CE Declaration CE Deklaration of Conformity - EN 14904 ... - Junckers

224 K. You, H. Zhan / Advances in Water Resources 52 (2013) 221–2313. Field applicationIn this section, the developed model 1D-VT is applied to interpreta field study of TCE transport in natural attenuation at PicatinnyArsenal in Morris County, New Jersey, USA. About1.26 10 5 m 2 of the unconfined aquifer at this site was contaminatedby TCE with a concentration of greater than 10 mg L 1resulting from improper disposal practices of metal degreasingand cleaning operations from 1960 to 1981 [29]. Extensive fieldstudies have been conducted to investigate the TCE contaminationin the unsaturated zone at the site. The unsaturated zone mainlyconsists of medium-to-coarse sand with organic carbon contentranging from 0.080% to 1.020% [29]. The sampling event took placein 1998 at a site where the unsaturated zone had an average thicknessof 3.2 m, a porosity of 0.32 and an average moisture content of0.117 [29]. No precipitation occurred during the sampling period,and the magnitude of the water table fluctuation was less than2cm[29]. The detailed site and sampling information can be foundin Smith et al. [36] and Choi et al. [29].In our study, the molecular weight, dimensionless Henry’s lawconstant, organic carbon partition coefficient and gas phase diffusioncoefficient of TCE are set to 131.4 g mol 1 , 0.28,0.735 m 3 Kg 1 and 0.0787 cm 2 s 1 , respectively, as can be foundfrom Chiao et al. [43]. TCE gas phase concentration at the water tableis set to be 2 mg L 1 , and initially it decreases linearly to 0 atthe ground surface with a constant concentration gradient of0.625 mg L 1 m 1 , as used by Choi et al. [29]. Besides the dataabove, the gas phase permeability and the period and magnitudeof the water table fluctuation are also required to predict the transportof TCE in the unsaturated zone. We used the atmosphericpressure data from August 19 to 21 in 1998 for this site and thecorresponding subsurface pressure data at the depth of 2.45 m toconduct parameter optimization to search for the optimal valuesfor the gas phase permeability k g , the magnitude A 2 and periodT 2 of the water table fluctuation. The simplex method is employedto do the parameter optimization because of its simplicity and rapidconvergence [11,12]. In order to obtain an unique pair of results,a good initial guess of the parameters to be optimized isindispensable [11,12]. We set the initial values of k g , A 2 and T 2 to3.5 10 14 m 2 , 0.01 m and 1 day, respectively, which are basedon the estimated and field observed data in Choi et al. [29].Fig. 1 shows the comparison of the measured and predictedsubsurface gas pressures at the depth of 2.45 m. The predicted subsurfacegas pressures are calculated using the optimal values of k g ,A 2 and T 2 , which are 3.8 10 14 m 2 , 0.001 m and 0.8 days, respectively.These values are close to those estimated by Choi et al. [29].As shown in Fig. 1, by taking into account the water table fluctuationthe predicted and measured subsurface gas pressures matchbetter with each other than those in Choi et al. [29]. The root meansquare error is 21 Pa, which is only 0.021% of the average total gaspressure. This demonstrates the accuracy of the developed solution1D-VT.Fig. 2 shows the change of the gas phase concentration of TCEwith time at the depth of 0.16, 1.6 and 3.0 m, respectively, whichare drawn with the same scales. At the shallow depth (Fig. 2(a)),the change of the concentration is mainly controlled by the atmosphericpressure fluctuation. During this brief time period, theincreasing atmospheric pressure causes the downward pressuredriven flux, which decreases the gas phase concentration, and viceversa. At the depth close to the water table (Fig. 2(c)), the change ofthe concentration is mainly controlled by the water table fluctuation.The upward moving water table increases the gas phase concentrationthere, while the downward moving water tabledecreases it. At the middle depth (Fig. 2(b)), because both theatmospheric pressure and water table fluctuations are retarded,the gas phase concentration fluctuates slightly there.Fig. 3 compares the change of the diffusive and advective fluxeswith time at the depth of 0.16 m. According to Fig. 3, both the diffusiveand the advective fluxes fluctuate with the fluctuations ofthe atmospheric pressure and water table. The change of the diffusiveflux is contrary to that of the gas phase concentration at thesame depth (z = 0.16 m) (Fig. 2(a) and Fig. 3(a)). The increasingatmospheric pressure leads to the decreased gas phase concentrationand increased concentration gradient at the shallow depth,which increases the upward diffusive flux, and vice versa. Themagnitude of the advective flux is about one order of magnitudeless than the diffusive flux. However, the variation in diffusive fluxis much smaller than that of the advective flux. Fig. 4 displays thedepth distribution of the two-day averaged diffusive and advectivefluxes. The averaged advective flux has a maximum value at themiddle depth (z = 2 m). During these two days, the net contributionof the advective flux is to move TCE downward toward the watertable. The averaged diffusive flux is constant across the depth exceptat the place close to the water table and ground surface, wherethe averaged diffusive flux decreases. As seen from Fig. 2, duringmost of the time in these two days, the gas phase concentrationof TCE close to the ground surface is lower than its initial value,while that close to the water table is higher than its initial value.At the same time, the concentrations of TCE at the boundariesare fixed to be constants. Therefore, the concentration gradientsclose to the two boundaries decrease, which results in decreasedaverage diffusive fluxes there. The averaged diffusive flux is abouttwo orders of magnitude greater than the averaged advective flux.Although the transient advective flux may be great, the net contributionof the advective flux induced by atmospheric pressure andwater table fluctuations can be neglected compared with thediffusive flux.1.006 x 105 t (day)1.005Pressure (Pa)1.0041.0031.0021.001calculated pressure at 2.45 mmeasured pressure at 2.45 m1atmospheric pressurecalculated pressure at 2.45 m by Choi et al., 20020.9990 0.5 1 1.5 2Fig. 1. Comparison of the calculated and field measured subsurface gas pressures atthe depth of 2.45 m at Picatinny Arsenal in Morris County, New Jersey, USA.Concentration (mg L -1 )0.10010.09980.09951.00020.99990.99961.9004(a) z=0.16 m(b) z=1.6 m1.9001(c) z=3.0 m1.89980 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2t (day)Fig. 2. Change of the gas phase concentration of TCE with time at (a) z = 0.16 m, (b)z = 1.6 m, and (c) z = 3.0 m.