Correlative multiple porosimetries for reservoir sandstones with adoption of a new reference-sample-guided computed-tomographic method

One of the main interests in petroleum geology and reservoir engineering is to quantify the porosity of reservoir beds as accurately as possible. A variety of direct measurements, including methods of mercury intrusion, helium injection and petrographic image analysis, have been developed; however, their application frequently yields equivocal results because these methods are different in theoretical bases, means of measurement, and causes of measurement errors. Here, we present a set of porosities measured in Berea Sandstone samples by the multiple methods, in particular with adoption of a new method using computed tomography and reference samples. The multiple porosimetric data show a marked correlativeness among different methods, suggesting that these methods are compatible with each other. The new method of reference-sample-guided computed tomography is more effective than the previous methods when the accompanied merits such as experimental conveniences are taken into account.


Supplementary Figure S2
Two successive core plugs (segments) were obtained from an original Berea Sandstone core, which were designated BS1-* and BS2-* respectively. A total of six mm-scale cores were extracted from the first plug (BS1-*), and then four among the six were used for the porosimetric experiment of helium injection method followed by mercury intrusion method. The rest two were for the porosimetries of CT and SEM_PIA methods. The top of the first plug was almost entirely ruined by the dense subsampling thus we had to secure the second plug (BS2-*) to extract additional mm-scale cores for further experiment of different items.

Supplementary Figure S3
The second core plug from an individual Berea Sandstone core, i.e., BS2-*, was used to extract two mm-scale core plugs, both for the porosity measurement using computed tomography (CT) under significantly improved resolution control, compared to the CT experiment of low resolution in the first sample set. Here one plug was used only for the sample of porosity measurement (measurement sample, MS), whereas the other was used only for the sample of reference (reference sample, RS). Adopting the concept of the reference-sample-guided computed tomography (Jin et al., 2013), a measurement sample and a reference sample were always paired to be scanned concurrently using the CT system (the same experimental procedure as it was in the first sample set). After the CT scanning, the mm-scale core plugs were manufactured into polished section for the application of the SEM_PIA method.

Supplementary Figure S4
Mineralogic composition was examined using the XRD equipment D8 ADVANCE and accompanied softwares DIFFRAC.EVA and Bruker TOPAS. Quartz is predominant throughout all the samples.
Quartz contents are low in sample group numbers BS1-1 and BS1-3 and high in BS1-2 and BS1-4, similarly with the trend of porosities (See also the main Figs. 1 and 2 for porosities). In case of the sum of other minerals, the reverse trend is true.

Supplementary Figure S5
Results of petrographic image analysis on scanning electron microscopic images (SEM_PIA) are shown. Note sample group numbers belonging to either the first sample set (light gray squares) or the second one (dark gray and black squares). The same gray-tone squares in each sample group represent a single polished section and an individual square corresponds to a different quarter of the polished section surface. Note also a large box with the dashed line which envelops all the squares in each sample group. The clusters of squares indicate a zigzagging variation of porosities along with the arranged sample group numbers, similarly with the trend of porosities measured by other porosimetric methods. Results of porosity measurements using mercury intrusion (MI) and helium injection (HI) methods are shown. Both methods were applied together in a series to each sample using different equipments, i.e., first helium injection by the pycnometer AccuPyc TM II 1330 and second mercury intrusion by AutoPore IV 9520. In calculation of HI porosities, the data of bulk volume (or bulk density) were brought from the records of the MI method. * The value in parenthesis of individual RS numbers represents the porosity value adopted from the MI porosity of practically the same sample and assigned to the RS in order to derive the measured porosity from individual RSs.  Raw blocks vary in size. They can be as large as 7-9 ft in length, 3-5 ft in width, and 2-4 ½' in height. Samples that are tested at Weatherford Labs are only 1 ½" dia x 2" long cut from these large blocks. The values we receive after Weatherford tests the samples are the n assigned to the whole block. That is why the cores cut from the blocks are sometimes q uite different from the values assigned, and why we say the values are for reference only, and not to be used for testing purposes.

Results
Paul.Lincoln@weatherfordlabs.com (The laboratory testing Berea sandstone samples) The grain volume of plug samples is measured by helium injection using the Boyle's Law method. Bulk volume determination is made by mercury immersion, Archimedes, or caliper measurements of length and diameter. Ambient porosity is then calculated using the formula: