| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Soil and Water Science Unit, Univ. of California, Riverside, CA 92521
Corresponding author (john.letey{at}ucr.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONTHEORYMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
3.1 for each treated sand. The infiltration rate increased with increased time for lower h0 values, but decreased with increased time for higher h0 values. The transition from increasing to decreasing infiltration rates with time occurred when h0/hp was approximately equal to 2.6. The infiltration rate behavior of an aqueous ethanol solution was consistent with theoretical relationships based on liquid surface tension. A positive hydraulic head was created at the interface of an overlying wettable and underlying water-repellent layer that affected the infiltration rate consistent with the effects of h0 on a nonlayered water-repellent sand. The following mechanism is proposed to explain the increase in infiltration rate with time. In water-repellent materials, positive hydraulic heads can be created within the profile during infiltration, which can increase as the depth to the wetting front increases. The higher hydraulic head induces an increase in hydraulic conductivity, which contributes to increased infiltration rate. Alternatively, if the depth of ponded water is sufficient to cause a hydraulic conductivity equal to that of the wettable material, the infiltration rate behavior is the same as traditionally observed for wettable soils.
Abbreviations: h0, depth of ponded water • hp, water entry pressure head • K, hydraulic conductivity • WDPT, water drop penetration time • 
l, liquid surface tension • ND, 90° surface tension • s, solid–air surface tension • sl, solid–liquid surface tension • 
, liquid–solid contact angle • 
, liquid density
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
, >90°, is slow during the initial phase of infiltration and increases with time. Water repellency can create unstable water flow within the soil matrix, which is related to an increased risk of groundwater contamination (Hendrickx et al., 1988; DeJonge et al., 1999).
Water-repellent soils have been found in many agricultural and natural areas throughout the world (Ritsema, 1998). A water-repellent soil does not wet spontaneously when water is applied at zero or negative pressure potential because is >90°. A positive pressure must be applied to force liquid into the soil. This pressure is the water entry pressure head, which has also been referred to as the breakthrough pressure head, hp. The numerical value of hp depends on both and the effective pore radius. It increases as increases and decreases as the pore radius decreases. The degree of water repellency of some soils changes with time after contact with water. For cases where the water does not initially penetrate the soil, but does so after some time of contact with water; changed from an initial value of greater than 90° to less than 90°.
The reason for different temporal infiltration behavior into water-repellent and wettable soil has not been thoroughly examined. Carrillo et al. (2000a)(2000b) investigated the effects of a water-repellent soil layer on unstable water flow (finger formation) through wettable soil below the repellent layer. They found that water flow through and below the water-repellent layer became more uniform as the combined depth of ponded water and depth to the repellent layer increased. The hydraulic conductivity of a wettable sand is independent of the depth of ponded water. However, Carrillo et al. (2000a) observed that the hydraulic conductivity of a water-repellent material increased with increase in depth of ponded water. The increase in hydraulic conductivity with increased ponding depth was associated with increased water content with increased ponding depth. We found no reports that investigated the effects of ponding depth on infiltration into a nonlayered, water-repellent soil.
The objectives of this study were (i) to determine the effect of ponding depth on hydraulic conductivity of two water-repellent sands and compare them with wettable sand; (ii) to examine infiltration rate as a function of time into dry water-repellent sands as affected by water ponding depth; (iii) to verify theoretical relationships between liquid surface tension, water repellency, and infiltration rate by using ethanol–water solutions; and (iv) to examine the effects of an overlying wettable layer on infiltration into a lower lying water-repellent layer.
|
THEORY |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
 | (1) |
is the liquid–air surface tension (hereafter referred to as the liquid surface tension), is the liquid–solid contact angle, r is the capillary radius, is the liquid density, and g is the gravitational constant.
When is greater than 90°, cos is negative and water will not enter the tube unless a positive pressure is applied. This pressure is referred to as the water entry pressure, hp.
The interfacial tensions between liquid and solid phases were given by Good and Girifalco (1960).
 | (2) |
sl is the solid–liquid surface tension, s is the solid–air surface tension, l is the liquid–air surface tension, and 
is a function of molecular properties of the solid and liquid. For a water–hydrocarbon system, is approximately unity.
Young's (1805) equation is
 | (3) |
Combining Eq. [2] and [3] and assuming = 1 leads to
 | (4) |
Selecting (= 90°) so that cos equals zero, and thus l is the 90° surface tension, ND, leads to
 | (5) |
Substituting Eq. [5] into Eq. [4] results in
 | (6) |
The contact angle can be directly calculated by Eq. [6] after ND is measured. Substituting Eq. [6] into Eq. [1] results in
 | (7) |
The capillary radius, r, can be determined based on Eq. [7] after hp and ND are measured.
Equation [7] can be used to calculate the solution entry pressure head for aqueous ethanol solutions by inserting the appropriate values of l and for the solutions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
A series of aqueous ethanol solutions producing various surface tensions was prepared for the ND measurement. A plot of percentage ethanol vs. surface tension at 23°C was produced by measuring the surface tension of each mixture using a surface tensiomat (Fisher Scientific Co., Pittsburg, PA). Twenty grams of sand were placed in an agar plate and leveled. Then three drops (40 µL) of these solutions were applied on the sand surface with a Pasteur pipette and the time taken for droplets to completely penetrate the sand were noted. The lowest ethanol concentration (highest surface tension) that penetrated in 5 s was taken to be ND as specified by Watson and Letey (1970). This value was used to calculate the water–sand contact angle.
The water entry pressure, hp, was measured using the technique of Carrillo et al. (1999). The treated sand was packed into a column and the ponding depth at which the water started infiltrating into the sand was measured and recorded as the water entry pressure.
Polypropylene tubes with 5-cm i.d. and a wall thickness of 0.4 cm were used to measure infiltration rate, i, and hydraulic conductivity, K. The tube holding the sand sample was either 20 or 40 cm long. A fine wire screen at the bottom of the tube retained the sand and allowed air to escape. The upper tube in which liquids were ponded was 60 cm long and connected to the sample tube. A polypropylene tube (6-cm i.d. and 2.5-cm long) was glued to the top tube as a connector. A 0.3-cm-deep groove was cut into the middle of the connector and a rubber O-ring was put in the groove. The bottom tube was pushed into the connector. The O-ring provided a water-tight seal. A port located 2 cm above the sand surface was used for water application to the top tube.
To prevent water from preferentially flowing between the sand and the tube wall, the tube was coated with a Teflon-based dry film lubricant before each measurement. A "packing apparatus" (Glass et al., 1989) was used to uniformly pack sand in the tube. The packing apparatus was a tube that contained three coarse wire mesh grates at 2, 11, and 21 cm from the bottom of the tube. These grates randomized the sand as it fell through the small tube into the soil sample tube. The average bulk density of the packed sand was 1.45 g cm-3. A strip of ruler was attached along the sand column tube to measure the wetting front depth.
A Marriott bottle was used to add distilled water or ethanol solutions to the column and maintain a constant depth of ponding. The opening at the bottom of the Marriott bottle was connected to the port of the tube above the sand by a plastic tube that was clamped until liquid was to be introduced. A tube through the stopper at the top of the Marriott bottle was placed at a position to maintain the desired constant head depth.
The Marriott bottle was placed on a balance (Model 7300, Pennsylvania Scale, Leola, PA; capacity 11.34 kg, precision 0.001 kg), and the weight recorded by a computer as a function of time. The infiltration rate was calculated from incremental time changes in weight of the Marriott bottle. Hydraulic conductivity measurements were done in a similar manner except that flow was allowed to occur until it reached a steady-state value. All measurements were done in a laboratory where the temperature was 23 ± 1°C.
For the first objective, the hydraulic conductivities of the treated and untreated sand columns were measured for different water ponding depths, h0. Standard procedures were used where the depth of ponded water was established and the steady-state water flow was measured. The hydraulic conductivity was calculated by dividing the steady-state water flow rate by the hydraulic head gradient across the sand column.
The second objective involved measuring infiltration rate as a function of time for different depths of ponded water. The treated sand was packed into the 20-cm tube as previously described and connected to the top tube. Water was introduced from the Marriott bottle and allowed to pond at a predetermined depth. The infiltration rate was calculated from the rate of water flow from the Marriott bottle to the column to maintain the constant head. Time zero was selected at the time when the desired head was achieved (a few seconds after the tube was unclamped transmitting water from the bottle to the column).
The infiltration rates were measured with h0 values of 15.0, 20.0, 24.5, and 30.5 cm for Treatment 1, and h0 values of 5.2, 7.8, 9.0, 10.0, and 19.0 cm for Treatment 2. Infiltration rate was measured in untreated sand with h0 = 5.2 cm.
The third objective was to verify theoretical relationships between liquid surface tension, water repellency, and infiltration rate. The experimental technique for this objective was the same as for the second objective except that an aqueous ethanol solution was used instead of distilled water for the infiltration rate measurements for the Treatment 1 sand. Based on the measured ND listed in Table 1 and using Eq. [5], s values of Treatment 1 and Treatment 2 were 0.0112 and 0.0132 N m-1, respectively. Since r and g are constant for a given sand, Eq. [7] can be used to calculate hp for a solution of known and . A 9% (v/v) ethanol solution with l equal to 0.056 N m-1 was used as the infiltration liquid. Substituting r, l, s, and into Eq. [7] the value of hp for Treatment 1 sand was determined to be 3.5 cm for the aqueous ethanol solution as compared to 8.4 cm for water. Theoretically the temporal infiltration behavior of this aqueous ethanol solution should be the same as water when the ponding depth of the ethanol solution is 4.9 cm less than water. In order to compare infiltration behavior for ethanol solutions with water for the same (h0 - hp) values, infiltration measurements were made for the ethanol solutions with h0 values of 10.0, 15.0, 19.5, and 25.5 cm.
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
ND and surface tension of water. Contact angles of both treatments were >90°; thus, a positive pressure, hp, is required for infiltration to occur in the two repellent sands. Treatment 1 was more water repellent than Treatment 2. Figure 1
illustrates that the hydraulic conductivity, K, in water-repellent sands depends on the depth of ponded water, h0. The increase in h0 up to a critical value resulted in an increased K, which is consistent with the results of Carrillo et al. (2000a). The K in the water-repellent sand attained the value of untreated sand if h0 was high enough. The ratio of h0/hp that resulted in maximum K equivalent to the untreated K was 3.1 for each treatment. Carrillo et al. (2000a) determined that the average water content in the sand decreased as the value of h0 decreased. Thus the decrease in K is caused by a decrease in the water content. The water entry pressure increases as the pore size decreases. Therefore the increase in water content could be the result of smaller pores being filled with water as h0 increased. Alternatively, finger flow through water-repellent sand could have caused only a portion of the sand to be wet. Whether the decrease in average water content was uniformly distributed in the sand or as result of finger flow where only a fraction of the sand was wet was not conclusively determined.
|
|
|
2.6) produced nearly constant infiltration rate as a function of time. This is also the approximate h0/hp value that produced hydraulic conductivity values close to that of untreated sand. Letey et al. (1962) attributed the increasing infiltration rate with time to organic coatings on the soil dissolving into the infiltrating water and reducing its surface tension. That explanation is not valid for the present findings because the treated sands had a very stable water repellency that did not change with time after contact with water. Furthermore, that explanation is not consistent with the observed effect of changing h0. Thus another mechanism is responsible for the dependence of the infiltration rate on h0 that will be discussed below.
The infiltration rate for the aqueous ethanol solution that had equal to 0.056 N m-1 is plotted as a function of time for the Treatment 1 sand in Fig. 4
. As with water, the shape of the infiltration rate curve was different for different values of h0. The critical h0 value for which the infiltration rate shifted from increasing to decreasing values with increased time is consistent with the theoretical estimates based on surface tension. Theoretically, the shape of the ethanol solution infiltration rate curve would be similar to that of water when the h0 value for water was 5 cm greater than for ethanol solution. A comparison between the results in Fig. 2 and 4 reveals that the expected behavior was observed. Note that h0 of ethanol solution equal to 19.5 cm produced the infiltration curve shape comparable with h0 of water equal to 24.5 cm, but opposite to that of h0 of water equal to 20 cm. The absolute infiltration rate values of the ethanol solution differed from those of water because of viscosity differences.
|
14 cm. Possibly because of packing or other variations this treated sample apparently had a higher water entry pressure than in other experiments on the same treated sand. Nevertheless, the interface hydraulic head increased and reached a stable value of 17 cm after 35 min. The infiltration rate continued to increase with time even with a constant head to a steady state value of 1.2 mm min-1 when water was seeping from the bottom of the column.
|
The hydraulic head at the interface between the upper 5 cm of untreated sand over a 20-cm layer of treated sand and the infiltration rate are plotted as a function of time in Fig. 6
for the case of ponded water depth equal to 23.5 cm. The hydraulic head at the interface rose rapidly until 9 cm and infiltration rate into the treated layer began. Thereafter, the hydraulic head continued to rise at a more gradual rate until a steady value of 24 cm was reached.
|
2.5 mm min-1. Using the hydraulic conductivity value of h0 equal to 24 cm and the hydraulic head gradient across the treated layer resulted in the calculated flow rate of 2.7 mm min-1, which is only slightly higher than the measured steady-state flow rate. The following mechanism is proposed to explain the increase in infiltration rate with time. In water-repellent materials, positive hydraulic heads can be created within the profile during infiltration. As the depth to wetting front increases, the depth of ponded water plus depth to the wetting front increases, which can induce a higher hydraulic head at the wetting front than for shallower wetting front. The higher hydraulic head induces an increase in water content of a water-repellent sand with a consequent increase in hydraulic conductivity that further contributes to increased infiltration rate. Alternatively, if the depth of ponded water is sufficient to cause a hydraulic conductivity equal to the wettable material, the infiltration rate behavior is the same as traditionally observed for wettable soils.
Preferential (finger) flow has been reported in water-repellent soils (Hendrickx et al., 1988, 1993; Ritsema et al., 1993; DeBano, 1971; Bauters et al., 1998; Wang et al., 1998; and Carrillo et al., 2000a, 2000b). Raats (1973) noted that the stability of a wetting front depended on the rate of change of the wetting front velocity with depth. Instability is maintained when the velocity increases with depth and tends to disappear when the velocity decreases with depth. The observed increase in infiltration rate with time for the lower ponded water depths (Fig. 2 and 3) would be consistent with unstable flow according to Raats theory. Although we did not make direct measurements, for the lower ponding depth runs the sand columns appeared to be preferentially wet in some zones.
|
SUMMARY |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
Water did not infiltrate until the depth of ponding exceeded the water entry pressure. Thereafter the infiltration rate as a function of time was dependent on the depth of ponding. At lower ponding depths the infiltration rate increased with increased time. At the larger ponding depths the infiltration rate decreased with increased time as is typical for wettable soils. Modifying the liquid surface tension by ethanol resulted in a behavior consistent with the theoretical change in the solution entry pressure. Specifically lowering the liquid surface tension decreased the depth of ponding where the infiltration curve shifted from increasing to decreasing infiltration as time increased.
Placing an untreated layer over the treated layer caused a positive hydraulic head to occur at the interface between the two layers. Thereafter the infiltration rate behavior was consistent with the hydraulic heads created at the interface. Increasing the hydraulic head at the interface caused the hydraulic conductivity of the repellent underlying layer to increase.
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 10, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONTHEORYMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARYREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Arye, I. Nadav, and Y. Chen Short-term Reestablishment of Soil Water Repellency after Wetting: Effect on Capillary Pressure-Saturation Relationship Soil Sci. Soc. Am. J., April 5, 2007; 71(3): 692 - 702. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Gilboa, J. Bachmann, S. K. Woche, and Y. Chen Applicability of Interfacial Theories of Surface Tension to Water-Repellent Soils Soil Sci. Soc. Am. J., August 3, 2006; 70(5): 1417 - 1429. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. L. Feng, J. Letey, and L. Wu The Influence of Two Surfactants on Infiltration into a Water-Repellent Soil Soil Sci. Soc. Am. J., March 1, 2002; 66(2): 361 - 367. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
