Implementing the new kelvin 2

Short Name: InK 2, Project Number: 15SIB02
Image showing a worker in a steel factory
Worker in steel industry

Developing methods for a comprehensive and low-uncertainty temperature dataset to redefine the kelvin

The International System of Units (‘SI’) is comprised of seven base units, including the kelvin - the unit of temperature. Since the signing of the metre convention in 1875, these units have been constantly revised and improved.

From 1954, the kelvin (K) was defined as 1/273.16 of the thermodynamic temperature of the triple point of water. This was derived using water cells and depended on the purity and isotopic composition of the water used. As it isn’t possible to prepare two exactly identical cells, measurements of the kelvin varied. In 2018, the world metrological community voted to revise four units of the SI, including the kelvin. The re-definition would link its value to the Boltzmann constant (k), making it independent from any material properties.

However, rigorous metrological data was required, along with uniform definitions and instructions to allow the kelvin to be realised at the highest level. This new ‘mise en pratique’ for the kelvin (MeP-K) required foundation on primary thermometry methods and results. In particular, a robust dataset, detailing deviations between temperatures founded on the two current temperatures in use: the International Temperature Scale of 1990 (ITS-90) and the Provisional Temperature Scale of 2000 (PLTS-2000).

The previous EMRP project InK developed a primary thermometry research infrastructure, covering measurements in the temperature regions 0.02 K to 1 K and ~80 K to ~300 K, and established thermodynamic temperatures for a selection of high temperature fixed points.


This project improved three primary thermometers for ultra-low temperature ranges and used these to derive new low-uncertainty thermodynamic temperature values for the two temperature scales. It also showed, for the first time, that PLTS-2000 could be thermodynamically inaccurate by more than 6 % at the lowest temperatures and revised the temperature data over the range 0.9 mK to 1 K.

The project also investigated three novel thermometry methods, indicating Doppler Broadening Thermometry (DBT) as a promising additional primary thermometry technique.

The new methods and data developed completed the knowledge required for a comprehensive low-uncertainty dataset. The thermodynamic measurement techniques were incorporated into the MeP-K documentation, ensuring the re-definition was successfully achieved and disseminated. 


As well as facilitating the redefinition of the kelvin in 2019, which will have a long-term effect on industry requiring improved temperature measurements, project results impacted on manufacturers of cryogenic equipment ,such as dilution refrigerators and cryogen free ultra-low temperature systems.


The EMPIR project Real-K continued this work, examining the extremes of the current temperature scales - temperatures greater than 1300 K and less than 25K – to make the entire temperature range simpler and easier to apply. This has enabled in-situ traceability at lower cost, in applications such as manufacturing and the nuclear power sector.


Project website
Other Participants
Aalto-korkeakoulusäätiö sr (Finland)
Federal State - Owned Unitary Enterprise "All Russian Research Institute for Optical and Physical Measurements" (Russian Federation)
National Institute of Metrology - NIM (China)
Royal Holloway and Bedford New College (United Kingdom)
Technical Institute of Physics and Chemistry of the Chinese Academy of Sciences (China)
Università degli studi della Campania Luigi Vanvitelli (Italy)
Université Paris 13 (France)