The Oil Drum | DrumBeat: February 28, 2008

238 comments on DrumBeat: February 28, 2008

Comments can no longer be added to this story.

Show without comments | PDF version

chapterwon on February 28, 2008 - 9:54am

The Reuters article "Loss of wind causes Texas power grid emergency" has a very misleading headline, since the article itself indicates that loss of wind was only a minor cause of the emergency.

From the article:

wind production fell from more than 1,700 megawatts, before the event, to 300 MW when the emergency was declared...

At the time of the emergency, ERCOT demand increased from 31,200 MW to a peak of 35,612 MW, about half the total generating capacity in the region,

1. Supply decreased by 1400 MW. Demand increased 4412 MW. So (Supply - Demand) incrased by 4812. 24% of this was due to loss of wind, and 76% was due to increase in demand. So it would have been much more accurate to title the article "increase in demand led to power emergency."

2. The total demand was half the generating capacity in the region. I understand that the wind can't just be switched back on. But why couldn't the coal plants and natural gas plants be turned back on at a moment's notice? Isn't this supposed to be their strength? I suspect there are extra causes of the emergency lurking here.

Comments can no longer be added to this story.

DaveMart on February 28, 2008 - 10:14am

The point of the piece is that the wind power dropped just when it was most needed.

EON who run a lot of the gird in Germany has been saying for years that keeping the grid running with a high wind input is tricky, and if you plan on the right back up, expensive too.

Your understanding of how the grid works is faulty - nuclear plants are used for baseload, ie they run all the time, as the marginal price of power from them is low.

This can cause additional problems when combined with wind power, as when the wind blows, which it doesn't all the time you obviously need to use it.

Coal and gas plants too can't be switched on at the drop of a hat, unless they are already turning over, in which case they are burning some fuel and form part of what they call the spinning capacity, which can be switched on very quickly but runs at reduced efficiency for the first half hour or so.

IOW you need fairly substantial back-up if you are using wind power, and keeping that ready to fire up costs a lot of fuel.

Comments can no longer be added to this story.

chapterwon on February 28, 2008 - 11:00am

My point was that one cause-- loss of wind-- was being singled out for blame for an emergency that clearly had multiple causes.

The biggest cause was the sudden unexpected spike in demand due to colder temperatures moving into the state. From the numbers in the article, it seems that *if wind hadn't dropped*, this alone would have caused a stage 1 emergency. Why was this the temperature drop and power spike unexpected? Large changes in temp can typically be predicted hours, even days in advance.

The article also says "multiple power suppliers fell below the amount of power they were scheduled to produce on Tuesday." Details, please, Reuters! How did the magnitude of these shortfalls compare to the magnitude of the wind shortfall? How many suppliers fell short? Why? What was their power source?

Given that a large spike in power was (or should have been) expected, and that multiple (non-wind) power suppliers were falling short, why wasn't standby backup increased, even at the cost of wasting fuel? When you have some problems, and more are on the way, that's the time to increase your safety margin.

And why wasn't the wind shortfall predicted? I'm no meteorologist, but it would seem to me that a large decrease in wind over a large area (i.e., movement of fronts) should be predictable on the timescale of hours.

So I think that the bottom line is that yes, a wind shortfall did cause problems. But the emergency happened because a lot of things went wrong; the drop in the wind was only a (relatively minor) cause.

Comments can no longer be added to this story.

DaveMart on February 28, 2008 - 11:16am

I don't seek to write wind off on the basis of one outage, but the problems are rather tougher than for some other sources.

Fossil fuel or nuclear can usually be just scheduled for maintenance, and taken off line at a time known way in advance, although of course you can have the occasional breakdown, but wind is by it's nature variable, so is tougher to balance all the time, and the back up needed for any given level of security of supply would tend to be higher, although by no means one for one.

German and Danish infrastructure is 'probably' usually run at higher safety margins than is common in America, just like the highways, so a higher penetration of wind may be easier than in most places in the states.

With the typical low-levels of infrastructure investment in the States it seems likely that there could be more frequent problems where wind increases it's share of generation much.

Comments can no longer be added to this story.

chapterwon on February 28, 2008 - 1:01pm

My original comment had been concerned with bias in the Reuters article, not the broader issue of suitability of wind. But the latter is interesting to talk about.

Your post perpetuates a common myth: "Fossil fuel or nuclear can usually be just scheduled for maintenance, and taken off line at a time known way in advance, although of course you can have the occasional breakdown"

Unscheduled maintenance is just a serious problem for fossil (at least coal). According to this study

because coal plants were shut down for scheduled maintenance 6.5% of the year and unscheduled maintenance or forced outage for another 6% of the year on average in the United States from 2000 to 2004, coal energy from a given plant is guaranteed only 87.5% of the year, with a typical range of 79%–92% (North American Electric Reliability Council 2005; Giebel 2000)

I don't know the comparable figures for nuclear, but remember that when nuclear unexpectedly goes down, a lot goes down all at once. Witness the power outages in florida earlier this week, and similar unscheduled shutdowns in Spain and Japan last year.

As the study cited above notes, simply by linking geographically diverse wind turbines, one can achieve a "baseload supply" (i.e., availability comparable to coal) of one third of nameplate wind capacity. Again, that's without any form of storage.

You are right that infrastructure investments are necessary. But investments are necessary to build the wind turbines in the first place. Think of the extra infrastructure as part of the cost.

Comments can no longer be added to this story.

xeroid on February 28, 2008 - 1:44pm

As I always say ... 'watch and learn' - as we inevitably move to more and more alternatives to FF expect instability and intermittent power, grids are complex machines - this is how many parts of the world have to work already, individuals in those countries plan so as to mitigate it's effects ... you can too!

Comments can no longer be added to this story.

DaveMart on February 28, 2008 - 3:03pm

I don't want to come across as some sort of anti-wind loon - I try to judge every suggestion on it's own merits.
In fact, for the UK at least, variability may be less of a problem than is currently realised:
http://www.eci.ox.ac.uk/publications/downloads/sinden05-dtiwindreport.pdf
sinden05-dtiwindreport.pdf

However, that does not mean that the variability, both over the short-term and the rather longer term will not cause any problems, and particularly in the US with it's history of minimal investment in infrastructure that may cause problems.

It is up to you to make your own judgement, but it seems to me that a high level of penetration of wind power in the States might cause more problems than a similar level in Europe.

Comments can no longer be added to this story.

Robert Marston on February 28, 2008 - 3:29pm

With wind having a rather high EROI already, has anyone done an HONEST feasibility study on what the costs and options for storage might be? Pressurized air wind energy storage? Heat?

Given wind's strong advantages in other areas I don't think it would be all too difficult to mitigate the intermittency issue with a little planning and forethought.

With these renewables it just looks like we're going have to build some storage capacity. I don't necessarily see that as a bad thing. Just part of the solution.

Comments can no longer be added to this story.

CalGuy on February 28, 2008 - 3:45pm

I favor pumped hydro as cost effective. People will say that you have to have the land for it and not everyone has the vertical rise. Even in the flat lands of Texas, there could be enough of a plateau for this. It is pretty efficient and can provide water storage for crops and homes.

Comments can no longer be added to this story.

DaveMart on February 29, 2008 - 9:21am

There was a recent study in Scientific American which sought to show that it would be possible to generate all the power for the US using solar energy, and transmitting it as needed from the South-West to other areas.

Amongst other issues discovered was the proposed storage mechanism to make up for overnight capacity.

They wanted to use compressed air.

The problem is with that is when you come to use it you have to re-heat it, and they were going to use natural gas for the purpose.

The burn would have been huge, if my memory is correct much larger than current gas use.

Other storage proposals run into similar problems.

It boils down to that unless you are very lucky and have access to hydroelectric capacity, as Denmark does from Scandanavia, then effectively you have to have large amounts of FF capacity to back-up, and that a proportion of it will be spinning capacity, ie fired up but idling, and burning fuel in the process.
When it is brought into action it is then much less efficient for the first half hour or so.

You might also perhaps build vast polders in the ocean or great lakes, and pump water in them, but the costs would be added to the already large costs of windpower, and so are pretty impractical for the foreseeable future.

Recent costs for the UK's proposed 33GW nameplate off-shore wind build were given as £66bn - and you only get around 10-11GW of power per hour from that on average, so it works out at around £6.6bn GW - hugely expensive without building fancy storage, and at least twice the price of a nuclear build.

You can reduce that of course if you have somewhere to put the turbines on land with good wind resources, as is the case in many areas of the States, but it is still pricey before you start building huge amounts of storage.

T Boone Pickens is currently building the biggest wind farm in the world in Texas, 4GW nameplate for $10bn.

The problem is of course that the actual energy flow on average per hour on a generous figure of 35% average capacity is only around 1.4GW.

Let's round that up for convenience and to be very fair and call it 1.5GW.

That is $6.6bn GW - about half the cost of off-shore but still dear and around the price of a nuclear build which would not suffer from the same need for storage at high rates of penetration.

So the wind option can be useful, particularly where it tracks well with peak use, but is a very long way indeed from being able to power most of the grid at any reasonable cost.

Comments can no longer be added to this story.

Joseph Somsel on February 28, 2008 - 8:37pm

There are two concepts intermingled here. First is capacity factor. That is just the total electrical production over a year divided by the nameplate rating x the hours in the year. The US nuclear fleet is hitting over 90% regularly. The initial design economic assumptions for these large nukes was 70% or 80% so they are doing really well.

The second is forced outage rate. That's when unplanned shutdowns occur. It used to be that a nuke would scram once or twice a year. Scrams seldom happen anymore. Most US nukes startup and run 18 months without hiccup or trip at 100% capacity.

They do have to be shutdown once in a while to remove old fuel and add fresh. Most plants do this over an 18 month fuel cycle. Refueling takes from 12 to 30 days depending on other work underway and the specific plant design features. New nukes are supposed to do 17 days with all required maintenance.

Coal plants have the capacity factor mentioned above but since they are fueled continuously, they don't need to be suhtdown for refueling. Hence, almost all of their downtime (but not all) is due to forced outages. Typically a boiler tube will blow or a burner with clog up, stuff like that.

We nukes call coal plants "dirt burners."

Your cited claim that wind power can claim 33% capacity is incorrect and is academic wishful thinking. In Texas, for example, ERCOT only allows 8% credit and that's generous.

Comments can no longer be added to this story.

jokuhl on February 28, 2008 - 1:34pm

Chapterwon and Dave;
I agree that the emphasis was unnecessarily harsh on windpower, which we know will be a variable supply. My contention is that a great portion of the fault lies with our system being built on the 'Assumption of Steady, Continuous Energy' Petroleum has fed us this myth for decades, and we are looking to all the other sources to back up that promise.

Petroleum and Gas will prove their intermittency soon enough, and in the rearview mirror of history it will look like One On, and then One Off.. Solar and Wind go away, but they keep coming back. Nuclear looks sort of steady from this 5 decade window at the height of Petroleum's abundance, but I am very skeptical that it can survive in a world without such a meaty petroleum backup supply. I've said it before.. I think Nuclear DEMANDS as much of a 'Baseload' of energy around it as it seems to offer.

The core problem is that we need to feed our towns, homes and businesses like anything else in the natural world eats.. you have to store energy and be able to ride out periods of scarcity. Life doesn't operate on this 'baseload' fantasy, we've just been able to mimic that euclidean ideal concept for a little while since we found that motherlode. It has been a fun little mask that we wore to pretend that we aren't children of 'Mother Nature'.. she has never even been that far away, we just acted like we were Astronauts for a spell.

Bob Fiske

Comments can no longer be added to this story.

Robert Marston on February 28, 2008 - 3:35pm

Well said sir. And at 90 billion per major plant nuclear does seem to be a very intense form of energy. I'm not completely adverse to nuclear the way some are, though. I could be wrong. But as for wind and solar, I'm in complete agreement. We need to build the storage.

Comments can no longer be added to this story.

joule on February 28, 2008 - 11:13am

DaveMart -

It is becoming increasingly apparent that the more that wind power increases as a percentage of the total power fed to a particular grid, the more we are going have serious problems with maintaining a stable power supply, such as the recent event in Texas.

Gas turbine back-up held in a constant state of standby is a solution, albeit an expensive one, both in capital investment and operating cost.

A more elegant solution would be some sort of a system of 'super capacitors' that could store something like an hour's worth of maximum wind farm output and then almost instantaneously begin releasing that power in the event of a sudden drop in wind power. This concept could be further refined to also include a bank of empty capacitors designed to take an hour's worth of excess wind power during short periods when too much wind power is being generated. The stored electricity could then be released back into the grid slowly as needed. Of course, large capacitors are very expensive, and more development work would be needed to achieve large inexpensive capacitors. While large batteries would also work, they wouldn't quite have the quick response of a capacitor, particularly during the charging part of the cycle.

So, essentially what is needed to make wind power more user friendly is an energy 'flywheel' to help smooth out the peaks and valleys. (In fact, actual mechanical flywheels have at various times been investigated for energy storage, and like everything else, they have their own set of pros and cons.) But as things stand right now, it appears that once wind power becomes even a rather small fraction of a grid's total power input, serious stability problems develop.

Comments can no longer be added to this story.

jokuhl on February 28, 2008 - 1:44pm

Can anyone comment on the status of Flywheel Storage at this point?

I keep wondering if the limitation is that we try to load them up to max capacity, introducing bearing, centrifugal stress and stability issues, if it isn't reasonable to 'Undercrank' them, and run more of them instead.. ?

Seems like a very simple and mass-producible storage system.. but I don't know what their Achilles' Heel is currently.

Bob

Comments can no longer be added to this story.

joule on February 28, 2008 - 2:29pm

jokuhl -

There was a great deal of work done on flywheel energy storage in the 1970s and early 1980s, however I haven't heard of all that much work being done on this subject since that general time frame. There are a number of companies that make flywheels for energy storage, but these appear to be for much smaller applications than large power generating systems.

The modern energy storage flywheel typically runs at very high speeds, has a carbon fiber rotor, is enclosed in a vacuum chamber to elminate air drag, and has special high-tech bearings. The high speed is essential, as the amount of energy stored is proportional to the square of the rotational speed. These can be scaled up in size, but only to a certain point, and then some serious design constraints come into play. Thus, if you were to attempt flywheel storage for a large wind farm, you would probably need several times the number of flywheels as you have turbines, and that could get pretty expensive. But honestly, I no longer have a feel for what the current economics are.

Maybe for really large applications, a larger but lower-tech flywheel (as you suggested) might have some merit. Still, it's going to take a huge spinning mass to store even a few minutes worth of the energy produced by a single large wind turbine.

Comments can no longer be added to this story.

mcgurme on February 28, 2008 - 2:55pm

Beacon Power specifically designs large flywheel systems for grid load management. http://www.beaconpower.com/products/EnergyStorageSystems/flywheels.htm

It appears they are just in the stages of testing these technologies in real life.

Comments can no longer be added to this story.

xeroid on February 28, 2008 - 2:52pm

I don't know what their Achilles' Heel is currently

As with any dense way of storing energy the problem is the failure mode!

Gas typically explodes, a serious failure mode - you don't want to be sitting in a house which has a gas leak when somebody switches on a light switch that makes a spark!

Gasoline typically burns fiercely - so is somewhat safer than gas.

Flywheels tend to leave the building in a marked manner! - massive armour plating is required, also they're not really suitable for a mobile application eg: aircraft with powerful rotating engine components can do strange/frightening things when close to the ground.

However flywheels are good for storing large amounts of power required at very short notice - they are often used in particle accelerators such as at CERN.

snip; The PS's rotating machine comprises a motor coupled to a generator. The generator's rotor acts like a flywheel, supplying high-power pulses of 40 to 50 megawatts to the PS magnets. The 6 megawatt motor drives the installation at 1000 revolutions per minute and compensates only for variations in speed. /snip http://bulletin.cern.ch/eng/articles.php?bullno=23/2006&base=art

Comments can no longer be added to this story.

DaveMart on February 29, 2008 - 9:30am

I don't see short duration fluctuations as being a major problem for wind power.

We already have batteries available which serve for this sort of purpose, although one utility got caught out recently when their supplier ceased trading, so it was left with several million dollars worth of batteries which had proved unreliable.

Ultracapacitors perhaps in conjunction with a battery would also help.

All this adds to the already high costs of wind though.

Comments can no longer be added to this story.

Joseph Somsel on February 28, 2008 - 8:57pm

One should remember that energy storage will always lever cost and reliability differences in inputs. That means that it is imperative to use the lowest cost, most reliable electrical inputs to energy storage. Use of expensive power inputs will result in disporportionaltely higher output costs.

Remember, that any storage is less than perfect. More goes in than comes out. The overhead capital and O&M cost of the storage facility means that the input source must be reliable too. There are only certain times when charging is feasible and if the source is not available, then the cost per unit output goes up as the overhead is spread over a smaller number of output units.

That means that nuclear or coal are the better sources for storage since they are 1) cheaper, and 2) more reliable. Solar and wind would just make more expensive stored energy.

Storage works AGAINST renewables, not for them.

Comments can no longer be added to this story.

WNC Observer on February 28, 2008 - 12:12pm

Obviously, some sort of buffering is needed. Wind has worked great for pumping water up into a tank. Make the tank big enough, and no wind for a little while doesn't matter. Various suggestions have been made for pumped storage schemes or to use wind power for hydrogen electrolysis; one or the other of these will have to be incorporated for wind to work on a long-term basis.

Comments can no longer be added to this story.

Starship Trooper on February 28, 2008 - 12:29pm

Actually, simple cycle CTs can be "turned on" at a drop of a hat, but they are about the only thing that can, particularly the aero-derivative units.

They cannot, however, be turned on and put into "pre-mix" operation. It takes time to heat the cans, stabilize, and to switch over to from diffusion to pre-mix mode. The newest units I've worked with have the ability to have generator turning as a motor (already synchronized) and the turbine "clutched" and not "spinning" until the start-cycle is initiated. But they still can only be started in diffusion flame mode with lots of NOx and lots of excess air to keep from melting components in the expansion section.

You are correct that standard fossil fuel-fired units cannot be started immediately and really even the ones that are operating at less than full power have maybe an immediate 5-10 MW "all call" added capacity with the remaining available capacity reserve capacity governed by the refractory heating rates and how fast you can increase the heat transfer through the various sections of the boiler. Other than their inherent efficiency advantage (and the fun of starting them up), I'm not aware of any particular advantage that a supercritical unit would have a subcritical one or visa versa.

As for nuclear units...slow to respond and there are also grid balance issues associated with their operation (in addition to NRC requirements associated around the loading of these units relative to the grid). The sort of situation desribed above makes balancing just a real joy (I note that my tongue is firmly planted in cheek).

I agree with you. It isn't as easy as it seems (or as simple as turning on a switch).

Comments can no longer be added to this story.

Joseph Somsel on February 28, 2008 - 8:47pm

Let me give you some hard numbers on gas turbine startup times. For a distillate fueled aeroderiviative, a 9 MWe unit needs 2 minutes. A 25 MW unit needs about 10 minutes. Now this is for emergency duty so proper operations would take it slower.

Nukes can load follow. The last one I helped build could swing from 25% to 100% back to 25% over about an hour, on autopilot and that's with all control rods out. The economics of large capital cost, low fuel cost generation encourages one to run them flat out all the time but one can only put as much power into a grid as the customers want to take out (barring storage.)

To bring a large coal or nuclear unit online from cold shutdown takes 12 to 24 hours or more (more if you can to avoid metal fatigue issues.)

Comments can no longer be added to this story.