I am injecting a known mass of CO2 in 2 wells into a 6 m thick 200mx100m aquifer using X2t to calculate mass balance between rock and water. Pretty simple 75% qtz, 15% K-spar, 10% calcite rock composition, looking at dissolution/precipitation reactions with calcite. I have run models for 1 to 0.001 m3/m2/yr discharge rates and 22 g/l and 104 g/l salinity. The high 104 g/l salinity models indicate low CO2 solubility, near zero calcite precipitation and some calcite dissolution. But I am having trouble accounting for the mass of CO2 inject in the fluid or rock using 'elemental composition of rock and fluid' outputs. Some CO2 is leaving the model domain but I still cannot find the CO2, even after accounting for outflow. I am running the models for 10,000 years plus, and am wondering if it is an iteration error? I thought I had figured it out by assuming that the lack of CO2 in rock or fluid compared to what I had injected (corrected for outflow), was accounted for as a gas phase and thus should have a high CO2 fugacity, but the high salinity model at 1 m3/m2/yr has the lowest CO2 fugacity, and this model has the most CO2 unaccounted for...i.e. the mass balance gets worse with higher salinity and groundwater flow rate. My epsilon is 1e-7 which is low so again I think iteration error. I also tried summing the individual dissolved CO2 species to check if the 'total C in fluid' values were similar and they were close but not exact. Overall the 'total carbon in fluid + rock' output says carbon is leaving the system, but I cannot figure out where it is going. Any suggestions?