The UFA can be used to test materials for various flow and transport properties before final decisions or costly field testing.(Conca, J. L. and J. V. Wright. 1992. Applied Hydrogeology 1:5-24.) The UFA can determine optimal operating parameters of other technologies, e.g., testing nutrient formulas for in situ bioremediation; determining thermal, vapor and hydraulic conductivities versus moisture content for in situ heating and vapor extraction; measuring tank sludge waste behavior for mitigating waste tank safety issues; determining saturated and unsaturated retardation factors and leach rates of heavy metals, uranium and other actinides for emplacing permeable reactive barriers; and measuring fluid conductivity for carbon tetrachloride, ethylene glycol and other organic fluids. These direct measurements can provide remediation strategies and defensible performance assessments on schedule and at lower cost.
Under unsaturated conditions, gravel acts as an effective hydraulic barrier to inflow of water from the surrounding environment and can be used to isolate subsurface disposal facilities or other structures (Figure 1).
This use is referred to as a diversion barrier, capillary barrier or Richards Barrier.(Frind, E. O., R. W. Gillham, and J. F. Pickens. 1976. Chapter 3 of Finite Elements in Water Resources, edited by W. Gray and G. Pinder, Pentech Press, London.) A Richards Barrier consists of a slightly sloped layer of gravel below a layer of finer-grained material such as sand or silt. In Figure 2, Kr, Ks, and Kg refer to the hydraulic conductivities of the rock matrix, the sand, and a tuff gravel, respectively, at their respective steady-state unsaturated water contents, 
.
Similarly, D refers to their respective diffusion coefficients. At hydraulic steady-state, the matric potential, or capillary pressure, is equal across the boundary, 
g = s. A positive pressure in the direction of flow is necessary for flow across the boundary from the sand into the gravel, a condition requiring the bottom of the sand layer to become completely saturated. If the hydraulic conductivity of the sand is sufficiently high, even a slight slope prevents the boundary from ever saturating. The flow rates required to saturate an ordinary sand or coarse silt are greater than occurs in any unsaturated zone saturated, i.e., recharges over 3 x 105 cm3/cm2/yr or over 10,000 inches of rain per year. Thus, the sand will never become saturated and no advective flow will occur into the gravel or onto the waste packages it is protecting. The size and nature of the gravel and sand must be chosen to optimize the relative conductivities and potential characteristics of each layer. The UFA was used to test various sands and gravels prior to pilot scale tests of a Richards Barrier for the high-level nuclear waste repository at Yucca Mountain. Different designs of a Richards Barrier made from candidate gravels and soils were run in large Percolation Boxes. Various water fluxes could be put into the top of the system to replicate desired recharge conditions, either as dominant matrix flow, periodic or continuous discreet fracture flow, or both. Table 1 shows results for some candidate soils and gravels. The clayey loam is a poor candidate soil because it has a relatively low saturated hydraulic conductivity and the barrier failed at relatively low recharge ates. On the other hand, the sandy loam was ideal. The sizeof the gravel had only secondary effect.
Structural offsets from seismic events were investigated using the Percolation Box by incorporating 10 cm vertical offsets along the sand/gravel boundary oriented to pool the flow and cause failure. Since this is a dynamic flow system, the flux or flow rate that can be handled by any particular height of offset should be predictable using the unsaturated hydraulic conductivity curve, K(), together with the water content/height distribution, h() or (), for the soil adjacent and upgradient to the offset. To test the predictive capability of the UFA, the hydraulic conductivity and matric potential of each soil was obtained using the UFA. Figure 3 shows the results for the sandy loam.
The K() curve gives the steady-state flux or recharge for this soil at any water content. Choosing a particular offset, e.g., 10 cm, the respective volumetric water content will be 44% and the maximum flux that this soil can handle is about 10-3 cm/s. Normalizing to the flow area across the top of the offset in the Percolation Box, that translates to a predicted fracture flow rate of 500 ml/hr at which the 10 cm offset should fail. The 10 cm offset actually failed at 800 ml/hr. This is eqivalent to a recharge of 3.6 x 103 cm/yr, a truly robust performance.
The UFA can be used in similar ways to predict or test the performance of various materials and remediation strategies for any situation involving flow and transport.
©1996-12 UFA Ventures Inc. UFA Ventures Inc.
