methane leak and plume migration in subsurface from hydrocarbon wellbores
I would like to know what module would be best to simulate methane leakage and plume migration in the subsurface from hydrocarbon wellbores? I did some digging and came across some publications where the authors had used EOS7C for this purpose. However, I'd like to confirm that. Any suggestions would be greatly appreciated.
Also, does anybody know if I could incorporate isotherms/absorptions onto the solid and organic matter in that potential module? Does EOS7C has that capability? I do not know how significant methane sorption onto clay mineral would be specifically in a relatively shallow wellbore and its migration path to the ground surface but I need to include that.
The user's guide of EOS7C V.1.0 does not mention the possibility to simulate gas adsorption. Other TOUGH2 modules or codes simulating gas adsorption are:
- EOS7C-ECBM : Webb S.W. (2011). EOS7C-ECBM Version 1.0: Additions for Enhanced Coal Bed Methane Including the Dusty Gas Model. Canyon Ridge Consulting Report CRC2011-0002.
- RealGasBrine (TOUGH+): Moridis G.J., Freeman C.M. (2014). User’s manual for the RealGasBrine v1.0 option of TOUGH+ v1.5: a code for the simulation of system behavior in gas-bearing geologic media. LBNL-6870E, Lawrence Berkeley National Laboratory, Berkeley, CA 94720.
- an EOS developed for TOUGH2 by Zarrouk S.J. (2004). Simulation of complex multiphase, multi-component, reacting flows in porous media. PhD thesis in Engineering Science, The University of Auckland, New Zealand.
To study the leakage of methane from a gas storage reservoir towards the surface, with accumulation in shallow confined groundwater aquifers, I used TMGAS (Battistelli and Marcolini, 2009; an improved version of TMVOC for real gas mixtures), that at the time included simple adsorption isotherms relating the adsorbed gas mass to the concentration of gas dissolved in the aqueous phase.
Now TMGAS uses gas adsorption isotherms with an approach similar to that followed by Zarrouk and Webb. It has been developed to simulate the expulsion of gaseous hydrocarbons and inorganic gases from source rocks and their subsequent migration at basin scale.
Probably EOS7C-ECBM could be the code version suitable for your needs.
TMVOC V.1.0 (Pruess and Battistelli, 2002) can model a mixture of water and CH4, but CH4 has to be included as an NCG, not a VOC because its critical temperature is lower than subsurface T at which CH4 migration should be simulated. And adsorption, simulated as a simple linear adsorption isotherm on organic matter, in TMVOC V.1.0 is limited to VOC and is not available for NCG.
In addition, TMVOC V.1.0 was developed for shallow subsurface conditions such those found in contaminated sites. Thus, the thermodynamic equilibrium is suitable for shallow P&T conditions, not those found in deep natural gas reservoirs.
The adsorption of NCGs, a linear adsorption isotherm on organic matter, is simulated in the TMVOC version supported by TOUGH3 which actually implements the TMVOCBio code version supporting the simulation of biodegradation reactions (Battistelli, 2004 VZJ; Jung and Battistelli, 2016). If biodegradation is not invoked, TOUGH3-TMVOC should basically work as the TOUGH2-TMVOC version, with the same limitation about P&T conditions.
You might check the accuracy of methane properties calculation in TOUGH3-TMGAS at the P&T range of your application and verify if the accuracy is acceptable or not.
A version of TMVOC implementing correlations for high P&T properties of VOC and NCG (but still using Henry's law for gas solubility in the aqueous phase) was developed by importing the GasEOS routines (basically those used by EOS7C) int TMVOC (provisionally called V.2). Examples of code application are the papers by Battistelli et al. (2003) "Modeling gas reservoir properties with TMVOC V.2.0" and Battistelli (2008, VZJ). TMVOC V.2.0 incorporates the Peng-Robinson (PR) EOS with binary interaction coefficients and a correlation to compute more accurate viscosities for gas mixtures at high pressure. It can also simulate the presence of dissolved solids, including NaCl whose concentration can affect brine thermophysical properties, and generalized adsorption isotherms for all the mass components.
TMVOC V.2.0 became TMGAS (Battistelli and Marcolini, 2009), but was also further improved at LBNL to model the injection of gas mixtures at high pressures (Zhang et al., 2011, Modeling of fate and transport of coinjection of H2S with CO2 in deep saline formations).
To conclude: I would take EOS7C and I would add CH4 adsorption with a Langmuir isotherm using the partial pressure of methane as shown by Zarrouk (2004).