Using Rock Physics for Constructing Synthetic Sonic Logs

Detailed knowledge of velocity-depth trends is essential for constructing well-to seismic ties, determining background velocity and generating synthetic seismic traces for interpreting seismic amplitudes in terms of porosity, lithology and pore fluid type. The acoustic properties of shales are of particular importance in the exploration setting as shales constitute more than 80% of sediments and rocks in siliclastic environments. Seismic and sonic log data may be used for estimating overpressure in shales in order to maintain borehole stability during drilling. In addition, velocities in shales can be used for predicting its mechanical properties and hence, borehole collapse. Despite of the apparent importance of understanding shale acoustics, sonic logs in shales are frequently unavailable or unreliable especially for older wells. Hence, it is essential to compute pseudo-velocity logs from other available information. This can be accomplished by using empirical relationships or rock physics transforms. In this study we have compared the predictive power of resistivity (Faust 1951, Hubert 2008) and porosity based models (Wyllie 1956, Holt & Fjær 2003) for computing synthetic P-wave velocity logs using wireline log data from mudstone intervals at three North Sea wells. The wells 6507/2-1, 6507/2-2 and 6507/2-1 are located at the Haltenbanken area at the Norwegian Sea. The wells targeted hydrocarbon potential at the upper Cretaceous and lower Tertiary sandstone sequences. The sandstone beds are overlain by thick Tertiary and Cretaceous sequences of mudstones. Since our primary goal is to model acoustic velocities in shale, we used wireline data from the Nordland (NL), Rogaland (RL) and Hordaland (HL) groups that are mainly composed of mudstone. The mudstone beds were identified from the well log data by using natural gamma log, sonic and resistivity data as well as information from the Norwegian Petroleum Directorate website. The volume of shale was estimated from the natural gamma log and the data points with values higher than 0.65 were discarded. Similarly, the data points with shale volume less than 40 % were discarded. Furthermore, exceptionally high resistivity values amounting to more than 6 Ohm m were not included in the analysis. Porosities were determined from the density log assuming 2.7 and 1.05 g/cm3 for matrix and fluid densities, respectively. The depth that was used for calculating Vp from Faust (1953) was the depth below mudline in order to the compare of well data from wells that have been drilled at different depths (Hubert 2008).