We combine forward stratigraphic models with a suite of uncertainty quantification and stochastic model calibration algorithms for the characterization of sedimentary successions in large scale systems. The analysis focuses on the information value provided by a probabilistic approach in the modelling of large-scale sedimentary basins. Stratigraphic forward models (SFMs) require a large number of input parameters usually affected by uncertainty. Thus, model calibration requires considerable time both in terms of human and computational resources, an issue currently limiting the applications of SFMs. Our work tackles this issue through the combination of sensitivity analysis, model reduction techniques and machine learning-based optimization algorithms. We first employ a two-step parameter screening procedure to identify relevant parameters and their assumed probability distributions. After selecting a restricted set of important parameters these are calibrated against available information, i.e., the depth of interpreted stratigraphic surfaces. Because of the large costs associated with SFM simulations, probability distributions of model parameters and outputs are obtained through a data driven reduced complexity model. Our study demonstrates the numerical approaches by considering a portion of the Porcupine Basin, Ireland. Results of the analysis are postprocessed to assess (i) the uncertainty and practical identifiability of model parameters given a set of observations, (ii) spatial distribution of lithologies. We analyse here the occurrences of sand bodies pinching against the continental slope, these systems likely resulting from gravity driven processes in deep sea environment.

Damage in subsurface formations caused by mineral precipitation decreases the porosity and permeability, eventually reducing the production rate of wells in plants producing oil, gas or geothermal fluids. A possible solution to this problem consists in stopping the production followed by the injection of inhibiting species that slow down the precipitation process. In this work we model inhibitor injection and quantify the impact of a set of model parameters on the outputs of the system. The parameters investigated concern three key factors contributing to the success of the treatment: i) the inhibitor affinity, described by an adsorption Langmuir isotherm, ii) the concentration and time related to the injection and iii) the efficiency of the inhibitor in preventing mineral precipitation. Our simulations are set in a stochastic framework where these inputs are characterized in probabilistic terms. Forward simulations rely on a purpose-built code based on finite differences approximation of the reactive transport setup in radial coordinates. We explore the sensitivity diverse outputs, encompassing the well bottom pressure and space-time scales characterizing the transport of the inhibitor. We find that practically relevant output variables, such as inhibitor lifetime and well bottom pressure, display a diverse response to input uncertainties and display poor mutual dependence. Our results quantify the probability of treatment failure for diverse scenarios of inhibitor-rock affinity. We find that treatment optimization based on single outputs may lead to high failure probability when evaluated in a multi-objective framework. For instance, employing an inhibitor displaying an appropriate lifetime may fail in satisfying criteria set in terms of well-bottom pressure history or injected inhibitor mass.