The evaluation of limestone formations by well logs has brought many interesting problems to light, some of which are as yet unresolved. While a major purpose of well logging is to ensure that zones containing producible hydrocarbons are not bypassed, this paper is more concerned with the further information that can be derived from the logs to assist in more clearly defining reservoir characteristics, fluid content and performance. In the hard limestones of Southeast Arabia sufficient evidence has not yet been gathered to allow any large-scale tailoring of the logging program; a number of logs are still run which ostensibly can be used to compute the same parameter. The validity of some of these measurements is discussed, and some tentative interpretations are made concerning their significance. In this paper, the five selected topics are discussed in light of their application to the hard limestone reservoirs under review.
Well-log analysis lends itself readily to the determination of water saturation and, hence, to computation of hydrocarbon reserves. However, it is advisable to obtain confirmation of the values so obtained by alternative methods because several factors influence the basic data from which these values are derived. As has been pointed out elsewhere, water saturations of 2 to 50 per cent of the available pore space have been observed in oil-producing formations. This wide range invalidates the possibility of choosing a reasonable value by inspecting the available choices and, thus, makes corroboration from other sources essential.
The methods used to determine connate-water saturation in the formations under review were core analysis of formations cut under oil-base mud, measurement of capillary pressures of cores and log analysis. Each of these methods suffers from limitations, and only where mutual agreement exists may a value be accepted with confidence. Since the log-analysis method offers distinct advantages over the other two, affording as it does a measurement under reservoir conditions, it was of interest to attempt to resolve the uncertainties inherent in the method by comparing results obtained on cores in the laboratory.