For most conventional reservoirs, matrix permeability determines the rock's ability to transmit fluids. It is a crucial input for the operator's completion design. Unfortunately, direct measurement of permeability is hard to obtain through well logging. Traditionally, a continuous permeability log is estimated from other petrophysical properties derived from well logs, such as porosity. The porosity-to-permeability relationship is typically established from core measurements. The drawback of this method is that porosity alone is insufficient to characterize the full variability in permeability. Also, the application of this method is limited by availability of core measurements representative of the reservoir in question.
In this report, we examine a siliciclastic oil reservoir in South China Sea with interbedded sandstone and shale. In such reservoirs, heterogeneity in permeability exists at a small scale. In order to capture the full variability of such reservoirs, it is critical to obtain a continuous permeability curve of the rock matrix.
It has been shown previously that permeability can be accurately computed from surface area-to-pore volume ratio (k?). This ratio can be obtained from lithology weight fractions by associating a specific surface area (S0) with each lithology. In general, the S0 values for clean sand and carbonate are well defined and stable. However, S0 for clay depends largely on clay type and varies from sample to sample. Since clay has the most significant effect on permeability, it is critical to select the correct k? for clay in the model.
In modern logging programs, formation pressure tests are commonly included for reservoir characterization. By observing pressure behavior versus time in a pretest, one can obtain direct measurement of fluid mobility at discrete stations along the well. This mobility can be converted to permeability if fluid viscosity is known. We can then use these measured values to calibrate clay S0 in our permeability model. Here is the workflow:
Zone the well such that the clay type is consistent within each zone.
Using a univariate optimization routine, find the clay S0 that minimizes the difference between k? and permeability from pressure tests.
Plug the calibrated S0 for clay into k? model to compute a continuous permeability log.
The above method was validated using a set of LWD logs acquired while drilling and LWD pressure tests acquired while reaming.