Short Stories

Visualization experiments of biodegradation in porous media and calculation of the biodegradation rate

Visualization experiments of biodegradation in porous media and calculation of the biodegradation rate
of 17
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  Visualization experiments of biodegradation in porous media andcalculation of the biodegradation rate D.V. Vayenas  a,b,c , E. Michalopoulou  a , G.N. Constantinides  a , S. Pavlou  a,b ,A.C. Payatakes  a,b,* a Institute of Chemical Engineering and High Temperature Chemical Processes-FORTH, P.O. BOX 1414, GR-26500, Patras, Greece b Department of Chemical Engineering, University of Patras, Patras, Greece c Department of Environmental and Natural Resources Management, School of Natural Resources and Enterprise Management, Agrinion,University of Ioannina, Greece Received 23 November 2000; received in revised form 29 April 2001 Abstract Biodegradation in porous media is studied with carefully controlled and well-characterized experiments in model porous mediaconstructed of etched glass. Porous media of this type allow visual observation of the phenomena that take place at pore scale. Anaqueous solution of five organic pollutants (toluene, phenol,  o -cresol, naphthalene and 1,2,3-trimethylbenzene) was used as a modelNAPL (representing creosote). The bacteria used were  Pseudomonas fluorescens , which are indigenous (even predominant) in manycontaminated soils. The maximum aqueous concentrations of the specific organic substances, below which biodegradation becomespossible, were determined as a function of temperature from toxicity experiments. Visualization experiments were made undervarious flow velocities and organic loadings to study the morphology and thickness of the biofilm as a function of the pore size andthe distance from the entrance, and the efficiency of biodegradation. The efficiency of biodegradation decreased as the aqueousconcentration of NAPL at the inlet increased and/or as the flow velocity increased. The thickness of biofilm decreased as the distancefrom the inlet increased and/or the pore diameter decreased. A quasi-steady-state theoretical model of biodegradation was used tocalculate the values of the mesoscopic biochemical rates and to predict the profile of NAPL concentration in the porous medium andthe thickness of biofilm in pores. The agreement between experimental data and model predictions is quite satisfactory.    2002Elsevier Science Ltd. All rights reserved. Keywords:  Bioremediation; Biodegradation; Biofilm thickness; NAPL; DNAPL; Toxicity; Porous media; Visualization experiments;Modeling 1. Introduction Biodegradation of organic liquids in porous media isa complex process encountered in many fields of prac-tical interest, such as natural attenuation of organicpollutants in soil and aquifers [11], microbially enhancedoil recovery [8], and wastewater biotreatment in tricklingfilters and similar devices [17]. Here, we focus on thebiodegradation of organic contaminants by indigenousbacteria in soil. The aim of this work is to investigate thegrowth of biofilm in pores of the porous medium forvarious typical flow velocities and organic loadings, tocorrelate the biofilm growth with the biodegradationefficiency, and to develop a method for the calculationof the kinetics of biodegradation. This work is part of abroader study of natural attenuation occurring at anabandoned tar factory site in Ringe, Denmark. This siteis being used for systematic studies by Geological Surveyof Denmark and Greenland (GEUS). An importantfeature of the present work is that the model dense non-aqueous phase liquid (DNAPL) used is a well-charac-terized mixture of five typical organic contaminants thatis representative of the creosote-derived contaminationof the Ringe site. Furthermore, the bacteria used areindigenous to this site.Leakages of organic liquids from underground stor-age tanks and pipelines along with irresponsible disposalof waste materials are primary sources of groundwatercontamination. NAPL penetrates into the soil driven bygravity and capillary forces, and may reach the in Water Resources 25 (2002) 203–219 * Corresponding author. E-mail address: (A.C. Payatakes).0309-1708/02/$ - see front matter    2002 Elsevier Science Ltd. All rights reserved.PII: S0309-1708(01)00023-9  groundwater table, in which case a large plume of dis-solved contaminants may be formed. The extent of thisplume is controlled by several processes, namely,amount and rate of release of NAPL from the source,penetration and spreading of NAPL into the soil,gradual dissolution of NAPL, transport of dissolvedNAPL and mixing with uncontaminated groundwater,volatilization, biodegradation of NAPL by indigenousbacteria (intrinsic biodegradation), etc. [44]. Biodegra-dation is the most important natural process contrib-uting to DNAPL removal from contaminatedgroundwater. Most organic compounds in crude oil andfuels are known to be biodegradable under aerobicconditions (e.g. [44]). Biodegradation is also the keyprocess of several remediation techniques (e.g. [22]).Despite the key importance of biodegradation of or-ganic liquids in porous media, the process has not beeninvestigated adequately [43]; lack of adequate under-standing of the mechanisms of action of bacteria on thescale from one to a few thousand pores (including co-operative effects) is a hindrance to devising methods forstimulating the biodegradation process to make it moreeffective for dealing with recalcitrant chemicals.In many cases, bacteria adhere to the pore walls.Although the reason for this adhesion is not fully un-derstood, it seems that exopolymer-producing strainsare more likely to adhere and develop thick biofilms [33].Development of biofilms and microbial activity (biode-gradation of organic pollutants) take place only whenthe local conditions are favorable [4]. For example, inthis work we found that the bacteria  Pseudomonas flu-orescens , indigenous to the Ringe site, survive and de-grade organic pollutants only if the concentration of organic pollutants in the aqueous phase is  smaller  than acritical value which depends on the chemicals involvedand temperature. On the other hand, a  sufficient supply of nutrients is another requirement for the growth of biofilms. 1.1. Review of literature Cunningham et al. [10] studied the effect of biofilmgrowth on the flow through porous media. They showedthat the growth of biofilms causes a significant decreaseof porosity and the permeability of the porous medium.Rittmann [46] investigated the relationship betweensubstrate loading and biofilm morphology. As substrateloading increases, the average biofilm thickness alsoincreases, resulting in a corresponding decrease of theeffective pore space. As the biofilm thickness increases,the interstitial velocity increases, if the flowrate throughthe porous medium remains constant, whereas the in-terstitial velocity decreases, if the pressure gradient re-mains constant. Increased thickness may result indepletion of nutrients within the biofilm structure.Stoodley et al. [50] investigated liquid flow in a modelbiofilm system using particle image velocimetry andscanning confocal laser microscopy. Analysis of particlevelocity data showed that the accumulation of biomassin a porous medium in the form of biofilms has a sig-nificant effect on local flow velocities and shear stresses.Some channels became completely blocked, whereasothers were only partially restricted.Wanner et al. [57] compared experimental measure-ments with theoretical predictions to conclude that, asbiofilm accumulation progresses, the free pore space isreduced, so as to form a series of channels of small di-ameter, which maintain advective mass transport Nomenclature a  media specific surface area  ð m 2 = m 3 Þ C   NAPL concentration (mg/l) C  0  NAPL concentration at the inlet (mg/l) C  i concentration of the i organic substance of theNAPL (mg/l) C  i0  inlet concentration of the i organic substanceof the NAPL (mg/l)  D c  chamber diameter  ð l m Þ  D NAPL  coefficient of diffusion of NAPL through thebiofilm  ð cm 2 = d Þ  D O  coefficient of diffusion of oxygen through thebiofilm  ð cm 2 = d Þ  D t  throat diameter (width)  ð l m Þ  K  d  bacterial biodegradation constant  ð d  1 Þ  K  O  oxygen saturation constant  ð mg = cm 3 Þ  K  NAPL  NAPL saturation constant  ð mg = cm 3 Þ  K  ð  z  Þ  fraction of alive bacteria at position  z  L a ; max  maximum active biofilm thickness  ð l m Þ  L in ; max  maximum inactive biofilm thickness  ð l m Þ  L min  minimumbiofilmthicknessofamonolayer ð l m Þ Q  volumetric flowrate  ð m 3 = d Þ r  C  consumption rate of NAPL or oxygen (mg/l d) V   total volume of the porous medium  ð l m 3 Þ w  mean pore depth  ð l m Þ W    N   mean biofilm thickness in zone  N   ð l m Þ x  biomass density  ð mg = cm 3 Þ Y   yield factor Y  O  oxygen yield factor (g-biomass/g-O 2 ) Y  NAPL  NAPL effectiveness factor (g-biomass/g-NAPL) Greek symbols e  porosity g  effectiveness of biodegradation l  specific growth rate  ð d  1 Þ l max  maximum specific growth rate  ð d  1 Þ r  coordination number 204  D.V. Vayenas et al. / Advances in Water Resources 25 (2002) 203–219  through the porous medium. Sharp et al. [49] studied thestructure and accumulation of thick biofilms in porousmedia and their effects on hydrodynamics and masstransfer. Imaging results indicated the creation andplugging of flow channels within the biofilm/porousmedia matrix and the slight increase of the hydrody-namic dispersivity as the biofilm matrix was developed.Kim and Fogler [26,27] studied the evolution of biomass and permeability in porous media starting with nutrient-rich feed, followed by starvation conditions for bacteria.They reported rapid accumulation of biomass and sig-nificant reduction of permeability during the first step,and biofilm sloughing and increase of permeability whenthe nutrient supply was stopped.Severalexperimentaldevicesandtechniqueshavebeendeveloped and used to study the morphology of biofilms,such as miniature probes and electrical conductivitymethods [36], soft X-ray microscopy [19], particle image velocimetry [50] and scanning confocal laser microscopy[16,28,50]. The growth of biofilms has been investigated in column reactors [16,57], silicon pore imaging elements (SPIEs) [13,14], flat plate porous media reactors [12,49], coaxial cylinders [29,30], membrane bioreactors [18,37], fluidized bed reactors [6], or glass model porous media[26,27,55]. In the present work we use model porous media etched in glass, which allow direct observation of the biofilm growth and the local flow field.Mathematical modeling plays a crucial role in un-derstanding biofilm dynamics and it is a key tool for thelinkage of microscale phenomena occurring within thepores and biofilm with macroscale parameters. Currentmodels take into account the formation of microcolon-ies, the development of heterogeneous colonizationpatterns, the sloughing of large biofilm sections, etc. Themacroscopic models should use  as few adjustable  pa-rameters as possible. Adjustable parameters should havea physical meaning, in order to be able to estimate theirimportance to the description of whole process. More-over, the mathematical evaluation of parameter signifi-cance is essential in determining the required level of accuracy of experimental measurements [31].Significant work on the modeling of biodegradationhas been done during the last few years. The first models[3,47] assumed steady-state biofilmsofuniformthicknesscontaining a single type of organism. In these modelsmass transport and biochemical transformations weredescribed by one-dimensional equations. Later, Wannerand Gujer [56] developed a general dynamic model, inwhich variable distributions of the various species and anumber of limiting substrates were considered. Vayenasand Lyberatos [51] developed a simple quasi-steady-statemodel, which could predict the concentration profile, aswell as the mean biofilm thickness profile along the po-rous medium using recursive analytical equations. Vay-enas et al. [52,53] developed dynamic models to describe nitrification in trickling filters. Their results showed thatnutrient concentration profiles reach the steady-statevalues in a few minutes, whereas biofilm thickness re-quires a very long time to reach steady state. Bishop andRittmann [7] reported that multi-dimensional modelingisrequiredforthepredictionofbiofilmheterogeneity.Tothis end a cellular automata approach has been used todescribe the three-dimensional growth of multispeciesanaerobic biofilms on a submerged flat surface[20,32,39,40,59]. Hermanowicz [21] used cellular auto- mata to simulate a two-dimensional biofilm, and studiedthe effects of external environmental conditions on bio-film development. A fully quantitative two- and three-dimensional approach for biofilm growth and structuralformation has been developed by Picioreanu et al. [41].This model simulates flow over the irregular surface,convective and diffusive mass transfer of substrate, bac-terial growth, and biomass spreading. 1.2. Objectives The aim of the present study is to understand themechanisms of biodegradation on the pore and biofilmscale as well as the associated cooperative effects on themesoscopic scale, in order to determine the mesoscopicbiodegradation rate under conditions similar to thoseencountered in soil. The present work has two featuresthat distinguish it from previous ones. First, in mostprevious works the bacteria used were not indigenous.Second, in previous works in order to accelerate thegrowth of biofilms, bacteria were fed with glucose orfructose, which favor their growth. Consequently, phe-nomena reported in the aforementioned works (such asbiofilm sloughing, etc.) may be absent in real soils. In thepresent work we selected the types of chemicals andbacteria, the aqueous concentration of NAPL and theflow velocity so as to approximate the conditions thatare encountered in the actual polluted site. To this endwe use: •  A well-characterized mixture of five typical organicpollutants (namely, toluene, phenol,  o -cresol, naph-thalene and 1,2,3-trimethylbenzene) as modelDNAPL approximating creosote. (The first four of these substances are common in most plumes; thefifth is usually hard to biodegrade and is used as atracer.) •  Bacteria of the type  Pseudomonas fluorescens , whichwere isolated from the contaminated site at Ringe.These bacteria are also encountered in many contam-inated sites (see below). •  Typical superficial velocities for aquifers of the orderof magnitude of 1 m/d.Toxicity experiments (see Appendix A) showed that thespecific indigenous bacteria  Pseudomonas fluorescens  arecapable to degrade the aforementioned organic chemi-cals, only if the chemicals are dissolved in the aqueousphase and their concentration is relatively low. Due to D.V. Vayenas et al. / Advances in Water Resources 25 (2002) 203–219  205  the low concentration of the organic pollutants and therelatively low specific growth rate of the specific bacteria(which was calculated using the quasi-steady-statemodel), the growth of biofilms is relatively slow and theexperiments lasted several months. Experiments wereperformed in model porous media to observe the growthand the morphology of biofilms in the pores visually, tomeasure the biofilm thickness, and to obtain quantita-tive information on biofilm evolution and pore clogging.The consumption of nutrients (NAPL) along the porousmedium was correlated with the biofilm thickness. Theexperimental results were used to validate the quasi-steady-state model developed by Vayenas and Lyberatos[51]. This model was adopted because of its simplicity,and because it consists of analytical expressions. Therate constants at the mesoscopic scale were determinedusing the experimental measurements and the afore-mentioned model. The calculated biodegradation ratecan be used to predict the fate of plumes in contami-nated sites and to make reliable and realistic toxico-logical and ecotoxicological risk assessment. 2. Materials and experimental procedure  2.1. Bacteria We have chosen to work with pure cultures of bac-teria of a single type to avoid any complications fromsynergistic or antagonistic effects. The bacteria, Pseudomonas fluorescens , were isolated from a contam-inated site at an abandoned tar factory at Ringe, Den-mark [35].  Pseudomonas fluorescens  are typicalindigenous bacteria and are encountered in many con-taminated sites [15,23,25,34,45,48]. For this reason the bacteria  Pseudomonas fluorescens  have been studied ex-tensively in connection with environmental applications(e.g. [9] and references therein).  2.2. Liquid phase The model non-aqueous phase liquid (model-NAPL)is a mixture of five representative pure organic pollu-tants, namely, toluene, phenol,  o -cresol, naphthaleneand 1,2,3-trimethylbenzene (analytical grade; suppliedby MERCK). The first four substances are very com-mon contaminants and are encountered in most plumes,whereas the fifth substance (1,2,3-trimethylbenzene) is arecalcitrant organic substance and it is usually used as atracer in field studies [58]. For the experiments, anaqueous solution of the model-NAPL is prepared asfollows. A mixture of the five compounds was put incontact with distilled water for several days. The solu-bilities of these substances in water are shown in Table1. Phenol and  o -cresol have relatively high solubilities,whereas the rest of the substances have relatively lowsolubilities. The final concentrations of all five organicsubstances in the aqueous solution are shown in Table 1.The density of the mixture of organic compounds dis-solved in water is 1.048 g/l (Table 1), and so the non-aqueous phase is DNAPL (Dense NAPL). This  initial  solution is used as stock in preparing feed solutions withthe addition of salts (see below).For simplicity, the model-NAPL is treated as a singlesubstance in the present work. Hence, the concentrationof the model-NAPL in the aqueous phase is given as asingle number, which is the sum of the aqueous con-centrations of all five organic species.Several inorganic salts are added into the stock so-lution of the organic compounds. The salts used andtheir concentration in the aqueous phase are shown inTable 2. These inorganic salts are necessary for thegrowth of bacteria. The feed solutions used in the ex-periments (see below) are prepared by mixing appro-priate amounts of the stock solution of Table 1 and of  Table 1Substances of model-NAPLSubstance MW (g/mol) Solubility in water,g/l (20  C) (given byMerck)Concentration inaqueous solution,mg/l  ð 20  C Þ Density, g/l(given by Merck)Concentrationfractions at theinlet of the porousmedium (%)Toluene 92.14 0.5 54.2 0.87 0.001Phenol 94.11 82.0 2220.0 1.06 68.72 o -Cresol 108.14 20.0 1852.0 1.04 31.26Naphthalene 128.16 0.03 1.7 1.15 0.0181,2,3-trimethylbenzene 120.20 0.02 9.4 0.894 0.001Model-NAPL 4137.3 1.048 100.0Table 2Model aqueous phaseSubstance MW (g/mol) Concentration(mmol/l)Concentration(mg/l)CaSO 4   2H 2 O 172.17 0.74 127.40CaCl 2  110.99 2.58 286.33Ca ð HCO 3 Þ 2  162.09 0.03 4.86NaNO 3  84.99 0.56 47.60MgCl 2  95.21 0.42 39.98KCl 74.56 0.06 4.47206  D.V. Vayenas et al. / Advances in Water Resources 25 (2002) 203–219  the aqueous phase of Table 2, and so the concentrationsof salts in the feed solutions are constant and equal tothose shown in Table 2.A necessary first step in our study is to ensure that Pseudomonas fluorescens bacteriadegradetheNAPLandto determine the maximum concentration of the model-NAPL in the aqueous solutions for which bacteria sur-vive, grow and function. The experiments of this type(toxicity experiments) and their results are given in Ap-pendixA.Themainconclusionfromtheseexperimentsisthat  Pseudomonas fluorescens  bacteria can degrade theNAPL only if the concentrations in dissolved species arerelatively small, Figs. 14 and 15 in Appendix A.  2.3. Model porous medium The model porous medium used in this study is anetwork of pores of the chamber-and-throat type etchedin glass. Porous media of this type allow direct visualobservation of phenomena taking place at the porescale. Additionally, it is possible to measure the biofilmthickness in pores and to correlate it with the pore di-mensions and the local flow field. The porous mediummodel is constructed using the photo-etching techniquedescribed in [54]. The skeleton of the network is a planarsquare lattice (coordination number,  r  ¼  4) with anode-to-node distance of 1704  l m and it is arranged sothat the macroscopic flow is along a family of diagonals.The pore network comprises 50    40  ¼  2000 chambersand approximately 4000 throats. The chamber diameterdistribution and the throat diameter (width) distributionhave discrete nearly normal distributions with (nominal)mean values  h  D c i ¼  850  l m,  h  D t i ¼  200  l m and stan-dard deviations equal to 1/4 of the corresponding meanvalue. The sizes of these pores are comparable to thoseof microfractures that penetrate the soil of the pollutedsite. (The pore size of the soil ‘‘matrix’’ is much smaller.)Mirror images of the pore network are etched on twoglass plates and after precise alignment they are sinteredin a programmable muffle furnace. All pores have alenticular cross-section [54] and a nearly uniform max-imum pore depth equal to ca 140  l m. The surface (top-view) porosity of the network is 0.30. The values of thegeometrical values of the pore network model aresummarized in Table 3. Due to glass plate heterogene-ities the etched pores are usually 50–100  l m larger thantheir nominal sizes and the size distributions becomenearly continuous.Two channels of width approximately 1 cm and depthapproximately 140  l m are etched at the inlet and theoutlet of the pore network. A hole of 5 mm diameter isdrilled (through the top glass plate) at the center of eachchannel to serve as inlet or outlet port. Three additionalholes of 5 mm diameter are drilled along the centerlineof the pore network, which is parallel to the direction of the macroscopic flow, at distances equal to 2.5, 4.5 and6.5 cm from the inlet of the network to serve as samplingports.  2.4. Experimental procedure To start a continuous operation experiment, themodel porous medium is inoculated with the bacteria Pseudomonas fluorescens . To achieve the attachment of the first bacteria on the pore walls, an LB medium(nutrient solution with 10 g/l tryptone, 5 g/l yeast ex-tract, 10 g/l NaCl) containing bacteria was recirculatedcontinuously through the porous medium (batch mode) only  during the first 12 h of the experiment. After this,the experiment was switched to the continuous flowmode using an appropriately diluted amount of thestock solution of NAPL (Table 1) as feed medium.Adjustments of the flowrate were done using a syringe-type pump (Harvard Apparatus, model 940A). Theflowrates were selected so that the superficial velocity isrepresentative of the superficial velocity in aquifers. Forexample, the flowrate 9.3 ml/d, which was used in theexperiments (see below), corresponds to a superficialvelocity equal to 1.53 m/d. The flowrate was kept con-stant until ‘‘steady-state’’ conditions (with regard to theaqueous concentrations of the organics at the outlet) areachieved. Then, step variations were made both in or-ganic loading and/or flow velocity. The duration of eachstep was about 20–40 days, so that steady-state condi-tions could be achieved. During each stage, samplesfrom the three sampling ports and the outlet were col-lected for chemical analysis every one or two days. Themeasurement of the aqueous concentrations of the or-ganic substances in the samples was performed using aGas Chromatograph (GC 17A, Shimadzu, with Petro-col e  DH 50.2 capillary column by Supelco). All ex-periments were performed at room temperature ð 20    2  C Þ .The oxygen concentrations at the inlet and the outletwere measured using a portable dissolved oxygen meter(Hanna Instruments, model HI 8543). The values of these concentrations during the whole experiment wereconstant and equal to ca 8 mg/l, ensuring fully aerobicconditions.The porous medium model was placed under a mi-croscope (Zeiss D-7082) for detailed observation of the Table 3Theoretical discrete distribution of chambers and throat sizes andfrequencies of appearanceChamber diameter  ð l m Þ  750 800 850 900 950Throat diameter  ð l m Þ  100 150 200 250 300Appearance frequency (%) 16 21 26 21 16Node-to-node distance,  l  ¼  1704  l m; surface porosity,  e  ¼  0 : 3;coordination number,  r  ¼  4; mean pore depth,  w  ¼  140  l m; totalvolume of the porous medium,  V    ¼  0 : 7 cm 3 ; specific surface area, a  ¼  112 cm 2 = cm 3 . D.V. Vayenas et al. / Advances in Water Resources 25 (2002) 203–219  207
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!