Flywheel based electrical storage

Re: Flywheel based electrical storage

The reason that I "guessed" how much power running at 100% full standby is because that was the only spec. I could find.

Because many of the frictional losses may be related to the square of the rotational velocity and others are linear--it is difficult to say how long the flywheel may effectively store power.

If we assume that 8% standby loss is proportional to the rotational velocity--you could guess that the amount of power remaining at any point in time might be (as an exponential decay):

100%*(0.92^t)= % energy remaining in t hours...

92% at 1 hour
85% at 2 hours
66% at 5 hours
43% at 10 hours
19% at 20 hours
1.5% at 50 hours

Don't have a clue if the above is accurate or not--but it is probably as close as you will come to comparing the energy storage to a "battery" without calling Beacon up and getting more specifications.

And, after reading the specs. more closely--I believe that the 8%/12.5 hour number is correct. This is a typical method used by marketing to bury the negatives of a system (use % power losses without clearly stating any real numbers and spec system out to 30 minutes of use--but define power output as a maximum amount over 15). Not quite deceptive--but defining strengths of a system in the best ways possible.

-Bill

Re: Flywheel based electrical storage

I am going by 2% loss based on maximum power rating--this is a rate, so you just multiply by hours if you want kWhr losses:

2% standby losses of rated power * 20MW (15 minute peak rating) = 0.4 MW or 400 kWatt required to keep the flywheel spinning at 100% speed? At $0.10 per kWhr--that is $4 per hour standby losses.

Sounds really high--but this is for a 200 "unit" installation. So the cost per flywheel is 2kW per flywheel ($0.20 per per hour per flywheel) for a 100kW peak in/out per unit. Still a fair amount of heat being released (wasted energy).



-Bill

But, I should also add that we "budget" 10-20% loss, per cycle for AGM/Flooded cell lead acid battery losses. If you are cycling a lot of power to even out (like from a wind farm) over short periods--the losses from a flywheel system may end up being a lot less.

Re: Flywheel based electrical storage

MarkP wrote: »

I have read most of the links provided about these flywheel storage units and I give up trying to decipher their "efficiency". It seems almost intentionally ambiguous.

Yep, when I made many of those posts, it was very unclear how they calculated efficiency--But I was pretty sure I understood the model and the actual losses were pretty high.

If your storage battery went "flat" in hours or a day--You would not be very happy.

And that is the issue, the larger the system, the more power it "burns" to keep up to speed and ready to help stabilize the loads/generation.

The standby losses shouldn't be linked in any way to the power output of the modules, which would necessarily be a maximum number. In other words they can put out a maximum power of 100 kw but as with batteries they can put out anything from 0 to 100 kw.

The answer is "it depends"... If the stored energy is there to run the local power for ~5 minutes until a diesel/natural gas genset can get up to speed and take the load--Then you would run the flywheels near maximum speed (and, as I understand, there are a some flywheel based storage systems for computer centers).

If you where trying to "smooth out" a wind farm--Then you would be looking more at running at 50% capacity so that it can be a load or a source as needed to better match loads to variable generation power.

It makes more sense to say they store 25 kwh when they are at maximum speed and how fast you take it out is up to you as long as it is less than 100 kw. At that point I would expect to see a curve showing how many watts are dissipated vs rotational speed. Some calculus would then give you how the power remaining vs time (as long as you were not drawing power). This would be per 25 kwh unit. The fact that no such curve is provided says more about the whole concept than anything else.

From what I can tell, this post (of mine :-)) is still about as accurate as I have seen:

BB. wrote: »

The reason that I "guessed" how much power running at 100% full standby is because that was the only spec. I could find.

Because many of the frictional losses may be related to the square of the rotational velocity and others are linear--it is difficult to say how long the flywheel may effectively store power.

If we assume that 8% standby loss is proportional to the rotational velocity--you could guess that the amount of power remaining at any point in time might be (as an exponential decay):

100%*(0.92^t)= % energy remaining in t hours...

92% at 1 hour
85% at 2 hours
66% at 5 hours
43% at 10 hours
19% at 20 hours
1.5% at 50 hours

Don't have a clue if the above is accurate or not--but it is probably as close as you will come to comparing the energy storage to a "battery" without calling Beacon up and getting more specifications.

And, after reading the specs. more closely--I believe that the 8%/12.5 hour number is correct. This is a typical method used by marketing to bury the negatives of a system (use % power losses without clearly stating any real numbers and spec system out to 30 minutes of use--but define power output as a maximum amount over 15). Not quite deceptive--but defining strengths of a system in the best ways possible.

-Bill

Also that the only actual installation was for such short term storage says a lot and this is consistent with the advertised rate of discharge of 100 kw for 15 minutes. The manufacturer of my golf cart batteries advertised how many hours they would run, not how many minutes.

So--At this point, flywheels would be for short term storage and, ideally, would have better cycle life than a big bank of lead acid batteries (or even super caps, etc.).

Otherwise, AGM Batteries and Flywheels have similar capabilities (~15 minutes discharge/charging cycle "best case" specifications). It is really the electronics (and any electromagnetics of the flywheels) that are limiting the actual peak output (i.e., make the flywheel support a 1 minute charge/discharge cycle--15x peak wattage--Sounds great, but are the variations in, for example wind farm output, on a 1 minute time constant, a 15 minute scale, or what).

You can go to the California Independent System Operators web site and see the variability in wind on a state wide scale (CAISO is ~85% of the electric power in the state)... It looks like 15 minutes would be about the minimum useful storage time to manage variable wind farm output.

It would be interesting to see what the "fine grained" variability is (on a local/regional level, and down to the minute or less resolution)...

Anyway--At this point, I don't think flywheels are going to be used for short term power/load management at this time.

-Bill

Re: Flywheel based electrical storage

Here is their current marketing presentation... Nothing on efficiency, but quite a bit of detail on short term power variations and overall cycle life:

http://www.beaconpower.com/files/bp_intro.pdf

-Bill