You are on page 1of 29

Laboratory evaluation of activated carbon for liquid phase applications

Laboratory evaluation of activated carbon for liquid phase applications

Corporate Headquarters 400 Calgon Carbon Drive Pittsburgh, PA 15205, USA Tel +1 (0) 412 787 6700 Fax +1 (0) 412 787 6713

European Operations of Calgon Carbon Corporation Zoning Industriel C de Feluy B-7181, Feluy, Belgium. Tl +32 (0)64 51 18 11 Fax +32 (0)64 54 15 91

www.chemvironcarbon.com

Preface
We have developed this book based on our experience in activated carbon, acquired since the early years of the 20th century. This booklet enables customers to run tests under suitable conditions and to evaluate the test results scientifically. This work is intended as a guideline and not as a substitute for personal contact with our experienced technical staff.

Table Of Contents
1 1.1 1.2 1.3 1.4 1.5 Introduction History Activated Carbon Forms of Activated Carbon Adsorption Recycling by Thermal Reactivation 6 6 8 9 10 12 3.3 3.4 3.5 3.2.4 3.2.5 3.2.6 Pump Flow Meter and Totaliser Connections 35 35 35 35 36 38 38 38 39 39 39 40 42 43 46 47 48

Liquid Pre-treatment Carbon Pre-treatment Column Operation 3.5.1 Temperature Linear Velocity of the Liquid in the Bed Duration of Column Testing Displacement, Sampling and Analysis

2 2.1 2.2

Isotherms Theory Experimental 2.2.1 2.2.2 2.2.3 General Description Selection of Experimental Conditions Test Procedure

13 13 14 14 15 17 19 19 20 23 25 4 3.7 3.8 3.9 3.10 3.6

3.5.2 3.5.3 3.5.4

Interpretation of Pilot Data 3.6.1 3.6.2 Mass Transfer Zone Determination Minimum Carbon Consumption

2.3

Interpretation of the Isotherm 2.3.1 2.3.2 2.3.3 Linear Isotherm Comparison of Different Isotherms Non-Linear Isotherms

Fixed Bed: Single, Series or Parallel Pulse Bed Accelerated Column Test (ACT) Sizing of the Adsorption System

2.4

Summary

Conclusion Appendix 1 : Format For Reporting Isotherms Data Appendix 2: Format For Reporting Pilot Data Glossary of terms commonly used in Activated Carbon Technology

49 50 51 52

3 3.1 3.2

Adsorption Studies in Columns Introduction Experimental 3.2.1 3.2.2 3.2.3 Test Location Selection of Equipment Columns

26 26 28 28 28 32 5

Headquarters and Technical Department

55

1 Introduction
1.1 History

Egyptian Hieroglyphics

Adsorption on porous carbons was described as early as 1550 B.C. in an ancient Egyptian papyrus, and later by Hippocrates and Pliny the Elder, mainly for medicinal purposes. In the 18th century, carbons made from blood, wood and animals were used for the purification of liquids. All of these materials, which can be considered as precursors of activated carbons, were only available as powders. The typical application was batch contact treatment where a measured quantity of carbon and the liquid to be treated were mixed and, after a certain contact time, separated by filtration or sedimentation. At the beginning of the 19th century, the decolourisation power of bone char was detected and used in the sugar industry in England. Bone char was available as a granular material which allowed the use of percolation technology, where the liquid to be treated was continuously passed through a column. Bone char, however, consists mainly of calcium phosphate and a small percentage of carbon; this material was only used for sugar purification.

In the early years of the 20th century, the first processes were developed to produce activated carbons with defined properties on an industrial scale. However, with both manufacturing processes, i.e. steam activation (V. Ostreijko, 1900 and 1901) and chemical activation (Bayer, 1915), only powdered activated carbon could be produced. During the First World War, steam activation of coconut char was developed in the United States for use in gas masks. This activated carbon type contains mainly fine adsorption pore structures suited for gas phase applications. Carbons with more transport pore structures, necessary for decolourisation of liquids, were only available as powders until the middle of the 20th century. After World War II, Calgon Carbon Corporation (the parent company of Chemviron Carbon) succeeded in developing coal-based, granular activated carbons with a substantial content of transport pore structure and good mechanical hardness. This combination allowed the use of activated carbon in continuous decolourisation processes, which, for many end users, such as sugar decolourisation, is an ideal combination.

In addition, Calgon Carbon Corporation and Chemviron Carbon pioneered a great deal of the work on the optimisation of granular carbon reactivation. Today many users are switching from the traditional use of powdered activated carbon as a disposable chemical to continuous adsorption processes, and using granular activated carbon plus reactivation. In doing this they are moving towards a more sustainable process that recycles and minimises waste.

1.2 Activated Carbon


Activated carbon is a porous material made up of a random structure of graphite platelets. It is produced from various carbonaceous materials such as different grades of coal, coconut shells, wood and peat. The majority of activated carbons are produced from coal. When these materials are chemically or thermally treated, changes occur in the precursors original structure. When sufficient thermal treatment is applied, the carbon atoms from the precursor begin to realign and form plate-like structures similar to graphite, however, they do not form the extended, uniformly packed structure that graphite has. Figure 1 and Figure 2 provide schematic diagrams of graphite and activated carbon. A review of these diagrams reveals the ordered structure of graphite and the limited ordering in activated carbon. Note that the pores in activated carbon are cracks and crevices.

Aliphatic dislocation of platelet


Graphic plates

3.354 ngstrms interlayer distance

Graphite platelet

100 ngstrms

Inter bonding of plates

Figure 1: Graphite structure

Figure 2: Activated carbon structure

The carbon precursor usually has an inherent pore structure associated with it. The activation process enlarges these pores, which ultimately provide channels for the adsorbates to penetrate into the activated carbon. When an adsorbate finds an area where the graphite platelets intersect, or are close to each other, adsorption occurs. Where the platelets intersect are the micropores (adsorption pores) and the enlarged pores are the macropores (transport pores). Adsorption only occurs in the micropores (high energy sites). Due to its graphite platelet structure, activated carbon has a very large surface area which gives it excellent properties of adsorption of organic compounds. These are removed from solution and fixed on the activated carbon surface by adsorption forces.

Activated Carbon Cloth - Activated Carbon is also available in special forms such as cloth and fibres.

Briquettes - Activated carbons can be agglomerated and formed into a variety of briquettes.

1.3 Forms of Activated Carbon


There are a number of different forms of activated carbon as described below. The form will affect the tests used in its characterisation. Granular Activated Carbon (GAC) - Irregular shaped particles with sizes ranging from 0.2 to 5 mm. This type is used in both liquid and gas phase applications.

The customer can have any colour he wants so long as its black Henry Ford - For some applications it is now possible to produce activated carbon in a range of different colours.

Granular and pellet activated carbons are used in fixed beds with the liquid or gas passing through the bed. For powder activated carbon, the powder is added to the liquid stream, mixed and then removed by filtration or decantation. Activated carbon is also available in special forms such as cloth and fibres.

1.4 Adsorption
Powder Activated Carbon (PAC) - Pulverised carbon with a size predominantly less than 0.18 mm (US Mesh 80). These are mainly used in liquid phase applications and for flue gas treatment. Adsorption is a naturally occurring phenomenon in which molecules of a liquid or gas are attached to, and trapped by, either an external or internal surface of a solid. Within a solid like activated carbon, each molecule is aligned with another so that equal forces act upon it, resulting in a balanced state. The molecules on the surface are not completely surrounded by other molecules and, consequently, are not in a balanced state. As those molecules seek to reach a balanced state, surface tension or energy is created. This surface energy is an attractive force and, if it is strong enough to overcome the energy of other passing molecules, they will adhere to, or be adsorbed onto, that surface. These forces are known as "London Dispersion Forces" (see Figure 3).

Extruded Activated Carbon - Extruded and cylindrical shaped pellets with diameters from 0.8 to 5 mm. These are mainly used for gas phase applications because of their low pressure drop, high mechanical strength and low dust content.

10

Additivity of London Dispersion Forces (10,000,000 x Magnification) The adsorption force present at the adsorption site is the sum of all the individual interactions between carbon atoms and the adsorbate molecules.
Carbon Atom

Adsorbate Molecule

As a result of this high adsorption capacity, activated carbon can be used to purify liquids in any part of plant operations. It is used for raw material purification to improve final product quality, or to treat effluent to remove undesirable compounds. The efficiency of powdered activated carbon used in batch processes can be easily predicted from pre-trials which require only a short time. In most cases, it can be immediately scaled up from the laboratory to the industrial equipment. The adsorption kinetics on granular activated carbon are primarily regulated by diffusion processes. Percolation tests with the granular material should, therefore, be conducted and evaluated using certain scientific procedures in order to obtain suitable information for the design of an industrial plant.

Figure 3: Relation between structure and adsorption force

In simpler terms, for example, if you have a contaminant in water and the energy exerted by the activated carbon for adsorption is stronger than the affinity of the contaminant and the water, then the contaminant will stick to the surface of the carbon in preference to remaining in the water.

1.5 Recycling by Thermal Reactivation


Once granular activated carbon is saturated or the treatment objective is reached, it can be recycled by thermal reactivation for reuse. Reactivation involves treating the spent carbon in a high temperature reactivation furnace to over 800C. During this treatment process, the undesirable organics on the carbon are thermally destroyed. Recycling by thermal reactivation is a highly skilled process to ensure that spent carbon is returned to a reusable quality. Chemviron Carbon operates Europes largest reactivation facilities and recycles daily large quantities of spent carbon for a diverse range of customers. Recycling activated carbon by thermal reactivation meets the environmental need to minimise waste, reducing CO2 emissions and limiting the use of the worlds resources.

Fine transparent pore

Aliphatic dislocation of platelet

Adsorbate in a liquid-like state

Graphite platelet

100 ngstrms

Inter bonding of plates

Figure 4: Molecular structure of a loaded activated carbon

Feluy reactivation facility

11

12

2 Isotherms
2.1 Theory
An adsorption isotherm is a relatively simple method for determining the feasibility of using granular activated carbon for a particular application. A liquid phase isotherm shows the distribution of impurities between the adsorbed phase and the bulk solution at equilibrium. It is a plot of the amount of impurity adsorbed per unit weight of carbon versus that remaining in the solution (Figure 5).

Taking logarithms, we obtain: log X/M = log k + 1/n log C This is the equation of a straight line whose slope is 1/n and whose intercept is k at C = 1. Therefore, if X/M is plotted against C on log-log paper, a straight line should be obtained. There are, however, occasions when deviations will occur; for example, isotherms of liquids containing multi-components may have a line with multiple slopes.

2.2 Experimental
2.2.1 General Description Liquid phase isotherms are normally based on a Freundlich isotherm which is an adsorption test showing the relationship between the residual concentration of a compound against the loading of the compound on the carbon. This is sometimes referred to as the liquid phase concentration versus the solid phase concentration. First, the carbon is ground up to reduce kinetic effects allowing the carbon to reach equilibrium as quickly as possible. Thereafter, increasing amounts of carbon are added to different flasks containing fixed volumes of the liquid to be treated. One of the flasks does not contain carbon - this is the blank. The principle of the isotherm is illustrated in Figure 6.

100

(X/M) co

Loading X/M (mg/g)

10

0.1

10

C0

100

Concentration C (mg/l)

Figure 5: Adsorption isotherm

Generally, straight line plots can be obtained by making use of the empirical Freundlich equation, which relates the amount of impurity in the liquid phase to that amount adsorbed onto the activated carbon by using the expression: X/M = kC1/n where
Increasing dosage of carbon Blank

X/M = loading, is the amount of impurity adsorbed (X) per unit weight of carbon (M); C = equilibrium concentration after adsorption; k, n = constants.

Figure 6: The Freundlich Isotherm procedure

13

14

The flasks are then stirred, normally overnight, at a constant temperature to reach equilibrium. The carbon is then filtered from the sample and the remaining liquid analysed for the concentration of the compound being adsorbed. The concentration in each sample and the blank is then determined. From these measurements, all the values necessary to plot an isotherm may be calculated (Appendix 1) and plotted onto a log-log graph. The Freundlich Equation assumes a straight line, but in practice a curved or unusually shaped line may result. An isotherm can be carried out either using a sample of liquid to be treated (containing the contaminant), or a synthetic isotherm is carried out using distilled water, spiked with the contaminant. Though an isotherm is useful for indicating whether or not a substance is adsorbable and can be used to estimate the consumption of activated carbon, it does not give required design information such as the contact time. 2.2.2 Selection of Experimental Conditions There are a number of variables that can affect the efficiency of adsorption when performing isotherms. Adsorbate solubility and liquid viscosity are key parameters in carbon adsorption, and both are affected by temperature. In order to obtain data that is relevant to the process, it is essential to run the isotherm under the same conditions as that of the process. As described previously, a liquid phase isotherm shows the distribution of impurity at equilibrium. In order to obtain equilibrium, the carbon and liquid must be in contact for sufficient time, preferably overnight for water and longer for liquids with higher viscosities. The solution should be tested at a pH which is consistent with the process. However, it may be worthwhile to evaluate the effect of pH and determine if the carbon loading is increased sufficiently to warrant changing the pH in the process. The solution pH may change by coming into contact with the carbon, either as a result of the dissolving of a portion of its ash, or by preferential adsorption of acidic or alkaline components. Since some impurities in solutions are sensitive to pH changes, it is imperative that the final pH should be adjusted to the same value as the untreated liquid before any concentration measurements are made.

It is known that the kinetics of adsorption onto activated carbon is controlled by the process of diffusion. The transfer of the impurities from the bulk solution to the internal surface of the carbon proceeds through three stages as indicated in Figure 7:
mm

10 nm

Bulk diffusion

Internal diffusion

micron

1 nm

Film diffusion
Figure 7: Diffusion in an activated carbon particle

Adsorption

1) Bulk diffusion of the compound from the liquid to the film around the carbon particle 2) Diffusion through this surface film and 3) Diffusion through the internal structure to the adsorption sites in the carbon. It is known from experiments that step 2 is normally the rate determining step. In order to increase the kinetics of the rate determining step, it is necessary to pulverise the granular carbon for testing by means of a ball mill or other suitable mechanical device so that 95 % will pass through a 45 m (325 US Mesh) screen. Carbon dosages are dependent upon the quantity of impurities present in the liquid and sufficient dosages should be used to establish an accurate curve. Experience has shown that dosages ranging from 0.1 g/l to 10 g/l are sufficient for most applications.

15

16

Stoppered Erlenmeyer flasks serve as satisfactory containers for the carbon solution mixture. Manual agitation, even at frequent intervals, is usually inadequate and a mechanical shaker or magnetic stirrers are preferred. After the required contact time has elapsed, the carbon must be removed from the liquid prior to analysis. This can be performed using either vacuum or pressure filtration. 2.2.3 Test Procedure 1. Pulverise a representative sample of the appropriate granular activated carbon (a 10-20 g sample is usually adequate) so that 95 % will pass through a 45 m screen (325 US Mesh), or use material pre-prepared and supplied by Chemviron Carbon. Oven dry the pulverised sample for three hours at 150C. 2. Obtain a representative sample of the liquid to be tested. Any suspended matter should be removed by filtration using a glass microfiber filter (grade GF/C). pH adjustment, if required, should be performed on the sample before filtration. 3. Transfer a minimum of six different weights of the oven-dried pulverised sample to separate Erlenmeyer flasks. Sample weights should be chosen in accordance with the carbon dosage suggestions given above. 4. To each container add 100 ml of test solution (V) from a graduated cylinder, and clamp the flasks on the shaker. If a constant temperature bath is used, the shaker should be so designed as to permit immersion of the lower part of the flask containing the carbon solution mixture in the bath. 5. Agitate the mixture for a minimum of four hours, preferably overnight, in some instances the equilibrium agitation time may exceed 24 hours. If the sample contains a volatile constituent, special care should be taken during manipulation to avoid loss of the constituent. The same volume of solution should be added to a container without carbon and subjected to the same procedure in order to obtain a blank reading. 6. After the chosen contact time has elapsed, filter the contents of the flask. One should discard the first and last portions of the filtrate and save only the middle portion for analysis. The blank should be filtered in the same manner as the other samples.

7. Determine the residual concentration of the contaminant in the filtrate by the relevant analytical technique. 8. Tabulate the data as shown in Table 1. The residual solution concentration, C, is obtained directly from the filtrate analysis. The amount adsorbed on the carbon X is obtained by subtracting the value of CV from the value of CoV, the influent concentration. The amount of constituent adsorbed per unit weight of carbon, X/M is calculated by dividing X by the weight of carbon M used as follows: X/M = (CoV - CV)/M 9. On log-log paper plot C, on the horizontal axis, against X/M, on the vertical axis, and draw the best straight line through the points, as illustrated in Figure 5.

M Carbon dosage remaining in solution (g/l) 0 0.1 0.2 0.5 1 2 5 10


Table1: Isotherm data table

C Concentration adsorbed (mg/l) 18.15 15.90 14.12 10.25 7.15 3.93 1.80 0.95

X = (Co - C)V Constituent adsorbed per unit weight (mg) 2.25 4.03 7.90 11.00 14.22 16.35 17.20

X/M Constituent

(mg/g) 22.50 20.15 15.80 11.00 7.11 3.27 1.72

17

18

2.3 Interpretation of the Isotherm


2.3.1 Linear Isotherm If a vertical line is drawn (Figure 5) from the point on the horizontal axis corresponding to the influent concentration (Co), and the best line through the data is extrapolated to intersect this (Co) line, the loading X/M of the activated carbon can be read from the vertical scale. This value, termed (X/M)Co represents the amount of impurity adsorbed per unit weight of carbon when the carbon is in equilibrium with the influent concentration. Since this equilibrium can be reached in a properly designed system, it represents the ultimate adsorption capacity of the carbon. From the value of (X/M)Co, the carbon dosage can be calculated for complete removal of impurities: Theoretical carbon dosage = C o /(X/M)Co where theoretical carbon dosage = theoretical amount of carbon used per volume of liquid; Co = influent concentration; (X/M)Co = extrapolated loading per unit weight of carbon at the influent concentration C o. The following is an example for a capacity calculation: If we plot the isotherm as shown in Figure 5 and extrapolate the data, we obtain a loading (X/M)Co = 26.5 mg/g at an influent concentration Co = 18.15 mg/l. From these figures, the theoretical granular carbon dosage or usage rate to treat the liquid may be obtained. In this example, the grammes of carbon required to treat one litre of solution are: Theoretical carbon dosage = 18.15 mg/l = 0.685 g carbon/l solution 26.5 mg/l However, there are some applications where the carbon dosage will be higher than this value because the carbon which is removed from the system will not be fully saturated. Note that if we use a powdered activated carbon for the same application, we can see from Table 1 that we need a carbon dosage of 5 g/l to obtain only 90 % purification.

2.3.2 Comparison of Different Isotherms The performance of two carbons for the same application can easily be compared by the examination of isotherms obtained under identical conditions. Isotherms obtained for two carbons (A and B) using the same test solution are plotted in Figure 8. The numbers adjacent to the experimental points in Figure 8 indicate the grammes of carbon used in the test and the relative performances of the two carbons.

100

(X/M) co

10

Loading X/M (mg/g)

5 g GAC/I

5 g GAC/I

1 0.1

10

C0

100

Concentration C (mg/l) Carbon A Carbon B

Figure 8: Parallel isotherms

The two carbons can be compared at any desired degree of purification by drawing a vertical line for that concentration from the horizontal axis and reading the X/M values corresponding to the points of intersection of that line with the isotherms. Such a line has been drawn at Co influent concentration. On Figure 8, the following X/M values are obtained: (X/M)A = 26.5 mg/g (X/M)B = 41 mg/g The relative efficiency (RE) of carbon A with respect to carbon B can be calculated from the equation: % RE = 26.5 mg/g x100 = 64.6 % 41 mg/g

19

20

Parallel isotherms were obtained in the case illustrated in figure 8. Consequently, the relative efficiency will be independent of the purification achieved. However, if the lines are not parallel, the relative efficiency of each carbon will vary depending on the treatment objective required. It should be emphasised that misleading results can be obtained by making the erroneous assumption that two carbons can be compared for column application by evaluating only the purification achieved by equivalent dosages. In the example (Figure 8), 5 grammes of carbon B achieve 93 % removal. The same amount of carbon A achieves 90 % removal. On this basis, the two carbons appear almost equally efficient. However, the relative efficiency shows that in a well designed column system carbon B will have a capacity X/M about one and a half times greater than that of carbon A at any degree of purification. In other cases, the isotherms for two carbons will cross as illustrated in Figure 9.

At concentrations higher than C1, carbon A has a greater capacity than carbon B; at lower concentrations, the reverse is true. Carbon A would be preferable for column operation because of its higher capacity at influent concentration. It should be noted, however, that for batch treatment, carbon B would be preferable. In general, the steeper the slope of the isotherm exhibited by a particular carbon, the greater will be its efficiency in column operation. This example also illustrates the necessity of obtaining isotherms that cover the entire concentration range before comparing two carbons for a particular application.

100

(X/M) co

Loading X/M (mg/g)

10

1 0.1

C1

10

C0

100

Concentration C (mg/l) Carbon A Carbon B

Figure 9: Crossed isotherms

21

22

2.3.3 Non-Linear Isotherms It was previously stated that, ideally, straight line isotherm plots are obtained. However, there may be departures from linearity.

(c)
100

Loading X/M (mg/g)

A curve similar to that shown in Figure 10a may be obtained if a non adsorbable impurity is present in the liquid being treated. Subtracting C, concentration of non adsorbable compounds, from C0 and replotting, the isotherm will usually yield a straight line.
(a)
100

10

1 0.1

10

100

Concentration C (mg/l)
Loading X/M (mg/g)

10

(d)
100

Concentration C (mg/l)

Loading X/M (mg/g)

1 0.1

10

100

10

(b)

100

1 0.1

10

100

Concentration C (mg/l)
Loading X/M (mg/g)

Loading Z

Figure 10: Non-linear isotherms


10

1 0.1

10

100

Concentration C (mg/l)

A sudden change in slope, as illustrated in Figure 10b, could indicate that at least two different types of adsorbates are present, showing markedly different capacities. The plot in Figure 10c, similar to Figure 10b, illustrates a situation in which at least three types of adsorbates are present, e.g. undefined "colour", but when all are adsorbed similarly, straight line isotherms should be obtained. The curve shown in Figure 10d indicates that the adsorbate has reached its maximum surface coverage at loading Z.

23

24

2.4 Summary
From the isotherm adsorption capacity, an estimate of the granular carbon usage rate necessary to meet the treatment objective can be obtained. Isotherm tests also afford a convenient method for comparing different carbons and for investigating the effects of pH and temperature. It is essential to realise that the lowest possible carbon usage rate is predicted by a properly run isotherm test. If a lower usage is observed in column tests or in the plant, then one or more of the following is true: the isotherm test was not at equilibrium volatile components were lost during the isotherm trial the interpretation of the isotherm was done incorrectly there was an error in the analytical techniques carbon fines may not have been completely removed from the solution the influent liquids tested in the isotherm and column tests were not identical there are micro-organisms present on the carbon in the column which are causing biological degradation of the adsorbed organic substances compounds, thereby reducing the carbon usage the pH of the isotherm solution may have been altered the isotherm was run under different process conditions compared to the actual process.

3 Adsorption Studies in Columns


3.1 Introduction
If an isotherm study indicates that the liquid can be treated to the desired purity level using granular activated carbon at a reasonable dosage, then the next step is to evaluate the carbon treatment in a dynamic test. The purpose of the column test is to obtain the "breakthrough curves under dynamic conditions, which show how the concentration of the solutes in the effluent will vary with the volume of liquid treated. This data will allow Chemviron Carbon to establish the design criteria for a full scale plant: the effective capacity of the granular activated carbon in an operating system the shape of the mass transfer zone and the bed depth to be taken into account for optimum adsorption the contact time necessary to achieve the treatment objective in the most economical manner influence of suspended solids or gas formation on the operating conditions for the adsorption columns colloidal suspended solids will have to be removed prior to the adsorbers or the system will have to use the granular activated carbon column as a mechanical filter as well and include a backwash sequence in the operation cycle. If gas evolution occurs, the adsorption system must be designed to prevent formation of gas pockets within the carbon bed. Depending on the application and flow, different types of adsorbers are used. Therefore, certain additional information must be obtained before the column test can be started. The technical department of Chemviron Carbon has extensive experience in purification of liquids and is able to provide advice on the type of adsorber to be used. In order to do this evaluation, the operating capacity as well as the desired flow rate and bed depth must be established to determine the optimum dimensions and the number of columns necessary for continuous treatment.

25

26

The rate at which the solutes are adsorbed by the carbon can only be determined by dynamic column tests. This is the key to the system design. Ideally, it would be desirable to be able to use mathematical techniques to predict the carbon performance from the equilibrium data. There have been many studies conducted and models developed, however, they all have their limitations. The studies used to develop these models were based on the presence, or assumed presence, of one or more of the following elements: uniform carbon particle size ideal isotherm dilute solutions single component adsorbate molecule was too large to enter the fine pore structure, so only film diffusion was controlling the kinetics flow rate was high so pore diffusion was controlled comparisons of predicted versus actual data were made on early part of breakthrough curve before carbon was saturated systems with relatively short mass transfer zones (MTZ) were studied. Until such time as a universal model is developed, it is still necessary to test in a dynamic system and use empirical methods to design the plant scale units. Prior to conducting a column test many aspects must be considered. Some of these are listed below and then discussed separately: location of the test size and type of pilot system column, diameter, number of columns, fixed or pulse bed carbon type and particle size linear velocity of liquid in the carbon bed temperature pH treatment objective presence of suspended solids heat of adsorption for organics during the wetting.

3.2 Experimental
3.2.1 Test Location The tests should be conducted at the plant if any of the following conditions exist: the liquid deteriorates or changes in character during shipment and/or storage the carbon dosage is so low that large quantities of liquid would be required to conduct the test the conditions of the liquid stream being treated vary widely. 3.2.2 Selection of Equipment The pilot test can be carried out in either fixed beds or in a single pulse bed column. Depending on the application, it should be discussed which type of adsorber should be considered. A simple way to conduct a pilot test is by using fixed bed columns. From this test we can obtain design data for either a fixed bed or a pulse bed system. 3.2.2.1 Fixed Bed Systems In the fixed bed vessel (Figure 11), the solution enters the top of the vessels and passes through the carbon until the carbon is spent or the outlet specification is exceeded. Then, the entire volume of carbon is removed from the vessel and replaced with either virgin or reactivated carbon. The fixed bed system will have fluctuation in the outlet quality. Initially, the carbon depth is sufficient to contain the mass transfer zone. However, as the carbon becomes spent the remaining carbon is not sufficient to remove all the impurities and some of them begin exiting the vessel; breakthrough occurs.

Figure 11: Fixed bed adsorber

27

28

fluctuations in the inlet contaminant concentration while maintaining a consistent concentration in the outlet. Periodically, the flow of liquid is stopped. Spent carbon is withdrawn (pulsed) from the bottom and virgin or reactivated carbon is added into the top of the adsorber.
Figure 13: Pulse bed adsorber
Time on stream or volume treated Fixed bed outlet concentration Treatment objective
Concentration

Figure 12: The Pulse Bed System

The pulse bed system (Figure 13) provides the best control over the outlet quality compared to any available technology. It is operated in a true counter current fashion. The liquid to be treated enters the bottom of the column, passing through the most spent carbon first. As the liquid rises through the column it is purified; exiting the column at the top where the freshest carbon resides. Because the adsorption process is equilibrium based, this operation best utilises the carbon adsorption capacity. Carbon which has been in equilibrium with a lower impurity concentration (in the top portion of the column) can come to a new equilibrium with a liquid (in the lower portion of the column) having a higher concentration than the original. This enables carbon to better accommodate

Concentration

3.2.2.2 Pulse Bed Systems

Time on stream or volume treated Pulse bed outlet concentration Treatment objective

Figure 14: Pulse bed adsorber

29

30

3.2.2.3 Fixed Bed vs. Pulse Bed System The choice of system depends on many factors. Some of the advantages and disadvantages of each system are given below.

3.2.3 Columns The pilot unit consists of either fixed beds in series or a single pulse bed column. Usually, full scale granular activated carbon beds are from 1.0 to 10 metres in depth and from 0.3 to 4.0 metres in diameter. In a laboratory, the diameter of the columns can be scaled down to 50 mm; a smaller diameter should not be selected in order to avoid excessive wall effects. By an appropriate reduction in flow rate, plant conditions can be simulated. The carbon bed depth or superficial contact time, should be the same as, or greater than the full scale system. If a pulse bed pilot column is used, full size plant conditions are rigorously simulated. Strictly speaking, fixed beds in series do not reproduce pulse bed conditions, however data obtained from such pilot systems can provide a reliable basis for design. Another way to conduct a column test is to use one of Chemviron Carbon's range of mobile adsorbers (see Figure 15), from the Cyclesorb series of vessels. They offer several advantages compared to traditional pilot columns.

Advantages Fixed Beds Lowest building profile Little operator attention required Self-filtering system Easy to inspect vessel interiors Requires lower inlet pressure to pass through the system Can handle large flow Can be designed for backwashing.

Disadvantages Requires greater plot area Higher capital investment than a pulse bed system for the equivalent amount of carbon on stream If a single adsorber is used effluent quality will vary as a function of the on-stream time of the column The treatment objective is reached at the same time as the column is disconnected from the system. Cannot treat liquids containing significant quantities of suspended solids Inspection and/or repairs can only be done by taking an adsorber off line and emptying the entire carbon content Effluent contains carbon fines after each pulsing operation and should be filtered.

Pulse Beds

Requires less plot area Can be controlled to have the effluent quality close to the specification value Has the lowest carbon dosage in applications having a long mass transfer zone No peak consumption of water during the recovery of product If gas is present, or can be generated in the carbon bed, an up-flow design is the better alternative.

Figure 15: Cyclesorb MA20 adsorbers in operation

31

32

The Cyclesorb service is a mobile adsorption service combining adsorber and transport vessel for the treatment of food grade, non-corrosive and corrosive liquid streams. A pre-loaded Cyclesorb is delivered to a site and larger vessels can be moved by forklift where required. The unit is operated until saturation of the activated carbon. A second unit is then delivered after saturation, and the exhausted unit is picked up at the same time, in order to minimise transport costs. Once the unit is returned to a Chemviron Carbon facility it is discharged, emptied of carbon, cleaned, inspected and refilled with either virgin or reactivated carbon. The Cyclesorb units range in size from bench scale to full industrial scale. These carbon vessels provide great flexibility of operation, and can eliminate the wall effects and other problems found using small scale pilot equipment. The liquid volume treated during these tests is much greater than with a classic pilot test so that plant variations can be evaluated more thoroughly. Due to their design, the Cyclesorb range is also very easy to handle and to start up. Since the carbon is pre-wetted and contained within the Cyclesorb there is no need for the operator to come in contact with the carbon.

For a classic laboratory study, columns are mounted vertically and connected in series for down-flow operation as shown in Figure 16. Up-flow treatment can also be considered, but there is no inherent advantage. Provision must be made for taking samples between each column so that purification can be studied as a function of bed depth, as well as of time and volume of solution treated. The carbon bed is supported by a stainless steel or plastic screen, glass wool, or porous plate at the base of each column. Manometers or pressure gauges should be used to monitor the pressure drop characteristics of the operation. The liquid is fed to the adsorbers using a positive displacement pump to allow accurate flow control. If necessary, the adsorbers can be jacketed or heat traced to avoid a temperature gradient across the beds. Figure 17 shows a set-up for a pulse bed pilot plant. The liquid is fed into the bottom of the adsorber and flows upwards through the carbon bed. The liquid is sampled before the adsorber inlet (sample 0), at different levels in the carbon bed (samples 1, 2, 3) and at the adsorber outlet (sample 4). The adsorber can be jacketed, if necessary. At regular intervals a fixed amount of carbon is withdrawn from the bottom of the adsorber, and an equal volume of fresh carbon is added to the top of the adsorber.

Siphon-break
PI

Vents

Charge Tank

Siphon-break

Outlet

(3)

(2)

(4)

(1) (4)

Pump Prefilter
Pump
INFLUENT

PI
INFLUENT

(1) (0)

(2)

(3)

(0)

Sample Ports

Spent carbon

Figure 16: Fixed bed pilot plant arrangement

Figure 17: Pulse bed pilot plant arrangement

33

34

3.2.4 Pump A pump capable of uniform delivery rates with positive displacement is preferable. The most important criterion is the continuous delivery of a constant flow with no introduction of air. 3.2.5 Flow Meter and Totaliser A flow meter and totaliser should be installed so that the flow can be controlled and the treated volume recorded. 3.2.6 Connections Rubber, glass, plastic or metal tubing can be used, depending on the corrosive characteristics of the material to be processed. Usually, rubber and glass tubing connections with glass tees and pinch clamps are the most convenient.

3.4 Carbon Pre-treatment


One of the most common errors in adsorption column test work which leads to poor results is that the carbon is not degassed prior to the adsorption test. When this is not done, air pockets form in the column and result in: (1) channelling, (2) high pressure drop, and (3) premature breakthrough of adsorbate. The time required to degas carbon is a function of liquid temperature, as shown in Figure 18.

3.3 Liquid Pre-treatment


If the isotherm investigation indicates that solids removal or pH adjustment is necessary before carbon treatment, then the same treatment should be made to the liquid before the carbon column studies. If the test liquid has been stored at a low temperature for preservation purposes, subsequent heating to room temperature or higher, may result in degassing the liquid in the carbon bed. When these conditions exist, the liquid must be degassed prior to pumping it through the carbon columns.

% Wetted (% of available internal volume filled)

100

0 0 12 24 36 48

25oC

Time (hours)

80oC

Figure 18: Wetting time with water

The carbon should be wetted prior to being placed in the test columns. Indeed it is extremely difficult to remove air pockets from the packed carbon bed and contrary to popular belief, backwashing the carbon at low rates does not remove the air pockets. The amount of granular activated carbon required to fill each column may be determined as follows: Weight carbon/column = V x AD where V = volume of carbon in column; AD = apparent density of carbon

35

36

The required amount of granular activated carbon for each column should be weighed into a suitable container and degassed or wetted using one of the following procedures: soak at ambient temperature for a minimum of 24 hours boil for 2 hours. The degassed carbon is charged to the columns as a slurry in small increments, keeping a thin layer of supernatant liquid above the carbon during charging. This is best accomplished by filling the column one third full with the wetting liquid before starting to charge the carbon slurry. This charging procedure should be repeated for each column in series, all connecting tubing and other void space must be filled with liquid in order to avoid pushing an air pocket through the carbon bed when starting the test. Any gas build-up can be released through the vents at the top of the columns.

3.5 Column Operation


3.5.1 Temperature As adsorption is a function of diffusion rate and as diffusion is affected by the liquid viscosity, the columns should be operated at the plant process temperature to eliminate this variable from the data analysis. 3.5.2 Linear Velocity of the Liquid in the Bed

Many studies have been conducted to examine if and how linear velocity affects diffusion rate. These tests confirm that an optimum linear velocity exists and must be considered for each application individually. Velocities that are too high can extend the mass transfer zone and therefore the breakthrough curve (Figure 19). Consequently, it is recommended that the linear velocity in the pilot column is the same as that expected to be used in the full scale plant. If time and money permit, it is advisable to investigate one or two other linear velocities to ensure that the optimum is being used.

100

Concentration %
0

Time on stream or volume treated


High linear velocity Low linear velocity Inlet concentration Treatment objective

Figure 19: Influence of linear velocity on the mass transfer zone

37

38

3.5.3 Duration of Column Testing Ideally, the test should be run until the last column is saturated, or until the treatment objective of the liquid in the last column has been reached. If time and/or the quantity of test liquids available do not permit the test to be run to ideal completion, it should be run to achieve at least saturation in the first column and breakthrough to 50 % of influent concentration in the second column. 3.5.4 Displacement, Sampling and Analysis So that accurate breakthrough curves can be defined, the wetting liquid must be displaced from the columns in order to establish a treatment starting point for each column used in the study.
0 100

Concentration %

After the wetting liquid is displaced from the first column, the collection of data to determine the breakthrough curve can be initiated. The breakthrough curve is determined by sampling at regular intervals; analysing the samples for impurities, and by measuring and recording the volume throughput (Appendix 2). The displacement procedure is repeated for each column until the wetting liquid from every column has been displaced. The breakthrough curve for each column must then be defined as previously described.

V1
Sample 1 Sample 4

Time on stream or volume treated


Sample 2 Inlet concentration Sample 3 Treatment objective

Figure 20: Typical breakthrough curves

From Figure 20 one can deduce: the length of the mass transfer zone the adsorption capacity of the carbon for the impurities to be adsorbed at breakthrough and saturation conditions. 3.6.1 Mass Transfer Zone Determination During the adsorption cycle in a column, the upper section of the bed will be saturated with impurity while the lower section is still free of impurity. Between these two extremes lies a zone in which the adsorption is actually occurring. This is referred to as the mass transfer zone or MTZ. As the column becomes saturated, this adsorption zone moves downwards through the bed and can be regarded as an adsorption wave front moving through the column (see Figure 21).

3.6 Interpretation of Pilot Data


Samples are taken at the influent of the first column and at the effluent of each carbon bed. Analytical data is plotted versus cumulative throughput or time. The plot of this data gives the so-called breakthrough curve for each column. Figure 20 shows typical breakthrough curves for complex multi-components where the influent impurity concentration is constant. The sample (1) curve corresponds to a superficial contact time t . Assuming all columns contain equal amounts of carbon, the curves (2), (3) and (4) correspond to superficial contact times of 2t, 3t and 4t respectively.

39

40

Inlet

Inlet

Inlet

Inlet
(0)
100

(1)

MTZ

Headspace

Saturated MTZ

Virgin

Outlet

Outlet Start test

Outlet Middle of test Short Mass Transfer Zone

Outlet Breakthrough
0

Concentration %

MTZ

MTZ

(2) (3) (4)

Long Mass Transfer Zone

t tm

2t
Superficial Contact Time

3t

4t

Figure 21: Mass transfer zone concepts

Treatment objective

Inlet concentration

The length of time required for the appearance of the impurity and the shape of the breakthrough curve provides an indication of the relative length of the adsorption zone. If considerable time elapses before the impurity appears, and the breakthrough curve is sharply defined and steep, then the adsorption zone is short in relation to the overall bed depth. If, on the other hand, breakthrough occurs almost immediately, a relatively long adsorption zone is indicated. After treating a volume of liquid V1 (Figure 20), column no.1 no longer participates in the adsorption process because it is saturated. The impurities adsorbed on the carbon are in equilibrium with the impurities in the liquid at influent concentration. The mass transfer then takes place mainly in columns 2 and 3, and to a minor extent in column 4. The variation in concentration of impurity through the adsorbers after treating a volume (V1) can be represented as shown in Figure 22. This graph shows that the minimum superficial contact time to purify the liquid from its value C0 down to Ce is tm.

Figure 22: Concentration versus contact time

3.6.2 Minimum Carbon Consumption The carbon saturation rate is primarily a function of the working conditions, and the amount and nature of impurity present in the liquid to be treated. Since carbon consumption depends on the superficial contact time, a good method of presenting data (Figure 23) is achieved by plotting the carbon consumption for a specific treatment objective against the individual superficial contact time of each column. The carbon consumption requirement is calculated by dividing the weight of carbon in each column by the volume of liquid treated at the point at which the treatment objective is reached. "The operating line" is the name given to the curve of such a configuration i.e. a single fixed bed. The operating line allows us to determine the optimum combination of superficial contact time and carbon saturation rate.

41

42

Inlet
100

Carbon Consumption %

Inlet

2t

3t

4t

Superficial Contact Time


A

MTZ

Isotherm

Column

Outlet

Short single column

Long single column

Figure 23 : Operating line of a carbon system

MTZ Outlet

Inlet

Outlet

3.7 Fixed Bed: Single, Series or Parallel


In a single fixed bed system, the carbon contained in the column is only partially saturated when the effluent impurity concentration reaches the treatment objective (Figure 24A). The bed is taken off-stream to be reactivated or replaced and the amount of carbon per unit of liquid treated exceeds the minimum value obtained from an isotherm test. As the depth of the carbon bed is extended beyond the length of the mass transfer zone, the relative amount of fully saturated carbon increases, and the carbon usage rate decreases (Figure 24B).
MTZ C Outlet Series mode Inlet Outlet

D Parallel mode

Figure 24: Single and multiple column Mass Transfer Zones

43

MTZ

MTZ

MTZ

44

A system of two fixed beds in series permits a more efficient usage of the carbon. Indeed when the effluent impurity reaches the imposed limit, the lead adsorber is taken off stream and the carbon is reactivated. An adsorber containing fresh or reactivated carbon is then put on stream in the polishing position (Figure 24C). In this system, more fully saturated carbon is discharged. As a result, the operating line for this system will indicate lower carbon exhaustion rates. Similar operating lines can be established for a system of fixed beds in parallel (Figure 24D). This figure represents a system of parallel adsorbers that have been installed at different points in time. This set-up can be used when a part or the full stream has to be treated at all points in time. The choice of that system configuration may be investigated due to one or more of the following factors: if the flow rate is too high, then the size of the adsorbers in a single pass would be too large to be economical space limitation could prevent the use of an extremely high column a minimum pressure drop would be required the adsorption process must be continuous with limited process fluctuation blending of effluent product is desirable. Furthermore, a combination of adsorbers in series-parallel is desirable when a lower carbon dosage has to be combined with the benefit of parallel operation.

3.8 Pulse Bed


A pulse bed operates as a large number of beds in series and therefore, is suitable for applications having high carbon consumption and long mass transfer zone. Some carbon is removed at a predetermined rate from the bottom of the column and replaced at the top by fresh carbon to ensure that the mass transfer zone remains within the column length, thereby ensuring full saturation of the carbon before discharge. This pulse rate is selected based on the experience of Chemviron Carbon in similar applications. The column is sampled at points (0), (1), (2), (3), (4) and analytical data is plotted versus cumulative volume or versus time. Figure 25 shows a representative set of results obtained from a pilot pulse bed.

100

Concentration %
0

Time or volume
Sample (1) Sample (2) Sample (3) Sample (4) Inlet concentration

Figure 25: Pulse bed test results

If the chosen pulsing frequency is too low, the mass transfer zone gradually moves through the bed with the risk of exceeding the treatment objective. On the other hand, if the selected pulsing rate is too high, the withdrawn carbon will not be fully saturated and the resulting operating costs will be higher than necessary.

45

46

3.9 Accelerated Column Test (ACT)


Pilot testing as described above predicts accurate carbon usage and produces data appropriate for design, it can however be very time consuming. Therefore, Chemviron Carbon has developed the Accelerated Column Test, an improved technique for testing the removal of organic impurities in water solutions. This test combines speed with the accuracy of a pilot column. Acceleration of the carbon adsorption cycle is achieved through a scaling down of the conventional column testing hardware. Except for their reduced scale, the other components of the test system and the overall system design are essentially identical to larger scale laboratory or field evaluation systems (Figure 26).

3.10 Sizing of the Adsorption System


The sizing of a granular activated carbon system depends primarily on the length of the mass transfer zone, the carbon saturation rate and the linear velocity. These parameters are determined by the column tests. Once the operating line is established, it is possible to select the combination of carbon consumption and superficial contact time which will give optimum results at the lowest cost. Since both capital and operating costs are involved, an annual operating cost must be established to compare the systems. The capital investment cost of the adsorption system is primarily dependent on the superficial contact time. The operating cost depends mainly on the carbon exhaustion rate. If the annual capital depreciation and operating costs are plotted as a function of superficial contact time, the summation of the two curves will give the total cost. The optimum design is indicated by the minimum of the total cost curve as shown on Figure 27.

Figure 26: Accelerated Column Test equipment


0.5 2 3 4

The technology behind the ACT was developed by Calgon Carbon Corporation, the parent company of Chemviron Carbon. This development enabled Calgon Carbon scientists to further develop a computer model for the column adsorption process. The Accelerated Column Test is based on this computer model. With this computer programme the breakthrough curves for full scale adsorption systems can be readily calculated from the data generated by the scaled down Accelerated Column Test.

Cost

Superficial Contact Time


Total cost Investment cost Operation cost

Figure 27: Annual costs

47

48

4 Conclusion
In conclusion, it should be emphasised that in many cases there are alternative ways in which a laboratory test can be carried out for a given application. The relative advantages of each method can be ascertained only by carefully working through the problem, keeping in mind the particular requirements of the process to be tested. Please contact Chemviron Carbon for further information prior to scheduling any tests. As a Producer and Reactivator of activated carbon since the middle of the last century, Chemviron Carbon has extensive experience in the application of activated carbon in almost every field, and can assist in establishing process and operational guidelines to answer your needs.

Appendix 1 : Format For Reporting Isotherms Data

Treatment Objective

Application

Date

Carbon sample reference number

49

M Carbon dosage (g/l)

Temperature (C)

Volume solution

Customer name

Type of carbon

Agitation time

Concentration

Contaminant

pH

C Concentration remaining in solution (mg/l)

X = (C0 - C)V Constituent adsorbed (mg)

X/M Constituent adsorbed per unit weight (mg/g)

50

Appendix 2: Format for Reporting Pilot Data

Glossary of terms commonly used in Activated Carbon Technology


Abrasion number The abrasion number of granular carbon defines the resistance of the particles to degrade on handling. It is calculated by contacting a carbon sample with steel balls on a Ro-Tap machine and determining the percent ratio of the final mean particle diameter. Carbon which has been contacted with an acid solution with the purpose of removing some of the ash in the activated carbon. Activated carbon is a crude form of graphite, with a random or amorphous structure, which is highly porous, over a broad range of pore sizes, from visible cracks and crevices to cracks and crevices of molecular dimensions. Any substance that is or can be adsorbed on the adsorbent. Any solid having the ability to concentrate significant quantities of other substances on its surface. Activated carbon is an adsorbent. A vessel designed to hold granular activated carbon. The phenomenon whereby molecules adhere to a surface with which they come in contact. A measurement of adsorption determined at a constant temperature by varying the amount of carbon used or the concentration of the impurity in contact with the carbon. The finest pores in the carbon structure. Pores which have adsorption capabilities. The weight per unit volume of activated carbon. The mineral oxide constituents of activated carbon. It is normally defined as a weight percent basis after a given amount of sample is oxidized. An operating method used to remove suspended solids from the carbon bed. Water is pumped into the bottom of the adsorber, flows upward through the carbon bed, and exits through the backwash outlet. The upward flow expands the bed and removes the suspended solids, carbons fines and entrained air. The percentage bed expansion time required for backwashing is a function of the backwash rate and water temperature. The amount of carbon expressed in length units, which is parallel to the flow of the stream and through which the stream must pass.

Remarks

Acid washed activated carbon Activated carbon

Superficial contact time (min)

Sample Volume Analytical data

Linear velocity (m/h)

Pilot description

Adsorbate Adsorbent

Application

Flow (l/h)

Date

Adsorber Adsorption Adsorption isotherms

Volume throughput

Adsorption pores Apparent density Ash

Carbon sample reference number

Time

Backwash

Volume of carbon/column

Weight of carbon/column

Treatment objective

Temperature (C)

Customer name

Type of carbon

Concentration

Contaminant

Sample N

pH

Date

Bed depth

51

52

Bed expansion

The effect produced during backwashing: the carbon particles become separated and rise in the column. The expansion of the bed due to the increase of the space between carbon particles may be controlled by regulating backwash flow. The first appearance of adsorbate material in the solution flowing from an activated carbon unit. Breakthrough is a first indication that reactivation or regeneration of the carbon is necessary. A column filled with granular activated carbon whose primary function is the preferential adsorption of a particular type or types of molecules. The flow of water or regenerant taking the line of least resistance through an activated carbon bed, as opposed to the usual distributed flow through all passages of the bed. Channelling may be due to fouling of the bed, poor distribution design or low flow. Adsorption where the forces holding the adsorbate to the adsorbent are chemical (valence) instead of physical (London's). Complex molecules which impart colour (usually undesirable) to a solution. Carbon adsorption is often used in such colour removal applications. In a carbon column the critical bed depth is the depth of granular carbon which is partially spent. It lies between the fresh carbon and the spent carbon and is the zone where adsorption takes place. In a single column system, this is the amount of carbon that is not completely utilised. The area of activated carbon through which the stream flow is perpendicular. The process of removing air (gases) from a bed of carbon. In a volume of 1m3 of activated carbon, there are approximately 400 l of void space, 400 l of pore volume and 200 l of skeleton carbon. The opposite of adsorption. A phenomenon where an adsorbed molecule leaves the surface of the adsorbent. The heat given off when molecules are adsorbed. The iodine number is the milligrammes of iodine adsorbed from a 0.1N solution by one gram of carbon at a filtrate concentration of 0.02N iodine.

Mass Transfer Zone (MTZ)

The adsorption gradient that exists in the carbon bed. It corresponds to the gradual transition of the carbon from "spent" to "fresh". The particle size of granular activated carbon as determined by a U.S. Sieve Series. Particle size distribution within a Mesh series is given in the specifications of the particular CHEMVIRON CARBON carbon. The percent by weight of water adsorbed on activated carbon. A decrease in fluid pressure during its flow due to internal friction between fluid molecules, and external friction due to irregularities or roughness in surfaces past which the fluid flows. In addition, accumulation of solids on the activated carbon can cause pressure drop. A unique application which a single carbon column offers the efficiency of several columns in series. This is accomplished by the removal of exhausted carbon from one end of the carbon bed and the addition of fresh carbon at the other end with little or no interruption in the process. The removal of adsorbates from spent granular activated carbon which will allow the carbon to be reused. This is also called revivification. It refers to any desorption process other than reactivation (including hot gases, solvents, supercritical fluids and displacing compounds) commonly used to remove the adsorbate from the adsorption space within the carbon. The surface area of granular activated carbon is determined by the Brunauer, Emmet and Teller Method (BET Method), which utilises the adsorption of nitrogen at liquid nitrogen temperatures in the calculation. Surface area is usually expressed in square metres. Pores larger than the largest adsorption pores. They function as a diffusion path to transport adsorbates. Adsorption does not occur in these locations even at near saturated conditions. The percent by volume of the interstices to total bed volume. The wave front is the carbon loading gradient that exists in the critical bed depth. It outlines the gradual transition of the carbon from "virgin" to "exhausted".

Mesh size

Breakthrough

Moisture Pressure drop

Carbon column

Channelling

Pulse bed

Chemisorption

Reactivation

Colour bodies

Regeneration

Critical bed depth

Surface area

Cross sectional bed area Deaeration (wetting)

Transport pores

Voids Wave front

Desorption Heat of adsorption Iodine number

53

54

5 Headquarters and Technical Department


European Headquarters Chemviron Carbon, Zoning Industriel C de Feluy, B-7181 Feluy Tel +32 (0) 64 511 811 Corporate Office Fax +32 (0) 64 541 591

Calgon Carbon Corporation, 400 Calgon Carbon Drive Pittsburg, PA 15205 Tel +1 (0) 412 787 6700 Fax +1 (0) 412 787 6713

Website

www.calgoncarbon.com

www.chemvironcarbon.com

55