An Open-Source Code for Fluid Flow Simulations in Unconventional Fractured Reservoirs
Round 1
Reviewer 1 Report
With all respect I have not seen much improvements from the first submission, having said so since in new version the source codes made available I have tried all the examples but non of them seems working. the read me file was also brief without actual direction on how to execute the examples and get results as presented in the manuscript. therefore there was absolutely no way to explore the codes and run the examples to at least be able to regenerate the results provided in the manuscript. there have been several Errors and missing parts in the source code provided that makes it not practical. therefore my decision to reject the manuscript in current format. the source code must be provided in such a manner that all the examples are executable and results of the manuscript can be reproduced.
Author Response
Response to Reviewer 1 CommentsWith all respect I have not seen much improvements from the first submission, having said so since in new version the source codes made available I have tried all the examples but non of them seems working. the read me file was also brief without actual direction on how to execute the examples and get results as presented in the manuscript. therefore there was absolutely no way to explore the codes and run the examples to at least be able to regenerate the results provided in the manuscript. there have been several Errors and missing parts in the source code provided that makes it not practical. therefore my decision to reject the manuscript in current format. the source code must be provided in such a manner that all the examples are executable and results of the manuscript can be reproduced.
We are sorry for the confusion. I believe these problems are due to the improper installation of our module. Given that this code is an extension module based on MRST (https://www.sintef.no/Projectweb/MRST/), we assume the user should know the basics of MRST before using our module.
To help reader to run our code, we have added (in orange) more details about instillation and usage of our code in the \texttt{github} page (https://github.com/BinWang0213/MRST\_Shale/tree/ShOpen).
Reviewer 2 Report
That is an interesting paper in which an open source code for simulating unconventional reservoirs with fractures (either hydraulic fractures or natural) is presented. It is well organised and written. The link to the code repository is given.
Many physical processes (gas flow, Forcheimer, hydro mechanics etc...) processes are accounted for. Fractures are accounted for via the EDFM method.
The presentation of the codes is clear and concise, in the MRST matlab framework.
Note that all these physical processes imply laws involving many semi empirical coefficients, the accuracy of which may be discussed, for example fractionnal derivatives may be of interest when considering shale gas flo (Albineli et al, see below). These aspects can be put in relation with the large uncertainties encountered while considering fractured reservoirs. As uncertainty studies imply running many forward simulations, could the authors give some information about the performance of their code : running time, number of mesh elements and fractures that can be considered. Authors use an automatic differencation module, a very interesting feature to compute overall sensitivity coefficient, did they use it in that direction?
All the developments are for 2D cases, I think that 3D cases may be discussed, especially because 3D modelling of flow in fractured media are becoming more and more popular both for shale gas recovery applications, or in the geothermal context (Ngo et al, Fourno et al, Fumagalli and Keilegavlen see below) .
IN conclusion, appart these remarks, this is a good paper
Modeling of 1D Anomalous Diffusion in Fractured Nanoporous Media Ali Albinali, Ralf Holy, Hulya Sarak and Erdal Ozkan Oil & Gas Science and Technology – Rev. IFP Energies nouvelles (2016) 71, 56
Ngo, T. D., Fourno, A., & Noetinger, B. (2017). Modeling of transport processes through large-scale discrete fracture networks using conforming meshes and open-source software. Journal of Hydrology, 554, 66-79.
Fourno, A., Ngo, T. D., Noetinger, B., & La Borderie, C. (2019). FraC: A new conforming mesh method for discrete fracture networks. Journal of Computational Physics, 376, 713-732.
Dual Virtual Element Methods for Discrete Fracture Matrix models Alessio Fumagalli and Eirik Keilegavlen Oil & Gas Science and Technology - Rev. IFP Energies nouvelles 74, 41 (2019)
Author Response
Response to Reviewer 2 Comments
That is an interesting paper in which an open source code for simulating unconventional reservoirs with fractures (either hydraulic fractures or natural) is presented. It is well organized and written. The link to the code repository is given. Many physical processes (gas flow, Forcheimer, hydro mechanics etc...) processes are accounted for. Fractures are accounted for via the EDFM method. The presentation of the codes is clear and concise, in the MRST matlab framework.
In conclusion, apart these remarks, this is a good paper.
Point 1: Note that all these physical processes imply laws involving many semi empirical coefficients, the accuracy of which may be discussed, for example fractional derivatives may be of interest when considering shale gas flow (Albineli et al, see below). These aspects can be put in relation with the large uncertainties encountered while considering fractured reservoirs.
Response 1: Thanks for your suggestion. We agree with that semi-empirical coefficients will greatly affect the accuracy of our code when using it for field applications. As you mentioned, this involves running large number of simulations to discuss the effect of these coefficients on production performance; however, this investigation would go beyond the scope of this manuscript, and we may consider it for future works.
Point 2: As uncertainty studies imply running many forward simulations, could the authors give some information about the performance of their code: running time, number of mesh elements and fractures that can be considered. Authors use an automatic differentiation module, a very interesting feature to compute overall sensitivity coefficient, did they use it in that direction?
Response 2: As this code is an extension module based on MRST (https://www.sintef.no/Projectweb/MRST/), the performance is consistent with the MRST. As far as the usage of automatic differentiation in sensitivity coefficient evaluation is concerned, we believe it is not necessary in this stage, however we consider it in the next developments.
Based on your suggestion, we have added info concerning the performance of the code in Section 5.2 (in orange).
Point 3: All the developments are for 2D cases, I think that 3D cases may be discussed, especially because 3D modelling of flow in fractured media are becoming more and more popular both for shale gas recovery applications, or in the geothermal context (Ngo et al, Fourno et al, Fumagalli and Keilegavlen see below).
Response 3: We agree; in fact, as reported in the conclusions, 3D implementation is very important in practical applications. So, we will work on the code to support 3D cases. The new code and a tutorial will available in an upcoming book (Advanced Modelling with the MATLAB Reservoir Simulation Toolbox (MRST)).
Reviewer 3 Report
The paper is very well written. It is well structured and scientifically sound. I only have very few comments:
- In the introduction, I would consider removing some linebreaks to achieve larger paragraphs to improve readability. However, that's personal taste.
- line 45: the EDFM has BEEN recently proposed.
- Sec. 2.4: Please specify a bit about the mechanics. From your text, I assume that the stress field is static. Or is the model resolved by poroelasticity?
- line 241f: This sentence is already given in the introduction line 60-62.
- Throughout the text, the use of font-color and font type is not consistent.
- While the presented paper is well for its purpose, the toolbox itself would benefit from a more detailed manual covering the first steps after the installation.
Author Response
Response to Reviewer 3 Comments
The paper is very well written. It is well structured and scientifically sound. I only have very few comments
Point 1: In the introduction, I would consider removing some linebreaks to achieve larger paragraphs to improve readability. However, that's personal taste.
Line 45: the EDFM has BEEN recently proposed.
Line 241f: This sentence is already given in the introduction line 60-62. Throughout the text, the use of font-color and font type is not consistent.
Response 1: Thank you for pointing these out. We have amended the manuscript following your suggestions. The changes are marked in orange.
Point 2: Sec. 2.4: Please specify a bit about the mechanics. From your text, I assume that the stress field is static. Or is the model resolved by poroelasticity?
Response 2: We resort to empirical models relating matrix and fracture permeability with the change of pore pressure (or effective stress), given the total stress field unchanged. Based on your suggestion, we have added additional explanations in the same section, marked in orange in the revised manuscript.
Point 3: While the presented paper is well for its purpose, the toolbox itself would benefit from a more detailed manual covering the first steps after the installation.
Response 3: We agree. To helps the reader to run their first example, we have added more details about install and usage in the \texttt{github} page (in orange).
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
- All methods used in this study have been already published in the literature. There is no any new fundamental contribution to shale gas simulation.
- In addition, the method of EDFM can only handle 2D hydraulic fractures, which is very weak. There are many many papers published in the literature to easily handle 3D any complex hydraulic fractures.
-
The model cannot handle more realistic 3D natural fractures, leading to the results are not so valuable. There are many powerful EDFM models in the literature which can handle 3D any complex natural fractures. Hence, the model in this study is not realistic. Please consider the 3D natural fracture networks with any dip angles.
- More importantly, the method in this study cannot handle two phase flow of gas and water in shale gas simulation, which cannot be ignored.
- This paper cannot handle more realistic geology model with corner point.
- There are no any validation of hydraulic fracture geometry with fracture propagation model.
Reviewer 2 Report
1- My major concern is in validation of the code where CMG and in-house simulator is used.
it's not clear what mechanisms are considered when performing these comparison. For example CMG does not take into account slippage effect, its not also consider BET adsorption you can have langmuir adsorption only, the geomechanical effect and stress dependent perm is completely different than what is considered in this paper and there is no discussion on what is considered in in-house simulator. to be able to have a validation the basis of comparison should be at least similar.
2- In conclusion of the validation, figures 11, 13, and 14 EDFM model sees to fail the validation mostly on rate predictions while EFM is more accurate. however in abstract and conclusion EDFM is proposed.
3- in figures 11, 13, and 14 the fracture conductivity seems to not showing important effect on cum production changing from 10000md-ft to 50 md-ft which is something that needs to be discussed.
