• Stephen Biss

International Std OIML R126 Mandates Maximum Permissible Error

Sample Cross-examination of CFS expert on OIML standard R126 (2012) respecting maximum permissible error in evidentiary breath instruments - new and after repair

Note: The cross-examiner erred in his calculations of MPE as required by OIML. The chart during cross-examination is included as well as the corrected version. The error was mathematical, entirely the fault of the cross-examiner in thinking R 126 MPE standard was drafted as blood rather than breath.


To identify that the international standard for MPE for instruments in service, 6 mg% at target 100 is much tighter than 10 mg%.

To identify that the international standard for MPE for instruments after repair is tighter than instruments in service.

To raise the concept of MPE after repair and in service at various target values across the measuring interval.

Q. Yes. Now, on the last day, ... I was having some difficulty in asking my question of you. And I think that perhaps you thought that what I was doing was

suggesting that item 5.2.2 requires control checks at a variety of different levels and an instrument response checking error at each of those. But rather what I wanted to ask you was item 5.2.2 on page 10, varies the required maximum permissible error depending on the concentration that is used for a control check, for an inspection, for a verification. Anytime a control check is used in that kind of context. A. Yes. Q. So, and that variation is across 0 to 200 milligrams per 100 mils if we look at 5.1 but the upper – a greater upper limit may be defined by the manufacturer.

A. Yes. Q. So, in this particular case, C.M.I. defines its upper limit in its specifications as 600 milligrams per 100 mils. A. For testing purposes, yes. Yes. Or, sorry, the range that it can measure, the actual blood alcohol concentration as opposed to the assessment range. Q. Yes. The measuring interval.

A. Correct. Q. And so, I take it then the–your understanding of the concept of maximum permissible error in the O-I-M-L document is that when an instrument is being inspected, an instrument – specifically, an instrument being in service under 5.2.2 the expectation is that the – the results of control tests are going to be within the maximum permissible error under 5.2.2. Is that how you read 5.2.2? A. All right. Could you repeat that now? Q. Under 5.2.2 my suggestion is that if – as you read it, the expectation is that the response of the

instrument to using a control solution at whatever concentration it is, the maximum permissible error for an instrument in service would be the numbers that are indicated there, and they’re – it’s a calculation in each case, under 5.2.2. A. So, as I read it here, 5.2.1 deals – or 5.2 relates to the operating condition, as opposed to service. Are you referring to trying to get the – like, when the instrument is sent in to service? To the manufacturer, where they’re having to restore it to original specifications? Or are you talking about operational parameters when the instrument’s being used for testing in the field? Q. All right. Under 5.2 it indicates the following M-P-Es apply within the rated operating conditions. A. Yes. Q. Which are specified. 5.2.1 talks about when an instrument is being submitted for type approval or initial verification or verification after repair. A. Correct.

[Erroneous Calculation of Maximum Permissible Error (MPE) under R 126 shown to witness at time of cross-examination - error by S. Biss but chart accepted by witness]

[Corrected version of table given to Court after allowing for breath rather than blood basis for standard - the error was identified by Mr. Kupferschmidt when he read the transcript1.]

Q. Item 5.2.2 refers to breath alcohol analysers in service. A. In service, correct. Yes. Q. So, in terms of the actual numbers the expectation is, and I think we used – there’s Exhibit Number 34, that was this document. I don’t know whether you still have your copy? Yes. That’s Exhibit Number 34. You indicated that you – you thought my calculations were correct there. The expectation is, that for an instrument that is in service, that if it is tested at 40 milligrams per 100 mils the expectation is that it will be within plus or minus 3 milligrams per 100 mils. If it’s tested at 50 it would be within plus or minus 3 milligrams per 100 mils, but if it’s

tested at 100 it would be within plus or minus 7 milligrams per 100 mils. A. Seven point five percent. Q. All right, the 5.3 indicates that a measured value to three decimal places – oh, down to two decimal places, yes, I’m sorry. My mistake. So, it would be 7.5. A. Yes.


MPE is a concept that can be found throughout international metrology literature. The point is that it needs to be considered as different amounts at different values within the measuring interval. An error of 12 mg/100mls at 40 target value is much more serious to linearity than an error of 12 mg/100mls at 300 target value.

1. Here is the paragraph that the cross-examiner misunderstood and resulted in the erroneous table:

#crossex #MPE #OIML #R126

14 views0 comments