This essay is an inquiry into the formulation of models in the science of petroleum and natural gas engineering. The underlying questions of this essay how adequate some of the fundamental models of this science really are, and what assumptions have been made as the models were created. Our claim is that a good account of the adequacy of models must be strongly pragmatist, for these questions cannot be answered properly without strict attention to human purposes. These purposes include not only the search for a better understanding of geological formations, their natural history and structure, but also classroom instruction for university students, and economically feasible extraction of petroleum and natural gas. These models include machines as well as natural formations, and so too raise the interesting question of how we (pragmatically) model machines. We claim that many of the distortions and over-simplifications in these models are in fact intentional and useful, when we examine the models in light of their pragmatic aims.