Abstract:
We introduce a novel application of the Markov Chain Monte Carlo (MCMC) method to address the problem of characterising uncertainties in modelled thermal histories. As real data are never perfect, uncertainties in measurement will be propagated through to final estimates of hydrocarbon volumes, types and locations, and therefore it is important to characterise the nature of the uncertainty at each stage of the modelling process. Here we specifically consider heat flow parameters used to generate thermal history models and illustrate the use of the MCMC method on both synthetic and real well data. We show that this technique is able to both recover a good data-fitting model and, perhaps more importantly, to characterise the uncertainties on every heat flow model parameter. The approach described here is generally applicable for characterising the parameter uncertainties in various other aspects of basin modelling.