Sample 6: Kinetic adsorption and biodegradation

This application was originally presented in the User's Manual for the original HBGC code (Yeh et al., 1998). The following problem description and the figures were taken directly from that document.

The input file is "kin_ads.in", the superfile is "hbgc123d.ka".

This application simulates bioremediation of a contaminated soil through the supply of an appropriate microorganism at the injection well. The biogeochemical reaction system for this application is adapted from the one-dimensional reactive transport benchmark problem developed by Valocchi and Tebes (1997), and simulated in sample problem 5 ("Microbiological and chemical reactions coupled with transport (1D, 2D, and 3D"). The region of interest is assumed to be homogeneous, having porosity of 0.05 and a permeability of 1.5E-13 m². One central injection well and four evenly placed extraction wells are screened over 1200 dm and flow is distributed evenly over this depth. The injection and extraction rates are 2.52 L/s. Because of symmetry of the flow field, only one quarter of the entire region is simulated using one quarter of the total injection and extraction rates. This quarter region is 150 dm on each side and is discretized using 30 x 30 uniformly sized elements. An impervious boundary condition is imposed on each side of the region. The simulation is performed in two dimensions using a unit thickness of 1200 dm depth and a steady-state flow field. The injection and extraction wells serve as a means of bioremediating this hypothetical field site.

A 50 x 50 dm region adjacent to the injection well is assumed to be initially contaminated with cobalt and nitrilotriacetate (NTA), a chelating agent. The remainder of the region is initially free from these species. Seven components are used to characterize the system: H⁺, H₂CO₃^*, NH₄⁺, O₂, NTA^3-, Co²⁺, and an adsorbent surface site >SOH. Table 1 summarizes the total concentrations of these components present initially in the region. The fluid composition is simulated using 14 soluble complexed species. Formation of all species are assumed to be equilibrium reactions. Table 2 summarizes the reaction tableau for these aqueous species and the equilibrium constants for their formation.

Table 1. Total concentration of components and kinetic species initially in the matrix fluid and in the injection fluid. All chemical species concentrations are expressed in mol/dm³ of media. Biomass concentration is expressed in kg/dm³ of media.

Table 2. Tableau of equilibrium aqueous speciation reactions for bioremediation application.

The mobility of two aqueous species, Co²⁺ and CoNTA^-, is retarded due to kinetic adsorption to an oxide coating on the porous media. The distribution and charge of the adsorption sites on the oxide coating is pH dependent as given by reactions (1) and (2). The two kinetic chemical adsorption reactions are given by (3) and (4).

(1) >SOH + H⁺ <-> >SOH₂⁺, log K_eq = 5.6

(2) >SOH - H⁺ <-> SO^-, log K_eq = -11.6

(3) >SOH₂⁺ + CoNTA^-_(aq) <-> >SOH₂-CoNTA_(ads), k_f1 and k_b1

(4) 2(>SO^-) + Co²⁺_(aq) <-> (>SO)₂-Co_(ads), k_f2 and k_b2

where k_f1 = 0.8000 /h, k_b1 = 0.005260 /h, and k_f2 = 0.2667 /h, k_b2 = 0.05260 /h. The adsorption parameters were selected to make this a significant kinetic process in the system; they do not necessarily represent the true adsorption behavior for Co²⁺ or CoNTA^-.

A preliminary simulation was performed without flow due to pumping to allow for this adsorption process to occur until a steady-state distribution of Co²⁺ and NTA^3- between the aqueous and adsorbed phases occurred. The purpose was to approximate a preremediation condition in which some species' mobility is hindered by adsorption to the matrix. Figure 6.1 shows the results of this kinetic adsorption process. Because of the pH of the system, the reactive adsorbent sites are primarily positively charged, and >SOH₂- CoNTA(ads) is the primary surface species. Negligible amounts of >(SO)₂-Co_(ads) are formed. By approximately 1000 hours, a steady- state condition is reached with 40% of the total Co²⁺ and NTA^3- adsorbed to the media.

After the adsorption process was allowed to reach steady state, the pumping wells were activated. For a period of 48 hours, the injection fluid contains a microorganism capable of degrading NTA. After that initial period, injection of water continues to maintain flow through the region but supply of the microorganism is ceased. The simulation is run for a total of 2500 hours. Table 1 summarizes the chemical and microbial concentrations in the injection solution. Both the matrix fluid and the injection fluid are assumed to have a density of 1 kg/dm³.

Biodegradation can alter the chemical distribution in this system. It is assumed that the microorganisms can use nitrolotriacetate as a growth substrate but can degrade only one of the aqueous complexed forms, HNTA^2-. The biodegradation reaction is:

(5) HNTA^2- + 1.620 O₂ + 1.272 H₂O + 2.424 H⁺ -> 17.307 C₅H7O₂N + 3.120 H₂CO₃^* + 0.424 NH₄⁺

where K_S = 7.64E-7 mol/L, K_O = 6.25E-6 mol/L, µ_max = 0.0916519 /h, b₀ = 5.44E-5 g/dm³ of media, and K_d = 0.00208 /h (Valocchi and Tebes, 1997). It is assumed that the microorganisms are in the aqueous phase in the injection solution, but that they may adsorb to the porous media once in the matrix. A kinetic reaction is used to describe this adsorption process:

(6) C₅H₇O₂N_(aq) <-> C₅H₇O₂N_(ads), with k_f and k_b

where k_f = 0.2667 /h and k_b = 0.05260 /h. Microorganisms both in solution and adsorbed to the media are assumed capable of degrading HNTA by reaction (5) and at the same rate.

Figure 6.2 depicts the total aqueous Co²⁺ and >SOH₂-CoNTA(ads) in the system after 44 hours and 304 hours of pumping. Transport of the aqueous species is apparent. A portion of the adsorbed species is removed from the originally contaminated zone, and some is formed in the initially clean region as aqueous Co²⁺ and NTA^3- are transported toward the extraction well. Figures 6.3 and 6.4 show the aqueous and adsorbed microbial species concentrations at the same times and contrasts them to a "no growth" condition. This "no growth condition" reflects a separate simulation (input file not included) performed with injection of the aqueous microorganism and its subsequent transport through and adsorption to the porous media; biodegradation reaction (5) was not included. Enhanced levels of biomass both in solution and adsorbed to the media result when reaction (5) is allowed to proceed.

Figure 6.5 shows the concentration of total aqueous Co²⁺ and NTA^3- over time at the recovery well. Three conditions are contrasted to allow the impact of different processes to be assessed: (1) recovery of a non-reactive tracer present initially in the same amount and spatial distribution as Co²⁺ and NTA^3-, (2) recovery of Co²⁺ and NTA^3- when flow is maintained through the region but microorganisms are not injected into the system, and (3) recovery of Co²⁺ and NTA^3- when microorganisms are injected into the contaminated zone and biodegradation can proceed. Contrasting conditions 1 and 2 shows the retardation of Co²⁺ and NTA^3- due to their adsorption to the oxide coating, and indicates that flushing of the contaminated zone with clean water will achieve cleanup under the conditions simulated. Contrasting conditions 2 and 3 helps make evident the slight acceleration of remediation that can be achieved with biodegradation in this system. NTA is removed from the system resulting in less recovered at the extraction well. Co²⁺ arrives slightly earlier as a result. Cobalt is retained in the media because of the combined effects of its equilibrium aqueous complexation with NTA^3- and its subsequent kinetic adsorption to the porous media surface as >SOH₂-CoNTA_(ads). As biodegradation removes NTA from the system, less is available to complex with Co²⁺ and adsorb to the surface. Cobalt's mobility is enhanced as a result when biodegradation occurs. Though the impact is small for the particular parameters chosen for this example, this result helps to illustrate the value of the capability of simulating kinetic geochemical and microbiological effects simultaneously.

References

Valocchi, A.J. and C. Tebes (1997): Benchmark Problems: A Workshop on Subsurface Reactive Transport Modeling, October 29 - November 1, 1997. Pacific Northwest Laboratory, Richland, WA.

Yeh, G.-T., Salvage, K. M., Gwo, J. P., Zachara, J. M., and Szecsody, J. E. (1998): HydroBioGeoChem: A Coupled Model of Hydrologic Transport and Mixed Biogeochemical Kinetic/Equilibrium Reactions in Saturated-Unsaturated Media. Report ORNL/TM-13668. Oak Ridge National Laboratory, Oak Ridge, TN.