IEEE - Can Synthetic Inertia from Wind Power Stabilize Grids?

[gary]

In an article (http://spectrum.ieee.org/energywise/energy/renewables/can-synthetic-inertia-stabilize-power-grids) in the Institute of Electrical and Electronic Engineers (IEEE)
Spectrum Magazine, contributing editor Peter Fairley reports on how
Canadian power utility operators are using a technology called "synthetic
inertia" that can turn wind turbines from being "potential liabilities to
power grid stability into substantial contributors to it."

[Shiraz]

thanks Gary - really interesting

[glend]

Yes, the IEEE article mentioned the SA recent outage i see. Is synthetic inertia just another buzz word for the old fly wheels. I recall an incident in a big Telstra exchange many years ago, where a power system flywheel jumped its mount and ran through the bottom of the building causing great damage. :eyepop:

[xelasnave]

Yes thanks for posting Gary.
I was surprised as to how efficient a well designed flywheel is and how its ability to store energy is, as I recall and sorry for no referenced authority other than my most unreliable memory, to be superior to a battery.
What type of battery of course but the thought of a flywheel spinning away for an extended period and losing less energy than a battery really surprised me.
Their potential to run around and do damage had not occurred to me.

Alex

[Shiraz]

maybe the synthetic flywheel effect could be used to allow time for something like this to take up transient loads and keep the grid stable without any synchronous generators at all.https://en.wikipedia.org/wiki/Dinorwig_Power_Station

guess lots of batteries would be better though - the sooner we go to grid connected household batteries and electric cars the better.

[gary]

Hi Alex,

The synthetic inertia systems on wind turbines discussed by the authors
don't employ a separate flywheel.

Instead they use the mechanical inertia of the wind turbine itself coupled
with control systems added to the power electronics.

By way of background, historically, power generation by steam turbines
uses synchronous generators. That is, the rotational speed of the generator
is synchronous with respect the (here in Australia) 50Hz alternating
current of the grid.

As wind turbine technology has evolved, rather than connect their generators
directly to the grid, the trend has been to employ power electronics that
go between its generator and the grid.

This has been made possible by huge advances in power electronics
over the past thirty or forty years.

For example, some systems employ double converters that convert the
AC from the wind generator to DC and another converter that converts
that DC back to (50Hz in Australia) AC which is synchronized with the
grid.

There are additional closed-loop control systems that feedback from the
grid back to the converters and to the turbine itself.

Power grid systems and power integrity on a grid is surprisingly complex.

Rather than there being just an alternating voltage waveform there is
also an alternating current waveform. Since power is the product of
voltage and current, on the one hand, ideally the two waveforms would be in
phase to deliver maximum power to a load. This is referred to as unity
power factor. When voltage and current is not in phase it is referred to
as reactive power. However, some amount of reactive power is required for
some motors to operate and reactive power plays an important role
in maintaining the voltage on the grid. It is a careful balancing act
between the amounts of normal power and reactive power on the grid.

Ideally the generators on the grid should be able to produce or
absorb reactive power to or from the grid itself.

So the synthetic inertia technology, which has been evolving in recent
years, is a combination of the inertial mass of the rotating machinery,
the power electronics conversions and the control systems used in
the feedback loops.

Control systems in (electrical) engineering is a complex area in itself with
deep roots in mathematics. That is also true literally as many control
systems utilize the mathematics of matrices of complex numbers
(electrical engineers love using complex numbers, that is, where the
square root of -1 is defined as the complex number "i" but electrical
engineers prefer to denote with the letter "j" because we already use "i" to
denote current).

We have had over 100 years experience in designing power distribution
systems using synchronous rotating machines. With the advent of
renewable power sources whose inherent generation schemes may
be asynchronous and even DC, there is less experience to draw upon
but many talented professionals are evolving the field quickly.

For example, the advances in semiconductors used in power electronics
continues to rapidly advance.

There was the famous debate between Edison and Tesla
on DC versus AC. Rotating machinery and passive transformers made
the obvious choice based on the technology available at the time.
Based on the technology available today, the debate would probably be
even harder fought if we were starting afresh.

[julianh72]

ANY form of stored energy has the potential to do damage if the energy is released in an uncontrolled fashion.

Flywheels store energy as kinetic energy - so if there is some sort of mechanical failure, that energy has to go somewhere.

Batteries store electrical energy as chemical energy - short out a battery, and things can get pretty hot pretty quickly.

Fuel tanks store chemical energy - an uncontrolled release can be explosive.

Hydro-electric dams store vast amounts of energy as gravitational potential energy - if the turbine or penstock fails, the water flow can destroy the power station, and the homes of thousands of residents below the dam.

[xelasnave]

Thanks Gary for taking the time to post all that information.

There is a small wind turbine at the new place.
I have no idea what it produces but maybe 200 watts at 12 vlts.
It is set high and gets good wind.

Not putting that system together myself I feel a little intimidated by it where as my previous system I built and altered over a 20 year period.

But the big luxury was a 2 kva Genny, from Bunnings $750 that has key or promote start.
That unit one could only dream about once.
Runs a flash gaming computer and so far no problems.

I am mostly just down the road from you these days and think of you each time I go thru My K for a visit at Brooklyn.
Again thanks.
Alex

[xelasnave]

Hi Julian
Thanks for the run down.
I have learnt the hard way about energy release in a battery bank.
I left heavy leads connected to my battery bank laying on the ground.
Well whilst out my dogs rolled around on them and I came home to burnt cables, didn't ruin the battery as something must have burnt thru or one of the dogs had the sence to separate them, maybe not but I know Lassie would have, and lucky there was no fire.
Really lucky all round.

I am toying with adding a large water tank up the hill to be filled via solar and add a peltie wheel.
But I have not got around to working any figures out...they use a lot of water but another production unit would be good.
I need to measure the fall next time as a starting point.

Alex

[Eratosthenes]

can you provide an example of "something" that doesn't store energy?

[julianh72]

No - but that's not my point.

Alex commented that the potential for flywheels "to run around and do damage had not occurred to me", and I was just making the point that ALL energy storage systems have the potential for destruction if the stored energy is released in an uncontrolled fashion. It's the quantity of stored energy that matters, rather than the mode of storage (although some storage systems are inherently safer / more controllable than others.)

It's probably helpful to think about how much useful energy is stored in a couple of common storage scenarios:

AA NiMH Rechargeable Battery: 2500 mA.hr x 1.2 volts = 11 kJ approx

Typical 12 volt Car Battery: 40 A.hr x 12 volt = 1.7 MJ (or 160 AA batteries)

Tesla Powerwall Mk II ("can power a two-bedroom home for a full day" according to the publicity material): 14 kW.hr = 50 MJ (equivalent to approx. 4700 AA rechargeable batteries or 30 car batteries)

Typical Car Fuel Tank: 60 litres @ 46.4 MJ/kg ~ 2100 MJ (approx. 200,000 x AA batteries, or 40 fully-charged Tesla Powerwalls - people tend to forget just how much energy we deal with every day when we drive around in our cars!)

Wivenhoe Pumped Storage Hydro Power Station: 2 x 250 MW x 10 hr = 18 TJ (1.6 billion AA batteries, or 350,000 Tesla Powerwalls, or 8,000 full car fuel tanks)

Typical Commercial Flywheel: 2 kW / 6 kW.hr / 100,000 rpm Flywheel UPS, e.g. as used in Telecomms UPS applications: 6 kW.hr = 22 MJ (2000 AA batteries, or 12 car batteries, or a bit less than half a Tesla Powerwall)

[Eratosthenes]

...and yet a single rain drop at rest can potentially unleash about 9,000,000,000,000 Joules in energy

[xelasnave]

A flat battery.

Alex

[julianh72]

Interestingly - that's roughly the same amount of energy stored and then re-released by Wivenhoe Pumped Storage Hydro Power Station in each 24-hour cycle - but each day, it has to lift about 28 GL of water by 100 metres elevation to store a comparable amount of energy!

http://www.csenergy.com.au/content-(168)-wivenhoe.htm

[Eratosthenes]

interesting....the energy required to lift 28 GL of water by 100m elevation is ~ (28,000,000,000 * 9.81 * 100) = 27.468 TJ of energy

How much of this energy is recovered when it is later unleashed and runs through the turbines?

[Steffen]

Yes, back then AC at the rotational speed of generators was the only way of efficiently distributing electrical energy. That is because it could be transformed up and down to higher or lower voltages with relatively low loss, hence overland transmission lines could be operated at much higher voltages than what households were supplied, commensurate with much lower currents (and transmission losses) than otherwise.

Modern alternative power generation methods necessarily decouple themselves from AC sync restraints. Solar power plants are free to generate DC, as wind farms will be free to generate AC at the frequency the wind dictates. Thanks to modern power electronics they can both feed into the grid at the required specs and, as the article shows, even to the benefit of the grid.

With the new speculative very high voltage DC distribution grids Edison may eventually triumph over Tesla, as much as it pains me to say that :)

From the control theory point of view I don't think there are many open questions in principle, however the economics of distributed power generation pose interesting information processing and legal questions, to which cryptographic distributed ledgers could provide an answer.

[julianh72]

Wivenhoe Pumped Storage Hydro Power Station is rated at 2 x 250 MW generating units x 10 hours generating time / 14 hours pumping time per 24-hour cycle. The recoverable energy is thus 5 GW.hr or about 18 TJ per 24 hour cycle, but the losses in pumping and pipeline friction etc mean that you need to expend more energy pumping the water uphill than you recover in the generating phase. The usable stored energy is "roughly the same" as the 9 TJ of theoretical fusion energy contained in a single raindrop.

I don't have the exact efficiency number for Wivenhoe, but the 250 MW generator / motor units, and the pump / turbine units, would have very similar electro-mechanical efficiency in both modes of operation; a typical 24-hour cycle is 14 hours of pumping followed by 10 hours of generating, giving an indicative overall "round-trip" efficiency of about 71%. Industry standard figures for large-scale pumped storage of 70% - 80% is typical, so the numbers stack up.
https://en.wikipedia.org/wiki/Pumped-storage_hydroelectricity

Modern rechargeable battery technologies are more efficient (80% - 90%, say) but we don't have any electro-chemical battery technology that can scale to GW.hr capacity yet.

The point of pumped storage systems is to act as giant batteries, which can store "base load" scale energy whenever and wherever it is generated, and release it when it is needed. Wivenhoe was built to allow Queensland's coal-fired power stations to operate at maximum base load efficiency round-the-clock, "pumping electricity uphill" in the off-peak periods, and releasing it in peak demand periods, but the same concept works with solar and wind power, for example, which by their very nature are not necessarily generated at times of peak demand.

[sjastro]

Really care to do a back of an envelope calculation for the distance travelled (ignoring atmospheric resistance) in order for a typical rain drop (volume 0.05ml) to release 9x10^12 Joules.

Then try calculating the energy incorporating atmospheric resistance where the resultant force on the rain drop is no longer conservative.
This latter calculation will give a much more realistic value.

I leave it to you as an exercise.

[julianh72]

I'm pretty sure Eratosthenes was talking about the potential fusion energy contained in a rain-drop, not the kinetic energy of a falling raindrop.

0.05 mL = 50 mg of water (i.e. 20 raindrops per gram).
Molecular weight of H2O is 18 (2 x Hydrogen = 2, 1 x Oxygen = 16), so a drop of rainwater contains about 2/18 x 50 mg = 5.5 mg of hydrogen.
About 0.7% of the mass of fusing hydrogen is converted to energy in the process of combining 4 hydrogen nuclei to make one helium nucleus.

E = mc^2 = (0.7% x 5.5E-6 kg) x 3E8 x 3E8 = 3.5 GJ per raindrop

(I guess Eratosthenes must have assumed slightly larger raindrops than 0.05 mL each to arrive at his 9 TJ figure!)

[sjastro]

I doubt that very much given that his calculation for the energy to lift 28 Gl of water is based on the formula for gravitational potential energy
PE= mgh.
I'm pretty sure he had the same idea in mind with rain drops otherwise changing the subject from rest mass energy to gravitational potential energy midstream in a thread seems quite unusual to me.

But then again what do I know....

[julianh72]

Eratosthenes' first post was simply a claim about the amount of energy in a single raindrop, with no other context:
...and yet a single rain drop at rest can potentially unleash about 9,000,000,000,000 Joules in energy

I took it as self-evident that this referred to the potential fusion energy, as this is a totally unfeasible amount of kinetic or gravitational potential energy for a single raindrop. (But my later "back-of-the-envelope" calculation suggests a more credible figure is actually about 9 GJ not 9 TJ.)

I then made the comment that the useful stored energy of 18 TJ in the Wivenhoe Pumped Storage Hydro Scheme is "about the same" as Eratosthenes' figure for a single raindrop - and yes, the stored energy of Wivenhoe pumped storage is simply the gravitational potential energy of 28 GL of water sitting 100 metres higher than the main dam reservoir.

However, it seems that it would actually take the fusion of the hydrogen in about 1,000 raindrops to liberate the same amount of energy - so it's actually a one-shot espresso cupful (50 mL), rather than a single raindrop ....

Even with these "back-of-the-envelope" figures, it's pretty obvious why nuclear fusion of hydrogen is such an attractive source of energy for the future.

[sjastro]

If you believe that is what he meant I respect your decision.

On a different subject your own calculations are wrong.
It looks as if all you have done is to take p-p reaction mechanism for solar fusion and simply assumed it applies here on Earth.
It's not that simple.
A parameter used in nuclear physics is the scattering cross section for fusion which is the probability that two colliding nuclei will undergo fusion.
This depends on conditions such as temperature, pressure and density.
The fact is that none of these conditions can be replicated on Earth so your figure of 0.7% conversion is rather fanciful.

Secondly the source of the protons are from your rain drops, so it has no relevance to the p-p chain mechanism.
You need to use energy to completely ionize the water in order to minimize the Coulomb barrier, then there is the complication of how the Coulomb barrier from O nuclei will effect the scattering cross section of the protons.

Do you know, does anyone know?

[julianh72]

Well, according to the pundits, we're "about 20 to 50 years" from cracking the problem of commercial nuclear fusion as a viable energy source. (Of course, we've been "about 20 to 50 years" away for at least the last 50 years!)

EuroFusion reckons that they are on track for a commercial demonstration multi-hundred megawatt grid-connected fusion reactor by about 2050 - we'll just have to wait and see if they accomplish that goal!
https://www.euro-fusion.org/programme/

Even allowing for the enormous "parasitic losses" required to create the temperatures and pressures required to sustain the plasma, pump the coolants, etc, such a plant would only need to "burn" a few grams of hydrogen per hour for a net usable energy output of say 1000 MW.

[Eratosthenes]

....this morning it was raining quite heavily....I am not clear as to how many raindrops there were in total, but there seemed to be quite a few of them coming down at very similar angles and velocities

[Shiraz]

interesting paper on grid stabilisation http://link.springer.com/article/10.1007/s40565-015-0143-x

hadn't clicked that wind turbines can have a stabilising potential through both the rotor inertia and possibly by keeping a reserve of extra wind power available through the use of conservative blade pitch settings. this requires AC/DC/AC, but it seems that is common these days.