Soil Science Society of America Journal 65:1675-1679 (2001)
© 2001 Soil Science Society of America
DIVISION S-1 - NOTES
Comparison of the thermal properties of four wettable and four water-repellent soils
J. Bachmanna,
R. Horton*,b,
T. Rena,c and
R. R. van der Ploega
a Str. 2, 30419 Hannover, Germany
b Dept. of Agronomy, Iowa State Univ., Ames, IA 50011
c Soil and Fertilizer Institute, Hebei Academy of Agricultural Sciences, Hebei, China
* Corresponding author (rhorton{at}iastate.edu)
 |
ABSTRACT
|
|---|
Water repellency is a widespread phenomenon, but its effect on most physical soil properties is still unknown. The present study deals with the thermal conductivity and volumetric heat capacity of wettable and water-repellent (hydrophobic) sandy and silty soils. Thermal conductivity and volumetric heat capacity were determined by heat pulse measurements. Heat pulse measurements were made on samples of eight soils (four pairs of wettable soil and the corresponding water-repellent counterpart). Water repellency either was caused by soil organic matter showing natural repellency or induced by chemical treatment of wettable soils with dichlorodimethylsilane (C2H6SiCl2). Thermal conductivity was also predicted with the models of de Vries and of Campbell. Almost all measured conductivities were larger than those predicted by the de Vries model. For the wettable soils, most of the conductivities in the soil water saturation range between 0.20 and 0.75 differed by more than 0.5 W m-1 K-1. The hydrophobic soils however, showed only in the range around a saturation degree of 0.20 to 0.50 values that deviated more than 0.5 W m-1 K-1 from the predictions. The Campbell model underestimated the conductivity at low saturation for wettable and hydrophobic soil, but overestimated it at high saturation for the wettable soil. Thermal conductivity for either dry or water-saturated soil was predicted satisfactorily by both models. It was found further that soil wettability had no systematic impact on heat capacity. It is concluded that soil thermal conductivity decreased as soil hydrophobicity increased, whereas the volumetric heat capacity was not affected by soil wettability.
|
INTRODUCTION
|
|---|
NUMEROUS STUDIES in past decades have pointed out that in many soils the presence of soil organic matter is associated with water repellency (Jaramillo et al., 2000). In some cases, water repellency appears in the humic topsoil, which is also the domain of intensive heat and mass exchange processes. In a recent study, Bachmann et al. (2001) showed with use of a column experiment that water repellency reduced evaporation. Induced repellency thus may have practical applications. Because water repellency affects the distribution of water in soil and water distribution affects some of the thermal properties, water repellency might affect soil thermal properties. To date, however, data on thermal properties of unsaturated water-repellent soils have not been reported. Therefore, a study of water-repellency effects on soil thermal properties is needed. The heat pulse method, as developed and tested in a number of recent studies (Campbell et al., 1991; Kluitenberg et al., 1993, 1995; Tarara and Ham, 1997; Bristow, 1998; Song et al., 1998) was used for this purpose as a reliable and handy technique.
The main objectives of this paper are to present measurements of thermal properties for wettable soils and comparable water-repellent soils and to investigate the applicability of two thermal conductivity models. Generally, for mineral soils, the thermal conductivity can be predicted satisfactorily (Sepaskhah and Boersma, 1979; Hopmans and Dane, 1986) with the standard model of de Vries (1963), and is thus used in numerous theoretical and experimental studies (Blom and Troelstra, 1972; Milly, 1984; Nassar and Horton, 1997). The de Vries model is physically based and includes the heat conductance because of vapor transport. Different bulk densities, mineral compositions, water contents, or temperatures are considered explicitly by the model. Because the de Vries model requires an extensive amount of input parameters, we also tested the empirical model described by Campbell (1985), which has a considerably lower demand of input data.
|
Materials and Methods
|
|---|
Soils
Because the mineral composition and the shape of the solid phase have a considerable impact on the thermal conductivity (de Vries, 1963), we conducted our experiments with a series of wettable soils and with the corresponding water-repellent soils that were almost identical with the wettable soils in texture and mineral composition. Thermal properties of the eight samples (four soils paired as wettable–water repellent) were studied with soil water content ranging from air-dry to saturation. All samples were taken from soils around Hannover in Northern Germany. One soil was a Spodosol under agricultural use, developed on glacial till sand (humic loamy sand). The contact angle of samples taken from the Ah horizon of this soil varied considerably, indicating good wettability and strong water repellency. In addition to this pair of natural soils, three wettable soils were made hydrophobic in the laboratory by coating the grain surfaces with Dichlorodimethylsilane (Shaw, 1975). Based upon soil texture, the amount of applied Silane varied between 50 to 90 mL kg-1 soil. The soil materials were a laboratory quartz sand (sand), a loamy subsoil sand from a quaternary river deposit in horticultural use (loamy sand), and a subsoil silt from a Weichselian loess in agricultural use (silt). Basic soil data are given in Table 1.
The degree of repellency of the soil materials was assessed with use of a sessile drop method (Bachmann et al., 2000a,b). Contact angles were measured at room temperature (20 ± 1°C) using a microscope fitted with a goniometer scale. Twelve contact angle readings were made for each soil material.
Heat Pulse Method
The dual-probe heat-pulse device (Campbell et al., 1991; Welch et al., 1996; Tarara and Ham, 1997) was used to determine thermal properties of the soils. The sensor consisted of two stainless steel needles (0.00127 m o.d., 0.028 m in length, and spaced 0.006 m apart) mounted parallel in a polyvinyl Cl (PVC) block. One needle contained a line-source heating element and the other a thermocouple. The heater was made from enameled wire with a diameter of 0.079 mm (Nichrome 80 Alloy; Pelican Wire Co., Naples, FL). Both needles were filled with high thermal conductivity epoxy (RBC-4300 and A-121 epoxy hardener, RBC Industries, Warwick, RI). We refer to Tarara and Ham (1997) for details about the probe design and construction. When a heat pulse is applied to the heater, the temperature change at the sensor needle as a function of time can be expressed (Welch et al., 1996) as:
 | [1] |
where 
T(t, r) is the temperature change (K); t (s) is the time after the heater is turned on; t0 (s) is the heating time; r (m) is the radial distance from the line heater (m), q' is the energy input of the heater (W m-1), 
c and 
are the volumetric heat capacity (J m-3 K-1) and thermal diffusivity (m2 s-1) of the medium, respectively, and –Ei(-x) is the exponential integral. The values of c and can be determined from the T(t, r) curve using a nonlinear procedure (Welch et al., 1996). Soil thermal conductivity, 
, (W m-1 K-1) is then calculated as the product of c and . Preliminary experiments were conducted with a silt loam at a volumetric water content of 10% to evaluate the precision of the heat pulse method. This preliminary experiment was carried out with two sets of bulk densities and five replicates. Results showed that the thermal conductivity of the samples with a bulk density of 1.214 ± 0.003 Mg m-3 was 0.532 W m-1 K-1 with a standard deviation of ±0.007 W m-1 K-1. For the samples with a bulk density of 1.525 ± 0.002 Mg m-3, the corresponding value was 0.687 ± 0.011 W m-1 K-1.
Air-dried samples were prepared by packing dry soils directly into cylinders of 0.052-m i.d. and 0.06-m height. The samples were saturated by slowly wetting the air-dry samples from the bottom of a cylinder. To check the degree of saturation, sample water content was measured after the thermal properties were determined. These measurements showed that the anticipated water content (saturation) corresponded to the porosity of the samples. For samples with intermediate water contents, the soil materials were moistened with distilled water to different water contents, mixed thoroughly, and then packed into the cylinders. Eight cylinders (one for each soil) were prepared. For various soil materials and water contents, the mean packed bulk densities were reasonably similar (Table 1).
During a measurement, the dual-probe heat-pulse sensor was inserted into the soil column from the surface. A constant current was applied from a direct current supply (Model 1635, B&K-Precision, Maxtec International Corp., Chicago, IL) to the heater for 15 s to generate the heat pulse. The heater was turned on automatically by the data logger at t = 0 and turned off at t = 15 s. The temperature data at the sensor needle were collected at 1-s intervals for 180 s (see Fig. 1)
. A data logger (Model 21X, Campbell Scientific Inc., Logan, UT) monitored the applied voltage along with the temperature as a function of time in the sensor needle (Fig. 1).
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 1. Measured changes in soil temperature as a function of time at the sensor needle located 0.006 m from the heater. The heating pulse was 15 s with an energy of 21.1 W m-1.
|
|
The data logger recorded the voltage drop across a 0.11-W precision resistor, from which the heating power was calculated. For the instrumentation employed in this study, the temperature measurement resolution was about 0.0167 K. The energy applied to the heater varied from 12.5 to 36.0 W m-1, depending on soil water content. As a result, the maximum temperature change at the sensor needle was in the range of 0.55 to 0.8 K. Soil thermal properties were calculated by analyzing the temperature change versus time data using a nonlinear curve-fitting technique (Welch et al., 1996). For one particular cylinder at one specific water content two measurements of thermal properties were made. In total 48 pairs of measurements for the wettable soils and 48 pairs for the water-repellent soils were made. Finally, gravimetric soil water content and bulk density were determined for each cylinder.
Prediction of the Thermal Conductivity
Apparent thermal conductivities, , were calculated according to de Vries (1963). For different ranges of the soil water content, 
( = 0.0, 0.0 < < k, and > k, with k being the critical soil water content), different submodels for the calculation of were used. The critical soil water content, k, is generally considered as the transition water content of a drying soil, at which the liquid soil water continuity ends. In this study, k was defined as the equilibrium water content at a pressure potential of -55 kPa.
For > k (Region 1), was calculated from the thermal conductivity of the main individual soil constituents as:
 | [2] |
where ki denotes the ratio of the average temperature gradient in the granular component i and the corresponding gradient in the bulk soil; xi is the volume fraction of component i; and i is the thermal conductivity of component i. Five components (n = 5) were considered: liquid soil water, moist soil air, quartz, other soil minerals, and soil organic matter.
For = 0.0 (Region 3), was calculated as:
 | [3] |
where the subscript a denotes dry air and the other symbols are used as before. For Region 2 (0 < 
k), values for were calculated by linear interpolation from values for for = k and = 0.0. The quantity ki was calculated according to Kimball et al. (1976). For the depolarization (shape) factor of the soil particles, a value of 0.333 was assumed, which applies for spherical soil particles. The water-content dependent shape factors for the air-filled pores ga, were taken as (Blom and Troelstra, 1972):
 | [4] |
where a is the air-filled porosity. The calculations of the diffusion coefficient of water vapor in air were carried out according to Kimball et al. (1976).
The second model used to estimate soil thermal conductivities for comparison with measured thermal conductivities was the empirical model of Campbell (1985). The Campbell model has the following form:
 | [5] |
where the parameters A, B, C, D, and E can be computed from bulk density, clay content, and volume fractions for quartz and other minerals as described by Campbell (1985).
|
Results and Discussion
|
|---|
Figure 2
shows the measured conductivities for all soils. The figure shows for both the wettable and the repellent soils a least-square fitted second-order polynomial curve. For each curve, the 95% confidence interval is also shown. Figure 2 seems to indicate that the
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 2. Measured apparent thermal conductivity for the wettable and the hydrophobic soils. The second-order polynomial curves were determined by least squares regression.
|
|
conductivities for the wettable and the water-repellent soils differed from each other and that the differences between the conductivities became larger with an increasing degree of saturation.
Because the bulk densities of the wettable soils differed only slightly from the corresponding bulk densities of the water-repellent soils (Table 1), we compared the measured thermal conductivity values directly with those based on predictions of either the de Vries or the Campbell model. With exception of the dry and the saturated soil, the measured conductivities were generally larger than the conductivities predicted by the de Vries model. In this respect, our results differ from previous studies in which most discrepancies were observed because of overestimation of thermal conductivity by the de Vries model (Kimball et al., 1976; Horton and Wierenga, 1984).
Figure 3
shows measured and predicted (de Vries model) conductivity values for both the wettable (Fig. 3a) and the repellent (Fig. 3b) humic loamy sand soil. In each case, a second-order curve-fitted polynomial is used to describe the conductivity as a function of the degree of saturation. Figure 3a shows that for intermediate degrees of saturation (between 0.15 and 0.75) considerably larger conductivities for the wettable soil were measured than expected from the de Vries model. The deviations were appreciably smaller for the corresponding water-repellent soil (Fig. 3b).
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 3. Measured and de Vries model-predicted thermal conductivity values for (a) wettable and (b) water-repellent humic sandy loam soil.
|
|
The results of Fig. 3 also hold when the corresponding data from all soils are combined (Fig. 4)
. In contrast to Fig. 3, however, now the differences (measured - predicted) as a function of the degree of saturation rather than -values are plotted. These differences were calculated both with use of the de Vries and the Campbell models. Figure 4 includes second-order polynomial fits of the data. Figure 4 shows that the measured conductivities match the de Vries model predictions best for dry soil. For these conditions the deviations are <0.2 W m-1 K-1. With two exceptions (hydrophilic sand and hydrophilic loamy sand) the predictions also are close (<0.25 W m-1 K-1) to the measurements at soil saturation. For intermediate water contents, however, Fig. 4 indicates that the measurements for water-repellent soils were closer to the estimates. The mean difference is considerably larger for the wettable soils.
The close agreement at saturation (except the wettable sand and the wettable loamy sand) indicates that the shape factors of the solid phase were chosen properly (see de Vries and Philip, 1986). The prediction at air dryness was also satisfactory for the mean of all measurements as indicated by the regression (Fig. 4). Larger deviations between measurement and prediction, as observed for the unsaturated soil, may have been caused by an improper estimate of the air-shape factor in the three-phase system as discussed by de Vries and Philip (1986). Air-shape factor has been modified by various authors. Neiss (1982) suggested to increase the shape factor in the wet range and to decrease it in the dry range, while Kimball et al. (1976) proposed ga() functions below the curve postulated by de Vries (1963) resulting in smaller conductivities. However, none of these suggestions did increase the precision of the model estimates for our soils.
Figure 4 also shows the deviations between estimates of the Campbell model and the measurements. In this case, the estimates were smaller for saturation degrees <0.5 for all soils, but larger for saturation degrees between 0.5 and 0.8 for the wettable soils. Again, the hydrophobic soils showed the tendency for slightly lower differences between estimated and measured conductivities. Because of the lack of further information, some speculations about possible mechanisms are appropriate.
A review of wettability effects on fluids in porous media indicates that the location of the wetting phase (e.g., water in wettable soils) or the location of the nonwetting phase (e.g., water in repellent soils) depends on the wettability of the fluid related to the surface of the solid (Dullien, 1979). For unsaturated soil–water systems this could have the consequence that, with increasing hydrophobicity, the heat conduction through the water menisci is decreasing with a decreasing amount of water that is located in the edges between the contact points of the solid phase. This feature may also affect the average diffusive pathway for water vapor. A third process involved might be the formation of macroscopic liquid domains in hydrophobic media and, correspondingly, the formation of relatively dry areas, which may reduce heat conduction in a way similar to aggregate formation of fine-textured soils (Hadas, 1977). Because of the physical nature of the de Vries model and the systematic deviations from the measurements, we conclude that the enhancement of heat transfer is caused by a physical effect, which is not represented through the parameters of this model. This phenomenon can affect heat transport via conduction through soil particles and liquid phase or heat transport because of the transport of vapor in the gaseous phase. Additional experiments might give an improved insight into the physical mechanisms that are involved. Hiraiwa and Kasubuchi (2000) found that the thermal conductivity resulting from latent heat transfer could be separated from the apparent thermal conductivity by subtracting the thermal conductivity at a temperature near 0°C from that at higher temperatures. Provided that the contact angle does not change considerably with temperature, it may be concluded that measurements carried out close to 0°C may help to separate the effect of the air-shape factor from the heat transfer coupled to the water vapor transport. Environmental scanning electron microscopy (ESEM) might also be a tool to locate the liquid phase within the unsaturated soil matrix.
In contrast to the conductivities, there was agreement between the measurements of the volumetric heat capacity of wettable and nonwettable soil (Fig. 5)
. Systematic differences of heat capacities between wettable and hydrophobic soils, comparable with the thermal conductivities, were not observed. This is consistent with our understanding that volumetric heat capacity is affected by fractions of solid, liquid, and gas, but not by the spatial arrangement of the phases.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 5. Measured volumetric heat capacities of all wettable and water-repellent soils. The two second-order polynomial curves were determined by least squares regression.
|
|
|
Conclusions
|
|---|
This study is considered as a first investigation of the thermal conductivity in water- repellent soils. Results show that the thermal conductivity is larger for wettable soils than for corresponding nonwettable soils. A comparison of measured thermal conductivities with two models resulted in considerable deviations between predicted and measured values. The de Vries model underestimated the thermal conductivity of the wettable soil by more than 0.5 W m-1 K-1; whereas for the hydrophobic soils the deviations were considerably smaller. These results suggest that an individual calibration of the underlying thermal conductivity model is recommended for each soil. Especially the thermal properties of wettable humic topsoils need to be examined before reliable predictions of soil heat flux can be made. A surprising result of this study is that the prediction of the thermal conductivity matches the measured conductivity data better for the hydrophobic soils than for wettable soils, which built the original database for the development of heat conductivity models.
|
NOTES
|
|---|
Journal Paper no. J-19034 of the Iowa Agric. and Home Econ. Exp. Stn., Ames, Iowa; Project No. 3287, and supported in part by Hatch Act and State of Iowa.
Received for publication October 3, 2000.
|
REFERENCES
|
|---|
- Bachmann, J., A. Ellies, and K.H. Hartge. 2000a. Development and application of a new sessile drop contact angle method to assess soil water repellency. J. Hydrol. 231/32:66–75.
- Bachmann, J., R. Horton, R.R. van der Ploeg, and S. Woche. 2000b. Modified sessile-drop method assessing initial soil–water contact angle of sandy soil. Soil Sci. Soc. Am. J. 64:564–567.[Abstract/Free Full Text]
- Bachmann, J., R. Horton, and R.R. van der Ploeg. 2001. Isothermal and nonisothermal evaporation from four sandy soils of different water repellency. Soil Sci. Soc. Am. J. 65:1599–1607 (this issue).[Abstract/Free Full Text]
- Blom, Th.J.M., and S.R. Troelstra. 1972. A simulation model of the combined transport of water and heat produced by a thermal gradient in porous media. Rep. No. 6. Dep. of Theoretical Production Ecology, Wageningen, the Netherlands.
- Bristow, K.L. 1998. Measurement of thermal properties and water content of unsaturated sandy soil using dual-probe heat-pulse probes. Agric. For. Meteorol. 89:75–84.
- Campbell, G.S. 1985. Soil physics with BASIC. Elsevier, Amsterdam, the Netherlands.
- Campbell, G.S., C. Calissendorf, and J.H. Williams. 1991. Probe for measuring soil specific heat using a heat-pulse method. Soil Sci. Soc. Am. J. 55:291–293.[Abstract/Free Full Text]
- de Vries, D.A. 1963. Thermal properties of soils. p. 210–235. In W.R. van Wijk (ed.) Physics of plant environment. John Wiley and Sons, Inc., New York.
- de Vries, D.A., and J.R. Philip. 1986. Soil heat flux, thermal conductivity, and the null-alignment method. Soil Sci. Soc. Am. J. 50:12–18.[Abstract/Free Full Text]
- Dullien, F.A.L. 1979. Porous media. Fluid transport and pore structure. Academic Press, New York.
- Hadas, A. 1977. Heat transfer in dry aggregated soil: I. Heat conduction. Soil Sci. Soc. Am. J. 41:1055–1059.[Abstract/Free Full Text]
- Hiraiwa, Y., and T. Kasubuchi. 2000. Temperature dependence of thermal conductivity of soil over a wide range of temperature (5-75°C). Eur. J. Soil Sci. 51:211–218.
- Horton, R., and P.J. Wierenga. 1984. The effect of column wetting on soil thermal conductivity. Soil Sci. 138:102–107.
- Hopmans, W.J., and J.H. Dane. 1986. Thermal conductivity of two porous media as a function of water content, temperature and density. Soil Sci. 142:187–195.
- Jaramillo, D.F., L.W. Dekker, C.J. Ritsema, and J.M.H. Hendrickx. 2000. Occurrence of soil water repellency in arid and humid climates. J. Hydrol. 231/232:105–111.
- Kimball, B.A., R.D. Jackson, R.J. Reginato, F.S. Nakayama, and S.B. Idso. 1976. Comparison of field-measured and calculated soil-heat fluxes. Soil Sci. Soc. Am. Proc. 40:18–25.
- Kluitenberg, G.J., K.L. Ham, and K.L. Bristow. 1993. Error analysis of the heat pulse method for measuring soil volumetric heat capacity. Soil Sci. Soc. Am. J. 57:1444–1451.[Abstract/Free Full Text]
- Kluitenberg, G.J., K.L. Bristow, and B.S. Das. 1995. Error analysis of heat pulse method for measuring soil heat capacity, diffusivity, and conductivity. Soil Sci. Soc. Am. J. 59:719–726.[Abstract/Free Full Text]
- Milly, P.C.D. 1984. A simulation analysis of thermal effects on evaporation from soil. Water Resour. Res. 20:1087–1098.
- Nassar, I.N., and R. Horton. 1997. Heat, water, and solute transfer in unsaturated porous media: I. Theory development and transport coefficient evaluation. Transp. Porous Media 27:17–38.
- Neiss, J. 1982. Numerical simulation of heat and water movement and of the formation of ice in soils. (in German.) VDI Verlag, Düsseldorf, Germany.
- Sepaskhah, A.R., and L. Boersma. 1979. Thermal conductivity of soils as a function of temperature and water content. Soil Sci. Soc Am. J. 43:439–444.[Abstract/Free Full Text]
- Shaw, D.J. 1975. Introduction to colloid and surface chemistry. Butterworths, London.
- Song, Y., M.B. Kirkham, J.M. Ham, and G.J. Kluitenberg. 1998. Measuring soil water content under turfgrass using the technique. J. Am. Soc. Hortic. Sci. 123:937–941.[Abstract/Free Full Text]
- Tarara, J.M., and J.M. Ham. 1997. Measuring soil water content in the laboratory and field with dual-probe heat-capacity sensors. Agron. J. 89:535–542.[Abstract/Free Full Text]
- Welch, S.M., G.J. Kluitenberg, and K.L. Bristow. 1996. Rapid numerical estimation of soil thermal properties for a broad class of heat pulse emitter geometries. Meas. Sci. Technol. 7:932–938.[ISI]
This article has been cited by other articles:
|
|
|
|
 
S. Lu, T. Ren, Y. Gong, and R. Horton
An Improved Model for Predicting Soil Thermal Conductivity from Water Content at Room Temperature
Soil Sci. Soc. Am. J.,
January 1, 2007;
71(1):
8 - 14.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
