Highlights of MRST 2021b

The new release comes with some improvements in functionality and minor bug fixes. In particular we would like to highlight the following:

• Network models. The network-models module has been introduced to add functionality for creating reduced simulation models that represent the reservoir as a network of 1D flow paths connecting wells. 
• Flexible code for adjoint-based computation of model parameters. The new ModelParameter class and accompanying utilities can be used to set up model tuning optimization problems which can be solved using MRST-internal or external optimizers. New examples include tuning of both coarsened models and network-type models

New modules

network-models

With the new network-models module you can create reduced simulation models that represent the reservoir as a network of 1D flow paths connecting wells. The resulting network model can subsequently be calibrated to match observed or simulated well responses using an adjoint-based or ensemble-based method. The interwell connections can be set up in different ways: all-to-all, injectors to producers, user prescribed, or identified by flow diagnostics pre- or postprocessing of an underlying simulation model. The functionality of the module is demonstrated on the Egg, Norne, and SAIGUP cases. 

ad-micp

The ad-micp module has been contributed by David Landa-Marbán and co-authors at NORCE. It extends the solvers found in ad-core and ad-blackoil with capability for simulating microbially induced calcite precipitation (MICP) applications. MICP is a new and sustainable technology which utilizes biochemical processes to create barriers by calcium carbonate cementation; therefore, this technology has a potential to be used for sealing leakage zones in geological formations. The main mechanisms in the conceptual model are as follow: microbial solution is injected which results in microbes attaching themselves to the pore walls to form biofilm, growth solution is injected to stimulate the development of biofilm, cementation solution is injected which is used by the biofilm for production of calcite, and the calcite reduce the pore space which in turn decreases the rock permeability. 
 
Further information on the mathematical model can be found in the two following publications: 
 
[A] Landa-Marbán, D., Tveit, S., Kumar, K., Gasda, S.E., 2021. Practical approaches to study microbially induced calcite precipitation at the field scale. Int. J. Greenh. Gas Control 106, 103256. https://doi.org/10.1016/j.ijggc.2021.103256. 
 
[B] Landa-Marbán, D., Kumar, K., Tveit, S., Gasda, S.E. Numerical studies of CO₂ leakage remediation by micp-based plugging technology. In: Røkke, N.A. and Knuutila, H.K. (Eds) Short Papers from the 11th International Trondheim CCS conference, ISBN: 978-82-536-1714-5, 284-290. 
 
For additional scripts to run the numerical studies in the aforementioned publications and information about compatibility with GNU Octave, visit: 
 
https://github.com/daavid00/ad-micp 

Changes to existing modules

ad-blackoil

• Setup functions for Egg, SPE1 and SPE9 have been moved into the example-suite module. You may need to add the example-suite module to your code if you have used setup functions for these models. 

optimization

• Several improvements have been made to the MRST-optimizer unitBoxBFGS for increased robustness and convergence. 

diagnostics

• Setup functions for the Egg model have been moved into the example-suite module. You may need to add the example-suite module to your code if you have used setup functions for the Egg model.