Latest regulations, industry case studies and unbiased articles written by experts
Latest regulations, news and unbiased articles written by experts
The best articles written by experts
The Journal for Employee Protection
  • Latest Issue
  • Trending
  • Press
  • Videos
  • Events

... All I Need is the Air that I Breathe

Published: 10th Jan 2004


Here’s a paradox. Situations where there is the greatest danger from inhalation hazards need the most complex forms of respiratory protective equipment (RPE) – breathing apparatus (BA).

The more complex such a device is, the more potential there is for some failure in one or more of its functions. The consequences of a failure in these situations are inevitably very serious. This is a vicious circle of escalating risk and consequence, which needs careful consideration and control.

This article describes some of the main stumbling blocks which the Health and Safety Executive (HSE) have identified in the investigation of workplace accidents, and during the development of guidance on the selection, use and maintenance of RPE. The lessons reported here are useful for all those who currently use, or are considering using, BA.

What is BA?

BA is any form of RPE which receives its breathing air (or less commonly other gas mixtures) from a source independent of the atmosphere immediately surrounding the user. BA includes fresh air hose and compressed airline supplied equipment, and selfcontained BA (SCBA) incorporating breathing gas storage or generating components. Perhaps the most familiar form of BA is that used by firefighters, with an air cylinder carried on the back and a full face mask, but BA doesn't have to be this complex. BA excludes any form of device which cleans the air surrounding the wearer using filters; these are respirators, and are a different story altogether.

The advantage of BA over respirators lies in its ability to be used against virtually any and all contaminants and in oxygen deficient atmospheres where filtering devices would be totally inadequate. It avoids all the difficulties of filter selection for use against unknown challenges or mixtures of hazardous substances, and of prediction of safe filter life. But there is of course a price to be paid for this advantage. There are snakes as well as ladders in this game - more on this later.

As with other forms of RPE, there are many types and classes of BA available, providing varying levels of protection, which in the UK is classified in terms of the Assigned Protection Factor (APF - see BS 4275:1997 for details). It is this APF value which should be used when selecting BA capable of controlling exposure in a given situation. Table 1 summarises the common types, based on their current European Standard classifications.

The main factors controlling the level of protection provided are the type of facepiece, and the way in which it is supplied with breathing air, but not all types of facepiece can be used with all types of supply. The penalty you pay for not having to worry about filters is that breathing air quantity and quality must be closely controlled.

For any equipment with a tight-fitting mask (half or full face), it is a pre-requisite that the mask should fit the wearer without leakage to be able to apply the APF values quoted in Table 1. How this can be demonstrated is described later.

APF Fresh air hose (EN 138 EN 269) Light duty airline Mask (EN 12419) Light Duty Airline Hood (EN 1835) Constant Flow Airline Hood (EN 270/271) or mask (EN 139) Demand Flow Airline Mask (EN139) Self-Contained
10 Half mask LDH1
20 LDM1, LDM2 LDH2 Half mask
40 Full mask, Hood LDM3 LDH3 Hood, Blasting Helmet, Full Mask Negetive Demand, Full Mask Negetive Demand, Full Mask
200 Suit
2000 Positive Demand, Full Mask Positive Demand, Full Mask

Table 1. Types and classes of BA

What can go wrong?

Based on the 122 investigations we have carried out where the various types of BA have been implicated in dangerous occurrences, failures and fatalities, the root causes fall into a relatively small number of categories (see Figure 1), although some come under more than one heading. The great majority of these incidents (80%) involved positive demand SCBA. This is probably a combined consequence of the complexity of this equipment, the levels of awareness necessary for users, and the extreme environments to which it is often exposed.

Deficient maintenance

In roughly one third of all these investigations, problems arose because the equipment involved had been subjected to inadequate, incorrect or inappropriate maintenance and cleaning regimes. These ranged from total neglect (Figure 2), through not rigorously following cleaning and maintenance procedures, to wholesale replacement of specific components with inappropriate alternatives (Figures 3 and 4), which were simply not up to the job. Components that are particularly vulnerable in this respect are mask exhalation valves, and o-rings in high pressure parts of air supply systems, both of which have caused serious failures in use.

Incorrect use

This category makes up about a quarter of the incidents studied, and includes instances where correct procedures have not been followed, through lack of knowledge, inadequate training, or simple human error. These include something as fundamental as not fully opening, or mistakenly closing off, the valve on a BA cylinder (understandable when you consider that the valve must be operated when it is out of sight, mounted on the wearer's back), or failing to recognise that a 'free flow' condition on a demand BA set was due to the supplementary supply being turned on - this deliberately overrides the breath-responsive function of the demand valve and continuously flushes air through the mask instead.

Design and manufacturing faults

These are fairly infrequent, but they do continue to occur. Usually they can be attributed to temporary lapses in production quality control - changes in the specification of raw materials which go unnoticed, bad batches of adhesive, or imperfectly moulded or machined components. Only when something goes wrong during use or testing do they come to light. In such cases, involvement of the enforcing authority has worked wonders for getting matters sorted out quickly.

Quantity / quality of breathing air supply

Six percent of investigations identified poor quality or inadequate supplies of breathing air as a contributing factor. These include not only contamination with dirt, oil or water (in some instances to such an extent that flow control valves became clogged and unable to provide enough air for the user), but also connection to non-breathable gas supplies, such as nitrogen purging lines, instead of air - there are a number of recorded instances of this.

The effects of inadequate quantity of breathing air differ, depending on the type of face piece being used. With visors, hoods, helmets and blouses, low air flow may initially go unnoticed, but results in inadvertent exposure to the airborne contaminant, the consequences of which may not be apparent for some time - possibly years. For tight fitting masks, the acute distress of not having enough air to breathe usually forces the wearer to take the mask off - in several instances this led to instant collapse, as the BA was being used in oxygen deficient environments.

No fault found

In twenty of the incidents studied, no faults with the performance or operation of the equipment were detected when it was studied in our laboratory. In many of these cases this came as no particular surprise, for example:

  • Escape BA which was available but not used when a worker was overcome by a build-up of nitrogen gas during generation of artificial snow indoors
  • A man collapsed while wearing a closed circuit BA set during a mine rescue. Ambient conditions in the mine were extremely hot and humid, and the work was strenuous. The cause of the collapse was heat stress, and was not directly related to the BA set
  • Several instances where BA was being worn during fire fighting, but wearers were overcome by other events - collapsing walls or explosions

The remaining half dozen or so instances of 'no fault found' are largely unexplained. It may be significant that several of these were reported as 'free flow' or 'smelt smoke' while wearing positive demand BA. Both of these events could be attributable to a mask that did not fit an individual well, while there is actually nothing wrong with the BA itself.


Clearly, maintenance and servicing of BA is vital for ensuring the safety of the wearer, and is required by law. BA servicing may be done in-house or it can be contracted out. Increasingly, this is carried out under a 'Total Care Package' (TCP). The driving force behind such packages is economy; it is generally cheaper to contract out this function to specialists than to provide the necessary staff and facilities in-house. However, there are hidden disadvantages to this approach unless it is managed effectively.

One issue that may cause concern arises when the TCP is provided by the manufacturer of the BA. If faults are identified during maintenance, will they be notified to the user, to the Notified Body who certified the equipment, or indeed to enforcement authorities for possible investigation and enforcement? While we have to have faith in the integrity of manufacturers, we cannot help feeling that a large body of interesting and important information may well not come to light. This seems to be borne out by the falling numbers of BArelated incidents which are being reported to us.

In order for a TCP to work effectively, the user must know what it is that they need from the service, and what full and complete records must be provided, so they can check and audit the system. They cannot just hand responsibility over to the service provider and expect them to take care of it all - the legal responsibility ultimately remains with the employer of the person who is required to use the BA. Employers have to be fully involved at all stages of setting up the contract and in reviewing it on a regular basis to ensure it meets their needs. One of our investigations highlights some of the possible pitfalls of this type of system:-

During a training exercise, a firefighter was attempting to rescue a dummy via an upper floor window, and his BA mask became trapped against the window frame. The visor was forced out of the mask body. This was reported to us and we investigated. The brigade had recently moved over to a TCP provided by the BA manufacturer. Our investigation showed that a small retaining screw in the visor clamp was missing, allowing the visor to eventually dislodge from the mask body. How long the screw had been missing could not be determined, but records were requested from the manufacturer to see if any recent servicing had taken place and whether any problems had been highlighted.

The initial response was inadequate in that the records did not show what components had been replaced at the last service. In addition, the set in question appeared to be made up of parts from two different BA sets; identification numbers were different on some of the components. Neither the brigade nor the manufacturer could explain this satisfactorily. Subsequently, further records were produced detailing the replaced components. They were highlighted on the reverse side of the sheet that had been initially presented, but had not been photocopied at the time of the request.

In this case, both the brigade and the manufacturer had been somewhat lax in their running of the system. The brigade was unsure of what they wanted from the deal, and the manufacturer provided incomplete information. Human error was not identified when parts from two different sets were put together, casting doubts on the effectiveness of the audit system, and suggesting that incorrect records may be being kept. Discussions with the brigade highlighted our areas of concern and they subsequently took these forward with the manufacturer.

Air quantity and quality

Airline and powered fresh air BA must have a means of checking that the air flow rate is adequate before use. But ask yourself when you last saw this flow checker, or anyone using it. Higher protection BA tends to have warning systems (e.g. low pressure/flow whistles) built in; these can't get lost and are hard to ignore.

Compressed air quality should be assured. Ideally, compressor systems should monitor and control quality continuously (systems are available which claim to do this), but in practice assurance is often achieved by testing the air at least once every three months, or more frequently when the quality of the air supplied is likely to vary. For mobile compressors, air quality should be assured just before the first use in any new location - either by installed engineering controls or by testing. For fuel driven compressors, make sure that exhaust gases cannot be drawn into the breathing air intake - this was among a catalogue of faults in one of our investigations. In all cases, BS EN 12021 (1999) should be complied with for lubricants, odour and taste, O2, CO2, CO and moisture. It is not reasonably practicable, or necessary, to test for all contaminants; the employer should use the risk assessment required by health and safety legislation (e.g. COSHH in the UK) to decide what other substances should be tested for.

Water content of compressed air often gives rise to queries - why does it have to be so dry? The limits set are largely to preserve the integrity and function of compressor systems and any attached BA. Condensed moisture leads to corrosion, which can weaken pipes and tanks, and also deplete oxygen levels in stored air. When compressed air flows through a restricting valve, it produces cooling. If the air is moist, this can lead to condensation and freezing, possibly cutting off airflow. You don't want this to happen in a critical life-supporting system like BA. Accumulations of condensed water can act as breeding grounds for micro-organisms, possibly leading to oxygen depletion, smells, exposure to toxins, allergens and infectious organisms including legionella.

The effects of breathing dry air are minimal for healthy people; the Tuareg and the Innuit do it all their lives. Discomfort due to drying of the upper respiratory tract linings is simply relieved by occasional breaks and liquid intake. Only for those with preexisting respiratory illness (e.g. asthma) is there likely to be any difficulty, but it is questionable whether these people should be required to use BA anyway.

Fit testing

For any form of tight-fitting facepiece, the wearer should ensure there is a good seal between their face and the mask. This is essential to achieve the optimum level of protection. This is most obvious for 'negative pressure' devices like unassisted fresh air hose and negative demand masks. Because these need a pressure below atmospheric inside the mask to pull breathing air in, a bad fit will inevitably lead to inward leakage. Less obvious is the fact that continuous flow BA systems can also leak in this way, when supply is unable to match peak inhalation rates during heavy work.

It is sometimes argued that positive pressure demand BA doesn't need to fit the wearer particularly well, because if there is a leak it will always be outwards. There is an element of truth in this - for a given size of leakage path, the exposure consequences to the wearer will be lower for a positive demand system than for any other. However, it is not true that such leaks are unimportant:

  • In-mask pressure can still briefly drop below that outside, until the demand valve has time to respond
  • If there is a leakage path, the demand valve will attempt to maintain in-mask pressure above that outside the mask, by supplying air. For self-contained sets, this depletes the reserve, reduces expected working duration, and could lead to the wearer running out of air prematurely. For airline BA, the increased consumption of compressed air also boosts operating costs for compressor systems

Advice on UK regulations concerning the use of RPE and BA recognises the importance of face fit for masks. HSE Information Document 282/28 ( provides details on how to go about demonstrating that a given mask fits the person who has to wear it. In general half mask fit can be demonstrated by either qualitative or quantitative tests; full face masks require a quantitative test.

Recently, the UK fire services inspectorate, in collaboration with HSE, have developed an enhanced donning routine for their users of positive pressure demand BA, as an alternative to a quantitative fit test. Based on available laboratory evidence, this routine is able to detect whether there are any significant leaks in the air supply system, including at the mask face seal. It consists of a brief series of positive and negative pressure tests, which is carried out rigorously and in full every time the equipment is issued for use and, if practicable, every time it is put on. It is much more sensitive than the old 'suck down' test. Details of the procedure can be found in Dear Chief Officer Letter (DCOL) 1/2004 Item C, published by the Office of the Deputy Prime Minister (ODPM), and distributed by the Fire Policy Division of ODPM to all Chief Fire Officers. Copies are obtainable either from ODPM ( or your local Chief Fire Officer.

If this test procedure is to be adopted by non-fire service users of positive pressure demand BA, they will need to demonstrate comparable levels of training, discipline and competence, and have suitable procedures in place to deal with the inevitable test failures.

In summary

When BA is used for protection, an appropriate support system should be in place to enable its continued safe use. This system must include:

  • examination, cleaning, servicing and maintenance that is designed to ensure that the equipment is kept in full working order ready for the wearer, should it be required
  • training of wearers in the correct donning and use procedures, including what to do in case of emergency
  • properly designed and maintained support systems, such as compressors or cylinder filling stations
  • comprehensive records of maintenance and periodic examination Any weak link in this chain is likely to break when put under stress, and we'd rather you didn't appear in our statistics the next time we review the data.

References BS 4275:1997 Guide to implementing an effective respiratory protective device programme. BSI, London BS EN 12021:1999 Respiratory protective devices - Compressed air for breathing apparatus. BSI, London

For more information on Breathing Apparatus please visit

Published: 10th Jan 2004 in Health and Safety International

Share this article with your friends