Include VAT

One of the most confusing things for the potential purchaser in the wonderful world of Breathalyzers is establishing exactly what sort of level of accuracy these devices have.

There are numerous companies out there (some with far higher marketing budgets than accurate production cost interests) who would have you believe their £25 unit is state-of-the-art, with levels of accuracy that rival the Police grade units.   When you consider that an ACTUAL police grade unit is at least £700, you would think that it would be clear that a device that cheap can give little more than a few beeps and a colourful light display.  But let’s look at these accuracy claims.

As a rule, devices which cost upto about the £60 mark are generally quoted as having an accuracy level of about “+/-0.02%BAC at the drink drive limit” – all of which sounds very impressive.  But put into perspective, what it ACTUALLY means is that if the drink limit reads as 80mg, the device can be accurate by plus or minus 20mg – therefore a variable accuracy rate of 25%.  Thats a big stretch.  And that’s at the drink drive limit.  At a quarter of the drink drive limit – the limit that anyone who works in Air, Rail or Maritime industries has to comply to – the accuracy rate technically would be plus or minus 100%.  Ouch.

ACTUAL police grade detectors (such as the Home Office Approved Draeger 6510 or the AlcoDigital 3000 which uses the same sensor) have an accuracy rate of plus or minus upto 2%.   In otherwords, that equivalent 80mg readout would vary by UPTO plus or minus 1.6mg.

There’s good reason these devices cost more.  That’s because they work.



  1. Spotty Welshman 30th January 2011 Reply

    I have been looking at buying a breath tester for some time and am interested in the alcosens 1 but cant see it on your sight ? It is priced at around the mark that you mention but this model has passed and exceeded the accuracy tests required by law so I am confused by your blogg. Alcosense 1 seems very acurate. Surely if it has passed the tests then it must be pretty good ? Their is also a video on the other website of an alcosens breath tester giving the same reading as a Police tester (not sure what the tester woz though).

    • suzannah 2nd February 2011 Reply

      We discontinued the AlcoSense products last year, however the AlcoSense 1 that you are talking about is essentially a re-badged device called ‘SafeMate’ which we used to stock about 6 years ago, and is produced by the Korean manufacturer KPro.
      In terms of accuracy, whilst the SafeMate (NOT the AlcoSense 1) is listed as a recognised american ASD (Alcohol Screening Device) this essentially means it is able to detect the presence of alcohol but gives no indication of accuracy. AlcoSense quote all their products as having an accuracy of +/- 0.02%BAC so (under laboratory conditions) at the UK limit of 0.08 it could have an accuracy variance of 25%, at the Transport limit of 0.02% it could have a variable accuracy of 100%, in common with all semi-conductor devices at this cost level.
      I’m not sure what you refer to with regard to ‘accuracy tests required by law’ – in the UK there is only one level of testing and that is the Home Office Approval, required for example by the units supplied to UK Police, but the accuracy of these approved devices is reflected in the price (in the region of £700 plus). None of the units marketed at Personal Users fall under Home Office Approval however the sensor of the AlcoDigital 3000 is at least the same as the approved Draeger 6510, even if the unit itself has not been through the approval process.
      Whilst we would hope that common sense would dictate that the potential accuracy of a device that costs about £20 could in no way compare to an approved device which costs £700, the fact that there simply isn’t any other Personal Device approval rating in the UK unfortunately exacerbates the difficulty in differentiating between wildly optimistic claims and reality which can be a very different matter!

Leave a Comment

Your email address will not be published.