Identifying Fluid Type and Contacts in Carbonate Reservoirs

Abstract Present-day fluid type and contacts in carbonate reservoirs can be difficult to determine from standard formation evaluation techniques because of complex rock properties and variable fluid compositions. In such situations, integrating novel rock-based geochemical analyses of adsorbed and inclusion-trapped fluids helps reduce fluid contact uncertainty and evaluate the probability of various fluid types. The rock-based analyses include three techniques that can be applied to either core or cuttings samples. First, volatile compounds adsorbed or trapped in pore spaces are measured by mass spectrometry using a patented pumpdown volatiles (PDV) technique. Second, fluid inclusion volatiles (FIV) analysis also uses mass spectrometry to characterize volatile compounds released from fluid inclusions when samples are crushed. Both analyses are rapid and inexpensive and therefore are frequently applied to entire wells to allow stratigraphic correlation of responses for mapping fluid types and contacts. However, because FIV signatures include both present and paleo fluids, additional analyses are needed when filling history is complicated (e.g., gas displaces oil). For example, PDV and FIV interpretations can be confirmed and refined with a third technique, thermal desorption gas chromatography/mass spectrometry, Iatroscan, and/or Rock Eval pyrolysis. In addition to these analytical techniques, a statistical modeling tool has been developed for quantitative probability predictions of reservoir fluid type from complex FIV and petrophysical signatures. The model is constructed by calibrating FIV and petrophysical data to known test results, and then applying it to predict fluid type in wells where test results are absent or ambiguous. Besides providing an integrated approach to fluid type and contact evaluation, this tool allows multiple scenarios and quantification of uncertainty. This paper summarizes methodologies and key applications of rock-based techniques for accurate resource evaluation, improved completion decisions, and optimized exploration, development, and production strategies in carbonate reservoirs. Introduction Using traditional well logs or seismic to identify present-day fluid type and contacts (FT&C) in carbonate reservoirs can be challenging due to complex and heterogeneous reservoir lithology and properties, aqueous pore fluid composition, and low-resolution seismic data. In petroleum reservoirs, rocks may adsorb small quantities of the surrounding fluid medium or the fluids may be trapped within mineral cements in the form of fluid inclusions. These fluids are often detectable using techniques based on mass spectrometry, including analyses of ExxonMobil patented pumpdown volatiles (PDV) (1) and fluid inclusion volatiles (FIV) (2), often combined with thermal desorption gas chromatography/mass spectrometry (TD-GC/MS). These rock-based techniqes can be used to identify FT&C and are most successful using closely spaced conventional or sidewall core samples, but cuttings have also been used successfully. The techniques are not applicable for samples drilled with oil-based mud and some drilling additives in water-based mud systems may also complicate interpretations. The data are typically interpreted qualitatively based on the chemical results concerning the location of hydrocarbon migration pathways, seals, and fluid type. However, statistical analysis and modeling can be used to give quantitative probabilities of fluid type (e.g., gas or oil) by calibrating geochemical responses against known test results and applying to other wells where test results were absent or ambiguous. In this paper, we summarize these methodologies and show examples of how we integrate these rock-based techniques with petrophysical evaluation and statistical approaches to determine FT&C. The results reduce uncertainty regarding the hydrocarbon phases and fluid contacts present in the carbonate reservoirs.