Stochastic simulation of karstic systems

Yves Frantz. ( 2021 )
Université de Lorraine

Abstract

Karstic systems are geological structures that strongly impact underground flows. Despite intensive explorations by speleologists, they remain partially described as many conduits are not accessible to humans. Paleokarsts are buried karstic systems with a significant reservoir potential. But they are not easily identifiable on seismic images, especially if the affected formations are located below a salt canopy as in pre-salt lacustrine carbonates offshore Brasil. In these contexts, a huge uncertainty subsists on the network location and the conduit geometry. Stochastic simulations help to better assess that uncertainty. The difficulty is to reproduce the system connectivity at different scales while integrating as much geological knowledge as possible without involving poorly constrained parameters (e.g. paleo-climate, boundary conditions...). Recent works of the RING team have focused on the simulation of channelized systems [Ruiu, 2015; Rongier, 2016; Parquer, 2018]. Karst conduits, like channels, have the elongated connected shapes that drive the system connectivity. But contrary to channelized systems, karst can also have a non-negligible vertical dimension (some systems extends their galleries over more than 5 km depth). Recent statistical works have also enhanced the diversity of encountered geometries and it appears quite difficult to dress one typical schema of organization [Collon et al., 2017]. Their architecture is in fact influenced by several parameters: existence of tectonic or stratigraphic inception surfaces, speleogenesis processes (hypogenic or epigenic origine, coastal location), recharge, time evolution... The work of this PhD will aim at developing new methodologies to simulate karstic systems with a focus on the respect of available data, in particular connectivity data and geological knowledge (e.g., inception features). It will focus in priority on the large scale of the karstic networks. At this scale, connectivity data are provided by tracer tests (or well/production test for buried systems). Complementary to the identification of existing connections, such tests also provide travel times. Interpretations of the signal even give some approximation of the average section of conduits [Dewaide et al., 2016]. Recently developed Lidar and aerial photography techniques have also demonstrated their potentiality to identify surface evidences of the karst presence [Alexander et al., 2013; Weishampel et al., 2011; Zhu et al., 2014]. For paleokarst systems, seismic can provide equivalent information [James et al., 2011; Zeng et al., 2011; Yang et al. 2012]. Thus, the simulator would aim at honoring both these hard and soft data. The local scale of conduits could also be considered if time and data allow it. At this scale, the reconstruction of volumic conduits has already been explored trough the development of a dedicated version of the ODSIM method [Henrion et al., 2010] for karstic conduits: e-ODSIM [Rongier et al., 2014]. Integrating regularly sampled sections, like it is now possible with optical laser device [Schiller and Renard, 2016], still remains a challenge that would be interesting to explore with gradual deformation techniques or NURBS object deformation.

Download / Links

BibTeX Reference

@PHDTHESIS{frantz_using_2021,
    author = { Frantz, Yves },
     title = { Stochastic simulation of karstic systems },
      year = { 2021 },
    school = { Université de Lorraine },
       url = { https://theses.fr/s285937 },
  abstract = { Karstic systems are geological structures that strongly impact underground flows. Despite intensive explorations by speleologists, they remain partially described as many conduits are not accessible to humans. Paleokarsts are buried karstic systems with a significant reservoir potential. But they are not easily identifiable on seismic images, especially if the affected formations are located below a salt canopy as in pre-salt lacustrine carbonates offshore Brasil. In these contexts, a huge uncertainty subsists on the network location and the conduit geometry. Stochastic simulations help to better assess that uncertainty. The difficulty is to reproduce the system connectivity at different scales while integrating as much geological knowledge as possible without involving poorly constrained parameters (e.g. paleo-climate, boundary conditions...). Recent works of the RING team have focused on the simulation of channelized systems [Ruiu, 2015; Rongier, 2016; Parquer, 2018]. Karst conduits, like channels, have the elongated connected shapes that drive the system connectivity. But contrary to channelized systems, karst can also have a non-negligible vertical dimension (some systems extends their galleries over more than 5 km depth). Recent statistical works have also enhanced the diversity of encountered geometries and it appears quite difficult to dress one typical schema of organization [Collon et al., 2017]. Their architecture is in fact influenced by several parameters: existence of tectonic or stratigraphic inception surfaces, speleogenesis processes (hypogenic or epigenic origine, coastal location), recharge, time evolution... The work of this PhD will aim at developing new methodologies to simulate karstic systems with a focus on the respect of available data, in particular connectivity data and geological knowledge (e.g., inception features). It will focus in priority on the large scale of the karstic networks. At this scale, connectivity data are provided by tracer tests (or well/production test for buried systems). Complementary to the identification of existing connections, such tests also provide travel times. Interpretations of the signal even give some approximation of the average section of conduits [Dewaide et al., 2016]. Recently developed Lidar and aerial photography techniques have also demonstrated their potentiality to identify surface evidences of the karst presence [Alexander et al., 2013; Weishampel et al., 2011; Zhu et al., 2014]. For paleokarst systems, seismic can provide equivalent information [James et al., 2011; Zeng et al., 2011; Yang et al. 2012]. Thus, the simulator would aim at honoring both these hard and soft data. The local scale of conduits could also be considered if time and data allow it. At this scale, the reconstruction of volumic conduits has already been explored trough the development of a dedicated version of the ODSIM method [Henrion et al., 2010] for karstic conduits: e-ODSIM [Rongier et al., 2014]. Integrating regularly sampled sections, like it is now possible with optical laser device [Schiller and Renard, 2016], still remains a challenge that would be interesting to explore with gradual deformation techniques or NURBS object deformation. }
}