Comparison of preparation analytical capabilities for liquified hydrocarbon mixtures in constant pressure (piston) cylinders

Project Description

Traceable liquid hydrocarbon mixtures are required in order to underpin measurements of the composition and other physical properties of LPG (liquefied petroleum gas) and LNG (liquefied natural gas), thus meeting the needs of an increasingly large European industrial sector.

The development of traceable liquid hydrocarbon standards by NMIs is still at a relatively early stage. NPL and VSL have recently developed methods for the preparation and analysis of such standards in constant pressure (piston) cylinders, but the comparability of standards and measurements between the NMIs has not been assessed, and no there are no CMCs for these mixtures.

This proposed bilateral comparison therefore aims to compare the preparation and analytical capabilities of NPL & VSL for these mixtures.

VSL and NPL will each preapare a mixture of the following compostion in a constant pressure (piston) cylinders:

Propane 20;    propene 20;   iso-butane 708;   n-butane 80;   iso-butene 70;   but-1-ene 80;   i-pentane 8;   n-pentane 6;   1,3-butadiene 8  (all units are mmol/mol).
The cylinders will then be exchanged and the amount fraction of each component in  mixture will be measured using in-house standards.
Details of the cylinders to be used will be communicated in advance of  start of the comparison, in order to ensure compatability.
It is planned that the cylinders will be exchanged in August - September 2011.

Final Report 2013-07-30

The comparison has been completed and the results are published in the KCBD.

The results of the comparison show a good agreement between the laboratories’ results and the comparison reference values for the six components with amount fractions greater than 1.0 cmol/mol (propane, propene, iso-butene, n-butane, iso-butane and 1-butene). Measurement of the three components with lower amount fractions (1,3-butadiene, iso-pentane and n-pentane) proved more challenging. In all but one case, the differences from the comparison reference values for these three components were greater than the expanded measurement uncertainty.