Nuclear Magnetic Resonance Logging in the Delaware Formation

Logan, D. (Schlumberger Wireline & Testing) | Morriss, C.E. (Schlumberger Wireline & Testing) | Williams, R. (GeoQuest)

OnePetro 

Abstract

The Delaware formation is a fine grained sandstone located in West Texas and Southeast New Mexico. Resistivity- based log interpretation in this formation has proved unreliable in many cases. In particular, deep invasion and high irreducible water volumes result in calculated water saturations that rarely reflect future production. Because of this, pay zones are identified primarily from mud logs and sidewall cores. However, our studies have demonstrated that borehole nuclear magnetic resonance (NMR) measurements are useful for evaluating the Delaware formation. The general use of NMR measurements for the estimation of porosity, pore size, permeability, producible porosity and bound-fluid volume and for the identification of pay zones has been previously described.

Often, the interpretation of borehole NMR data is enhanced by the results of lab NMR measurements on core samples. For this reason, NMR measurements were made on 20 water-saturated cores, a crude oil sample from the Delaware and a partially oil-saturated core sample.

Water-saturated samples have narrow T2-distributions and maximum T2 values in the order of 200 ms. A typical distribution is shown in Figure la. The relatively short T2 values reflect the fine grains and small pore sizes typical of this formation. However, significant pore size variation between the samples results in a generally poor correlation between permeability and porosity. Permeability estimation is improved when an NMR parameter, such as logarithmic mean T2 or bound-fluid volume, is used together with porosity. Lastly, NMR free-fluid porosities were found to be in good agreement with the volume of water expelled from the core by centrifuging at 100-psi air-brine capillary pressure. The free-fluid porosities were computed from the T2-distributions using a 33-ms cutoff.

T2-distributions for the crude oil and a core sample partially saturated with the crude oil are shown in Figures 1b and lc. T2 values for the crude oil are long, predominantly in the 100 to 5000 ms range, reflecting the low viscosity of the oil. For the partially oil-saturated sample, T2-distributions were found to be considerably broader than for the case when the sample was completely saturated with water. Maximum T2 values increased from 200 ms to about 1000 ms. In addition, the mean T2 and also the free-fluid and bound-fluid porosities are significantly different when the sample is partially oil saturated; hence, permeability estimation using these parameters will be affected by the presence of oil.

NMR logs were recorded in the Delaware formation in four wells (A, B, C and D) located in Southeast New Mexico. The wells were logged with either the NML Nuclear Magnetism Logging tool or CMR Combinable Magnetic Resonance tool.

For Well A, irreducible water volume and permeability estimates from the NML log were in good agreement with core data and subsequent production, except in thin-bed intervals where the vertical resolution (about 4 ft.) of the NML log is inadequate. NML station logs were also successful in differentiating between known oil and water zones.

For Well B, T2-distributions from CMR station logs were used to identify oil-productive intervals. As observed with the lab measurements, water-saturated intervals have T2-distributions that end at approximately 200 ms whereas oil bearing intervals have maximum T2 values on the order of 500ms.

Oil bearing intervals may also be identified from T2-distributions obtained during continuous depth logging with the CMR tool. Figure 2 shows a comparison between CMR data and the total gas measurement from a mud log obtained in Well C.

P. 997