SPOILER ALERT:
The official UK and USA metrology websites are INACCURATE
Since the inception of the International System of Units (SI) in 1960, it has always been recognised that some non-SI units, such as the litre, minute and hour, are “widely used and are expected to continue to be used for many years”.
The SI caters for this reality by giving special status to a small number of such units, allowing them to be used alongside SI units for the foreseeable future. For the avoidance of confusion over which non-SI units can be used alongside SI units, the SI Brochure includes an internationally agreed list of “non-SI units that are accepted for use with the SI”.
Anyone in doubt over whether any given non-SI unit, such as the “calorie”, “bar”, “are”, “nautical mile” or “astronomical unit”, can be used with the SI, can simply check the list in Table 8 of the SI Brochure.
The 2019 Redefinition of the SI
An often-overlooked detail of the redefinition of the SI base units in 2019, was the major rationalisation that also took place at that time of the non-SI units listed in the SI Brochure – the definitive reference for the metric system.
The 2019 revision resulted in a greatly simplified list of “non-SI units that are accepted for use with the SI”. These units, together with all SI units, form what is generally known as the modern metric system. This process of rationalisation has been under way since the inception of the SI.
The previous (8th) edition of the SI Brochure, published in 2006, included definitions for no less than 42 non-SI units, of which only 12 were described as being accepted for use with the SI. These non-SI units were split into different categories across four separate tables.
In contrast, the current (9th) edition of the SI Brochure, published in 2019, includes definitions, in a single table, for just 15 non-SI units, all of which are described as being “accepted for use with the SI”.
Previous editions of the SI Brochure included definitions for some non-SI units that are not accepted for use with the SI. Categories detailing examples of such units, e.g. “Other non-SI units” in the 8th edition of the SI Brochure, no longer exist. The rule is now much simpler – any non-SI unit not included in the SI Brochure cannot be considered to be part of the modern metric system.
The 15 non-SI units that are accepted for use with the SI
The table below reproduces the complete list of all “non-SI units that are accepted for use with the SI”, as given by Table 8 of the current SI Brochure.
| Name | Symbol | Value |
| minute | min | 1 min = 60 s |
| hour | h | 1 h = 60 min = 3600 s |
| day | d | 1 d = 24 h = 86 400 s |
| astronomical unit | au | 1 au = 149 597 870 700 m |
| degree | ° | 1° = (π/180) rad |
| minute | ′ | 1′ = (1/60)° = (π/10 800) rad |
| second | ″ | 1″ = (1/60)′ = (π/648 000) rad |
| hectare | ha | 1 ha = 1 hm2 = 104 m2 |
| litre | l, L | 1 l = 1 L = 1 dm3 = 103 cm3 = 10-3 m3 |
| tonne | t | 1 t = 103 kg |
| dalton | Da | 1 Da = 1.660 539 066 60(50) × 10-27 kg |
| electronvolt | eV | 1 eV = 1.602 176 634 × 10-19 J |
| neper | Np | |
| bel | B | |
| decibel | dB |
Whilst national metrology institutions were quick to update their websites with respect to the new definitions for SI base units, five years later some of these same organisations have still not updated their websites to take account of the latest definitions for non-SI units accepted for use with the SI.
National Physical Laboratory (NPL)
NPL is the home of the UK’s National Metrology Institute (NMI). Amongst other things, NPL was responsible for proposing the recent additions to the SI prefixes that were internationally approved in 2022.
However, NPL’s list of “Non-SI units accepted for use with the International System” is far from ideal:
The table on their website omits the following non-SI units that are accepted for use with the SI:
- hectare
- electronvolt
- astronomical unit
- dalton
- decibel
The table also has two identical entries for “tonne”, and a definition for “minute of plane angle” which is just another name for “minute of arc”.
National Institute of Standards and Technology (NIST)
NIST is an agency of the United States Department of Commerce. It is the national metrology institute of the USA.
Table 6 on their webpage,
https://physics.nist.gov/cgi-bin/cuu/Info/Units/outside.html
lists “Units outside the SI that are accepted for use with the SI”.
It defines the electronvolt with an approximate value, stating that “the value must be obtained by experiment”. However, since 2019, it has been defined with an exact numeric value, owing to the redefinition of the ampere.
The table also omits the following non-SI units that are accepted for use with the SI:
- hectare
- decibel
Table 6 neglects to state that the “unified atomic mass unit” is now referred to as the “dalton”.
On the same webpage, Table 7 incorrectly lists the following units as being “currently accepted for use with the SI”, whereas in fact they are obsolete and are not accepted for use with the SI:
- nautical mile
- knot
- are
- bar
- ångström
- barn
- curie
- roentgen
- rad
- rem
The “are” in particular has never been accepted for use with the SI.
NIST have a number of other webpages detailing non-SI units that are accepted for use with the SI. The following webpage seems to be up-to-date:
https://www.nist.gov/pml/special-publication-330/sp-330-section-4
But this page has numerous inaccuracies:
https://www.nist.gov/pml/special-publication-811/nist-guide-si-chapter-5-units-outside-si
To add to the confusion, in Table 9, NIST makes a distinction between “non-SI units that are accepted for use with the SI” as defined by the CIPM, and “non-SI units that are accepted for use with the SI” as defined by their own Guide.
Examples of units that NIST’s Guide state are accepted for use with the SI, regardless of the CIPM, include old CGS radiometry units such as the “curie”, “roentgen” and “rad”.
Table 9 also incorrectly lists the following units as being “non-SI units accepted for use with the SI by the CIPM”:
- nautical mile
- knot
- ångström
- barn
- bar
- millimetre of mercury
Conclusion
Both NPL and NIST are national metrology institutes. As such, the information on their websites is assumed to be correct and up-to-date. It’s quite likely that other websites and sources of information about the metric system cite theses websites as authoritative sources of information. It is therefore shocking that the information to be found there about non-SI units has seemingly remained incorrect for so long.
Further reading
SI Brochure – BIPM
https://www.bipm.org/en/publications/si-brochure
Non-SI units – UKMA
https://ukma.org.uk/what-is-metric/definitions/non-si-units/
Non-SI units accepted for use with the SI – Metric System
https://metricsystem.net/non-si-units/accepted-for-use-with-si/
Editor: Please note that since this article was originally published, the NPL website has been corrected.
Metric Views 
In my view, NIST, NPL and BIPM are all at fault. In particular the 9th edition of the SI Brochure does not draw to attention material that was present in the 8th edition, but which has been deleted. In many standards, when a definition is deleted, that definition continues to be present in the standard, but is flagged as being “deprecated” with the implication that it should not be used in new applications unless it needs to be used to maintain backward compatibility.
ISO 31 is an example of how deprecation works. ISO 31-1 catalogues units of measure that are derived from length such as length itself, area, volume, capacity. This catalogue includes both the symbol and the definition of the unit as well as the suggested symbol for specific instances of the unit. For example, if I wrote A=πr², the standard catalogues suggested symbols for area and radius as being A and r. In ISO 31-1:1978 imperial and customary units appeared alongside their SI counterparts, but in ISO 31-1:1992, the units of measure that were based on the foot rather than the metre were relegated to Appendix A of the standard and were flagged as having been deprecated.
In a similar manner, ISO 31-3 catalogued units of measure that were based on mass (as well as time and length), thus covering mass, force, pressure, energy, power etc. Again the 1978 version catalogued units that were based on metric, imperial and customary units while in the 1992 version, only units based on metric quantities were catalogued in the standard itself while those units that were based on the pound were relegated to Appendix B and flagged as being deprecated.
In 2009 ISO 31 and ISO 1000 (definitions of metric prefixes) were merged into a single set of documents ISO 8000. An inspection of the ISO 8000 contents page shows that the deprecated units in ISO 31-1:1992 and ISO 31-3:1992 are no longer catalogued.
References
https://pdfcoffee.com/iso-31-1-engl-1992-quantities-and-units-part-1-pdf-free.html
https://pdfcoffee.com/iso-31-3-engl-1992pdf-pdf-free.html
https://www.intertekinform.com/preview/98700315736.pdf?sku=877033_SAIG_NSAI_NSAI_2084431
LikeLike
The SI Brochure is compiled by scientists, for the benefit of scientists, whose allegiance has been with metric and SI units for many years, despite the “traditions” of UK, USA and others. This allegiance is very likely to continue. Such esoteric matters are beyond the interests of Joseph Public or Josephine Public.
For my part, I would be very pleased to see road distances and speed limits metricated and beer served in litres, or sub-multiples thereof. UKMA could say, “Job done!” and take a well-earned retirement.
If the erudite members of NPL and NIST, among others, think that their publications are out of step with the SI Brochure, I suggest we leave them to argue this out among themselves.
LikeLike
Good article, Isaac.
Martin, I agree. My experience is that people are more likely to update text if I can quote an authority that says a unit or term is deprecated (such as when commenting on draft ISO standards). Version 5 of the brochure had table 12 for ‘units generally deprecated’.
The brochure is developed by the ‘Consultative Committee for Units’ (CCU). See ‘Meeting reports’ at: https://www.bipm.org/en/committees/cc/ccu
You can search for discussions by using something like:
site:bipm.org ccu deprecated
LikeLiked by 1 person
The problem with the NIST pages is that they have not been updated from edition 8 to edition 9 of the SI Brochure. The page quoted in the article is part of a series of pages on the SI. The “Introduction” page carries the warning note:
“Note
The information on the SI below was current up until May 20, 2019 (World Metrology Day). Since then the SI has undergone major changes. Hence the pages below are out of date. Updated pages will soon become available.”
Since it is 2024, I think “soon” is over. They should update, withdraw the pages, or at least carry the disclaimer on every page. I did send a note to a NIST contact, but Monday is a holiday here so a response is unlikely for a few days.
NIST SP 330 (the US edition of SI Brochure) is updated to edition 9, SP 811 is not and many of NIST’s webpages are not.
LikeLike
I checked the Cambridge International, AQA, OCR and Edexcel Physics A Level syllabuses. None of them mentioned the dalton, but they all mention the atomic mass unit, ether explicitly or by its symbol “u”.
LikeLike
I don’t think that the inaccuracies on the NPL and NIST websites can be explained simply by a failure to take account of changes made after the 8th edition of the SI Brochure.
Also, two of the NIST webpages quoted in the article state that they were last updated in 2023:
https://www.nist.gov/pml/special-publication-330/sp-330-section-4
“Created August 27, 2019, Updated September 14, 2023“
https://www.nist.gov/pml/special-publication-811/nist-guide-si-chapter-5-units-outside-si
“Created January 28, 2016, Updated December 1, 2023“
The old CGS radiometry units (curie, roentgen, rad, rem), that NIST incorrectly describe as being “accepted for use with the SI”, were already completely absent from the 8th edition. Of those CGS units that were listed in the 8th edition of the SI Brochure, none were described as being “accepted for use with the SI”.
There is of course nothing to stop NIST from including old CGS units, or any other non-SI units, in their Guide, but they really ought not to be stating that they are “accepted for use with the SI”.
For reference here are the complete lists of the “Non-SI units accepted for use with the SI” in the 9th and 8th editions:
9th Edition
minute
hour
day
astronomical unit
degree
minute (angle)
second
hectare
litre
tonne
dalton
electronvolt
neper
bel
decibel
8th Edition
minute
hour
day
degree
minute (angle)
second
hectare
litre
tonne
electronvolt
dalton / unified atomic mass unit
astronomical unit
Additionally, here are the only CGS units present in the 8th Edition (where they were described as: “Non-SI units associated with the CGS and the CGS-Gaussian system of units“):
erg
dyne
poise
stokes
stilb
phot
gal
maxwell
gauss
oersted
LikeLike
@Isaac
The tables alone don’t fully represent the obsolete radiation units in the 8th edition of NIST SP 330. The text states the following about them:
“4.3 The curie, roentgen, rad, and rem
This section and Table 10 below have been added to the United States version of the BIPM SI Brochure because, although the curie, roentgen rad, and rem are not
accepted by the CIPM for use with the SI, they are widely used in the United States, especially in regulatory documents dealing with health and safety. The interpretation of the SI for the United States given in the Federal Register notice referenced in footnote (f) of Table 6, p. 32, does in fact accept their use with the SI. Nevertheless, that notice strongly discourages the continued use of the curie, roentgen, rad and rem and recommends that the corresponding SI units should be used whenever possible, with values of relevant quantities given in terms of these outdated units in parentheses only if necessary.”
US law treated them as reluctantly acceptable for use with the SI within the US. However, they have been removed from the 9th edition of NIST SP 330.
LikeLike
@jmsteele9027
Thanks. Lets hope that NPL and NIST both fix their websites.
I had messaged them both anonymously about this over 6 months ago, using their contact pages. However, neither responded, hence this article.
LikeLike
@Isaac
I will leave NPL to you, as I doubt they would listen to me.
As a USMA member, I contacted our contact person at NIST with a link to the essay and my personal comments about it. I hope to get a response later this week. I also reviewed and reported our USMA pages which have some of the same problems.
However, I suspect they will cite budgetary constraints. I have been complaining about the lack of update to NIST SP811 for a few years, and they say they don’t have the headcount to get to it.
LikeLike
I noticed in the posts above that the NIST does not use the International Time/Date Standard ISO-8601. Strange because I would think that a reputable standards organisation would at least make an attempt to use all agreed upon standards such that there is no excuse for them not to use the ISO-8601 standard for dates and times.
LikeLike
One of the most noteworthy “missing” items in the list of non-SI units that are acceptable for use with SI is the year. I suspect that this is deliberate because different people have different understandings of what a year is. For some, it is 365.25 days, for others it is sometime 365 and sometime 366 days and for yet others it is 365.23 days. Thus if people wish to use years alongside other SI units, they will have to specify what they mean by a “year”.
Geologists seem to have taken matters into their own hands and use the symbol “a” (annum) for years when they talk of certain rocks being 100 Ma old. It should be noted of course that the symbol “a” is also the symbol for the “are” (an area of 100 square metres, hence the word hectare).
LikeLiked by 1 person
@ Daniel
If you spell or abbreviate the month, ISO 8601 is inapplicable; it is a numeric-only format, and does not apply to any “text” form of the date.
Also note that NIST supports US exceptions to the SI Brochure (NIST SP 330), including spelling, and clearly stating American preferences,( L vs l for liter, dalton vs unified atomic mass unit for atomic mass, etc), and explains the BIPM position in footnotes or marginalia.
LikeLike
jmsteele9027:
So, why does the month need to be spelled out or abbreviated? So what if there is a numeric only format used? Are employees of the NIST incapable of understanding an all number date/time format? I would find that strange for people who should be highly intelligent and an all numeric date/time format should be completely understandable by them.
Also the ISO 8601 standard is independent of the SI Brochure, so why do the employees at the NIST feel a need for an exception? Is there a logical reason that those at the NIST can’t handle universal standards that don’t yell out American exceptionalism?
Does this have anything to do with the year coming first instead of last?
LikeLike
Last week I attended a lecture by one of the managers from NPL. After his lecture, I spoke to him about the discrepancy in the NPL website, but he did not seem too concerned about it (he was not directly connected with the formulation of SI, though he did tell me that the colour chart showing the seven base units and seven defining constants actually originated at the NPL. I got the impression that the main concerns at the NPL about SI was the purist view that the candela, the mole and possibly temperature were not real base units. I have long held that view that since the candela is a function of the human eye it is not a real base units. Similarly the mole is the result of a conversion factor the links atomic scale physics and macro scale physics. The lecturer told me that some at teh NPL held the view that temperature was a manifestation of energy transfer rather than a base unit in its own right. Food for thought.
LikeLiked by 1 person
An NPL manager is not concerned about errors in a table of units on the NPL website? Really?
LikeLike
Another commentary on why the USA has not adopted the metric system. What do y’all think?
Ezra aka punditgi
LikeLike
UPDATE: The NPL website has been fixed now.
LikeLike
It is certainly interesting as Martin M. mentions, that an NPL manager considers the candela, mole and temperature (kelvin) “not base units”. I have often been confused about the candela while the mole is just a number and the temperature is a molecular energy (via Boltzmann’s constant).
LikeLike
This guy presents all the SI units (so he says). Did he get everything right?
Ezra aka punditgi
LikeLike