(87d) Advanced Thermodynamic Property Models for Natural Gas, LNG and Related Mixtures

In the 1990’s the availability of new equation of state models led to new strategies in natural gas custody transfer. On-line density measurements for the conversion of measured volume flows into mass and finally energy flows were replaced by temperature, pressure and composition measurements via gas chromatography. Densities at pipeline states are calculated rather than measured. Similar developments took place for custody transfer in the LNG business. However, the equation of state model commonly used for LNG-density calculations, the revised Klosek and McKinley correlation, is based on a much weaker data base and is strictly speaking valid only for saturated boiling liquid, not for pressurized LNG.

With the GERG-2008 equation of state a thermodynamic property model became available, which yields a highly accurate description not only of densities but also of caloric properties for natural gas and related mixtures in a broad temperature and pressure range – from typical pipelining states over supercritical states relevant, e.g., for LNG processing down to liquid states at cryogenic temperatures. In the pipelining range the GERG-2008 model is based on very extensive and accurate data sets –its accuracy is very high in this range. In other ranges of temperature, pressure and composition it is still the most accurate model available, but its accuracy is lower due to limited data bases. It has been shown that the use of highly accurate property models is advantageous not only for custody transfer calculations but, by far more demanding, also for process design and optimization.

Limitations of the GERG-2008 model became obvious as well, particularly as a result of the discussion on CO₂-rich mixtures relevant for carbon capture and storage (CCS) applications. Today, there are international efforts to develop property models which are improved for certain applications, but which still build on the GERG-2008 model or are at least compatible to it to make use of its tremendous achievements for the description of natural-gas like multi-component mixtures. Today one may distinguish three different developments:

Improvements for mixtures different from natural gases: The GERG-2008 model covers an extremely broad range of components, but mixtures with some of them received rather little intention because they are typically considered minor components. When talking about different applications, these components may become main components, e.g. CO₂-rich mixtures with water when talking about CCS. In this case limited improvements are possible even based on the existing sets of experimental data. For CCS relevant mixtures international measuring campaigns were launched, which will soon lead to a substantially improved data base.

Extension to applications not covered by highly accurate property models to date: To allow for a consistent description of effects like hydrate formation, state of the art fluid models like GERG-2008 are combined with models describing different (very often solid) phases or components like electrolytes.

Improvement for natural gas like mixtures at temperatures and pressures not (highly) accurately described yet: In fact this point is very relevant for LNG applications, but difficult to address. Without a substantial body of new accurate experimental data no significant improvements will be possible.

The following talk by Richter et al. will present first results of a project aiming at the generation of highly accurate experimental data in temperature and pressure ranges relevant for LNG applications. Political and financial support by the LNG community is required to push such projects not only at Ruhr-University Bochum.