How does a drink driving breathalyser work?


What you need to know about breath test machines.

Can breath test machines accurately measure alcohol levels?

If you have been charged with drink driving, the chances are you were breath tested. The reason we can be so confident of that is because 87% of motorists charged with drink driving provide a sample of breath. But how do you know if the breath result is correct? 

The large, evidential breath machines found at police stations are Type-Approved. This means that they have undergone a series of rigorous technical tests and trials before being approved for use. There are currently three devices approved for use in England and Wales;

  • The Lion Intoxilyser 6000UK
  • The Intox EX/CR
  • The Camic Datamaster 


Three minutes to blow

When the breath machine begins its operational sequence, it will give the subject three minutes to provide the first specimen of breath. If successful, a second three minute period will begin allowing the subject to provide a second sample. During any three minute period the subject may have several unsuccessful attempts. If the subject provides an incomplete or unsatisfactory sample during the first three minute period, the instrument will time-out and abort the cycle. 

Deep Lung Air

When the subject blows into the machine it continuously monitors the alcohol content in the whole of the breath blown through the machine. Initially, this comes from the upper respiratory tract and, as the blowing continues, deep lung air begins to pass through the machine. At this stage the alcohol content of the breath is equivalent to the alcohol in the capillaries of the lungs.

During the blowing process the machine ‘sees’ an increase in the breath alcohol level as the air is exhaled. As deep lung air is reached the value reaches a plateau and this is the figure recorded as a valid breath sample.

The evidential devices used will also record values for incomplete breath samples by measuring the length of time the subject has blown, the volume of breath blown into the machine and the breath alcohol content at the time blowing ceased. It should be noted that the values recorded for these ‘partial’ samples are usually lower than the true breath alcohol value (found in deep lung air).

You can read the 'Guide to Type Approval Procedures' here

Mouth Alcohol

Oesophageal reflux can give rise to eructation, regurgitation and/or nausea. This means that vapours, liquids and solids present in the stomach may be brought up into the mouth. So that if alcohol is present in the stomach active gastric reflux means that some of that alcohol comes back into the mouth and it will contribute to any recorded breath alcohol value. So that the measured value for the breath alcohol will be falsely elevated.

Another feature of mouth alcohol is that it diminishes very rapidly at a rate complying with a half-life law. Studies have shown that the numerical mouth alcohol level drops by 50% roughly every minute. So that an initial false value of, say, 80 μg drops to 40 μg after one minute, 20 μg after two minutes, 10 μg after 3 minutes and then 5 μg, 2 μg and 1 μg.

There is no long term retention of a mouth alcohol value and even very high false values drop to zero within about 10 minutes. This is the reason that the police will ask motorists when they last had a drink and then allow up to 20 minutes if necessary to ensure any residual mouth alcohol has dissipated before the screening test.

It also means that if two valid breath samples were to be taken within a matter of two to three minutes of each other there would be a large difference in numerical value between the two results.

Elimination Rates 

Based on a person’s height, weight and gender, it is possible to determine the proportion of the body into which alcohol is distributed. This allows the contribution of a stated set of drinks to that person’s alcohol level to be determined. 

In performing alcohol calculations it is necessary to make certain assumptions. In most cases it is assumed that all the alcohol consumed was fully absorbed prior to the time of the evidential breath test. The rate at which alcohol is eliminated can vary between individuals, typically from 4.4μg% to 10.9μg% per hour, with an average figure of 7.8μg% per hour.