Hydraulic Fracturing: From Modeling and Simulation to Reconstruction and Characterization

The past two decades have witnessed an enormous expansion of hydraulic fracturing (HF) in the subsurface as a tool to stimulate hydrocarbon reservoirs and, in particular, gas-rich shales formations. To date, however, concerns persist as to whether hydraulic fractures could breach the impermeable barriers that isolate aquifers from hydrocarbon-bearing formations, thus jeopardizing the supply of fresh groundwater, and whether they could induce significant seismic events. Questions are also mounting as to whether the engineering process has been designed to optimize the recovery of hydrocarbons and geothermal energy. The lack of simulation tools to realistically predict the geometry and hydraulic conductivity of the created fractures, and of effective techniques to characterize such fractures from in-situ measurements, is fueling these concerns. 

Despite a long series of investigations on the subject since the 1950s, it is only over the last two decades that a coherent understanding of the critical mechanisms at play has emerged. Moreover, technical issues brought about by the strong nonlinearity of the governing equations, due in part to the fluid-solid coupling, as well as those by the moving boundary nature of the problem, have hindered the development of efficient computational algorithms to simulate: (i) the propagation of hydraulic fractures in complex media, and (ii) the transport of proppant particles that are added to the injected fluid to create a permeable pathway for hydrocarbon. 

The theory and applications of inverse problems (IP) have long made a silent imprint in science and technology as a critical tool in establishing the link between physical reality and its mathematical model. The need for considering the “inverse” side of the coin is especially pressing in the context of HF applications where mathematical and computational models are deployed to describe a complex multi-physics process, often evolving kilometers beneath the earth’s surface where only a limited amount of sensory data are available. To complicate the problem further, the data on the process are often multi-modal and may include diverse sets, such as (active or passive) seismic signals, inclinometer measurements, and information on the flow rate of the fracturing fluid. Among a diverse array of pressing inverse HF problems to be tackled are: (i) geometric reconstruction of an evolving fracture surface, (ii) quantifying the distribution of proppant therein, and (iii) tracking the movement of fluid through a fractured rock mass.

Despite the clear need, interaction between the HF and IP communities has been limited at best. In this spirit, the intent of this workshop is to initiate and foster the synergy between the two groups, with an emphasis on exposing the IP community to a new class of applications while highlighting the importance of system characterization among HF researchers.