Page 210 - Machine Learning for Subsurface Characterization

Dimensionality reduction and clustering techniques Chapter 6 179

using miscible light-hydrocarbon injection. The three EOR-efficiency indices,
developed in this study, exhibit consistent predictions of the recovery potential
in the 200-ft-thick upper, middle, and lower shale formations. Among these
indices, the microscopic displacement (MD) index generates predictions at
the highest resolution. MD-index calculation requires accurate estimations of
miscible free-oil volume, free water-filled porosity, and bound fluid-filled
porosity. Miscible free-oil volume estimate accounts for the alteration in the
minimum miscibility pressure due to the pore-confinement effects. Free
water-filled and bound fluid-filled porosities are estimated using factor
analysis (a dimensionality reduction technique) of the NMR T2 distribution
response measured over 64 discrete T2 bins. KC-index is computed using
K-means clustering, which is consistent with the MD-index. KC-index has
the lowest resolution in the low-productivity upper and lower shales,
whereas the R-index has the lowest resolution in the productive middle
shale. These indices predict that 50% and 20% of the middle shale should
exhibit high and intermediate recovery potentials, respectively. Indices also
indicate that the recovery potentials of the upper and lower shales are
drastically lower than that of the middle shale formation.

References
[1] Han Y, Misra S, Simpson G, editors. Dielectric dispersion log interpretation in Bakken
petroleum system. SPWLA 58th annual logging symposium. Society of Petrophysicists and
Well-Log Analysts; 2017.
[2] Hawthorne SB, Gorecki CD, Sorensen JA, Steadman EN, Harju JA, Melzer S, editors.
Hydrocarbon mobilization mechanisms from upper, middle, and lower Bakken reservoir
rocks exposed to CO. SPE unconventional resources conference Canada; 2013 5–7
November. Calgary, Alberta, Canada: Society of Petroleum Engineers; 2013.
[3] Sorensen JA, Braunberger JR, Liu G, Smith SA, Hawthorne SA, Steadman EN, et al.
Characterization and evaluation of the Bakken petroleum system for CO 2 enhanced oil
recovery. In: Unconventional resources technology conference; 20–22 July; San antonio.
Texas, USA. URTEC: Unconventional Resources Technology Conference; 2015. p. SPE-
178659-MS.
[4] Taber JJ, Martin FD, Seright RS. EOR screening criteria revisited—part 1: introduction to
screening criteria and enhanced recovery field projects. SPE Reserv Eng 1997;12(03):189–98.
[5] Thomas B. Proposed screening criteria for gas injection evaluation. J Can Pet Technol 1998;
37(11):14–20.
[6] Al-Adasani A, Bai B. Recent developments and updated screening criteria of enhanced oil
recovery techniques. In: International oil and gas conference and exhibition in China; 8–10
June; Beijing, China. SPE: Society of Petroleum Engineers; 2010. p. SPE-130726-MS.
[7] Bourdarot G, Ghedan SG. Modified EOR screening criteria as applied to a group of offshore
carbonate oil reservoirs. In: SPE reservoir characterisation and simulation conference and
exhibition; 9–11 October; Abu Dhabi, UAE. SPE: Society of Petroleum Engineers; 2011.
[8] Rivas O, Embid S, Bolivar F. Ranking reservoirs for carbon dioxide flooding processes. SPE
Adv Technol Ser 1994;2(01):95–103.
[9] Diaz D, Bassiouni Z, Kimbrell W, Wolcott J. Screening criteria for application of carbon
dioxide miscible displacement in waterflooded reservoirs containing light oil. In: SPE/DOE