What if electricity were like the Internet?

Power Lines
Photo Credit Bob Fornal

So, Nick Carr writes regularly about cloud computing and how the Internet is heading more and more towards the same model as the utilities. And he’s right. And this is a good thing.

Now let’s turn that on its head.

When will the utilities start to become more like the Internet? Specifically, when will disparate, disconnected electrical grids join up to give us one global electricity super-grid?

Can you imagine the resilience of a massively connected super-grid? One which can route around problems.

Then think about how much more stable the super-grid would be if the excess energy produced by, for instance, Scandanavian wind farms on windy nights could simply be sold to meet capacity shortages in the US as people arrive home from work, or in Japan as they start to wake up.

What if the grid were smart, publishing prices in real-time, based on supply and demand fluctuations?

And further, what if smart meters in homes and businesses could adjust appliances based on the real-time pricing (thermostats up/down, devices on/off, etc.)?

And what if, again like the Internet, the super-grid were read/write i.e. if you could be a producer as well as a consumer? Think plug-in hybrid vehicles, for example. In times of more abundant electrical supply when energy is cheap or negatively priced (sun shining on Spanish PhotoVoltaic arrays and/or wind blowing on Northern European wind farms), plug-in hybrids could suck in electricity and act as a distributed electrical storage mechanism. Several hours later, if the wind dies down, or the sun sets electricity prices jump and the smart meters realise it is now financially advantageous for plug-in hybrids to sell electricity back to the grid. So they do.

What if most of the technologies to make this happen already existed? How long will it be before the utilities embrace the Internet model in the same way the Internet is jumping on the utility model?

UPDATE: Simon Wardley writes that:

According to wikipedia, “the concept of an interconnected global grid linked to renewable resources was first suggested by Buckminster Fuller” in the 1970s.

There are even organisations such as the Global Energy Network Institute (GENI) who apparently have been working on the “viability of the interconnection of electric power networks between nations”.

This subject deserves a higher profile.

It seems I am in august company!


  1. Thomas says

    What if that makes us and all the networks interdependent and therefore less resilient?

    I think it would make load/frequency balancing incredibly complex. These days, we don’t even have common plugs and power levels on the consumer end, let alone at the transmission and distribution side…

  2. Brian Miles says

    From what I know about transmission, and the inefficiencies thereof, routing electricity across the planet might not make a lot of sense.

    Then there’s also the issue of would grid operators actually go for this. In the US at least, with locational marginal pricing, congestion on the grid seems to allow wholesalers to make great profits–robustness be damned (cf. the August 2003 black out in the Northeast).

    Public citizen has a report titled, “The Failure of Electricity Deregulation”, from May 2007, from which I’m drawing most of this (

  3. says

    If you want to use an internet metaphor, I think we should rather look at P2P-power: Local, distributed power generation. Why transport energy if you can generate it in your backyard, on your own roof, and share (or take from) the surplus across town?

  4. says

    @Thomas – I wouldn’t get hung up on technological difficulties, these can be engineered around. Today my mobile phone has 8gb of storage. 10 years ago only high end servers had that kind of capacity.

    @Brian – Transmission losses are now less of a factor than they were previously and new materials have the potential to reduce these even further.

    With the increasing price of oil, the economic imperatives are in place to make this happen

    @Meryn – absolutely, this falls in perfectly with the thesis I was putting forward. The plug-in hybrid cars are only one of the read/write options, microgeneration is obviously another.

  5. says

    Tom — the huge difference I see is the size and establishment of the big players in the game. They make plenty of money as is, and don’t have a big incentive to change. Early computer networks weren’t nearly this established, and the benefits of interconnection were clear, and costs relatively low. Also, regulation and public policy play a dominant role in power transmission; this could also be the savior in that public policy can force changes that the incumbents wouldn’t choose. I think we’ll see better work within nations than between them, for this reason. The US is big enough to implement something like the widely-distributed-renewables picture within its own borders, especially if we pair that with a solid, reliable way to feed energy in from massively distributed sources.

  6. says

    I completely agree Tom. I think part of the reason that the incumbents don’t see this as an option is the traditional view that energy supply and distribution is a massive requirement that requires massive solutions – economies of scale etc. As with early computers, the focus was on delivering bigger and bigger computers before the mini-computer and then the micro-computer led to a whole new model of networked computers and then the internet.

    Just as the energy companies are now seeing the benefits of selling efficiency to their customers, where previously the idea of selling less was definitely less, so I think they will come around to the idea of distributed networks being good for everyone and giving them new markets and new possibilities.

    Maybe it will follow the model of the software companies who are effectively out-sourcing much of their R&D by providing access to their software via APIs for amateur programmers to play with. The electricity supply/distribution companies can open their networks so that customers can provide storage, peak shaving and generation and the companies make money by managing the system…

  7. says

    Most of what you express above exist already, but this doesn’t mean that your points are not valid: innovation is often the sum of several precursors.

    Most of the complexities lie in the commercial aspects, here’s what I personally understand on the software aspects.

    1. Read/Write -some countries mandate a tariff to buy back electricity from small producers. It can be perverse, like in France it is degressive meaning it encourages people to build lots of small wind farms instead of more efficient larger ones.
    ETRM (Energy Risk and Trading Management) is the software that is needed to manage those complex transactions.

    2. Reliability. A very personal comment here: much like the trains privatisation, the benefits of the deregulation of utilities are not really perceived by consumers, at least not as much as say telephone. While the cost of telcos has dropped enormously (remember dialling from payphones with calling cards and waiting for the call back?), I can’t say I see similar benefits for rail users and electricity or gas: it’s often more expensive and crucially capital investments in the network are often not sustained. The UK rail system is broken all over and is in a state where fixing is nearly impossible without major disruptions. The US has and will continue to experience black-outs and brown-outs.
    Germany and France on the other hand have kept regulated electricity providers and from my own personal experience, the network seems better. Plus of course, generating 80% from nuclear gives France a big advantage now -we outta thank de Gaulle’s vision… Oh, and do check Wikipedia on HVDC.

    3. Finally, you’re talking about SmartMeters. Oracle demonstrated those at CS Week in San Antonio and at PowerGen in Milan, the issue is about open standards to connect the meter to the appliances (tell the fridge to be a bit warmer for 1h) and the meter to the utilties billing system.

  8. says

    Hi Tom,

    People that work in electrical power generation and transmission generally aren’t familiar with the Internet and visa-versa. I have a degree in electrical engineering, a background in large grid sychronisation for wind turbines and I am the founder and Managinf Director of a SaaS company. I am unusual in that I have a leg in both camps.

    You are absolutely right! Electrical Transmission Systems traditionally look like an old IT model. Generators = Mainframes. Separated Grids with limited interconnect = Closed LANs. The new buzz word is the ‘Smart Grid’. This looks much more like the Internet. If you own a PV solar panel or a wind turbine then you become a read-write user of the Internet, sorry I meant Smart Grid.

    I agree with all of the points above, from other comments, that support your argument. And I agree partly with some of the points that counter your argument. Power distribution is different from data distriution in one key respect and we need to bear that in mind.

    Moore’s law will not apply to electricity distribution. Technological advances will not reduce the cost or increase the power of electrical distribution at anything like the same rate as for data processing and transmission. Moore’s law is really about processing data with lower and lower amounts of energy per MIP. You can’t process energy with lower and lower amounts of energy. Therefore implementing smarter grids will involve large amounts of capital infrastructure investment. Research TREC as an example of one such proposed project.

    In supply chain management there is a phenomenon called the ‘bullwhip effect’. This phenomenon is taught in SCM using the Beer Game. This bullwhip effect is caused by the surprising fact that data and inventory are interchangeable within the laws of diminishing returns. Similar to this effect, the application of data to electricity grids can have an amazing effect on the price reduction of electricity, increase in renewables usage and the stability of the network. The magic key to this utopia is creating large open markets for electricity where anyone can create or consume electricity when the price reaches their threshold for either. The Internet used in parallel with the grid will enable the trading in electricity by ever smaller organisations until each household has a ‘current account’. I see the grids of the future like a giant eBay for electricity.

    Keep scratching this surface, Tom. Pandora’s box will open.

  9. says

    Hi Tom,

    I’d like to see more on this, with specific reference to the energy and transport section of the Zero Carbon Britain report. See

    Also, there will be lots of specialist energy consultancies looking into these questions, e.g. see Large utility companies are still reeling from the massive shock of deregulation in the early nineties, and to change the game so fundamentally in the way you suggest would require huge multilateral cooperation and transposition into national law. The questions aren’t easy – e.g. which type of feed-in tariff to go for.

    Keep plugging away though, I’ll keep an eye on this.

  10. says

    In the early days of USENET, uucp and tape transfer were very popular. Innovations in storing and transferring electricity will power the Moore’s law equivalent in Smart Power Grids.


  1. Cloud Electricity…

    Tom Raftery of Greenmonk asks “What if electricity were like the Internet?” that is, what if electrical transmission grids distributed electricity in much the same way as the Internet replaced contiguous point-to-point communication circuits (POTS) w…