Field Applications of a New Nuclear Magnetic Resonance Fluid Characterization Method
- R. Freedman (Schlumberger Sugar Land Product Center) | N. Heaton (Schlumberger Sugar Land Product Center) | M. Flaum (Schlumberger Sugar Land Product Center)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Evaluation & Engineering
- Publication Date
- December 2002
- Document Type
- Journal Paper
- 455 - 464
- 2002. Society of Petroleum Engineers
- 1.11 Drilling Fluids and Materials, 4.3.4 Scale, 5.8.7 Carbonate Reservoir, 4.1.5 Processing Equipment, 1.6 Drilling Operations, 5.2 Fluid Characterization, 5.2 Reservoir Fluid Dynamics, 5.3.4 Reduction of Residual Oil Saturation, 5.6.1 Open hole/cased hole log analysis, 4.1.2 Separation and Treating, 5.2.2 Fluid Modeling, Equations of State, 5.6.4 Drillstem/Well Testing, 5.2.1 Phase Behavior and PVT Measurements
- 3 in the last 30 days
- 513 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
This paper discusses the application of a new magnetic resonance fluid (MRF) characterization method to field data acquired by a nuclear magnetic resonance (NMR) well-logging tool. The theoretical and experimental foundation underlying the method was discussed in a previous publication.1 The MRF method provides a detailed formation evaluation of the near-wellbore region investigated by modern NMR logging tools. This information includes total and bound-fluid porosities, hydrocarbon saturations, oil viscosities, and hydrocarbon-corrected permeabilities. This paper presents MRF oil saturation and viscosity estimates in oil-bearing wells.
In this paper, we first provide a short review of the main ideas and key equations that make up the MRF method. We discuss the generalization of the MRF method to logging data acquired by the Combinable Magnetic Resonance (CMR)** tool and apply the method to mixture data acquired by a CMR tool on crude-oil and water samples.
We also discuss the existing database of MRF results on partially saturated rocks and present new laboratory results that include blind tests on Yates field dolomite rocks measured at initial and residual oil saturation.
The sources of uncertainty inherent in NMR diffusion-based viscosity estimation in reservoir rocks are discussed, and we propose a strategy that accounts for the uncertainties and provides reasonable estimates of oil saturations and viscosities. The strategy is applied to the MRF analysis of the field data in this paper.
The MRF method is applied to CMR station log data acquired in three wells. Analysis of the MRF data and the results from each of these wells is discussed.
It is well known that NMR logging tools provide a wealth of information to formation evaluation that is simply not obtainable from other well-logging measurements. Foremost among the many physical properties probed by NMR is molecular diffusion. Because water molecules typically diffuse much faster than oil molecules and much slower than gas molecules, NMR diffusion measurements provide a means for detection and differentiation of reservoir fluids. This capability has generated much excitement since the introduction of modern pulsed NMR logging tools and has led to dozens of industry publications on different fluid-typing methods and applications. Papers by Looyestijn2 and Slijkerman et al.3 discussed methods based on inversion of a multifluid relaxation model. While the use of a multifluid relaxation model represented an improvement over earlier methods, the model discussed in these papers did not properly account for the fact that crude oils have broad distributions of molecular diffusion coefficients.
Despite a considerable effort on the part of service companies and oil companies, the development of an accurate and reliable NMR fluid-typing method has been limited by the lack of a detailed understanding of molecular diffusion and NMR relaxation in hydrocarbon mixtures.
The goal of this paper is to present some field applications of a new MRF characterization method. This paper builds on the theoretical and experimental work of Freedman et al.1
The next section discusses the main ideas and equations underlying the method. This includes a review of the constituent viscosity model (CVM) and the associated MRF method multifluid relaxation model. It also discusses how the MRF multifluid relaxation model is generalized to account for the magnetic field gradient distribution of the CMR tools. The "Laboratory Experiments" section of the paper presents some laboratory results on the analysis of water/crude-oil mixture data acquired by a CMR tool. This section also presents some new laboratory results on the application of the MRF method to rocks partially saturated with crude oils and shows our current database of laboratory results on partially saturated rocks. The section entitled "Accuracy of NMRBased Oil Viscosity Estimates" discusses the uncertainties in NMR-based viscosity estimation. The final section, "Field-Test Data," discusses the application of the MRF method to field data from several wells.
Summary of the MRF Method
The MRF method provides, by simultaneous inversion of a suite of NMR data, a detailed formation evaluation of the near-wellbore region. The MRF method provides the following answer products:
Flushed-zone fluid saturations and volumes.
Total NMR porosities.
Bulk volumes of irreducible water.
Brine T2 distributions and T1/T2 ratios.
Crude-oil T2 and diffusion coefficient (D) distributions.
These outputs are computed by simultaneous inversion of a simple suite of NMR data using the MRF multifluid relaxation model. The computed brine and crude-oil T2 distributions are diffusion- free because the relaxation model explicitly accounts for molecular diffusion in the magnetic field gradient of the logging tool. The most important, powerful, and unique capability of the MRF method is its ability to accurately determine flushed-zone water saturations and oil viscosities. The MRF method also can be used to evaluate gas-bearing formations; however, the Density- Magnetic Resonance (DMR)** interpretation method4 offers an equally reliable and simpler method for gas-zone evaluation.
Two key ingredients underlying the MRF method are:
A new microscopic CVM that provides a link between diffusion- free NMR relaxation times and molecular diffusion coefficients in live and dead crude oils.
A new multifluid relaxation model that incorporates the CVM.
|File Size||2 MB||Number of Pages||10|