Environmental Fluid Mechanics - Chapter 16 (end) doc

16 Remediation Issues 16.1 INTRODUCTION As awareness of environmental problems has evolved over the years, there has been an increasing focus not only on reducing ongoing sources of pollu- tion but also on remediation of sites that are already contaminated. A clear example of this is with the so-called Superfund program set up in the United States during the 1980s to clean up large hazardous waste sites causing serious contamination of groundwater resources. Another example is with efforts in the Great Lakes basin of North America to reverse the trend of eutrophica- tion that had been growing steadily in the middle part of the last century. Reduction of nutrient loadings such as phosphorus and nitrogen have reduced drastically the productivity in the lakes, and now the concern is with cleaning residual sites of contamination, in large part associated with contaminated sediments. In general, the initial phase of remediation involves controlling contam- inant disposal into the particular environment under consideration. This may mean complete cessation of all source loadings or at least reduction of loads to a suitable level to allow recovery of the system, either by natural or by engineered processes. According to the type of environment and the time and length scales of the problem, the remediation may include containment and treatment. Most phases of environmental remediation require an under- standing of fluid flow and transport phenomena by advection and diffusion, as discussed in previous chapters of this text. Remediation strategies also may incorporate the use of chemical agents and biodegradation of contami- nants. Methodologies that have been developed and used for many decades for water and wastewater treatment can often be adapted for environmental remediation. Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved. 16.2 SOIL AND AQUIFER REMEDIATION 16.2.1 General Aspects In cases of soil and aquifer contamination, the source of contaminant is usually located above ground, or in a shallow depth below the ground surface. The contaminant penetrates first to the vadose zone, and later it reaches the ground- water. Therefore the issue of aquifer and groundwater contamination is usually associated with site contamination. On the other hand, if the contaminant pene- tration is limited to the vadose zone, then site contamination is not necessarily accompanied with aquifer and groundwater contamination. Figure 16.1 shows a schematic description of site contamination, which originates from a typical landfill. In addition to landfills, other sources of groundwater contamination include spills of soluble substances, which become completely sorbed in the vadose zone and then are gradually released by percolating runoff water. An important category of potentially spilled materials includes a variety of hydrocarbons, such as oils and fuels, and these are collectively known as nonaqueous phase liquids (NAPLs). When a NAPL has a density less than that of water, it is referred to as a light nonaqueous phase liquid,orLNAPL. When NAPL is denser than water, it is called dense nonaqueous phase liquid (DNAPL). When LNAPLs are released at the soil surface, they percolate through the vadose zone and eventually float on top of the groundwater table and the capillary zone, while gradually releasing dissolved hydrocarbon into the flowing groundwater. DNAPLs sink through the water layer and rest on the bottom of the aquifer, except for material that may be adsorbed onto soils, Figure 16.1 A typical site contamination originating from a landfill. Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved. either in the vadose zone or in the water layer itself. LNAPLs also may sorb onto soils. Assuming that any ongoing source of contamination can be removed, there are different alternatives to consider in deciding how to remediate a given site. Sometimes removal of the contaminated vadose zone is desir- able and feasible. In cases of relatively small oil spills it is quite common to excavate the soil contaminated by the spill, to avoid contact between the oil spill and groundwater. However, removal of the contaminated soil intro- duces an additional problem of disposal of the removed soil. If the amount of contaminated soil is not too large, then incineration may be appropriate. For large quantities of contaminated soil, a more common reclamation method involves soil excavation and deposition in a bioreactor, where biodegradation of the contaminant can be achieved in a comparatively short time period. This requires controlling the appropriate supply of moisture, oxygen, and nutrients for the enhancement of the microorganism development and growth. 16.2.2 Containment of the Contaminated Site If the contaminated site cannot be excavated economically or technically, then it may be appropriate to contain it and to apply technologies of in-situ remediation. Containment of the contaminated site is obtained by surrounding the contaminated site by cutoff walls, or vertical barriers, as shown in Fig. 16.2. For the analysis of the vertical barrier performance, it is possible to adopt a one-dimensional conceptual model as shown in Fig. 16.3. The barrier consists of a porous medium with very low permeability and it separates the Figure 16.2 Containment of the contaminated site by vertical barriers. Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved. Figure 16.3 Conceptual model of flow with a vertical barrier. contaminated groundwater from fresh groundwater. According to Eq. (11.3.1), the one-dimensional differential equation of contaminant transport through the barrier is R ∂C ∂t C v ∂C ∂x D D ∂ 2 C ∂x 2  C 16.2.1 where R is the retardation factor, C is the contaminant concentration, t is the time, v is the interstitial fluid velocity, x is the horizontal coordinate, D is the dispersion coefficient, and  is the decay coefficient for the contaminant. The contaminant is transported by advection and diffusion through the barrier. At the upstream boundary, namely at x D 0, the contaminant concen- trationisassumedtobeC en , which may be time dependent. For example, C en may gradually decrease if the contained area is subject to remediation treatment. However, for a conservative calculation, we may assume that C en is constant. At a downstream cross section the contaminant concentration is C ex . The value of C ex increases with time due to the contaminant flux through the barrier. The increasing value of C ex has no effect on the advective contam- inant flux through the barrier, but it may lead to a decreasing diffusive flux, due to a smaller concentration gradient. Therefore, again for a conservative calculation, we consider that C ex D 0, and its value is kept constant. Such an assumption has no effect on the advective contaminant flux through the barrier, but it maintains the maximum possible diffusive flux of the contami- nant. The instantaneous contaminant flux F, at any cross section of the barrier, is given by F D  vC  D ∂C ∂x 16.2.2 where  is the porosity, v is the interstitial flow velocity, C is the contaminant concentration, D is the dispersion coefficient (including molecular diffusion and mechanical dispersion), and x is the longitudinal coordinate. We may refer to differences between values of F at the entrance cross section, where x D 0, and values of F at the exit cross section, where x D Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved. L. At the entrance cross section, the contaminant is subject to advection, represented by the first term on the right-hand side of Eq. (16.2.2), as well as dispersion, which is represented by the second term on the right-hand side of Eq. (16.2.2). At the exit cross section, due to the assumption of C ex D 0, the advective contaminant flux vanishes, and the contaminant is transported solely by dispersion. Even under unsteady-state conditions, the advective flux is assumed to be kept constant at the entrance cross section of the barrier. On the other hand, the dispersive flux gradually decreases, as noted above. Initially it is very large, when the contaminant concentration gradient is large. On the other hand, at the exit cross section the dispersive flux gradually increases from an initial value of zero. Therefore calculation of steady-state conditions may provide an estimate of the maximum contaminant flux that can be expected at the exit cross section of the barrier. Of course, from a practical view point, it is also appropriate to provide an estimate of the time period needed to develop the maximum steady-state flux. For a conservative contaminant, under steady state conditions the flux F is constant at every cross section of the barrier. Under such conditions, Eq. (16.2.2) is obtained by direct integration of Eq. (16.2.1). A further integration of Eq. (16.2.2) then gives ln  K  F v  C  D vx D 16.2.3 where K is an integration constant. Applying the boundary conditions of C D C en at x D 0, and C D 0atx D L, then shows that K D 1 F/v  C en F D  vC en 1  expvL/D 16.2.4 The Peclet number of the barrier is defined by Pe b D vL D 16.2.5 If Pe b is high, then Eq. (16.2.4) can be approximated by F ³  vC en 16.2.6 If Pe b is very small, then Eq. (16.2.4) can be approximated by F ³ D C en L 16.2.7 By introducing the expressions of K and F (Eq. 16.2.4) into Eq. (16.2.3), the contaminant distribution in the barrier is found as C D C en  1  exp[v/DL x] 1  expPe b   16.2.8 Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved. For large values of Pe b , this expression reduces to C D C en  1  exp  Pe b  1  L x  16.2.9 For small values of Pe b , we use the series expansion of the exponential terms of Eq. (16.2.8) to obtain C D C en  1  x L  16.2.10 Also, if Pe b is large, then steady-state conditions of contaminant transport through the barrier are established after an approximate time period of T ³ L v 16.2.11 If Pe b is small, then the time period required for the establishment of steady- state conditions can be estimated using analytical solutions of the diffusion equation. 16.2.3 Pump-and-Treat of Contaminated Groundwater Following containment of a contaminated site, appropriate treatment tech- nologies are generally needed to bring the site to full reclamation. Ground- water of the contaminated site can be pumped into a treatment plant and later reinjected into the aquifer. Sometimes the treated groundwater can be used directly, mainly for irrigation purposes. A variety of treatment methods are classified as in-situ treatment methods. These can sometimes be applied without physical barriers. A common approach is to apply hydrodynamic isolation approaches, rather than physical barriers, to contain the contami- nated portion of the aquifer. Hydrodynamic isolation applies various types of injection and extraction well combinations that do not allow the migra- tion of groundwater from the contaminated site to neighboring aquifers. In Fig. 16.4, schematics of two common options of hydrodynamic isolation are shown. Calculation of flow conditions in the two examples of Fig. 16.4 can be done using potential flow theory and well hydraulics, as detailed in Chap. 11. Each of these examples is associated with the separation of the aquifer flow into two regions. One region incorporates mainly the fresh groundwater. The other region incorporates a comparatively small portion of the fresh ground- water flow and also the flow of contaminated groundwater. A well-defined line of separation represents the interface between these two regions. The schematic of Fig. 16.4 shows two examples of pumping of contaminated groundwater for its possible treatment by conventional methods of waste Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved. Figure 16.4 Hydrodynamic isolation of a contaminated site: (a) isolation by a single pumping well; and (b) isolation by a combination of a pumping well and an injection well. Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved. treatment. Such an approach is called pump-and-treat. Hydrodynamic isola- tion, incorporated with pump-and-treat, can be a comparatively inexpensive method, which leads to gradual reclamation of the contaminated portion of the aquifer. To obtain high efficiency of the systems shown in Fig. 16.4, it is impor- tant to avoid conditions of significant dispersion and mixing between the fresh and contaminated groundwater. However, problems arise if contamination of the aquifer is associated with significant sorption–desorption capacity onto the soil of the aquifer. For example, in cases of soil contaminated with LNAPL, then in the case described by Fig. 16.4a, groundwater is continuously contam- inated by the residual adsorbed or entrapped material. It should be noted that in cases of NAPL entrapment, different agents to enhance the remediation, such as surfactants and nutrients for microbial activity, can be added to the water injected into the aquifer. However, such materials should be chosen so as not to cause other types of aquifer pollution. If the contaminated site of Fig. 16.4b is rich with adsorbed or entrapped contaminant, then injected water is subject to contamination prior to its pumping by the pumping well. Figure 16.5 shows a schematic of a pump-and-treat system, in which the aquifer is contaminated by NAPL. Figure 16.5a illustrates a problem of contamination by LNAPL where, due to seasonal and annual fluctuations of the groundwater, some quantities of the LNAPL are entrapped within the top layers of the aquifer. The flow induced by the pump-and-treat system is associated with dissolution and solubilization of the entrapped NAPL, as well as with penetration of the dissolved constituents into the deeper portions of the aquifer. In the case described by Fig. 16.5b, DNAPL is entrapped throughout the entire thickness of the aquifer. Induced groundwater flow of the pump-and-treat system is associated with the dissolution of the entrapped DNAPL. Calculations of the performance of the pump-and-treat system shown in Fig. 16.5 can be done using a conceptual one-dimensional flow model. Under such conditions, the process of NAPL dissolution and mass transfer from the entrapped NAPL ganglia to the flowing aqueous phase can be calcu- lated using the approach presented in Sec. 11.5. The pump-and-treat system of Fig. 16.5a then appears to be inefficient, as most of the induced ground- water flow cannot be in contact with the entrapped LNAPL. Furthermore, the induced groundwater flow enhances transverse dispersion, which leads to penetration of dissolved constituents into deeper layers of the aquifer. As an alternative, the pump-and-treat system of Fig. 16.6 is based on the use of a single pumping well. The discharge of the well causes a drawdown of the groundwater table and an associated cone of depression. The cone of depression contains the lens of LNAPL and avoids the uncontrolled migration Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved. Figure 16.5 Examples of pump-and-treat systems: (a) aquifer contamination by LNAPL; and (b) aquifer contamination by DNAPL. of NAPL. The floating lens of LNAPL flows towards the pumping well in the region of the cone of depression. Various techniques can then be applied to collect the floating LNAPL in that region, by various types of membranes and floating pumps. Following the pumping of the contaminated groundwater, it must be treated. The appropriate treatment of the extracted groundwater depends on the type of contaminant. In cases of inorganic contaminants, precipitation is an attractive treatment method. Precipitation is governed by the pH value, which Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved. Figure 16.6 Containment of the LNAPL lens by a single pumping well. may be adjusted by adding lime to the treatment stream. Sometimes aeration of metals creates salts with faster precipitation. Dissolved organic materials can be removed by air stripping. Organic compounds, which have low volatility, cannot be removed efficiently by air stripping. Instead, they can be sorbed onto activated carbon. Other compounds can be treated by biological methods similar to the treatment of domestic wastes. 16.2.4 In-Situ Remediation In various cases, remediation by pump-and-treat is not feasible or is not the optimal method. For example, when a volatile organic compound is spilled into the unsaturated zone, it partitions between the liquid and vapor state. The vapors may migrate through the vadose zone and accumulate in underground structures like basements, where they pose a threat of fire or explosion. In such cases soil-vapor extraction (SVE) methods can provide an appropriate measure of in-situ remediation. According to these methods, wells are installed in the vadose zone and are used to pump air and vapor. Other SVE systems may incorporate air injection wells and air-vapor extracting wells. Such systems can also be used if the contaminant volatility is comparatively low. In such cases the injected humid air enhances the bioactivity and in-situ bioreme- diation. These kinds of systems are sometimes referred to as air-sparging systems. Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved. [...]... growth-limiting substrate In the limiting case of Cs × Kc , Eq (16. 3.3) yields D 16. 3.4 max When this occurs, the reaction is called a zero-order reaction Alternatively, in the limiting case of Cs − Kc , Eq (16. 3.3) yields D max Kc Cs Copyright 2001 by Marcel Dekker, Inc All Rights Reserved 16. 3.5 This is called a first-order reaction, with the first-order rate constant given by max /Kc If we let M represent... concentration of the growth-limiting substrate If several growthlimiting substrates should be considered, then the concentrations and halfsaturation constants of all growth-limiting substrates should be incorporated in Eq (16. 3.3) by products of terms similar to that of Eq (16. 3.3) Usually, besides the organic substrate, it is appropriate at least to consider oxygen as another growth-limiting substrate In... expressed as dCs max MCs D dt Y Kc C Cs 16. 3.6 The ratio between max and Y represents the maximum contaminant utilization rate per unit mass of microorganisms Equation (16. 3.6) should be accompanied by the equation of biological mass transport, growth and decay, dM D dt max MY Cs Kc C Cs bM 16. 3.7 where b is a first-order decay coefficient, representing cell death 16. 3.3 Modeling of Biodegradation The... yield coefficient, Kc is the half-saturation constant, KG is the half-saturation constant of the limiting growth nutrient, CO is the concentration of the natural organic carbon, F is the ratio of limiting growth nutrient to hydrocarbon consumed, and b is the microbial decay rate The set of Eqs (16. 3.8)– (16. 3.10) considers that the organic contaminant and oxygen are the growth-limiting substrates Prior to... first-order decay rate is 0.5 day 1 , what is the steady-state concentration in the lake? Assume well-mixed conditions (c) Assuming that the lake is at steady state with regard to the chemical concentration, calculate the time required for the lake to reach 10% of the steady-state concentration if the inflow concentration is reduced to zero (d) How long would it take for the lake to reach a new steady-state... corresponding to retention times of (a) 1 day, (b) 2 days, (c) 5 days, and (d) 10 days Problem 16. 12 In Eq (16. 4.6), explain the implication of taking the second summation over all the inflows to the bay In other words, what is the implicit assumption that has been used? Problem 16. 13 Under steady-state conditions of a well-mixed bay in a large lake or coastal area, with no inflows to the bay and no net flow across... is reduced to zero Setting Cin D 0 in Eq (16. 4.1) and Figure 16. 10 Schematic for problem in which a lake has one inlet and one outlet Copyright 2001 by Marcel Dekker, Inc All Rights Reserved integrating, along with the initial condition that C D C0 at t D 0, gives C D C0 exp Q Ck t 8 16. 4.3 Or, setting C D C1 and solving for t, we have tD Q Ck 8 1 ln C1 C0 16. 4.4 As should be expected, the required... expression, while assuming an oxygen half-saturation constant of 0.1 mg/L, a benzene half Copyright 2001 by Marcel Dekker, Inc All Rights Reserved saturation constant of 22.6 mg/L, a maximum utilization rate of 9.3/day-mg, and a microorganism population of 0.05 mg/L (c) Repeat the calculation of part (a), but apply a first-order decay expression, while assuming a half-life for benzene of 5 days Solution... Problem 16. 4 Consider the steady-state distribution of a radioactive contaminant in a barrier Assume that at the barrier entrance the contaminant concentration is C0 At the exit of the barrier, the contaminant distribution vanishes The retardation factor is R, the flow velocity is V, the dispersion coefficient is D, and the decay rate is (a) Derive the analytical solution of Eq (16. 2.1) for steady-state... Problem 16. 5 For the conditions specified in problem 16. 4, assume R D 1.5, h D 1 m, L D 3 m, K D 10 3 m/d, D D 10 8 m2 /s, and the half-life of the radioactive contaminant is t1/2 D 10 yr Also assume that at the barrier entrance the contaminant concentration is C0 D 2 ppm At the exit of the barrier, as before, the contaminant concentration vanishes (a) Apply the numerical scheme of problem 16. 1 to . expvL/D 16. 2.4 The Peclet number of the barrier is defined by Pe b D vL D 16. 2.5 If Pe b is high, then Eq. (16. 2.4) can be approximated by F ³  vC en 16. 2.6 If Pe b is very small, then Eq. (16. 2.4). All Rights Reserved. treatment. Such an approach is called pump-and-treat. Hydrodynamic isola- tion, incorporated with pump-and-treat, can be a comparatively inexpensive method, which leads to. aqueous phase can be calcu- lated using the approach presented in Sec. 11.5. The pump-and-treat system of Fig. 16. 5a then appears to be inefficient, as most of the induced ground- water flow cannot be