Hydraulic interface properties of mixed-flow karst systems

The parameter study shows the different influences of the hydraulic parameters on the drawdown and the derivative curve. The different set-ups have in common that a radial flow period is reached at the end of the pumping test characterizing the hydraulic permeability on regional scale. Differences between the described parameter settings can be detected during the storage period as well as the transition period. The different parametrizations also influence the duration and shape of the linear flow regime. Beside the discussed parameters the linear flow period is also influenced by the conduit length and the development of the conduit network. Here, these effects are excluded. As already mentioned the transition period characterizes the pressure equilibrium stage between the conduit and the matrix systems. Changes of the hydraulic parameters therefore influence the shape of the drawdown curve. Based on this conclusion a collection of different type curves correspond to different hydraulic settings can be created characterizing the influence of the interface between conduit and fissured matrix. All set ups consider a single conduit of the length ^L = 3000 m and constant diameter. The laminar Hagen–Poiseuille flow equation is applied to achieve quasi-infinite conductivity along the conduit.

Figure 5.5 consist of 9 different diagnostic plots representing different interface properties. Every diagnostic plot combines two different pairs of drawdown and derivative curve: (a) the black curves present the drawdown behavior associated to changes of the hydraulic matrix parameters and (b) the red curves present the drawdown behavior associated to changes of the hydraulic conduit parameters. Both collections of diagnostic plots can also be found in Appendix I (matrix parameter variation) and Appendix II (conduit parameter variation) supplemented by the parameter values. The colored dots represent the time of pumping start. Therefore the curves are not congruent but

they represent the same flow behavior (in different time intervals), which will be further discussed.

Fig. 5.5: Dimensionless type curves for a single conduit of L = 3000 m depending on the dimensionless wellbore storage and the skin damage factor. Black lines represent conduit associated parameter changes; red lines represent the fissured matrix associated parameter changes

Along the rows the dimensionless wellbore storage decreases from left to right with values ranging of CD = 0.01 to CD = 0.0001. Values of CD = 0 would represent the double-fissured porosity approach (DROGUE,1992) without any storage associated to the conduit system. For small CD values it can be assumed that the conduit is mainly intersected by joints, faults and fissures with marginal porosity (e.g. WORTHINGTON ET AL., 2000, BOURDET, 2001). For high CD

values a fast-responding storage is provided by fractures or solution enlarged features hydraulically connected to the conduit (e.g. MCCONNELL,1993). Along the columns the type curves change as a function of the skin damage factor,

ranging from Sf = 0.01 to Sf = 1. Small skin damage factors of Sf = 0.01 do not add an additional interfacial pressure drop and the conductivity differences between interface and matrix are high. With increasing skin damage values the flow restriction of the interface increases.

The diagnostic plots of Figure 5.5 represent the diversity of mixed flow karst systems. Depending on the hydraulic properties of the (fissured) matrix and the conduit system, characteristic flow field conditions can be observed. Those changes can be detected on different catchment scales as presented by Figure 5.6.

Fig. 5.6: Schematic representation of different degrees of local and regional karstification.

Based on the hydraulic properties different conceptual karst systems can be defined. The models of Figure 5.5a–c are characterized by a flow restriction of the interface between the conduit and the (fissured) matrix. A low permeability on a local scale (Fig. 5.6) that can be associated with a low degree of macroscopic

of the conduit. The transition period between storage period and linear flow is, compared to the other diagnostic plots, extended. In contrast, the cone of depression inside the matrix has a shallow slope due to the generally high hydraulic matrix conductivity, which is characteristic for a uniformly karstified aquifer. The generally high degree of karstification is also an additional explanation for the high hydraulic gradients between the conduit and the matrix, i.e. on a local scale. That can be the result of turbulent exchange flow or restricted inflow from additional karstic features in the vicinity of the conduit.

Under certain circumstances the interface can also be affected by debris load or collapsed structures.

The diagnostic plots of Figure 5.5g-i represent karst systems with high skin permeability (cf. Fig. 5.6). The hydraulic gradient in the vicinity of the conduit is low resulting in short transition periods. Due to the low hydraulic conductivity of the fissured matrix the cone of depression has a shallow slope on regional scale. The conceptual model is dominated by localized karstification and flow restriction caused by the matrix. The characteristics are similar to karst catchments referred to as matrix restricted karst flow systems.

Apart from the interface permeability the diagnostic plots of Figure 5.5 present differences of storage distribution. The fast-responding storage is characterized by the storage area instead of volume. Therefore, the storage volume, related to the recent precipitation history, is not considered. An increased fast-responding storage (Fig. 5.5.a/d/g) extends the duration of the storage period. This effect is intensified by high skin damage factors. Hence the linear flow period is nearly completely masked by the storage period and afterwards the diagnostic plots of Figure 5.5a tend to radial flow (cf. Theis type curve). In case of a small fast-responding storage area combined with a low skin damage factor the storage period is absent. Changes of the matrix storativity influence the diffusivity of the matrix. With decreased matrix storage and hence increased diffusivity (Fig. 5.5a/d/g) the cone of depression extends more rapidly inside the matrix. This results in a short linear flow period.

Figure 5.5 shows that the influences of the interface hydraulics are multifaceted. Different degrees of storativity and permeability contrasts between the highly conductive conduits and the (fissured) matrix change the overall flow behavior during pumping tests. Those differences can be related to

different conceptual models of the local scale of the karst aquifers as illustrated in Figure 5.6.