What would happen if an energy storage device failed?

[D Tibbets]

Deleted, see next post.

Dan Tibbets

[D Tibbets]

The real question is not if a few people will be killed (BP cared little about that aspect in their Texas refinery), but the cost of the accident.

I think that in developed country such as USA or Japan, etc. each occurred death is very very costly for BP or any other company e.g. TEPCO....

Actually you are wrong concerning BP. Concerning the Texas refinery explosion and fire a few years ago. Aside from stupid operational decisions, they had planed costs of the plant with certain conditions. One was that they assigned a monetary value to the workers lives. The had a limit where if making the system safer cost more than ~ 1.5 to 3 million dollars per worker, then it was financially reasonable to avoid this cost. After the accident, and the ongoing lawsuits, I suspect this attitude is regretted, but I doubt their policies have changed much other than perhaps better hiding of the decision process.

And your assertion that much smaller devices cost as much as large devices is mostly wrong. There are variations, but costs generally scale as ~ the 3rd power of size. By this scaling a Polywell ~ 1 /10th the volume (or less) ^3rd power or ~ 1000 times less. Even assuming a 2nd power scaling , a 100 MW Polywell would cost ~ 1/100th of a 10 GW Tokamak (this is probably close to the smallest tokamak that could be made). At first glance you would consider them equivalent. But there are grid considerations, failure time scales, life time scales, and consequences of single failure r off line intervals. I have heard repeatedly that electrical providers hate the Tokamak for these reasons. It is just too big for practical application.

Also, mention of ~ 10 Tons of TNT equivalent my be reasonable. That size of explosion would destroy the entire plant. But, you say, the rate of the explosion is slower and this greatly mitigates the blast. True, but there are two adjustments. How much slower? An assumption that a high explosive burns in ~ 10 micro seconds, and a quench and ohmic heating lasts 10 milliseconds may result in a blast effect of ~ 1/1000th (I don't know if it is a linear relationship). This would result in a blast effect similar to ~ 20 lbs (~10 KG) of TNT. That is a significant blast. Secondly, this ignores secondary effects- rapid expansion of a superheated liquid (steam explosion of the liquid helium), pressurized water steam pipes, hot liquid lithium, possible vacuum vessel rupture and implosion, etc.

PS: If a range of superconducting magnet energy storage is the topic, then Tokamaks are not excluded. They are a reasonable example of such a system scale where failure is becoming increasingly significant for the local and non local risks, and catastrophic loss of a very expensive machine, both in terms of money, and the time necessary to repair or replace it. Even if only one magnet is damaged, the repair and replacement costs may be many hundreds of millions of dollars and a year off line.

Another superconductor that hasn't been mentioned is the Navy's consideration of using an ~ 30,000 HP superconducting electric motor in warships. How robust is it ? Would a torpedo that put the ship at risk, instead ensure that the ship was sunk, or at least totally disabled (no motive capacity).

Dan Tibbets

[Skipjack]

Please describe, hypothetically, a safety measure that would prevent the explosive release of said hypothetical energy if a hypothetical terrorist shoots a hypothetic shaped charge thru the hypothetical device.

The hypothetical safety measures that would prevent the hypothetical explosive release of said hypothetical energy if a hypothetical terrorist shoots a hypothetic shaped charge thru the hypothetical device would hypothetically be several instead of one.
Hypothetically, one could split up such a hypothetical energy storage device in several smaller hypothetical devices that are separated by hypothetical armor, or other means of hypothetically preventing any hypothetical bullets or other hypothetical devices from penetrating all of the hypothetical energy storage devices at once.
This should, hypothetically, keep the size of the hypothetically resulting explosion small enough that it can be contained to a small part of this hypothetical energy storage plant.

[Joseph Chikva]

Energy releases in milliseconds vs. nano or microseconds in high explosive case (1000-1000000 times slower), energy released in heat form by the current oscillating in resistive matrix of superconducting cable. That energy is enough only for heating of that matrix only on 17 deg K. Not enough?

Absolutely not. The heat does NOT appear evenly throughout the matrix. It appears at the break. And that is enough to make the material at the break go KABOOM!

The links that ladajo provided abonve show the result of a measily 1MJ (ONE mega joule) of energy dumping across a gap. That kind of EXPLOSION (read kaboom) kills many people a year around the world. Dumping TJ (Tera Joules) across gap would be a bigger kaboom.
Actually, TJ is WAY too small. 10TWHr = 36000TJ = 36 PJ (peta Joules)

Yes, energy appears at the break or at any switching off. Then we have transient process. And do you know that alternate electric current flows through the dielectric (read e.g. capacitor)? And in considering here case ITER's SC cable weighing 365 t (or real numbers) there is not any other way that energy will not transfer in the heat form.

If you speak about the appearing gap, can you mention that dimension of that gap and than order of magnitude of self-EDF? This important if you bother on electric arc. And it is very interesting for me numbers.
Then we can discuss how 41GJ may be equivalently to 10t of very dangerous HE energy and at the same time the same energy is equivalently to 0.95 t of less dangerous e.g. jet fuel’s energy. For you note the typical charge of fuel in commercial passenger aircraft 40 t of jet fuel.

10TWHr is too fancy energy storage for inductive and all the more capacitive energy storage.

If you bother that energy of MJ order can kill, recall that 5.56x45 bullet of M4 has only 1.5 kJ and also can kill.

[Joseph Chikva]

I have heard repeatedly that electrical providers hate the Tokamak for these reasons. It is just too big for practical application.

Please, show me such supplers who hates TOKAMAK. As I am sure that any supplier of custom design electric components will then include ITER with the pride into their reference-list (past projects).

But, you say, the rate of the explosion is slower and this greatly mitigates the blast. True, but there are two adjustments. How much slower?

Lets estimate together.
Composition B has density 1650 kg/m^3, velocity of detonation: 8,050 m/s and provides 0.703 m^3/kg of gases.

Admit that the bomb has a shape of sphere.
10'000 kg has a volume 6.06 m^3
Volume of sphere V=4/3*πR^3
So, R=cubic root of (3/4*V/π) = 1.13 m
So, the time of propagetion of shock wave from bomb's surfece to the center we can estimate as 1.13/8050=140microseconds

So, in 140 microsecond 10 t composition B provides in the sphere diameter 2.26 m 7030 m^3 of gases.
Now please describe the same for Inductive Energy Storage or Tokamak as 41GJ of energy is also equivalently to 1.36 t of butter.

[rcain]

....
Now please describe the same for Inductive Energy Storage or Tokamak as 41GJ of energy is also equivalently to 1.36 t of butter.

"European Atomic Energy Commission disperses EU Butter Mountain in ITER Tokamak" - that sounds not so much dangerous as downright messy to me JC

(though perversely not far from the truth of recent EU/ITER news).

[Joseph Chikva]

"European Atomic Energy Commission disperses EU Butter Mountain in ITER Tokamak" - that sounds not so much dangerous as downright messy to me JC

(though perversely not far from the truth of recent EU/ITER news).

I did not say that it is not dangerous. But the following statement of Dr. Bussard does not correspond to true:

However, many magnetic fusion power systems, especially those using large fields over large volumes, have been designed to use superconducting magnets in order to reduce drive power requirements and thus improve system power balances. In such systems, coil failures can cause the stored magnetic energy to be dumped precipitously as ohmic heating in “normal going” superconductor coils, potentially leading to coil material meltdown. If the coil inductance is; or becomes suciently small during failure, this can even result in quasiexplosive vaporization of coil material. The stored magnetic energy in such systems is thus a very large hazard and potential source of
system catastrophe. Typically stored energy in large scale superconducting magnetic fusion systems; e.g. large tokamaks; is of the order of 10^12 joules;10^6 MJ; 4 MJ is equivalent to energy content of 1 kg of high explosive.

Where he saw "Typically stored energy in large scale superconducting magnetic fusion systems; e.g. large tokamaks; is of the order of 10^12 joules" in the beginning of 90s? If ITER - the largest TOKAMAK will have only 25 times lower stored and 25 years after.
Why "the stored magnetic energy to be dumped precipitously as ohmic heating in “normal going” superconductor coils, potentially leading to coil material meltdown" if that coil material will have enough heat capacity to be not molten?
Why "quasiexplosive vaporization" and not temperature increase on several tens Kelvins?

[Giorgio]

Where he saw "Typically stored energy in large scale superconducting magnetic fusion systems; e.g. large tokamaks; is of the order of 10^12 joules" in the beginning of 90s? If ITER - the largest TOKAMAK will have only 25 times lower stored and 25 years after.

He was not referring to an experimental or existing machine, he was probably referring to the operational parameters of some of the designs for commercial working Tokamaks of the 90's.
If we spend some time looking we might even identify to what particular design he was referring, but I really do not see the need to do that.

Why "the stored magnetic energy to be dumped precipitously as ohmic heating in “normal going” superconductor coils, potentially leading to coil material meltdown" if that coil material will have enough heat capacity to be not molten?
Why "quasiexplosive vaporization" and not temperature increase on several tens Kelvins?

Because an uniform increase of temperature over all the mass of the reactor will require an uniform and instantaneous failure over all the superconducting medium, while in fact we know this type of accidents are normally localized, so there is little to no chance that a spot of material can handle all that energy without vaporizing. Just remember what happened in the LHC.

[ladajo]

You took the words out of my mouth Giorgio.
Joseph seems to focus his argument on a dispersed failure. Whereas real failures tend to be local events, and in that once the failure starts it tends to feed itself exponentially in the focus of available energy to its mode of failing. Energy is looking for an out, and invariably, a failure provides a path, that then becomes a better path, and better again (at least briefly) until the energy source runs out of push.

At this point, I am not sure what Joseph is trying to argue. It would seem he is trying to say that failures can never happen, because energy seeks a homogenious rest point. Of course he is not seeming to recognize that if the point is somewhere else, and the door is small and fragile...

[Joseph Chikva]

If we spend some time looking we might even identify to what particular design he was referring, but I really do not see the need to do that.

I need not to spend time for learning 90s or earlier or later projects.
Typical is plasma volume 1000 m^3, on axis field 6-7 T or smaller machines with field up to 14T.
I can spend some time for estimation the volume of magnetic field as that exceeds plasma volume and, so, then to estimate stored energy.
But do not see necessity.
As statement that superconducting cable will obligatory be molten down in case of loss of superconducting condition is wrong. Not obligatory in case of big enough mass of conventionally conducting matrix. And we have successfully shown that 365 t is enough in ITER’s case.

Because an uniform increase of temperature over all the mass of the reactor will require an uniform and instantaneous failure over all the superconducting medium

We have not supeconducting medium but have thin superconducting elements (filaments or how they called in English) in resistive but well conductive medium. As I know that medium is copper matrix. Why not uniformity?

[Joseph Chikva]

Whereas real failures tend to be local events,...

Real (failures) explosions of pressure vessels e.g. steam boilers taught us how to make them more safe. And now we are not bothering about pressure. On the contrary, every newer design of steam turbine uses higher parameters of steam: temperature & pressure.

[ladajo]

Joseph, the point is that the failure starts at the point of weakness for the energy boundary. It does not start uniformly across the entire construct. Once that failure point defines, it then renders previous engineering calculations mostly mute, as they were not applied to the "new" structure created at the failure point.

In simple terms, once it starts to bust, all previous bets are off.

If you provide a grounding point for a Superconducting magnet, The energy contained within will seek to use the exit point, and more than likely in a spectacular fashion if the point was not planned for.

The manifested failure paths are not ones normally capable of carrying the load because they were not planned for.
Yes you can predict and engineer for failures, however, at some point, a failure will occur that was not in the plan.

Whoever thought that a mouse would lead to a car burning down and exploding on the side of the road? But it has happened, mouse takes up residence, chews on things that look tasty, car gets driven, things leak, catch fire, and explode.

[KitemanSA]

Please describe, hypothetically, a safety measure that would prevent the explosive release of said hypothetical energy if a hypothetical terrorist shoots a hypothetic shaped charge thru the hypothetical device.

The hypothetical safety measures that would prevent the hypothetical explosive release of said hypothetical energy if a hypothetical terrorist shoots a hypothetic shaped charge thru the hypothetical device would hypothetically be several instead of one.
Hypothetically, one could split up such a hypothetical energy storage device in several smaller hypothetical devices that are separated by hypothetical armor, or other means of hypothetically preventing any hypothetical bullets or other hypothetical devices from penetrating all of the hypothetical energy storage devices at once.
This should, hypothetically, keep the size of the hypothetically resulting explosion small enough that it can be contained to a small part of this hypothetical energy storage plant.

Seems your answer amounts to "you can't so don't build it". Well mdeminico, that is the answer afforded by skipjack.
I figured if prodded hard enough we might actually get an answer fronm one of them .

[Joseph Chikva]

Joseph, the point is that the failure starts at the point of weakness for the energy boundary.

First let's consider the case when superconducting condition losses via neutron flux.

[Joseph Chikva]

Seems your answer amounts to "you can't so don't build it". Well mdeminico, that is the answer afforded by skipjack.
I figured if prodded hard enough we might actually get an answer fronm one of them .

I think that if you put the wire into the plug, result would not be pleasant.