J Cham Tech Biotechnol 1981,31, 717-731 Design Models for Adsorption Systems in Wastewater Treatment Gordon McKay Department of Industrial Chemistry, The Queen's University of Belfast, 21 Chlorine Gardens, Belfat BT9 5DL (Paper received 27 February 1981 and accepted 19 June 1981) The application of mathematical models to the design of adsorption systems as a method of purifying wastewaters is considered The design of batch, fixed-bed, pulsedbed, moving-bed and fluidised-bed systems for effluent treatment is discussed A brief review of the wide range of industrial effluents and adsorbents that may be utilised in such systems is considered Reactivation procedures are presented and the principles of minimising design costs by optimising design are given Introduction Adsorption operations exploit the ability of certain solids to concentrate specific substances from solution onto their surfaces Typical liquid separations include the removal of moisture dissolved in gasoline, decolorisation of petroleum products and aqueous sugar solutions, removal of objectionable taste and odour from water, and the fractionation of mixtures of aromatic and paraffinic hydrocarbons The treatment of wastewaters by adsorption techniques is receiving growing attention since the standards for the quality of a waste effluent are gradually becoming more rigid, and new methods of treatment are continuously being developed Certain industrial wastes cannot be treated by conventional methods of aerobic digestion and adsorption offers an attractive method for solving these problems Most of the work has been based on studies with activated carbonl-14 which has found wide application in wastewater treatment operations Activated carbon is available in powdered or granular form; the powder is usually dry-fed or slurry-fed to the water undergoing treatment, whereas the granular material is normally used in pressure-fed adsorption units similar to pressure filters Specific applications of adsorption include the removal of dialysable organic compounds in solution in sewage effluents.15 Other examples of the adsorption of organic compounds from dilute solutions are available, sucb as butanol adsorption on carbonla and detergents with carbon and macroporous zeolites and gel-type resins.l7.1* The substances which are most necessary to remove, for example pesticides, tastes and odours, generally constitute only a small fraction of the organic matter in the ~ a t e r ~ Treatment of textile wastes by adsorption techniques has received considerable attention An activated carbon plant,20 designed to treat 2.52 dms s-1 (40g.p.m.), incorporates a continuous on-line operation for the treatment of textile effluent and has two sets of three adsorption columns in series While one set of columns is in operation for the effluent treatment, the remaining columns are undergoing regeneration The use of other adsorbents in treating textile effluents is continually being investigated Activated silica made from sodium silicate has been usedal-2s as an important adjunct in the treatment of water supplies as well as waste effluents from many industrial processes Silica gel has been shown to be an effective adsorbent for basic dyes24325 in fixed-bed and fluidised-bed systems Some workers 28-28 have found Fuller's earth and bauxite successful as adsorbents for textile effluent 0142-0356/81/1200-0717 $02.00 01981 Society of Chemical Industry 717 G McKay 718 treatment, but considerable flow problems wereencountered in the fixed-bed operation Williamson29 tested several cheap commercially available materials for the treatment of wastewaters The use of peat and wood for wastewater treatment is of considerable intere~t.~O-~O Adsorption systems It is important to consider the modes of contacting the solid adsorbent and the wastewater when applying the adsorption system to large-scale treatment Five types of contacting systems are usually encountered: (i) batch-type contact; (ii) fixed-bed type process; (iii) pulsed beds; (iv) steadystate moving beds; (v) fluidised-bed type contact Batch-type processes are usually limited to the treatment of small volumes of effluent, whereas fixed-bed systems have an advantage because adsorption depends on the concentration of solute in the solution being treated The adsorbent is continuously in contact with fresh solution; thus the concentration in the solution in contact with a given layer of adsorbent in a column is relatively constant Conversely, the concentration of solute in contact with a given quantity of adsorbent is continuously changing due to the solute being adsorbed Fluidised beds for effluent removal result in high mass-transfer rates, but suffer from relatively short residence times 2.1 Batch adsorption The schematic diagram for a single-stage adsorption process is shown in Figure The solution to be treated contains G kg of solvent and the pollutant concentration is reduced from YO to Y1 kg of solute kg-1 of solvent The amount of adsorbent added is L kg of adsorbate-free solid and the solute concentration increases from XO to X1 kg of solute kg-l of adsorbent If fresh adsorbent is used, Xo=O The mass balance equates the solute removed from the liquid to that picked up by the solid G( Yo - Y l )= L(X1- XO) = L X (1) The Freundlich isotherm data from equation (2) may now be applied to (1) L g of adsorbent X , g of solute g-' of adsorbent G g of solvent G g of solvent C, g of solute g-' of solvent C, g of solute 9-l of solvent L g of adsorbent XI g of solute g-' of adsorbent Figure Single-stage batch adsorption 719 Design models for adsorption systems in wastewater treatment Figure Contact filtration Schematic arrangement for singlestage batch adsorption A, Agitated contacting tank; B, filter press; C, filtrate storage tank; D, steam where k and n are the Freundlich constants, Xe represents the equilibrium concentration of solute (kg) on adsorbent (kg) and Ye is the equilibrium solute concentration Substituting equation (2) into (1) for unit mass of adsorbent gives equation (3) L YO- Ye c=o'ln (3) Equation (3) permits calculation of the adsorbent, solution ratio for a given change in solution concentration, YO-Ye A typical batch process is shown in Figure The effluent to be treated and the adsorbent are intimately mixed in the treating tank for a set time to enable the system to approach equilibrium, following which the thin slurry is filtered to separate the solid adsorbent and adsorbate from the liquid The equipment may be made multistage by providing additional tanks and filters If the operation is to be made continuous, centrifuges or a continuous rotary filter may be substituted for the filter press The adsorbent is usually applied in the form of a finely ground powder and the time required for the adsorbent and liquid to come to substantial equilibrium depends primarily on the concentration and particle size of the solid, the viscosity of the liquid and the extent of agitation Agitation should be vigorous to ensure rapid contact of the adsorbent particles with the liquid 2.2 Fixed-bedadsorption Normally, in practical cases, the concept of fixed-bed adsorbers is expressed in graphical terms, to allow for deviation from the ideal case The dynamic adsorption system is represented in Figures and 4, where the wavefront length (mass-transfer zone or MTZ) can be expressed in terms of time, In the MTZ the concentration of solute in the effluent will change from YOto Ye In Figure the area below the wave (a, g, d, e, a) reflects unused adsorbent capacity and the (a, g, d, e, a)/ (a, b, d, e, a) ratio is the fraction of unused adsorbent in the MTZ I Figure The Mass-transfer zone I I / I/ V I I I G McKay 720 Effluent v breakthrough curve 71 yv - - - yo e eb - - v eb e, e Onstream t i m e ( ) Figure Progress of a stable mass-transfer front through an absorbent bed(a) and position of the stoicheiometric front relative to the stable mass-transfer front during dynamic adsorption (b) A vertical straight line drawn through g, the 50% breakthrough point, such that area (a, f, c, b, a) =(a, g, d, b, a) and (f, e, d, c, f) = (a, g, d, e, a), yields an equivalent stoicheiometric front for the system at time, 0, The rectangle (h, f, c, k, h) corresponds to adsorbent at its equilibrium loading, X,; it is defined as the equivalent equilibrium section and is specified in terms of the length of this section, ZE.The area (f, e, d, c, f) corresponds to adsorbent at its initial loading, X O ;it is specified in terms of the length of unused bed, Z, The adsorbent bed at breakthrough is actually comprised of an equilibrium zone and a masstransfer zone The system is shown schematically in Figure where (a) depicts a stable MTZ moving at uniform velocity, U,through an adsorber and (b) shows the stoicheiometric transfer front superimposed over the actual transfer front The total amount of solute removed, w, from the bulk fluid between time and time (0+ A0) is expressed in equation (4) (4) w = G ( Y e - YO)A As w kg of soluted are adsorbed, the shape of the MTZ is unaltered although its position changes and there is an increase in bed length, AZ, and w = ( AZ) a,( Xe - XO) (5) where p is the bulk density of the adsorbent and a is the surface area for adsorption Combining equations (4) and ( ) and rearranging gives equation (6) AZ= ( A0 ' ) or apAX AZ= UA8 During steady-state operation, the values of G, p, AX and A Y are fixed and so at any time, 0, the position of the stoicheiometric front, Z,, is given by equation (7) zs= u0 and by definition at the breakthrough time, 0b, zs=zE Therefore zE= u8b At the stoicheiometric time, 0,, z, =zo Therefore z o = ue, (7) Design models for adsorption systems in wastewater treatment 721 Combining equations (9), (11) and (12) and eliminating U gives z u = z o (l+) Thus simple breakthrough curves provide the data required for the correlation of rate data in dynamic systems and equation (8) is the basic equation which may be used for the analysis of breakthrough data Several examples of the use of the MTZ theory33-37 may be found in the literature A simplified approach for fixed-bed adsorbers is available to correlate the Service time with the operation variables One such model is the Bed-Depth Service Time (BDST)modelas (a simplified form of a previous theory)3gand states that the Service time of a column is given by the following equation (14): e,=" [Z-mln V ($-I)] xo v where NOis the adsorption capacity, k the rate constant of adsorption and V the superficial volumetric flow rate At 50% breakthrough, Xo/X=2 and 8= 80.5, the final term in equation (14) becomes zero giving the relationship: 80.5= No xo v Z= constant Z ~ Consequently a plot of BDST at 50% breakthrough against bed height should be a straight line passing through the origin The model has been applied successfullys~3sprovided certain conditions apply; namely, that the MTZ is small compared with the adsorbent bed height and also that the breakthrough curve is symmetrical about the stoicheiometric front A further column-design method38 is based on two concepts: the height of an equivalent transfer unit (HETU) and the maximum permissible space rate at which full adsorption takes place (R) The HETU unit is based on the variation in the void fraction, E , with variation in mean equivalent particle diameter, dp,and packing density It reaches a maximum of 0.12 m when E is 0.50 and dp is 0.008 m and methods for estimating HETU for any adsorbent are a ~ a i l a b l e ~The ~ - HETU ~~ for any adsorbent can be determined from Figure and the minimum adsorbent bed height for maximum product quality is 55-60 HETU; thus Zmin=(55-60) HETU (16) Equation (16) normally results in minimum bed heights between and m Bed heights>Zminare desirable economically for longer on-stream lifetimes, but values of Z> 10 m should not be used because of difficulties in maintaining packing stability Column diameters are limited to constructional details For diameters c m columns are usually made from pipes and for diameters 1-4 m (a maximum value for stable packing) plate fabrication is normally used The maximum flow rate, R (m3 of solution per m3 of adsorbent h-1) may be calculated from equations (17) and (18) with certain limitations as described by Johnston (See reference49.) log R=2.0896-38.7 p + b g (0.00308-0.305 dp) For p > 0.5 x 10-3 Nsm-2 and dp< 0.00305 m R=101.2-3.08~ S p + ~10gp2-219.5dp For 0.05 x lO-3< p < 0.50 x 10-3 Nsm-2 and dpup to 0.008 m G McKay 122 0.003 I 0.34 I I 0.36 0.38 I 0.40 Void fraction 0.42 ( D packed ~ I 0.44 I I 0.46 0.48 density ) Figure HETU against void fraction Nsm and dp=0.61x m are substituted into equation (18) However, if a value of 0.40x a value of -779 m3 of feed per m3 of adsorbent h-l is obtained Furthermore suitable values of p and dp in equation (17) give low results [Limited information in reference 38 makes it impossible to derive the correct equations for Rx; the author therefore does not recommend the use of equations (17) and (18).] The volume of bed is obtained by dividing the throughput by R and thus the amount of adsorbent, diameter and height of bed can be determined The maximum permissible space rate is a function of the feed absolute viscosity, the pressure 42 drop and dp.Pressure drop and flow distribution problems have been di~cussed.~l Figures and showed schematically the operation of fixed-bed adsorbers in series and in parallel respectively 2.3 Pulsed beds A moving or pulsed-bed system may be used, where some carbon is removed at intervals from the bottom of the column and replaced at the top by fresh adsorbent, when the adsorption wave front In out Figure Fixed adsorption beds in parallel Design models for adsorption systems in wastewater treatment In Figure Fixed adsorption beds in series 723 I1 t L out begins to leave the system Consequently, the rate at which the wave front moves through the bed will be the rate for determining when a ‘pulse’ of fresh adsorbent must be removed The BDST equation may be varied to cover pulse systems by determining the wave-front rate from pilot-plant tests The modified equation is developed by assuming that the moving bed is pulsed just as the wave front begins to leave the column The design equation based on equation (14) may be expressed as: eb= U Z - /I where the constants No Yo v and /I= k-1n Yo (;-I) Effluent Freeboard Media Influent Figure Schematic diagram of pulsed-fed adsorption unit Air inlet G McKay 124 For changes in flow rate, VI, then the constant a is replaced by al, where (20) and for changes in feed concentration, XI*then a and b are replaced by u1 and where and YA is the new effluent concentration A typical process configuration which has been developed43 is shown in Figure The system uses a column or bed of a fine medium with wastewater and air flowing concurrently upward through the bed The air supplies oxygen for biological oxygen demand (BOD)and also provides a pulsing agitation action to the bed to prevent clogging L L X G G Y L I+dX YtdY n L XO G YO Figure Continuous-contact moving-bed adsorption 725 Design models for adsorption systems in wastewater treatment 2.4 Steady-state moving-bed adsorbers Steady-state conditions require continuous movement of both fluid and adsorbent through the equipment at a constant rate, with no change in composition at any point in the system with passage of time If parallel flows of solid and fluid exist, the net result is at best a condition of equilibrium between effluent streams or the equivalent of one theoretical stage However, countercurrent operations of such systems enable separations equivalent to many stages to be developed A schematic diagram of a continuous counter-current adsorption system is shown by Figure A solute balance around the entire tower is given by equation (21) G( Yo- Y I )=L( Xo - XI) (21) The mass balance around the upper part of the column is, G( Y - Y I ) =L ( X - X I ) From the mass-balance equations, the operating line can be constructed on the equilibrium curve (Figure 10) The operating line is the straight line of slope L/G, joining the terminal conditions v, Y Figure 10 Operating diagram for continuous-contact adsorption R (XO, YO)and ( X I , Y I ) The concentrations of solute, X and Y, at any height within the column will lie on this line By assuming that the column operates under approximately isothermal conditions and that film and internal diffusional resistances can be represented by an overall mass-transfer term, ky,then the incremental mass-transfer rate is, LdX= Gd Y = kN( Y - Ye) d Z (23) where d Z is an incremental height in the adsorber, a is the external surface of the adsorbent particles and Ye is the equilibrium solute concentration in liquid corresponding to composition X The driving force, A Y = Y - Ye, is obtained from the vertical distance between the operating line and the equilibrium curve in Figure 10 The number of mass-transfer units, N, may be defined by equation (24) (24) G McKay 726 The number of mass-transfer units may be obtained by graphical integration and the height of the bed, 2, is obtained by combining equations (23) and (24) to give, l.0 dY kyadZ_ -2 Y - Ye G Ht where the height of the mass-transfer unit, H, is defined by Mass-transfer correlations, for evaluating k,, have been reviewed.44 Large-scale devices for the continuous contacting of granular solids and fluids have been developed after certain operational problems had been overcome These difficulties included obtaining uniform flow of solid particles and fluid without channelling or local irregularities, as well as those of introducing and removing the solid continuously into the vessel to be used Such devices include the hyper~orber~ and ~-~ the ~ Higgins contactor,*8~49the latter being specifically designed for solidliquid ion-exchange operations The previous systems utilise the principals of fluidisation and recent work on fluidisation has been applied specifically to colour removal 2.5 Fluidised beds In wastewater treatment systems it is advantageous to keep the particle size as small as possible, so that high rates of adsorption may be achieved Columnar operation, in which the solution to be treated flows upward through an expanded bed of the particulate adsorbent, is one method of taking advantage of small particle size The main problems of the use of small particles in fixed beds are excessive head loss, air-binding and fouling with particulate matter The design of a continuously operating counter-current fluidised bed is relatively ~imple,~O-52 although the solution of the design equations can become quite ~ o m p l e x ~ ~ - ~ ~ Determination of the mass-transfer coefficient, k,, provides the fundamental design parameter for fluidised-bed systems An intraparticle diffusion-controlled adsorption model has been prop0sed5~~59 and also a ‘completely mixed column’ model has been used for the removal of alkylbenzene sulphonate from wastewater,OO with a fluidised bed of carbon, and of basic dye from effluent,24 with a fluidised bed of silica By assuming steady-state conditions, the rate of mass transfer to the adsorbent across a film is, Rate= Driving force Resistance The driving force is the product of the concentration gradient between adsorbent and solution and the total area across which transfer is taking place The resistance is assumed to be uniform throughout the system and the reciprocal of this resistance is the transfer coefficient, k y Therefore r = k y A M Ay (28) where : A =effective transfer area per unit mass of adsorbent; M = mass of adsorbent; Ay =concentration difference across the film; r = rate of transfer Since the investigations24, 60 were performed under non-steady-state conditions, the rate of transfer may be expressed in tcrms of the concentration in solution at any time and the eon&khtkdtioh gradient across the film at that time Due to the strong affinity of the adsorbate for the adsorbent a partition coefficient, P,is incorporated to increase the concentration gradient term Its value is the mass concentration of adsorbate in solution, at time 0, divided by the mass concentration in the adsorbent at time 0, at equilibrium: P= Y equilibrium - Ys 727 Design models for adsorption systems in wastewater treatment where: Q=mass flow rate of solution; k,=mass-transfer coefficient per unit driving force ( y represents the instantaneous concentration gradient term and is difficult to measure experimentally) The term, ys, may be evaluated by graphical integration of the term (yo-y), since the time may easily be found from the experimental data: where W is total weight of adsorbent By rearranging equation (30): Qro Qr Pys=y+-kyAM kyAM ~ (32) and differentiating with respect to y : (33) Now substitution into equation (31): (34) Hence : (35) Integrating equation (35) gives: Defining E by equation (37): and rearranging equation (36) gives: yo-y=yoexp ( - E e l (38) Therefore a plot of In (1 -y/yo) against time will be a straight line of slope - E Final values of ky are then calculated by EWQ k '-A~(QP- E W ) (39) where AT= AM= total surface area of adsorbent The oyerdl height of a transfer unit, Ht, is then obtained from equation (40): Ht=- G kYaP (40) The Ht obtained is based on a 'single-resistance' model and neglects two additional processes, namely: (i) transfer between the fluidised bed and the wall (other particles); (ii) transfer between lean and dense phases in fluidised beds 50 128 C McKay Ht may be affected by several orders of magnitude based on the single-resistance model, but mathematical analysis is complex.61-68 Regeneration The economics of adsorption systems are very dependent on the ability to regenerate the adsorbent For many adsorbents regeneration by refractory techniques is possible and for carbon this results in a l0-15% loss of adsorbent.67 In all five adsorption systems discussed previously the size of the regeneration system depends on the carbon-use rate For batch adsorbers and fixed-bed column systems the mass of carbon per batch contractor/or column is needed and this value may be used to predict the minimum capacity of the furnace required For pulsed and other types of moving-bed contactors, the carbon usage rate may be estimated from the slope a to calculate a new slope a’ in the BDST equation (14), given by: where f is a moving-bed factor 20.30 The cost of regeneration is based on the capital cost of the furnace and associated equipment A typical system is shown in Figure 11 ; in addition, bins and conveying equipment are required To adsorber From adsor.her F Figure 11 Reactivation plant A, Spent carbon; B, dewatering screw; C, reactivation furnace; D, quench tank; E, reactivated carbon tank: FI,Fz, Fs, eductors Operating costs are based on labour, make-up carbon, power, fuel oil (or other heating medium) and steam if a stripping regeneration process is envisaged The heat input, HR, required for regeneration is based on equation (42) where HI is the thermal capacity of the furnace, H2 of adsorbent and H3 of retained fluid in particles Thermal capacities are based on a temperature drop, AT, which is the temperature of regeneration minus the adsorbent particle temperature Other conventional heat losses from the furnace and the flue gases will occur and a combustion efficiency for the fuel of 45 % has been reported.49 As shown in Figure 11 adsorbent is fed from the adsorbers to a spent carbon tank and from there via a dewatering screw-feed system into the furnace The reactivated adsorbent is quenched with treated water and transferred to a storage tank where it is held for re-use An alternative regeneration system utilising purge gas has been reported,6e in which an efficiency factor, vefr, for thermal regeneration has been correlated against the function, BHZ/cpgG, where BH is the overall particle heat transfer coefficient, Z is the packed-bed height, Cpg is the specific heat of gas and G’ is the superficial gas mass velocity Total regenerating time may be calculated from the heating and cooling cycle times and the carbonusage rate must be balanced against these to supply sufficient regenerated adsorbent Regeneration with solutions is often used, although the method produces a small additional effluent A major development described previously 2o is the implementation of biological regenera- Design models for adsorption systems in wastewater treatment 729 tion Certain adsorbents, such as peat and wood, are much cheaper than most of their competitors and, whereas they could not withstand higher temperature regeneration, the material could be burned and used as a source of low-grade heat Costing and optimisation From the basic design equations, the minimum costs for the system can be obtained by optimising the size of the adsorption and regeneration units Bed depth and flow rate determine residence time; the total costs, direct and indirect, may be plotted against residence time Figure 12 shows this effect and indicates a minimum total operating cost for a specific residence time u) t u) 0 -C t e 0) Direct cost CI Figure 12 Optimisation to determinz residence time Residence time Conclusions The various methods of solid-liquid contacting in adsorption systems have been considered In fluidised bed systems higher mass-transfer rates are obtained and smaller adsorbent particles may be used thus giving a large external surface area for adsorption The control of such systems is more difficult than in a fixed-bed column and there is a lack of flexibility in operation Fixed-bed adsorbers are most widely used and the effluent is continuously coming into contact with fresh adsorbent and therefore the driving force to reach equilibrium is high In batch processes a relatively long contact time is needed to reach equilibrium and a further solid-liquid separation stage is needed at the end of the process The economics of adsorption systems depend on the selection of the correct adsorbent, the method of contacting and the regeneration system Several methods of designing adsorption systems for wastewater treatment have been considered The design of the reactivation unit is presented and the technique of minimising the cost of adsorption plant has been discussed The procedures outlined adopt a ‘simplified’ approach to design in the sense that detailed kinetic and mass-transfer steps have not been incorporated into the models Nevertheless there is considerable evidence in the literature that the methods may be used successfully and should be used with some basic design ‘common sense’ It is usually beneficial in wastewater treatment plants to allow some overdesign to account for certain factors: (i) flow rate surges; (ii) changes in feed concentration; (iii) increased G.McKay 730 throughput requirements or increased levels of effluent quality; (iv) decreased adsorbent performance due to numerous reactivations References I 10 1I 12 13 14 IS 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 SO 51 52 53 54 55 56 Rizzo, J L Chem Engng Progr Symp Ser 1971, 67,466 MacCrum J M Proc Amer Assoc Text Chem Colorists Symp Atlanta, 1971, p 133 Molvar, A E.; Rodman C A.; Shunney, E L Text Chem Colourist 1970, 2, 36 Activated Carbon Reclaims Textile Industry's Waste Waters Enoiron Sci Tech 1969, 3, 314 Hassler, J W Activated Carbon, Chemical Publishing Co., Inc., New York, 1963 Evaluation of Granular Carbon for Chemical Process Applications D-116, ICI America, Inc., May, 1962 Euahtatioti of Granular Carbon in Columns for Wasrewater Treatment PC-3, ICI America, Inc., May, 1972 Davies, R A , ; Hartmut, J K.; Clemens, M M Chem Indy 1973, 1, 827 Fornwalt, H J.; Helbig, W A.; Scheffler, G H Br Chem Engng August, 1963 Fornwalt, H J.; Hutchins, R A Chem Engng April, 1966 Joyce, R S.;Valentine, A S Activated Carbon in Wastewater Treatment Seminar on Advanced Waste Treatment Research, Cincinnati, Ohio, 1962 Hager D G.; Fulker, R D War Treat Exam 1968, 17,41 Gauatlett, R @.;,Packham,R F Chem Indy 1973, 1, 812 W e e r >W J.; M,&nis,J C J Am 'Vht.W k s Ass 1964, 56, 447 Piihter, H A Chem In8y 1973, 1, 818 Bartell, F E.; Sloan, C K J Am Chem SOC.1929, 51, 1643 Gustafson, L Ind Engng Chem Prod Res Dew 1968, 7, 107 Oehme, C.; Martinola, F Chem Indy 1973, 1, 823 Robeck, G G.; Dostal, K A.; Cohen, J M.; Kreissl, J F J Am War Wks Ass 1965, 57, 181 Fram Corporation Bio-Regenerated Activated Carbon Treatment of Textile Dye Wastwater Wat Poll Contr Res Ser., Project No 12090 DWM, 1971 Mittelstaedt, R Tappi 1968, 51, 61A Pitman, R W.; Wells, C W J Am War W k s Ass 1968, 60, 1167 Middleton, A B Am DyestuflReptr August, 1971, 26 McKay, G.; Alexander, F Chem Engng 1977, 319,244 McKay, G.; Poots, V J P.; Alexander, F Ind Engng Chem Process Res Dev 1978, 17, 406 Mutch, N Q.J Pharm Pharmac 1946,19,490 Parsons, C L Min Technol Bull 71, US Bureau of Mines, 1913, p 32 Thornton, H A.; Moore, J R Sewage Ind Wastes 1953, 23,497 Williamson, J N.; Heit, A H.; Calrnon, C Public Health Services Publications, No 999-WP-14, vol 12, 1964 McKay, G.; Otterburn, M S.;Sweeney, A G J.S.D.C Aug, 1978, 357 Silvo, E Proc 4th Internat Peat Congr Otaniemi, Finland, Vol 4, 1972, 31 Tinh, V Q.; Le Blanc, R.; Janssens, J M.; Ruel, M Canadian Inst Min Metall Bull 1971, 64,99 Farnham, R S.:Brown, J L Proc 4th Infernat Peat Congr Otaniemi, Finland, Vol 4, 1972, 271 Ottemeyer, W Gesundh Ingr 1930, 53, 185 Poots, V J P.: McKay, G.; Healy, J J Water Res 1976, 10, 1061 Poots, V J P., McKat, G.; Healy, J J Scient Proc Roy Dublin Soc., Ser A 1978, 6, 61 Leslie, M Am Dyesrufls Reptr 1974, 63, 15 Poots, V J 0.; McKay, G Proc R Irish Acad 1977, 47, 539 Poots, V J P.; McKay, G.; Healy, J J Water Res 1976, 10, 1067 Poots, V J P.; McKay, G.; Healy, J J W.P.C.F May, 1978, 926 Collins, J J Chem Engng Prog Symp Series No 74 Michaels, A S Ind Engng Chem 1952.44, 1922 Rosen, J B Ind Engng Chem 1954, 46, 1590 Heister, N K.; Vermeulen, T.; Klein, G Chemical Engineers Handbook (Perry, R H., Ed.), McGraw-Hill, 1977 Lukchis, G M Chem Engng 1973,80, I I Lukchis, G M Chem Engng 1973,80,83 Mair, J B.: Westhaver, J W.; Rossini, F D Ind Engng Chem 1950, 42, 1279 White, G E Chem Engng Prog 1950, 46, 363 Johnston, W A Chem Engng 1972,79,87 Allen, H V Petr ReJ July, 1944 Strain, H H Ind Engng Chem 1950, 42, 1307 Johnston, R L.;Baumann, E R J W.P.C.F 1971, 43, 1640 Berg, C Trans Am Ind Chem Engng 1946,42,665; Chem Engr Prog 1951,47 585 Kehde, H.; Faifield, R G.; Frank, J C.; Zahnstecker, L W Chem Engng Prog 1948,44, 575 Swinton, E A.; Weiss, D E.; etal Austral J appl Sci 1953, 4, pp 316,510,519, 530, 543 Hancher, C W.; Jury, S H Chem Engng Prog Symp Ser 55 N o 24, 1959, p 87 Deaip models for adsorption systems in wastewater treatment 57 58 59* 60 61 62 63 64 65 66 67 68 69 731 Higgins, R.; Roberts, J T Chem Engng Prog Symp Ser 50 No 14, 1954, p 87 Anon Chem Engng April 15, 1963,10, 92 Errnenc, E D Chem Engng 1961, 68,87 Weber, W J.; Morris, J C J War Poll Control Fed 1965, 31, 425 Etherington, L D.; Fritz, R J.; Nicholson, E W.; Scheeline, H W Chem Engng Prog 1956, 52, 274 Leva M Fluidination McGraw-Hill New York 1959 Zenz, F A In Modern Chemical Engineering (Acrivos, A., Ed.), Reinhold Publishing Corporation, New York, 1963 Zenz, F A.; Othrner, D F Fluidisation and Fluid Particle Sysrems Reinhold Publishing Corporation, New York, 1960 Davidson, J F.; Harrison, D Fluidisation Academic Press, London and New York, 1971 Kunii, D.; Levenspiel, Fluidisation Engineering J Wiley and Co., New York, 1969 Rosen, J B J Phys Chem 1952, U ,387 Amundsen, N R.; Karten, P R.; Lapidus, L J J Phys Chem 1952, 56, 683 Culp, G L.; Culp, R L New Concepts in Water Purification Van Nostrand Reinhold Environmental Engineering Series, New York [...]... systems for wastewater treatment have been considered The design of the reactivation unit is presented and the technique of minimising the cost of adsorption plant has been discussed The procedures outlined adopt a ‘simplified’ approach to design in the sense that detailed kinetic and mass-transfer steps have not been incorporated into the models Nevertheless there is considerable evidence in the literature... adsorbent By rearranging equation (30): Qro Qr Pys=y+-kyAM kyAM ~ (32) and differentiating with respect to y : (33) Now substitution into equation (31): (34) Hence : (35) Integrating equation (35) gives: Defining E by equation (37): and rearranging equation (36) gives: yo-y=yoexp ( - E e l (38) Therefore a plot of In (1 -y/yo) against time will be a straight line of slope - E Final values of ky are...727 Design models for adsorption systems in wastewater treatment where: Q=mass flow rate of solution; k,=mass-transfer coefficient per unit driving force ( y represents the instantaneous concentration gradient term and is difficult to measure experimentally) The term, ys, may be evaluated by graphical integration of the term (yo-y), since the time may easily be found... been considered In fluidised bed systems higher mass-transfer rates are obtained and smaller adsorbent particles may be used thus giving a large external surface area for adsorption The control of such systems is more difficult than in a fixed-bed column and there is a lack of flexibility in operation Fixed-bed adsorbers are most widely used and the effluent is continuously coming into contact with... therefore the driving force to reach equilibrium is high In batch processes a relatively long contact time is needed to reach equilibrium and a further solid-liquid separation stage is needed at the end of the process The economics of adsorption systems depend on the selection of the correct adsorbent, the method of contacting and the regeneration system Several methods of designing adsorption systems for. .. rate determine residence time; the total costs, direct and indirect, may be plotted against residence time Figure 12 shows this effect and indicates a minimum total operating cost for a specific residence time u) t u) 0 0 -C t e 0) 2 Direct cost CI 0 Figure 12 Optimisation to determinz residence time Residence time 5 Conclusions The various methods of solid-liquid contacting in adsorption systems have... Engng August, 1963 Fornwalt, H J.; Hutchins, R A Chem Engng April, 1966 Joyce, R S.;Valentine, A S Activated Carbon in Wastewater Treatment Seminar on Advanced Waste Treatment Research, Cincinnati, Ohio, 1962 Hager D G.; Fulker, R D War Treat Exam 1968, 17,41 Gauatlett, R @.;,Packham,R F Chem Indy 1973, 1, 812 W e e r >W J.; M,&nis,J C J Am 'Vht.W k s Ass 1964, 56, 447 Piihter, H A Chem In8 y 1973, 1, 818... successfully and should be used with some basic design ‘common sense’ It is usually beneficial in wastewater treatment plants to allow some overdesign to account for certain factors: (i) flow rate surges; (ii) changes in feed concentration; (iii) increased G.McKay 730 throughput requirements or increased levels of effluent quality; (iv) decreased adsorbent performance due to numerous reactivations References... Industry's Waste Waters Enoiron Sci Tech 1969, 3, 314 Hassler, J W Activated Carbon, Chemical Publishing Co., Inc., New York, 1963 Evaluation of Granular Carbon for Chemical Process Applications D-116, ICI America, Inc., May, 1962 Euahtatioti of Granular Carbon in Columns for Wasrewater Treatment PC-3, ICI America, Inc., May, 1972 Davies, R A , ; Hartmut, J K.; Clemens, M M Chem Indy 1973, 1, 827 Fornwalt,... regenerating time may be calculated from the heating and cooling cycle times and the carbonusage rate must be balanced against these to supply sufficient regenerated adsorbent Regeneration with solutions is often used, although the method produces a small additional effluent A major development described previously 2o is the implementation of biological regenera- Design models for adsorption systems in wastewater