Kinetic parameters for specific source rocks provide the explorationist a way to obtain a more accurate assessment of the timing and quantity of hydrocarbon generation. This is attributed to the fact that organic matter decomposes into hydrocarbons at a rate that is dependent on its inherent chemical composition and structure. Thus, kerogen types I, II, II-S, II/III, and III decompose at different rates. The utility of this approach has been shown in numerous studies where the kinetic behavior of individual kerogen is documented by observing the shape and position of the pyrolysis S2 peak at different laboratory heating rates. Decomposition curves captured during the kerogen pyrolysis at multiple heating rates allow calculation of the activation energy (E) and the rate constant (A) that are required for the Arrhenius equation. This type of data has been proven to be quite successful in solving specific exploration problems. The essence of this approach is to provide data to describe the generation of hydrocarbons that is specific for a particular source rock or organic facies, instead of reliance on a “generic” oil window.

Numerous mathematical models are available for calculating kinetic parameters. The most important for source rock kinetic studies are the Discrete, Gaussian, and 3-Parameter Narrow Profile models. In the discrete model, the shift in peak pyrolysis temperatures from multiple non-isothermal experiments is used to determine the initial A factor range. Linear and non-linear regression is then used to determine the optimum A factor and fractions characterized by each E value. The selection of the E spacing is critical in that errors can be introduced into the kinetic calculation if the E spacing is too large; a value of 1 kCal mol^-1 was used in this study. The resulting kinetic calculations provide a single A value, multiple E values (if present), and the percentage of each E value that describes a particular decomposition rate. In contrast, the Gaussian model presumes a normal distribution of activation energies about a principal value. The initial A and E values are taken from the approximate calculation and the best fitting curves are obtained by nested linear regression until the calculated error function is the lowest possible value. The calculated values include an A and principal E value, as well as a Gaussian distribution parameter assuming a first order reaction. The 3-Parameter Narrow Profile model is used to describe organic matter with 100% decomposition at a single activation energy over a narrow temperature window. As this final calculation method is not applicable in this case study, it will not be further discussed.

Laboratory analysis of source rocks that charge the Maikop/Diatom-Productive Series (!) Petroleum System (Wavrek et al., 1998) display a significant range of E and A values. The discrete model results obtained in this study display a range of E values between 48 to 60 kCal mol^-1 with sample mediums around 53 kCal mol^-1; these results are comparable to those reported by Abrams and Narimanov (1997). However, it is noted that the discrete kinetic model is not suitable for all kerogen types; particularly those that are characterized by a relatively uniform organic matter composition. Given the documentation that the oils of the basin are a strikingly similar, it follows that the effective source rock facies responsible for the oil charge is fairly uniform in composition. This latter point is confirmed in the optical and chemical analytical results of the effective source rock facies and that the strikingly similar oils are generated over a relatively narrow range of thermal stress (Wavrek et al., 1996, 1998; Abrams and Narimanov, 1997). It is our experience that the Gaussian model typically provides results that suggest a slightly faster rate of organic matter decomposition (compared to the Discrete model) for the relatively uniform kerogen, and that this generally provides a better match to independent measurements within the petroleum system. Indeed, this experience is confirmed in this study where the Gaussian model calculation of the kerogen decomposition results provides an E value of 49.8 kCal mol^-1 and corresponding A value of 7.50 x 10¹² sec^-1. This is in excellent agreement with the results derived from an inverse calculation based on crude oil thermal stress analysis (Wavrek et al., 1996).

The most important issues addressed in this case study of the Maikop/Diatom-Productive Series (!) Petroleum System include: selection of the appropriate calculation model to describe the kerogen decomposition results (forward modeling), and the use of the inverse modeling methods (based on crude oil thermal stress analysis) as a guide when multiple kerogen decomposition experiments do not provide converging results. These diverse approaches are essential to select basin modeling calculation variables that adequately describe the specific petroleum system.

AAPG Search and Discovery Article #90937©1998 AAPG Annual Convention and Exhibition, Salt Lake City, Utah