Site-scale Numerical Simulation Study Of CO₂ Leakage Impact On Shallow Groundwater Quality

The deep stored carbon dioxide would leak through the unknown fault, fracturein the cap rock and abandoned wells through the cap rock of storage formation, andreact with the original groundwater in the aquifer and mineral composition in theaquifer media through complicate geochemistry reactions, which would impact thegroundwater quality in shallow aquifer, and reduce the water supplement function ofthe drinking aquifer; and the natural remediation by the aquifer itself and artificialinjection/extraction would impact the concentration of groundwater species withdifferent degrees after the leakage stopped.In this study, a site has been chosen for the injection test, the groundwater withsaturated carbon dioxide have been injected into the target aquifer through injectionwell to simulation the leakage of carbon dioxide from deep formations, and monitorthe concentration variations of main species in groundwater. Based on the survey ofgeology, hydrogeology conditions, analysis the baseline of groundwater species intarget aquifer and the concentration variations during the injection test, establish areactive transport model to simulation the injection test, and inverse thehyrogeochemistry mechanisms, and compared effects with different remediationmethods.The baseline of groundwater quality has been taken based on the monitoringresults for 20 months before the injection test, and 102 samples and 45 species havebeen detected, and the results show that the species concentrations are stable; pHvaries between 7.5 and 8.5. According to the comparison results among the averagedata group and 10 typical samples have been chosen, the saturation index of calciteare between -0.09 and -0.95, which means the calcite are unsaturated and maybe thecontrol mineral of groundwater quality; the saturation index of dolomite are between-2.41 and -4.12, which means the dolomite are unsaturated, but it couldn't be thecontrol mineral of groundwater quality, the saturation index of siderite and Fe(OH)₃ are near zero, which means they maybe impact the groundwater quality; theoxidation-reduction potential analysis results show that, the Eh of HS-/SO₄^2-arebetween -0.322 and -0.263V, the Eh of Fe²⁺/Fe³⁺are between 0.101 and 0.243V, thedifference means the aquifer is not in the equilibrium state. The averageconcentrations for every species could be used as the baseline of target aquifer.The field injection test began on Oct. 18, 2011, and the monitoring results ofMW3 well show that, pH and the concentrations of Ca, Mg, Sr, Mn, K and Na occurobvious changes during the injection test, the peak times are around Nov. 13, 2011,and the pH and concentration of Mn exceed the maximum control limit; according tothe correlation analysis results between species in groundwater during test, thespecies could be divided into 2 groups: (1) Ca, Mg, Sr and Mn, their concentrationhave good linear relationship with concentration of Ca species, the slope are near 1.0,which mean that they should be affected by the same mechanism; (2) Na and K, theirconcentration has good linear relationship with concentration of Ca species too, butcompare with the first group, the correlation worse than first group, maybe becausethey have different resources.Based on the field survey results and the monitoring results during the injectiontest, considering the dissolution and precipitation of controlling minerals and ionexchange affect, a reactive transport model has been established withTOUGHREACT 2.0. The simulation results show that, because of the injection withcarbon dioxide saturated groundwater, the pH of target aquifer decreased, and thedissolution of calcite enhance the concentration of Ca in groundwater, the increasingCa could be the driving force in the ion exchange process, and the concentrations ofMg, Sr, Mn, K and Na increased. The good fitting results show that, the establishedmodel could understand the hydrogeochemistry mechanism during the injection testwell. The simulation results show that, at beginning of the injection test, the spatialdistribution of Ca represented as the highest concentration around the injection welland decreased to the surround, as the test going, the weaken dissolution of calciteand the dilute of injected groundwater with baseline, the spatial distribution of Carepresented as the lowest concentration around the injection well and increased tothe surround until reach the peak concentration, and then decreased to the baseline at the edge of affected area, the affected area enlarged as the test going on; the initialvolume fraction and dissolution rate sensitivity analysis results of calcite show that,with the increasing of initial volume fraction, the peak concentration of Ca ingroundwater increased, and the peak time delay, with the increasing of dissolutionrate of calcite, the dissolution finishing quicker and quicker, and the peakconcentration of Ca decreased; the sensitivity analysis of Cation Exchange Capacity(CEC) results show that, when the CEC is larger than 12 meq/100g, the change ofCEC will impact the concentrations of groundwater dissolution species significantly,when the CEC is smaller than 12 meq/100g, there are no significant changes ofgroundwater species, and consider the composition of aquifer media as coarse sandand gravel, organic matters with high CEC, the initial number of CEC used in themodel do make sense.When the leakage of carbon dioxide stopped, the target aquifer could recoveryby itself under the natural condition or artificial injection and extraction. In this study,based on the established reactive transport model, consider the spatial and temporaldistribution of concentration at the 21stday after the beginning of injection test as theinitial conditions, there are 6 different remediation method through the change ofinjection and extraction conditions, and assess the remediation results under naturalcondition and artificial injection and extractions, and the results show that, (1) thedistribution of pH and Ca would move to the downstream under natural condition,and the circle formation would transform into oval, and affect new area, because ofthe dissolution of calcite, the area with high pH reduced, but the peak concentrationof Ca enlarged one order; (2) although the concentration of main species decreasedwhen keep present injection and extraction condition, but because the distance ofextraction, enlarged the affected area during the remediation, and several area withhigh concentration occurred around the extraction well; (3) the concentration of mainspecies under the condition of sole injection in IW1 are similar with case 2 at thebeginning, and then move to the downstream, but new area affected in this case; (4)the concentration of main species eliminate quickly under the condition of soleextraction in IW1, and all the concentrations decreased to baseline after 6 months; (5)there would be a local area with high concentration around the extraction well underthe condition of injection of MW1, MW2, MW4 and extraction of IW1, although it's a small area, but the concentration of Ca could be 1.5 time of the beginningconcentration, and this case is not better than case 4; (6) under the condition ofextraction of MW1, MW2, MW4 and injection of IW1, the remediation time wouldlonger than case 4, and enlarged the affected area, several area with highconcentration occurred around the extraction wells. According the discussionaforementioned, the case 4 with sole extraction of IW1would the best method for theremediation of carbon dioxide leakage affected aquifer, but the extractedgroundwater should be further treated.