The latest government data reveals that in 2017 250 people were killed in a crash where at least one driver was over the drink-drive limit.
That is the most fatalities for eight years and the figures have prompted calls for tougher enforcement and lower drink-drive limits to improve safety.
There’s lots of received wisdom about what is and isn’t a safe or legal amount to drink before driving, often based on outdated information and rough guesswork.
There are alternatives, however, in the shape of home breathalyser units which claim to accurately tell whether you’re near or over the drink-drive limit.
With dozens on the market, ranging in price from a couple of pounds to a couple of hundred pounds, we decided to test how accurate home breathalsyer units are with the help of the UK’s biggest manufacturer, AlcoSense Laboratories.
It makes consumer units for home use as well as supplying industry and single use breathalysers to several police forces.
Its founder and managing director, Hunter Abbott explains the differences between different types of unit.
“The most basic ones are single-use tubes with reactive crystals, which basically give you a pass or fail. Above them are semi-conductors which are cost-effective to make but not as accurate or long lasting as fuel cell units. Fuel cell units use the ethanol in the sample as a fuel to create electricity and measures this to determine the amount of alcohol present. They are much more accurate and have a longer life but are more expensive due to their use of precious metals like platinum.”
At the top of the tree are the breathalysers which use expensive infrared spectrometers for ultimate accuracy and reliability, one of which we’ll be using as a benchmark.
Sadly, testing the units doesn’t involve borrowing a police breathalyser, necking a few pints and blowing into them one after the other.
The human body, as it turns out, is pretty useless at performing controlled, consistent, repeatable tests. A few minutes’ difference, how deeply you breathe and how many samples you’ve already given all affect any reading. Instead, we used breath alcohol simulators which produce a constant and repeatable sample of alcohol in air at the same humidity as human breath.
For the purpose of the test the solutions in the simulators were made up by an independent supplier to the Scottish drink-drive limit – 0.22 milligrammes of alcohol per litre of breath (mg/L) – which many want to see introduced in the rest of the UK.
The apparatus looks a little like something from a cheap sci-fi film. Glass jars are linked to a small air pump and the breathalysers via a network of rubber hoses, with various valves, switches and LCD screens controlling and monitoring everything from the temperature of the samples (vital for accuracy) to the pressure of the air passing into the unit.
It works by pumping air through the solution at a set pressure and volume. When the air passes through the solution it is heated and ‘picks up’ alcohol and water vapour at a fixed value, replicating the action of a person blowing into the unit but without any of the pesky person-related variables.
To set a control reading, we used a £13,000 Drager Alcotest 9510 which is a police evidential breathalyser, not the handheld devices used at the road side. If you’re taken to a police station to give an evidential breath sample – the reading used if you’re taken to court – you’ll probably be faced with one of these.
All units have a built-in tolerance. The Drager allows for three per cent misread either way and police will round down in the drivers’ favour. The AlcoSense units also have a margin of error but are calibrated to take this into account and overread slightly, so users can’t accidentally end up over the limit.
Using the samples prepared to the 0.22mg/L limit the Drager returned a result of 0.229mg/L.
We then tested four of AlcoSense’s systems, ranging from the basic semi-conductor Lite 2.0 to the fuel cell Excel, Pro and Ultra, which uses the same sensor as police handheld units.
In normal settings the devices round up the readings to two decimal places and add the tolerance to avoid any risk of underreading. For our testing we were able to access the raw figures on the AlcoSense Excel, Pro and Ultra.
The Excel, showed 0.233mg, the Pro 0.212mg and the Ultra matched the Drager exactly with a reading of 0.229mg. In normal mode (the readings you would see when using them), each one read slightly higher than the police evidential unit.
The entry-level AlcoSense Lite 2.0, which uses the more basic sensor proved less accurate, with a reading of 0.4mg/L, but at least overread rather than risking a falsely low reading which could put an over-the-limit driver behind the wheel.
So, for the purposes of using one to check yourself the morning after a big night out, all four proved more than accurate enough.
The results did, however, show the gap between different types of sensor and Hunter warns that not all breathalsyers are created equal.
“The making and selling of consumer units isn’t regulated in the UK,” he says. “There is a very good Pan-European consumer breathalyser standard which is based on a cut down version of the European police standard. This is what we design and build our fuel cell and single-use products to, but as it’s not a legal requirement to comply with it, it means standards in the marketplace are variable to say the least.”
Fuel cells can vary in size and quality. The key active ingredient in a fuel cell sensor is platinum – the most expensive metal in the world – so an easy way to cut cost is to use less platinum, but at the expense of accuracy. There’s also the question of calibration, which if done wrong at the factory will render the unit useless. Likewise, without annual recalibration even a good sensor will lose accuracy over time.
Further complicating the problem is that many limits are set in an alcohol in blood reading. The breathalyser has to convert the breath reading to a blood reading, and different countries calculate the conversion differently, so depending on where a unit is made and where it is designed to be used, the numbers might not line up. For example, a breathalyser designed for French use will under read by 13 per cent if used in the UK.
It makes the market a potential minefield but choose a reputable unit from a recognised manufacturer and you should be able to rely on it. Just remember that the components in these are expensive and you get what you pay for, so don’t expect a £5 unit from Amazon to perform nearly as well.