Electricity: measuring and testing – Particle precession resonance – Using well logging device
Reexamination Certificate
1999-10-29
2001-05-08
Oda, Christine (Department: 2862)
Electricity: measuring and testing
Particle precession resonance
Using well logging device
C324S300000, C324S307000, C324S312000
Reexamination Certificate
active
06229308
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to nuclear magnetic resonance techniques and, more particularly, to a method for evaluating a formation using nuclear magnetic resonance measurements.
BACKGROUND OF THE INVENTION
Conventional formation evaluation methods for detecting hydrocarbons have relied largely on electrical log measurements of either formation resistivity or conductivity and on measurements of formation porosity from neutron, density or sonic tool measurements. The utility of electrical log measurements for hydrocarbon detection, and also for quantitative estimation of hydrocarbon saturations, is dependent on empirically based saturation equations like the well known Archie equation or others including the Waxman-Smits and Dual-Water models. In many environments this traditional approach to formation evaluation provides accurate reservoir predictions. Nonetheless, the following types of occurrences are not uncommon: missed pay zones, inaccurate estimates of hydrocarbons in place and costly completions of zones that are not commercial. The causes of these occurrences include formation waters of unknown or variable salinity, fresh formation waters, clay conductance effects on measured resistivity, inaccurate inversion of resistivity data, and formations that have anomalous values for the Archie parameters.
The evaluation of hydrocarbon reservoirs using pulsed NMR logging tools offers the potential to provide a solution to the problem in formnation evaluation stemming from the fact that many hydrocarbon reservoirs can be misinterpreted or even missed altogether by conventional resistivity based evaluation methods.
An NMR approach that uses a “differential methodology” was proposed in publications of Akkurt, et al. (NMR Logging Of Natural Gas Reservoirs, Paper N presented at the 36
th
Annual Meeting of the Society of Professional Well Log Analysts, 1995). This methodology involves making two NMR spin-echo measurements with different wait times; that is, different times for polarization or re-polarization of the spins. The raw measurements (detected spin echo signals), or the T
2
-distributions computed from these measurements, are subtracted to yield a “differential signal” (either a differential T
2
spectrum or echo train) that can be further processed to estimate hydrocarbon filled porosity. In the NMR well logging literature, some of the differential methods are called differential spectrum method (DSM) and time domain analysis (TDA). The wait times of the methods are selected so that the differential signal contains small contributions from the brine in the formation. In order to select proper wait times so that the brine contribution is canceled, knowledge of the NMR properties of the fluids in the formation is required. This is a limitation of these methods for oil exploration logging. Moreover, the interpretation of the technique requires that the T, distribution of the brine phase not overlap with the T
1
spectra of the hydrocarbon phases. In carbonate reservoirs and in reservoirs containing light to intermediate viscosity oils (e.g., 1-50 cp), the brine and hydrocarbon TI-distributions can overlap. This limits the applicability of the differential methods to shaly sands containing very low viscosity oils and gas. A recent paper, Akkurt et al. (Enhanced Diffusion: Expanding the Range of NMR Direct Hydrocarbon-Typing Methods, Paper GG presented at the 39
th
Annual Meeting of the Society of Professional Well Log Analysts, 1998) noted the limitations of the DSM and TDA methods for oils with intermediate viscosities and proposed a method called the Enhanced Diffusion Method (EDM) that attempts to exploit the fact that the brine phase is more diffusive than intermediate viscosity oils. By increasing the echo spacing so that diffusion dominates the T
2
relaxation of the brine, an upper limit (T
2DW
) on the apparent T
2
can be achieved. To obtain the oil filled porosity, the Akkurt, et al. 1998 paper proposed integrating the apparent T
2
-distribution for relaxation times greater than T
2WD
. Although the basic concept of the EDM is believed to be valid, there are complications in practice that limit its reliability for detection of oil, including: (1) the apparent T
2
-distributions are broadened by the regularization (smoothing) that is applied by the processing to reduce noise artifacts, so integrating the apparent T
2
-distributions from a sharp brine cutoff can lead to predictions of oil in water zones; (2) the oil signal can have a short relaxation time tail that extends into the brine signal; (3) in exploration wells it cannot be assumed that the diffusivity of formation oils is less than that of water; and (4) in wells drilled with oilbase muds, it is difficult using the EDM concept to separate the filtrate signal from that of the native oil.
A recent paper by Chen et al. (Estimation of Hydrocarbon Viscosity With Multiple TE Dual Wait-Time MRIL Logs, Paper 49009 in the Transactions of the 1998 SPE Annual Technical Conference and Exhibition, 1998), proposes a method for combining dual-wait time and multiple echo spacing data to estimate oil viscosity. The differential methodology is used to combine the different measurements. The spin-echo trains from long and short wait time data acquired with the same echo spacing are subtracted to eliminate the water brine signal. This method has the limitations discussed above. Furthermore, the subtraction of the differential signal increases the noise by a factor of 1.4 which is also one of the drawbacks of the TDA method.
A fundamental weakness of the aforementioned inversion methods is that separation of the measured data into brine and hydrocarbon signals is only done in an ad hoc manner in fitting the differential signal. An approach that makes this separation at the outset is disclosed in Looyestijn (Determination of Oil Saturation from Diffusion NMR Logs, Paper SS presented at the 37th Annual Meeting of the Society of Professional Well Log Analysts, 1996), which uses “diffusion processing” to compute the oil saturation from NMR data acquired with different echo spacings. Looyestijn fits the measured data to a model that explicitly includes the brine and oil signals. The model used five simple exponentials for the brine phase and a stretched exponential for the oil phase and was applied to NMR log data from a development well drilled with a waterbase mud. The oil relaxation times were known from lab measurements on produced samples and the oil and brine filled porosities were computed from the log data. In a Published PCT International Patent Application further describing the work of Looyestijn and his colleagues (WO 97/34166 of R. Bonnie, M. Johannes, P. Hofstra, W. Looyestijn, , R. Sandor and J. Karl), there is disclosed a technique for determining the fraction of a fluid selected from at least two fluids in a formation that includes the following steps: selecting a relationship between the NMR echo response from the fluids, the fractions of the fluids, and at least one variable which affects the NMR echo response in a manner dependent on the fractions of the fluids, varying the at least one variable, such as wait time or pulse spacing, in the course of an NMR measurement to thereby affect the NMR echo response in a manner dependent on the fractions of the fluids, and determining the fraction of the selected fluid by fitting the NMR echo response to the selected relationship. An example set forth in WO 97/34166 involves the determination of water saturation in a rock formation containing a medium gravity oil and water by applying a gradient magnetic field NMR measurements on a sample of the rock formation. The water was modeled with two transverse relaxation times and two corresponding volume fractions, the component with the short relaxation time representing bound water and the component with the long relaxation time representing movable water. The oil was modeled by one transverse relaxation time and one corresponding volume fraction. The WO 97/34166 Publication states that by repeating their meth
Jeffery Brigitte L.
Oda Christine
Ryberg John J.
Schlumberger Technology Corporation
Shrivastav Brij B.
LandOfFree
Formation evaluation using magnetic resonance logging... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Formation evaluation using magnetic resonance logging..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Formation evaluation using magnetic resonance logging... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2491694