294-7: Effects of Polyacrylamide (PAM) Treated Soils on Water Seepage in Unlined Water Delivery Canals Jianting (Julian) Zhu 1, Michael H. Young 2 and Ernesto Moran 3 Desert Research Institute, Division of Hydrologic Sciences, Las Vegas, NV ( Introduction Objectives Acknowledgment: The financial support by U.S. Bureau of Reclamation, under cooperative agreement 05-FC , is greatly appreciated. Contents do not reflect the views of this agency. Viscosity / Surface sealing / Plugging of pores Experiments were designed to explain Ks reduction by quantifying role of viscosity changes, conductivity of PAM layer, and plugging of large soil pores. Analysis of Steady State Infiltration When PAM is added to full-scale canal, it hydrates, reacts with suspended sediment and settles to the bottom of canal ‘prism.’ Flocculate reduces infiltration and thus seepage loss. Full-scale canals can exhibit seepage in bottom and sidewalls – this analysis considers only vertical infiltration. 1D flow process considers two-layer soils with distinct and contrasting hydraulic properties (see figure on the right). For steady state conditions, Darcy’s law through each layer gives The saturated hydraulic conductivities before PAM treatment and effective hydraulic conductivities after PAM treatment were determined from constant head experiments. Experimental design included 3 suspended concentrations of kaolinite, and soils of 3 textures (Moran, 2006), including: C33 coarse sand #70 mesh sand Loam-textured soil Graphs on left are averaged Ks values for given experimental conditions. Ks decreases with increases in both PAM application rates (expressed in lbs/acre) and suspended sediment concentrations (SSC). Largest reduction in Ks observed in C33 and #70 mesh sand. Small reduction observed in loam. Results Using Laboratory Data Laboratory Experiments to Determine Hydraulic Conductivities of PAM Treated Soils where z 1 is the height of the layer 1:layer 2 interface above the water table, z 1 =L-d, where d is the thickness of PAM-treated layer. When d<<L, z 1 ~ L, and the equation governing infiltration rate into a two-layer soil becomes where  =  2 /  1,  =K s2 /K s1. i.e., the ratios of hydraulic parameters after and before PAM treatments of soils. The infiltration rate, p, can be obtained solving iteratively for p/K s1 For special cases where  = 1,  = 1 and d = 0, the result reduces to one-layer scenario (i.e., without PAM [WOP]). Substituting (6) into (5) and manipulating, we obtain For another special case where  = 1, but K s2 < K s1, Equation (8) reduces to The ratio of seepage rates for the two-layer system can be defined as where z is the elevation,  is the suction head (a positive quantity for unsaturated soil),  =  i at location z i, q is the Darcian velocity (flux rate) and K(  ) the unsaturated hydraulic conductivity function. Using the Gardner model for hydraulic conductivity function: where Ks is the saturated hydraulic conductivity, and  is a parameter. For Gardner hydraulic conductivity function, the general solution can be shown as (see the above figure for symbol meaning) For the bottom layer with no PAM, the steady state flow equation can be written as: where q is the vertical flux rate of water which is the same for both layers (for infiltration, q is positive). For the top (PAM treated) layer, the steady state flow equation can be written as where L-z 1 is equal to the thickness of PAM-treated soil layer, d = L-z 1. In the case of ponding,  2 equals the negative ponding depth. From Equation (4), we can obtain Note that infiltration rate (q) is negative, i.e., p = -q, and that negative suction at canal bottom is replaced with ponding depth h, i.e.,  2 = -h. (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) The ratio, r, in Equation (11) can be considered as a measure of efficacy of PAM applications for reducing water seepage loss. Smaller r values correspond to higher efficiency of PAM application. We assume typical values of  for the 3 soil types used in the experiments; Carsel and Parrish (1988) suggest  = (1/cm) for C33 sand, and  = (1/cm) for loam. We assume a value of 0.1 (1/cm) for the #70 mesh sand, which is between C33 sand and loam groups but is closer to the sand group. Graphs below illustrate predicted seepage ratio for several combinations of PAM and SSC concentrations. For each graph, water table depth L=400 (cm) and canal water depth h = 50 (cm).  Results show that seepage loss rates are quite sensitive to . Some method should be developed to characterize  for the thin PAM layer.  Groundwater table depth does not affect seepage loss ratio between PAM-treated and untreated soils, but depth to water table and water depth in canal would influence the actual water seepage loss. The groundwater table depth is only important for determining seepage loss ratio when water table depth converges on canal bottom.  Canal water depth influences seepage loss ratio of PAM-treated and untreated soils, if  values of the two layers differ. Concluding Remarks Ks ratio after PAM treatment: (a) C33 sand, (b) #70 mesh sand, (c) loam soil Seepage ratio vs. canal water depth for C33 sand for the scenario of PAM = 40 lbs/acre and SSC = 300ppm Investigate the effects of applying a thin PAM layer at the bottom of water delivery canals on water seepage into subsurface material. Determine the parameter sensitivity for reducing water seepage loss through canal bottoms. Persistent drought in the west and mid-west US has created a significant need to improve efficiency of water delivery canal systems. Polyacrylamide (PAM) has been used extensively in furrow irrigation applications, but recently is undergoing scientific scrutiny as a potential water conservation tool for unlined canals. Laboratory, field and numerical experiments are being conducted to better understand the benefits (efficacy, longevity) and risks (environmental and human impacts) of PAM usage in canals. The research described herein uses laboratory results and a new analytical approach for estimating seepage loss reduction after PAM is applied to soil. Sensitivity Analysis PAM layer thickness d = 0.1 (cm) PAM layer thickness d = 0.05 (cm) (i = 1, 2)