• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Gas blast: What the?

arcticpenguin

Philosopher
A
Joined
Sep 18, 2002
Messages
5,687
http://www.theaustralian.news.com.au/common/story_page/0,5744,7457952%5E1702,00.html

A GAS explosion at a hospital in eastern Japan injured at least 10 staff and patients today, police said.

The blast at the hospital in Iwaki city occurred when six hospital staff tried to remove helium gas from magnetic resonance imaging equipment, a Fukushima prefectural (state) police spokesman said on condition of anonymity.

Most of the injured, which included at least four patients, were cut by flying glass and a few had broken bones, the spokesman said.
Click to expand...
Helium gas will not ignite. It it really was helium it must have been something like the pressure building up in a closed container without a safety valve.
 
arcticpenguin said:
http://www.theaustralian.news.com.au/common/story_page/0,5744,7457952%5E1702,00.html


Helium gas will not ignite. It it really was helium it must have been something like the pressure building up in a closed container without a safety valve.
Click to expand...

Yeah, sounds like a combination of pressure and a crack in a tank or something. I still remember when my next door neioghbor's big air compressor blew up. The welds on one end of the tank gave way. Blew a big hole right through the side of the garage.
 
Sounds like somebody forgot to turn off the main tank valve before removing the regulator...
 
Sounds to me like it might be a little different than any of those situations. MRI uses liquid helium to cool the magnet down to superconducting temperature. You get boiloff from that liquid, of course, which can provide plenty of pressure if not vented properly, but it might not have involved a gas cylinder at all.
 
Arctic---Maybe it's simply a case of someone not knowing their gas...from a hole in the ground!:wink8:
 
The same sort of thing was reported during the situation at Three Mile Island.
Pressure was building up in the containment vessel--a buildup of hydrogen. After the pressure was eased off, a Washington, D.C., newscaster reported that "The danger of a hydrogen bomb explosion has diminished." :eek:
 
I think Ziggurat is right, it was probably liquid helium used to cool some materials down to superconductor levels. I remember in physics lab, we were doing tests with liquid helium dewars by sticking temperature probes down in the liquid. If you lowered it too fast, the helium would vaporize quickly and overpressure the inside of the container. The prof said someone let the probe fall once and the subsequent pressure increase shot the thing out and into the ceiling.
 
Ziggurat and zer0vector have already commented on the possibility of pressure building from the vaporization of liquid helium.

I'll add here that a superconducting magnet system (as used for MRI) has the potential to "quench", dumping the (really substantial) energy stored in the magnet into the liquid helium bath. When this happens, a lot of liquid helium becomes very cold gas (with the appropriate change in volume!) in a relatively short period of time. Depending on the system, thousands of liters of liquid may be vaporized in a time frame best measured in seconds.

It's also worth noting that the cryostat that holds the magnet relies in part on a fairly good vacuum to allow things to live at cryogenic temperatures with an acceptable normal rate of boil-off. So, another way to ruin the day is to "soften" the vacuum substantially - cryogen boil-off can increase dramatically. If the liquid level gets low enough, the magnet will eventually "quench" when part of it warms up enough that it ceases to be superconducting.

Such systems are designed with pressure relief devices (burst disks, and sometimes also check valves) suitable for handling a quench without rupturing the vessel in which the magnet resides. Depending on where the magnet is sited, it may also include ducting suitable to dump it outside the building (suffocation hazard, among other things).

The temperatures in the helium vessel are such that, if ambient air is allowed to get in, much of what comprises air will condense and freeze. So, it's possible, through continued sloppy procedures (or a leak), to obstruct pathways in the cryostat to an extent that normal boil-off can escape without trouble, but the gas evolved in a quench cannot.

I'm not suggesting that this is what happened here, just pointing out that such potential exists in a superconducting magnet system.

_Q_
 
This is a little unrelated, but since _Q_ knows a bit about it, I thought I'd ask. Assume you have a superconducting system, and you're running a fair amount of current through it. If the material suddenly becomes non superconducting( R becomes greater than zero), then by:

P = I^2 R

The power that needs to be dissipated by the material increases dramatically. I don't know if this could cause an explosion, but it could certainly melt down any superconducting components.
 
zer0vector said:
This is a little unrelated, but since _Q_ knows a bit about it, I thought I'd ask. Assume you have a superconducting system, and you're running a fair amount of current through it. If the material suddenly becomes non superconducting( R becomes greater than zero), then by:

P = I^2 R

The power that needs to be dissipated by the material increases dramatically. I don't know if this could cause an explosion, but it could certainly melt down any superconducting components.
Click to expand...

As _Q_ mentioned in his description of a magnet quench, the heat dumped by a quenched magnet can be quite significant. But since it's in a cryogenic environment to begin with, and so starting out cold and in a cold environment, actually melting the components is generally not a concern. Superconducting magnet systems are built to withstand quenching (and are usually test-quenched to make sure they can stand the stress) because accidents happen. But I think simple thermal expansion (which can be very anisotropic - the heat dump may be fairly localized on the magnet) is probably a bigger worry than melting in terms of damage to the magnet itself during a quench (assuming coolant venting works correctly).
 
Ziggurat said:


As _Q_ mentioned in his description of a magnet quench, the heat dumped by a quenched magnet can be quite significant. But since it's in a cryogenic environment to begin with, and so starting out cold and in a cold environment, actually melting the components is generally not a concern. Superconducting magnet systems are built to withstand quenching (and are usually test-quenched to make sure they can stand the stress) because accidents happen. But I think simple thermal expansion (which can be very anisotropic - the heat dump may be fairly localized on the magnet) is probably a bigger worry than melting in terms of damage to the magnet itself during a quench (assuming coolant venting works correctly).
Click to expand...
You're right, local heating could damage the coil, and damage to the coil is devoutly to be avoided.

In at least some systems, a network of diodes and resistors (located in the helium bath) is connected to the magnet in such a way that no isolated section of the magnet has to take it on the chin in the event of a quench. The energy is still dumped into the bath.

Something that I neglected to mention in my earlier message is that such magnet systems usually include some facility for intentionally initiating a quench by energizing small heaters located within the magnet coil. This is a safety feature, allowing the magnet to quickly be deenergized in an emergency (life-threatening) situation.

_Q_
 
zer0vector said:
This is a little unrelated, but since _Q_ knows a bit about it, I thought I'd ask. Assume you have a superconducting system, and you're running a fair amount of current through it. If the material suddenly becomes non superconducting( R becomes greater than zero), then by:

P = I^2 R

The power that needs to be dissipated by the material increases dramatically. I don't know if this could cause an explosion, but it could certainly melt down any superconducting components.
Click to expand...
It's not unheard of for some (small) part of such a magnet system to become resistive enough to be a problem, but not enough so as to cause a quench. Yes, I^2R heating is observed. It might be severe enough that the helium boil-off caused by this heating far exceeds the normal boil-off of the cryostat, or it might be subtle enough to only be observed as an unacceptable rate of decay of the magnetic field.

_Q_
 
_Q_ said:

In at least some systems, a network of diodes and resistors (located in the helium bath) is connected to the magnet in such a way that no isolated section of the magnet has to take it on the chin in the event of a quench. The energy is still dumped into the bath.

Something that I neglected to mention in my earlier message is that such magnet systems usually include some facility for intentionally initiating a quench by energizing small heaters located within the magnet coil. This is a safety feature, allowing the magnet to quickly be deenergized in an emergency (life-threatening) situation.
Click to expand...

The superconducting magnets I've worked with were all for SQUID magnetometers (devices used for measuring the magnetic responses of small samples of materials), so although they go to high field (7 tesla), they're quite small (spatially) compared to an MRI. I don't recall noticing anything in the manual about resistors to dump the heat in a quench, but since they're much smaller, you also wouldn't have near as much energy to dump as for an MRI (those not familiar with this, energy in the magnet is proportional to the integral of the magnetic field squared over all of space). It makes some sense that an MRI would require fancier safety features, but the resistor network is actually something I hadn't thought of before. We also had no way of performing an emergency quench, but you'd also be hard-pressed to create a life-threatening situation with a SQUID mangetometer, and since there's a lot less energy in the field than for an MRI, I'm guessing it's probably also a lot quicker to ramp the field down normally without quenching.
 
You must log in or register to reply here.

Back
Top Bottom