Wednesday, August 27, 2008

Building a better power grid in a hurry 

Both Glenn Reynolds and Tom Maguire are writing about our low-tech electricity grid, and the limits that it imposes on our ability to get wind and solar power from the windy and sunny places to the cities and the suburbs. The occasion for their posts is an article on the subject in today's New York Times.

The dirty secret of clean energy is that while generating it is getting easier, moving it to market is not.

The grid today, according to experts, is a system conceived 100 years ago to let utilities prop each other up, reducing blackouts and sharing power in small regions. It resembles a network of streets, avenues and country roads.

As it happens, this was also the subject of a special survey on energy published in The Economist of June 21st 2008. This passage seems particularly on point:
[E]lectricity grids are about to become bigger and smarter.

Bigger means transcontinental.... The new grids would use direct, rather than alternating, current. AC was adopted as standard over a century ago, when the electrical world was rather different. But DC is better suited to transporting power over long distances. Less power is lost, even on land. And DC cables can also be laid on the seabed (the presence of all that water would dissipate an AC current very quickly). In the right geographical circumstances that eliminates both the difficulty of obtaining wayleaves to cross private land and the not-in-my-backyard objections that power lines are ugly. Indeed, there is already a plan to use underwater cables to ship wind power from Maine to Boston in this way.

As it happens, Europe already has the embryo of a DC grid. It links Scandinavia, northern Germany and the Netherlands, and there is talk of extending it across the North Sea to the British Isles, another notoriously windy part of Europe. By connecting distant points, this grid not only delivers power to market, it also allows the system some slack. It matters less that the wind does not blow all the time because it blows at different times in different places. The grid also permits surplus power to be used to pump water uphill in Norwegian hydroelectric plants (a system known as pumped storage), ready for use when demand spikes.

Smarter grids, however, would help to smooth out such spikes in the first place. The ability to accommodate inherently intermittent sources such as wind is only one of several reasons for wanting to do this, but it is an important one.

A smart grid will constantly monitor its load and (this is the smart bit) take particular consumers offline, with their prior agreement and in exchange for a lower price, if that load surges beyond a preset level. For this purpose, a consumer may not necessarily be the same as a customer. The grid’s software would be able to identify particular circuits, or even particular appliances, in a home, office or factory. Their owners would decide in what circumstances they should shut down or boost up, and the smart grid’s software would then do the job. Water heaters and air-conditioners might stock up on heat or cold in anticipation of such shutdowns. Fridges would know how long they could manage without power before they had to switch on again.

Reducing spikes in demand that way will cut the need for what are known in the industry as “peakers”—small power plants such as pumped-storage systems that exist solely to deal with such spikes. Parts of America’s existing dumb and fragmentary electricity grid are so vulnerable to load variations that their owners think they may be able to cope with no more than about 2% of intermittent wind power. Clearly peaks will never be eliminated entirely. However, Mr Abate reckons that a combination of smart grids and gas-fired peakers should push the potential for wind power up a long way.

As it also happens, just two days ago I had a conversation that touched on this topic with Carter F. Bales, a retired McKinsey & Co. partner who led that firm's climate change initiative and who authored an optimistic article on "Containing Climate Change" in the forthcoming issue of Foreign Affairs.


From the discussion in news media, it does seem that one need not believe in anthropogenic global warming to support a reconstruction of our national power grid. Not only is it essential to the exploitation of energy that will reduce our dependence on foreign oil (a salutory goal in and of itself), but it would save an enormous amount of energy regardless of its source (making the economy much more efficient). The question is, how do you rebuild the grid in a hurry? Well, the same way you build a lot of anything in a hurry -- you grant a very attractive tax incentive to people with capital. My proposal is simply this: Authorize accelerated depreciation of qualifying grid infrastructure for federal income tax purposes and full deductibility of the resulting passive losses against ordinary income. Just as these twin devices, which together constitute a "tax shelter," resulted in a massive boom in construction spending back in the 1970s and 1980s, they would unleash a gusher of capital into "smart" grid redevelopment. Yes, they would amount to "tax cuts for the rich" which at least one presidential candidate purports to oppose, but there is no question that they would get the grid rebuilt far faster than any centrally managed government program or process that depended on the balance sheets of public utilities.


By Anonymous Anonymous, at Wed Aug 27, 10:28:00 PM:

How are they now able to send DC power over long distances? I thought AC was originally adopted because DC couldn't be transmitted more than a couple of miles. Then there will be the NIMBYs. They should be able to block the construction of these lines for years.  

By Anonymous Anonymous, at Wed Aug 27, 10:48:00 PM:

In the US today, you can already sign up with the poco for a voluntary shutdown during peaks in exchange for reduced rates. FPL has been doing this for years.  

By Anonymous Anonymous, at Wed Aug 27, 11:31:00 PM:

feeblemind, AC won out over DC because the voltage was easily transformed. Power is equal to the voltage multiplied by the current. High current causes high heat. High heat means power loss. Therefore, to transmit large amounts of power without significant power losses, you need to efficiently change the voltage.

DC systems need to rectify the power. This means that they are converted to AC and then the voltage is transformed. There are rectifying losses, so the question becomes which is the greatest loss, rectifying or AC power transmission.  

By Blogger Papa Ray, at Thu Aug 28, 12:35:00 AM:

"but there is no question that they would get the grid rebuilt far faster than any centrally managed government program or process that depended on the balance sheets of public utilities."

One only needs to look at the mess that the government has made of the southern border fence to know that we don't want the government to have anything to do with this problem/project.

Papa Ray  

By Anonymous Anonymous, at Thu Aug 28, 01:21:00 AM:

This comment has nothing to do with this post but a wishlist related to this blog.

1. The title of each post says only "TigerHawk". When I bookmark 10 articles, I see 10x TigerHawk in my bookmarks and nothing else. Is it possible to put the actual title of the post in the title section of the html document?

2. Print format. I have a tablet computer (Nokia N800), and it is practically impossible to read this blog on it in the normal layout. I travel close to 90 minutes each day (one direction), so I would be happy, if there were an alternative simple format sans images, sidebars, etc. Something like that I can find at Richard Fernandez' blog. Then I can also read TH on my way home.

Thanks for your time and for the great blog,


By Anonymous Anonymous, at Thu Aug 28, 08:50:00 AM:

Feeblemind is correct, AC was chosen for its' superior ability (low transmission loses) to transmit large power over long distances. DC loses become unacceptable anywhere from 2 to 10 miles depending on voltage (higher voltages = less loss). These is no limiting factor to connecting wind/solar/tidal power to the present grid system, only sufficient ampacity of the cables and suitable coordination with neighboring power generation is required.  

By Anonymous Anonymous, at Thu Aug 28, 09:07:00 AM:

The vapid blathering of a certain presidential candidate for hope and change has found a home with the alternative energy nitwits. Given the gullibility on energy issues of americans, they will mandate politicians steal all their money for hope, change, and endless useless pork spending.  

By Blogger buck smith, at Thu Aug 28, 09:12:00 AM:

DC is already used to transmit Power over long distance, for example down the west coast:


The way to improve electric distribution in general is to open a new fontier underground. Auction off rights to develop a competing distribution system under ground.  

By Anonymous Anonymous, at Thu Aug 28, 09:12:00 AM:

California has passed a law that says they have to go to 30% "renewables" by, I think, 2020. Jim Detmers, who runs the California ISO (aka, "the grid") has said in public that he thinks the grid can do 20%, but that they simply don't have the technology to make it to 30%.

It's really hard to make the grid "smart".

Oh, and wind power is a complete scam.

Longer essay with supporting links here:


By Blogger Ray, at Thu Aug 28, 10:07:00 AM:

Edison beats Westinghouse! 100 years later!  

By Blogger M. Simon, at Thu Aug 28, 10:30:00 AM:

Semiconductors make the AC/DC/AC conversion possible.

It has very little to do with rectification losses. The big losses are in the DC to AC conversion which couldn't be done efficiently at high voltage until the advent of semiconductors.  

By Blogger M. Simon, at Thu Aug 28, 10:36:00 AM:

AC has the problem of circulating currents which cause losses with no power delivered. DC does not have that problem.

If we go to 2 MV DC continent spanning transmission is possible.

As to "smart grid" - that is already done in aircraft. It is also done on the current grid with dispatchable loads like electric steel furnaces.

The problem is getting it down to smaller loads. It will require a communication infrastructure as well as load control.  

By Anonymous Anonymous, at Thu Aug 28, 10:54:00 AM:

Reading the various discussions on this unfortunately reinforces my view that we're more likely to see big power users migrating to where the energy is than successfully reengineering the grid, at least any time soon. Our political, budgetary and legal gridlock are formidable obstacles.  

By Blogger Derf, at Thu Aug 28, 10:54:00 AM:

M. Simon Thank you for the only comment that made technical sense to me. My EE degree is four decades old and this discussion is boggling my mind. Do you have some references I can read on the creation and long distance transmission of high voltage DC?  

By Anonymous Anonymous, at Thu Aug 28, 11:13:00 AM:

Long-distance DC lines aren't going to go underground. We don't have cables capable of that kind of voltage. If we did, they'd be unbelievably expensive and almost impossible to maintain.

Unless we try to place the windmills in the Great Plains states (i.e. the Pickens plan), there's no need for a national grid.

I think a lot of people misunderstand why the grid is "brittle" It's not that there's anything deficient in the transmission lines themselves, it's that there isn't enough generating capacity to handle an emergency shutdown of a major generator. When that happens, the whole thing comes down like a house of cards. More and more diversified generation capacity actually helps stabilize the grid. We're a long ways from physically stressing the transmission lines. If you run more current through them, they just lose more power.

Certain corridors may need to be beefed up, but a "national grid" would accomplish nothing.  

By Anonymous Anonymous, at Thu Aug 28, 11:14:00 AM:

I live in Nebraska, where T. Boone and his cohorts want to build thousands of windmills and string miles of new transmission lines to take the power to the coasts. My problem, why should I have my currently low electric rates (and taxes) raised, likely double, so that the coasts can take the power? T. Boone wants the taxpayers to build the new transmission system at a cost of over $60 billion. Build nuclear, clean coal and natural gas generation sources close to the users, and upgrade their local lines, at their expense - I don't want to pay for their problems.  

By Anonymous Anonymous, at Thu Aug 28, 12:02:00 PM:

If you really want to improve the power grid in a hurry, forget about tax credits and so on. The big obstacle to transmission lines comes from Nimbyism in the form of environmental reviews, endangered species act, local and state siting land use reviews, and on and on. Cut through all that morass and power lines will be built quickly. Not willing to give up your right to oppose projects you don't like for whatever reason? Then you will get more of the same.

By Blogger Porkov, at Thu Aug 28, 12:35:00 PM:

With the progress being made at Northeastern on nanomanufacturing (especially as it applies to graphine) and the work at Cornell on high-temperature superconductivity and the promising work being done on Polywell fusion, maybe we ought not rush into building a grid that will be obsolete before it is halfway finished.  

By Blogger Neil, at Thu Aug 28, 12:58:00 PM:

Although there are good reasons behind the historical choice of AC transmission, DC transmission lines can deliver more power than AC transmission lines. The problem, as M.Simon points out, is reactive losses.

For equal amounts of power, transmission losses decrease linearly (roughly, ignoring reactance) as voltage increases.

AC transmission was originally chosen because it is very simple to use transformers to step-up voltage on the production end of the power line, and step-down again at the user end of the power line. That way you can transmit power over relatively long distances at high voltage, without having to pipe 2000 Volts (for example) into your house.

The problem is that when the current in a transmission line changes (say for example, as a 60Hz sine wave), it has to buck its own magnetic field, causing "reactive" losses. DC transmission, where voltage and current stay relatively constant, does not have that problem. The only losses are the resistance of the copper.

With modern semiconductors, it is feasible (though still not cheap) to step-up and step-down DC, and convert it from/to AC, so DC transmission is feasible and has lower losses if the AC transmission losses outweigh the DC heat loss in the semiconductors.


By Anonymous Anonymous, at Thu Aug 28, 01:20:00 PM:

DC loses become unacceptable anywhere from 2 to 10 miles depending on voltage (higher voltages = less loss).

Modern DC transmission really cranks up the voltage (450,000 volts on the Quebec-New England Transmission), which transmits just fine over long distances (800 miles for Quebec-New England Transmission).

And we can do it piecemeal. The reason Quebec-New England was implemented as HVDC to start with was because the Quebec and New England grids are unsynchronized. So replacing various current interconnects in the grid with HVDC would allow smaller AC grids that don't have to be synchronized with each other, but which can draw power from each other. That improves reliability right there.  

By Blogger Ray - SoCal, at Thu Aug 28, 01:34:00 PM:

DC is already being used in underwater cables per the Economist article and is planned for a DC grid linking Norway and other countries, so I don't see a problem using them below ground.

DC at higher voltages has the following advantages over AC Voltage Transmission

1. DC Power can be transmitted at twice the voltage of AC. This means lower losses

2. DC Power uses the entire cable, where AC Power at high voltages only uses the outside due to the skin affect.

3. DC Power lines would give the ability to have a large scale grid, allowing for wind power to be used more effectively. Since different areas have wind at different times, a big grid would allow sharing to even out the supply. Same idea as having a portfolio of companies or industries to reduce the risks in stocks.

4. DC Power can be easily transmitted on the ground, where AC due to ground can't be. This also allows for underwater DC Power cables.

5. Using modern power electronics (IGBT's is my guess), Voltage conversion for DC can be done as economically as AC is done with transformers.

6. No frequencies to sync.

7. No reactance in the lines - so no coupling issues.  

By Blogger Brett, at Thu Aug 28, 02:28:00 PM:

"2. DC Power uses the entire cable, where AC Power at high voltages only uses the outside due to the skin affect."

Admittedly, I studied electrical engineering back in the 70's, but I'm sure this hasn't changed since: The "skin effect" is a function of high frequencies, not high voltages. In copper, the skin depth for 60 cycles is a bit over 8mm, regardless of voltage. This is more of a disadvange for buried lines than aerial, since the aerial lines have steel cores for mechanical purposes anyway, while buried lines don't need to sustain such high mechanical stresses, and can be conductor all the way through.

To add another advantage of DC, I would imagine leveling short spikes with capacitors would be simpler.  

By Anonymous Anonymous, at Thu Aug 28, 03:38:00 PM:

Hey Nebraska, do we once again have to remind you that your state is on wlefare paid by the coast states? Since 1983 Nebraska has taken more federal money than it has put in, $0.10 extra for every dollar in 2005 to be exact, whereas we Californians have PAID more in taxes than we receive since 1983 to the tune of receiving only $0.78 for every dollar in 2005.

I'm from the Midwest so I know your issues, but if you don't want to help the coast states with our power issues, please, please, pay your own way and pay us back (with interest please) the money we've put into Nebraska for the past 23 years. Thanks.  

Post a Comment

This page is powered by Blogger. Isn't yours?