D. F. Cannon and G. R. Coates
Log interpretation is a combination of art and science. Ideally, log measurements would uniquely describe the answer, but this is seldom the case because most answers are under-determined ( more unknowns than measurements). The art of interpretation and interpretation-models are used to fill this gap between the known and the unknown. When this is the case, log analysts use intuition based on past experience to provide the missing data. Art is evident in both the obvious form of “from my experience this means” and the less obvious form of “interpretation models”. Model constraints and measurement interdependency’s, such as localized shale relationships and invasion relationships, are examples of formalized art that are based on past wisdom.
Such wisdom frequently migrates over time from “the best guess we have today” to “rules” of log interpretation. When new measurements become available they often challenge such “rules”. This was the case when recently developed clastic-mineral-analysis techniques for core and logs were compared with traditional concepts. One reason for the conflict was that traditional techniques work with rock types not minerals (e.g. sand vs. quartz, clay vs. illite, smectite, kaolinite). The conversion from rock-based to mineral-based interpretations is necessary because it permits formation evaluations that were previously not possible and it brings log and core answers closer together.
Two typical procedures for evaluating shaly sands are “Logs-only” and “Core-calibrated”. Logs-only is most common and is very attractive since log information is often all that is available. In such cases analysts select parameters from log data using crossplots as in-situ calibrations for a predefined petrophysical model. Such approaches require the presence of, often hard to find, in-situ calibrations zones (zones of 100% of what is being calibrated). The lack of insitu calibrations is reflected in inconsistent answers between analysts who have had to extrapolate data to non-obvious calibration end points. Core-calibrated interpretations utilizing core derived properties, such as cation-exchange capacitance (CEC) for entry into a petrophysical model, are often as inadequate as the Log-only technique since the full benefit of this approach requires every formation of interest to be fully cored and analyzed, including the analysis of the shales.
A new mineral-based formation evaluation method has been developed that combines the best of both Logs-only and Core-calibrated approaches. It uses mineral knowledge gathered from core analysis of key formations (via XRD, FTIR, etc.) to select or confirm the appropriate mineral model for log interpretation. This core validated mineral model can then be used in similar zones from well to well.
This is not a force fit of logs to core in the historical sense of making logs yield a porosity or some other response equal to core. The only corrections applied to the log data are the appropriate environmental corrections. Core data is not required on every zone and the rocks need not stay constant. All that is required is that the model used contain all the major mineral components present in the formations being evaluated.