Managing the Quality of the Impounded Water

Abstract

Design, construction, and operation of a dam should involve planning and careful consideration not only of the foundation and mass of the dam itself but also of the proper management of the reservoir, and of communities displaced by the reservoir, and impacted in any way upstream or downstream. Many management problems involve a reservoir’s density stratification, resulting in low oxygen, phosphorus release, and hydrogen sulfide (H₂S) in the lower layers. Control measures include selective withdrawal and artificial aeration. Case examples are given. Other problems introduced by damming are often best dealt with by measures slow and well-considered, as illustrated by examples. References for further study are provided.

1. Introduction

A dam is the parent of an impoundment, a reservoir, a lake not formerly present. If such a lake develops quality problems, its parent, the dam, bears some responsibility. A range of lake problems can be foreseen, as discussed in the following sections.

Many problems involve the density stratification of a lake (Section 2). Density stratification often leads to thermal stratification, depleted oxygen in the lower layer, and the release from bottom sediments of nutrients causing algal blooms. To address these problems, mitigating measures can be provided in the design and operation of the dam, as well as in control of contaminant inflows to the reservoir (Section 3).

In turn, the mitigating measures usually come with side effects, good or bad, and sometimes costs. Management measures upstream, in, and downstream of reservoirs should be undertaken slowly and carefully (Section 4).

This brief chapter may serve as a literature survey of the management and environmental impact issues introduced.

2. Stratified reservoirs

Many reservoirs, natural or created by dams, become density-stratified.

2.1 Definitions

Figure 1 is a schematic profile of a density-stratified reservoir. The dam is on the left. It has a gated spillway in its upper parts, and a valve and metered withdrawal exit port at its lowest point, i.e. a low-level sluice. An intake tower with ports at several elevations to abstract water for treatment and production is shown at the right in this schematic diagram, although it may actually be placed next to the dam. (Note that usually only one abstraction port is open at a time).

Figure 1.

Schematic profile of a density-stratified reservoir.

The upper part of the reservoir is the epilimnion. Below it lies the slightly denser hypolimnion. They are separated by the pycnocline, a layer where the rate of change in density with respect to depth is greatest.

Interestingly, the pycnocline is a region that often sees a high concentration of plankton. Harder [1] has observed that plankton, originally homogeneously distributed throughout a volume containing two bodies of water of different densities, will tend to congregate at the density interface, i.e. the pycnocline. Fish are there, too, to feed on the plankton.

2.2 Causes

Density stratification is often a natural phenomenon, due to:

• Solar warming of the epilimnion, rendering it less dense (except at near-freezing temperatures);

• Tributary inflows of cooler water from upstream, which upon entering the reservoir plunge to join the hypolimnion;

• Salinity intrusion from downstream, if the dam is separating the reservoir from the sea or a saline estuary.

2.3 Problems caused by density stratification

In turn, density stratification inhibits vertical mixing, especially across the pycnocline. Impaired vertical mixing leads to the reduction of dissolved oxygen in the hypolimnion. The (impaired) downward diffusion of oxygen through the pycnocline may be estimated as follows: assume the diffusive flux to be F = −Kdc/dy, where K is the vertical diffusion coefficient and dc/dy is the vertical gradient of dissolved oxygen concentration. While in general K is a function of vertical gradients in ambient velocity as well as density gradients, Koh and Fan [2] showed that, as a rough approximation in the ocean, K can be related to the vertical density gradient alone:

$$\text{K}\approx \left({10}^{-4}\right)/\text{e}$$E1

where e = (1/ρ)(dρ/dy) is the vertical relative density gradient expressed in m⁻¹, and K is expressed in cm²/s; and where 4 × 10⁻⁷ m⁻¹ ≤ e ≤ 10⁻² m⁻¹.

Low DO leads, in turn, to problems such as the release of phosphorus from sediments, which in turn enables algae blooms on the surface. In extreme cases, the reduction in dissolved oxygen results in fish kills, unpleasant odors, and even downright dangerous hydrogen sulfide (H₂S) concentrations onshore near the reservoir, and unattractive quality if hypolimnion water is withdrawn for production. This issue is discussed further by Thomann and Mueller [3], among others.

3. Management measures

3.1 Selective withdrawal, selective abstraction

A reservoir should be able to spill downstream, and not just from a surface spillway. Water for use should be withdrawn, and not just from the reservoir bottom.

If a reservoir is stratified, the poorer quality water is usually (but not always) found in the lower layers.

Figure 2(a) illustrates a discharge strong enough to overcome buoyancy forces both in the hypolimnion and in the pycnocline. The flow pattern approaches that of potential flow in a homogeneous medium. Water from below and above the outlet elevation, perhaps even epilimnion water, is withdrawn. This arrangement is partially satisfactory in the example of Reservoirs C and E, described below, and would greatly improve the product of Reservoir D.

Figure 2.

Withdrawal flow patterns. From Imberger, in Fischer et al. [4]. (a) Discharge is strong enough to overcome buoyancy forces both in the hypolimnion and in the pycnocline. (b) Strong stratification in the hypolimnion, with small discharges.

Figure 2(b) illustrates a condition in which withdrawal is exclusively from the hypolimnion, without greatly disturbing the pycnocline. The withdrawal through an outlet port is kept sufficiently small that it “sips” water from a thin layer at the outlet level, somewhat like extracting very few paper sheets from the middle of a stack of papers. The water removed is replaced by the waters above, from the surface down through the epilimnion and the pycnocline to the withdrawal level, descending nearly as a block (except as disturbed by wind or inflows).

Selective withdrawal is discussed in more detail by Imberger in Chapter 6 of Fischer et al. [4], and by Bryant and Wood [5], among others.

3.2 Examples

Reservoir A. This artificial body of water, 85 ha in area and 22 m deep, tends to stratify due to solar heating of its epilimnion. Its withdrawal system consists of a tower such as shown at the right side of Figure 1, with a choice of several depths from which water can be withdrawn for treatment and production. Epilimnion water is usually of the best quality, but it may be advantageous to withdraw from lower layers, as long as they are acceptable, to keep the hypolimnion from becoming a stagnant body of water with no exit.

As will be discussed below, this reservoir also has an aeration system working in tandem with the withdrawal tower, to keep the lower layers sufficiently aerobic [6].

Reservoir B was formed by damming a tidal estuary. It soon encountered water quality problems, including high salinity in the water withdrawn for production. The problems were attributed to (a) the sill of the gated spillway being set so low, and the gate seals insufficiently tight, that they admitted seawater at high tide; and (b) there was no low-level sluice available to preferentially void the most saline water at low tide. Recognizing these faults led to improved design in a later attempt to convert a tidal estuary to a freshwater reservoir (Reservoir E, below) [6].

Reservoir C receives only freshwater inflows, but thermal stratification leads to low dissolved oxygen in the hypolimnion. Bottom withdrawal for production serves to drain the least good water and keeps the hypolimnion from becoming any worse than it already is. However, the depressed oxygen level is insufficient to prevent the reduction of PO₄, hence, the release of P from sediments. This results in major algal growths on the surface.

Expensive and somewhat problematic large-scale dosing of CuSO₄ has been undertaken to control the algal growths. Also, major efforts are undertaken to control the inflow to the reservoir of raw sewage and its oxygen-demanding constituents by upgrading sewerage and treatment in bordering slum areas [7, 8].

Reservoir D, a few kilometers distant from Reservoir C, has a withdrawal point only at mid-depth, near its dam. Its hypolimnion waters are of very poor quality, and lack of a low-level sluice has meant that the hypoxic lower layer could not be easily removed, with taste and odor problems propagating upward even to the water take-off level [6].

Reservoir E: Conversion from seawater to freshwater. A freshwater impoundment can be created by placing a dam across the mouth of a saline tidal estuary to convert it to a freshwater reservoir. When completed, the pool level should be kept higher than the highest local sea tide. The spillway crest is set well above mean tide level, and the spillway gates are made as water-tight as feasible. A low-level sluice is provided to release water from the lowest point in the impoundment out to sea. After the conversion is complete, freshwater for production is to be withdrawn from either of two depths.

The low-level sluice is equipped with a valve, a venturi flow gauge, and a salinometer. Whenever the salinity reading exceeds that suitable for potable water, and there is a net pressure gradient in the sluice towards the sea, the valve can be carefully opened to release the impounded saline water to sea.

This arrangement permits a new impoundment to be converted from seawater to potable freshwater within a time period depending on the rate of freshwater inflow and the volume of the reservoir, without the need to first pump the impoundment dry of its original seawater. At Lower Seletar Reservoir in Singapore, the conversion took about 1 year to accomplish (Figure 3) [9].

Figure 3.

Measured and predicted rate of desalination in Reservoir E.

3.3 Air-bubble mixing and hypolimnetic oxygenation

A mechanism not directly connected to the dam but often operated in association with withdrawal strategies is air-bubble induced circulation. A relatively simple system consisting of air blowers/compressors on shore, an air-feed tube leading from the blowers to a local deep spot on the lake bed, and an air diffuser at that spot on the lake bed, provide a means both to enhance mixing of epilimnion and hypolimnion waters through the pycnocline and to oxygenate the hypolimnion (Figure 4).

Figure 4.

Air bubble column from air diffuser on lake bed, inducing water circulation patterns.

In 1970 Cederwall and Ditmars [10] undertook an early study of the mixing of a water body by an air-bubble plume. In 2012 Lima Neto [11] modeled the liquid volume flux in bubbly jets using a simple integral approach. An extensive discussion of air-bubble columns, hypolimnetic oxygenation, and the air flow rates in many actual installations is provided by Cooke et al. [12].

3.4 Examples

Charles River Basin. Originally the Charles River flowing between Boston and Cambridge, Massachusetts, USA had been a saline tidal estuary to Boston Harbor and the Atlantic Ocean, with a 2-m tide range. As a cosmetic measure, a dam was constructed near its mouth ca. 1910 to maintain a consistently high pool level. However, the dam still allowed a very saline hypolimnion to persist—devoid of dissolved oxygen, and rich in hydrogen sulfide. The density difference was so great that boils from the hypolimnion rarely surfaced, and nuisance sulfide odors at the surface were very rare.

However, plans to build a new and better dam led to fears that the density difference would be weakened, allowing more frequent occasions of hypolimnion water reaching the surface and causing widespread nuisance and perhaps even dangerous H₂S conditions in the heart of the city. A means was sought to eliminate the H₂S from the lower layer.

Following a successful test, a set of six air-bubble diffusers was deployed over a 3-km length of the Basin. A bed of crushed stone supports each diffuser assembly consisting of a fiberglass cradle and an adjustable rack. The rack holds the diffuser in place and is so constructed that it can be adjusted vertically by about 1 m to provide equal air distribution along the entire length of the diffuser. The diffuser is the 5-m long end of the 75-mm diameter polyethylene pipe from shore. It has four rows of 3-mm diameter holes placed at quarter points around the diffuser pipe’s circumference. The holes are spaced 50 mm in the center, and each row is aligned off-center to its adjacent row. A reverse flow check valve is at the intake end of the diffuser and a cleanout flange is provided at the other, normally closed, end. The airflow rate to each of the diffusers is of the order of 0.14 standard m³/s.

This diffuser system did not fully mix the very strongly stratified upper and lower layers, but it did largely eliminate the H₂S and replace it with some dissolved oxygen [6, 13, 14, 15].

Reservoir A. In addition to the multi-level abstraction tower described for Reservoir A above, it was deemed wise to provide a means to ensure vertical mixing for adequate dissolved oxygen at all levels in this 22-m deep impoundment. A single air diffuser, of design similar to those in the Charles River Basin described above, was installed at the deepest point. It successfully kept DO levels adequately high throughout the whole depth and 85-hectare lateral extent of the water body.

Based on this success, similar aerators were installed in other reservoirs in the neighborhood, and have been operating there for nearly four decades [6].

Running costs, it should be noted, especially for electricity, can become significant for an aeration system kept operating on a long-term basis. For most efficient operation, it is best to choose blowers/compressors that are rated for the pressure and flow rate they face. This can be difficult for water depths of the order of 20 m, which make for pressures high for most blowers, but low for most compressors. Again, see the extensive discussion of air-bubble columns, hypolimnetic oxygenation, and the airflow rates in many actual installations by Cooke et al. [12].

An example of aeration failure was the attempt to eliminate H₂S odors emanating from the Kai Tak Nullah, a heavily polluted waterway right next to the runway at Hong Kong’s old airport. The failure of a pilot aeration program in the waterway to raise the near-zero dissolved oxygen level was tentatively attributed to the relatively high tidal current to and fro above the diffuser [6].

4. Impacts upstream and downstream of a dam

Large dam projects are often planned and debated for a number of years before being built, in order to optimize their benefits and to minimize their adverse impacts; yet some less desirable results still ensue. Obeng [16] describes four large dam projects in Africa. Lake Kariba was formed by damming the Zambezi River in 1958. Although hydroelectric power was realized and fish grew abundantly in the reservoir, there were problems including population resettlement, deoxygenation, seismic tremors, an explosive bloom of Salvinia auriculata, and schistosomiasis. The impacts on local populations by the Kariba Dam on the Zambezi River bordering Zambia and Zimbabwe have been studied extensively over the past half-century by Dr. Thayer Scudder and his associates [17].

Kainji Dam on the Niger River in Nigeria was intended for hydroelectric power, flood control, navigation, irrigation, and fisheries [16]. It has met many of these objectives; advance planning has helped to mitigate, although not entirely obviate adverse effects such as algal blooms and schistosomiasis.

Lake Nasser, formed by The Aswan High Dam on the Nile River in Upper Egypt formed an impoundment known as Lake Nasser in Egypt and Lake Nubia in Sudan. As intended, the dam has contained floods, produced hydroelectricity, stored water, and contributed to fisheries, but it is the adverse effects: environmental, affecting the health of the people, and threatening the limestone foundations of many of the Pharaonic monuments that have brought the most criticism of the project [6, 16].

The dam project on the Volta River in Ghana, proposed as far back as the early 1920s, was well studied and planned. However, perhaps due to pressing economic interests, the forestalling of possible environmental impacts did not receive adequate attention during construction [16]. The resettlement of some 80,000 displaced people, though well planned beforehand, was in the event rushed and poorly executed. Eventually, however, the dam fulfilled the intentions for which it was built.

For another example, in 2021 the quick filling of the reservoir behind the dam on the Blue Nile in Ethiopia has been of heated international concern to Sudan and Egypt [18]. Öziş [19] has discussed the quantity and quality implications in the upstream and downstream dimensions of damming on the Euphrates and Tigris Rivers in the Middle East. Anderson [20] discusses the impacts of the Three Gorges Dam on the Yangtze River in China.

The operators of a reservoir should anticipate a large storm inflow and may decide to draw down the reservoir to accommodate it. However, the suddenness of the drawdown can leave riparian groundwater levels unbalanced by the reservoir, to the extent that major landslides into the reservoir cannot be held back, and the suddenly displaced water overtops the dam and causes catastrophic flooding downstream. A prime example is the Vajont Dam flood near Belluno, Italy in 1963, caused by 260 million cubic meters of rock sliding abruptly into the reservoir [21].

A low coastal dam can be called a barrage. Burt & Watts [22] have edited the proceedings of an international conference held in 1996 in Cardiff, UK titled, “Barrages: Engineering Design and Environmental Impacts.” The 39 professional papers cover Concepts and Issues; Environmental Implications; Changing Estuaries; Cardiff Bay’s Hydraulic Regime; the Cardiff Bay Barrage; Ecological Issues; Water Quality; Costs, Benefits and Decisions; Tawe Barrage; River Flows and Control; and Planning and Control.

5. Conclusion

Density stratification often happens normally in a natural lake or an impoundment behind a dam, due to factors such as solar heating of the upper layers or colder or saltier water intruding the lower layers. Density stratification inhibits vertical turbulent mixing, particularly across the pycnocline, the elevation where the change in density with respect to depth is greatest.

Inhibited vertical mixing can lead to depressed dissolved oxygen levels in the hypolimnion (lower layers), particularly if the lake is subject to sewage or other oxygen-demanding inflow. Low oxygen, in turn, releases phosphorus from sediments to feed algal and other phyto-blooms on the surface.

The adverse effects of stratification can be ameliorated by the selective withdrawal of water, to waste, or to abstraction, from the upper or the lower layers, as deemed wise. Air-bubble columns have been used to reduce the intensity of stratification and to increase the oxygen level in the lower layers.

It is important to try to anticipate the possible adverse impacts that the construction of a dam, and the creation of its impoundment, may have on the affected population, environment, and geology nearby, upstream, and downstream. As discussed above, the possible impacts are many and varied. Field conditions from site to site may vary sufficiently that simple import of one solution to a problem, successful at one place, may not necessarily work well at another, as noted for Kai Tak Nullah. The best advice is to move forward slowly and carefully with implementation.