International Round Robin for Coke Reactivity Index, Coke Strength after Reaction and I600.

The quality of a coke is commonly defined by its reactivity and strength, these parameters indicate the utilisation behavior in the blast furnace. In a laboratory environment these parameters are described by the coke reactivity index (CRI), coke strength after reaction (CSR) as well as the various drum tests (ASTM, MICUM/IRSID and JIS) conducted at ambient temperature.

This project investigated the repeatability and reproducibility of ISO 18894:2018 Determination of coke reactivity index (CRI) and coke strength after reaction (CSR) in practice. It also investigated the feasibility of using the same test portion (-22.4+19mm) to measure room temperature strength known as I600. Subsamples from two cokes with varying quality (CSR/CRI) were produced in the Research Coke Oven at CSIRO and despatched to laboratories for analysis. A total of 12 laboratories returned results, five were from within Australia and the remaining seven from overseas.

This round robin exercise demonstrated that:

• The variability in results is greater for a sample with lower quality (indicated by lower CSR and higher CRI).
• Within a laboratory, variability between replicates (repeatability) is within acceptable limits based on ISO 18894.
• For the higher quality sample, the variability between laboratories (reproducibility) is considered acceptable although for lower quality samples variability in test results is greater than the limits listed in the ISO standard.

This may be explained by the range of equipment and practices still considered 'allowed' within the standard although for pragmatic reasons further tightening of critical parameters would be undesirable. However, from the results obtained it was clear that selection of a single screen aperture for use in the test would be an area for potential improvement or at the very least further investigation. On the whole, ISO 18894 :2018 appears to have improved standardisation and further 'tightening' of the published repeatability and reproducibility limits is unnecessary. The test results from this round robin showed lower calculated reproducibility than the round robin conducted in 2016 as part of ACARP project C24057 with a similar number of participants. Future round robin exercises assessing CSR and CRI should carefully consider coke selection to avoid results crossing precision ranges. In addition, wider participation in future exercises, particularly from mainland China would be desirable.

Not all participants conducted the I600 test, and most are not routinely asked to conduct the test. Analysis of the gathered test data showed minimal bias in the test results and no obvious test parameters impacting the results. The test results can provide a good indicator of coke abrasion strength when there is insufficient sample to complete conventional tumble tests, I600 could be considered for inclusion in ISO 18894 at future meetings of ISC TC27/SC3.