sciencebriefs
13:00in productionCh. 1 · Four units, four constants/ 13:00 · ceiling 15 min
Physics · Engineering

2019 revision of the SI

2018

On 20 May 2019 the kilogram stopped being a chunk of metal in France and became a consequence of the Planck constant. The vote took eleven years to arrive and left a few loose mathematical ends behind.

The 2019 revision fixed the numerical values of four constants — the Planck, elementary charge, Boltzmann and Avogadro constants — so that the kilogram, ampere, kelvin and mole no longer depend on a physical object or an idealised experiment. The main trigger was a measured drift in the kilogram's reference cylinder; the cost was a few exact relationships between units that are now only approximate.

Chapters & takeaways6
  1. 0:08
    Four units, four constants

    The kilogram, ampere, kelvin and mole were each re-anchored to a fixed numerical value of a fundamental constant instead of an object or an idealised setup.

  2. 2:10
    An eleven-year approval process

    From a 2007 mandate to a unanimous 2018 vote, the change moved only once independent experiments agreed closely enough.

  3. 4:20
    Why the kilogram had to go

    The reference cylinder in Sèvres was found to drift by tens of micrograms a year against its official copies.

  4. 6:30
    The balances that made it possible

    Kibble balances and the Avogadro project supplied Planck-constant measurements precise enough to meet the threshold the CGPM had set.

  5. 8:40
    The seam the fix left behind

    A handful of previously exact relationships between units, including the dalton and the mole, are now only approximately true.

  6. 10:50
    Worth understanding, not worth dread

    The physics is genuinely intricate, but the outcome for anyone using a scale or thermometer is nothing at all.

Worth your time?

Selectively. Start with the brief, then study the parts we point at.

4/ 5
What works
  • lays out a specific, falsifiable trigger — measured mass drift in the kilogram prototype
  • gives the acceptance threshold the CGPM actually used before it would vote
  • is honest about the exact relationships the change broke rather than presenting it as costless
What does not
  • does not make the Planck-constant definition of mass intuitive for a general reader
  • leaves the size of the practical consequences, if any, for ordinary weighing understated
Study it if
  • readers who want the mechanics behind a change they may have heard about in passing
  • anyone curious how a committee decides a physical constant is 'known well enough'
  • people interested in the trade-offs a redefinition like this creates
Skip it if
  • readers who want a plain restatement of what a kilogram is now, without the history
  • anyone uncomfortable with constants expressed to nine or more significant figures
The written brief4 min read

Four units, four constants

The claim is narrow and specific: that four of the seven SI base units — the kilogram, ampere, kelvin and mole — could be, and as of 20 May 2019 were, redefined by fixing the exact numerical value of a fundamental constant, rather than by a physical object or an idealised experimental setup. The kilogram now follows from the Planck constant, the ampere from the elementary electric charge, the kelvin from the Boltzmann constant, and the mole from the Avogadro constant. The second, metre and candela had already been defined this way in earlier decades. The stated aim was to make the base units as stable as the laws of physics rather than as stable as a manufactured artefact, while keeping every unit’s actual size unchanged for continuity.

An eleven-year approval process

Getting there took over a decade of formal process. The General Conference on Weights and Measures mandated the investigation in 2007; by 2010 proposals existed but the International Committee for Weights and Measures judged the underlying measurements not yet good enough, and the 2011 and 2014 conferences accepted the principle while deferring the vote for the same reason. The threshold was explicit: acceptance required at least three independent experiments producing values for the Planck constant with a relative uncertainty of no more than five parts in a hundred million. That bar was cleared by 2016, CODATA published agreed recommended values for the constants in October 2017, and the General Conference voted unanimously to proceed on 16 November 2018, with the change taking effect the following May.

Why the kilogram had to go

What holds up is the reason for acting in the first place. The kilogram’s definition rested on a single platinum-iridium cylinder held at the International Bureau of Weights and Measures, and comparisons against its official copies found mass drifting by as much as twenty micrograms a year in some of them — a small figure, but one with no floor, since nothing about a physical cylinder guarantees it stops moving. The ampere’s old definition, based on the force between two infinitely long parallel wires carrying current, could never be realised in an actual laboratory. The kelvin’s reliance on the triple point of water broke down at very low and very high temperatures. Fixing constants instead removes each of these specific weaknesses at once.

The balances that made it possible

The technical achievement that made the vote possible was measurement, not policy. Kibble balances — instruments that balance a weight electromagnetically and were formerly known as watt balances — and the Avogadro project, which counts atoms in a nearly perfect silicon sphere, gave independent routes to the Planck constant that converged closely enough to satisfy the CGPM’s threshold. Multiple groups working by different physical methods reaching compatible values was the specific evidence the committee had asked for, and it is what separates this redefinition from a purely administrative decision: the constants were pinned down experimentally before they were fixed by decree, not the other way round.

The seam the fix left behind

The change did not come free. Because the mole and the kilogram were previously linked through the dalton and the Avogadro constant in a way that made certain conversions exact, fixing the Avogadro constant on its own broke that exactness — the relationship now holds only approximately, and the molar mass constant is no longer precisely one gram per mole, though the resulting discrepancy is reported as extremely small. Fixing the elementary charge similarly means that the vacuum permeability and permittivity, previously exact by definition, now carry a small measured uncertainty tied to the fine-structure constant. Critics have also questioned, on philosophical rather than practical grounds, whether the mole and candela were ever base physical units in the same sense as the others.

Worth understanding, not worth dread

For a general reader this is worth the time mainly as a case study in how a standards body actually changes its mind: not on a single measurement, but on convergent evidence from different instruments crossing a threshold fixed years in advance. It will not change how a kitchen scale or a lab balance behaves, and the paper trail of committee votes and CODATA values is dry by design. But for anyone who wants to see the gap between a physical constant being ‘discovered’ and being formally adopted as the basis of a unit, and who is willing to sit with a few broken exact relationships as the price of the change, the account rewards the attention. It is not a story built for drama, and does not need to be.

Same field · Physics4 of 183
Up next in Science

REM rebound

· 13:00

William Dement's REM-deprivation experiments showed the brain pushes back when a sleep stage is withheld, and that push-back, not sleep loss in general, is what REM rebound actually measures.

13:00