Formatted Title
The Challenges with Amendment Delivery in Fractured Rock: Using Discrete Fracture Matrix Simulation to Quantify Reach
Background/Objectives
Contaminant remediation in fractured porous rock is difficult due to complexities of fracture network connectivity and aperture variability that create strong heterogeneities with respect to flow and solute transport processes. Common remediation methods rely on the ability to deliver some type of amendment to enhance existing biological activity or deliver reactive solids or solutes to sequester or degrade contaminants in situ. While fractures can act as amendment delivery pathways, there is no guarantee that the amendment will reach its intended extent due to channelized flow. Furthermore, there is a dichotomy that exists within fractured porous rock: the most accessible parts are the fractures (with apertures that are typically 10s to a few 100 microns) which contain only a fraction of the contaminant mass, whereas the majority of the mass resides within the low permeability rock matrix blocks between the fractures (only accessible through diffusion). Numerical modeling and/or pilot studies may be implemented to estimate the effectiveness of a remediation method at fractured porous rock field sites. This study posits that discrete fracture matrix (DFM) modelling is a valuable supporting tool for visualization and quantitative forecasts of remedial actions that require an injection of an amendment. DFM models explicitly represent the flow and transport processes relevant at the fracture-matrix interface, as well as the variability in the network itself using field data on fracture aperture, frequency, and length.
Approach/Activities
A series of 3-D DFM simulations were created to illustrate the use of DFM modelling in quantifying the reach of an amendment in a targeted treatment zone. Hydrogeosphere (HGS) was used to model five daily injections of a bromide tracer into an 8 m well screen centered within a 20 x 20 x 10 m fractured sandstone domain, representative of a research site with matrix and fracture properties informed by field and lab measurements. Solute reach (i.e., volume invaded) and losses out of the target zone were tracked during the forced injection episodes and thereafter for two months. Fifty-one statistically equivalent 3-D realizations were created to produce ensemble values of tracer reach, given the site-specific fracture frequency data. For all realizations, a series of observation wells was placed radially away from the injection well to evaluate distribution/delivery of the bromide tracer.
Results/Lessons Learned
Across the 51 realizations, an average of 63% of the bromide mass exited the target zone (domain boundary) during the injection period. Analysis of the bromide mass remaining within the target domains revealed that 70% of the fracture volume experienced some concentration of the injected tracer above 10-3 [C/C0]. However, at the end of the two-month observation period, only 10% of the matrix pore volume received the solute at relevant concentrations. Furthermore, delivery of the bromide to the observation wells depended on a direct connection to the injection well via the fracture network, which became unlikely even at 7.5 m away from the injection well.
This work illustrates many challenges associated with amendment delivery. The spatial distribution of the amendment is highly dependent on the fracture network pathways’ connection to the injection well. Much of the injected mass exits the target zone through these pathways. Finally, the mass remaining within the target volume is delivered superficially relative to the contaminant location, deeper within the rock matrix blocks, thus requiring slow diffusive processes to complete the delivery task.
Additional scenarios are being explored where the injection well takes different positions within a single 30 x 30 x 10 m DFM domain realization, examining the potential sensitivity of reach to the fractures intercepting specific injection locations. Analysis within this larger domain could illuminate the radial influence of an injection well given certain injection pressures. Understanding the reach of a delivered solute could aid in optimizing remediation designs for contaminated fractured rock. Paired with plume scale and reactive transport modelling, the spatial efficacy of a remediation method can be estimated. Future work should incorporate undulating apertures, further complicating the pattern of solute delivery, to better represent the realities of remediation in fractured porous rock.