294-7: Effects of Polyacrylamide (PAM) Treated Soils on Water Seepage in Unlined Water Delivery Canals

1
294-7 Effects of Polyacrylamide (PAM) Treated
Soils on Water Seepage in Unlined Water Delivery
Canals Jianting (Julian) Zhu1, Michael H. Young2
and Ernesto Moran3 Desert Research Institute,
Division of Hydrologic Sciences, Las Vegas, NV
89119 (1Jianting.Zhu_at_dri.edu 2Michael.Young_at_dri.e
du 3Ernesto.Moran_at_dri.edu)
Substituting (6) into (5) and manipulating, we
obtain
Results Using Laboratory Data
Introduction
Analysis of Steady State Infiltration
(7)

• 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.

• 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 a for the 3 soil
  types used in the experiments Carsel and Parrish
  (1988) suggest ? 0.145 (1/cm) for C33 sand, and
  ? 0.036 (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 L400 (cm) and canal water depth h 50
  (cm).

• 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).

where z1 is the height of the layer 1layer 2
interface above the water table, z1L-d, where d
is the thickness of PAM-treated layer. When dltltL,
z1 L, and the equation governing infiltration
rate into a two-layer soil becomes
(8)
where ??2/?1, ?Ks2/Ks1. i.e., the ratios of
hydraulic parameters after and before PAM
treatments of soils.
PAM layer thickness d 0.1 (cm)
PAM layer thickness d 0.05 (cm)
The infiltration rate, p, can be obtained solving
iteratively for p/Ks1 For special cases where ?
1, ? 1 and d 0, the result reduces to
one-layer scenario (i.e., without PAM WOP).
Objectives
(9)

• 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.

For steady state conditions, Darcys law through
each layer gives
For another special case where ? 1, but Ks2 lt
Ks1, Equation (8) reduces to
(1)
where z is the elevation, ? is the suction head
(a positive quantity for unsaturated soil), ?
?i at location zi, q is the Darcian velocity
(flux rate) and K(?) the unsaturated hydraulic
conductivity function.
(10)
The ratio of seepage rates for the two-layer
system can be defined as
Laboratory Experiments to Determine Hydraulic
Conductivities of PAM Treated Soils
Using the Gardner model for hydraulic
conductivity function
(2)
(11)
where Ks is the saturated hydraulic conductivity,
and ? is a parameter.
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.
For Gardner hydraulic conductivity function, the
general solution can be shown as (see the above
figure for symbol meaning)
Sensitivity Analysis
(i 1, 2)
(3)
For the bottom layer with no PAM, the steady
state flow equation can be written as
(4)

• 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.

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
(5)
Concluding Remarks
where L-z1 is equal to the thickness of
PAM-treated soil layer, d L-z1. In the case of
ponding, ?2 equals the negative ponding depth.

• Results show that seepage loss rates are quite
  sensitive to ?. Some method should be developed
  to characterize a 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.

From Equation (4), we can obtain
Seepage ratio vs. canal water depth for C33 sand
for the scenario of PAM 40 lbs/acre and SSC
300ppm
Ks ratio after PAM treatment (a) C33 sand, (b)
70 mesh sand, (c) loam soil
(6)
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.
Acknowledgment The financial support by U.S.
Bureau of Reclamation, under cooperative
agreement 05-FC-81-1165, is greatly appreciated.
Contents do not reflect the views of this agency.