SPE Journal
Volume 14, Number 1, March 2009, pp. 50-53

SPE-104342-PA

Numerical Method of Evaluating Elemental Content of Oil-/Water-Saturated Formations Based on Pulsed-Neutron Gamma Inelastic Log Data

View full textPDF ( 199 KB )

DOI  More information 10.2118/104342-PA http://dx.doi.org/10.2118/104342-PA

Citation

  • Khisamutdinov, A.I. and Phedorin, M.A. 2009. Numerical Method of Evaluating Elemental Content of Oil-/Water-Saturated Formations Based on Pulsed-Neutron Gamma Inelastic Log Data. SPE J.  14 (1): 50-53. SPE-104342-PA.

Discipline Categories

  • 6 Reservoir Description and Dynamics

Summary

A numerical method is proposed for solving an inverse problem for a pulsed-neutron gamma inelastic log. The objective is to determine a composition of the in-pore fluid and of the rock skeleton by gamma ray measurement data in spectral windows of four elements: C, O, Ca, and Si. Inversion of the data is based on an iteration procedure, on each step of which the corresponding direct problem is solved for transport equations of neutrons and gamma quanta. Results of numerical experiments are presented that confirm a convergence to exact solution for most of the unknowns. The proposed numerical method is the development of an approach for solving inverse problems in nuclear geophysics (Khisamutdinov and Blankov 1989; Khisamutdinov and Minbaev 1995; Khisamutdinov 1999; Khisamutdinov and Phedorin 2003; Khisamutdinov 2005).

Introduction

Inversion methods are an important component of nuclear-geophysical technologies. The methods commonly used now are approximate. In some methods, the transport equations are replaced with diffusion equations; in others, measured gamma ray spectra are approximated with a linear combination of given standard monoelemental spectra, using laboratory measurements for further evaluating the elemental content of a formation. We do not review inversion methods here, but we provide a few references (Hertzog 1980; Herron 1988; Roscoe et al. 1991; Gilchrist et al. 1999). In this paper, we develop a rigorous method, contrary to approximate ones.

The problem of pulsed-neutron gamma logging with measurements of inelastic scattering gamma-quanta (PNG-I) is discussed. This type of logging is based on irradiating the rock by monoenergetic neutrons emitted by a pulsed generator located in the borehole and subsequently registering the gamma ray energy spectra. This gamma radiation is caused by inelastic scattering of neutrons on the atoms of the medium. Measurements are taken during the pulse duration of the generator when most neutrons have not been slowed down yet. This enables us to distinguish the gamma ray spectrum of inelastic scattering from the capture gamma ray spectrum. Gamma ray energy that a nucleus irradiates at inelastic scattering of a neutron is uniquely characteristic for the element (the isotope) where the scattering has occurred. Modern PNG-I tools mainly are aimed at measuring spectral yields of four elements, C, O, Ca, and Si, which is related to high cross sections of inelastic scattering at these elements or to their high abundance in rocks and fluids.

Quantitative interpretation of experimental data is difficult for several reasons. Both the neutron field and the scattered gamma-quanta field depend not only on concentrations of the previously mentioned four elements in the formation but also on concentrations of other elements and on wellbore properties (including the fluid that fills it). In this paper, the numerical method is constructed and one particular problem of interpreting pulsed-neutron gamma-log data is solved. We treat the interpretation problem as an inverse problem of evaluating parameters for the transport equation. This approach to inversion of measured nuclear-geophysical data was developed in Khisamutdinov and Blankov 1989, Khisamutdinov and Minbaev 1995, Khisamutdinov 1999, Khisamutdinov and Phedorin 2003, and Khisamutdinov 2005; and it can be summed up as follows:

  • Qualitative elemental composition of the formation being investigated is chosen properly, the set of unknown parameters is fixed, and constraint equations are written out.
  • Equations for measurements are defined; in combination with constraint equations, they form the system of nonlinear equations for unknown parameters.
  • Taking into account the first and second items, a numerical method is constructed to solve the system of equations for the inverse problem, and this method should be convergent to the exact solution.

Results of numerical experiments shown in this paper seem to confirm the convergence of this method to the exact solution for oil/water saturation and for most mineral components of the formation. Note that unknown parameters are linearly contained in the coefficients of transport equations (i.e., in the cross sections of interactions between particles and the medium).

The present work relates to both geophysics and applied mathematics. We hold the following method: A realistic problem containing all key points is chosen to be solved, and we construct an inversion method for this problem. It is important that the method can be generalized to more-complicated and integrated problems. At first, a model of medium and measured values should be formulated as key objects. The method can be generalized, in particular, to a case of joint inversion of gamma spectra from inelastic scattering and radiative capture of neutrons.

View full textPDF ( 199 KB )

History

  • Original manuscript received: 17 September 2006
  • Meeting paper published: 3 October 2006
  • Revised manuscript received: 5 May 2008
  • Manuscript approved: 21 August 2008
  • Published online: 16 March 2009
  • Version of record: 1 March 2009