| Front page | | Contents | | Previous | | Next |

Model assessment of reductive dechlorination as a remediation technology for contaminant sources in fractured clay: Modeling tool

Appendix J Sensitivity analysis

• J.1 Sensitivity to fracture aperture and spacing
• J.2 Global sensitivity analysis

J.1 Sensitivity to fracture aperture and spacing

The model response to change in fracture aperture and spacing is assessed for the three different degradation scenarios by analyzing the output parameters shown on Figure J.1: time to remove 90% of the initial total contaminant mass, average time for the fracture outlet concentration to be lower than a limit concentration and finally average peak concentrations for the two daughter products.

The different geometric configurations of Table G.1 are used with the transport and degradation parameters from the base case scenario.

Figure J.1 - Parameters for sensitivity analysis of fracture aperture and spacing – total mass on right axis (blue line)

C_lim is defined for each of the chlorinated compounds as a function of the groundwater quality standard (C_standard) recommended in Miljøstyrelsen [2005] and an assumed dilution factor of 10 between the concentration at the fracture outlet and the resulting concentration in the underlying aquifer. Hence C_lim = C_standard*10. The dilution factor of 10 is assumed here, but does not change the main conclusions from this part.

Table J.1 - Groundwater quality standard and output parameters from model

Click here to see Figure J.2

Figure J.2 – Average time to reach Ci < C_lim for no degradation (a), degradation in fracture (b), degradation in fracture and reaction zone in matrix (c) and degradation in fracture and the whole matrix (d), note the log vertical scale

Click here to see Figure J.3

Figure J.3 – Average time to remove 90% of the initial contaminant mass, for no degradation (a), degradation in fracture (b), degradation in fracture and reaction zone in matrix (c) and degradation in fracture and the whole matrix (d), note the log vertical scale

Click here to see Figure J.4

Figure J.4 – Average maximum concentration for the daughter products (DCE and VC), for degradation in fracture (a), degradation in fracture and reaction zone in matrix (b) and degradation in fracture and the whole matrix (c), note the log vertical and horizontal scale

The single fracture/matrix model is not sensitive to the fracture aperture, except for the maximum daughter products concentration in case of degradation in fracture (Figure J.4 (a)) and a limited sensitivity for the average time to reach an output concentration below C_lim in case of degradation in the fracture (Figure J.2 (b)). In these two cases, aperture reduction results in a higher water velocity in the fracture and therefore the daughter products do not have the time to be produced. Otherwise the non-sensitivity of this parameter is explained by the definition of flow in the fracture, which depends on the fracture spacing only (see Section 5.1.2 in the main report). On the contrary the model results are very sensitive to the fracture spacing: the mass removal time increases with fracture spacing, as well as the time to reach an output concentration below C_lim. Furthermore, assuming degradation in the whole matrix leads to a decrease of the clean-up times. In this case, the model is less sensitive to fracture spacing.

J.2 Global sensitivity analysis

The sensitivity analysis is performed on all independent parameters (Table J.2), of which the value is changed by +/- 20%. The resulting change in the three output parameters (time to remove 90% of the initial contaminant mass, average time to reach C_i < C_lim and average peak concentration of daughter products) is normalized to calculate the sensitivity index (see Appendix D). The sensitivity analysis is performed on the third degradation scenario (degradation in the fracture and in a reaction zone in the matrix) with a homogenous TCE concentration as initial condition. In order to be able to compare the different simulations, this initial concentration is corrected in order to maintain the same initial total mass.

Table J.2 - Independent parameters for sensitivity analysis (in orange transport parameters, in green degradation parameters)

The four most sensitive independent parameters are the same for the different output considered, matrix porosity, fracture spacing, net recharge and TCE sorption coefficient. The most sensitive parameters are then the ones controlling transport, especially diffusion/sorption processes, and not dechlorination. The limiting process in this system is the counter diffusion out of the matrix, which is controlled by sorption, flushing of the fracture and fracture spacing.

The two least sensitive parameters are the fracture aperture and longitudinal dispersivity in the fracture. The low sensitivity of this last parameter is due to the fact that transport in the fracture is mainly advective and not dispersive.

Table J.3 - Sensitivity index for the three output parameters

Sensitivity analysis has also been performed by varying the parameters in the typical ranges found in the literature and the same conclusions can be done, concerning the most and least sensitive parameters.

| Front page | | Contents | | Previous | | Next | | Top |

Version 1.0 July 2009, © Danish Environmental Protection Agency