To assess the potential toxicity of sodium in the soil-water system, the SAR is used.
The SAR is determined by obtaining soil solution from the soil after saturating it with water, removing the solution by vacuum, and analyzing it for Na, Ca and magnesium (Mg) in milligrams per liter. The following formula is then used to estimate the SAR:
SAR = (Na/23)/(Ca/40 + Mg/24)1I2 (11.5) where 23, 40, and 24 are the atomic weights of Na+, Ca2+, and Mg2+, respectively.
Generally, SAR values greater than 15 are considered potentially toxic to plants.
The SAR magnitude reflects the quantity of sodium on the exchange sites of the soil. Most arid-region soils with SAR values of 15 have approximately 15% of their CEC loaded with sodium. This sodium load is known as the exchangeable sodium percentage or ESP. Soils with an ESP greater than 15 would be considered unproduc-
412 SALT-AFFECTED SOILS AND BRACKISH WATERS
tive and, depending on the magnitude of ESP, such soils may also be classified as toxic.
This information, however, comes from soils of the arid regions of the western United States, and one cannot be sure that the critical SAR-ESP threshold of these soils also applies to soils in humid regions.
11.2.3 SAR-ESP Relationships
The sodium adsorption potentials of two humid soils are presented in some detail and the SAR-ESP data are shown in Figures 11.1-11.4. For comparison purposes, these figures also include the SAR-ESP relationship of salt-affected soils found in arid-region soils (western United States). Figures 11.1-11.4 show that for any given SAR, the ESP for either one of the two humid soils is greater than the ESP of the western U.S. soils. This indicates that the two humid soils adsorb sodium on their exchange complex more effectively than the western U.S. soils.
Some differences in the SAR-ESP relationship between the two humid soils (Figs.
11.1-11.4) are also apparent. The data indicate that the SAR-ESP relationship of the Pembroke soil is independent of chloride (en and to some degree pH, but this is not true for the Uniontown soil. It appears that as pH increases, the Uniontown soil shows a strong adsorption preference for Na+, but as Cl-concentration increases, it shows a
100 a... 90
C/) I.IJ 80 ...
.~ E 70
'0 C/) 0 .&:. 60
-~ 50
'0
Q)
'0 0 40
...J 0
U 30
I.IJ
U 20
-::!! 0
0 10
0
0 25 50 75 100
o C I = 175 mg L-I A CI = 1750 mg L-I
o CI '" 7000 mg L-I
125 150 175 Sodium Adsorption Ratio (SAR)
200
Figure 11.1. Relationship between percentage of CEC loaded with sodium (ESP) and SAR at three chloride concentrations of Pembroke soil at pH 4.3 (the solid line without data represents most salt-affected soils in the western United States; it was produced using Eq. 11.4) (from Marsi and Evangelou, 1991a, with permission).
11.2 SALTS AND SOURCES
a. III W
E ::::I
=0 0 III .l:'.
-
'j
"tJ
Q)
"tJ 0 0
...J U w
-u ::l! 0
0
100 90 80 70 60 50 40 30 20 10 0 0
OCI .175mgL- 1 A CI .. 1750 mg L-I
o CI .. 7000 mg L-I
20 40 60 80 100 120 140 160 180 200 Sodium Adsorption Ratio (SAR)
413
Figure 11.2. Relationship between percentage of CEC loaded with sodium (ESP) and SAR at three chloride concentrations of Uniontown soil at pH 4.3 (the solid line without data represents most salt-affected soils in the western United States; it was produced using Eq. 11.4) (from Marsi and Evangelou. 1991a. with permission).
100
-III a. 90
w 80
~
.~ E 70
"tJ III 0 .l:'. 60
-~
"tJ 50
Q)
"tJ
40
0 0 ...J
U 30
w u
-::l! 0 20
• 10
0
0 6 12 18 24 30
Sodium Adsorption Ratio (SAR)
Figure 11.3. Relationship between percentage of CEC loaded with sodium (ESP) and SAR at a chloride concentration of 175 mg L-1 of Uniontown soil at three pH values (the solid line without data represents most salt-affected soils in the western United States; it was produced using Eq. 11.4) (from Marsi and Evangeiou. 1991a. with permission).
414
a..
CJ)
--ILl E
-=
"0 (/) 0 .s: -~
"0 til
"0 0 0 .J
u ILl
-U :.e 0
0
100 90 80 70 60 50 40 30 20 10 0
0
SALT-AFFECTED SOILS AND BRACKISH WATERS
10 20 30 40 50 60 70 80 90 100
Sodium Adsorption Ratio (SAR)
Figure 11.4. Relationship between percentage of CEC loaded with sodium (ESP) and SAR at a chloride concentration of 175 mg L -1 of Pembroke soil at three pH values (the solid line without data represents most salt-affected soils in the western United States (from Evangelou, 1998, unpublished data).
weak preference for Na+. These observations playa very important role in decisions about managing brine discharges onto agricultural soils or into streams and lakes.
In summary, if one discharges brine onto a soil with a strong Na+ adsorption potential, this soil will protect the groundwater from Na+ contamination at the expense of its own potential Na+ contamination. However, a soil with low Na+ adsorption potential will protect itself from Na+ contamination at the expense of potential groundwater contamination.
11.2.4 Adverse Effects ofNa+ in the Soil-Water Environment
Sodium adsorbs more water molecules per unit (mole) of charge than most other metal ions (K+, Mg2+, Ca2+) commonly found in the soil-water environment. Hence, when brine (NaCl) is discharged in the soil-water environment, clay and organic particles tend to adsorb fully hydrated N a ions. This causes the particles to become waterborne, a process also known as dispersion. Under dispersion, soils become impermeable to water; lakes, streams, and rivers experience large increases in suspended solids (clays and organics). Most soil clays undergo dispersion at an ESP of around 15. At this ESP level, soils appear to be toxic because they lose the potential to function as porous media (water infiltration and gas exchange are restricted).
11.2 SALTS AND SOURCES 100
90 o Pembroke
~ 0 80 I:;. Uniontown
"O~
~~ c .- 70 Y = 162.44 -1.83 X
"- .2:
:l -
60
_ u C :l (/)"0 C 50
QJ 0
> 0
"';: u
0 ã - 40
"Q)"S 30 a:: c ...
-g, 20
I
10
OL-.--.I..---L--L-.--I..~..LJ..j~Lo<)()I--...I...-~~ti:Jl
o 10 20 30 40 50 60 70 80 90 100 Dispersion Index (%J
415
Figure 11.S. Relationships between relative saturated hydraulic conductivity and percent dispersion index of two Kentucky soils (dispersion index = percent of total clay remaining waterborne after 1 hr of settling in an Imhoff cone) (from Marsi and Evangelou, 1991c, with permission).
The dispersion phenomenon in the two humid soils (Pembroke and Uniontown) was evaluated through the use of an Imhoff cone test and a permeameter. The Imhoff cone is commonly used by engineers to determine settleable solids (see Chapter 9).
The results of clay dispersion obtained by the Imhoff cone test are expressed as a dispersion index (percent of total clays in the soil sample dispersed), which is correlated with relative saturated hydraulic conductivity. This is shown in Figure 11.5.
It demonstrates that each of the soils, depending on its clay content (Pembroke 59%;
Uniontown 20%), exhibits unique saturated hydraulic conductivity behavior with respect to the dispersion index. Also, in each ofthe soils, various mechanisms (different line slopes) appear to control saturated hydraulic conductivity.
The phenomenon of soil dispersion with respect to Na+ loads (magnitude of ESP or SAR) appears to be unique to all soils on at least one particular point. As the total salt or Cl- concentration in the water increases, the dispersion index decreases and the saturated hydraulic conductivity increases (Fig. 11.6). When this occurs, the soil- water system becomes toxic to plants and organisms owing to high osmotic pressures.
When chloride concentration in solution increases beyond 6000 mg L -I, Na ions near clay surfaces begin to dehydrate because of high osmotic pressure in the surrounding solution. This causes clay particles to flocculate (flocculation is the reverse of disper- sion) and, consequently, the saturated hydraulic conductivity of the soil increases.
11.2.5 Brine Chloride and Bromide
The chloride-bromide concentration in brine is greater than 1000 mg L-1 and varies widely among wells. At concentrations greater than 106 mg L -I in water, it can be toxic to crops.
416
~ 0
,,-Q) >.
-o .:;; ~ : : J -.- -
+- u
o ::J
(1)"0
c::
Q) 0
>U . .;= u
0 ' -
- -Q) ::J
0::: e
"0
>.
:::c 100
90 80 70 60 50 40 30 20 10 0 0
SALT-AFFECTED SOILS AND BRACKISH WATERS
o CI :: 7000 mg L-1
6 C I = 1750 mg L-I
o CI:: 175 mg L-1
10 20 30 40 50 60 70 80 90 100 Exchangeable Sodium Percentage (ESP)
Figure 11.6. Relationship between relative saturated hydraulic conductivity and ESP at three levels of chloride (from Marsi and Evangelou, unpublished data).
Chloride plus bromide is also toxic to crops at elevated concentrations due to the salting-out effect (high osmotic pressure). This salting-out effect appears to become important at EC levels greater than 2 mmhos cm-I. A level of 1 mmhos cm-I is equal to approximately 640 mg L -I dissolved solids or approximately 350 mg L -I Cl-I. Water normally used for human consumption has an EC value significantly less than 1 mmhos cm-I. Because salting-out effects are generally independent of salt type, they can be caused by either sodium chloride plus bromide salt or calcium chloride plus bromide.
11.2.6 Heavy Metals
Although the concentration of heavy metals in brines is usually not high enough to cause alarm, iron can sometimes be quite high (10-100 mg L -I), but it quickly oxidizes and precipitates out because of the high pH of the brine.
11.2.7 Boron
Boron concentration in brines can vary from 10 to 100 mg L -I. In the soil solution, a boron concentration as low as 4 mg L -I is toxic to some crops.
11.2.8 Alkalinity
As such, alkalinity does not cause toxicity. However, at concentrations greater than 90 mg L-I, alkalinity can dramatically increase the toxicity of sodium by removing
11.2 SALTS AND SOURCES 417 calcium from the water as calcium carbonate. If this occurs in an agricultural field owing to brine disposal, the soil-water environment could become highly toxic to crops.
Alkalinity in brines can be evaluated through the use of the pHc. The mathematical expression for the pHc (Langelier or saturation index) is as follows:
Saturation index = pH - pHc (11.6) where pH denotes measured-solution pH, and pHc denotes equilibrium pH for CaC03 under a given set of conditions (under a pC02 of 0.0003 and pure CaC03, pHc = 8.4).
When the saturation index >0, CaC03 precipitation is expected; when the saturation index <0, CaC03 dissolution is expected. The saturation index can be derived by estimating pHc as follows:
(11.7) and
(H+)(Ca-)
K = c 3 = 10.0-10.33
2 (HC03")
(11.8)
where Ksp is the solubility product constant of CaC03, K2 is the second dissociation constant of H2C03, and the parentheses denote solution ion activity. Rearranging and substituting Equation 11.7 into Equation 11.8 gives
(11.9)
"
Taking logarithms on both siOes of Equation 11.9 gives
-log K2 = -log H~ -log Ksp + log Ca2+ + log HC03 (11.10)
Rearranging,
(11.11) A practical approach to estimating the pHc of water moving through soil is as follows:
pHc = (pKz - p~) + p(Ca + Mg) + pAlk (11.12) where p~ and p~ represent pK2 and pKc (p~ = pKsp) corrected for ionic strength (see Table 11.1). An estimated pHc (using Eq. 11.12) ofless than 8.4 suggests that Ca2
+
will precipitate as limestone (CaC03). An estimated pHc (using Equation 11.12) of greater than 8.4 suggests that CaC03, if present, will dissolve. The values of
418 SALT-AFFECTED SOILS AND BRACKISH WATERS p~ - p~, p(Ca + Mg), and pAlk are obtained from Table 11.1 after analyzing the water for Ca, Mg, Na, RC03, and C03. The concentration values in Table 11.1, columns 1,3, and 5 are in milliequivalents per liter (meq L-1). The values in columns 4 and 6 represent the negative logarithms of the corresponding values in columns 3 and 5, respectively. The values for p~ - p~ are obtained from column 2 through the corresponding sum of N a, Ca, and Mg in column 1.
An example using Table 11.1 is demonstrated below. Assuming analysis of a water sample gives
TABLE 11.1. Tables for CalCUlating pHc Values of Waters
Concentration
Concentration Concentration C03 + HC03
Ca+Mg+Na pK2-pK~ Ca+Mg p(Ca+ Mg) (Alkalinity) pArk
(1) (2) (3) (4) (5) (6)
0.5 2.11 0.05 4.60 0.05 4.30
0.7 2.12 0.10 4.30 0.10 4.00
0.9 2.13 0.15 4.12 0.15 3.82
1.2 2.14 0.2 4.00 0.20 3.70
1.6 2.15 0.25 3.90 0.25 3.60
1.9 2.16 0.32 3.80 0.31 3.51
2.4 2.17 0.39 3.70 0.40 3.40
2.8 2.18 0.50 3.60 0.50 3.30
3.3 2.19 0.63 3.50 0.63 3.20
3.9 2.20 0.79 3.40 0.79 3.10
4.5 2.21 1.00 3.30 0.99 3.00
5.1 2.22 1.25 3.20 1.25 2.90
5.8 2.23 1.58 3.10 1.57 2.80
6.6 2.24 1.98 3.00 1.98 2.70
7.4 2.25 2.49 2.90 2.49 2.60
8.3 2.26 3.14 2.80 3.13 2.50
9.2 2.27 3.90 2.70 4.0 2.40
11 ~?-8 4.97 2.60 5.0 2.30
13 2.3b 6.30 2.50 6.3 2.20
15 2.32 7.90 2.40 7.9 2.10
18 2.34 10.00 2.30 9.9 2.00
22 2.36 12.50 2.20 12.5 1.90
25 2.38 15.80 2.10 15.7 1.80
29 2.40 19.80 2.00 19.8 1.70
34 2.42
39 2.44
45 2.46
51 2.48
59 2.50
67 2.52
76 2.54
Source: From Ayers, 1977, with permission.