a b s t r a c tDuring studies of unconventional natural gas reservoirs of Silurian and Ordovician age in the northern Appalachian basin we observed complete reversal of the normal trend of carbon isotopic composition, such that d 13 C methane (C 1 ) >d 13C ethane (C 2 ) >d 13 C propane (C 3 ). In addition, we have observed isotopic reversals in the d 2 H in the deepest samples. Isotopic reversals cannot be explained by current models of hydrocarbon gas generation. Previous observations of partial isotopic reversals have been explained by mixing between gases from different sources and thermal maturities. We have constructed a model which, in addition to mixing, requires Rayleigh fractionation of C 2 and C 3 to cause enrichment in 13 C and create reversals. In the deepest samples, the normal trend of increasing enrichment of 13 C and 2 H in methane with increasing depth reverses and 2 H becomes depleted as 13 C becomes enriched. We propose that the reactions that drive Rayleigh fractionation of C 2 and C 3 involve redox reactions with transition metals and water at late stages of catagenesis at temperatures on the order of 250-300°C. Published ab initio calculated fractionation factors for C-C bond breaking in ethane at these temperatures are consistent with our observations. The reversed trend in d 2 H in methane appears to be caused by isotopic exchange with formation water at the same temperatures. Our interpretation that Rayleigh fractionation during redox reactions is causing isotopic reversals has important implications for natural gas resources in deeply buried sedimentary basins.Published by Elsevier Ltd.

No abstract

No abstract

is a senior geologic scientist with the Pennsylvania Geological Survey, where he has worked since 1980. He worked as a geophysical analyst for the Western Geophysical [Company] in Houston, Texas, before taking his present position in Pittsburgh, Pennsylvania. He also teaches sandstone petrology at the University of Pittsburgh. Laughrey's research interests include isotope and organic geochemistry, carbonate and clastic petrology, and borehole geophysics.

Using examples from the Permian Basin of Texas, the North Slope of Alaska, and the Bergan Field of Kuwait, this paper describes how oil geochemical fingerprinting can be applied to diagnose quickly and easily three production problems that may affect highly deviated wells. High-Resolution Gas Chromatography can be used to quantify ~1,000 different compounds in an oil, and the relative abundances of those compounds form a geochemical fingerprint. Geochemical differences between fluids in adjacent reservoirs can serve as natural tracers for fluid origin, allowing changes in production in highly deviated wells to be understood. Application 1: In wells that are fracture stimulated, oil fingerprinting can be used to assess whether induced fractures have propagated out of the target interval and into overlying or underlying formations. Oil fingerprinting can be used to quantify what percentage of the produced oil and gas is coming from each interval and how the effective stimulated rock volume changes through time. This concept is illustrated here with a Permian Basin example. Application 2: In wells with multiple laterals in the same well (such as those in certain North Slope, Alaska fields), sand can settle out of the production stream and form sand bridges that obstruct production from one or more of the laterals. In addition, sand co-produced with oil from shallower laterals can settle at the bottom of the vertical section during regular production and obstruct the entry to a deeper lateral. Geochemical fingerprinting can be used to determine quantitatively the contribution of each of several zones to a commingled oil stream. This technique allows the operator to identify sanded-out intervals for fill cleanout (FCO). Application 3: If two reservoirs are both oil bearing, but are of very different permeability, horizontal wells with an intended landing target in the tighter reservoir may be adversely affected if the well path contacts the more permeable reservoir. The Mauddud reservoir in Kuwait provides examples of this phenomenon. The Mauddud carbonate occurs between two massive clastic reservoirs, the Wara and the Burgan. Average Mauddud porosity is 18% with low permeability (1-10 mD), characteristics which make this reservoir a candidate for horizontal drilling. However, some lateral wells in this carbonate may encounter the adjacent, more permeable reservoirs over a short portion of the well path. In such cases, production from the adjacent reservoir may account for virtually all of the well's production, even though the well was intended to be completed solely in the tighter reservoir. Oil fingerprinting can be used to identify wells affected by this problem. A common theme unifies these three applications: Geochemical differences between in-situ fluids in adjacent reservoirs can serve as natural tracers for fluid movement. However, these techniques have been under-applied as tools for optimization of production from highly deviated wells. This paper illustrates the application of this technology to that well type in a variety of play types.