PROJ00326 - Laserfiche WebLink

Home Browse Search

POROSITY ESTIMATES FROM GEOPHYSICAL LOGS Porosity data derived from geophysical logs are of value in calculating specific yield from the specific- retention/grain-size regression model and also are an integral part of other techniques for estimating specific yield from geophysical logs. Because geophysical logs can be subject to error or application limitations, the effects of log errors on porosity and specific-yield estimates need to be investigated. The porosity of a formation can be determined by use of geophysical logs, such as sonic logs, neutron logs, or density logs. Each of these logs can accurately estimate the porosity of a formation if equipment and conditions in the borehole closely approximate the assumptions used in developing the logs. Sonic logs are based on a number of assumptions that cannot be evaluated readily in the borehole without analyses of core samples, This limitation makes the sonic log the poorest indicator of porosity among the three logs (Keys, 1990). Neutron logs and density logs can pro- duce comparable results. However, neutron logs are best suited to measuring porosity of dense materials of small porosity, and density logs are best suited to mea- suring porosity of low density materials of !luge poros- ity. The low density, large porosity clastic materials of most aquifers are appropriate for porosity evaluation by density logs. This conclusion is further supported by Patchett and Coalson (1979, p. 11) who reported that "multiple porosity logs do not necessarily improve the ability to obtain porosity from logs in sandstones," and by Johnson and Linke (1978) who concluded that, for sandstones in the MacKenzie Delta, the density log alone yielded more accurate porosity estimates than multi-porosity methods. Density logs were selected as the principal porosity log for the work described in this report because of the potential accuracy of the log in sandstone environments, the general ease of log inter- pretation, and the greater availability of this log in the water-well industry. Additional information on log- ging procedures and log interpretations is contained in Hearst and Nelson (1985) and TIttmann (1986). A density log measures the bulk density of a for- mation by use of Compton scattered gamma radiation. A radioactive source that emits medium-energy gamma radiation is held against the borehole wall by the log- ging tool. Gamma radiation interacts with electrons in the atoms of the formation in a process called Compton scattering. In this process, the medium energy gamma radiation is absorbed by the electrons and lower energy gamma radiation is emitted. This lower energy radia- tion is detected and counted by the logging tool as an indication of the electron density in the formation, The known atomic and molecular structure of common water-filled geologic materials enables bulk density to be calculated from electron density. A density porosity log is calculated from the bulk density log by use of the equation: _ Pg- Pb $-p-p' g f (2) where $ = the porosity, P g = the grain density, Pb = the bulk density, and Pf = the fluid density. In freshwater aquifers composed of clastic mate- rials, grain density is assumed to equal the mean grain density of most clastic sediments (2.65 glcm3) and fluid density is assumed to be 1.0 glcm3. Some errors in density porosity logs can be iden- tified, and remedial steps can be taken to decrease the effect of the error on specific yield estimated from the log. Errors in density porosity logs primarily are caused by: I. The quality control of the logging equipment and procedures. 2. The stochastic nature of the radioactive logging process. 3. The assumed mean grain density of 2.65 glcm3. 4. The irregularity (rugosity) of the borehole. Quality-Control Errors Quality control in geophysical logging has been reported by Patchell and Coalson (1979) to be some- what variable. Density porosity logs were shown to range from grossly erroneous (logs indicating porosity less than zero) to a set of 93 logs from the Wattenberg gas field in Colorado, 48 of which indicated identical porosities in a calibration interval. Only 16 of the 93 logs required a bulk-density correction greater than :t 0.02 glcm3 to also indicate the correct porosity in the calibration interval. The log corrections had a mean of zero, a standard deviation of 0.018 glcm3, range of -0.08 to +0.03 glcm3, and were approximately nor- mally distributed. For comparison, the bulk density was calculated for a 20-ft sandstone interval from 1,915 to 1,935 ft, using logs from wells A3, USGS, and holes CI and CIA. The four logs were run at different times by dif- 10 Techniques lor Estimating Specific Yield and Specific Retention from Grain-Size Data and Geophysical Logs from Clastic Bedrock Aquifers