In the Classroom
Teaching Estuarine Chemical Processes by Laboratory Simulation T. Ortega,* J. M. Forja, and A. Gómez-Parra Departamento de Química Física, Facultad de Ciencias del Mar, Universidad de Cádiz, Campus Río San Pedro s/n, 11510 Puerto Real, Cádiz, Spain; *
[email protected] and substances, regulating their residence time in the ocean Courses in the chemistry of marine and environmental (6 ). Logistically, however, sampling in estuaries requires sciences at the higher education level require the student to facilities and resources that are often not available to teachers, perform many laboratory experiments. It is also essential that such as for the transport of numbers of students to places that the experimental work includes field practice. are often distant from schools and colleges. Generally, learning activities related to environmental The objectives of the experiments described here are (i) chemistry or chemical oceanography are difficult to carry out, to reproduce the longitudinal gradient in the stationary state for several reasons: (i) they are expensive to organize; (ii) they of a given estuary; (ii) to determine whether the behavior of require a lot of time and the participation of many teachers; various chemical species is conservative or nonconservative (iii) they are difficult to coordinate with the rest of the teaching during their passage through the estuary; (iii) for those species activities during the period of the course; (iv) they are very presenting nonconservative behavior, to measure the losses or dependent on weather conditions. However, without denying gains undergone by the dissolved phase as the ionic strength the need to work on site for this kind of study, it is possible of the medium varies; and (iv) to establish the dependence of to significantly reduce the amount of time spent on field the solubility of oxygen on the salinity and temperature. The experimentation by employing laboratory simulations. This simulation is carried out by means of stationary gradients of offers additional advantages, since under conditions that are salinity. It allows us to study the reactivity of chemical species carefully controlled in the laboratory, it is possible to reproinvolved in the water mixing processes and to characterize the duce certain processes that occur only sporadically or unprebehavior of chemical species in the stationary state. We did dictably in the field, and it is possible to shorten the time not consider tidal influences. span of very slow natural processes, such as those subject to seasonal variation. This article presents a device that simulates the process of The Estuary Simulator estuarine mixing and the behavior of numerous substances as Figure 1 shows a diagram of the simulator. The simulation they pass through an estuary. The simulator is similar to that of the estuarine mixing is achieved by the countercurrent described by Bale and Morris (1), although its dimensions mixing of sea water with river water. The mixing is performed and the features offered are considerably greater. Since its in a series of eight tanks situated at ascending levels, the water design in 1994 it has been used in the practical teaching of more of greatest salinity being at the lowest level. The lowest tank than 800 students of Chemical Oceanography in the Faculty of Marine Sciences of the University of Cádiz. The results obtained from this empirical trial are compiled in this paper. QR Q2 There are several reasons why the simulation of chemical processes in estuaries is interesting. Estuaries Q3 Q'1 constitute the parts of the ocean 1 River Water where the greatest salinity gradients are found. Although the current Qi Q'2 2 definition of salinity produces an adimensional number (2), in this paper the traditional definition was considered: salinity is a measure of Q'i i Qn the mass in grams of dissolved salts in a given kilogram of seawater (‰) Qi+1 Q'i-1 (3). Changes in salinity within a Qi Q SW relatively few kilometers produce a n - 1 Q'n-1 variation in the ionic strength from Q'i 0.01 to 0.7 M, which causes a n Q'n i strong level of reactivity for many chemical substances transported by rivers; in many cases, this involves their passing from the dissolved to the particulate phase, or vice versa Waste Sea Water (4, 5). In this way, estuaries act as a chemical barrier for many elements Figure 1. Diagram of estuary simulator. JChemEd.chem.wisc.edu • Vol. 78 No. 6 June 2001 • Journal of Chemical Education
771
In the Classroom
(1)
Q i+1 × Si+1 + Q ′i–1 × Si –1 = (Qi + Q ′i) × Si
(2)
in which, for i = 1, Q′i–1 = QR and Qi–1 = 0; for i = 8, Q i+1 = QSW. Implicitly, it is assumed that the estuary is in a stationary state [(dS/dt)x = 0)] (8). To facilitate the calculation, the density has been considered equal to one for all tanks. By choosing convenient relationships between the flows Qi , Q SW, and QR, it is possible to generate a large number of longitudinal gradients of salinity, and by using the function describing the relationship between QR and the volume of the tanks, residence times similar to those of real estuaries can be obtained. Conceptually, the simulation of the process of mixing in the estuary consists of substituting the continuous variation in salinity that is produced in a real estuary by a series of stages in which the salinity is constant and between which there exists a sharp variation in the salinity. The degree of approximation to the real process of mixing depends, therefore, on the number of tanks used or on the difference in the salinity of the water added to the first and last tanks, which may be different from SR and SSW. In this case, the device would be simulating the mixing process in only one part of the estuary. In the experiments described here, the river water used is from a river flowing into the Bay of Cádiz (SW Spain) and the sea water is taken from a well with a salinity of 36.324 (‰). The tanks were aerated to ensure the saturation of oxygen, and approximately 4 L of sediment was placed in each; this sediment had been washed with water of the same salinity as that the tank was going to hold. The silicon tubes of the peristaltic pump were changed daily and the flow rates QR, QSW , and Q i were also checked daily. Figure 2 shows the evolution of the salinity gradient of the estuary obtained as forecast by eqs 1 and 2. The operating conditions are shown in Table 1. Behavior of Chemical Species Passing through the Simulated Estuary After a period equivalent to five times the residence time of the river water in the simulator, the stationary state is reached. Then the pumps are switched off and aeration is stopped. 772
Expected Salinity 0h 4h 18 h 28 h 44 h Final Salinity (70 h)
15
10
5
0 2
1
3
4
5
6
7
8
Tank Figure 2. Evolution of salinity gradient, from start of trial.
Table 1. Operating Conditions for the Simulation Trials Variable
Value
No. of tanks
8
Vi
20 L
QR
26.1 mL min᎑1
QSW
25 mL min᎑1
Qi
25 mL min᎑1
τ1
a
7.65 h
τ8
a
3.96 h
τtotal a
2.22 h
Suspended solids 110–60 mg L᎑1 aτ
[Cl− ] / (g kg −1)
Q i+1 + Q ′i–1 = Q i + Q ′i
20
Salinity
receives sea water, at a flow rate of QSW, and the highest tank receives river water at a flow rate of QR. To each intermediate tank, i, a flow (Q i+1) of water of less salinity is pumped from the tank just below it, while from it, by overflow, water is returned to the same tank below at a flow rate of Q′i . The intermediate tank also receives the overflow from the tank next above it (Q′i –1), to which it in turn sends by pumping a flow of Q i. Each tank is agitated by a small submersible pump to ensure uniform salinity. In this way, the simulator reproduces the mixing process that would be produced in a vertically and laterally homogeneous estuary (type A in the classification of Beer [7]). To produce the ascending flows (Q i ) a multichannel peristaltic pump is used, so that all the flows are equal. The flows QSW and QR are produced using independent pumps. The relationship between the salinity in each tank and the water flows between tanks can be established by means of a general and particular material balance. This requires the solution of the following system of 16 equations (1≤ i ≤ 8):
is residence time.
8
4
0 0
5
10
15
20
Salinity Figure 3. Variation of chloride concentration in the dissolved phase with salinity.
After waiting one hour to eliminate any over-saturation of the dissolved gases, samples are taken for analysis of dissolved oxygen, some of the major elements of sea water, and nutrients. The determination of the major ions and nutrients is carried out after filtering the samples through a membrane of 0.22 µm (Millipore GSWP 04700). The analyses were conducted using the methods described by Grasshoff (9). Figure 3 shows the linear relationship between the concentration of chloride in the dissolved phase and the salinity. This conservative behavior is also observed for other major ions in the sea water (e.g. F ᎑, Ca2+, and SO42᎑). This is not surprising; it is why the salinity can be used as a conservative index of mixing.
Journal of Chemical Education • Vol. 78 No. 6 June 2001 • JChemEd.chem.wisc.edu
[Silicate] / (µmol L−1)
[PO43− ] / (µmol L−1)
In the Classroom
12
9
6 0
10
20
240
180
120
60 0
30
10
20
30
Salinity
Salinity
Figure 4. Variation of the concentration of HPO42᎑ and SiO2 in the dissolved phase with salinity in the simulation assays.
Concentration
C RW
dC dS
the difference between the flow of this substance (χ) across the two sections delimiting this stretch. Bearing in mind that this dynamic mixer simulates a vertically and laterally homogeneous estuary, and accepting that the flow F of the substance χ across a section A of the estuary, in which its concentration in the dissolved phase is C, would be given by the sum of the transport by advection and by diffusion (15): F (χ) = Q × C ± k × A × (dC/dx) where Q is the flow rate of the river, x is the longitudinal dimension of the estuary, and k is the coefficient of longitudinal diffusion. If the flow is considered positive (greater than zero) from the river toward the sea, a minus sign for the sum would imply that it refers to a substance whose concentration is higher in the river than in the sea, and vice versa. Accepting that in the stationary state the transport of salts along the estuary would be zero, it is found that
S = Si
Ci −1
Q × S – k × A × (dS/dx) = 0 Ci
where S is the salinity. Combining these last two expressions gives
Ci +1 C SW
F (χ) = Q × [C ± S(dC/dS)] SRW Si −1
Si
Si +1
SSW
Salinity Figure 5. Example of nonconservative behavior.
Table 2. Cumulative Loss of Phosphate along the Salinity Gradient Salinity Range
Phosphate Loss (%)a
0.50–0.60
16.87
0.50–0.74
17.49
0.50–1.01
21.20
0.50–1.54
26.40
0.50–2.61
31.58
0.50–4.91
33.94
0.50–9.22
35.15
0.50–7.79
42.77
aBased on quantity entering the estuary from the river.
The nutrients (inorganic salts of P, N, and Si coming from the rivers to the ocean) can have different behaviors as a function of their reactivity. The concentration of silicate varies linearly with the salinity and it presents the same pattern that is identified in several estuaries around the world (10, 11). However, the behavior of phosphate in the simulator (Fig. 4) is nonconservative. This result, which also is in agreement with those obtained in field studies (12–14), may be attributed to the adsorption of HPO42᎑ onto the precipitates of iron oxyhydroxide or to apatite formation, which occurs particularly in regions of low salinity. In a case such as this, it is simple to calculate the loss of these substances in the dissolved phase in one stretch of the estuary. This is given by
and expressed as a percentage of the total quantity of χ supplied by the river (Q × C0), F (χ) = (100/C0)[C ± S × (dC/dS)] Therefore, to calculate the proportion of the substance χ that has been lost in a stretch of the estuary, it is only necessary to know the concentration of χ and the salinity at the extremes of this stretch and the values of the derivative C = f (S ) at these salinity points. This last determination can be done either analytically or graphically from the curves, as shown in Figure 5. In many cases, it may be a valid approximation to consider as the value of (dC/dS ) for Si the mean value of the slope of the straight line joining the experimental points (Ci–1, Si –1), (Ci , Si) and (Ci+1, Si+1). By proceeding analytically, the losses of phosphate given in Table 2 may be obtained. Figure 6 shows the results obtained by teams of students who used the simulator to measure the concentration of oxygen in the different tanks. Taking into account that the sea water had a relatively constant temperature (19.3 ± 0.5 °C), the figure shows the dependence of the solubility of oxygen on the salinity given by the empirical equation of Weiss (16 ); the error intervals are also included. The students’ results for the solubility of oxygen (Fig. 6) are usually lower than the saturation data, especially in those tanks with higher salinity values. Although these results should be used with care, since the students were not experienced in analysis, they can be related to the characteristics of the sea water used (“fossil” marine water with a high concentration of ammonia). The reduction in the concentration of oxygen could have been caused by consumption of oxygen in the oxidation of NH4+ to NO3᎑. To assess the influence of nitrification on the dissolved oxygen concentration, other assays were carried out to measure the concentration of inorganic nitrogen compounds. The water used was natural sea water and river water enriched with inorganic nitrogen. To enhance the speed of nitrification, which in natural systems is slow, the residence time of the water in the simulator was increased and a bacterial inoculum from a sewage plant was added to the river water. The results ob-
JChemEd.chem.wisc.edu • Vol. 78 No. 6 June 2001 • Journal of Chemical Education
773
[NO3− ] / (µmol L−1)
275
120
80 0
10
5
15
18
12
6
0
20
0
5
225
0
10
20
30
40
Salinity Figure 6. Mean values of the concentration of dissolved oxygen obtained in the estuary simulator. Error bars represent statistical errors in the analytical method obtained by different groups of student. The lines show the variation of the solubility of oxygen with the salinity, for the working temperature range (19.3 ± 0.5 ºC).
tained for the evolution of the concentrations of NH4+, NO2᎑, NO3᎑, and oxygen percent of saturation throughout the simulator are consistent with this explanation (Fig. 7). Final Remarks The device constructed is clearly not a hydraulic simulator of the processes of mixing of the bodies of water coming together in the estuary, but it does allow the simulation of the chemical reactivity of any substance passing through the estuary in a realistic manner. In fact, the assumption that chemical reactivity does not vary with any change of scale is implicit in all chemical analysis. Further, the elimination of any effect derived from tidal transport within the estuary, since the simulator operates in the stationary state, does not affect the chemical reactivity; this only affects the geographical location within the estuary where this reactivity takes place. The experiments described here deal with the behavior of substances that are naturally present in the sea and river waters being mixed. It is, however, quite feasible to apply this simulator to investigate the behavior of other substances, particularly contaminants; these only need to be introduced to the river water, or to whichever of the tanks represents the specific zone of the estuary involved. The substances introduced and analyzed may be ones that have never before been present in the natural environment; therefore this simulator will be useful for carrying out a priori studies of environmental impact. Acknowledgments We thank M. F. Osta and P. Vidal for their assistance in the initial setup of the estuary simulator. Thanks are also due to S. Rubio for his help during the graphic design of the estuary
16 12 8 4 0 0
5
10
10
15
20
15
20
Salinity
15
20
Dissolved O2 (% satn)
Salinity
200
774
160
250
[NH4+ ] / (µmol L−1)
[O2] / (µmol L−1)
300
[NO2− ] / (µmol L−1)
In the Classroom
102
100
98
96
Salinity
0
5
10
Salinity
Figure 7. Variation of the concentrations of NO3᎑, NO2᎑, NH4+, and percent of saturation of dissolved oxygen with salinity obtained in the estuary simulator.
simulator. This research was funded by grants (AMB98-0782 and MAR97-1028) from the Spanish CICYT. Literature Cited 1. Bale, A. J.; Morris, A. W. Estuarine Coastal Shelf Sci. 1981, 13, 1–10. 2. UNESCO Technical Papers in Marine Science, Vol. 36; UNESCO: Paris, 1981. 3. Millero, F. J.; Sohn, M. L. In Chemical Oceanography; Millero, F. J.; Sohn, M. L, Eds.; CRC Press: Boca Raton, FL, 1992; pp 59–113. 4. Burton, J. D. In Estuarine Chemistry; Burton, J. D.; Liss, P. S., Eds; Academic: London, 1976; pp 1–36. 5. Nowicki, B. L.; Candace, A. O. Mar. Ecol. Progr. Ser. 1990, 66, 131–146. 6. Barth, T. W. Theoretical Petrology; Wiley: New York, 1952. 7. Beer, T. In Environmental Oceanography; Pergamon: Oxford, 1983; p 226. 8. Morris, A. W. In Practical Estuarine Chemistry; Head, P. C., Ed.; Cambridge University Press: Cambridge, 1985; pp 1–60. 9. Methods of Seawater Analysis; Grasshoff, K.; Ehrhardt, M.; Kremling, K., Eds.; Verlag Chemie: Weinheim, 1983. 10. Máeda, H. Publ. Seto. Mar. Biol. Lab. 1952, 2, 249–255. 11. Máeda, H.; Takesue, K. In Chemistry and Biogeochemistry of Estuaries; Olausson, E.; Cato, I., Eds; Wiley: London, 1980; pp 233–262. 12. Liss, P. S. In Estuarine Chemistry; Burton, J. D.; Liss, P. S., Eds.; Academic: London, 1976; pp 99–130. 13. Lebo, M. E.; Sharp, J. H. Estuarine Coastal Shelf Sci. 1992, 35, 235–252. 14. Balls, Ph. W. Estuarine Coastal Shelf Sci. 1994, 39, 329–352. 15. Stommel, H. Sewage Ind. Wastes 1953, 25, 1065–1071. 16. Weiss, R. F. Deep Sea Res. 1970, 20, 291–303.
Journal of Chemical Education • Vol. 78 No. 6 June 2001 • JChemEd.chem.wisc.edu