Culling criteria for “standard” formation water analyses

Abstract

Although many schemes have been proposed for sorting and classifying formation waters, none contains procedures for automatically culling erroneous analyses prior to computer processing or graphical display. Techniques are described for automatic (electronic) culling of formation water analyses on initial entry into a data base, based on chemical analysis and on the method of production and sampling. With respect to the chemical analysis, this can be rejected for a variety of reasons, including incomplete or poor data, and contamination by mud filtrate, acid wash or cement wash. With respect to the method of production and sampling, certain methods of production are excluded, as are samples from multiple drillstem tests or from drillstem tests where the recovered fluid is dominantly non-aqueous; culling is also based on samples subject to corrosion in separators and treaters. The culling criteria act sequentially and using the more than 141,000 formation water analyses from the entire Western Canada Sedimentary Basin as a test case, 68.8% of the analyses were rejected from further study. Following assignment to the pertinent stratigraphic unit, a second series of conditional culling criteria can be applied at the aquifer-specific and area-specific level, including rejection because of KCl mud contamination. Final culling (third series of criteria) is done manually by a geochemist using a variety of methods such as cumulative frequency plots of individual ions and regional concentration variation maps.