Major technical issue of GWB-14

[Mojo]

Dear Brian and Jia,

I have been working with GWB react module for the past few months and I just wanted to let you know that the program does not report the right porosity values when the system gets a little bit complex. For example, when we include evaporation of water into the program for a simple system (water + quartz), as water evaporates during simulation time, porosity decreases! The way porosity is calculated in GWB seems to be wrong. Porosity should be calculated from:

                                                                                      phi= pore volume/ bulk volume= (bulk volume- mineral volume)/bulk volume

In a porous media, the bulk volume a rock/porous media is always constant, but in GWB it changes! Even when we have reactions, the bulk volume should remain constant. What should change is the mineral volume. For example, when a reaction led to the dissolution of minerals, the mineral volume decreases while the pore volume increases. Therefore, the bulk will remain constant. This is the real physic of porous media. As such, porosity will evolve.

Now, for the case of evaporation that I mentioned, since GWB assume pore volume = fluid volume, as soon as evaporation starts, porosity decrease which is wrong and misleading. In reality, the porosity of rock will remain unchanged by evaporation of its pore fluid unless the evaporation of formation fluid led to the precipitation of minerals and an increase in mineral volume. Note that for the simple example that I mentioned, you will see mineral volume will remain unchanged.

Another example is the temperature drop. If you make a simple model (water+ quartz) and drop the temperature, a reduction in the volume of fluid would occur which is expected. However, since GWB assumes fluid volume is equal to pore volume, it gives a reduction in system porosity which again is wrong. If there is no reaction or change in mineral volume, porosity should remain constant.

Another example is the dissolution of CO2 in brine. If we have a sliding CO2 fugacity in a system without any reactions. We will see fluid volume increase linearly over time as CO2 dissolve in brine and therefore porosity increases! This is wrong if mineral volume remains unchanged. 

Do you have any feedback on this? Please let me know since it is a major problem that can mislead a lot of users and gives wrong data. This is of great importance for users using GWB for porous media applications such as scaling issues in reservoirs.

 

I look forward to hearing from you.

 

Kind Regards

Mojtaba

 

[Jia Wang]

Hello Mojtaba, 

The bulk volume in the GWB is calculated as the sum of the mineral volume, fluid volume,and inert volume. The GWB calculates the porosity by dividing the fluid volume by bulk volume. Note that this is slightly different from your description stated above. GWB apps are designed to look at the saturated portion of your domain. In other words, you should really think about this as the porosity for the saturated fraction of the porous media. For more information on this, please look at the volume and porosity commands in section 6.116 and 6.71 in the GWB Command Reference guide.  

If you are using React and your model precipitates a mineral, the porosity reported would decrease as the fraction of mineral volume increases. If you decrease the fluid volume though (i.e. by evaporation), the bulk volume calculated by the program will decrease with the decreasing fluid volume which leads to a decrease in the porosity.  If you are trying to simulate scaling with React, you will want to look at the change in mineral volume and not the fluid volume for the results you expect

On the other hand, you might consider working in the reactive transport model applications X1t or X2t to model scaling effect. In X1t and X2t, the bulk volume is fixed, because of the size of the nodal blocks specify by the constraints in the domain pane so they should produce results closer to what you are expecting. In particular, the fixed bulk volumes should help with the variation in density due to temperature and salinity. Please note that these are still saturated models. The porosity calculated are the same as described above. If you try to evaporate as you did in React, this might not perform as expected. For more details on how porosity evolves in reactive transport models, please see section 2.12 Porosity Evolution in the Reactive Transport Modeling guide. 

Hope this helps,
Jia Wang