Integrating structural geological data into the inverse modelling framework of iTOUGH2
J. Florian Wellmann a,n, Stefan Finsterle b, Adrian Croucher c a CSIRO Earth Science and Resource Engineering, 26 Dick Perry Ave., Kensington, 6151 WA, Australia b Earth Sciences Division, Lawrence Berkeley National Laboratory, University of California, Berkeley, USA c Department of Engineering Science, University of Auckland, Private Bag 92019, Auckland, New Zealand a r t i c l e i n f o
Received 11 March 2013
Received in revised form 13 September 2013
Accepted 22 October 2013
Available online 31 October 2013
Implicit geological modelling
Multiphase flow simulation
Monte Carlo a b s t r a c t
The validity of subsurface flow simulations strongly depends on the accuracy of relevant rock property values and their distribution in space. In realistic simulations, this spatial distribution is based on two geological considerations: (1) the subsurface structural setting, and (2) smaller-scale heterogeneity within a hydrostratigraphic unit. Both aspects are subject to uncertainty, whereas techniques to address heterogeneity are well established, no general method exists to evaluate the influence of structural uncertainties. We present a method to include structural geological data (e.g. observations of geological contacts and faults) directly into an inversion framework, with the aim of enabling the inversion routine to adapt a full 3-D geological model with a set of geological parameters. In order to achieve this aim, we use a set of Python modules to combine several pre-existing codes into one workflow, to facilitate the consideration of a structural model in the typical model evaluation steps of sensitivity analysis, parameter estimation, and uncertainty propagation analysis. In a synthetic study, we then test the application of these three steps to analyse CO2 injection into an anticlinal structure with the potential of leakage through a fault zone. We consider several parts of the structural setting as uncertain, most importantly the position of the fault zone. We then evaluate (1) how sensitive CO2 arriving in several observation wells would be with respect to the geological parameters, (2) if it would be possible to determine the leak location from observations in shallow wells, and (3) how parametric uncertainty affects the expected CO2 leakage amount. In all these cases, our main focus is to consider the influence of the primary geological data on model outputs. We demonstrate that the integration of structural data into the iTOUGH2 framework enables the inversion routines to adapt the geological model, i.e. to re-generate the entire structural model based on changes in several sensitive geological parameters. Our workflow is a step towards a combined analysis of uncertainties not only in local heterogeneities but in the structural setting as well, for a more complete integration of geological knowledge into conceptual and numerical models. & 2013 Elsevier Ltd. All rights reserved. 1. Introduction
Structural geological models are commonly used to incorporate information about major geological units and their rock properties into flow simulations. It is well known that these geological models contain uncertainties (e.g. Mann, 1993; Bárdossy and Fodor, 2001; Thore et al., 2002; Turner, 2006; Suzuki et al., 2008; Caumon, 2010; Wellmann et al., 2010; Caers, 2011; Cherpeau et al., 2012; Lindsay et al., 2012) and it is reasonable to assume that simulated flow fields are sensitive to changes in the structural geological model.
We propose a framework to test sensitivities of simulated flow fields with respect to structural parameters derived from geological data, and to use observed flow field responses to invert for these structural parameters. We establish an automated link between structural geological modelling (using an implicit geological modelling method) and multi-phase flow simulations (using the general-purpose flow simulator TOUGH2). TOUGH2 is used for a wide range of applications, from hydrogeological studies and contaminant transport, to carbon sequestration, geothermal reservoir engineering and nuclear waste disposal (Pruess et al., 2011).
The link to TOUGH2 is computationally enabled via PyTOUGH, a set of Python modules offering pre- and postprocessing routines for TOUGH2 simulations (Wellmann et al., 2011). Our forward workflow from structural data to flow simulations is then integrated into an inverse modelling framework, iTOUGH2 (Finsterle, 1999), to use these data as parameters in inversions as well as sensitivity and uncertainty analyses.
The evaluation of sensitivities of simulation results to input parameters is often performed using a manual procedure, for example by testing the influence of minimal and maximal parameter values. Although this procedure can provide insights into the model behaviour, the overall informational value of the analysis is restricted (e.g. Carrera et al., 2005). A systematic analysis of sensitivities based
Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/cageo
Computers & Geosciences 0098-3004/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.cageo.2013.10.014 n Corresponding author. Tel.: þ61 8 6436 8826.
E-mail addresses: firstname.lastname@example.org, email@example.com (J.F. Wellmann).
Computers & Geosciences 65 (2014) 95–109 on mathematical principles delivers (in addition to quantitative sensitivity measures) a detailed error evaluation, including information on parameter covariances and correlations (Refsgaard, 1997; Sun and Sun, 2006). For flow simulations with TOUGH2, this functionality is implemented in iTOUGH2. In addition to sensitivity analysis, iTOUGH2 provides methods for parameter estimation (or inversion) from observed data and for uncertainty propagation analyses (e.g.