Technical question regarding mass conservation in TOUGH3
Dears,
I am trying to model CO2 injection in brine saturated aquifer using a 1D radial model. CO2 is injected with a constant rate (10 kg/s). Rate distribution throughout the reservoir in my case is obtained using the COFT block data type as illustrated in the attached figure. I have a problem with interpreting the mass rate distribution because I find the mass rate (mainly liquid mass rate) exceeds the injected mass rate ahead of the flooded front (at r = 10 m) which is not in agreement with the mass conservation principle; the mass rate at any location in the reservoir is supposed to be lower/or equal to the injected mass rate. These rates are outputted in the OUTPUT_CONNE excel file. I can find FLOW, FLOW_G, and FLOW_L which are supposed to be total, gas, and liquid mass rate, respectively.
Thanks in advance
2 replies
-
Dear Refaat,
Very nice plot! I have run the radial CO2 injection problem many times, but I never thought to plot the results that way. I just reran it and I see the same effect as you - liquid mass flow rate is greater than CO2 injection rate. I think the code is behaving correctly. Remember that CO2 density is much less than water density. So when you inject 10 kg/s of CO2 it fills up a lot more space than 10 kg/s of water would. So when the CO2 pushes on the water ahead of the front, since water compressibility is low, the volumetric flow rate of water would be about the same as the volumetric flow rate of CO2. With water's greater density, the water mass flow rate is bigger than the CO2 mass flow rate. Conservation of mass is not violated - you are just decreasing the total mass in this region as lower-density CO2 replaces higher-density water. Does that make sense?
Christine
P.S. I think I'll add this way of plotting next time I teach the radial CO2 injection problem in the TOUGH Beginner's Short Course - this is an interesting effect.
-
Christine Doughty Dear Dr. Doughty,
Thanks so much for your reply. Sure, you can do it. You are absolutely right. I ran the same case with two different scenarios. The first one is water injection in a brine saturated reservoir and the second one is a gas injection in a gas-saturated reservoir. I made the salinity of the native brine to be zero for the water injection case and I have also deactivated the energy balance solution through the MULTI data block to exclude the effect of temperature on the rate distribution. I obtained the attached results for both cases.
-
