower and incorporate the consumer. The ability to incorporate consumer equipment and behavior in the design and operation of the grid.

  • Tolerant of attack. A grid that mitigates and stands resilient to physical and cyber security attacks.

  • Provides power quality needed by 21st century users. A grid that provides a quality of power consistent with consumer and industry needs.

  • Accommodates a wide variety of generation options. A grid that accommodates a wide variety of local and regional generation technologies (including green power).

  • Fully enables maturing electricity markets. Allows competitive markets for those who want them.

  • Optimizes assets. A grid that uses IT and monitoring to continually optimize its capital assets while minimizing operations and maintenance costs.

    Several factors are driving the need for a "smart grid." For example, deferred maintenance that can no longer be ignored is mandating billions in upgrades. Regulatory changes mandate still more new hardware and software.

    Still another driver is the substitution of "bits" for "iron" -- using smart systems to delay or reduce the need for expensive capital assets:

    Smart technologies can reduce the need for power plants, power lines, and substations. To name just four examples:

  • Demand response programs that shave peak loads, reducing the need for expensive (and polluting) peaking power plants

  • Sensors and meters that show exactly where power is being used, so utilities can expand only where needed and when needed

  • Electronics and control software that monitor power fl ows in real time, to run existing lines much closer to capacity without compromising reliability

  • Sensors and software to remotely monitor expensive equipment to know when it really needs to be replaced
  • According to studies by PNNL, the Rand Corporation and others, the savings from measures like these could be $50-100 billion over the next 20 years.

    Skyrocketing prices for oil and natural gas are bringing a new sense of urgency to all energy issues, including the grid, and this report represents a call to action to consumer groups, trade associations, utilities, scientists, and environmental organizations, among others. As we continue to electrify everything, and increasingly feed in electricity from countless solar, wind, and other distributed installations -- and do it in a 24/7 world -- the needs for a sturdier, smarter grid will grow daily. And our failure to upgrade our electricity infrastructure could threaten our economic -- not to mention our national -- security.

    As we noted in our 2003 report, "Given that the cleanest energy plant is the one that you don't have to build, grid-optimization represents the ultimate clean-energy play."

  • Bookmark and Share


    Improving the grid is fine, but it should be acknowledged the current electric grid (or the 3 interconnected grids in North America, with equivalents elsewhere) is one of the most impressive and, as you note, vast, creations of human engineering; problems are, relative to the enormous benefit it provides, extremely rare. I'm also not sure why its vastness implies "vulnerability" - rather the "3,000 electric utilities, 2,000 independent power producers, and hundreds of related organizations" I would think provide a robustness unmatched by most other industries. An attack on any individual utility or producer has little effect on electric supply since we have so many others interconnected already.

    And it should also be acknowledged that the "smart grid" proposals have the effect of increasing the effective number of "utilities ... [and] power producers" - by orders of magnitude. So we're going from a vast system of great complexity, to one that's even more complex! One hopes we can make it smart enough to confine problems instead of cascading them...

    The point at the end, that "the cleanest energy plant is the one that you don't have to build," is also, I think, misleading. Grid losses are really quite small and it's not clear to me a "smart grid" would be able to lower them any further, so the issue isn't really inefficiency of the grid as far as energy goes. The real issue is low capacity factors and high capital cost because with the current grid we need lots of excess capacity to meet variations in demand (and supply for intermittent or seasonal sources like hydro, wind, and solar) throughout the day, and the year.

    But saving money by lowering excess capacity INCREASES vulnerability, because we're running everything closer to the edge. You can't both save money and make things less vulnerable; the smart grid may help with one or the other, but it can't do both.

    Posted by: Arthur Smith on 25 Oct 05

    A smart grid with DSM would probably carry a more consistent amount of power and would have lower losses per unit (resistance losses scaling as current squared).

    saving money by lowering excess capacity INCREASES vulnerability, because we're running everything closer to the edge.
    Preventing demand from getting too close to the system limits, and especially being able to shed demand smoothly when equipment runs too hot or fails, would greatly reduce vulnerability.  As we saw on 8/14/03, a "dumb" system is vulnerable to cascading failures.  A "smart" system need not be.

    Posted by: Engineer-Poet on 25 Oct 05

    If plug-in hybrids or all electric vehicles are to become a large scale reality, I would think the grid would need to be able to tell those cars when to charge themselves. Imagine I have my car in my garage overnight--I would pay less if I allow for "optimized" charging times dictated by the grid. I wiouldn't care as long as it charged by morning. Since electric cars would be a huge new addition to the electrical consumption, I fear it just won't happen without some grid smarts (and some significant new power sources overall, of course.)

    Posted by: Jeff Rusch on 25 Oct 05

    Vulnerability to short-term disruptions would be lower with a "smart grid", but most demand reduction measures are short term; people have to have their power back within 24 hours or sooner, or things start getting difficult. That makes operating at the edge of capacity more vulnerable to long-term disruptions: you can't just bring in extra supply from outside an area affected for months at a time while rebuilding things, because there is no excess supply elsewhere.

    The 8-14-03 event was exacerbated by some grid operations being a bit "too smart" - i.e. automation caused some of the probelms. A large number of nuclear plants shut down automatically with the grid fluctuations; some could not be restarted for more than two weeks. In the interim, power was supplied by backup capacity at coal-fired and other plants available. If we are operating closer to the edge (in order to save money on building capacity, as advertised), such backup capacity would not be available on short notice, and the effect of power plant shutdowns would be there until they were able to restart, i.e. possibly at least for weeks.

    To make things robust and invulnerable you build in safety factors, like excess power plant and transmission line capacity. If you want to save money, that's a separate goal.

    Posted by: Arthur Smith on 25 Oct 05

    A "smart grid" would be able to do things like maintaining its spinning reserve as a capability of e.g. cutting large parts of its A/C load down to half power in 200 msec, or tweaking the chargers of the region's plug-in hybrid cars as fast as the radio commands could reach them.  This fine matching of generation and load would allow all generators to be working at their optimum power, boosting efficiency of both the generators and the grid.

    Having a large number of e.g. plug-in hybrid cars bidding for energy would make the utilities' jobs far easier.  Instead of having to guess how much energy they needed to make overnight, they'd have bids in hand; once they know how many megawatt-hours they need to supply by 7 AM, they fire up the least-cost mix of plants to make it (or the least-cost set of independent generators wins the bids to supply it).  Minute-to-minute glitches in demand can be handled by adjusting the charging load, or loads like A/C systems delivering chilled water.  A reserve tank of water could be a "bank" that the utility could fill if there was a temporary power surplus and drain if there was a deficit, adjusting load without affecting the customer.  All it needs to work is the data networks to permit good DSM.

    This kind of thing works in AE systems; solar and off-grid microhydro systems often have automatic load shedding for overload conditions and dump loads to take excess power.  All we need to do is bring it to a larger scale.  There will be new problems to solve, but what can we do that won't throw up any challenges?

    Posted by: Engineer-Poet on 26 Oct 05

    I had no idea there was that much power line in america. The Olduvai Theory states the fragility or our electric grid, that a number of major problems could cause the golbal grid to just die, in 2005, you cannot do anything without electricity, if it were all to die, every last power plant on earth, we would be back in the stone age. The Olduvai Theory says permanent blackouts will begin to occur between 2008 and 2012, being completely dead by 2030. I dont believe this theory but it is unrefuted, but since 2008 is so close, we'll see what happens won't we?

    Posted by: David on 26 Oct 05

    Won't happen.  NG shortages may cause rolling blackouts and prohibitions on air conditioning and wasteful lighting in areas served by gas-fired generation, but that's not going to be a collapse.  Wind power will start hitting the market in a much bigger way as siting objections are overridden with eminent domain (after Kelo, it'll happen) and nukes will be pumping watts hard by perhaps 2015.

    Coal-fired generation will fill at least some of the gap.  If necessary, the coal trains can be switched from diesel fuel to coal slurry for the diesels; we're unlikely to run out of coal to make steam unless we do some really stupid things.

    Posted by: Engineer-Poet on 26 Oct 05



    MESSAGE (optional):

    Search Worldchanging

    Worldchanging Newsletter Get good news for a change —
    Click here to sign up!


    Website Design by Eben Design | Logo Design by Egg Hosting | Hosted by Amazon AWS | Problems with the site? Send email to tech /at/
    Architecture for Humanity - all rights reserved except where otherwise indicated.

    Find_us_on_facebook_badge.gif twitter-logo.jpg