November 16, 2008
Variable Electric Generator Improves Wind Power Efficiency

An interesting article in MIT's Technology Review reports on a generator for wind turbines that can harvest electric power over a wider range of wind speeds.

ExRo Technologies, a startup based in Vancouver, BC, has developed a new kind of generator that's well suited to harvesting energy from wind. It could lower the cost of wind turbines while increasing their power output by 50 percent.

The new generator runs efficiently over a wider range of conditions than conventional generators do. When the shaft running through an ordinary generator is turning at the optimal rate, more than 90 percent of its energy can be converted into electricity. But if it speeds up or slows down, the generator's efficiency drops dramatically.

If the cost delta for this generator design is small enough then the 50% boost in electric power could greatly improve the economics of wind electric power. The generator works by switching in more magnetic coils as turbine speed increases.

ExRo's new design replaces a mechanical transmission with what amounts to an electronic one. That increases the range of wind speeds at which it can operate efficiently and makes it more responsive to sudden gusts and lulls.

Here's the company's description of their Variable Input Electric Generator (VIEG):

Rather than layering individual legacy machines one on top of the next, the VIEG uses a series of coils, configured in "balanced stages".The magnetic balancing allows the use of permanent magnets, yet still reduces cogging torque to a bare minimum, which allows the VIEG to operate at extremely low wind speeds (near zero).As available energy increases, the VIEG matches generator resistance to source energy by electronically adding generator stages. Conversely, the VIEG is able to drop stages as available energy (wind speed) drops, cycling up and down without hesitation and without mechanical friction.The need for a gearbox is eliminated, and a single VIEG generator scales up and down with available energy in a way that would take almost 70 individual generators to match.

My guess is that wipespread deployment of this generator would also reduce the problem of wind intermittency since wind power wouldn't drop off suddenly below a threshold. The power supplied would decay more slowly. This would tend to make load balancing easier I would think. Anyone know if this intuition is correct?

Share |      Randall Parker, 2008 November 16 07:59 AM  Energy Wind

Volkman said at November 16, 2008 4:04 PM:

Wind power is inherently unreliable. This cute little innovation may eventually help in a miniscule way, but without large scale storage, load leveling for wind will be hellish. Even worse you still need coal or gas backup. That's no way to run a utility.

Fat Man said at November 16, 2008 9:30 PM:

It still won't generate electricity when the wind speed is zero, or even very much when it is 3 mph. I still want to see the costs for a full system with transmission and storage.

Jerry Martinson said at November 16, 2008 10:41 PM:

Has anyone (i.e. EPRI, DoE, or something) published anything showing how much of the grid can be powered with wind before there's problems with low-wind brownouts or having to shut down aluminum smelters, papermills, etc... How much investment is needed in transmission power lines to reap the benefits of geographic diversity so that the dips and valleys of generation capacity is averaged out more?

It would seem that this is central to the question of how much duplicate generation capacity and transmission investment is required to get more than today's few percent generated by wind power. Innovations such as this one might be much more valuable than they first appear when looked at from this perspective as it might reduce the amount of duplicate capacity required. Since the market is a dirty mix of regulation and free-market pricing signals, I'm not sure how things like this get priced correctly so that the invisible hand guides investment in the right direction. Too bad all the people who are skilled in the complicated Monte Carlo/Markov simulation techniques needed to price this stuff right all apparently work on the Wall Street arbitrage casino on other problems.

Nick G said at November 17, 2008 10:55 AM:

Well, power is the cube of wind speed, so power would still drop off fairly quickly with falling wind speed.

On the other hand, if a 50% increase in power comes from improved efficiency at lower speeds, than this would have to reduce output variance. OTOH, if it comes from improved efficiency at higher speeds (i.e., if the current sweet spot for generator efficency is significantly below maximum wind speed), then it might increase variance.

Either way, 50% increase in power is always a good thing. At worst, there would be a very small % of times when a windfarm's peak output would be lost - not really a big deal.

Intermittency is exaggerated as a problem: 1) geographical diversity greatly reduces it (for example, one wind turbine can easily reach 100% of rated power, but a wind farm almost never goes over 85% of rated power). 2) Solar is negatively correlated with wind. 3) demand management will solve most of this problem, especially given that EV/PHEV/ErEVs (whose dynamically schedulable charging can buffer supply variance, even without V2G) are going to grow at roughly the same speed as wind power.

I wonder about the status of the turbine blade design based on whale fins? That was also supposed to increase efficiency, and it looked extremely promising.

I think we're far from diminishing returns for wind power tech.

Shay said at November 17, 2008 2:21 PM:

There has also been a recent discovery of the stabilizing effect that whale "lumps" provide to turbines. These mimic the leading edge lumps on whale fins and flukes. I recall it lowered the stall speed of the turbines by a large margin.

Shawn Levasseur said at November 17, 2008 6:28 PM:

Does this make smaller scale wind power more or less practical? (By smaller scale, I mean having wind turbines on smaller hills or on top of buildings)

Any developments on the load balancing front?

B Dubya said at November 18, 2008 9:08 AM:

"The generator works by switching in more magnetic coils as turbine speed increases."
Why do I think you got this backwards?
In a synchronous generator, the rotational speed to make a voltage is directly proprtional to the number of pole pairs.
So...if this thing is to work, as the turbine speed decreases, pole pairs would be switched in ( I am assuming through some sort of static switch arrangment, such as a thyristor trigggered ny RPM and V(out) signals)), field exitation would increase and voltage would remain within a relatively narrow band.
But P(out)=P(in)-P(losses). If the turbine speed is a function of wind energy (it is), P(out) will decline regardless of the number of poles. All the pole switching appears to do is sustain voltage output, with an additional cost interms of increased excitation current to the generator field (to maintain stator voltage). I think you would need some sort of auto kickout relay scheme to prevent reversing power as the generator voltage drops when wind speed (and energy) is too low to maintain voltage. In a DC system, dropping voltage sheds load, until you eventually motorize the low voltage generator.
At any rate, the greater the number of pole pairs in the generator, the lower the required RPM is to produce any given voltage, because it's relative motion between the conductors that generates the voltage (V=K*Flux*N, where K is a constant describing the generator winding characteristic that includes pole pairs, flux strength describes the relative power of the magnets (also affected by pole pairs), and N is speed in RPM).

For wind to work, even as a supplimental power source, I don't see how you can operate except in a DC mode, floating on a big (large capacity, high voltage) battery in order to smooth out the power dips and surges, with a big old solid state inverter to interface with the AC distribution system, which is where we all get our house power.

Me? I'm going with the fission of heavy isotopes to make steam. Or better yet, He3 fusion.

Knot said at November 18, 2008 2:01 PM:

RE: intuition - I don't think it will make a huge difference in the power dropoff / threshold.
Unless being paid by the .gov, utilities only locate these things where there is good wind
most of the time anyway. (Wyoming plains, Seacoast w/ steady on (or off) shore winds, etc.)

As for many of the above comments. . .
The efficiency gained is soley a result of better torque vs. power curve matching, so that
power output is maximized. This is a physical implementation of Maximum Power Point matching
for solar cells, or . . . impedance matching transformers for your car speakers. The short
explanation is that even a 90% efficient electric generator isn't 90% efficient if it is ran
at the wrong speed. That is why old farmer windmills aren't used commercially for wind power,
even though it is an existing, cheaper technology.
(No Off-grid automotive alternator conversion anecdotes, please)
The primary benefit of this technology is that it makes new and existing installations more efficient,
so we have less capital expenditure for each mW/Hr produced.
And, while load balancing will occur due to diversity just as power consumption normalized according
to load diversity, . . . I agree with Volkman in the first comment.

Unless we expose ourselves to HUGE loss of reliability due to natural disaster, (hail storm on the solar farm)
(Still day in Wyoming. . . It DOES happen sometimes), the coal fired plants will have to up from a
five percent spinning reserve to five percent plus a significant percent of the total wind & solar
on the base load, in order to take up the load in the event of outage.
Summary: Unless we (or our government) is willing to have more frequent outages and cascading blackouts, we
will have to burn more coal the more wind and solar we have on our grid.

Those with more practical grid experience and diversity modelling expertise are encouraged to tell me
why I'm wrong.

Nick G said at November 18, 2008 3:24 PM:

"HUGE loss of reliability due to natural disaster, (hail storm on the solar farm)"

Are you suggesting that solar farms are vulnerable to major damage from hail? They're designed to withstand hail. Do you have information to the contrary?

"we will have to burn more coal the more wind and solar we have on our grid."

I think in a grid with a large % of wind (greater than 20%) some coal plants might be needed on standby, ready to power up at, say, 12 hours notice, when wind shortfalls were predicted. How much coal is burned by a coal plant on standby? How does that compare to the coal burned at full nameplate output? Is it more than 2-3%?

Knot said at November 19, 2008 8:22 AM:


In inverse order. . . And I don't intend to be pedantic, so forgive me if my writing comes out that way.

Power plants work by running steam through turbine connected to generator. Since the connected grid may
drop a power plant at any time, the grid authority (WAPA here in southwest) requires each power station to
maintain what is called 'spinning reserve' of 5%. All the different powerplants carrying 5% extra capacity
have to come on line instantly during the sag to make up for the dropped plant. The mechanical implementation of
this is the voltage sags and the steam valves throttling the steam open up until either the voltage recovers or
the valve is fully open.
Summary: All powerplants have 5% more steam than they need to fulfill instantaneous demand of their regular service area.
Further note: The powerplants are most efficient running at what you are calling full nameplate output. That is where
the least losses due to throttling are going to happen. So, local peaking is almost always done with gas turbines, which
can be up in 30 seconds. However, they are capital intense, and not the appropriate technology to supply the base load.
(And, you have just put a demand loading on the natural gas piping system)

You can't (fully) fire a powerplant in 12 hours. Not happening ever. Think about a basic control volume with that
much thermal mass. The kind of power input for a 12 hour rise time would require some pretty extreme heat input. . .
which would require some pretty extreme materials. I think even steam powered (non-nuclear) ships required 36 hour
orders to make weigh for the engineering crew, so they can start to heat things up ahead of time.
Summary: While you don't have to be running flat out, you can't make instant steam. So, YES it sure as heck is more than 2-3%
Note: Do the calculations for 2-3% of a coal fired powerplant. Translate to solar or wind farm capacity. Think about it.

Diversity of supply is what offsets the need for spinning reserve.

RE: solar cell susceptibility to hail. Are you kidding? Or do you live somewhere where you don't routinely see golfball to
softball sized hail mixed with tornados every summer during afternoon heating thunderstorms? I don't have specific info
on specific farms, but. . .
I routinely price check for putting PV solar on my house, and have designed and redesigned the system many times. In my research
for this, it has been (repeatedly) pointed out that the (current) solar cells are pretty brittle. They have some thin protection, but
anything that is going to put one inch dents in automotive sheet metal is likely to either punch through the protection or vibrate
the frame and substrate so hard that the brittle material is going to be subject to fatigue over the long run. And the cells are so
expensive that you really need them to last forever. They don't repay capital cost for residential grid-connected service for 32 years.
The best warranties are for 20 years. Doesn't matter for the reliability argument.
The capacity to produce soley from the gray radiation during a severe storm
is still equivelent to putting 75% of the solar cells out of service.

Brian H said at November 19, 2008 2:45 PM:

B Dubya;
About the variable speed issue: It is my impression that kicking in more coils as wind speed increases would smoothly increase resistance, so that turbine speed would remain constant within a narrow band.

My money is on 's Focus Fusion reactor-generators: small 5MW $250,000 home-garage sized installations, using pulsed aneutronic proton-boron fusion. Output pricing expected to be in the KWH range. Time to working production model estimated at 5-8 yrs.

Economics alone will render virtually all other generating models and plants obsolete and unsustainable.

Nick G said at November 19, 2008 4:41 PM:


How much power does maintaining that 5% spinning reserver cost to maintain? My understanding of steam systems is that a certain amount of power is needed to produce the steam pressure, and then a small % is needed to replace heat leakage and maintain pressure, but that a 5% reserve doesn't increase coal consumption by 5%. I would expect that it would increase it by much less, perhaps .2% (4% of 5%).

Keep in mind, the purpose of a fullscale renewable buildout would be to replace coal production - that means you'll have a lot of underutilized coal plants around. There's a cost to that, but it's unavoidable if you want to reduce coal electrical generation. Given that, the cost of backup for renewables is only the additional overhead due to the cost of plant operations & maintenance being spread over fewer KWHs in underutilized coal plants, and the additional fuel cost of running at lower thermal efficiencies due to running at lower temperatures.

How long would it take to raise power output from a coal plant from 50% of capacity (and close to optimal temps), to 100%? more than 12 hours?

RE golfball sized hail: U/L certification requires the ability to withstand that. Softball sized hail? That's very rare in most places. Heck, that would break windows - how often does the average homeowner have to replace windows due to hail??

" local peaking is almost always done with gas turbines, which can be up in 30 seconds. However, they are capital intense"

My sense is that they usually cost about 4-500/KW capacity. That doesn't seem too bad. Again, keep in mind that we're hoping to displace fossil fueled generation (heck, that's the heart of Picken's plan): that's going to leave a lot of underutilized NG generating capacity.

"Do the calculations for 2-3% of a coal fired powerplant. Translate to solar or wind farm capacity."

I'm not sure what you mean. 2-3% of a 400MW coal plant would be 8-12MW. That would require 24-36MW of wind name-plate capacity, on average. So, having figured that out...what are you getting at? While you're thinking about that, keep in mind that you don't have to provide 100% backup, as geographical diversity, weather forecasting, and demand management will greatly reduce the % of backup required. In other words, backing up a wind farm is very different from backing up a big coal or nuclear plant, which can go down very, very quickly.

Knot said at November 19, 2008 6:30 PM:

Nick G

I'll make this my last comment, unless you crank open something relevant on your site.
I don't want to take more of futurepundit's bandwidth for my long comments.
To build on the positive, we both agree about the need for generation diversity.
The problem is that as we build smaller and smaller (and more diverse) generation
capacity, our total expenditures go up. Restated, we are unable to take
advantage of economies of scale when we make anything smaller. The result
is that the $/watt number goes up.

The 2-3% comment was intended to point out how huge the spinning reserve is.
For our local plant, the 2% number comes out to 5.4 MW. That is a lot of power (profit).
The power industry considers fractional percentage points in efficiency gains to be
huge profit or loss makers.

RE: Storms. . . In both Texas and Oklahoma (500 miles apart) we had to replace home AND car
windows due to hail. We were happy both times, as the neighboring areas were hit by tornados.
We got off light. But you are generally right in that I guess most of America doesn't have
this weather as regularly as I've seen it in my life.

RE: Reliability -> Any power generation capacity can drop instantly. (Think joyriding a backhoe)
The reasons for the drop can be various typical cascade stuff internally, but all the systems I've
seen (From the industrial high reliability consumption side of things, incidentally) can drop off
the mains equally quick. The reason is that we don't want to kill our workers. Certain types of
faults are always going to take the system to a safe and de-energized state as quickly as possible.
While the wind & solar stuff won't have the issue of false alarming an emergency steam dump they
are equally susceptible to other internal faults that cause loss of service. Another issue to consider
is that power stations have multiple parallel feeders. Environmentalists hate these. Locally,
large wind farms are criticized on the basis of having to have even a single set of overhead wires
feeding them. Reliability suffers for obvious reasons without parallel feeders with
geographic diversity.

RE: Transfer function for coal usage to spinning reserve. You are correct that it isn't linear,
but don't forget that if we are keeping 13.5 MW spinning reserve, we have to be burning coal and
generating 13.5MW of steam. (Ignore system losses) I don't know the exact numbers, but I am guessing
you can back them out pretty easy. For my numbers from my local plant, efficiency is 42%. That was
the highest in the country when it was built. Use 38% efficiency for a baseline of 2% of load.
5.4/.038 =~14MW
Coal @ $105 / ton (Coal is tied to price of diesel due to delivery by train & misc. heavy equipment)
33 MJ/kg heat content for coal
I figure about 1 lb/sec required for 14 MW of power (round numbers).
{1.68 ton / hr * $105 /ton = $176 /hr). This is fuel cost only. No
equipment amortization for {onsite delivery, pulverizing/infeed nor
flue recovery costs}. This is back of the envelope estimate
for a 5.4 MW spinning reserve, 2% of 270 MW. Kind of fun to do some of that engineering
stuff we studied so hard at so long ago!

I read your page and will check up for a relevant post if you want to continue the dialog, but. . .
I am agnostic regarding anthropogenic global temperature effects. Also not a huge believer in
the whole Peak Oil movement. I think that we need to move from oil to renewable not so much because
we *have* to, but because it is for the best. If something can be efficient, than we should strive
to be as efficient as we can. Waste not, want not.

Besides, if we are going to take over the Universe, we are going to need to get this energy thing
figured out first.

Bilderschmidt said at June 4, 2014 6:50 PM:

Large wind farms cannot survive without massive government subsidies. Compare the subsidy per watt for wind vs. for nuclear. No comparison when you get down to the energy unit. The subsidy for wind is around 2.2 cents per kw hr. or more. A lot more if you include the subsidies to developers and the mandates for utilities to use wind first, and keep fossil plants running as backup.

Most government subsidy for fossil fuels goes to poor consumers to heat their homes in the winter, according to Robert Rapier, who appears quite trustworthy.

Wind promoters are either in on the take from government subsidy, or they are dilletante enthusiasts whose hands-on knowledge of energy and power is minimal to non-existent.

Nick G said at June 7, 2014 3:20 PM:


Wind power is cheaper than new coal, without subsidies.

The wind subsidy is only about 1 cent: - it's a 2.3 cents per kWh tax credit for the first 10 years of a 25 year life.

Sadly, Robert Rapier has mastered the art of selective reporting of information. The largest subsidies for fossil fuels are indirect. Coal has many external costs: occupational health costs, CO2, sulfur/acid rain, mercury in food, water consumption, adding up to $.18 per kWh ($345B/year):

"The United States' reliance on coal to generate almost half of its electricity, costs the economy about $345 billion a year in hidden expenses not borne by miners or utilities, including health problems in mining communities and pollution around power plants, a study found.

Those costs would effectively triple the price of electricity produced by coal-fired plants, which are prevalent in part due to their low cost of operation, the study led by a Harvard University researcher found.

"This is not borne by the coal industry, this is borne by us, in our taxes," said Paul Epstein, a Harvard Medical School instructor and the associate director of its Center for Health and the Global Environment, the study's lead author.

"The public cost is far greater than the cost of the coal itself. The impacts of this industry go way beyond just lighting our lights." or

or here's the link to the pay-wall-protected version if desired:

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

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