Simple kinetic models of petroleum formation. Part II: oil-gas cracking

Abstract

Listen

Translate article

Kinetic schemes describing oil-gas cracking are a required component of models which predict the masses, compositions and phases of petroleum expelled from source rocks; an additional application is in predicting the temperature zone over which reservoired oil deposits will be degraded to gas. This simple first-order kinetic model, calibrated on closed-system laboratory pyrolysis of 16 different source rock samples, describes the bulk conversion of oil (C 6+ molecular range) to gas (C 1–5 molecular range). Cracking rates vary markedly between samples, depending on the saturate to aromatic ratio of the generated oil. The initial hydrogen index (Hl 0) of the sample, taken as an indirect measure of the saturate to aromatic ratio of the oil it generates, correlates strongly with optimized rate constants for oil cracking in individual samples. Thus cracking rates are predictable using a simple, routinely performed geochemical screening measurement. Source rocks with high Hl 0 tend to generate saturate-rich oils with a high mean activation energy and a tight distribution of bond energies, which crack relatively slowly over a high but relatively narrow temperature range. Source rocks with low Hl 0 generate aromatic-rich oils with a low mean activation energy and a broad distribution of bond energies, which crack relatively rapidly over a lower, but relatively wide, temperature range. This is an important contributing factor in the observed gas-proneness of ‘type III’ source rocks with low Hl 0. The projected cracking window (defined limits at 10–90% of initial oil degraded; reference heating rate 2°C Ma −1) varies from 155 to 205°C in very high quality source rocks (Hl 0 > 600 mg g C −1) to 115–205°C in very poor quality source rocks (Hl 0 50 mg g C −1). An order of magnitude increase (decrease) in heating rate elevates (depresses) these temperature windows by ca. 15°C. The extrapolations to the geological subsurface are subject to a confidence limit no better than 7°C. Extrapolation to the reservoir environment requires caution because kerogen, an important potential catalyst and hydrogen donor allowing 100% cracking efficiency in the source rock, is absent. Furthermore, fractionation during expulsion causes saturate enrichment of expelled, ultimately reservoired oils. Thus in-reservoir cracking should be at least as slow as cracking in good quality source rocks, and coke or pyrobitumen formation is needed to conserve the hydrogen balance. A case history based on deep petroleum pools of the UK Central North Sea confirms these projections, demonstrating no positive evidence of significant in-reservoir cracking at temperatures as least as high as 174°C, and perhaps even 195°C. Previous studies have overestimated the importance of in-reservoir cracking. Future studies aiming to predict petroleum composition in traps should concentrate on understanding the charge composition expelled from the kitchen.