WATER QUALITY
Ambient Water Quality Criteria for Dissolved Oxygen
2.0 Forms and Transformations
2.1 Physical Properties
Oxygen is the most abundant element of the
earth's crust and waters combined. The combination of the divalent
oxygen atom with single valent hydrogen atom comprises the extremely
stable H2O molecule. Under natural conditions water
exists in several physical states, but the molecule itself dissociates
to a very limited extent as ions (H+ and OH-). Two OH- molecules
can, by covalent bonding, combine to form H2O2 or
hydrogen peroxide. Holm et al. (1986) state that there is evidence
that hydrogen
peroxide is formed and accumulates in the photo-oxidation of
organic compounds in surface and ground waters and in precipitation.
The decomposition of water to yield dissolved
oxygen normally would be outside the realm of ambient conditions;
an endothermic
reaction such as provided by electrolysis is required to produce
O2 and H2 gas. Photosynthesis is the only natural
process that oxidizes water to oxygen. Another reaction of oxygen
atoms is
the formation of ozone (O3), which occurs naturally
when mediated by absorption of ultraviolet light, or can be manufactured
artificially
using high electrical voltage. Ozone is highly unstable and is
normally confined to the upper atmosphere. Here, there is sufficient
intensity of ultraviolet light to split the stable oxygen molecules,
freeing oxygen atoms and promoting recombination with other molecules
to form ozone. Substantial qualities of ozone are increasingly
being found in areas where air quality is degraded. Ground-level
ozone is formed by the reaction of byproducts from fossil fuel
combustion (hydrocarbons and nitrogen oxides) in the presence
of sunlight.
The double bonded, two-atom molecule is the single form of oxygen
which has relevance to this discussion. Air contains approximately
20.9 percent oxygen gas by volume; however, the proportion of
dissolved oxygen in air dissolved in water is about 35 percent,
because nitrogen (the remainder) is less soluble in water. Oxygen
is considered to be moderately soluble in water and this solubility
is governed by a complex set of physical conditions that include
atmospheric and hydrostatic pressure, turbulence, temperature
and salinity (Wetzel, 1983). A brief description follows of how
these conditions relate to and influence dissolved oxygen.
Atmospheric and hydrostatic pressure-Henry's
Law states that the amount of oxygen which will remain dissolved
in a volume
of water, at constant temperature, is proportional to the ambient
pressure of oxygen gas with which it is in equilibrium (CREM,
1987). Air pressure at sea level under standard conditions (fully
saturated with oxygen and water vapour, 0 degrees Celsius) is
equal to 760 mm Hg (or 101.325 kilopascals) and the proportion
of this pressure attributable to oxygen is directly related to
the fraction of oxygen in air. Oxygen tension or partial pressure
(PO2) is equivalent to atmospheric pressure minus
a compensation factor for water vapour pressure (the latter is
available in
tables as in Colt, 1984), multiplied by the oxygen fraction in
air:
PO2 = (Atmos. Press. - Water Vapour
Press.) x % O2
Davis (1975) presented the following example at 10 degrees Celsius
and one atmosphere (sea level):
PO2 = (760 mm Hg - *9.2 mm Hg) x 20.95/100
= 157.3 mm Hg
* saturated water vapour pressure at 10 degrees Celsius (from
Table 1)
Thus, at any given barometric pressure and temperature (and
corresponding water vapour pressure) the oxygen partial pressure
can be calculated. At altitudes above sea level the gravitational
attraction of gas molecules becomes less and there is a progressive
reduction in barometric pressure. Tables are available (e.g.
Table 2 from NCASI, 1985) which provide correction factors for
computations of oxygen partial pressure at altitude.
For criteria purposes it is more common to
express oxygen content in terms of concentration rather than
partial pressure. This concentration usually is represented by
solubility in mg/L (ppm) or mL/L and these units have corresponding
pressure equivalents. In the previous example with freshwater
at 10 degrees Celsius and 157.3 mm Hg, the air-equilibrated solubility
is 7.90 mL/L or 11.29 mg/L from solubility tables (e.g., Table
3 from APHA, 1992-column one for freshwater). If this sample
was 50 percent saturated, the concentration and pressure equivalents
would simply be halved (Davis, 1975). The oxygen solubility values
in Table 3 represent full (100 percent) saturation of oxygen
under one set of conditions.
For barometric pressures other than 760 mm Hg (sea level), oxygen
solubilities can be computed from the following equation:
C* = C*760(Pt - p) / (760
- p) (from
Colt, 1984)
C* = oxygen solubility
C*760 = saturation value at 760 mm Hg (Table 3)
Pt= barometric pressure (mm Hg)
p = vapour pressure of water (Table 1)
Example: Give the oxygen
solubility at 15 degrees Celsius when the barometric pressure
is 29.33 in mm Hg.
Pt = 29.33 in Hg (25.4 mm/in) =
745 mm Hg
p = 12.79 mm Hg (form Table 1)
C* = 10.08 mg/L(745 mm Hg - 12.79 mm Hg) / (760 mm Hg - 12.79)
= 9.88 mg/L
Table 1. Vapour Pressure of Freshwater in mm Hg as a Function
of Temperature
Temp.C
|
0.0
|
0.1
|
|
0.3
|
0.4
|
0.5
|
0.6
|
0.7
|
0.8
|
0.9
|
0
|
4.58
|
4.62
|
4.65
|
4.68
|
4.72
|
4.75
|
4.79
|
4.82
|
4.86
|
4.89
|
1
|
4.93
|
4.96
|
5.00
|
5.04
|
5.07
|
5.11
|
5.14
|
5.18
|
5.22
|
5.26
|
2
|
5.29
|
5.33
|
5.37
|
5.41
|
5.45
|
5.49
|
5.53
|
5.57
|
5.60
|
5.64
|
3
|
5.68
|
5.73
|
5.77
|
5.81
|
5.85
|
5.89
|
5.93
|
5.97
|
6.02
|
6.06
|
4
|
6.10
|
6.14
|
6.19
|
6.23
|
6.27
|
6.32
|
6.36
|
6.41
|
6.45
|
6.50
|
5
|
6.54
|
6.59
|
6.64
|
6.68
|
6.73
|
6.78
|
6.82
|
6.87
|
6.92
|
6.97
|
6
|
7.01
|
7.06
|
7.11
|
7.16
|
7.21
|
7.26
|
7.31
|
7.36
|
7.41
|
7.46
|
7
|
7.51
|
7.57
|
7.62
|
7.67
|
7.72
|
7.78
|
7.83
|
7.88
|
7.94
|
7.99
|
8
|
8.05
|
8.10
|
8.16
|
8.21
|
8.27
|
8.32
|
8.38
|
8.44
|
8.49
|
8.55
|
9
|
8.61
|
8.67
|
8.73
|
8.79
|
8.85
|
8.91
|
8.97
|
9.03
|
9.09
|
9.15
|
10
|
9.21
|
9.27
|
9.33
|
9.40
|
9.46
|
9.52
|
9.59
|
9.65
|
9.72
|
9.78
|
11
|
9.85
|
9.91
|
9.98
|
10.04
|
10.11
|
10.18
|
10.24
|
10.31
|
10.38
|
10.45
|
12
|
10.52
|
10.59
|
10.66
|
10.73
|
10.80
|
10.87
|
10.94
|
11.01
|
11.09
|
11.16
|
13
|
11.23
|
11.31
|
11.38
|
11.46
|
11.53
|
11.61
|
11.68
|
11.76
|
11.83
|
11.91
|
14
|
11.99
|
12.07
|
12.15
|
12.23
|
12.30
|
12.38
|
12.46
|
12.55
|
12.63
|
12.71
|
15
|
12.79
|
12.87
|
12.96
|
13.04
|
13.12
|
13.21
|
13.29
|
13.38
|
13.46
|
13.55
|
16
|
13.64
|
13.73
|
13.81
|
13.90
|
13.99
|
14.08
|
14.17
|
14.26
|
14.35
|
14.44
|
17
|
14.53
|
14.63
|
14.72
|
14.81
|
14.91
|
15.00
|
15.10
|
15.19
|
15.29
|
15.38
|
18
|
15.48
|
15.58
|
15.68
|
15.78
|
15.88
|
15.97
|
16.08
|
16.18
|
16.28
|
16.38
|
19
|
16.48
|
16.59
|
16.69
|
16.79
|
16.90
|
17.00
|
17.11
|
17.22
|
17.32
|
17.43
|
20
|
17.54
|
17.65
|
17.76
|
17.87
|
17.98
|
18.09
|
18.20
|
18.31
|
18.43
|
18.54
|
21
|
18.66
|
18.77
|
18.89
|
19.00
|
19.12
|
19.24
|
19.36
|
19.47
|
19.59
|
19.71
|
22
|
19.83
|
19.96
|
20.08
|
20.20
|
20.32
|
20.45
|
20.57
|
20.70
|
20.82
|
20.95
|
23
|
21.08
|
21.20
|
21.33
|
21.46
|
21.59
|
21.72
|
21.85
|
21.99
|
22.12
|
22.25
|
24
|
22.39
|
22.52
|
22.66
|
22.79
|
22.93
|
23.07
|
23.21
|
23.34
|
23.48
|
23.63
|
25
|
23.77
|
23.91
|
24.05
|
24.19
|
24.34
|
24.48
|
24.63
|
24.78
|
24.962
|
25.07
|
26
|
25.22
|
25.37
|
25.52
|
25.67
|
25.82
|
25.98
|
25.13
|
26.28
|
26.44
|
26.59
|
27
|
26.75
|
26.91
|
27.07
|
27.23
|
27.39
|
27.55
|
27.71
|
27.87
|
28.03
|
28.20
|
28
|
28.36
|
28.53
|
28.69
|
28.86
|
29.03
|
29.20
|
29.37
|
29.54
|
29.71
|
29.88
|
29
|
30.06
|
30.23
|
30.41
|
30.58
|
30.76
|
30.94
|
31.12
|
31.30
|
31.48
|
31.66
|
30
|
34.84
|
32.02
|
32.21
|
32.39
|
32.58
|
32.77
|
32.95
|
33.14
|
33.33
|
33.52
|
31
|
33.71
|
33.91
|
34.10
|
34.29
|
34.49
|
34.69
|
34.88
|
35.08
|
35.28
|
35.48
|
32
|
35.68
|
35.89
|
36.09
|
36.29
|
36.50
|
36.70
|
36.991
|
37.12
|
37.33
|
37.54
|
33
|
37.75
|
37.96
|
38.18
|
38.39
|
38.61
|
38.82
|
39.04
|
39.26
|
39.48
|
39.70
|
34
|
39.92
|
40.14
|
40.37
|
40.59
|
40.82
|
41.05
|
41.28
|
41.51
|
41.74
|
41.97
|
35
|
42.20
|
42.43
|
42.67
|
42.91
|
43.14
|
43.38
|
43.62
|
43.86
|
44.10
|
44.35
|
36
|
44.59
|
44.84
|
45.08
|
45.33
|
45.58
|
45.83
|
46.08
|
46.33
|
46.59
|
46.84
|
37
|
47.10
|
47.35
|
47.61
|
47.87
|
48.13
|
48.40
|
48.66
|
48.92
|
49.19
|
49.46
|
38
|
49.72
|
49.99
|
50.27
|
50.54
|
50.81
|
51.09
|
51.36
|
51.64
|
51.92
|
52.20
|
39
|
52.48
|
52.76
|
53.04
|
53.33
|
53.62
|
53.90
|
54.19
|
54.48
|
54.78
|
55.07
|
40
|
55.36
|
55.66
|
55.96
|
56.25
|
56.55
|
56.86
|
57.16
|
57.46
|
57.77
|
58.07
|
Tabulated oxygen saturation values are available
as a function of barometric pressure and elevation over a range
of temperatures (e.g., as in Colt, 1984). The correction factors,
listed in Table 2, also can be applied directly to oxygen solubilities
at the range of elevations shown (this inverse relationship is
linear). For non-standard pressures and elevations, Wetzel (1983)
provides a nomogram from which oxygen solubility and percent
saturation can be derived at an observed temperature and oxygen
content. In special cases, when the composition of dissolved
gases under study differs from that of air, Bunsen coefficients
can be used to calculate solubilities (mole fractions of gases
must be known); Colt (1984) provides the necessary formulae and
tables for these calculations.
At any particular depth in a column of water, the amount of
gas that can be held in solution is determined by the combined
atmospheric and hydrostatic pressures, and is known as the absolute
saturation (Wetzel, 1983). Normally, saturation is considered
in relation to the pressure at the water's surface, at a specific
temperature and salinity. Supersaturation, a non-equilibrium
situation, is the term used when the partial pressures of gasses
(primarily nitrogen and oxygen) in solution exceed their equivalent
atmospheric pressures. Hydrostatic pressure increases rapidly
with depth and dissolved gas solubility doubles approximately
every 10 m (hence, the increased efficiency of aeration devices
at depth), while the degree of supersaturation decreases with
depth (Colt, 1984). For example, a gas supersaturation of 130
percent (surface measurement) is reduced to 100 percent saturation
at a depth of 3.0 m.
Table 2. Correction Factors for Barometric Pressure and Oxygen
Saturation at Altitude
ALTITUDE
|
CORRECTION
|
|
(feet)
|
(metres)
|
FACTOR
|
0
|
0
|
1.00
|
500
|
152
|
0.98
|
1000
|
305
|
0.96
|
1500
|
457
|
0.95
|
2000
|
610
|
0.93
|
2500
|
762
|
0.91
|
3000
|
914
|
0.89
|
3500
|
1067
|
0.88
|
4000
|
1219
|
0.86
|
4500
|
1372
|
0.84
|
5000
|
1524
|
0.82
|
5500
|
1676
|
0.81
|
6000
|
1829
|
0.80
|
Notes:
1. Multiply barometric pressure of dissolved oxygen solubility
at sea level for the appropriate temperature (Table 3) by the
correction
factor for your altitude.
2. Interpolate, using linear relationship, for greater accuracy.Source:
NCASI, 1985.
Table 3. Solubility of Oxygen in Water (Fresh and Saline) Exposed
to Water-saturated Air
at Sea Level (760 mm Hg (101.3 kPa)
| |
Oxygen
Solubility (mg/L)
|
|
Temp.
|
Chlorinity
(freshwater)
|
|
(C)
|
0
|
5.0
|
10.0
|
15.0
|
20.0
|
25.0
|
0.0
|
14.621
|
13.728
|
12.888
|
12.097
|
11.355
|
10.657
|
1.0
|
14.216
|
13.356
|
12.545
|
11.783
|
11.066
|
10.392
|
2.0
|
13.829
|
13.000
|
12.218
|
11.483
|
10.790
|
10.139
|
3.0
|
13.460
|
12.660
|
11.906
|
11.195
|
10.526
|
9.897
|
4.0
|
13.107
|
12.335
|
11.607
|
10.920
|
10.273
|
9.664
|
5.0
|
12.770
|
12.024
|
11.320
|
10.656
|
10.031
|
9.441
|
6.0
|
12.447
|
11.727
|
11.046
|
10.404
|
9.799
|
9.228
|
7.0
|
12.139
|
11.442
|
11.783
|
10.162
|
9.576
|
9.023
|
8.0
|
11.843
|
11.169
|
10.531
|
9.930
|
9.362
|
8.826
|
9.0
|
11.559
|
10.907
|
10.290
|
9.707
|
9.156
|
8.636
|
10.0
|
11.288
|
10.656
|
10.058
|
9.493
|
8.959
|
8.454
|
11.0
|
11.027
|
10.415
|
9.835
|
9.287
|
8.769
|
8.279
|
12.0
|
10.777
|
10.183
|
9.621
|
9.089
|
8.586
|
8.111
|
13.0
|
10.537
|
9.961
|
9.416
|
8.899
|
8.411
|
7.949
|
14.0
|
10.306
|
9.747
|
9.218
|
8.716
|
8.242
|
7.792
|
15.0
|
10.084
|
9.541
|
9.027
|
8.540
|
8.079
|
7.642
|
16.0
|
9.870
|
9.344
|
8.844
|
8.370
|
7.922
|
7.496
|
17.0
|
9.665
|
9.153
|
8.667
|
8.207
|
7.770
|
7.356
|
18.0
|
9.467
|
8.969
|
8.497
|
8.049
|
7.624
|
7.221
|
19.0
|
9.276
|
8.792
|
8.333
|
7.896
|
7.483
|
7.090
|
20.0
|
9.092
|
8.621
|
8.174
|
7.749
|
7.346
|
6.934
|
21.0
|
8.915
|
8.456
|
8.021
|
7.607
|
7.214
|
6.842
|
22.0
|
8.743
|
8.297
|
7.873
|
7.470
|
7.087
|
6.723
|
23.0
|
8.578
|
8.143
|
7.730
|
7.337
|
6.963
|
6.609
|
24.0
|
8.418
|
7.994
|
7.591
|
7.208
|
6.844
|
6.498
|
25.0
|
8.263
|
7.850
|
7.457
|
7.083
|
6.728
|
6.390
|
26.0
|
8.113
|
7.711
|
7.327
|
6.962
|
6.615
|
6.285
|
27.0
|
7.968
|
7.575
|
7.201
|
6.845
|
6.506
|
6.184
|
28.0
|
7.827
|
7.444
|
7.079
|
6.731
|
6.400
|
6.085
|
29.0
|
7.691
|
7.317
|
6.961
|
6.621
|
6.297
|
5.990
|
30.0
|
7.559
|
7.194
|
6.845
|
6.513
|
6.197
|
5.896
|
Notes:
1. Formulae are available for equilibrium oxygen concentration
at non-standard pressures and for all chlorinity values.
2. For wastewater, it is necessary to know the ions responsible
for the solution's electrical conductivity to correct for their
effect on oxygen solubility and use of the tabular value.
Source: APHA, 1992
The degree of oxygen supersaturation necessary for bubble
growth increases with depth. Ramsey (1962) explained that,
in the absence of turbulence, bubbles may form due to the partial
pressure of oxygen at depths of less than one metre. Below
four metres, oxygen will be maintained in solution by hydrostatic
pressure even when extremely supersaturated relative to the
pressure at the surface.
The entrainment of air below water falls or dam
spillways is a common cause of supersaturation that first came
to prominence as an environmental problem in the Pacific Northwest
in the Columbia River system (primarily in Washington State,
but also below the Hugh Keenleyside Dam near Castlegar). Gas
bubbles can develop in fish and invertebrates due to a large
imbalance between ambient and internal partial pressures, and
lethal or sublethal effects can result. Since oxygen usually
is not the principal gas of importance (nitrogen is) and total
gas pressure is more central to the issue of supersaturation,
gas bubble disease is dealt with in a separate criteria document
on total dissolved gases. The effect of hydrostatic pressure
also must be taken into account when measuring oxygen concentrations
at great depths by an electrode as opposed to a chemical technique.
Oxygen solubility remains effectively constant with depth whereas
partial pressure increases; therefore, a polarographic probe
which measures partial pressure rather than concentration must
be corrected accordingly. For example, at 400 m a correction
of 5 percent (less) is necessary (Hitchman, 1978).
Turbulence -
the diffusion of gas in water is slow and, for equilibrium
with
atmospheric oxygen to be established, circulation
must occur such as in the epilimnion of stratified lakes or
at periods of turnover. The rate of oxygen distribution and
equilibration is dependent on turbulence. Increased turbulence
forms a greater surface area from which excess gasses from
supersaturation can dissipate, and brings trapped subsurface
water to the surface (NCASI, 1985). In cases where the initial
dissolved oxygen concentrations at depth are not far from saturation,
equilibrium may occur in a few days. Alternatively, in deep
lakes complete oxygenation may never be achieved before thermal
stratification terminates circulation for a seasonal interval
(Wetzel, 1983). Oxygen distribution will be discussed further
in Section 3.1.
Temperature -Temperature,
more than any other physical condition in the aquatic environment,
affects the solubility potential
of dissolved oxygen. This relationship is non-linear as solubility
increases considerably in cold water (Wetzel, 1983). Freshwater
is saturated with 14.6 mg O2/L at 0 degrees Celsius, which
declines to 8.3 mg O2/L at 25 degrees Celsius (at sea level).
As solubility declines with increased water temperature, Davis
(1975) points out that oxygen partial pressure drops only slightly
due to increased molecular activity.
Oxygen solubility tables for a range of temperatures are available
from a number of sources; however, references prior to 1981
should be avoided due to updating of these solubilities. Table
3 was extracted from a larger table in APHA (1992) which lists
solubility values for dissolved oxygen in freshwater and saline
waters, equilibrated with air at one atmosphere (sea level).
Salinity -The
oxygen content of water decreases exponentially as salinity
increases, such that the difference between solubility
in seawater and freshwater is about 20 percent (Wetzel, 1983).
Tables (e.g., Table 3) and nomograms (e.g. Figure 1) are available
for deriving oxygen saturation in saline waters. The new definition
of salinity, which was adopted by the Standard Methods Committee
in 1985, is based on the electrical conductivity of seawater.
Specific conductance is converted against a known standard
(KCl in water) to chlorinity and then to total salinity by
a correction factor:
salinity = 1.80655 x chlorinity
The scale has no dimensions,
therefore parts per thousand (g/kg) no longer applies (APHA,
1989).
Figure 1. Nomogram of Oxygen Solubility in Air-saturated Water
at Different Salinities
Source: Hitchman,
1978
2.2 Analytical Methods
2.2.1 Surface Water
There are two common methods
for determining the solubility of oxygen in water: the Winkler
or iodometric method and its modifications, and the electrometric
method using membrane electrodes. The precision of other chemical
and colorimetric methods is invariably less than that for the
Winkler method (Hitchman, 1978). The Winkler method involves
the more precise titrimetric procedure based on the oxidizing
property of dissolved oxygen, while the membrane electrode
procedure is based on the rate of diffusion of molecular oxygen
across a membrane (APHA, 1992). Since the amount of oxygen
in water is dependent upon a complex set of physical properties
and biological processes, the method of measurement must be
suited to the source water. Temperature, salinity, turbulence,
pressure, photosynthetic activity, respiration and chemical
interferences (oxidizing or reducing compounds) affect the
concentration of dissolved oxygen in water.
Iodometric Procedures -
APHA (1992) describes four derivations of the Winkler method,
the selection of which is based on minimizing
the effects of interfering materials known to be present. For
example, the azide modification for most effluent and stream
measurements removes interferences caused by nitrite, the most
common interference in biologically-treated effluents. Zenon
Environmental Laboratories uses this method (reagents include
manganese sulphide, potassium salt and sulphuric acid) for
calibrating oxygen meters. A determination of 0.05 mg/L is
possible, which can ensure a meter accuracy of 0.1 mg/L (Heier,
1991). The other procedures described in APHA (1992) include
the permanganate modification used for samples containing ferric
and ferrous iron (e.g., acid mine drainage), the alum flocculation
modification which removes interferences from high suspended
solids, and the copper sulphate-sulfamic acid flocculation
modification for biological flocs (e.g., activated sludge)
which have high oxygen utilization rates. Further modifications
are available for the following: Pomeroy-Kirschman method when
high dissolved oxygen levels (> 15 mg/L) or high organic
content are present, Alkali-hypochlorite modification in the
presence of SO32-, S2 O32-, and polythionate, and the "Short" modification
for organic substances which are readily oxidized in strong
alkali or by the iodine in acid solution (Hitchman, 1978).
A major disadvantage of the above methods is that they are
not appropriate for in situ measurements. Samples should be
handled carefully to avoid agitation and contact with air,
and special equipment is necessary to eliminate changes in
pressure and temperature when sampling at depth. It is commonly
acknowledged that dissolved oxygen is best measured in the
field because of the changes in concentration that are likely
to occur between sampling and lab analysis. In some instances,
fixative agents (including sulphuric acid, sodium azide) can
be used by collectors to stabilize samples for transit to a
lab, but these chemicals are costly and extremely corrosive
and accuracy of the dissolved oxygen determination still would
be questionable. Equipment for measuring oxygen levels in the
field is described in the following sections.
Electrometric Procedures - Early oxygen sensors had to be
designed for each analytical situation, and electrodes used
were subject to direct exposure to the sample medium. The most
significant development in the design of efficient sensors
was achieved by Dr. Leland Clark whose membrane-covered electrode
reduced the risk of contamination and provided a more uniform
diffusion layer for oxygen to pass. Present generation meters
are a convenient size, simple to operate and reasonably rugged.
The submersible electrodes are particularly useful for continuous
monitoring, profiling dissolved oxygen with depth and testing
waters which have high interference values (effluents, particulates,
colour, etc.). Zenon Laboratories, for example, uses an Orion
meter for conducting continuous dissolved oxygen analyses for
biochemical oxygen demand (Heier, 1991). Some of the newest
models incorporate computerized remote control and interfacing
to download data (e.g., YSI Model 59). There are three variations
of membrane-covered probes commonly in use, each having specific
attributes. Figure 2 contains schematic diagrams and probe
reactions as examples of galvanic, polarographic and oxygen
balance sensors.
The galvanic sensor is self-polarizing and produces its own
electric current. A lead or silver anode and a silver cathode
reside within a potassium hydroxide electrolyte, and galvanic
potential is produced by the reduction of oxygen at the cathode.
The current generated is proportional to the rate of oxygen
diffusion through the membrane (which is dependent on the concentration
of molecular oxygen) (YSI, 1989).
The most common sensor is the polarographic probe, which employs
a silver anode and gold cathode in a potassium chloride electrolyte.
When voltage is applied, oxygen accepts electrons from the
cathode. For each molecule that is reduced, a proportional
current is registered that is converted to oxygen content.
A newer and more sophisticated
system is employed in oxygen balance sensors,
which were designed to address some of the shortcomings of
the previous probes.
Three electrodes (or more) operate in a potassium hydroxide
electrolyte. Oxygen still defuses through a membrane and is
reduced at the cathode(s); however, an equal quantity is generated
at the anode(s). This diffusion continues until the oxygen tension
is balanced on either side of the membrane, and the current
necessary to maintain this balance is converted to a read-out
of oxygen partial pressure (YSI, 1989).
Figure
2. Oxygen Sensors
Benefits / Drawbacks
- rugged - probe
consumes oxygen
- high current output facilitates long-term (water flow is
necessary) monitoring-electrode is consumed over time
- no warm-up required-membrane should be changed periodically
Benefits / Drawbacks
- Teflon membrane is easily-probe
consumes oxygen
changed in the field water flow is necessary
- requires several minutes to equilibrate and give a steady
read-out
Benefits / Drawbacks
- fast response-relatively
expensive
-no electrolyte/electrode consumption-if
membrane is fouled or damaged,
-membrane may be permanent type sensor must be replaced and
-accuracy is not dependent on water flow instrument recalibrated
since little if any oxygen is consumed
Source (Figures): YSI, 1989
All of the these sensors are
susceptible to various physical conditions which affect the
diffusion rate of oxygen through membranes. These influences
are (roughly in order of decreasing importance): temperature,
water flow, membrane fouling, salinity and barometric pressure.
With the exception of contamination, oxygen meters have the
compensation circuitry (manual or automatic) necessary to mitigate
these influences. Temperature is considered to have the most
significant affect on membrane permeability. APHA (1992) recommends
that temperature sensitivity be checked regularly against the
original calibration. A nomograph for temperature correction
is usually supplied with the instrument or one can be constructed.
Some meters compensate automatically for temperature using
thermistors; however, their accuracy over a wide temperature
range has been questionned (APHA, 1992). In YSI probes, the
temperature effect on molecular activity causes a three percent
change in diffusion rate for every degree Celsius change, even
though the oxygen pressure is constant. A temperature-sensitive
thermistor corrects this differential. An additional thermistor
is usually present to compensate for the varying solubility
of oxygen in water when measurements in mg/L are taken (i.e.,
oxygen content rather than partial pressure or percent saturation).
Water flow, such as created by stirring, is particularly important
for galvanic and polarographic probes which have oxygen-consumptive
reactions that can create a layer of depleted oxygen next to
the membrane. These probes are either fitted with stirrers
or must be moved through the water column at a minimum specified
rate if in static water. Salinity correction also may be necessary
to reflect the decline in oxygen carrying capacity with increased
salinity. Usually, salinity must be measured by the user and
is then manually adjusted on the meter. Finally, instruments
may be equipped with automatic barometric pressure compensation,
or a tabulated correction factor (Table 2) may be determined
and the meter calibrated accordingly after an oxygen reading
is taken.
Specific calibration procedures have been developed by manufacturers
and it is recommended that these be followed prior to each
daily sampling routine. The general rule is to calibrate an
oxygen probe under conditions most similar to the water being
sampled - preferably in the sample water itself. However, in
freshwater containing contaminants, calibration should be done
in distilled water. In saltwater, calibration can be done in
the water to be tested. Again, if pollutants are present, it
is necessary to use clean saltwater or water having the same
salt content / specific conductance (can be adjusted by adding
potassium chloride). In estuarine water or water with variable
ionic content, the sample chlorinity must be determined to
allow revision of the original calibration value taken in clear
water. Gasses such as hydrogen sulphide, sulphur dioxide and
carbon monoxide also will contaminate an oxygen sensor. Membranes
should be changed and meters calibrated frequently when the
presence of such gasses is suspected (APHA, 1992).
Manufacturers commonly present more than one method of adjusting
the oxygen read-out of their meter to a sample of known oxygen
content. The following calibration options are described for
YSI equipment, but can be considered standard methods for most
meters.
Winkler titration - A water
sample is subdivided into four parts, three of which are titrated
and the results
averaged. If one
of the values differs from the other two by more than 0.5 mg/L,
only the remaining two are averaged. A probe is placed in the
fourth sample for three to five minutes to reach thermal equilibrium
and then stirred at least 30 seconds before a reading is taken.
The reading is adjusted to the titration average. This relatively
complex procedure is accurate, but often impractical in the
field and is applicable only to freshwater with no interfering
ions.
Air-saturated water - A
sample of water (usually distilled) is aerated or stirred for
approximately 15 minutes
to saturation. The water temperature is measured and a solubility
table consulted for the appropriate oxygen content (correction
for atmospheric pressure or altitude may be necessary). A reading
is then taken with the probe and the meter adjusted to the
known tabulated value.
Water-saturated air - Air
calibration usually is the preferred procedure because of its
simplicity and reliability. Air-saturated
water and water-saturated air at sea level both have an oxygen
partial pressure of 160 mm Hg. However, there is less certainly
of the former being 100 percent saturated, while air is by
definition air-saturated. To achieve water-saturated air, the
probe can be placed in a bottomless container with a wet blotter
or a specially made calibration chamber with a few drops of
water. YSI's own calibration chamber has a long handle which
allows the sealed probe to be incubated underwater to insure
proper thermal equilibrium in the field where air/water temperature
differentials can be considerable (YSI, 1989).
2.2.2 Interstitial Water
The most long-standing technique for measuring the dissolved
oxygen content of sediment water in spawning media has involved
the use of standpipes. A standpipe is a length of pointed pipe
(usually steel or plastic) that is driven into the bottom sediments.
Holes drilled in the lower end accept sub-surface water only,
since the top of the pipe projects above the water surface.
In 1954, Wickett developed a standpipe apparatus (subsequently
referred to as the `Mark I') and procedures for calculating
interstitial oxygen content, which have persisted in modified
form to the present. His procedure was to drive the sampler
to a standard depth (e.g., 30 cm so that the perforations could
be at the egg deposition level of 25 cm), pour sand around
the pipe to reduce the exchange of surface water next to the
pipe, draw off the water within the pipe several times using
suction prior to taking a sample, use the Winkler titration
method of analysis, and measure the temperature of the pore
water. Terhune (1958) developed a `Mark VI' model, primarily
to improve the accuracy of measuring permeability using a dye
dilution rate technique. He reported consistent results in
determining oxygen content within five percent, which he considered
satisfactory in view of the natural variability that could
be as high as 100 percent in the same redd.
McNeil (1962) focused on improving field measurement accuracy
of dissolved oxygen concentration. He described detailed procedures
for the fixation and handling of sample water to improve precision,
which will not be reproduced here. Two necessary precautions
he advised were: 1.) leave the standpipe in the stream for
at least 24 hours before sampling to allow conditions to stabilize,
and 2.) remove only small water samples (about 30 mL). With
respect to the latter, the author showed that variability can
be introduced with relatively large withdrawals. If the sub-surface
water originated from highly oxygenated stream water at points
high in dissolved oxygen, replicate samples had higher readings,
wheras lower readings were found for second samples taken at
points having low oxygen values due to poorly oxygenated ground
water sources. With the development of accurate and reliable
dissolved oxygen meters, standpipes also can be used in conjunction
with a remote probe (preferably the non-consumptive type) to
avoid the possibility of oxygenating sample water during handling
(e.g., as in Woods, 1980). In Sowden and Power's (1985) study
of rainbow trout survival in a ground water-fed stream (reported
in Section 4.3.2.1), mini-piezometers used for measuring pressure
head also functioned as standpipes from which samples were
pumped out and analyzed with a polarographic probe. In Scrivener
and Brownlee's long-term study of forest harvesting effects
on Carnation Creek (1989), interstitial water was simply withdrawn
by stainless steel syringe from a depth of 20 cm and analysis
done by Hach Kit (reported accuracy of 0.1 mg/L).
In deeper water, the usual standpipe method has obvious limitations.
Thompson and Heimer (1967) developed a simple and inexpensive
method that utilized a thin (1 cm outside diameter) metal probe
perforated at one end and attached to rubber tubing at the
other (length adapted to water depth). A 20 cm collar mid-way
along the metal tube functioned in similar fashion to that
used with the `Mark VI' standpipe, to keep surface flow from
travelling down along the outside of the probe. Five millilitre
samples were withdrawn through the side of the rubber tubing
by syringe and analyzed by a modified micro-Winkler syringe
technique. Analyses of dissolved oxygen content in interstitial
water in lake and marine sediments are less commonly done and
necessarily are more complex. Brinkman et al. (1982) reviewed
three existing techniques for collecting pore water (coring,
dialysis and direct suction) and identified problems with oxidation,
disturbance and suspension of sediments. They decided to design
their own apparatus for use in shallow lakes, and their paper
details a frame with attached water sampling probe(s) which
can be pushed into the sediment. The investigators reported
that ambient characteristics were largely maintained, particularly
with respect to exclusion of oxidation effects.
