May 02, 2010
US National Electric Power Grid Fight

Writing in the New York Times Matthew Wald reports that lots of interest groups are making common cause against a big US national electric power grid whose main purpose is to bring wind electric power from the plains states toward where the demand lies.

After several years of debate, a coalition has emerged around the idea of a strong national electric grid, centrally planned and broadly financed, that would promote renewable energy. The group includes giant investor-owned utilities, public power entities, influential elected officials of both parties and state energy officials, and they speak with a single voice.

And they oppose it.

I have been wondering if or when this was going to happen. Really big bucks are at stake. So interest groups have gotten together to respond to this proposal. In a nutshell: the builders and operators of big wind farms in the US plains states want big electric power transmission lines (probably HVDC) built from the plains toward the East Coast to carry the power from where the wind is to where the people and industry are located. Naturally, the wind industry doesn't want to pay for these transmission lines that will boost the demand (and therefore the price and ROI) of their electric power. Rather, they want to see the costs spread around. Privatize profits, socialize costs. It is the American Way.

Of course, wind power delivered on lines paid by others would lower the prices of electricity generated by local (local to the Midwest and East Coast) nuclear, coal, natural gas, and other power sources including local wind. The opponents also dress up their opposition in high moral principle.

But opponents of a national plan say that it would be biased toward wind on the Great Plains, possibly to the exclusion of renewable resources closer to load centers, including offshore wind.

Of course both sides are promoting positions that boost their financial bottom line. So what should the policy be for long range electric power transmission lines? I do not know. How to make the construction of the lines driven by market forces that best reflect relative costs and benefits of various competing electric power sources? Local electric power delivery companies are regulated utilities that need reliable suppliers. Transmission line construction requires regulatory approval of lots of jurisdictions. This is an inherently politicized and technically complex market.

A picture is worth a thousand words. The purple and red are where the wind is:

A wind resource map of the United States. Click on this map to go to more wind maps.

You can see that the US Southeast especially has weak winds. In my view that part of the country needs nuclear power.

In a recent thread about wind power's economic effects on The Oil Drum quite a few people (myself included) challenged the post's assertion that (government subsidized) wind power lowers electric power costs overall. Wind's reliability problems increase the need for load-following peaker electric power generators that can turn on quickly when the wind fails. Some of the issues raised overlap with this post's topic.

I think it is important not to romanticize any energy source. Wind has many advantages including less pollution and less dependence on dwindling fossil fuels supplies. But it has disadvantages too. There's no energy panacea.

Share |      Randall Parker, 2010 May 02 04:39 PM  Energy Electric Generators


Comments
Fat Man said at May 2, 2010 10:30 PM:

I think I predicted this. Not only that, but it is probably a good thing, as wind power is wildly uneconomic. Best to stall them here than before we have sunk billions we don't have into technology that cannot provide usable power.

Elvenrunelord said at May 3, 2010 4:03 AM:

There's no energy panacea.......

I question that opinion and present the wave generation systems as an example of renewable energy that is for the most part consistent 24 hours a day. I've seen projections that claim that all the worlds energy needs could be solved by use of wave generation systems.

Offshore wind power seems to be steady as well.

Hong said at May 3, 2010 4:29 AM:

Windpower zealots looking for another taxpayer handout? Who knew?

Sean Lynch said at May 3, 2010 5:25 AM:

I'm kind of surprised nobody has pointed out the fatal flaw with this whole plan: the grid will go down, taking huge swaths of the nation with it. There is precisely one way to build a reliable power grid: don't connect pieces of it together. Not as efficient, but failure in one area doesn't propagate to the whole country.

We need distributed power generation and local storage or elastic use of surplus power. Of course, distributed generation is what we'll get with this plan anyway: everyone will need to buy a generator when the entire nation has a week-long blackout.

pond said at May 3, 2010 6:28 AM:

There is no panacea = there is no magic bullet. Very true; we will need many different solutions to make up for declining coal and oil supplies.

The traditional (historically) answer for 'wind alley' advocates is just to build up the wind turbines where the wind is, and connect as you can to the outer grid. Electricity then becomes cheaper and cheaper in the plains states, and google and other huge data centers will be built there, jobs in construction will follow, other manufacturers will move there ... population and jobs will follow the cheap energy.

If that scenario eventuates, long before it is complete, the other states, losing jobs and tax revenues, will be clamoring to lay HVDC lines into plains states where the power is. No need to plan these massive lines ahead of time.

I agree with the poster who said that distributed energy production is what we need, too. That and more-intelligent grids. But connecting everything together overcomes some of the problems of intermittency inherent in solar and wind, allowing the entire grid to act like a huge battery. Calm areas pull power from areas where the wind is blowing hard right now; when night falls, solar-powered areas can pull from wind-powered areas. They did a theoretical survey in Germany last year or so, and found that just in that country, over a few weeks, if the whole country were considered, wind and solar balanced out in this way.

Among the cows in Iowa said at May 3, 2010 9:23 AM:

"In a nutshell: the builders and operators of big wind farms in the US plains states want big electric power transmission lines (probably HVDC) built from the plains toward the East Coast to carry the power from where the wind is to where the people and industry are located. Naturally, the wind industry doesn't want to pay for these transmission lines that will boost the demand (and therefore the price and ROI) of their electric power. Rather, they want to see the costs spread around."

They can't take the risk of lines across multiple states. There is enough risk in wind farms less than 5 miles across. If the government can stop them, the government has to have skin in the game. Nuclear has the same problem.

It only costs 1/10 cent per KWH to send power from Kansas to NYC. This is about rent-seekers holding their monopoly.

"opponents of a national plan say that it would be biased toward wind on the Great Plains, possibly to the exclusion of renewable resources closer to load centers, including offshore wind."

We saw what happened to Cape Wind. If they want offshore, let them start building it right away. Put up or shut up.

"I'm kind of surprised nobody has pointed out the fatal flaw with this whole plan: the grid will go down, taking huge swaths of the nation with it."

This already happened in 2003 without a supergrid. It does not create new problems.

Aron said at May 3, 2010 11:34 AM:

I think the bottom-line is to not worry too much. Things like this are only a problem when you try to cram 'one big solution' from some top-down oversimplified view of reality. Alternative energy does need to grow rather organically, and there is economic value to having the source co-located to the demand that should be considered.

Are the plains states already consuming all the wind-power they can? Shouldn't that be step 1?

Nick G said at May 3, 2010 11:42 AM:

Earlier this year Senator Corker successfully offered an amendment to an energy bill at the committee level that would require that payment for transmission improvements be made by those who benefit.

But the coalition went further. “Voluntary interconnection-wide coordination should be a complement to, and not a substitute for, local and regional processes,’’ the group said.

In other words, we don't really care about how these transmission lines are paid for, we just don't want outside competition. If the cost-allocation tactic doesn't work, we'll keep trying other things.

Wind's reliability problems increase the need for load-following peaker electric power generators that can turn on quickly when the wind fails.

Not if we can build long-distance transmission, and use Demand Side Management. Both of these are much cheaper and more effective than more capital expenditure on new local generation. Of course, they aren't nearly as profitable for local utilities.

Nick G said at May 3, 2010 11:49 AM:

Alternative energy does need to grow rather organically

That's a prescription for very slow growth - we need much faster growth to deal with climate change and coal pollution.

there is economic value to having the source co-located to the demand that should be considered.

Transmission losses are small. What other economic value is there?

Are the plains states already consuming all the wind-power they can?

Iowa's population, like other nearby states, is small. Iowa is at 14-18% already: that's the point at which marginal costs start to rise. Besides, transmission takes years to build, so clearly now is the time. The nearby, obvious market is the Chicago area, but Chicago gets 50% of it's power from wind, which competes with windpower at night, so Exelon is fighting desperately to keep windpower out.

Nick G said at May 3, 2010 11:51 AM:

oops: Chicago gets 50% of it's power from nuclear

The other half comes from coal...

Bruce Dunn said at May 3, 2010 1:26 PM:

There may be some interesting possibilities in combining wind and solar power. The southern end of the US wind belt is adjacent to the south west areas of the US which have good solar power resources. Wind energy tends to be more prevalent in the winter, while solar energy is more prevalent in the summer. If the solar energy and wind energy areas were connected and treated as an integrated system, there would be more reliability than for either system by itself.

Nick G said at May 3, 2010 1:36 PM:

Absolutely: wind is slightly stronger at night, winter, and storms, when solar won't produce. They're natural partners.

Among the cows in Iowa said at May 3, 2010 2:20 PM:

Iowa wind could let Chicago use electric heat instead of natural gas. Strong winter winds mean lots of cheap energy. Just need wires.

You are probably right about Chicago. Electricity is very expensive there. I know people who steal power in protest. I would not think of it.

Among the cows in Iowa said at May 3, 2010 2:21 PM:

BTW, wind here is 16 knots right now. That is at 10 m up at the airport. At 80 m it's a lot more.

Randall Parker said at May 3, 2010 7:47 PM:

Among the cows,

You got a good source for cost estimates for transmitting electric power, say, 1000 miles?

It strikes me that long distance transmission lines used for wind will operate at lower average capacity than long distance lines used for more dependable power sources. I'm curious to know what the utilization rates will be.

Dan said at May 3, 2010 9:18 PM:

Hi Randal,

I don't know the cost for electric power transmission, but I have it on "good" sources (e.g. folks who would develop it if they could) that the primary thing holding back HVDC / other long-distance grid connections is the regulatory environment surrounding their construction. Apparently, if siting & regulatory issues were relaxed or "better" regulated, the economic case for long-haul power line construction is pretty sound w/o subsidies. Sorry, can't name sources and it's not my personal expertise .... so take it with a suitable grain of salt.

//dan.

Among the cows in Iowa said at May 4, 2010 8:38 AM:

"You got a good source for cost estimates for transmitting electric power, say, 1000 miles?"

ABB says about $620 million to send 2000 MW over 1500 km.
http://www2.internetcad.com/pub/energy/technology_abb.pdf page 6.
Another 100 km of lines adds $25 million. Call it $325 million per GW, $325 per KW.
Repay at 8% per year and it costs $26 per KW. Use it 35% of capacity, it adds about 1 cent a KWH. Use it at 50% and it adds about .7 cents. A line taking power from several states would do better than 1 state or 1 farm.

Among the cows in Iowa said at May 4, 2010 8:43 AM:

I was quoting another poster here for cost. I see that poster was wrong. ABB is probably right.

Joe Blow said at May 4, 2010 10:09 AM:

Single grid, single point of failure, particularly if "smartgrid" technology is fielded. The vulnerability to cyberattack in particular is terrifying. Check out the problems Brazil has had over the last year, which some attribute to a series of cyberattacks.

Nick G said at May 4, 2010 10:29 AM:

Here's an estimate for transmission costs in Texas of $.26/nameplate W): http://www.cnbc.com/id/25708335.

I believe there's a similar project in CA that was about $.25/W - I'll see if I can put my finger on it.

Junk Science Skeptic said at May 4, 2010 11:36 AM:

One needs to look at history for an answer here. Virtually all forms of natural energy that were exploited for practical purposes required some form of storage. Think about the millpond, or the farmer's elevated tank filled by a windmill driven pump.

To suggest that wind can make up for the shortcomings of solar, and vice versa, is a gross oversimplification. Commercial power grid operation is simply incompatible with "raw" wind or solar power.

So how do you store or package wind or solar power to make it commercially viable?

One answer, though not very efficient or cost-competitive per megawatt, is to use the solar and wind generated electricity at its source to electrolyze hydrogen from ambient air. The hydrogen could then be piped (piping the gas probably being more efficient than transmitting the electricity, it's the electrolysis part that's not very efficient) to fuel-cell electricity generation capacity located nearest the point of consumption.

The problem with wind and solar right now is that in order to low-ball the capital cost estimates, everybody is selling only half of the hardware needed for commercial viability. Think for a minute how practical it would be for a coal-fired power plant to operate without the space needed to stockpile coal, that's the sort of wind and solar technology you're being sold today.

As for the national grid system, the theory is fine, but in practice it works too much like a two-dimensional road grid with far too many potential choke points. The thinking needs to shift more toward the concept behind the internet; distributed capacity, redundant bandwidth flow paths, and fewer choke points.

For an example, let's say I have a high-rise office building. My electric utility, Zapco, calls me and tells me that due to a maintenance shutdown of one of its generating stations, it will need to charge me more for the peak-hours power I need to run my air conditioning this summer. With a web-like energy distribution network, and access to redundant sources, I can weigh Zapco's rate surcharge against what it would cost me to power my air conditioning with natural gas from T. Boone Pickens, or hydrogen from HydroIowa, or clean diesel generation from Caterpillar.

mishu said at May 4, 2010 12:18 PM:

Demand Side Management = having the power company decide how much power you can consume. That's not going to fly with the public. Try again.

Nick G said at May 4, 2010 12:36 PM:

One needs to look at history for an answer here.

Technology has changed a lot since then.

To suggest that wind can make up for the shortcomings of solar, and vice versa, is a gross oversimplification.

Yes, and no one is saying anything that simplistic. What they are saying is that the two are synergistic, and together have less output variance than alone.

Commercial power grid operation is simply incompatible with "raw" wind or solar power.

Now that's a simplistic statement, and not true. Wind power variance is very similar to demand variance, so a statement like that is like saying that utilities can't deal with demand variance.

Now, are the costs to mitigating wind's variance? Of course, but they're much smaller than you're thinking. Geographic dispersion reduces variance, and DSM is much cheaper than storage.

As for the national grid system, the theory is fine, but in practice it works too much like a two-dimensional road grid with far too many potential choke points. The thinking needs to shift more toward the concept behind the internet; distributed capacity, redundant bandwidth flow paths, and fewer choke points.

To a point, I agree. Keep in mind that balancing between parts of the grid only requires transmitting balancing amounts, not the whole load.

With a web-like energy distribution network, and access to redundant sources, I can weigh Zapco's rate surcharge against what it would cost me to power my air conditioning with natural gas from T. Boone Pickens, or hydrogen from HydroIowa, or clean diesel generation from Caterpillar.

I agree. That's compatible with expanding wind power.

Nick G said at May 4, 2010 12:38 PM:

Demand Side Management = having the power company decide how much power you can consume.

Sometimes: utilities often pay companies and households for the option of curtailing supply. It's cheap, it works, and everyone's happy.

OTOH, some schemes simply vary the price, and people can decide how or whether they want to respond. Seems to work well for cell phone service.

JIMV said at May 4, 2010 12:44 PM:

The problem with a national power grid is that it is run by politicians and if it fails, it fails everywhere...No thanks.

kyle said at May 4, 2010 12:44 PM:

"How to make a million dollars: First, get a million dollars." - Steve Martin

Makes you almost miss the days of the railroad tycoons. If it isn't worth doing it yourself, it probably isn't worth doing. Its summer time, and interstate 80 is ripped up from the trucking same as it always is. The electricity from the nuclear plant a couple dozen miles down the road is cheaper than green electricity, saving me enough money to keep buying my Fiji Artesian Water. Its responsible behavior to pay the luxury markup, as our local water supply in lake michigan could be put to much better use if they put in a pipeline to phoenix and california. Its for the children. If you don't have bills to pay, your desire to work diminishes. And idle minds are the devil's playground.

real reason said at May 4, 2010 12:45 PM:

Mishu, Replace power company with Obama administration.
Demand Side Management = having Obama decide how much power you can consume. That's not going to fly with the public. Try again. I think that is the real reason behind the "National Grid", not to promote wind power but to create artificial shortages so as to reduce potential CO2 emissions. Because only the Federal Govm't has the knowledge and authority to distribute load among generation facilities not the producers and consumers of the power.

MEC2 said at May 4, 2010 12:59 PM:

Yeah, we should link them all together, kind of like Christmas lights, in one large, tangled, electric Gordion Knot, that way when one goes out, they ALL go out.

Nick G said at May 4, 2010 1:10 PM:

JIMV,

The problem with a national power grid

A "national" grid isn't needed - just connections between grids.

run by politicians

Not anymore than now.

real reason

I don't see how any of that has anything to do with what I've said, or what the Original Post said, or what any reputable source of information has said.

Jim said at May 4, 2010 1:32 PM:

I question that opinion and present the wave generation systems as an example of renewable energy that is for the most part consistent 24 hours a day. I've seen projections that claim that all the worlds energy needs could be solved by use of wave generation systems.

Offshore wind power seems to be steady as well.

Britain thought so also. They built wind farms in the North Sea, arguably the most dependable wind in the world. Wind Farms produced no Electricity...

Water must be a better answer, since it's denser and more reliable, but the greens don't like the idea of dams.

LarryD said at May 4, 2010 1:53 PM:

"... Wind energy tends to be more prevalent in the winter, while solar energy is more prevalent in the summer. If the solar energy and wind energy areas were connected and treated as an integrated system, there would be more reliability than for either system by itself."

OK, lets try and cost out such a solution. Step one, pick how much power we commit to supply. This is just an intellectual exercise, so this is an arbitrary number, say 100 GWh.

I'll use the figures from EIA's 2016 Levelized Cost of New Generation Resources from the Annual Energy Outlook 2010. If you don't like their figures, provide another source.

On shore wind is 149.3 /MWh so that's 1,493,000
Solar thermal is 256.6 /MWh so that's 2,566,000
Total is 4,059,000

On the other hand, nuclear is 119.0 so that's 1,190,000

This is assuming that wind and solar actually do cover for each other, and no additional power needs to be provided. I still need convincing on that, sometimes it's both cloudy and still.

Nick G said at May 4, 2010 2:22 PM:

Jim,

You're link didn't come through. I should note that whenever you refer to ideas in terms of a group, like "the greens", you advertise that you've stopped thinking for yourself, and are just following the leadership of some other group.

LarryD,

The EIA 2016 projection makes very odd assumptions:

"• Proportions of total wind resources in each category vary by EMM region. For all thirteen EMM regions combined, 1.3 percent of windy land is available with no cost increase, 5.4 percent is available with a 20 percent cost increase, 11.2 percent is available with a 50 percent cost increase, 27.3 percent is available with a 100 percent cost increase, and almost 54.8 percent of windy land is assumed to be available with a 200 percent cost increase."

page 162, http://www.eia.doe.gov/oiaf/aeo/assumption/pdf/renewable.pdf

That's just silly. There's enough low-cost wind in the central wind-belt of the US to power the whole country, should we choose. They're assuming that Eastern states would prefer to pay for $.24/KWH power within their boundaries, rather than buying $.08/KWH power from Iowa. That makes no sense.

I'm not sure why you're adding up the wind and solar figures.

----------------------------

This is the International Energy Agency (hardly a pro-wind outfit) on long term average costs, in their 2008 Energy Outlook:

http://farm4.static.flickr.com/3413/3216687750_58ed243b8d_o.jpg

You'll see that in the US that wind beats everything but coal, and even coal is only slightly ahead.

Christy said at May 4, 2010 2:48 PM:

When disparate groups make common cause I look for the hidden agendas.

Has anyone had a transmission line go up anywhere near them in the last few years? Anyone? Remember, one of the lessons learned from the last big New York Blackout was the lack of reliable transmission lines. I've been out of the power utility business for nearly a decade now, so I'm not up to date, but I don't think we've seen much of an improvement in the distribution systems. Utility companies have an incredibly hard time getting transmission lines built. No one wants them in their own backyard. Public support is nil. Regulations are discouraging.

Wind power makes us feel all warm and fuzzy inside. I suspect the utilities would like to take advantage of "good" power to get these transmission lines built. Then, of course, they could use them to transmit "bad" power to the Midwest when the wind power doesn't produce as expected.

Christy said at May 4, 2010 2:51 PM:

Sorry, my previous comment was totally off base. Damn the speed-reading.

Randall Parker said at May 4, 2010 6:27 PM:

Among the cows,

You give the example of 1500 km. But I think 1500 miles is more representative of the problem. It is 1668 miles from Minot ND to Philadelphia PA. That's 2684 km. Then we are up to around $900 million for capacity to move the electricity. It is even further to New England. Though Amarillo TX to Atlanta GA is only 1100 miles or 1770 km. Colby KS to Charlottesville VA is over 1300 miles. These are the kinds of distances involved in moving US plains state wind to the East Coast. It seems especially problematic for New England.

What's the loss rate per distance with HVDC?

LarryD said at May 4, 2010 6:37 PM:

Nick G "I'm not sure why you're adding up the wind and solar figures."

"...If the solar energy and wind energy areas were connected and treated as an integrated system, there would be more reliability than for either system by itself."

The point of the exercise is to show how expensive such a combined system is.

Nick G "They're assuming that Eastern states would prefer to pay for $.24/KWH power within their boundaries, rather than buying $.08/KWH power from Iowa. That makes no sense."

The EIA report "As mentioned, the costs shown in the table are national averages. However, there is significant local variation in costs based on local labor markets and the cost and availability of fuel or energy resources such as windy sites. "

The IEA citation isn't very useful, it's a graphic with no exact figures, and when I went to the IEA web site and tried to download the 2008 report, what I got didn't even have that graph in it.

You may not think much of the EIA, but at least their report has some transparancy.

Greg F said at May 4, 2010 7:08 PM:
Now that's a simplistic statement, and not true. Wind power variance is very similar to demand variance, so a statement like that is like saying that utilities can't deal with demand variance.

No. Your statement is simply not true. The transient changes of wind are far faster then demand changes. The changes in the wind are not predictable, the demand changes are.

I get so tired of people who don't know basic electrical theory, much less understand the enormous complexity of the grid, offering up comic book solutions to electric power generation. They don't understand that the generators have to be carefully synced on the same frequency even when the plants are 100's of miles apart. The don't understand the concept of reactive power and how it has to be compensated for. They don't seem to understand there will be extended periods of time (mostly in summer and winter) when there would be no significant wind over the entire continental US. They resort to averages of wind power when the reality is that power has to be balanced over very short time spans (seconds). In order to stay alive you have to breath. If someone suggested that if you consumed your average requirement of oxygen for a day in 23 hours you don't have to breath for the last hour you would think they were crazy. And they would be. The same is true of power generation. The power has to be delivered on very short time scales or it becomes next to useless. Averages filter out the short time scales making wind power look attractive. It's a comic book reality.


Among the cows in Iowa said at May 4, 2010 7:21 PM:

"You give the example of 1500 km. But I think 1500 miles is more representative of the problem."

I gave 1600 km, 1000 miles. I added $25 million for extra distance. Read the ABB link for more.

"What's the loss rate per distance with HVDC?"

About 3% per 1000 km.
http://www.futurepundit.com/archives/007089.html#reply20100413082004

someone said at May 4, 2010 8:32 PM:

--
Sean Lynch said at May 3, 2010 5:25 AM:

I'm kind of surprised nobody has pointed out the fatal flaw with this whole plan: the grid will go down, taking huge swaths of the nation with it. There is precisely one way to build a reliable power grid: don't connect pieces of it together. Not as efficient, but failure in one area doesn't propagate to the whole country.

We need distributed power generation and local storage or elastic use of surplus power. Of course, distributed generation is what we'll get with this plan anyway: everyone will need to buy a generator when the entire nation has a week-long blackout.

Joe Blow said at May 4, 2010 10:09 AM:

Single grid, single point of failure, particularly if "smartgrid" technology is fielded. The vulnerability to cyberattack in particular is terrifying. Check out the problems Brazil has had over the last year, which some attribute to a series of cyberattacks.
--

Exactly. been wondering myself for years.

And what about warfare? a sustained cyber-attack (from inside?)? strategic EMP blasts? Boom. back to the stone age for more than a week you can be sure.

* I have the same concern over the FAA's move from ground-based radar to satellite. I wonder if we should pay to keep the ground-based system as a back-up in case of 'unforseen' events. If we were to lose GPS and/or want to rebuild the ground system, it would take a while.

I'm sure I'm being overly cautious, but it would be nice to hear these concerns publicly addressed (I don't recall they were at all), if even dismissingly.

Mark in Texas said at May 4, 2010 9:13 PM:

As Greg F pointed out electricity has to be produced within a fraction of a second of when it is consumed.

Demand patterns are pretty predictable. There is very little demand in the middle of the night, in fact there is usually less demand than the minimum that is produced at base line thermal and nuclear power plants. These are units that cannot be started up easily and are usually shut down only for scheduled maintenance. In the morning electric demand grows as people wake up, turn on their TV and cook breakfast then head to work. Peak demand is during the day when industry is in operation, then in the afternoon people head home, turn on their TV, computer air conditioner or heater and cook dinner and then demand drops off as people go to bed.

Currently, the United States does not have a lot of east-west transmission capacity. Since we have four time zones, this prevents wheeling power between time zones which would allow otherwise unused power to be utilized at both ends of the day. This would not require any additional smart grid technology, just additional transmission lines.

WindWatcher said at May 4, 2010 10:37 PM:

Wind doesn't need as much back up generation as you might think and it doesn't need to come from fossil resources. Battery technology is advancing rapidly as a resource for balancing wind generation. Wind power can be used to charge batteries when there is surplus wind and the storage used when the wind dies. Some utilities already have large battery projects. A single truck trailer located near the wind generation can provide several megawatts of power. This also solves part of the transmission problem. By siting the battery storage near the generation, transmission lines do not have to be upgraded to higher capacity voltage to transmit from a large windfarm. The surplus power just goes to the batteries so the capital that would have been used to upgrade the voltage can go to battery storage. Many startups in energy storage and many of the applications are just scalable mdoules of car batteries. The dual market for batteries is accelerating the research to find cheaper, faster charging, higher capacity batteries and nano technology (Altairnano and A123 Systems) is making better cheaper faster possible.

What about pairing wind and hydro? A Dutch company is designing an energy island in the North Sea that uses these together.. The island will be constructed of walls around an interior cavity. The clay bed of the north sea enables a water tight seal so wind turbines on the island can pump water out of the cavity when the wind blows and traditional hydro generation turbines will capture power when the water flows back into the cavity. The island can store up to 2 GW hours of electricty for use on demand.

These ideas won't work everywhere, but with smart grid technology, which is being deployed at an accelerated pace due to government stimulus funds, it is possible to pair small wind and storage and small solar and storage just about anywhere on the grid. It will make the grid not just a network but a smart network that can operates as a virtual power plant. The key will be giving customers incentives to participate. Some energy economists believe that power markets will be opened to competition once the grid is smart. There will be too many service companies (like Google) who wil be able to go around the utiity regulatory structure and offer value to energy users for our current system to work. Customers will demand choice.

Among the cows in Iowa said at May 5, 2010 9:12 AM:

"Single grid, single point of failure, particularly if "smartgrid" technology is fielded."

The HVDC papers say it stabilizes the grid because it is controllable.

"The vulnerability to cyberattack in particular is terrifying."

Getting grid control off the public Internet is a good idea. Air gap!

Big D said at May 5, 2010 12:10 PM:

I have heard that wind turbine power generation at full speed drops to as little as 40% of rated maximum during the summer months (due to the thinner, hotter ambient air--think about how a helicopter loses lift in high/hot conditions). Can anyone confirm/deny this number?

I have also heard that wind turbines produce relatively "dirty" (frequency/phase variations, not carbon emissions) electricity. Confirm/deny?

If correct, how does it make any sense to commit to any use of wind turbines beyond a niche where their drawbacks are lost in the sheer size of the non-wind electricity being generated, transmitted, and used?

Also, while I'm not going to proclaim it as a solution until and unless it proves itself, if polywell *does* somehow work out, I *would* call that a panacea...

Kevin Stein said at May 5, 2010 7:41 PM:

Any wind farm or solar array is going to have to deal with storage issues and that adds costs. How to level the grid is one of the most pressing issues and like transfering energy over long distances needs to be tackled from all sorts of angles. One of them would be using AFC (Alkeline Fuel Cells) to store the energy or to produce hydrogen which could also be used as a fuel for hydrogen based vehicles. There is no silver-bullet out there. It is getting all technologies to the level where this jig-saw puzzle fits together to meet our energy needs while protecting our environment.

Gridshift, a company based in Georgia, has just published a white paper about a no-platinum, highly efficient and high rate electrolysis unit which could be one part of the solution we are looking for. Anywone interested in how hydrogen production relates to the grid and solar/wind should check out this white paper at:


www.grid-shift.com/white_papers


Thanks again for the interesting article and I certainly enjoyed reading the lively comments.

Nick G said at May 5, 2010 7:54 PM:

LarryD,

I think you're adding up wind and solar when you should be averaging them. After all, for each of them you bought enough resource to provide all the KWHs in the system. If you're going to have a hybrid wind/solar system, then you need to buy 50% (or so) of each. You've doubled the cost of the system.

Per the EIA: However, there is significant local variation in costs based on local labor markets and the cost and availability of fuel or energy resources such as windy sites.

They're requiring that a local grid use very high-cost power within their geographic boundaries, rather than importing much cheaper power. That may make political sense, but...

The IEA citation isn't very useful

Yes, I borrowed that from Jerome A Paris, so I'm not really familiar with the source. I'll have to see if I can find something better as an authority. FWIW, all of the cost estimates I've seen for the US say that wind, nuclear and new coal are in the same neighborhood of costs (excluding external costs, of course).

Engineer-Poet said at May 6, 2010 7:36 AM:

LarryD did add figures in $ per 100 GWh, so the total would be 200 GWh.  The total should be divided by 2.  (Another triumph for unit analysis.)

HVDC has interesting properties.  First, the inverters can match any frequency and phase; IGBT inverters can produce or consume reactive power and start a grid from cold (thyristor inverters require external exciting voltage).  An AC line transmits power based on the phase difference between its ends and its impedance, which isn't always what's desired or useful (or even stable).  The power output of an HVDC inverter can be controlled in real time.  This means that an HVDC system can shift power between multiple points with ease and smoothness, which effectively shares resources like spinning reserve between them as well.  Add some fat capacitor banks at terminals and you'd get some extra resilience (I'd calculate the capacitance of the line itself but it's been too many years since physics class and I've got no time).

Sure, HVDC is vulnerable to EMP.  So's our current system.  I fail to see how a supergrid costs us anything in that regard.

Nick G said at May 6, 2010 9:02 AM:

Wind power variance is very similar to demand variance - Your statement is simply not true.

Yes, I believe it is. I didn't make that up - it was from published research. I'll look to find it again.

The transient changes of wind are far faster then demand changes.

Some of them are, some aren't. Keep in mind that "demand" is an aggregate of a lot of consumers, some of which change quite quickly. In the same way, we have to look at aggregate windpower production. The variance of aggregate windpower production depends on the level at which one analyzes it. I believe the research I have in mind looked at the aggregate of several local windfarms.

The changes in the wind are not predictable

Of course they are. Obviously, they aren't 100% predictable, but the value in question depends on the timeframe. Forecasting 3 days ahead is almost meaningless, but forecasting 1 hour ahead is extremely useful.

I get so tired of people who don't know basic electrical theory, much less understand the enormous complexity of the grid, offering up comic book solutions to electric power generation.

Gosh, how did those people at the National Renewable Energy Laboratories ever get their PhDs? How could they propose aggregating widely separated wind farms, as they did here: http://www.nrel.gov/wind/systemsintegration/ewits.html to find that 30% wind market share was feasible? How did those Stanford professors get tenure, when they did the same sort of analysis here:

http://news.stanford.edu/news/2007/december5/windfarm-120507.html

"Wind power, long considered to be as fickle as wind itself, can be groomed to become a steady, dependable source of electricity and delivered at a lower cost than at present, according to scientists at Stanford University.

The key is connecting wind farms throughout a given geographic area with transmission lines, thus combining the electric outputs of the farms into one powerful energy source. The findings are published in the November issue of the American Meteorological Society's Journal of Applied Meteorology and Climatology. "

They don't seem to understand there will be extended periods of time (mostly in summer and winter) when there would be no significant wind over the entire continental US.

Do you have some data for that? I'd love to see it. It will come as a big surprise to a lot of people.

They resort to averages of wind power when the reality is that power has to be balanced over very short time spans (seconds).

No, they really don't. They're looking at geographical averages, not time averages.

Nick G said at May 6, 2010 10:03 AM:

Mark in Texas.

There is very little demand in the middle of the night, in fact there is usually less demand than the minimum that is produced at base line thermal and nuclear power plants.

Do you have some data on that? I've been hearing anecdotal reports of wasted night-time generation for a while, and I haven't been able to track down real info.

Big D,

Are you from Dallas?

I have heard that wind turbine power generation at full speed drops to as little as 40% of rated maximum during the summer months

I don't think air density changes nearly that much from winter to summer.

Zero C: 1.292
25 C: 1.184

per http://en.wikipedia.org/wiki/Density_of_air

I have also heard that wind turbines produce relatively "dirty" (frequency/phase variations, not carbon emissions) electricity.

That was a problem with early turbines - it's been solved. Current turbines provide the normal "services" expected from a generator.

Among the cows in Iowa said at May 6, 2010 2:01 PM:

The Iowa senate voted to fund a study of a new nuclear plant in Iowa.
http://www.world-nuclear-news.org/NN-MidAmerican_nuclear_study_fee_clears_Iowa_Senate-1003105.html
Iowa has one nuclear plant, the Duane Arnold plant in Palo. This is upriver from me.

Iowa gets 77% of its electricity from coal. The national average is 50%. Iowa needs more wind just to catch up. Exporting wind power evens out the bumps.

Greg F said at May 6, 2010 6:32 PM:

Nick,

The variance of wind.
http://i116.photobucket.com/albums/o25/techtipmail/Wind%20Power/OntarioHourlyWind.jpg

The variance of demand.
http://i116.photobucket.com/albums/o25/techtipmail/Wind%20Power/OntarioHourlyDemand.jpg

See the difference Nick? Each vertical division represents a week. Even with hourly averages it's not even close. The variations in wind speed over short time frames are not predictable. You are stuck on the fallacy of averages. Power has to be delivered in very short time frames. Averages hide that inconvenient fact.

Gosh, how did those people at the National Renewable Energy Laboratories ever get their PhDs? How could they propose aggregating widely separated wind farms, as they did here:
Don't be so damn naive. First, how they got their PhD's is irrelevant. How they keep their job may be relevant. It's not usually a good career move to arrive at politically incorrect conclusions. It usually doesn't do much for your funding. I live within 10 miles of 4 colleges/ universities. Some of the most ignorant remarks I have ever heard come out of the mouths of people with PhD's. Either way all you are doing is an appeal to authority. Basically Nick your telling me you don't understand the technical issues because you never address them.

I said

They don't seem to understand there will be extended periods of time (mostly in summer and winter) when there would be no significant wind over the entire continental US.

Nick responds:

Do you have some data for that? I'd love to see it. It will come as a big surprise to a lot of people.

See the Ontario plots. They have wind generators that span the distance from central Wisconsin to central New York. Roughly 600 miles. Whether it surprises "a lot of people" is irrelevant. Wouldn't be the first time in human history that happened.

No, they really don't. They're looking at geographical averages, not time averages.
From the Stanford press release:
The researchers used hourly wind data, collected and quality-controlled by the National Weather Service, for the entire year of 2000 from the 19 sites.
Hourly data is marginally useful when seconds are important. It's an average. An average is a low pass filter which means they are filtering out the high frequency variability. Your using a magnifying glass when you need a microscope.

Next heat wave go to:
http://vortex.plymouth.edu/make.html

It's an ugly web site but you can get national wind data. The cold snap this past January I looked at the data there. Almost nothing over 15 MPH. You don't have to believe me, you do have to do more then cite press releases from people who have an incentive to get the right answer.

Engineer-Poet said at May 6, 2010 10:12 PM:

Greg F, I don't suppose you noticed that:

  1. The hourly demand graph is plotted at 500 megawatts per division, but
  2. the the hourly wind production graph is plotted at 50 megawatts per division?
That makes the wind look a LOT more variable than it is.

I tried to make a Google map to show exactly how Ontario's wind farms are distributed across the province, but the interface is very irritating for the first-time user.  I did see that the farms I found are between lakes Huron and Erie, and 600 miles is tiny compared to the geographic expanse of Canada (let alone the USA).

If you look on the scale of seconds, local grid demand jitters by a few megawatts.  Most of this is handled by short-term variations in frequency; the power demand of anything running on an induction or synchronous motor varies with grid frequency, so this automatically matches demand with supply.  Plants used for spinning reserve and regulation typically have response times of seconds to minutes.

I am curious about General Compression, specifically the technology they have which allows them to switch from storage to generation in seconds.  I'd like to see what lets them make this claim.  But if they can actually do it, the variability of wind power is a non-issue.

Randall Parker said at May 6, 2010 11:55 PM:

E-P,

There are different time scales over which wind's variability matters. There's the short term moment to moment. The ability to use stored power helps with that short term fluctuation. But storage device cost more the longer they store. Also, the availability to storage at low cost depends on geography and geology.

Wind that stops blowing for many hours requires back-up that is not from storage. Wind that stops blowing for days requires back-up that is more like baseload but has to be cheaper in capital cost than baseload and therefore higher in fuel cost and higher in kwh.

Then there are seasonal differences. Electric power is most valuable on hot summer afternoons. That's not wind's sweet spot.

Spaceman said at May 7, 2010 7:36 AM:

Irrespective of the benefits, there will be great challenge in obtaining right-of-way and constructing transmission lines. The Feds best be prepared to exempt the lines from the bulk of required approvals from various federal and state governments, environmental hurdles. Otherwise, the construction will be tied up in court for the next decade

Greg F said at May 7, 2010 1:54 PM:
Greg F, I don't suppose you noticed that: ...
Of course I noticed Engineer-Poet. Wind is but a small proportion of demand but it doesn't matter as were talking about the intermittency in time. Hypothetically increase the capacity at all the wind farms by 10 times and what changes? Only the scale of the amplitude would change. The graphs would look the same. The frequency spectra would be unaffected. If I have some time this weekend I will do a FFT of both. The wind generators I expect will have a random distribution of frequencies with the exception of the seasonal changes. On the demand side I am sure there will be a pronounced peak at 24 hours as well as seasonal changes. There is no doubt that 2 will look quite different in the frequency domain.
I did see that the farms I found are between lakes Huron and Erie, and 600 miles is tiny compared to the geographic expanse of Canada (let alone the USA).
I don't know what your point is. In practice power is not generally transported that far anyway (exception is dedicated high voltage transmissions lines). It's a grid. Electrically it's a mesh circuit of distributed voltage sources and loads. The percentage of power you get from any one generator is going to decrease with distance.
If you look on the scale of seconds, local grid demand jitters by a few megawatts.
Compared to what? If your total demand is 4 MW and it jitters by 2 MW you have a big problem. Then again if your total demand is 200 MW a couple of MW of jitter is not an issue.
I am curious about General Compression, specifically the technology they have which allows them to switch from storage to generation in seconds.
I would view this with a great deal of skepticism as industrial compressors are notoriously inefficient (10%). It wouldn't surprise me if there is significant rent seeking going on here. Glowing press releases go back 3 decades of promising advances that never materialized. Bottom line is you can't build a system hoping someone will invent a critical part you need. It's just foolish.
Nick G said at May 7, 2010 4:27 PM:

Greg,

I certainly understand the technical details of the situation. OTOH, I don't think my saying that is going to convince you. That's why I refer to experts in the field. You can dismiss any published literature you like as biased, but it doesn't seem to advance the discussion much.

Now, let's have a more serious discussion: I think the heart of the problem here is that you're not looking at the statistics of the situation.

Hypothetically increase the capacity at all the wind farms by 10 times and what changes? Only the scale of the amplitude would change

That's not at all true. Wind capacity isn't going to increase by multiplying existing capacity, it will increase through more supply in different places, which are not closely correlated. See http://en.wikipedia.org/wiki/Law_of_large_numbers

Wind in Ontario is what, 2% of overall supply? If you were to take a random sample of 2% of all consumers, their demand curve would look much, much more chaotic. Talk to any ER manager: any one emergency appears completely random, and yet they can predict demand reasonably well, and size helps: a big ER is much, much easier to staff for than a small one, because the ratio of variance to the mean will be much, much smaller.

Nick G said at May 8, 2010 10:26 AM:

Greg F,

They don't seem to understand there will be extended periods of time (mostly in summer and winter) when there would be no significant wind over the entire continental US. See the Ontario plots. They have wind generators that span the distance from central Wisconsin to central New York. Roughly 600 miles.

The funny thing is, we've had this conversation before. Here's what I said last time:

Canada is a funny place: it's provinces are enormous, but it's population is small, and concentrated in the south. In this case, the wind farms are concentrated around Windsor and Toronto in a relatively much smaller area. Further, as of 2008 they only had about 1.2GW wind (nameplate), which really isn't much.

First, you need to do a much more careful analysis than just eyeballing a few days: you need to look at the parts of the year and day where capacity is actually needed, and you have to use the statistical approach preferred by the local ISO to forecast capacity credit. This may not feel intuitive, but the absolute minimum isn't the correct approach. Keep in mind that all forms of generation have variance. Nuclear, for instance, can "trip" in seconds - that can remove a full GW from the grid for days. That's why Ireland, for instance, chooses not to use nuclear: that kind of variance is too much for a small grid.

2nd, if generation can't get above 85%, then you don't need to "back up" any more than that. Actually, you don't need to back it up at all, because you never plan for that kind of generation as part of peak capacity - you just have to manage the change in generation, as the winds die down and wind generation comes down from it's peak. The best way to do that is to use Demand Management: ramp up demand during that peak (say, by signaling to commercial refrigeration to set it's temp down a couple of degrees, and for electric vehicles to charge at maximum), and then ramp demand down as the wind dies. Very cheap, very effective, well tested, widely used. It's what Texas used recently to deal with such a problem.

hourly data is marginally useful when seconds are important. It's an average.

Ok, this is an important question. Everything I've seen suggests that's not the case. Have you seen actual data?

Next heat wave go to: http://vortex.plymouth.edu/make.html It's an ugly web site but you can get national wind data. The cold snap this past January I looked at the data there. Almost nothing over 15 MPH.

Greg, this site gives surface wind data. Wind speeds have to be measured at the height of the wind turbine, commonly 80 meters up. Wind speeds are much higher there.

In practice power is not generally transported that far anyway (exception is dedicated high voltage transmissions lines). It's a grid. Electrically it's a mesh circuit of distributed voltage sources and loads. The percentage of power you get from any one generator is going to decrease with distance.

That's true, but that's just an artifact of the way we currently generate. It doesn't have to be that way, and it shouldn't.

Engineer-Poet said at May 8, 2010 12:59 PM:

Quoth RP:

But storage device cost more the longer they store.
The cost of storing compressed air underground is minuscule.  Dakota Salts is just one of several efforts I've seen talking about 50 hours of storage.

Word on TOD is that NG prices are far below breakeven for the shale plays, but drilling is being driven by speculation.  Sooner or later production will slide until prices come back up to the $7-$8/mmBTU range (the USA doesn't have enough LNG capacity for imports to hold the price down).  The direct competition for CAES systems is simple-cycle gas turbines (they have similar response times; CCGT takes much longer to fire up).  If gas costs $9/mmBTU to the customer, the difference between a GE simple-cycle turbine at 46% efficiency and a CAES system at 80% gas-to-electric efficiency is about 2.8¢/kWh.  If you assume zero fuel cost 50% of the time (because wind generation is greater than demand and only compressors are operating), you can roughly double that difference for the system as a whole.

Quoth Greg F:

Hypothetically increase the capacity at all the wind farms by 10 times and what changes?
Why hypthesize the impossible?  Adding wind capacity means increasing the geographic dispersion, which in turn smooths the production curve.  Besides, Ontario shares power with the USA from Michigan all the way to the east coast.  (I've seen the lines across the Detroit and Niagara rivers.)
I don't know what your point is.
The point is that the correlation of wind-farm output decreases with increasing distance.  Ontario's real-time variation with Saskatchewan and the Dakotas is going to be close to nil.
Electrically it's a mesh circuit of distributed voltage sources and loads.
That's a hyper-simplification into DC terms, which doesn't apply usefully at such a scale.  Electrically it's a system of fractional-wavelength transmission lines with relatively few point sources of real and reactive power and much more finely distributed consumers of both.  Adding a supergrid of HVDC links with PWM inverters allows some of the transmission-line behavior to be bypassed, which increases grid stability.
If you look on the scale of seconds, local grid demand jitters by a few megawatts.
Compared to what?
Compared to the total load of multiple GW.
I would view this with a great deal of skepticism as industrial compressors are notoriously inefficient (10%).
If you're going to pull numbers out of there, at least wash them in hot water before you post them; it's unsanitary.
It wouldn't surprise me if there is significant rent seeking going on here.
You mis-spelled "arbitrage".  The Ludington pumped-storage facility on Lake Michigan has been doing that for decades, and something which can switch from load to source in seconds can also make money on fast-response regulation and spinning-reserve services as well.  That will only matter until the market for regulation is saturated by electric vehicles, but by then the investment should be amortized and the O&M can be paid by arbitrage between opposite supply/demand balances.

th said at May 8, 2010 5:06 PM:

"Word on TOD is that NG prices are far below breakeven for the shale plays, but drilling is being driven by speculation." Speculation on what? Speculation is usually what comes from garbage like goldman sachs, if these leeches can't force the price of it up, it must be an extreme oversupply? I also saw that the word on the oil drum is that huge debt is what is needed to fix the economy, lots of Wiemar republic fans over there, you one of them?

Engineer-Poet said at May 9, 2010 8:53 AM:
Speculation on what?
The shape of the production decline curves of the shale plays.  Nobody knows how much gas these wells are ultimately going to produce (there aren't many old ones yet, and none in many reservoirs), and some say that the ultimate recoverable resource is as little as a third of what's being touted.
th said at May 9, 2010 4:31 PM:

poet, barnett is still growing production, it's almost 30 years since the first well was drilled, you're reading simmons too much, marcellus and haynesworth and fayetteville and about 15 others are the same type of formation as barnett, by the way, what do you think of this little detail that always surface after its too late.


http://www.bostonherald.com/business/general/view.bg?articleid=1253263

Randall Parker said at May 9, 2010 4:32 PM:

th, The word on TOD is not that huge debt is needed to fix the economy. Rather, some of the writers are forecasting massive defaults, large scale failures of financial institutions, and financial collapse. I agree with them on the massive defaults. But I do not expect total collapse.

Try being less insulting and more accurate.

E-P,

I was amazed to read Henry Groppe agree on the expected big surge upward in natural gas prices. He's forecasting $8/mmBTU by the end of 2010. I was inspired to go buy some shares in EnCana.

As for storage in salts: What's the potential capacity for this form of storage? Also, where's it located?

When I say it costs more to store for longer period of times part of my point is that storing for many short cycles costs less per cycle than storing for one long cycle. How often the storage gets used determines how many storage cycles the capital costs can be spread over.

I see a big problem, for example, in trying to store solar electric power from the summer to use in January. The storage cycle is very long. So the capital costs have got to be low or it just isn't practical.

As I've noted, wind power output varies substantially between months and between years. I doubt storage mechanisms can cope with this variability. Back-up power generators and dynamic pricing have to do the job.

th said at May 9, 2010 5:47 PM:

parker, from the article...."Let's start with the 1981-1987 period. I have always wondered how the US was able to add all kinds of nuclear facilities, ramp up coal production, and improve auto mileage, back at the time oil consumption dropped in the early 1980s. Maybe the answer was more debt!"

"Back in 1981, oil was very expensive, in terms of the disposable personal income of Americans. But it gradually got much cheaper in the 1981 to 1987 period, as debt was ramped up. Part of this may have been other energy sources that were ramped up to help substitute for oil helped reduce the real demand for oil, and thus bring down it's price"

This woman seems to want to find good things debt does that debt didn't do, increased debt or better insulation or oil substitutes had nothing to do with the price of oil at that time, the cure for high prices is high prices, non-opec production grew by 10mmbbls/d while demand went down providing lots of excess oil on the market, that's pretty basic economics, we've now got the huge debt and high oil prices, the consequences of that debt will eventually destroy lots of things, oil demand will be one of them, so there's the connection, but not the one she's suggesting.

Engineer-Poet said at May 9, 2010 6:54 PM:

th, it would help if you paid attention.  The hydro-fracked horizontal wells are quite new and the projections of "asymptotic decline" are selling interests in wells but, according to the experts I've been reading, are not proven.  The initial flows are good but decline very quickly; the curve of the tail is the difference between a 10 BCF well and a 3 BCF well.

Randall, Dakota Salts is also a potash-mining operation.  They appear to be solution-mining deposits in the Williston basin near the borders with Saskatchewan and Montana, leaving cavities deep underground.  These are ideally suited for storage of compressed air, as they're essentially free.  That's a bit to the NW of the bulk of the purple areas of the wind map, but when the state has 2.9 TWH/yr of potential wind generation, running some lines from e.g. Ashby to Westby (400 miles by road) wouldn't cost all that much.  Figuring 8 parallel HVDC rights-of-way at $500,000 per mile and 100 GW capacity each† (for ND's total 770 GW of capacity plus extra), those runs would cost $3.2 billion or about 0.4¢/watt.

Yes, wind varies a lot by season and by year.  But that's not what the nay-sayers are complaining about; they are saying that variation on the scale of hours to days makes it impossible to integrate with the grid.  As you can see, storage turns that argument on its head; if you have as little as 50 hours of storage which averages half-full at any time, you can commit to large amounts of generation for a full day in advance no matter what the wind is doing.  It even allows the storage facility to provide spinning reserve.  When the wind is low, you burn the fuel you would have burned if you had no wind.

† I'm figuring that the ROW is the biggest expense and the cost of increasing the current capacity is mostly just stringing more wire; putting up bigger towers for all the extra wires would be another capital expense, but you wouldn't have to buy the insulators and wire until you needed them.

th said at May 10, 2010 4:44 PM:

poet, hydro-frac, horizontal drilling is what made barnett balloon into what it is now, it's working, anything above $3.50 gas before the late 90's was considered to be a boom, now it looks like $4 is the bottom, there's your difference between now and then even with hydro frac and sideways drilling, it isn't going back to $2.50 but hopefully it ain't going back to $12 which is good, unless the lefties and the AGA find a way to screw it up. The real test which I think is coming, will be when the hedge funds, speculators, goldman sachs and LNG import facilities all shut down and get the hell out of the nat gas market.... reliable, multi-year supply perceptions do that, all making simmons look a little hysterical.

th said at May 10, 2010 5:11 PM:

"As you can see, storage turns that argument on its head;" what storage, compressed air? What is the return on compressed air? Are you saying that the compressed air storage will turn the blades as well as the wind did, if only coakley had won in massachusetts this wouldn't be the embarrassment to all things wind.

Greg F said at May 10, 2010 8:08 PM:

In response to EP's complaint about the scale difference between the two graphs (demand and wind power generated in Ontario) I wrote:

Hypothetically increase the capacity at all the wind farms by 10 times and what changes?

The point was that increasing capacity won't change the intermittency. EP then waves his hands proclaiming:
Why hypthesize the impossible? Adding wind capacity means increasing the geographic dispersion, which in turn smooths the production curve.

It's the old assumption that wind is random and if it isn't blowing in one place it will in another. The assumption is dead wrong. Despite the fact that the Ontario wind farms are dispersed as far as 600 miles apart on 3 different Great lakes EP has continued to wave his hands proclaiming this assumption is delivered truth. First the readers will need a grasp of the power curve of a wind turbine. The following is for a 1.5 MW GE wind generator. The important part of this document is the power curve located on page 4.

http://www.gepower.com/prod_serv/products/wind_turbines/en/downloads/GEA14954C15-MW-Broch.pdf

Here is reality. With the power curve in mind examine below the four animations covering a 24 hour period for north America.

http://www.cam.cornell.edu/~pauljh/US_Composite_Radar/2009-12-18/wunderground.wind-speed.2xus_ws_anim-0.gif
http://www.cam.cornell.edu/~pauljh/US_Composite_Radar/2009-12-18/wunderground.wind-speed.2xus_ws_anim-1.gif
http://www.cam.cornell.edu/~pauljh/US_Composite_Radar/2009-12-18/wunderground.wind-speed.2xus_ws_anim-2.gif
http://www.cam.cornell.edu/~pauljh/US_Composite_Radar/2009-12-18/wunderground.wind-speed.2xus_ws_anim-3.gif

What you should have noticed is the vast majority of the North American continent is blue with wind speeds under 7 meter/sec. So much for smoothing the production curve. EP goes on with this assumption stating:

The point is that the correlation of wind-farm output decreases with increasing distance. Ontario's real-time variation with Saskatchewan and the Dakotas is going to be close to nil.

The point is that EP is wrong. Go back to the gif animations and pay special attention to Canada. Also, notice the wind direction that is shown by the arrows and the large geographical areas they encompass. To further emphasize the point more wind maps from this weekend.

http://i116.photobucket.com/albums/o25/techtipmail/Wind%20Power/Weekend1.jpg
http://i116.photobucket.com/albums/o25/techtipmail/Wind%20Power/Weekend2.jpg

EP then tries to dazel us with his brilliance. I wrote:

Electrically it's a mesh circuit of distributed voltage sources and loads.

EP now shows us how little he really knows about electrical theory states:
That's a hyper-simplification into DC terms, which doesn't apply usefully at such a scale.

First, EP's electrical theory is so weak he doesn't realize that mesh circuits don't only apply to DC circuits. See page 24 of the following from Purdue for an example:

www2.tech.purdue.edu/eet/courses/eet207/Lectures/23-MESH.pdf

Second, the second part of EP's statement is a straw man. He is telling us that something he doesn't understand doesn't apply. Priceless.

EP's post couldn't be complete without an arrogant snide remark. My comment was to the utility of compressed air for backup storage as EP, ever so naive, was promoting. I said:

I would view this with a great deal of skepticism as industrial compressors are notoriously inefficient (10%).

EP responds with an insult.
If you're going to pull numbers out of there, at least wash them in hot water before you post them; it's unsanitary.

Cute. If you don't have an argument just insult. But that isn't the end of it. He then follows by moving the goal posts.
The Ludington pumped-storage facility on Lake Michigan...

The Ludington pumped-storage facility is pumped water, not compressed air.

Greg F said at May 10, 2010 10:03 PM:

Nick you don't understand the technical details at all. You made a claim that "wind power variance is very similar to demand variance". I posted 2 graphs that a blind man could see that they weren't similar and what do I get from you? Nothing. You ignored it. There they are the statistics in graphical form and you have the audacity to accuse me of not looking at the statistics. You're projecting. You're not looking at the real data, the real statistics. In fact you're ignoring them.

I don't dismiss the published literature, I verify it. If it is at odds with reality it's wrong. It doesn't matter how many Phd's wrote it. It wouldn't be the first time the so called experts were wrong. Study the history of Eugenics. That atrocious time in history was driven by the "experts". And they were wrong. Spectacularly wrong. You on the other hand take the experts as gospel truth in spite of the contradictory evidence. Papers published by academics who have never worked in the power generation industry (no experience) that are peer reviewed by other academics who also have never worked in the power generation industry (more lack of experience) does not constitute an "expert".

The funny thing is, we've had this conversation before...
The funny thing Nick is you don't listen. Your assertion that "the wind farms are concentrated around Windsor and Toronto " (whatever "concentrated" means numerically) is unsupported. In fact your assertion is not defined in any meaningful way. What makes the hand waving even more silly is the distance from Windsor to Toronto is about 200 miles. Anyway here is a map with all of Ontario's wind farms.

http://www.ieso.ca/imoweb/siteshared/images/wind_generation_in_Ontario.pdf

Try doing more then waving your hands and making up facts.

Greg, this site gives surface wind data. Wind speeds have to be measured at the height of the wind turbine, commonly 80 meters up. Wind speeds are much higher there.
More hand waving Nick. The hub is 80 meters. The blades will be both higher and lower. The wind maps are produced with radar so the "surface" is not just 10' off the ground. Here is a little project for you that doesn't need any "experts". Gather hourly wind speed data for all or some of the wind turbine locations in Ontario. Use surface stations for real surface data and radar wind data for boundary layer wind. The hourly wind turbine data is published for each farm in Ontario. You should then be able to correlate the turbine output with the wind data on an hourly basis.
Randall Parker said at May 10, 2010 10:16 PM:

th, I've explained to you that the TOD folks (most notably Gail Tverberg, aka "Gail the Actuary") do not believe debt can save us from the effects of Peak Oil. Gail thinks the collapse of the banking system will undermine our ability to respond to Peak Oil and this failure of the banking system will make it hard to invest in alternatives.

The 1980s isn't Gail's model for the future of our response to fossil fuels supply problems. You have to read more if you want to find out what she really thinks the future holds in store.

I say this as someone who has repeatedly stated in the comments at TOD that we can handle the massive bankruptcies and finance the migration and that we can keep industrial civilization functioning. You can do Google for my comments on TOD and find my responses to her.

Hydro-fracing: Henry Groppe and Arthur Berman are both skeptical and argue that decline rates are too steep and therefore production costs are higher than some of the natural gas companies are stating. When the legendary Henry Groppe agrees with the position of Arthur Berman I get worried.

Randall Parker said at May 10, 2010 11:46 PM:

Greg F, Good pictures of wind patterns. A picture's worth a thousand words. I especially like the two graphs of output of Ontario wind farms. I'd like to see graphical data for wind output by US state by hour for recent periods of time.

I'd love to see a sped up animation of wind speed over the United States and Canada where each hour is about 10 seconds and it goes on for a week or two.

E-P, Nick G, I see Greg F as making reasonable arguments that you haven't refuted. Not saying he's right. Am saying he's being reasonable.

Engineer-Poet said at May 11, 2010 6:56 AM:

Quoth Greg F:

EP's electrical theory is so weak he doesn't realize that mesh circuits don't only apply to DC circuits.
You're so simplistic (or so trollish) you can't even get the question right.  The issue isn't "mesh", it's "voltage sources" (no generator in an AC grid can be usefully modelled as a voltage source).  You neglect to quote the words I was replying to, either because you're sloppy or in order to try to hide your tracks.
Second, the second part of EP's statement is a straw man. He is telling us that something he doesn't understand doesn't apply. Priceless.
Ironic.
Cute. If you don't have an argument just insult.
Let's have a cite for that 10% claim, and its relevance to the application here (e.g. not supply of shop air).  The isentropic efficiency of axial turbocompressors can be upwards of 87%.  If their efficiency were as low as you claim, gas-turbine powerplants would be impossible; the compressor back-work wouldn't let them run.
The Ludington pumped-storage facility on Lake Michigan...
The Ludington pumped-storage facility is pumped water, not compressed air.
The issue was not the fluid, but arbitrage (storing cheap off-peak juice, selling at peak).

Again you change the subject while trying to hide what you did.  This is extremely trollish.  Strike two.

Here is reality. With the power curve in mind examine below the four animations covering a 24 hour period for north America.

http://www.cam.cornell.edu/~pauljh/US_Composite_Radar/2009-12-18/wunderground.wind-speed.2xus_ws_anim-0.gif ....
The first thing I notice is that there's a lot of variation across the map.  The second thing I notice is that those 4 animations cover just 1 day (I'd like to dig for the maps for last week, if I have time).  Third, the Great Lakes are blank spots (they've got some Class 5 territory on them).  But the biggie took me a while to catch:  you're using wunderground figures.  The reported winds for weather maps are measured at a standard 10 meters above the ground.  I shouldn't have to mention that the typical wind-turbine hub is now placed at 80 meters.  Using the 7th-root approximation, the winds at 80 meters are about 35% higher than at 10 meters.  The amount of energy/area is ~140% greater, and the effect on the generation curve of the GE turbines you used as an example is going to be enormous; between 7.0 and 9.5 m/sec, the 1.5sle will go from ~650 MW to ~1350 MW output.

You picked a source which systematically understates the figure under debate by roughly a factor of two.  Strike three, you're out.

Greg F said at May 11, 2010 10:04 PM:

EP is off on one of his arrogant rants again so here goes. This is what I wrote:

In practice power is not generally transported that far anyway (exception is dedicated high voltage transmissions lines). It's a grid. Electrically it's a mesh circuit of distributed voltage sources and loads. The percentage of power you get from any one generator is going to decrease with distance.

Then EP makes assertion without evidence.

That's a hyper-simplification into DC terms, which doesn't apply usefully at such a scale.

First he assumed it was DC, which he now knows, is not a limitation on a mesh circuit. Then he follows with the straw man asserting facts without evidence. Just for fun:

Evaluating North American electric grid reliability using the Baraba´ si–Albert network model
http://giacs.isi.it/files/wp3_files/D%20P%20Chassin%20phys.%20A%20355.pdf

See figure 1 where it states:

Fig. 1. The transmission network is a mesh system (dark gray), but sub-transmission (light gray) and distribution (white area) portions are mainly radial structures.

The radial structures are "dedicated high voltage transmissions lines".

The part he accuses me of ignoring is just EP throwing something else at the wall hoping it will stick.

Electrically it's a system of fractional-wavelength transmission lines with relatively few point sources of real and reactive power and much more finely distributed consumers of both.

This is just hand waving. It makes no sense at all. That is why I ignored it.

You're so simplistic (or so trollish) you can't even get the question right.

Another insult. And the question is?

The issue isn't "mesh", it's "voltage sources" (no generator in an AC grid can be usefully modelled as a voltage source).

LMAO. If "no generator in an AC grid can be usefully modeled as a voltage source" there wouldn't be a grid. This is as bad as EP's DC assumption. Another one of EP's assertion without evidence

You neglect to quote the words I was replying to, either because you're sloppy or in order to try to hide your tracks.

EP can now read minds. I neglected them because it is nonsense. At 60 Hz the wavelength on a transmission line is 3100 miles. Even if we assume the velocity of propagation is 66% the wavelength is still 2050 miles. A line is considered electrically short when it is 1/4 of the wavelength (about 500 miles for our example). There is no utility in treating an electrically short line as a transmission line. Lumped parameters are sufficient. The only thing in the grid that approaches 500 miles are the "dedicated HV transmission lines".

EP never provides a cite but he sure likes to demand them.

Let's have a cite for that 10% claim, and its relevance to the application here (e.g. not supply of shop air).

The number I claimed was for industrial compressors. IOW shop air. That is what EP will get. Notice how EP attempts to dictate that I supply something different then what I claimed. He simply can't admit when he is wrong. Here is a reference.

http://www.mntap.umn.edu/energy/82-CompAir.htm

Compressed air is one of the most expensive uses of energy in a manufacturing plant. About eight horsepower of electricity is used to generate one horsepower of compressed air.

I was wrong. It isn't 10%, it's 12.5%. Beat me with a stick.

The isentropic efficiency of axial turbocompressors can be upwards of 87%.

Again EP asserts fact without evidence. Notice no cite.

If their efficiency were as low as you claim, gas-turbine powerplants would be impossible; the compressor back-work wouldn't let them run.

Another fact asserted without evidence. Funny, I just posted a reference that says they are as low as that.

The issue was not the fluid, but arbitrage (storing cheap off-peak juice, selling at peak).

Moving the goal posts. EP said "I am curious about General Compression" which is a company that is doing compressed air. Not pumped water. Pathetic.

The first thing I notice is ...

The first thing I noticed is the whole paragraph is hand waving. Not a single bit of data to support his assertions. Surprise surprise. The simple thing to do would be to get the hourly data from the wind farms in Ontario on those lakes for that 24 hour period. Guess EP didn't think of that. Then again maybe he did and it didn't produce the results he would have liked.

Greg F said at May 11, 2010 10:14 PM:

Thanks Randall. The problem with US wind power is most of it is proprietary even though the tax payers are paying the tab. Denying access to the data seems to be the method of operation for the environmental movement.

Among the cows in Iowa said at May 12, 2010 9:42 AM:

"The number I claimed was for industrial compressors. IOW shop air."

Shop air uses piston compressors.

"Funny, I just posted a reference that says they are as low as that."

But not anything you would use on a gas turbine.

Engineer-Poet said at May 12, 2010 5:13 PM:

Greg F, in pure trollish fashion, decides to belabor the assertion that the grid is a mesh circuit (it mostly is) which nobody disputed instead of addressing his claim that it can be modelled as voltage sources and loads (it can't, because powerplants are not voltage sources), which he attempts to make everyone forget by adding heaps and heaps of verbiage about unrelated matters.  He now adds that it's got a lot of radial (rather than mesh) components, which calls his original "mesh" assertion into question, but nobody was disputing that either.

Give it up.  You're just polluting the blog.

Electrically it's a system of fractional-wavelength transmission lines with relatively few point sources of real and reactive power and much more finely distributed consumers of both.
This is just hand waving. It makes no sense at all. That is why I ignored it.
You mean, your grasp of physics and circuit theory doesn't extend to complex mathematics, energy-storage elements and time delays.  No surprises there.  I could write out the math to describe what I'm talking about in exact terms, but you wouldn't understand that either.
If "no generator in an AC grid can be usefully modeled as a voltage source" there wouldn't be a grid.
How would you know?  A real alternator isn't a voltage source, it has a rather substantial impedance (as do the transformers and lines in the system).  A voltage source has zero impedance by definition.  This shows your utter and complete ignorance of even the simplest elements of DC circuit theory.

AC circuits do things like having phase differences between pieces, and having power transfer across a reactive link peak and go to zero as the phase difference increases.  That's one of the reasons you can't use DC theory to analyze non-trivial AC grids.

But let's get into the meat.

The number I claimed was for industrial compressors. IOW shop air. That is what EP will get....
http://www.mntap.umn.edu/energy/82-CompAir.htm
I was wrong. It isn't 10%, it's 12.5%. Beat me with a stick.
You think so.  I think not.
The issue was not the fluid, but arbitrage (storing cheap off-peak juice, selling at peak).
Moving the goal posts. EP said "I am curious about General Compression" which is a company that is doing compressed air. Not pumped water.

All right, you appear to have 3 issues here.
1.  Compressed-air energy storage is limited to efficiency on the order of 12.5%.
2.  Air compression cannot achieve efficiency anywhere near 80%.
3.  Bonus:  I show you where you mis-interpreted your University of Minnesota reference.

I tell you what.  I bet I can prove you wrong on all 3 points, WITH cites and a thermodynamic analysis.  If you're right on all 3, I pay a $1000 consulting fee and if I prove you wrong on all 3, you pay a $1000 consulting fee; 2/1 splits are worth $250.  Randall Parker is the judge and the agent, and he keeps 10% of any amount which goes through his hands.  All work is posted here, payments shall be made by money orders with the judge's fee separate.

Agreement is void if RP refuses to participate.  RP gets the judge's fee to help pay for the blog, and the work to archive here as permanent added value (either as reference or object lesson in hubris).

If either party fails to accept these terms if RP accepts, or carry through on them after acceptance, they never post here again.

Paying for education is something you should be happy to do.  Let's see you put your money where your mouth is.

Engineer-Poet said at May 12, 2010 7:31 PM:

Rather, you have 2 assertions which I believe I can prove wrong, and I win the bonus if I can show you how you read it wrong.

Engineer-Poet said at May 14, 2010 5:54 AM:

Don't have the courage of your convictions?

Tell you what, I'll add the wind increase with altitude (between 10 and 80 meters) as a double-or-nothing bet.

th said at May 14, 2010 4:18 PM:

This could be the start of something fun and interesting, a project of high school level competence to show strictly compressed air from any source, whether its wind or flywheels driven by cows running in a circle, is viable strictly on its own at 80% efficiency without an outside source of heat such as a gas turbine, isn't this the premise?

Engineer-Poet said at May 15, 2010 8:15 PM:

th, are you willing to take a $250 bet about the straw-man arguments in your 5/14 comment?  It will have to be settled after Greg F either accepts or defaults (I give him until Tuesday to respond).

th said at May 17, 2010 5:11 PM:

before we bet, one parameter is simulations don't count, there's so much to this you govt loving clowns don't get, but we'll see, you set the terms.

Engineer-Poet said at May 19, 2010 5:14 PM:

Greg F, having not returned to this discussion in a week and having not posted an e-mail address for contact, has abandoned the field.  My offer to bet him is withdrawn without prejudice.

th, nobody claimed an overall system efficiency of 80%, but individual components are better.  Three things must be agreed:

  1. total system efficiency is the product of the efficiencies of all process steps,
  2. Isentropic efficiency figures for existing commercial products are sufficient proof, and
  3. What General Compression is claiming is that their figures are feasible with such products.
You in?

Engineer-Poet said at May 26, 2010 5:55 PM:

And now th has been out for a week.  Offer withdrawn also without prejudice.

No longer following this thread.  If you want to return to it, mail me.

Post a comment
Comments:
Name (not anon or anonymous):
Email Address:
URL:
Remember info?

                       
Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©