Design, construction and performance reliability verification evaluation of a high pressure volumetric sorption system for CO₂ sorption in South African coals

High pressure sorption equipment are based on the gravimetric or volumetric principle method of gas sorption onto a specific sorbent. This paper discusses the design, construction and performance consistency testing that was conducted on a recently constructed in-house high pressure volumetric sorption system (HPVSS) in South Africa (SA) which was used for the evaluation of supercritical carbon dioxide (CO₂) sorption on SA coals. A detailed description of the equipment, the commissioning phase, leak testing evaluation, estimation of sample, reference and void volume, as well as the method of conducting the measurements of the sorption isotherms are explained. Six (6) South African (SA) coals were selected based on their myriad of properties i.e. coal rank, coal type, and petrographic and proximate analyses. The samples were used as the sorbing material for pure CO₂ in supercritical phase to conduct the isothermal sorption investigations. In order to evaluate the system, multiple reliability experimental runs were conducted and the isotherms were evaluated and the sorption data was verified using the one-way analysis of variance method (ANOVA) to ensure the equipment performance was statistically reliable. Both the, system temperature and pressures, were varied to determine the potential risks of leaks in order to validate the consistency of the system at varying conditions of supercritical gas injection. For the total number of experimental test runs that were conducted, was observed that very good repeatability of the sorption data was reproduced under all conditions. The data exhibited good homogenous variance in all data sets compared, thus demonstrating overall confidence in the repeatability of these CO₂ results, and that the reliability of the HPVSS has been confirmed in the test sorption experiments. A comparison to both national and international published data sets also confirmed good fit of the findings.