« Previous entry | Home | Next entry »

Katrina's lesson: We need an "Intelligrid"


IntelliGridlogo.jpgKatrina showed that much of our infrastructure can be bottlenecked, from our electrical grids, to our telecom networks. Some said media coverage seemed to work best, but that was decentralized -- as exemplified by the vibrant work of bloggers.

On one of the failings, for example the electric grid, there's a worrying lack of urgency. Again, it is fertile Silicon Valley that has ideas about how rejig our grid. The problem is building the political case, to get approval in Washington, and the states.

Our story, published today in the Mercury News (free registration), is about how venture capitalists and technologists in Silicon Valley agree with the proposal, by Palo Alto's Electric Power Research Institute, for something called the Intelligrid. It makes sense. Instead of hundreds of power plants, such as giant 1,200 megawatt nuclear power plants, powering our needs, we should have millions of solar powered...

...residences and workplaces -- or at least a power source closer to their destination -- so that a good fraction of the power doesn't get wasted in transportation.

And that way, a single error at one plant, or cut in an electrical transmission line, doesn't shut down the power of 2 million people, as it did last month yet again in Southern California.

Utilities would have to buy into idea, which arguably is against their interests -- they'd lose control. So our politicians need to be pressured. And they're not feeling that pressure. We're interested in feedback on the Intelligrid idea, from other cleantech experts like Rob Day. Is it really just politically dead in the water? (Update: See Rob's great response here.)

Locally, it is significant because the California Independent System Operator, which oversees most of the state's electricity system, has just approved a $300 million transmission line that brings power into S.F. from Pittsburg under the S.F. Bay. That's a whopping-big line. Sounds like the opposite of the Intelligrid, but maybe we're missing something. Apparently, it still needs environmental approval.

Meanwhile, check out the idea of Felix Kramer, co-founder of the Palo Alto company, CalCars -- retrofitting the Toyota Prius hybrid to form an efficient decentralized electric generator grid. The raft of comments suggest he's on to something.

Felix informed us this morning, after reading our story, that EPRI has actually been pushing this idea of the Intelligrid for decades, but simply came up with the name 4 years ago. It is the brainchild of EPRI's founder, Chauncey Starr now 93, and still going in to work.

Related to all this, there's an Red Herring interview with Nancy Floyd, co-founder of the San Francisco venture firm, Nth Power, and note her answer to the question about where the next big opportunity in cleantech:

The big infrastructure problem is the aging grid, and the whole automation area. The average age of transformers is 38 years, and their design life is 40. Almost every week transformers explode, causing outages, costing money, even killing people. And how do they find out if a transformer is going to fail? They send someone out to take a sample of oil from the transformer and send it to a lab to get results a week later. It䴜s just one poignant example of how antiquated the system is. I think about my boys instant messaging for fun, and we still send meter readers out. It䴜s something that has to change

Meanwhile, the NYT fell on another "decentralization" story, about telecom. It notes the drawbacks of the incumbent phone network, and advantages of the decentralized nature of WiFi. If cities had lots of little WiFi nodes, their communications might stay up during a Katrina-like catastrophe. Note, too, how these networks could best be powered by distributed solar or other energy sources.

There'd be work for Silicon Valley companies, too. The NYT piece says the cost of the WiFi network is laughable. It refers to the Silicon Valley (Sunnyvale) company, Tropos:

Alternatively, a city could simply hire a mesh-networking company like Tropos Networks, which estimates a cost of $70,000 to cover a square mile with DSL-speed connections. These numbers are so low that they are virtually rounding errors in any city's budget.

TrackBack URL for this entry:

Links to blogs that reference this entry:

From: SiliconBeat
Smart grid: A $45 billion opportunity?
Excerpt: Remember our rants about the need for a so-called Intelligrid, and about how local investors were starting to talk about this too, i.e., a more efficient way to distribute power, which would also help the environment? Well, the Global Environment Fund ...
Tracked: October 26, 2005 8:38 AM
From: SiliconBeat
Smart grid: A $45 billion opportunity?
Excerpt: Remember our rants about the need for a so-called Intelligrid, and about how local investors were starting to talk about this too, i.e., a more efficient way to distribute power, which would also help the environment? Well, the Global Environment Fund ...
Tracked: October 26, 2005 8:46 AM
From: Generator Portable
Generator Portable
Excerpt: Next , a subsidiary of durable heavy-duty vinyl Double stitched seams for “generator” genera...
Tracked: February 23, 2006 8:14 AM
From: Questions
Excerpt: They are arranged by topic and are updatedThis page uses frames, but your browser doesnt support them. Vessel Hull De...
Tracked: July 31, 2006 9:06 AM


The Intelligrid is something that we ought to get around to doing sooner rather than later. It's a critical part of a 21st century infrastructure. For what CalCars.org is promoting -- plug-in hybrids -- the current grid has off-peak capacity to charge tens of millions of cars. But a modernized grid would pave the way for the now futuristic "vehicle-to-grid" (described at the bottom of http://www.calcars.org/vehicles.html -- millions of cars as active nodes on a two-way national power network), with many additional benefits.

In the original article, I was surprised to see Ray Rothrock's statement that "wheeling losses" could reach 50% -- I'm not an engineer, but I've seen numbers more around 10%, for example: http://www.memagazine.org/backissues/june01/features/reality/reality.html .

Finally, I wanted to clarify that any hybrid, not just a Prius, could provide portable backup power, mobile for emergencies, in Silicon Valley driveways for local power outages. Comments especially welcome at my blog (Matt gave URL in his posting).
-- Felix Kramer, founder, California Cars Initiative

Felix Kramer on September 19, 2005 8:08 PM
Comment link


Regarding your critique of Ray's point about "50 percent loss," check out page 76 of September's Scientific American (link below). It is a graph in an article by Amory Lovin, which suggests the loss could be even greater.


(Thanks to Scott Love for the tip).

Matt Marshall on September 20, 2005 6:32 AM
Comment link

Hyper growth rates without marketing expense is always the critical gage.

A couple hours ago, I went out out for an espresso. My girlfriend picked a spot on Sunset and without exaggeration, I spotted maybe 25-30 Prius's in a 40 minute round trip car trip from Marina del Rey.

Energy activism is taking on cool appeal(without any "sans" marketing) in addition to economical benefits(accented by the post-Katrina psyche): that's key to adoption.

We have transitioned past early adoption. Hybrid vehicles are now mainstream: growth rates will be astounding going forward. Truly blowout.

Fuel cell remains in the garage while Hybrid has arrived.

Marina Architect on September 20, 2005 2:28 PM
Comment link

With regard to the NYT noting that the cost is "laughably small", a Tropos deployment is nothing but. As with many people, I think the author confuses the capabilities of a mesh ideal with the realities of existing mesh implementations.

Ideal mesh allows for seamless automatic routing among peers over 4-5 hops. Existing mesh solutions only offer mesh requiring fixed wire access every 2-3 nodes, limiting hops to 2 in order to achieve an acceptable level of service. The limit of number of hops also means that, at best, peers can speak with one another, but there is still no connectivity with the outside world. A benefit, but not on the level the NYT touts.

Secondly, $70,000 a square mile for equipment alone is not trivial for municipal wireless. For the Philadelphia project desiring 135 sq miles over a small portion of downtown, this comes to $9.45M. This does not include operating expenses or any deployment costs, including ripping up roads to install fixed wire access to 1/2 to 1/3 of the nodes.

Chris on September 21, 2005 1:33 PM
Comment link
Post a comment

Remember personal info?