skip to primary navigationskip to content


Arguably, liquefied helium and nitrogen pose the greatest threat to life in this Department. This threat exists not just for those who conduct experiments, but all who use the premises. Therefore everyone should understand the nature of the risks and what to do if things appear to be going wrong. At the very least, everyone should be familiar with the instructions on this page

A printable code of practice is available here.


This page gives advice on the control of risk from liquefied gases (principally helium and nitrogen), solid CO2 and the piped supply of nitrogen that comes from the main tanks.

On this page you will find


Effects Of liquid gases

  • Asphyxiation
  • Changes in material properties
  • Condensation of water
  • Condensation of liquid oxygen
  • Blockage of outlets
  • Burns, Frostbite

Legal Requirements

Accident Case Histories

Risk Assessment: Step 1 - Assessing The Hazards

  • Asphyxiation
  • Other properties
  • Solid carbon dioxide

Risk Assessment Step 2: Assessing The Risk

  • Asphyxiation
  • Assessing the risk from the other hazards
  • Carbon dioxide

Reducing The Risk

  • Asphyxiation
  • Damage to insulation and other materials, condensation of water
  • Condensation of liquid oxygen
  • Blockages of outlets to apparatus
  • Burns

Contingency Planning



The Department of Physics uses significant quantities of liquid nitrogen and liquid helium and also has a network of pipes throughout many of the buildings delivering gas from the tanks. This represents an effectively unlimited supply of asphyxiant gas. The following key risks are therefore present in the department:

  • Asphyxiation - due to the displacement of air,
  • Damage to the insulation on electric cables - which could lead to electrocution,
  • Condensation of water onto nearby electrical equipment,
  • Condensation of liquid oxygen onto the pipework, etc, when drawing off liquid nitrogen or helium - leading to a fire risk,
  • Blockage of outlets from cryogenic containers, leading to a risk of explosion,
  • Burns - due to touching cold metal, getting clothing soaked with cryogen, or direct contact with the liquid.

(this is not an exhaustive list!)

Effects Of Liquid Gases


You will find tables listing the symptoms and signs of reduced oxygen on the human body. Please treat these with extreme caution as they are misleading. The onset of asphyxiation can be rapid, and there are likely to be no obvious warning signs to the victim. There are no physical symptoms of distress, and the function of the brain is seriously impaired in an oxygen deficient atmosphere, without the person's knowledge. It can take as little as two breaths in an oxygen deficient atmosphere to cause unconsciousness, and death will occur within minutes. Below around 10%, death or brain damage may be inevitable, even if the casualty is rescued and resuscitation is attempted.

The atmosphere normally contains approximately 21% oxygen by volume. Atmospheres containing less than 18% oxygen are potentially dangerous, and entry into atmospheres containing less than 20% oxygen is not recommended.

Accident records show that, for almost every person who is overcome and dies in an atmosphere that does not support life, an additional person dies, unsuccessfully attempting a rescue.

It is probable that a visible vapour cloud is significantly depleted in oxygen. Personnel should be instructed never to enter a vapour cloud.

Changes in material properties

We have all seen people smashing rubber hose, etc, when doing demonstrations with liquid nitrogen. In polymers this is primarily due to the fact that cryogenic liquids take the material well below its glass transition temperature, into a regime where it is brittle. It recovers if it is warmed up - but it may break if it is disturbed while cold, or if it contains residual stresses. The most tragic demonstration of this change in material properties was in the loss of the Challenger in 1986.

Floor tiles crack, and may come away from the floor. These can cause trip hazards, and in any case are unsightly. Carpet tiles appear to be just as vulnerable as lino. Pneumatic tyres on vehicles that are exposed to liquid nitrogen may explode.

Carbon steels (i.e. not stainless) are body centred cubic, and also become dramatically more brittle at low temperatures. They are not suitable for low temperature service.

The drains cannot cope with liquid nitrogen - the materials can be damaged - expensively. If the damage is not immediately apparent, think what might happen if someone puts a chemical down the drain - where might the leaked material go to?

Condensation of water

Most of the time, the air is well laden with water vapour, and in consequence, wherever you are using cryogens, you will have water vapour condensing out and often forming significant ice burdens on pipes, etc. This can give rise to a risk of electrocution, or items being broken under the weight of the ice. Water may drip onto precious equipment.

Condensation of liquid oxygen

Since oxygen has a higher boiling point than nitrogen, it can condense into liquid nitrogen. Fortunately it is a light blue colour, so if it has condensed in a significant quantity into an open Dewar of nitrogen it is obvious. Be aware that liquid oxygen is a serious fire hazard, and pure oxygen can cause organic materials to ignite spontaneously or explode.

What is less well known is that the liquid that forms and runs off a nitrogen delivery tube from a Dewar when you are discharging nitrogen is also very rich in oxygen. Hence smoking is prohibited at nitrogen transfer points. Liquid that is allowed to dribble onto clothing can take a very long time to disperse, and the individual is in considerable danger while their clothing is rich in oxygen - do not get the liquid on your clothing, and do not go out and smoke immediately after conducting a nitrogen transfer.

Blockage of outlets

Liquid helium can readily cool air below its freezing point, and solid air is quite 'sticky'. It can block up bits of your apparatus (the author speaks from experience!). If the apparatus becomes blocked, then it is possible that an explosion may result, because the helium will evaporate before the blockage clears itself.

Blockages can form with nitrogen as well, if ice forms in sufficient quantity. Dewars must not be left outside when it is raining.


The standard solution to the problem of requiring an outlet that will not block is the Bunsen valve - a piece of rubber tube with a slit in it, cheap and very effective for nitrogen Dewars. Helium Dewars must be connected to the laboratory helium gas return line.

Burns, frostbite

Liquefied gases, particularly when soaked into clothing, can produce burns that are similar to heat burns. Unprotected skin may also stick to un-insulated items and flesh may be torn on removal.

More prolonged contact can cause the flesh to freeze. While frozen it will appear yellow and waxy, and will probably not hurt, but when it thaws, it is likely to give rise to intense pain. Such burns require immediate medical attention.

Persistent, superficial contact with liquid nitrogen can also give rise to chilblains - itchy, red, irritated patches of skin on the hands.

Breathing in the cold gas can trigger asthmatic attacks in susceptible individuals.

Legal Requirements

There is no legislation specific to cryogens. However, there are three basic legislative requirements:

  • to ensure so far as is reasonably practicable, that people are not exposed to risks to their health and safety
  • to undertake a risk assessment to define how this will be achieved
  • to be aware that, under certain circumstances, a room or area may become especially dangerous to life through lack of oxygen, to detect when this may happen, and take steps to prevent people from entering the danger area.

Accident Case Histories

One research facility 'lost' a number of the old narrow-necked Dewars (the tiltable ones that were mounted on wheeled trolleys). It turned out that each weekend they were taken out of doors. Some still had small quantities of nitrogen in them. It is surmised that the necks of the Dewars were being blocked by ice, and at some time over the weekend they were exploding!

In 1999 there was a fatal accident in a research establishment, where a technician was filling Dewars. He had around ten years experience. The boil-off during the filling process was sufficient to reduce the oxygen concentration in his room below that which would support life. His colleagues, who attempted to rescue him, nearly suffered the same fate and one was very lucky to survive.

In a workplace where they had both piped nitrogen and piped air, someone accidentally connected some breathing apparatus to the nitrogen line. A workman donned the breathing apparatus and immediately collapsed. He could not be resuscitated and died.

A workman put his head into the opening of a tank in which his colleagues had been working. He took a couple of breaths of the air inside the tank, and died, even though he only had his head in the opening.

These incidents should be borne in mind when planning your contingency arrangements.

Risk Assessment: Step 1 - Assessing The Hazards


Record the identity of the cryogen, what you are using it for, and the maximum quantity of liquid gas that you have in a single container in the room (whether inside the apparatus or inside the Dewar).

Calculate the volume of the room. Then calculate the minimum oxygen content, based on the assumption that the cryogen from a single container escapes and simply displaces air from the room. Nitrogen expands to 680 times the volume of liquid, and helium expands to 740 times the volume of the liquid.

If the cryogen is piped into the room, then record the maximum quantity that could be accidentally dumped in the room, and calculate the oxygen concentration in the same way.

There is a network of pipes bringing nitrogen gas from the liquid nitrogen tanks into many of the labs. Since the tanks contain many thousands of litres of liquid, this means that the supply is effectively unlimited. Thus, it is important to identify whether your room has such a supply and how it is being used.

There is a useful spreadsheet calculator that will do the calculations here. Simply select the correct tab for the gas you have, and enter in the data in the yellow boxes. If the result of your calculation is 18% or less, there may be a problem. For piped nitrogen, use the tab marked 'gas release'.

The spreadsheet model is limited. It assumes that the inert gas has leaked into the room, displacing air in equal amount, and that the inert gas has mixed with the air that is left. In reality, the atmosphere close to the item that is leaking is probably depleted in oxygen to a far greater extent than the above calculation would indicate. One could also argue that cold nitrogen will tend to accumulate near the floor, and helium above the head. However, if the gas is released rapidly, there will be turbulence and mixing.

Now that you have done the calculation for the room, please repeat it for the corridors where you leave the large Dewars, and the spaces through which cryogenic liquids are piped.

Consider now what will happen - how will the air be replenished? How quickly will it be replenished? Is the position the same at night, or at the weekends?

All of this can be recorded on a form that has been designed for the purpose: Word document; pdf document.

Other properties

These are usually fairly obvious, and can be seen from the layout of the equipment - in other words, there is a risk if the electrical cables are on the floor close to the nitrogen top-up tube, etc.

Have a good look at the way in which jobs are done, to spot the other sources of hazard that crop up.

Solid carbon dioxide

Carbon dioxide is toxic. While you might also suppose that it is an asphyxiant, high levels of carbon dioxide in the atmosphere will be fatal long before it reaches a value where it threatens the oxygen supply to your brain.

Risk Assessment Step 2: Assessing The Risk


The calculations done so far give you a simplistic account of the intrinsic hazard. You now have to work out the risk that the dangerous event might happen. To do this you need to work out how the cryogen could be released in that quantity. You need to consider such events as:

  • Fracture of a pipeline
  • A pressure relief valve sticking open
  • The loss of vacuum in an insulating jacket
  • Knocking over the helium or nitrogen Dewar
  • A magnet quenching
  • Opening the wrong tap on the nitrogen Dewar and leaving it open, etc.

Many of these have already happened at some time. Many of them pose a real danger to life.

Some of these will be sudden and spectacular, and if someone is present when they occur, they will perhaps react to them. However, the event may happen at night, or it may happen to a person who is injured or knocked unconscious. What will happen then? What if a Security guard is the first to find the problem?

While helium is lighter than air, and nitrogen is marginally denser, do not assume that they will be released in neat layers. If there is turbulence, they will mix, and once mixed, they will not separate again (you can buy cylinders of 5% hydrogen in nitrogen - they will be uniformly mixed indefinitely!)

In general, you only need to worry about a single failure mode - e.g. if you have two magnets in a room, only one may quench. Clearly if there are likely to be two failures that will be linked by cause and effect you need to consider them both.

In the case of potentially serious spills, such as the possibility that the contents of the storage tank will empty themselves via a hole in the pipeline, you would be wise to take into account the possibility of multiple failures, either in equipment or the actions of personnel. When the outcome is of a possibility of multiple deaths, you need to take this step.

If a particular sequence of events has happened once, then it becomes a known cause, and must be taken into account and planned against. For example, if it is known that someone has overfilled and over-pressurised a system in the past, whether it was due to utter carelessness on their part or not, it is something that you have to guard against in the future. (quote from a judge 'you have a duty of care not just to the careful and diligent worker, but also the lazy and indolent').

Other hazards

Nothing beats a good inspection to spot how the cryogen can give rise to problems. Write down your findings, and plan to reduce the risk so far as is reasonably practicable. Listen to the folk-lore of the area - what has gone wrong in the past? You cannot dismiss a risk because someone did something idiotic - you must expect this kind of behaviour from newcomers and those who are becoming over confident, or wish to cut corners.

Carbon dioxide

The gas calculator has a page for carbon dioxide. The data can be handled in the same way as the nitrogen or helium data.

Reducing The Risk


Now that you have the basic information about your system, you need to make some decisions. The hierarchy you should adopt is as follows:

  • Remove the problem altogether, which may be achieved by having less cryogen in the space (e.g., not leaving the Dewar in the room)
  • Ensure that cryogen from your apparatus is vented to a space that is large enough to accept it without threatening the oxygen level.
  • Ensure the ventilation level is adequate to disperse the gas (but check on the status of the ventilation system at night or at weekends).
  • Reduce the probability of accidental release by ensuring the stability of Dewars, maintaining the floor in good condition, training your personnel, using labels and diagrams, etc, to remind the forgetful.
  • Impose procedural control - such as prohibiting the transport of cryogens in a lift along with personnel, prohibiting the transfer of cryogens at night, unless at least two people are present.

The second step in the hierarchy is dealing with a risk that you cannot engineer out. e.g. Where you believe that it would be advisable, place oxygen monitors in the area, or give out personal oxygen monitors. You may choose this on the basis of, say

  • the sheer quantity of cryogen involved,
  • a high risk activity such as cryogen transfer,
  • the possibility of an unforeseen leak,
  • the possibility of a person being forgetful, or
  • the possibility of multiple failure.

You may, for instance, decide to interlock the oxygen monitor to an emergency ventilation system.

If you invest in oxygen monitors, do not forget to put in place a programme of maintenance and inspection. They are no use if they are not calibrated, or their batteries are flat!

If your hazard is carbon dioxide, remember that an oxygen monitor will only tell you of the danger AFTER the atmosphere has reached a fatal level. Invest instead in a carbon dioxide monitor!

Damage to insulation and other materials, condensation of water

This needs good experimental design. Ensure that cables are not placed where they are likely to be affected by liquid nitrogen spills. This may demand some ingenuity for the leads to the apparatus, and you may have to site the leads off the ground.

Condensation problems can be dealt with in the same way. Make sure that ice and condensation cannot lead to water mixing with electricity.

Train people not to pour liquid nitrogen down the drains, or over the floor tiles (it's good fun, but the floor does often get irrevocably damaged). Ensure that when you design widgets for cryogenic service, you choose suitable materials.

Condensation of liquid oxygen

Train people to avoid getting oxygen-enriched liquid on their clothing. Ensure there is a 'no-smoking' policy at places where cryogens are transferred.

Blockages of outlets to apparatus

Your apparatus is likely to have several over-pressure prevention measures such as safety valves or Bunsen valves for atmospheric pressure release, which will have been put in at the design stage.

Management of this risk is largely by training - people need to become familiar with the system that they are using, and this takes time. Don't let a newcomer cool down apparatus on their own until you have done it with them several times, ensured that they understand the layout of the system (a diagram and written instructions would be helpful!) and know the danger points in the procedure, and what to do if something goes wrong.


These usually arise when pouring nitrogen or transferring helium. Ensure that you train people about the risk of getting frozen to the pipework, etc.

Make sure the gloves you issue are the right type - loose fitting cryogenic gloves. It is no good giving people gloves that soak up nitrogen! Eye protection is essential when transferring cryogenic liquids into open buckets.

Ensure that people wear the gloves and goggles - it's important.

Contingency Planning

Even when you have done your best to devise a safe system of work, something can go wrong. It is important to ensure that everyone knows what to do.

What if?

  • You come across a vapour cloud?
  • You hear an oxygen meter alarming?
  • A Dewar falls over?
  • Someone collapses in a laboratory?
  • Liquid nitrogen is spilling uncontrollably from a Dewar?
  • Someone trips and breaks their ankle while carrying a liquid nitrogen container?

Some pointers (not exhaustive):

  • It is probable that a visible vapour cloud is significantly depleted in oxygen. Personnel should be instructed never to enter a vapour cloud.
  • People should know where the portable oxygen monitors are - they may need them.
  • If there are pipelines that have an effectively unlimited source of cryogen or asphyxiant gas (e.g. nitrogen delivery, helium return) the layout of the system needs to be demonstrated to people, along with some sort of warning system to alert people to the failure, and many people need to know how to isolate the problem - NOT JUST THE OWNER OF THE EQUIPMENT! Remember that the first person to find your equipment in a state of failure may be a Security Guard.
  • Evacuation procedures need to be thought through - both from the point of view of getting people out safely, but also keeping them out until it is safe.
  • Personnel need to know how to get a first aider quickly and efficiently.
  • People need to be trained to cope with a collapsed casualty, so that they do not become the second casualty.

Don't assume that an oxygen meter will automatically save people's lives. You still need to think through an action plan and train people. Otherwise, they may simply turn the meters off, to stop the horrible noise that they are making!

A sheet of emergency and first aid instructions is here.

This page was last updated 12 March 2015