The search for PPM and PPBs has to stop.

[mendip_discovery]

After a conversation on another thread, I thought I better post about it in a new thread.

With the 17025 laboratory, I work at we have been told our schedule can no longer use ppm or ppb as ILAC had removed it "Because of the ambiguity of definitions" so as metrologists we are now in search for μV/V or part per 10⁶. Of course, change the μV/V to μA/A or whatever tickles your fancy.

UKAS (the UK 17025 body) states LAB45 section 6.9.

It should be particularly noted that relative expressions, such as percentages, are not
permissible when the range of the quantity values includes, or is close to, zero. Under such
conditions, an absolute term must also be present; either on its own or in conjunction with the
relative term.

ILAC state in ILAC P14:09/2020 ILAC Policy for Measurement Uncertainty in Calibration

4.3 The uncertainty covered by the CMC shall be expressed as the expanded uncertainty having a coverage probability of approximately 95 %. The unit of the uncertainty shall always be the same as that of the measurand or in a term relative to the measurand, e.g., percent, μV/V or part per 10⁶. Because of the ambiguity of definitions, the use of terms “PPM” and “PPB” are not acceptable.

This is new(ish) to me and my auditor as it had snuck under the radar during the COVID confusion. As its on my schedule I am also making a point referring to it on my certificates, if it confuses customers then I will explain, I just like to work in 1 system at a time.

[TimFox]

Historically, "M" has sometimes been interpreted as 10³, as in the Roman numeral "M".
Also, there was confusion between "billion" as 10⁹ (normal usage in US) and 10¹² (normal usage in UK before 1975).
When I started physics grad school in 1970, a common joke was "BeV" = "Brookhaven electron-volt" (as in BNL) and "GeV" = "Geneva electron-volt" (as in CERN).

[David Hess]

Historically, "M" has sometimes been interpreted as 10³, as in the Roman numeral "M".
Also, there was confusion between "billion" as 10⁹ (normal usage in US) and 10¹² (normal usage in UK before 1975).
When I started physics grad school in 1970, a common joke was "BeV" = "Brookhaven electron-volt" (as in BNL) and "GeV" = "Geneva electron-volt" (as in CERN).

It makes sense, but if someone demanded that of me now, I would use dB just to spite them.  I have gotten quite used to converting between dB and uV/V in my head because old operational amplifier datasheets used both.

100dB = 10 ppm
106dB = 5 ppm
114dB = 2 ppm
120dB = 1 ppm
180dB = 1 ppb

[DH7DN]

Such revision don't happen "out of the blue". Either there have been some inconsistencies or contradictions in context of ISO/IEC 17025:2018 (Section 7.8 ) or there have been inconsistencies in reportings of measurement results or applications of CMCs.

My opinion on this is unfortunately biased for different reasons but as an accredited calibration lab, one is obliged to legal units (mostly SI unless you're in an exotic country such as North Korea) and technical standards/rules/policies. There is nothing confusing about stating the measurement results objectively and unambigously according to a harmonised terminology. Problems arise as soon one doesn't meet the requirements and leaves room for interpretation. Expressing results in µV/V or "parts in 10⁶" is valid and unambigous. Maybe a metrologist may give us some insight behind the PPM/PPB issue?

ISO 17025 doesn't say it explicitely but under certain circumstances the standard implies "The customer is not always right" and the lab is obliged to reject the (accredited) calibration order.

[Kleinstein]

Historically, "M" has sometimes been interpreted as 10³, as in the Roman numeral "M".
Also, there was confusion between "billion" as 10⁹ (normal usage in US) and 10¹² (normal usage in UK before 1975).
When I started physics grad school in 1970, a common joke was "BeV" = "Brookhaven electron-volt" (as in BNL) and "GeV" = "Geneva electron-volt" (as in CERN).

It makes sense, but if someone demanded that of me now, I would use dB just to spite them.  I have gotten quite used to converting between dB and uV/V in my head because old operational amplifier datasheets used both.

100dB = 10 ppm
106dB = 5 ppm
114dB = 2 ppm
120dB = 1 ppm
180dB = 1 ppb

The use of dB is problematic in some cases, where it is not clear if the value in question is more like a power or voltage. So use 10 dB or 20 dB for a factor of 10 change. It is ok for voltages, but how about pressure or optical power or concentrations.

With Si unit the capital M is used for millions. The japanese like it so much that some even use Mg for tons.

[penfold]

It was interesting to see this thread crop up, coincidentally I had to check some NIST document for something recently and stumbled across a prohibitory statement about 'ppm'/'ppb', I scratched my head a couple of times and thought "whaaat?", interestingly it referenced ISO31-0 (? I think, the obsolete one). I don't tend to write 'ppm' in an official sense, so to me it was entirely inconsequential, but I didn't get the impression it was an entirely new thing, and not even recently if ISO31-0 was withdrawn circa 2009.

I get that it's not entirely in keeping with SI, but there's just something that looks far more intuitive about ppm/°C than µV/V/K. I'd speculate it was something to do with confusion with 'physical parts' per million as particle counts etc as the ratio of two dimensionless quantities, hmm, what. about V/MV or mV/kV for symmetry?

[TimFox]

Despite the SI, in US wire gauges, after AWG 0, 00, 000, and 4/0, the larger sizes are given in “MCM” for “thousand circular mils”.  For example, 0.50 inch diameter is 250 MCM.  If the circular mil were an SI unit, then that would be 250 kcmil.
In the example given, the expression “nV/V” uses the correct SI units “nV” and “V”.  (Here, I avoided the letter “mu”.)

[TiN]

PPMs must live on. Not until inch and other nonsense units go away 

[TimFox]

Of course, "ppm" and "ppb" are not SI units, they are English-language abbreviations.

[JohanH]

Despite the SI, in US wire gauges, after AWG 0, 00, 000, and 4/0, the larger sizes are given in “MCM” for “thousand circular mils”.  For example, 0.50 inch diameter is 250 MCM.  If the circular mil were an SI unit, then that would be 250 kcmil.

And of course, in a large part of the rest of the world we use IEC 60228. AWG is totally unknown over here. "Mil" and "mils" is also an unknown abbreviation (easily confused with "mile"). In Europe the correct SI designation millimeter or mm is used.

[TimFox]

As I posted before in another thread, I was once confused during an installation in UK where the local engineer asked me about a wire size, using "mils", which I knew was not 0.001 inch.  Unfortunately, he did not mean "mm", he meant "mm²".

[Conrad Hoffman]

IMO, the demise of ppm is old news. I also lament the loss of micron, though it's still in so much common use that I don't think anybody's going to change anytime soon. In the US, a micrometer is a thing you use in the shop to measure with. I think angstrom is fading away but is still acceptable in the "outside SI" list at NISt.

[mendip_discovery]

Such revision don't happen "out of the blue". Either there have been some inconsistencies or contradictions in context of ISO/IEC 17025:2018 (Section 7.8 ) or there have been inconsistencies in reportings of measurement results or applications of CMCs.

I think it may be because it doesn't state what that ppm is referring to. For example, we may have a 10 mV to 100 mV range with an uncertainty of 23ppm + 12μV the ppm can be seen as unclear if it relates to mV or μV. Plus ppm is an English acronym and possibly they would like to reduce them.

Lab 45 (6.10) has this to say about the use of %

Particular care should be taken when the unit itself is normally expressed in percentage terms;
examples are relative humidity (%rh) and amplitude modulation (% AM). For example,
50 %rh ± 10 %rh means the boundaries are 40 %rh and 60 %rh, whereas
50 %rh ± 10 % means the boundaries are 45 %rh and 55 %rh.
 
Under circumstances of this nature the presentation of the CMCs must be such that there is no
ambiguity in interpretation.

The last bit being the interesting one. Its interesting that electrical is used as the example so it clearly is some sort of comment on them. Though as a mech lab that is going for some 17025 stuff I think the below quote may give a reason for the mud slinging that might be going on. ;-)

Lab 45 6.17

In dimensional calibration, the schedules of accreditation for the calibration of basic dimensional
measuring tools and equipment have historically reported a CMC for parameters most relevant
to  end  users.  This  method  however  does  not  provide  transparency  with  regards  to  auxiliary
measurements  such  as  flatness  and  parallelism  of  micrometer  measuring  faces  etc.  These
additional measurement techniques shall be listed on the schedule of accreditation along with
the corresponding measurement uncertainty.

For years labs only needed to say the error of length and not mention the capability for flatness or squareness. It is a bit like having a multimeter calibrated and the only Uncertainty quoted is for the Voltage.

PPMs must live on. Not until inch and other nonsense units go away 

Its interesting as officially you splitters have been using the metric system for a while, but for some reason subconsciously you still want to be back under British Rule 

As I posted before in another thread, I was once confused during an installation in UK where the local engineer asked me about a wire size, using "mils", which I knew was not 0.001 inch.  Unfortunately, he did not mean "mm", he meant "mm²".

Yeah, we use mil as an abbrev for mm. So "shave a few mils off that" and "I have some 2 mil wire" are common. I do get confuddled when you get places selling wire in Amperage, "yeah that is 12 amp wire". AWG, mm² and CSA at times do give me fun afternoons exp when a customer makes up some crimp samples for me but assumes I know the size of wire just by looking at it.

IMO, the demise of ppm is old news. I also lament the loss of micron, though it's still in so much common use that I don't think anybody's going to change anytime soon. In the US, a micrometer is a thing you use in the shop to measure with. I think angstrom is fading away but is still acceptable in the "outside SI" list at NISt.

So if PPM and PPB are old news then why are we getting people hear singing and dancing about how their meter is good to 1ppm.

[Vtile]

What is so sacret in ppm and ppb and ppt and what all there is. I will not cry after those. Why not to use proper units, they are rather usable.

[TiN]

Same goes to percents, it is just convinience when you only want a ratio between similar measurands, instead of dealing with 9 digits after decimal centipedes.

[mendip_discovery]

The same could be said for going down into the mV or k or M. Everything should be described as from the measurand.

So 100 mV 0.1 V and so on. A line needs to be drawn.

[David Hess]

You will never get my ppm blanket!