Moving to a post-fossil energy infrastructure is no small task. Leave aside the politics of the problem for a moment, and look at the logistics: replacing coal, oil and gas-fired power plants with cleaner, renewable technologies isn't simply a matter of unplugging one and plugging in the other. Renewable sources often requires wide spaces to generate useful amounts of power, and need to be situated in areas most conducive to their generation needs (sunny regions for solar, windy for turbines, the ocean for wave, etc.). Moreover, there is great value in adding in small, local generation (often referred to as micro-generation) to the mix, from wind micro-power, micro-hydro and rooftop solar panels to more exotic technologies like Stirling Engines, plug-in hybrids, and potential future developments like photovoltaic curtains.
Such a model of diverse, widespread sources of power generation is typically called "distributed energy," and it has some definite advantages over the current, largely centralized infrastructure. Distributed power can be more robust against accident or attack on the power grid: knocking down a 5 megawatt wind turbine would be bad, but not nearly as disastrous as abruptly taking a 1,000 megawatt coal power plant off the grid. Distributed power also allows greater resource flexibility: the more varied the resources used to generate electricity, the less likely are disruptions resulting from limited availability of one of them. This latter is particularly important due to the variable nature of wind and solar. Output from a given wind or solar farm will rise and fall with local conditions, but the overall availability of electricity from multiple locations and resources can still be consistent.
But distributed energy is currently more costly than centralized power (PDF). Some of that cost comes from managing the complexity of variable power generation, changing usage patterns, and a multiplicity of sources. Distributed energy resources will have to be managed more like a computer network, complete with abundant routers and switches. The success of distributed energy is ultimately dependent upon the increasing availability of computer-enabled power networks, or "smart grids." And smart grids for distributed power, in turn, will increasingly rely upon the availability of distributed computing.
It's likely that smart grids are coming, even without an aggressive shift to renewable energy. On top of dealing with variable, dispersed inputs, smart grids allow more efficient routing of power, with fewer idle or wasted generators; smart grids would, in principle, allow an overall lower level of generation to support continued levels of use (or, more hopefully, a growing level of use of in turn more efficient buildings and devices). Smart grids are, in the end, a fundamental part of building post-oil, bright green communities.
Like computer networks, a successful smart grid will take advantage of "end to end" topology, where the real smarts of the system can be found close to the points of use, not centralized. Telecom researcher David Isenberg referred to this system in the information world as a "stupid network" in a seminal 1997 article. A March 2004 article at Mechanical Engineering magazine, by Roger Anderson and Albert Boulanger, explored the potential value of stupid network-style smart grid development:
One implication of applying machine learning techniques to grid control is that the sensory intelligence has to be pushed to the "last mile" of the grid—to the point where it meets the customer—in order to make it sufficiently smart.Before the end of the decade, cheap silicon devices will be attached to most manufactured items—even inert things like steel plates. [...] The sensors, computer, and communicator will be self-contained and operational for the life of the asset, from its creation to its destruction.
It is crucial to add this kind of real-time sensing and control if the future grid is to fully exploit synergies among various power sources, including wind and solar. We believe the key to exploiting these synergies is ubiquitous silicon associated with all critical assets in the electric grid, from generation to storage, transmission, distribution, and finally consumption.
Cheap silicon in the field will rewrite the way we manage the electric grid. Widespread wireless computing on every critical node of the Smart Grid will deliver business intelligence. It will incorporate self-healing, self-organizing, Web services, and peer-to-peer computing among networks of connected assets. Each field computer will have enough memory so that it can capture its own best practices and be its own data historian.
What's needed, then, is more computing power, available in a manner which reflects the distributed power infrastructure. That is to say, what's needed is distributed (or grid) computing. E4Engineering reports that such a connection of smart grids and grid computing is already underway in the UK:
A European grid computing project worth £4.7m could solve the approaching problem of how to co-ordinate the electricity output of a proliferation of new wind farms and solar power stations. [...]'Today renewables contribute intermittently to the power grid, but in 15 years we're aiming for 30 per cent of our power from alternative sources and it's not viable to have them leaping on and off unexpectedly. We need a way to handle the change from monitoring a few hundred power stations with private networks, to controlling 30,000 alternative energy generators.'
The project aims to develop the equipment and software needed to build a grid computing network that could autonomously process the instrument data from thousands of energy sources, and allow the power industry to optimise the ebb and flow of electricity on their national grids.
What those working on the power industry version of end-to-end networks may not realize is that these technologies are ripe for end-user exploration and exploitation -- hacking, if you will, but not the malicious kind. The more the smarts of the smart grid can be found in individual appliances and devices, not at the power company's central office, the more opportunities there will be for alternative networks to form over the grid. A few months back I suggested that BitTorrent might be a model for an open variant of distributed power; the growing consideration of distributed computing resources as a way of managing the distributed electricity grid makes this all the more possible.
As more renewable energy production is connected to the general power grid, the more we will need smart systems managing the result. While difficult, it will ultimately be for the good. The efficiencies of smart grid management coupled with the sustainability of renewable energy will be a big win for us all.
Comments (1)
http://smallisprofitable.org/
by Amory Lovins and the Rocky Mountain Institute addresses the accounting methodology issues around distributed generation, and comes to the conclusion that the actual benefits are often under-estimated by a **factor of ten** by the accounting models used by the industry. Worth a look.
Posted by Vinay | February 25, 2005 1:19 PM
Posted on February 25, 2005 13:19