Accurate Thermal-Neutron Decay Time Measurements Using the Far Detector of the Dual-Spacing TDT A Laboratory Study

To improve the accuracy of saturation-change determinations in reservoirs, thermal-neutron-decay time was measured in the laboratory with a long-spacing TDT (trade mark) device. The far detector of a TDT-K sonde was used in 17.1 and 30.4% porosity sandstone formations for several formation-fluid salinities ranging from 0 to 247,000 ppm NaCl. A 7-in. (17.8-cm) casing cemented in a 10-in. (25.4-cm) borehole was used with and without a 2 7/8-in. (7.3-cm) tubing centered in a gravel pack. Complete decay curves are constructed from measurements made in successive channels of a multichannel analyzer. Values of formation intrinsic decay time calculated from nuclear-capture crosssections are compared with decay times measured with the far detector using the Scale Factor TM gating system. Results show That far-detector measurements are less influenced by diffusion and indicate the usefulness of the large source-detector spacing for determination of changes in water saturation () from logs run at different times in the history of a producing well. Although errors in values calculated producing well. Although errors in values calculated for are reduced by using values measured with the far detector, oar analysis shows that further improvement in accuracy can be obtained by using the correction data derived from the laboratory measurements. Introduction In the TDT time-lapse technique, changes in water saturation, Sw, are determined from the corresponding changes in the measured formation thermal-neutron-capture cross-section. The relationship used to find the saturation change, in an oil-bearing formation is ....................(1) where is the macroscopic thermal-neutron-capture cross-section of the formation at the time of a first measurement, is the cross-section at the time of a subsequent measurement, is the formation porosity, and and are the macroscopic porosity, and and are the macroscopic thermal-neutron-capture cross-sections of formation water and oil, respectively. It is clear from Eq. 1 that it is not necessary to know, the cross-section of the rock matrix. Furthermore, the accuracy of is not affected by systematic errors in the measured values of, if these errors are the same for the two measurements, and . When the errors in and are not equal, however, the effect on the result for from Eq. 1 may be significant. Therefore, for use in secondary- and tertiary-recovery programs, it is of interest to look for ways of improving accuracy available in the present state of the art. One possibility is to use a longer source-detector spacing for the TDT sonde. A longer spacing (632 mm) is available by using the far detector of the TDT-K tool. however, in most cases in practice, it is the near-detector recording that is shown for the curve on the log. This is because the higher counting rate available at this detector (spacing of 343 mm) is needed to obtain an acceptable statistical validity with a single pass in the well. The measurements with the longer spacing are useful when statistical uncertainty is reduced by recording slowly or by conducting several runs and averaging them. To investigate performance with the longer spacing, laboratory measurements determined the changes in systematic error caused by changes in the of the formation fluid when all other parameters were constant. We found that parameters were constant. We found that measurements with the far detector of the TDT-K sonde had smaller systematic errors that were less sensitive to changes in the formation fluid than those from the near detector. SPEJ P. 59