Use of a Void Structure Model to Obtain Additional Information from Mercury Intrusion Porosimetry

A software package called "Pore-Cor Research Suite" has been developed at the University of Plymouth by Peter Matthews ( pmatthews@plymouth.ac.uk ), Reader in Applied Physical Chemistry, and his research group. It approximates the void structure of porous media with a network geometry which is simple but nevertheless versatile enough to give a close match to mercury intrusion porosimetry. The modelling process leads to improved categorisation of plug samples, and additional insights such as an explanation of absolute permeability changes. For further information on the software, the research group and its publications, and to view some structures in Virtual Reality, see http://www.pore-cor.com .

As an example, we describe the modelling of a sample studied as part of a consultancy project commissioned by Oil Plus, Abingdon , UK . for an oilfield operated by ChevronTexaco, the results of which were presented at the Society of Core Analysts International Symposium in Pau , France , September 2003.

The plug comprised a poorly consolidated, lateral turbidite sandstone, with a porosity of 27.89%. The applied mercury pressures of a mercury intrusion measurement were converted to pore diameters using the Laplace equation, assuming the standard surface tension for mercury and an advancing contact angle of 140°. Ten fits were carried out using a Boltzmann-annealed Simplex, each involving a different stochastic realisation, Figure 1.

Figure 1: Comparison of Experimental and Simulated Mercury Intrusion Curves (Click for larger view)

Figure 2 shows the pore and throat size distributions, expressed as the number of features per cubic millimetre of sample, which generate one of the simulated mercury intrusion curves.

Figure 2: Pore and Throat Size Distributions for One Stochastic Realisation (Click for larger view)

The unit cell of the first stochastic realisation is shown in Figure 3, with 58.1% of the accessible void volume filled with mercury (shown grey). The green scale bar is of length 1 mm. Note that the simulated void structure is not homogeneous - it has slight vertical banding/lamination which is highlighted by the position of the mercury. Periodic boundary conditions allow the unit cell to repeat infinitely in each direction.

Figure 3: Unit Cell of the Simulated Void Structure, 58.1% Full of Mercury (Shown Grey) (Click for larger view)

The effect of occluding the structure is shown in Figure 4, in which the bottom axis shows the maximum allowable particle diameter (MAPD). When compared to the traditional capillary bundle approximation, also shown, the Pore-Cor model suggests a higher diameter can be occluded (a feature of 24 µm, or particle of 8 µm) before permeability reduction is significant, whereupon the permeability reduces more sharply, falling to zero at a particle diameter between 15 and 18 µm.

Figure 4: Effect of Formation Damage Shown for 10 Stochastic Realisations and Averaged, Assuming 3-Particle Bridging (Click for larger view)

Figure 5 shows a stochastic realisation of the unit cell at the point at which the colloidal inclusions are just severe enough to cause the permeability to drop to zero. The structure has then been intruded with a non-wetting fluid (mercury or oil) shown darker grey, from the top face, to the maximum extent which can be intruded. It can be seen that the extent of intrusion is low, with access to many large pores being blocked by colloidal inclusion of small throats.

Figure 5: Non-Wetting Fluid (e.g. Oil or Mercury) Entering the Sample, Which Has its Formation Damaged Just to the Extent that its Gas Permeability has Dropped to Zero. Increase in Pressure on the Non-Wetting Fluid Will Not Increase the Intrusion of the Non-Wetting Fluid, Unless Colloid is Displaced. (Click for larger view)

Figure 4 showed that there are significant variations between stochastic realisations, which would be reduced by the averaging effects of using a larger unit cell. Nevertheless, there is a clear trend and meaningful average obtained from the present model. The geometry is clearly much cruder than that usually used for pore-scale modelling, for example by Blunt and co-workers at Imperial College , reported in a recent issue of this newsletter. However, mercury intrusion porosimetry is a cheaper and more standard method than the micro CT scanning which Blunt's structures are based on. So the new software provides a cost-effective method of obtaining additional useful information from existing and new mercury intrusion curves.

Disclaimer: The material available on this website is designed to provide general information only. Whilst every effort has been made to ensure that the information provided is accurate, it does not constitute legal or other professional advice.
Please note: The Department of Trade and Industry cannot be held responsible for the contents of any pages referenced by an external link.