Oxygen monitors are standard practice when you have MRI machines and Helium cooling. I would like to assume that this monitoring was in place and it never raised any flags? https://www.pureairemonitoring.com/helium-mri-rooms-around-w...

The air’s usual O2 concentration is ~20.5% and it looks like safety monitors trigger at 19.5% and 18%: https://www.pureairemonitoring.com/all-categories-gas-monito...

Helium makes up less than 0.0005% of air naturally. I’m guessing that increasing He concentration 1000 times to 0.5%, for example, is enough to cause problems for iPhones, but not enough to trigger the alarm or pose any danger to people.

Breathing helium is completely safe as long as the oxygen concentration is near normal. Hospitals have occasionally even put patients with respiratory disorders on helium/oxygen breathing mixtures because the work of breathing is slightly lower than air.

Agreed! The only risk is that if you did release a massive amount of helium in an enclosed space, it could displace enough oxygen to drop the O2 concentration to dangerous levels.

That's why it seems like oxygen concentration monitors are a good idea for hospitals with liquid He. They don't really need to monitor helium levels since that's not the direct cause of problems. It's only an issue (but a big issue) if there's so much He released that it displaces enough air to meaningfully dilute O2 concentrations.

Helium is much lighter than Nitrogen and Oxygen and may not displace them rather it will float at the top of the room or escape to the atmosphere quite easily. Though if it were Carbon Di'oxide it would have been a different story.

This is completely wrong. You're neglecting gas mixing and the entropy gain of solution.

Our atmosphere does not contain mostly O2 and N2 in the middle, with a layer of CO2, Ar, etcetera near the surface. The sugar in a bottle of coke does not spontaneously sink to the bottom.

Sugar in coke doesn't sink to the bottom, indeed, but if you put a sugar cube in water, it first sinks to the bottom, and then (slowly) dissolves.

Likewise, if you release helium in air, it goes up first, and then (quickly) dissolves in the air.

Disperses would be a better word than dissolves.

Diffuse would be a better word than disperses.

Ahh true i didnt consider that. But the gas mixing would happen only at contact level, something like the meniscus between two fluids. i doubt they would homogenously mix unless the molecules have an affinity for each other, in heliums case it is an ideal gas.

Is it even possible to monitor fir helium levels? It’s an inert gas, so I’m not sure if it’s easily detectable directly.

Yes we use helium sensors for analyzing diving gas mixtures. They infer helium concentration based on the thermal conductivity of the gas; there's no chemical reaction.

Yes, with iPhones!

It used to be the most common treatment for asthma patients in fact.

Wasn't a helium mixture also not used for deep sea diving and (afterwards) decompression chambers?

Yup, technical divers that do extremely deep dives generally use heliox or trimix which can contain 70-90% and 30-45% helium respectively at the surface. The ratios change at depth due to how the various gasses compress, but it's still a much higher concentration of helium than what you'd experience in pretty much any other environment. Unlike carbon dioxide, helium doesn't cause respiratory distress which is why it's a good replacement for nitrogen during deep dives, and also why it's so popular for suicides

Considering that Trimix and Heliox used in deep commercial dives can have the majority (Aka >50%) of the breathable “air” be helium, your theory that there’s really any risk of danger to people seems a bit unwarranted to me.

Helium isn't toxic but it displaces oxygen. Heliox still contains enough oxygen to breathe (at sea level it's 21% oxygen just like normal atmosphere, and when diving the partial pressure of oxygen is the same as oxygen in normal atmosphere). But if you release helium into a room full of air, patients with already-compromised respiratory systems from injury or from respiratory diseases or from opioid painkillers could be at risk for brain damage or death if they were borderline already and not being monitored closely enough.

The author's cavalier "I bet the nurse’s voices were higher pitched that day!" was incredibly inappropriate given the potential danger.

> Helium isn't toxic but it displaces oxygen.

While technically correct in theory, in any real world environment Helium will not displace Oxygen unless you are dumping orders of magnitude more helium than in this scenario.

> Heliox still contains enough oxygen to breathe (at sea level it's 21% oxygen just like normal atmosphere, and when diving the partial pressure of oxygen is the same as oxygen in normal atmosphere).

There’s so much wrong with this (and the rest of your post) I can’t think of a polite way to respond, so I’ll just wish you a good day.

I’ll admit I haven’t gotten that far in my diving certs yet, but isn’t the point of Heliox so that you can have less than 21% O2 at high depths to avoid narcosis and a fatal ppO2?

To simplify it a bit.

The point is to keep partial pressure of oxygen below that of maximum safe limit of 1.2 to 1.6. For example, 100% oxygen at surface is 1.0 and thus safe but will very fast reach maximum when you increase the pressure just by a few meters. Regular air with 21% oxygen will start to be a problem at 50 meters, so divers that need to go deeper need to use mixes that has less than 21% oxygen. When a mix is less than 18% oxygen they become hypoxic and cannot safely be used at shallow depth.

So the point is not to get "the same as oxygen in normal atmosphere". It is to keep the partial pressure of oxygen within safe limits, usually between 0.18 pp02 and 1.4 pp02 depending on a multitude of factors and safety margins.

The only point I was trying to convey with that statement was that there might be much less oxygen by volume in deep-diving heliox but the partial pressure of oxygen is still maintained so that cellular respiration can continue normally. Was that the only thing you thought was outrageously wrong with my comment?

I am not jsjohnst, but the only additional nitpick is that Heliox is any blend that is exclusively helium and oxygen and the reason why heliox is more commonly used for deep dive is that helium has less resistance compared to nitrogen, meaning that pushing compressed air through ones respiratory system takes a bit less effort. Helium in heliox is thus used to displace the nitrogen, with the oxygen ratio being adjusted based on health, depth, bottom depth, purpose, decompression, surface interval and so on.

Kind of, a fatal ppO2. But in a hospital room you have the other kind of fatal ppO2, the kind where you pass out and need an O2 mask.

Hospital rooms aren't hermetically sealed. Helium dissipates so quickly that this isn't a realistic safety concern.

Helium rises. Not slowly. It really rises very quickly. Patients are in beds, horizontal, many also on oxygen feeds. The nurses are standing. Their voices would indead become higher in pitch long before patients were impacted.

If thier voices were rising, id suspect helium but would be worried that some reaction released hydrogen. Same squeeky voices but a serious fire hazard.

I remember these constant problems with helium tanks in rockets. It's way harder to keep helium in place.

You need ~0.2 Bar of oxygen partial pressure. The purpose of Trimix/Heliox is to keep oxygen partial pressure under 1 Bar while also reducing nitrogen to prevent narcosis. You can't breathe the 200m mix at sea level pressure.

> You can't breathe the 200m mix at sea level pressure.

What makes you think you can’t (other than it being hypoxic)? You most certainly can breath Heliox, in fact, it has been used medically[0] for almost 100 years, far longer than used for commercial diving.

[0] https://en.m.wikipedia.org/wiki/Heliox

Heliox doesn’t denote a specific concentration of helium and oxygen; even just within diving the mix varies depending on the depth it is to be used at.

Correct! A common ratio for deep diving is 12/88, but 21/79, 30/70 and 40/60 are also used at less depth.

In fact, Heliox would usually not be used for dives to 200m due to HPNS risk. Usually it would be Trimix for that.

Either way, GP is mostly wrong on the points they made.

Also, the oxygen partial pressure you need to stay under is 1.6 Bar for oxygen toxicity. Huge difference between 1.0 and 1.6, depending on oxygen percentage, the difference could be a >50% reduction in MOD.

Thank you for spelling "breathe" correctly. It seems these days I see it spelled wrong more than I see it spelled correctly.

The only risk is that the O2 levels drop too low from a massive release of He...and while I'll 100% agree this is a very minor risk, I'd still rather hospitals play it safe and install O2 deficiency monitors.

The article I linked mentioned that O2 monitors are legally required for hospitals in NYC. That regulation might have been created because in 2000 someone working on an MRI machine died from a Nitrogen leak (it's also used for cooling and like He, the only risk is that it displaces O2). It's likely this person wouldn't have died if an O2 monitor was in place that sounded an alarm fast enough for the victims to leave the area. http://www.nydailynews.com/archives/news/nitrogen-gas-leak-k...

edit: googling a bit more found some interesting references on dealing with "oxygen-displacing gases". O2 monitors are on the long list recommendations/requirements. One thing I found interesting is that at one laboratory part initial assessment involves a controlled release of the maximum amount of gas that will be stored and then measuring the drop in O2.

https://www.ors.od.nih.gov/sr/dohs/Documents/ProtocolOxygenM...

http://www-group.slac.stanford.edu/esh/eshmanual/references/...

It is the main reason you have to be careful with working with liquid nitrogen in closed rooms: the danger of displacing too much oxygen. While our breathing air consists of 80% nitrogen, when you evaporate large amounts of liquid nitrogen (like 10 liters in an instant) in a closed room, there is the danger of displacing oxygen to a dangerous level. Working with smaller amounts (like 1 liter) is quite safe though.

Not enough to directly endanger people, but a hospital is full of electronics that plenty of people need to live, and it sounds like they could potentially share the iPhone’s vulnerability.

If it makes you feel better:

Hospitals of reasonable size typically have an on-site clinical engineering team that handles those kinds of situations. Important (that is, capital expense category) hospital equipment will typically emit all kinds of warnings and alarms way before anything's actually a problem, because everyone would rather rely on the on-site engineering spending a little extra time silencing false positives then leave anything to chance.

All the nursing journals are full of articles about combating “alarm fatigue”. After 1 google query and 2 minutes of link-clicking:

> “An analysis of alarms at The John Hopkins Hospital, Baltimore, Maryland, revealed a total of more than 59 000 alarm conditions over a 12-day period-or 350 alarms per patient per day.1,2”

https://www.nursingcenter.com/journalarticle?Article_ID=1617...

Alarms are mostly ignored or turned off.

If you spend any time in an emergency room or a hospital, you'll soon realize that there's almost always an alarm going off somewhere, usually more than one and usually for long periods of time.

> Hospitals of reasonable size typically have an on-site clinical engineering team that handles those kinds of situations.

Not for superconducting magnets they don’t. They may be involved with MR scanners somewhere, but that would be extremely uncommon.

It's not an issue for most equipment though because it's only really the MEMS type oscillator that is sensitive to this type of gas intrusion. It's a very new type of clock and most other devices don't use them so there's basically zero risk.

Unfortunately the O2 sensors were being monitored by an iPad.

Or the O2 alarms just don't work and the hospital was negligent.

Occam's razor.

Indeed. I had to reread that paragraph. That's a LOT of helium (by volume).

"So, i noticed some of the workers' iwatch's weren't registering the users' heart-rates. At first, I thought it was because they had all died from suffocation from the massive gas leak. But, then I discovered something interesting... it was actually the helium molecules worming their way into the internal clock chip! Amazing day."

Yeah, but at the same time, it wasn't enough to set off the alarms for low oxygen content, which is at least a little disturbing, right?

Not necessarily.

Nobody could adequately explain to those of us in the CS dept why we had to take 3 semesters of Chemistry. Bits of it hung on and from other sources such as space exploration articles I recall this much:

Fires and lungs both operate on partial pressure of oxygen. As long as nothing else in the air is toxic, your body cares that it gets X oxygen molecules per cubic centimeter of air in your lungs, not Y parts per million. Those deep sea submariners are breathing mostly helium with a small fraction of oxygen in it. If you just compressed surface air there would be so much oxygen that your hair would explode when you ran your hand through it. Assuming the electronics didn't burst into flames first. And if you didn't set yourself on fire, that much nitrogen would kill you pretty quick.

So the question is, does helium displace air or mix in with it? I believe the answer is 'some of both'. If that's the case (and I think we can infer that from "the alarms didn't go off") then a good amount of helium might reduce the oxygen partial pressure less than going to 3000 feet above sea level. So what's the Venn diagram of COPD sufferers, in a hospital wing near the MRI machine, that aren't currently on supplemental oxygen?

Moreover, helium floats to the top and escapes through any opening (staircases, HVAC ducts, gaps in ceiling tiles) because it is lighter than air. This makes it much less dangerous than a heavier gas such as CO2 even if there's enough of it to displace a significant amount of air.

Not just any opening: just about any microscopic crack would let it out and most non-metal materials are effectively porous when trying to contain helium under pressure. It's even used to detect leaks in high vacuum chambers.

You don’t even need cracks. Helium (and hydrogen) can diffuse through metal, especially at high temperatures. Hydrogen embrittlement is a problem for processes that use high temp hiydrogen like the Haber process.

how is it typically dealt with? regular annealing?

I’m not sure about in modern processes, but Bosch’s original solution for the Haber-Bosch industrial process was to just periodical replace an expendable steel liner in the reaction vessel.

That is not true by any stretch of the imagination. If gases behaved like that then we'd all be choking on a layer of argon with all of the oxygen out of reach above us. No, gases diffuse and mix easily.

Gases diffuse and mix given enough time. There's not enough time when you have a massive indoor gas leak.

Even outdoors, large amounts of CO2 tend to sink to the ground and suffocate people in low-lying areas. Look up Lake Nyos for a particularly grizzly example of that.

Helium breathing mixes are for deep sea divers, not submariners. Most submarines run an interior gas mix nearly identical to regular air.

I didn’t mean navy boys, I meant research submarines at extreme depth.

Research submarines at extreme depths also use regular air. There's really no benefit to helium mixes for submarines. The breathing gas isn't compressed, it's at about 1 bar.

Odd, I assumed that subs would run at least a few atmospheres to compensate for pressure differential. Structural assistance not worth giving the crew the bends on fast ascents?

Why bother with 2-3x more air pressure and all the problems that go with it (crew decompression, lack of abort options, etc.) when the outside pressure is hundreds or thousands of times more? You're really not gaining anything engineering-wise.

One thousand times would make it 10km depth - so clearly not “thousands”. Other than that you are right that there is no gain in e.g. doubling the internal pressure.

Yes I made a mistake in my estimation there. Max pressure at the challenger deep is about 1,000 atmosphere, so that shouldn’t have been plural.

The pressure outside is +1 bar per 10 meters. If you're 1 km deep, that's 100 bars (+1 atmospheric bar) while it's ~1 bar inside the sub. The mechanical stress on the submarine depends on the pressure difference, not the ratio. So increasing air pressure (with all the complications and risks it would bring) would only reduce the material load by a few percentage points.

Then why all the videos of people who sound like garden gnomes?

Which videos?

I assume the first two semesters were first-year (freshman) chemistry. What was the third year? Organic?

As long as there is enough oxygen in the air, what's disturbing about it? People breathe helium/nitrogen/oxygen mixtures all the time.

https://en.wikipedia.org/wiki/Trimix_(breathing_gas)

I didn't mean that the air would be dangerous to breathe.

I meant that it's disturbing that the amount of He needed to disable the iPhone is low enough that the standard sensors around He (which measure 02 levels) don't regard it as a major leak for He to be at that level.

Virtually all (esp. consumer) devices assume 'normal' atmospheric conditions with 'normal' temperature ranges and 'normal' humidity ranges. As you start to go outside those ranges, strange things can happen. People at extreme altitudes and/or in extreme climates see this on some types of devices. It's not a big deal. If you have needs outside the typical ranges, you provide your specifications and be prepared to pay much more for your solution.

Why would that be disturbing? Electronics and organisms do not have identical environmental tolerances/intolerances.

Presumably, the sensors exist to protect humans, not electronics.

A device-disabling threat that normal threat sensors can't detect, even when those sensors have been deployed because of the chance of unusually high levels of that threat?

Yes, I'm allowed to be disturbed by that; why wouldn't you be?

Not at all.

The devices are tested for usual STP. Nothing else, unless specified. They are done so not for human comfort, but for tested conditions. No more.

If you want extreme conditions, pay through the nose.

Helium, unlike carbon dioxide, is lighter than both oxygen and nitrogen, so it would rise in the rooms and be very unlikely to displace oxygen in people and set off any oxygen alarms.

In addition, unlike carbon monoxide, helium doesn't bind to, well, anything and certainly not hemoglobin, so it doesn't present a momentary exposure hazard either.

See other reply -- I wasn't referring to danger to humans.

What the heck? I clarified that I was concerned about the silent danger to the iPhone, not to humans, and people keep assuring me there’s no danger to humans, and giving me flack just for saying that wasn’t my concern? This is a weird thread.

Quoting you:

> Yeah, but at the same time, it wasn't enough to set off the alarms for low oxygen content, which is at least a little disturbing, right?

First, alarms for low oxygen content are set for people and no other reason. There can be alarms for high oxygen content, but they are extremely rare and would likely not be present in very many places in a hospital.

Second, why in the world would low oxygen be a danger to an iPhone?

You can't make a complete non-sequitur of a logic leap and then expect everybody to be on board your train of thought.

I wasn’t comaplining about that, but for getting to -3 for pointing to a reply where I clarified the concern.

Furthermore, comments were rolling in assuring me of the nonthreat to humans after I clarified.

You called it "disturbing", so the clarification that you were not intentionally implying any danger whatsoever is kind of annoying.

“Device disabling threats that aren’t detectable by the usual sensors” can’t count as “disturbing” because a human didn’t die?

"threat" the way UV light is. So no, not disturbing for it to silently ignore the equivalent of popping a party balloon.

Bricking your iPhone for a week-plus is not the equivalent of popping a party balloon or a short dose of UV light!

And yes, I get it, the sensor exists to detect for threats to life, not threats to iphones. But that's the point: this is a threat we're not set up to watch for at all -- hence why it took so much investigation to root-cause it!

Are you still going to make it your hill-to-die-on that it's "not disturbing"?

Come to think of it, did that "lack of faith" scene in Star Wars (1977) also seem confusing to you?

> Bricking your iPhone for a week-plus is not the equivalent of popping a party balloon or a short dose of UV light!

You totally misunderstood my comment.

I mentioned popping a party balloon because it could also disable your phone. And UV light because it can also destroy certain electronics. But sensors that can pick up little bits of UV are normally not set to warn about it, and that's perfectly reasonable and not "disturbing".

We can use a different word if you want. You think it's a [significant] problem and I don't think it's a problem. Is that better?

I'm not here to die on the hill of word choice. I disagree with your underlying opinion.

Sorry -- do you have a citation on that? If popping party balloons could disable iphones, I think we'd hear about it more often. (I can't find anything about it on a quick search, and I hope I can be excused for missing an analogy that depended on an invalid premise!)

What made it unnoticed for so long was that it required the much larger MRI He release to trigger. What (I claim) makes it disturbing is that an expensive incident can happen (bricking a hospital's iphones for weeks) at levels that no one currently calibrates sensors to check for, because no one expects that level to have a negative effect of such magnitude.

Regardless of your preferred terminology, it seems odd to take an attitude of "oh, it only deleted everyone's second-brain for week, in a way that was hard to root-cause, no one died, no big deal" and therefore balk at the use of "disturbing".

I mean, do you consider undetectable, inexpensive phone-bricking techniques to be generally "non-disturbing", or just this one?

> Sorry -- do you have a citation on that? If popping party balloons could disable iphones, I think we'd hear about it more often. (I can't find anything about it on a quick search, and I hope I can be excused for missing an analogy that depended on an invalid premise!)

Some of the tests were just plastic bags with tiny amounts of helium in them. Unless you're asserting some exceptionally strange gas physics here, you can break an iphone with some balloons.

> Regardless of your preferred terminology, it seems odd to take an attitude of "oh, it only deleted everyone's second-brain for week, in a way that was hard to root-cause, no one died, no big deal" and therefore balk at the use of "disturbing".

> I mean, do you consider undetectable, inexpensive phone-bricking techniques to be generally "non-disturbing", or just this one?

Hang on. You were calling the calibration of the air quality sensors disturbing. That's what I objected to. I wasn't making any claims about the iphone bug itself.

And nobody was talking like Donald Duck either so I would guess it was leaking out of the room at a pretty hefty rate.

Depends on where the sensors are located. If at head height, the helium may occupy the layer above and tall people would have a hard time.

Sadly, the plant used Apple's iSensors which don't work in the presence of helium.

What a relief that must have been.

Another scary practice is venting natural gas indoors. Most of the time it's fine, but sometimes you get a catastrophic explosion: https://www.youtube.com/watch?v=rjxBtwl8-Tc (The US Chemical Safety Board has a bunch of interesting videos.)

This is nuts! Why would you even vent flammable gases indoors on a construction site?!

Because when you initially pressurize a newly installed portion of pipe, it's not filled with natural gas, it's filled with air. Air doesn't burn, something like a natural gas burner with a pilot light is going to detect that the pilot light went out and shut off the gas flow to prevent an explosion. That leads to a sort of catch 22, you can't purge the line without the burner going and the burner won't turn on until the line is purged.

Ideally you would pressure test and purge new work with nitrogen as well in order to prevent a potentially combustible mixture from forming inside the pipe while the air is being purged from it.

As to the safety of purging natural gas directly indoors, obviously that carries some risk but it's not as dangerous as you might assume. Below a certain concentration (5% by volume) natural gas mixed with air actually won't burn freely. Also above 15% natural gas by volume won't burn freely. The risk of an explosion is only present in that 10% range. If you're putting in a new gas stove in your kitchen and you crack open the valve with the hose off of the stove until you smell gas, that's not a ton of natural gas that you've let out into a room with a relatively huge volume. The pressure of a natural gas pipe in a house is actually very low, only around 0.25 psi. It's coming out slow enough that you've got plenty of time to turn off the gas before you come anywhere close to hitting that lower explosion limit where things get extremely dangerous.

They thought they were venting outside. It only went inside because of an undetected leak.

As long as he helium doesn’t displace the oxygen it is fine. Saturation diving has people living in “heliox” environments where the atmosphere is helium and oxygen.

There are more details in the original thread https://www.reddit.com/r/sysadmin/comments/9mk2o7/mri_disabl... on reddit:

I discovered that the helium leakage occurred while the new magnet was being ramped. Approximately 120 liters of liquid He were vented over the course of 5 hours. There was a vent in place that was functioning, but there must have been a leak. The MRI room is not on an isolated HVAC loop, so it shares air with most or all of the facility. We do not know how much of the 120 liters ended up going outdoors and how much ended up inside.

Except that Helium is not dangerous whatsoever and to actually displace oxygen you would need much much more than that.

Venting 120l of liquid helium (or about 84,000l of gaseous helium) really isn't that much of a health concern

I've seen multiple vents of ~1500l of liquid helium (~1,050,000l of gaseous) in a single day from a couple bad MRI units when I worked at a company that makes them

Only "dangerous" place to be was in the fragmentation path of the burst disc (which was enclosed in the top of the unit (at the time)), or directly in the path of the nearly-liquid (ie extremely cold) helium as it vented

It dissipates extremely rapidly

Yea, that sounds incredibly dangerous considering how easily you can suffocate inhaling helium

how easy is it? It floats, so it only fills a room from the top down... and if you fall over unconscious, you're automatically moving to a part of the room with more oxygen. it would have to displace all of the air in a room in order to suffocate someone, no?

It's fairly difficult, but definitely possible. It's hard to recognize because what is killing you is the lack of oxygen. People generally don't notice they are becoming hypoxic.

But there are plenty of things to help you out here. First is that helium is very light. So if you pass out and fall down you will be in a lower concentration of helium (assuming you are not in a sealed room). (Labs with heavy gases have vents on the floor) If the room is reasonable ventilated you should be fine. It dissipates fairly quickly. Helium permeates through most things, but this is a slow process so probably won't save you.

I meant more that you don't feel like you are suffocating when you are inhaling helium, so it can be hard to realize something is going wrong until it is too late. Definitely easier to suffocate with heavier-than-air gasses though.

Multi-story building. The floor could be, given enough gas, under.

If helium is small enough to get in a watertight phone, it is certainly small enough to dissipate in a probably not so watertight building.

Quenching MRI magnets can suffocate people in the room they quench, and potentially the entire floor. I don’t know about a whole building, but just because something can dissipate doesn’t mean it can’t do damage in the meantime.

Not true - the whole point of burst discs etc in MRIs is so that if a quench happens, the patients and any technicians in the room survive

Venting even a full MRI's worth of LHe (~1500l) as most will make you a little lightheaded for a few minutes

The risk of death from quench events is effectively non-existent

If you fall over unconscious from helium then you'll probably die without immediate help. Your natural breathing reflex is triggered by the presence of CO2, not the lack of oxygen. If you inhale a lot of helium and pass out, your body doesn't start breathing again because there is no CO2, so you just suffocate instead of inhaling oxygen.

You're continuously producing CO2 whenever you're alive. If you fall down and land in a normal atmosphere then your normal breathing reflex will save you. Low oxygen atmospheres don't stop you from breathing; they're dangerous because you continue breathing as normal. The CO2 is carried away from your lungs with each breath, so there's no feeling of suffocation.

> If you inhale a lot of helium and pass out, your body doesn't start breathing again because there is no CO2, so you just suffocate instead of inhaling oxygen.

Other posters have pointed out that this is false(you will continue producing CO2 while unconscious just fine, chemoreceptors will notice). I would like to add that I have personally witnessed people falling unconscious from helium. They recovered just fine.

Really, the greatest danger is the fall.

you don't breathe because there is CO2 in the air, you breathe because there is CO2 in your blood

If that were true then divers would have to bring canisters of CO2 down with them.

Well helium rises. So falling down is a good thing, as most vents are above you. If you go to a lab with heavy gases you'll notice vents on the floor for exactly this reason.

Right! It's not like an "allergy" to helium is going to be a problem for any user at any time unless, you know, there's a big helium leak, at which point the phone is the least of your problems

Wouldn't this affect people's voices?

Eventually. But I would expect that by the time the effect is noticeable, the concentrations will be too high to support consciousness and will have been for some time.

I'm not sure that's true.

If you displace 30% of the atmosphere with helium, the O2 concentration drops from 21% to 15%. That's low enough that you'll probably start to see mild cognitive impairment, but it's still plenty capable of supporting life. However, a 30/70 mixture of helium and air is about 67% as dense as normal air, which means voices would be pitched more than half an octave higher.

Considering the current helium shortage it's a shame they just wasted it.