Abdulla Rawand D., ENTROPY-REGULARIZED GEOCHEMICAL ALLOCATION: AN INFORMATION-THEORETIC FRAMEWORK FOR APPORTIONING COMMINGLED PETROLEUM FROM WHOLE-OIL FINGERPRINTS
https://doi.org/10.15407/mineraljournal.48.03.062
UDC 550.84:553.98
Abdulla Rawand D., PhD
E-mail: rawand.abdulla@epu.edu.iq; orcid: 0009-0006-9662-0942
Department of Petroleum Technology, Koya Technical Institute,
Erbil Polytechnic University, Erbil, Iraq
ENTROPY-REGULARIZED GEOCHEMICAL ALLOCATION:
AN INFORMATION-THEORETIC FRAMEWORK FOR APPORTIONING
COMMINGLED PETROLEUM FROM WHOLE-OIL FINGERPRINTS
Language: English
Mineralogical journal 2026, 48 (3): 62-71
Abstract: Allocating commingled petroleum production to its contributing reservoir zones is a core task of oil-field geochemistry, yet conventional fingerprinting relies on a few hand-picked peak ratios and ordinary least-squares unmixing that becomes unstable when zone oils are geochemically similar (collinear) and analytically noisy. We introduce Entropy-Regularized Geochemical Allocation (ERGA), an information-theoretic framework that treats the entire whole-oil gas-chromatographic trace as a probability distribution. ERGA couples a variance-weighted, simplex-constrained data term with a maximum-entropy prior on the allocation vector, and derives a Jensen-Shannon residual-divergence statistic that tests whether an unmodelled (missing) end-member contributes to the mixture. The method was evaluated against ordinary non-negative least squares (NNLS) in a reproducible Monte-Carlo simulation (synthetic three-zone libraries spanning condition numbers from 9 to 83; 400 realizations per noise level). For well-resolved zones ERGA reduced allocation error by about 1.5—1.6× at all noise levels. For realistically collinear zones it lowered the root-mean-square error from 12.6 to 8.5 percentage points at 7 % noise (R² 0.63 → 0.79), the gain increasing with noise once the inversion is ill-posed. The detection statistic flagged a hidden zone with ≥80 % power once it exceeded about 12 % of production, at a 5 % false-positive rate. ERGA provides a principled, uncertainty-aware and self-diagnosing route to dynamic production allocation, transferable to reservoir surveillance and environmental source apportionment.
Keywords: petroleum geochemistry, production allocation, geochemical fingerprinting, maximum entropy, information theory, commingled oil, gas chromatography.
References / Література
Aitchison, J. (1986), The Statistical Analysis of Compositional Data, Chapman and Hall, London, 416 p. https://doi.org/10.1007/978-94-009-4109-0
Cover, T.M. and Thomas, J.A. (2006), Elements of Information Theory, 2nd ed. Hoboken, Wiley-Interscience, 748 p.
Halpern, H.I. (1995), Development and applications of light-hydrocarbon-based star diagrams, AAPG Bulletin, Vol. 79 (6), pp. 801-815. https://doi.org/10.1306/8D2B1BB0-171E-11D7-8645000102C1865D
Harris, C.R., Millman, K.J., van der Walt, S.J. et al. (2020), Array programming with NumPy, Nature, Vol. 585, pp. 357-362. https://doi.org/10.1038/s41586-020-2649-2
Hwang, R.J., Baskin, D.K. and Teerman, S.C. (2000), Allocation of commingled pipeline oils to field production, Organic Geochemistry, Vol. 31 (12), pp. 1463-1474. https://doi.org/10.1016/S0146-6380(00)00123-6
Jaynes, E.T. (1957), Information theory and statistical mechanics, Physical Review, Vol. 106 (4), pp. 620-630. https://doi.org/10.1103/PhysRev.106.620
Kaufman, R.L., Ahmed, A.S. and Elsinger, R.J. (1990), Gas chromatography as a development and production tool for fingerprinting oils from individual reservoirs: applications in the Gulf of Mexico, in: Proc. 9th Annual Research Conf., GCSSEPM Foundation, pp. 263-282. https://doi.org/10.5724/gcs.90.09.0263
Larter, S.R. and Aplin, A.C. (1995), Reservoir geochemistry: methods, applications and opportunities, in: Cubitt, J.M. and England, W.A. (eds.), The Geochemistry of Reservoirs, Geol. Society, London, Spec. Publ., Vol. 86, pp. 5-32. https://doi.org/10.1144/GSL.SP.1995.086.01.02
Lawson, C.L. and Hanson, R.J. (1974), Solving Least Squares Problems, Englewood Cliffs, Prentice-Hall, 340 p.
Lee, D.D. and Seung, H.S. (1999), Learning the parts of objects by non-negative matrix factorization, Nature, Vol. 401 (6755), pp. 788-791. https://doi.org/10.1038/44565
Lin, J. (1991), Divergence measures based on the Shannon entropy, IEEE Transactions on Information Theory, Vol. 37 (1), pp. 145-151. https://doi.org/10.1109/18.61115
Peters, K.E., Walters, C.C. and Moldowan, J.M. (2005), The Biomarker Guide. 2nd ed., Cambridge, Cambridge Univ. Press. https://doi.org/10.1017/CBO9780511524868
Skilling, J. and Bryan, R.K. (1984), Maximum entropy image reconstruction: general algorithm, Monthly Notices of the Royal Astronomical Society, Vol. 211 (1), pp. 111-124. https://doi.org/10.1093/mnras/211.1.111
Virtanen, P., Gommers, R., Oliphant, T.E. et al. (2020), SciPy 1.0: fundamental algorithms for scientific computing in Python, Nature Methods, Vol. 17, pp. 261-272. https://doi.org/10.1038/s41592-019-0686-2