Abstract Four-dimensional petroleum system models within the Western Canada sedimentary basin were constructed using hydrous pyrolysis (HP) and Rock-Eval (RE) kinetic parameters for six of the major oil-prone source rocks in the basin. These source rocks include the Devonian Duvernay Member of the Woodbend Group; Devonian-Mississippian Exshaw Formation; Triassic Doig Formation; Gordondale Member; Poker Chip A shale, both of the Jurassic Fernie Group; and Ostracod Zone of the Lower Cretaceous Mannville Group. The Mannville Group coals also contributed oil to the oil sands (Higley et al., 2009) but are excluded herein because HP kinetics were used for both models with identical results. The locations of oil migration flowpaths are identical for the HP and RE models, with the exception of an earlier onset of generation and migration shown with the HP model. Both models show that the oil sands are located at focal points of the petroleum migration pathways. The principal differences between the models are the onset and extent of oil generation from the Jurassic Fernie source rocks (Gordondale Member and Poker Chip A shale) at about 85 Ma with the HP model and 65 Ma with the RE model. Earlier oil generation in the HP model is caused by the high sulfur content of the type IIS kerogen in the Jurassic source rocks. The influence of organic sulfur is accounted for in the HP kinetic parameters, but not the RE kinetic parameters. The cumulative volume of oil generated from the source rocks is 678 billion m 3 for the HP model and 444 billion m 3 for the RE model, or 65% of the HP volume. This difference is attributed to early generation from type IIS kerogen that resulted in much larger volumes of thermally mature source rocks for the Jurassic Fernie Group and consequently larger volumes of generated oil. The Gordondale Member in the HP model generated more than 550 times the volume of oil generated by the Gordondale Member in the RE model. The timing and generated volumes are comparable in the RE and HP models for source rocks that contain normal levels of organic sulfur (type II kerogen). The Duvernay is an exception because of the very low sulfur content of its type II kerogen. The result is higher HP kinetic than RE kinetic parameters, with associated greater thermal maturities required for HP than for RE oil generation. Consequently, there is less mature Duvernay source rock in the HP model than the RE model.

Abstract The amount of oil a pod of active source rock is capable of expelling is an important parameter in determining the ultimate oil potential of a petroleum system. Laboratory pyrolysis methods available for determining this parameter can be grouped into three categories: (1) hydrous pyrolysis , (2) closed anhydrous pyrolysis , and (3) open anhydrous pyrolysis. The processes responsible for expulsion of oil in nature are best simulated by hydrous pyrolysis. The presence of water in both nature and hydrous pyrolysis ensures the occurrence of a water-saturated bitumen within a source rock. This dissolved water reduces cleavage of bitumen molecules by terminating free radicals with water-derived hydrogen and allows the cleavages that do occur to form an immiscible oil. The importance of water is best demonstrated by its absence in closed anhydrous pyrolysis, which generates a bitumen that ultimately decomposes into an insoluble pyrobitumen rather than an expelled oil. As a result, this method is not useful in evaluating oil expulsion. Conversely, a vaporized oil may be generated in open anhydrous pyrolysis, but the amount of oil generated is dependent on experimental conditions, and the processes responsible for this evolved oil are not operative in nature.