Circuito de Cianuración - Planta de Beneficio

5.3 Planta de Beneficio

5.3.2 Circuito de Cianuración

General Theory of Biological Filtration

Biological filtration relies on physical, chemical, and biological treatment

mechanisms for removal of algal toxins of concern, similar to its predecessor of deep bed filtration. The main objective of engineered biological filtration is to establish consistent hydraulic throughput and optimum removal efficiencies of a wide array of water quality constituents. In the case of algal toxin removal, adequate hydraulic and environmental conditions should allow the formation of microcystin degrading bacterial biofilm

communities (that will concurrently degrade dissolved nutrients) that are stable over time (Ho et al. 2012a). Ultimately, a feedback “bioclogging” mechanism, combined with the accumulation of particulate and organic material or cell debris, influences the available pore space for water flow over time, which results in the requirement to backwash the filter grains and reestablish a relatively clean filter bed (Thullner et al. 2004, Engesgard et al. 2002, 2006). The backwash frequency and intensity are likely to be reduced as to not completely disturb the existing biofilm communities and to allow full and fast regeneration of the previous biofilm communities after filter start up (Emelko et al. 2006).

The efficiency of biological filtration is first governed by physical-chemical mechanisms of contact (transport) and attachment of bacterial and colloidal particles to the surface of the collector grains. Contact potential of either bacterial particles or colloids to filter grains is controlled by several physical transport mechanisms including straining, sedimentation, diffusion, hydrodynamic interactions, inertial impaction, and interception

(O’Melia and Stumm 1967, Huisman and Wood 1974, Keir et al. 2009). Straining relates to the mechanism of entrapment of bacterial particles or colloids that are too large to fit through the pore spaces of the sand grains and depends on the ratio of the diameter of the particle in question (dp) to the representative grain diameter of the media (d10) (O’Melia and Stumm 1967, Huisman and Wood 1974, Keir et al. 2009). Sedimentation refers to the physical settling of bacterial particles or colloids on the grain surface, due to density differences between the particle and fluid, and depends on the available surface area provided by the collector grains, the hydraulic loading rate, as well as the theoretical

settling velocity predicted by Stoke’s law (Yao et al. 1971, Huisman and Wood 1974, Keir et al. 2009). A comparison of the surface loading rate (HLR normalized by the theoretical surface area of grains) to the settling velocity gives some indication to the impacts of sedimentation for the removal of bacterial particles and colloidal matter (Huisman and Wood 1974).

Molecular diffusion relates to the random movement of generally submicron bacterial particles or colloids from areas of high concentration to low concentration and plays a role in mass transfer of particles or water quality constituents from the bulk solution of fluid to the biofilm or sand grain surfaces (Dullien 2012, Bear 2013). Contact between the particle or colloid of interest and the collector surface may also arise due to hydrodynamic actions, or the random, drift in motions associated with migration of spherical bacterial particles across non-uniform shear fields experienced in the soil pores (Keir et al. 2009). Interception occurs when the streamline of a particular bacterial particle or colloid exists within a given radius of the grain surface, resulting in an increased contact potential (Yao et al. 1971). Transport of bacterial particles to grain surfaces may also be

mediated by bacterial mobility (Keir et al. 2009). Finally, inertial impaction is related to contact with a grain surface as a result of the inertial forces of the particle/water quality constituent of interest that causes deviation from the fluid streamlines, thereby impacting the collector surface (Yao et al. 1971).

Once the bacterial particle or colloid of interest comes in contact with the filter grain surface, several mechanisms such as electrostatic interactions, London-Van der Waals forces, and the nature of the particle and grain surfaces influence the ultimate attachment and removal (O’melia and Stumm 1967, Keir et al. 2009). Electrostatic attraction between the particle and grain collector surface arise due to variations in the electric double layers of the particle or water constituent of interest, primarily as a function of ionic strength and pH (O’melia and Stumm 1967, Keir et al. 2009). Changes to the electric double layer may include adsorption of ions in solution or disassociation of functional groups, which may result in net attraction or repulsion to the media surface (O’melia and Stumm 1967). Most clean quartz or silica sands used in filtration applications carry a net negative surface charge, causing the initial deflection of net negatively charged bacteria, anions, and organic matter (Tufenkji and Elimelech 2005). However, as filtration progresses, charge reversal may occur due to oversaturation of the collector surface with positively charged particles, which increases relative adsorption of negatively charged species. Some studies have also described local heterogeneities in surface charge of the grain surface that influence

bacterial attachment, even at high activation energies; therefore, the charge distribution on a surface collector should not be considered completely uniform (Tufenkji and Elimelech 2005). The nature of the bacterial particle surface, such as the presence of organic

in solution and the grain surface, may also influence the ultimate attachment efficiency (Franchi and O’melia 2003, Bolster et al. 2001). London-Van der Waals dispersion forces are associated with dipole moments formed from the temporary asymmetrical distribution of electrons on atomic nuclei and are important attachment forces when the separation distance between the particle and grain surface is relatively small and the size of the molecules or particles of interest is relatively large (Grasso et al. 2002, Tufenkji 2007). Other forces influencing attachment include born and hydration forces. Born forces, repulsive in nature, occur when the electron clouds of two atoms or molecules overlap (McDowell-Boyer 1992, Keir et al. 2009). Hydration forces can arise due to a disruption or change in the molecules surrounding a surface, where repulsive hydration forces exist due to the affinity of water molecules to bind to hydrophilic surface groups on the media surface (hydroxyl groups and hydrated ions) (Elimelech, M., & O’Melia 1990).

The nature of the grain collector surface also plays a role in attachment of particles or water quality constituents of interest, including the relative hydrophobicity and

associated interactions due to phase affinity. For example, if a surface is more hydrophobic it is more likely to bind with constituents that have a natural affinity to dissolve or interact in that phase, such as nonpolar molecules in octanol. The increase in hydrophobicity of the surface also reduces the Gibbs free energy of the surface, which may promote bacterial attachment (as the net repulsive force is reduced significantly) (Scholl et al. 2003, Chen and Strevett, 2003). Similarly, surface roughness or the presence of natural or contrived surface coatings (i.e., metal hydroxides) of the collector grain surface may increase particle

The summation of the interactions among these forces, both attractive and repulsive in nature, termed the resultant adhesive force, influences the ultimate attachment efficiency on the grain surface and varies according to the separation distance of the particle and grain surface (Keir et al. 2009). Generally, from DLVO theory, the interaction energy between the bacteria particle and grain surface is attractive at small separation distances, whereas the interaction energy is repulsive at larger distances from the surface. As the particle approaches the surface, an energy barrier is present (primary maximum) that must be overcome before the particle officially attaches (and reaches) the primary minimum energy state. In some cases, a secondary energy minimum is also present at larger separation distances that may play a significant role in the irreversible attachment of bacterial cells with relatively low thermal energies (Redman et al. 2004). Ultimately, the ionic strength and the pH of the medium have a large impact on the distribution of the theoretical interaction energy curve, where solutions with high ionic strength (decrease double layer thickness) and result in lower activation energies (easier access to the primary minimum). High or low pH may also result in protonation or deprotonation of functional groups, which may increase the contribution of electric double layer repulsion over other forces, thereby increasing the activation energy barrier. However, studies have demonstrated that the effect of pH under operational conditions (5-8) on bacterial

attachment was diminished compared to the effect of ionic strength (Jewett et al. 1995). Once a sufficient bacterial population is established on the filter media (termed the ripening period), biodegradation of different water quality constituents may commence. The proper functioning of the biofilm community depends on the availability of both

by heterotrophic bacteria as energy and carbon sources for development of new biomass or production of extracellular polysaccharides (EPS) or other cellular byproducts. Of course, the presence of autotrophic bacteria (nitrifiers) and inert biomass also affects the structure and function of the biofilm matrix (Rittmann 1987, Rittmann et al. 2002). The function of the EPS coating in the biofilm community is to provide protection against environmental stresses and dehydration as well as to maintain attachment to a given surface under a variety of hydrodynamic conditions (Vu et al. 2009). Substrate utilization of bacterial biofilm communities is generally limited by the mass transfer of both electron donors and acceptors from the bulk liquid in the pore space to the biofilm surface

(Rittmann and McCarty 1980). Current models of biofilm subsistence have proposed that the mass transfer of either electron donors or acceptors to the biofilm surface depend on the diffusion coefficient in the pore space and the length of the diffusion layer film, which presents a resistance to mass transfer (Fick’s first law) (Rittmann and McCarty 1980, Rittmann 1982a). As the EA or ED is transported to the biofilm surface, a spatial gradient of either EA or ED from the surface of the biofilm to the grain surface is established, creating a concurrent diffusive flux of either the EA or ED into the biofilm matrix (Fick’s second law). The diffusive flux of substrate or electron acceptor into the biofilm matrix enables the growth of different microbial communities from the utilization of different substrates, which is often predicted using a Monod type, hyperbolic equation (where the substrate utilization is dependent on the specific microorganism concentration, cell yield, maximum growth rate, and half saturation constant). As long as there are no perturbations in the system, the biofilm community is assumed to reach steady state, where the net

accumulation or microorganisms balances the net decay or release of microorganisms (Rittmann and McCarty 1980, Rittmann 1982a).

The maintenance of the steady state biofilm depends on the rate of microbial detachment, as opposed to attachment, which can be physically, chemically, or biologically mediated. Of the myriad physico-chemical and biological mechanisms, erosion, abrasion, sloughing, predation, and filter backwashing are the main contributing factors (Liu and Tay 2001, Liu and Li 2008). Erosion of biomass within the biofilm results due to varying

hydrodynamic shear conditions within the porous media and is analogous to bits and pieces of the matrix being “shed” from the existing biomass (Liu and Tay 2001). It has been previously demonstrated that conditions of high hydrodynamic shear in porous media lead to a thin, stable, smooth, and dense biofilm matrix, preferable for biological filtration applications (Liu and Tay 2001). Physical collisions with external particles also lead to abrasion of the biofilm matrix (Chaudharry et al. 2003). Sloughing is another physical detachment mechanism related to the sudden loss of a large portion of the biofilm matrix on the physical size order of the length of the biofilm matrix, most likely a function of the hydrodynamic conditions (Telgmann et al. 2004). The potential of grazing by native protozoa on heterotrophic bacterial populations within a biofilm is yet another potential detachment mechanism that may occur in drinking water treatment biofilters. A more direct way, from an operational standpoint, to induce bacterial detachment is the advent of a backwashing system using air scour or fluidized bed techniques to promote

hydrodynamic shear and concomitant losses in biomass from filter grains (Emelko et al. 2006). The use of a backwashing process is often a necessary tool to ensure efficient volumetric throughput of water over the operational life of the filter system.

Microorganisms have also been known to actively detach from supporting media as a result of nutrient limitations, such as carbon or trace nutrient sources (Sawyer and Hermanowicz 1998).

Design and Operation of Conventional Biological Filtration Systems

The design of a conventional biological filtration system depends on the following criteria (but is not limited to): 1) media type and characteristics; 2) length of the media (L) and presence of alternating layers of media types (i.e., mono, dual media); 3) hydraulic (HLR), surface (SLR), and organic loading rates (OLR); 4) empty bed contact time (EBCT); 5) limiting substrate fluxes (Jdeep) and concentrations (Smin); 6) as well as the filter backwash techniques (Huck 2000, Chaudharry et al. 2003). Other significant parameters that should be taken into account are the source water quality such as temperature, ionic strength, pH, as well as the concentration of residual oxidants, and whether ozonation is applied to treat the source water (from pre or post ozonation processes).

The most important criteria for biological filtration include a surface supporting quick microbial growth, larger surface area to support more biomass growth, and adequate surface texture to ensure biomass stability, in which the ultimate selection has major cost implications (Chaudharry et al. 2003, Urfer et al. 1997, Huck 2000). The media types for most conventional systems range from quartz or silica sands, to anthracite, or to granular activated carbon (GAC). The specific surface area (unit surface/unit volume filter) of sands is typically higher than GAC, due to the fact that bacteria are not able to colonize the

microporous structures (1-100 nm) and the effective grain size for sand is lower than GAC (Urfer et al. 1997). Benefits of GAC over sand or anthracite include its macroporous

bacterial detachment from hydrodynamic shear. Further, GAC also adsorbs chemical

constituents in water, such as algal toxins, which sand or anthracite may fail to consistently remove (Urfer et al. 1997). In terms of performance, media selection (anthracite vs. sand) was determined not to have a significant effect on biodegradable organic matter (BOM) removal efficiency for full scale systems, despite the greater amount of biomass observed for full scale systems with GAC (Huck 2000). In addition, GAC-sand dual media systems were concluded to provide better aldehyde and carboxylic acid removals at colder temperatures, establish a BOM biofilm more rapidly, and provide increased protection against ozone or chlorine residuals than anthracite-sand systems. The greatest deterrent to GAC based systems is the initial capital cost and operational costs associated with

regeneration compared to anthracite or sand (Huck 2000).

Media characteristics may include the uniformity coefficient (UC), representative grain diameter (d10), and the grain size distribution of particles. In general, grains with low uniformity coefficient (1-1.4), medium to coarse grain size distributions (0.20 to 0.75 mm) and representative grain diameters are desirable for efficient biological filtration operation and water throughput. The depth of media for most conventional biofiltration systems ranges from 0.08 to 3.73 m, with an average depth of approximately 0.75 m for 21 WTPs surveyed in the U.S. (Evans et al. 2013a, 2013b). Little comparison has been made regarding the treatment performance of biological filters with varying grain sizes and depths, but it is expected there is an inherent tradeoff between treatment efficiency, media size, and water throughput. For example, if smaller representative sizes of sand were incorporated, treatment efficiency would most likely improve (due to an increase in specific surface area) at the expense of a smaller volume of water throughput. Clearly, the

greater depth of media would potentially increase treatment efficiency of water quality constituents but would require more physical resources and area to engineer. Lastly, a majority of biofilter systems in the U.S. are also configured as dual media as compared to mono and multi-media systems (Evans et al. 2013a, 2013b).

The theoretical design and operation of a biofilter can be improved through the use of several distinct “macroscopic” parameters including the hydraulic loading rate (HLR), the surface loading rate (SLR), the organic or substrate loading rate (OLR), and the empty bed contact time (EBCT) (Figure 4). These parameters are termed macroscopic in that they are relatively coarse (and at the continuum scale), where the level of control of the pore scale processes is variable. Hydraulic loading rate (HLR) is specified as the volume of water applied to the nominal surface area of the filter and is equivalent to the specific discharge in groundwater systems. Comparably, surface loading rates (SLR) are defined as the volume of water applied to the specific surface area of the filter grains. If spherical shapes are assumed the SLR can be approximated by Equation 1, where Q represents the

volumetric rate of water application to the filter (m3_{/sec). The organic loading rate (OLR) is} defined as the HLR multiplied by the concentration of species in the water, to obtain a mass loading per time (Equation 2). The empty bed contact time (EBCT) is simply the length of the filter normalized by the hydraulic loading rate (Equation 3). Typical HLRs for

conventional biofilter systems range from 5 to 30 m/hour, with corresponding EBCTs ranging from 5 to 20 minutes, similar to rapid sand filtration systems (Evans et al. 2013a, 2013b). A number of studies have demonstrated that EBCT is a significant operational and design parameter for biological filtration processes as compared to HLR alone (Carlson and Amy 1996, 1998). These studies also determined that for a given EBCT, BOM removal was

independent of the change in HLR, suggesting that external mass transfer (of nutrients or electron acceptors in the bulk liquid to the grain surface) does not play a large role in BOM removal (Urfer et al. 1997). Since BOM removal and influent concentration of BOM were directly proportional, increasing the EBCT was determined to have a less than proportional effect on BOM removal, but still positive nonetheless (Urfer et al 1997, Huck 2000).

𝑆𝑠 = 6 𝑑 (1 − 𝜃) ∗ 𝑉𝑟𝑒𝑎𝑐𝑡𝑜𝑟 (1) 𝑆𝐿𝑅 = 𝑄 𝑆𝑠 (2) 𝑂𝐿𝑅 = 𝐻𝐿𝑅 ∗ 𝐶_𝑤 (3) 𝐸𝐵𝐶𝑇 = 𝐿 𝐻𝐿𝑅 (4)

Figure 4. A schematic of the drinking water treatment biofiltration process and associated coarse design/operational parameters

Rittman (1982a, 1987, 1989) proposed additional, “microscopic” parameters,

focused at the pore scale, to aid in fine tuning the design and operation of aerobic biological treatment systems. These microscopic parameters were developed from the theory of biofilm kinetics in biological reactor systems. The bulk of the theory is focused on a substrate flux (Jdeep, the minimum flux to support a “deep biofilm”) that enters the biofilm, which is dependent on the mass transfer rate of substrate from the bulk liquid to the

the length of the biofilm, if the biofilm is at steady state (no net growth or loss of microorganisms). The utilization of substrate in the biofilm by heterotrophic

microorganisms creates a concentration gradient, driving a diffusive flux of substrate throughout the depth of the biofilm. Steady state biofilms were categorized into deep, shallow or fully penetrated in nature (Rittman and McCarty 1980). Biofilms that are considered deep (with sufficient thickness) often have a length scale where the concentration at the media surface can be assumed to be zero (and have the highest diffusive flux), whereas shallow and fully penetrated biofilms either have a concentration of substrate at the media surface that is nonzero or proportional to the concentration at the surface of the biofilm (Figure 5). From this theory, the biofilter can be designed to operate at the most efficient flux, J (J>Jdeep*3 for heterotrophic organisms) and substrate

concentration, S (S>Smin) to support an adequate population of degrading microorganisms (Equations 5, 6, and 7).

𝐽 =𝑄(𝑆0−𝑆𝑒)

𝐴𝐹𝑉 (5)

Where S0 and Se are the influent and effluent concentrations, Q is the volumetric flow rate of water, AF is the theoretical biofilm surface area, and V is the volume of the reactor.

𝐽_{𝑑𝑒𝑒𝑝}= {2[𝑆_𝐿𝐹− ln (1 + 𝑆_𝐿𝐹)]}1/2 ₍₆₎

Where SLF is the concentration of the substrate at the biofilm surface (which can be equivalent to the concentration in the bulk solution for well mixed systems). 𝑆_𝑚𝑖𝑛= 𝐾𝑠∗𝑏′

(𝑌𝜇𝑚𝑎𝑥−𝑏′) (7)

Where Ks is the half saturation constant, b’ is the overall first order biomass loss coefficient,

In document Labores de desarrollo y preparación para viabilizar la explotación de la veta Kathy entre los niveles 2000 – 2050. Minera Yanaquihua S.A.C.- Arequipa (página 165-170)