--> On the Laboratory Analysis of Natural Gas Hydrate, by John A. Ripmeester, Hailong Lu, Igor L. Moudrakovski, and Yu-Taek Seo; #90035 (2004)

Datapages, Inc.Print this page

ON THE LABORATORY ANALYSIS OF NATURAL GAS HYDRATE

John A. Ripmeester1, Hailong Lu2, Igor L. Moudrakovski1, and Yu-Taek Seo1
1 Steacie Institute for Molecular Sciences, National Research Council Canada, Ottawa, Ontario, Canada
1 Terrain Science Division, Geological Survey of Canada, Ottawa, Ontario, Canada

Natural gas hydrate samples, in massive form as well as in varying degrees of dispersion in mineral matrices, have now been recovered from a variety of locations. As considerable experience in sample handling and storage has been attained, and appropriate laboratory procedures have been developed, a suite of useful data is now beginning to emerge. In most instances it is relatively straightforward to derive structural information from a number of different instrumental techniques, and we can have confidence that results obtained for natural samples are reliable. However, if we consider that hydrates are likely to be quite heterogeneous, some care needs to be taken in analyzing hydrate samples for structure, composition and phase equilibrium properties, as it must be recognized that the results may well depend on the size of the sample and the details of the sample morphology. This is the case both for samples that are analyzed by decomposition, and those that are analyzed by instrumental methods. In this contribution we will examine the results obtained for a number of natural hydrate samples, discuss the limitations of the procedures that have been used so far, and make some suggestions for future improvements.

Hydrate structure has been determined from a number of different instrumental methods, including powder X-ray diffraction, 13C NMR spectroscopy and Raman spectroscopy. Powder X-ray diffraction methods as commonly practiced requires bulk samples, and the technique depends on the presence of a periodic lattice. It is limited only by a minimum size of the crystal as the individual reflections broaden as the crystal size decreases. So far, there has been little attempt to examine PXRD of hydrate in porous media. 13C NMR spectroscopy also requires samples in bulk (at least ~ 100mg) in order to obtain a spectrum in a reasonable time. In this case the analysis depends on the chemical shift of carbon-containing species, which depends not only on molecular structure but in many instances also on the size of the trapping site (hydrate cage). Therefore it is relatively straightforward to distinguish sI from sII hydrate, and mixtures can be identified quite readily. Since NMR spectra are inherently quantitative, it should be possible to determine the composition directly from the distribution of the guests over the cages using some of the classical hydrate models. However, this is only the case if the distribution of all of the guest species in the sample is known. If guests without an NMR signature are present, the procedure becomes more difficult, and some attempts should be made to obtain a complete inventory of the gases in the sample. Raman spectroscopy also uses a cage-dependent signature for specific vibrational modes of various guest species, especially in hydrocarbon molecules. Although easier to apply, the utility of Raman spectroscopy is more limited, as the spectra are inherently not quantitative, there is no progression of cage dependent shifts, and there is considerable overlap for regions of spectral interest (eg C-H vibrations in hydrocarbons). Except for pure hydrates, Raman spectroscopy will remain a qualitative technique unless a considerable amount of work is carried out to calibrate the Raman spectra of a variety of gas mixtures. However, a strong point in favour of the technique is that Raman spectroscopy interrogates a rather small sample volume, so that a macroscopic sample could be scanned to determine the degree of homogeneity of the hydrate phase.

Given the fact that hydrate samples may be heterogeneous, it becomes important to determine not only the gas composition, but also the stability field of the hydrate samples that are studied by the instrumental methods mentioned above. Although the general P, T conditions of recovered hydrate can be inferred from its location, or by measuring the P,T conditions for a macroscopic sample once it is recovered, this will give only information on the least stable hydrate in the sample. The local composition and stability fields may differ considerably and can give important information on the initial formation of hydrate and possible subsequent reformation/fractionation of gases. When this information is placed alongside the wet analysis, and visual assessment of the hydrate inside the core, we can expect to develop a much better understanding of the hydrate, its conditions of formation, and its history.

So far, most recovered hydrate samples are known to be primarily of sI with methane as the principal component, although a number of observations have suggested that small amounts of heavier gases are present as well. However, there are known instances of natural sII hydrate, and recent results obtained for some Cascadia samples suggest both sII and sH hydrates to be present. As there are many organic species present, some of which are trapped inside the hydrate, with others on the hydrate surface, it is clear that a lot of work remains to be done to try and understand the interplay of structural type, guest distribution and hydrocarbon fractionation. As well, a better understanding of local stability fields for natural hydrate will be necessary both for the estimation of resources, and for designing processes for recovery of hydrate from natural deposits.