The oxidation of sewage sludge in the liquid water phase at elevated temperatures and pressures : wet-air oxidation

The oxidation of sewage sludge in the liquid water phase at

elevated temperatures and pressures : wet-air oxidation

Citation for published version (APA):

Ploos V Amstel, J. J. A. (1971). The oxidation of sewage sludge in the liquid water phase at elevated temperatures and pressures : wet-air oxidation. Technische Hogeschool Eindhoven.

https://doi.org/10.6100/IR114081

DOI:

10.6100/IR114081

Document status and date: Published: 01/01/1971 Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne Take down policy

If you believe that this document breaches copyright please contact us at: openaccess@tue.nl

THE OXIDATION OF SEWAGE SLUDGE

IN THE LIQUID WATER PHASE

AT ELEVATED TEMPERATURES AND PRESSURES

(WET-AIR OXIDATION)

Fl t,;l-IT

WATER 'POLLUTION

THE OXIDATION OF SEWAGE SLUDGE

IN THE LIQUID WATER PHASE

AT ELEVATED TEMPERATURES AND PRESSURES

(WET-AIR OXIDATION)

PROEFSCHRIFT

TER VERKRIJGING VAN DE GRAAD VAN DOCTOR IN DE

.TECHNISCHE WETENSCHAPPEN AAN DE TECHNISCHE

HO-GESCHOOL TE EINDHOVEN, OP GEZAG VAN DE RECTOR

MAGNIFICUS, PROF.DR.IR.A.A.TH.M.VAN TRIER,VOOR EEN COMMISSIE UIT DE SENAAT IN HET OPENBAAR TE

VERDEDIGEN OP VRIJDAG 2 APRIL 1971 TE 16 UUR

DOOR

JOHANNES JACOBUS ASUERUS PLOOS VAN AMSTEL

DIT PROEFSCHRIFT IS GOEDGEKEURD DOOR DE PROMOTOR

PROF. DR. K. RIETEMA

CO-REFERENT

AAN OE KOMMUNE NUENEN EN IN HET BIJZONOER AAN PRINSES LISELORELEI

ACKNOWLEDGMENT

The thesis in this make-up would not have been completed but

for the contribution of many whom I would like to thank here.

I will mention in particular Mr. J. Bos for his interest in

the project and for his advice and assistance in the research

of the literature.

Thanks are also due to Mr. W.C. Koolmees who advised on the

design and construction of the apparatus, and to Messrs. P.A.

Hoskens and A.H. van der Stappen, who constructed and put

to-gether the equipment.

Furthermore I am indebted to Messrs A.W.C.M.van Alphen, F.C.M.

van de Berg, C.G.M. de Boer, F.H.J. Bukkems, P.J.A.M. Derks,

H. van Gool, G. Groen,

c.

van de Moesdijk, P.J.T. Samuels

'

H.J.C. Slegers, F.C.R.M. Smits, H.P.E. van de Venne and P. van Zutphen, who carried out most of the experiments.

Almost all technical drawings have been made by Mr.J.Boonstra, the remainder by Mr.Klein Wassink. Mrs. D.M.Vermeltfoort typed the text which was edited by Miss G.M. Kurten. I would like to

thank both ladies and both gentlemen for the accuracy with

which they carried out their work.

I am very grateful to Mr. H.J.A. van Beckum, who made this

thesis readable by correcting the language and to Ton Smits,

the famous artist,who made this thesis digestible by his witty cartoons.

I wish to thank in particular the directors of the Architekten en Ingenieurs Bureau of the N.V. Philips' Gloeilampenfabrieken

for the assistence offered during the final stages of this

work.

Finally, I include in my acknowledgement my wife and also

Mr. J. Waterman for their encouragement during the last weeks

of the preparation of the thesis.

CONTENTS

SUMMARY

1. TREATMENT AND DISPOSAL OF SEWAGE

2.

1.1. INTRODUCTION

1. 2. WASTE WATER TREATMENT 1.3. SEWAGE SLUDGES

1.4. SEWAGE SLUDGE TREATMENT AND DISPOSAL

THE WET-AIR OXIDATION PROCESS 2.1. EVOLUTION OF THE PROCESS 2.2. OXIDATION "ROUTES"

2.3. INFLUENCE OF PROCESS PARAMETERS 2.4. END PRODUCTS 2.5. COST OF PROCESS 2.6. LITERATURE REVIEUWS 2.7. CONCLUSIONS

x

1 1 2 3 5 8 8 9 10 13 13 14 14

3. THE OXIDATION OF A GLUCOSE SOLUTION AS A MODEL SLUDGE 15 3.1. PRELIMINARY EXPERIMENTS

3.1.1.

Apparatus and experimental details

3.1.2.

Thermal treatment and its influence on

the oxidation

3.1.3.

Influence of shaker frequency

3.1.4.

Homogeneous oxidation of glucose

3.1.5.

The maximum conversion

15 15 16 19 21 23 VII

3.1.6. Mathematical description of conversion rate

3.2. CONTINUOUS FLOW EXPERIMENTS

3.2.1. Apparatus and experimental details

3.2.2. Influence of temperature

3.2.3. Influence of COD concentration in the

24 25 25 27

feed and oxygen pressure 29

3.3. THEORETICAL ANALYSIS OF RESULTS AND DISCUSSION 30

3.3.1. Introduction

3.3.2. Model for the macro kinetics

3.3.3. Model for the micro kinetics

3.3.4. Speculations on the order of magnitude

30 32 36

of kinetic data 40

4. THE DISSOLUTION OF SLUDGE

4.1. APPARATUS AND EXPERIMENTAL DETAILS

4.2. THE HYDROLYSIS OF SLUDGE PARTICLES

4.2.1. The rate of hydrolysis of activated

sludge

46 46 47

47

4.2.2. The rate of hydrolysis of primary sludge 51

4.2.3. Repeated-hydrolysis 53

4.3. INFLUENCE OF CONCENTRATION AND OF OPERATING

CONDITIONS ON [cmax] 54

5. THE OXIDATION OF PRIMARY SLUDGE 57

5.1. PRELIMINARY BATCH EXPERIMENTS

5.2. SEMI-BATCH EXPERIMENTS

57

59

VIII

5.2.1. Apparatus and Experimental details 60

5.2.2. Model for oxidation in a semi~batch system 62

6. THE OXIDATION OF ACTIVATED SLUDGE

6 .1.

6.2. 6.3. 6.4.

EXTENSION OF THE MODEL

APPARATUS AND EXPERIMENTAL DETAILS

EXPERIMENTS WITH SLUDGES OF DIFFERENT ORIGINS FURTHER EXPERIMENTS WITH EINDHOVEN SLUDGE 6.4.1. Influence of temperature 74 74 76 77 79 79

6.4.2. Evaluation of kinetic data and discussion 81

6.4.3. Influence of pressure 86

7. GENERAL DISCUSSION AND CONCLUSIONS 89

7 .1.

7.2.

7.3.

COMPARISON BETWEEN MODEL SLUDGE AND OTHER SLUDGES

7.1.1. 7.1.2.

Reaction orders

Locus of oxidation and diffusion limitation

7.1.3. Reaction rate constants 7.1.4. Effect of dissolution

7.1.5. Sludges of different origins ANOMALOUS PHENOMENA

CALCULATIONS OF THE SIZE OF COMMERCIAL REACTORS REFERENCES NOMENCLATURE SAMENVATTING LEVENSBESCHRIJVING 89 89 90 90 92 93 93 94 102 107 110 113

SUMMARY

Wet-air oxidation is a method of sewage sludge treatment by

which the sludge is oxidised in a liquid water phase in the

0

presence of air at temperatures of about 200-300 c and

pres-sures of 40-120 atm. From an analysis of the literature on the

subject i t became clear that this process had apparently been

developed empirically and that only little insight into the

fundamental aspects was present.

Because of the attractiveness of wet-air oxidation, a research

project was carried out in which a process engineering

ap-proach was applied.

First the oxidation of a glucose solution as a model sludge

was investigated with semi-batch and continuous flow

ments at about 200°c and 50 atm.It followed from these experi-ments that the oxidation of the model sludge was fast compared

with diffusion of oxygen. This caused the oxidation to take

place within the diffusion layer around the gas bubbles. The

chemical reaction rate can be described as first order in

or-ganic matter and zero order in oxygen, while the reaction rate

-1 0

constant is about 2 sec at 200 c.

A model is presented for the conversion in the continuous flow reactor, including combined reaction and diffusion, mixing and

convection. This model gives a fair description of the

influ-ence of process parameters on the conversion and of the

con-centration profiles in the reactor.

At the high temperatures applied in practice real sludges

partly pass into solution by hydrolysis. Owing to this,

simul-taneous oxidation of hydrolised sludge and sludge particles

takes place.

In order to understand the contribution of the oxidation of

hydrolised sludge to the total conversion rate, the hydrolysis of sludge was investigated with batch experiments.

The effect of hydrolysis on the overall conversion rate was

studied, using primary sludge. Oxidation experiments were

carried out at 230°c and 100 atrn with sludge as such, with

hydrolised sludge and with a suspension of the solid residue

of hydrolysis.

From these investigations i t was concluded that hydrolysis

does not influence the overall conversion rate, since

hydro-lised sludge and solid sludge have almost the same reactivity.

The oxidation of activated and primary sludge proceeds more slowly than the oxidation of the model sludge, and the

conver-sion takes mainly place in the bulk of the sludge. The degree

of diffusion limitation of oxygen, or the extent to which the

conversion rate is reduced by oxygen transfer depends on

temp-erature. At 180°c mass transfer can be neglected, while at

290°c the conversion rate is largely determined by the rate of

mass transfer.

A model for the oxidation of sludges is presented.The starting

point of this model is that in the sludge two groups of

com-ponents can be distinguished which differ in reactivity and

which are oxidised simultaneously, while furthermore a third

group is present which is completely inactive. Experiments

have shown that activated sludge consists for about 65% of

more reactive matter, 25% of less reactive matter and 10% of

inactive matter. The chemical reaction rate was described as

first order in organic matter and first order in oxygen (at

relatively high oxygen pressures, the reaction becomes zero

order in oxygen). Mass transfer is also included in the model.

It was found that this model provides a fair description of

the experimental findings. The effect of temperature on the

reaction rate constants of both oxidisable groups in activated sludge can be described by Arrhenius' law,while the activation

energy is about 23 kcal/mol for both groups. At 255°c the

On the ground of the results of the research the size of a

commercial reactor was calculated. Depending on the amount of

surplus oxygen, the required residence time in a plug flow re-actor operated in co-current was two to three times as long as

in a plug flow reactor operated in counter-current, while the

required residence time in a reactor in which the liquid was

completely mixed, was 6 to 12 times as long as in a

counter-current reactor.

In practice, a co-current reactor is applied, in which the

mixing state will be somewhere between completely mixed and

plug flow, which will require a residence time of about five

times that in the counter-current plug flow reactor. For prac-tical reasons, however, the co-current reactor applied in com-mercial installations, will probably remain more attractive.

-~,,,

'

Chapter 1 TREATMENT AND DISPOSAL OF SEWAGE

1.1. INTRODUCTION

For many centuries sewage has been discharged into streams,

lakes and ponds, and even now this is a very normal procedure.

In these natural waters micro-organisms consume the discharged

contaminations, while at the same time the solids of the

sew-age settle, both mechanisms resulting in natural purificat.ion.

In order to avoid pollution of natural waters,

the discharge

of sewage should balance the natural purification capacity of

the receiving water. When the flow of waste water exceeds this

capacity,

the contaminations in the waste water must be

re-duced by a suitable treatment.

In general,

the pollution of natural waters is objectionable

for the possible hazard to public health and safety.Of a

less-er consequence,

but still very real,

is the aesthetic aspect

of the deterioration of surface water. Before long this aspect

will also weigh more heavily, because the increasing amount

of leisure time and wealth imply increased recreation at, on

and in the water (1, 2, 3).

For the reduction of contaminations in the waste water a

varie-ty of waste water treatment processes is available ( 1, 2, 4,

5, 6, 7). Such a treatment generally results in a flow of more or less clean water and in a second small flow containing

con-centrated suspended impurities known as sewage sludge. In the

present procedures the sewage sludge undergoes a further

treat-ment towards a form suitable for one or another method of

fi-nal disposal, like dumping or soil conditioning (11, 33, 34,

35, 36).

1.2. WASTE WATER TREATMENT

Figure 1.1 shows a conventional activated sludge installation.

sewage Figure 1.1. A B grit

D

to surf ace water air activated 1----~'--~~~~~~~--' sludge primary sludge

Sewage treatment plant.

A: Sereening; B: grit removal; C: primary settling; D: aeration; E: seeundary settling.

The treatment of domestic sewage and many other organic waste

waters may usually be divided into two steps: pretreatment and

biological oxidation.

Pretreatment includes screening,

grit

removal and sedimentation or flotation.

The sludge removed

from the settling or flotation tanks is called primary sludge.

The pretreated waste water can be further subjected to

biolog-ical oxidation,

resulting in a removal of colloidal and

dis-solved organic matter by the action of micro-organisms. In the

activated sludge system the pretreated waste is brought into

contact with the activated sludge, which consists of

floccu-lated micro-organisms and adsorbed contaminations, the process

being carried out in an aerated tank.

The mixture of activated sludge and treated water is subjected

to a .secondary sedimentation.

The activated sludge is partly

recycled to the aeration tank, maintaining stationary condi-

₁

tions. The surplus sludge is removed for further treatment andi

disposal.

The overflowing liquid of the secondary sedimentation tank is .

discharged into the natural waters or used to irrigate the .

land.

At present another step is sometimes added to reduce the

nitro-gen and phosphorus content (8, 9, 10).

1.3.

SEWAGE SLUDGES

Municipal sewage consists of aqueous discharges from kitchens,

bathrooms, lavatories and laundries,

and also of waste waters

from a variety of industries.

Since primary sludge consists of the settable contaminations

of the sewage,

its composition depends on the habits of the

population and on the kind of industry discharging on the

mu-nicipal sewerage.

A typical composition of the organic matter in a suspension of

primary sludge is as follows.

Protein Lipids Starch Crude fibre Volatile acid Total

kg/m3

6.2

10.5

3.6

13.5

1.3

35.1

Deviations from these figures have to be expected. The amount of inorganic matter in primary sludge is also subject to

vari-ations and may differ from plant to plant. The primary sludge

of Eindhoven contains an amount of inorganic matter of about

20 kg/m3.

The composition and physical properties of the sludge have a

great influence on the selection of the sewage sludge

treat-ment procedure.

Activated sludge arises spontaneously in an activated sludge

installation from the micro-organisms present in the waste

water. It consists of flocculated bacteria, protozoa, etc.

and adsorbed material.

The chemical composition is about C118H170051N17p

The bacteria have a diameter of 0.5 to 1.5µ and are seldom

longer than 10µ.

Their slimy skin causes them to be grouped together into

tenu-ous flocks which may have characteristic dimensions of 20 to

100µ.

The concentration of organic contaminations in sewage and sew-age sludge is often characterised by the biological oxygen de-mand (BOD). This is the amount of oxygen which is consumed per unit volume by the action of micro-organisms on the contamina-tions (77). In general the BOD is expressed in mg/1.

Another characterisation of the concentration which will be

used in this thesis is the chemical oxygen demand (COD).

The COD is the amount of oxygen necessary per unit volume for

oxidation with a dichromate-sulfuric acid mixture under

stand-ard conditions (77). In general the COD is also expressed in

mg/1. However, in this thesis practical units are used, so the COD is expressed in kg/m3.

Since the break-down by the dichromate-sulfuric acid mixture

in general proceeds further than by the action of the

be-tween BOD and COD will frequently vary.For domestic waste

wat-er Huntwat-er

<i27)

reports that

COD ::::: 2 BOD

1.4. SEWAGE SLUDGE TREATMENT AND DISPOSAL

The numerous

sludge treatment procedures may be grouped into

four major categories, indicated in figure 1.2.

raw sludge

,--concentration

digestion

dewatering

combustion

final disposal

Figure 1.2.

Basic steps in sewage sZudge treatment.

Which kind of combination of these procedures provides the

most economic and reliable

solut~on,

depends on the nature and

concentration of the sludge,

on the selected disposal method

and on local situations.

The concentration step of the sludge suspension, indicated in

figure 1.2, yields a more economic treatment in subsequent

p:ro-cesses ( 5, 14) •

Anaerobic sludge digestion. involves biological breakdown of

organic matter in the absence of oxygen by anaerobic

bacte-ria.

Dewatering by filtration, centrifuging or by the use of drying

beds, results in a sludge of a more or less solid state.

How-ever, flocculation agents have to be added (15, 16, 17, 18).

Increase of filtration rate can also be obtained by heating

the sludge for half an hour at 180-200°c and 10-15 atmospheres, resulting in the disappearance of the colloidal structure (13, 21, 22, 23, 119).

This heat treatment process produces a sludge with an

offen-sive smell, which, however, disappears when air is also fed to the reactor, resulting in partial oxidation (2, 4, 15).

Combustion reduces the volume of the solids. The final product is a mineral and odourless ash.

Two groups of combustion processes can be distinguished:

(i) without previous dewatering (20, 26, 27, 28, 38),

e.g. the wet-air oxidation process;

(ii) with previous dewatering and heat drying (20, 29, 30), e.g. combustion in a fluidised bed.

The wet-air oxidation process is a new and promising process

by which the organic matter in the sludge is oxidised in the

liquid water phase at elevated temperatures and pressures

(220-300°c, 60-125 atmospheres ) in the presence of air (27,

28, 38)· The process is diagrammatically represented in figure 1.3.

In the heat exchanger, sludge and air are preheated prior to

being admitted to the reactor. In the gas-bubble slurry

reac-tor exothermic oxidation of the sludge takes place by which

the temperature rises about 30°c. The heat content of the

ef-fluent of the reactor is used to preheat the sludge and air

feed. After heat exchange, gas and liquid phases are separated and expanded to atmospheric pressure. For a large plant

reduc-tion in compression energy cost can be accomplished by using

an expansion engine for the separated gases.

In selecting the method for the final disposal of the treated

sludge pump sludge

Figure 1.3.

eactor expansion engin stack gas air compressor oxidised sludge

Wet-air oxidation process.

that public heal th or safety· shall not be impaired and no new pollution problem is generated.

When modern treatment procedures like wet-air oxidation are

applied, the final disposal of the sterile inorganic ashes

Chapter 2

0

l

-THE WET-AIR OXIDATION PROCESS

2.1. EVOLUTION OF THE PROCESS

The history of wet-air oxidation starts in 1912 when

Strehlenert (41) patents a method for the treatment of spent

sulfite liquor from paper mills with compressed air at 180°c.

In later versions of the process the oxidation of the paper

mill effluent is performed at temperatures ranging from 230

-330°c. This method was first patented in Sweden in 1949 by

Cederquist (42, 44, 45); his process has not been applied in

practice (46).

Independently, Zinunermann patented nearly the same process(43) in the U.S.A. in 1950. This patent was followed by many others

for a diversity of operating conditions and process

perform-ances (e.g. 47/69). The development and promotion of his

pro-cess was carried out by the ZIMPRO (ZIMmermann PROcess)

divi-sion of Sterling Drug. The first conunercial installation for

after a

short time because of corrosion

(70, 71, 72).

Till

February 1969,

18 other installations were sold

(73).

The

largest installation is located near Chicago and has a

capac-ity of

200

tons of dry sludge per day.

A patent for a system of wet-air oxidation in a deep shaft

ex-tending into the earth was awarded to Bauer

(112).

The depth

of the shaft was made sufficient to provide the required high

pressure.

The first patents claimed nearly complete oxidation and

atten-tion was focussed on paper mill effluent.

Attention was changed to sewage sludge and the advised degree

of oxidation was gradually reduced.

Nowadays

5-20%

oxidation

is applied at relatively low temperatures of

180 - 200°c,

with

the prime object of obtaining a better drainable sludge so

that the original combustion process is transformed into a

conditioning process prior to dewatering.

However, owing to the attractive possibilities of the orig·inal

combustion version and owing to the fact that this process has

never been approached in a process-engineering manner, we

car-ried out

an investigation of the high temperature version,

which is embodied in this thesis.

2.2.

OXIDATION "ROUTES"

Sewage

sludge consists of solid particles suspended in waste

water. In the wet-air oxidation process the oxygen is supplied

as gaseous air; therefore, a three-phase system is provided in

the reactor.

The oxygen diffuses from the gas-bubbles through

the gas-liquid interface into the suspension, where it reacts

with the solid sludge particles.

At the elevated temperatures

the organic polymeric structures

of the sludge are hydrolised to smaller soluble molecules (23,

Teletzke (75) observed the formation of free amino acids, free fatty acids and lower sugar molecules.

Hurwitz ( 76 ) presents data from which i t follows that the

fraction of organic material which is dissolved increases rap-idly with temperature and approaches 1 at 260°c.

As a result of this dissolution the oxidation of sewage sludge

can proceed via the solid particles as well as via the

dis-solved matter.

Takamatsu (129) assumed that the oxidation only proceeds

through dissolved matter, but he did not prove this experimen-tally.

In this respect i t can be referred to a patent of a system of

wet-air oxidation in which the sludge is first hydrolised as

far as possible. After settling of the solid residue which

takes place under the high temperature and pressure, only the

sludge solution is oxidised (114).

By the wet-air oxidation volatile products like acetic acid

might be generated. Oxidation of volatile products only takes

place in the liquid; in the gas phase no oxidation was ever

observed at the applied conditions (70,74).

2.3. INFLUENCE OF PROCESS PARAMETERS

In fig. 2.1 the influence of the temperature and the reaction

time on the COD reduction of a sludge suspension is

repre-sented. The data have been taken from Zimmermann (70). Total

pressure, mentioned.

starting concentration and kind of sludge were not

Nevertheless i t follows from this graph that

com-plete

3oo

0

c.

fast

oxidation is not possible at temperatures lower than

It is also seen that there is at first a relatively

oxidation, which is followed ny a very slow reaction.

The COD reduction reached after the relatively fast reaction

can be considered as a maximum conversion. As follows from

figure 2.1, this maximum is a function of temperature. Nearly

ioor-~-:::::::==================:---i 80 ~OD reduction 60 (%) 40 20

Figure 2.1.

time (h)

3oo

0

c

250 200 150 100

InfZuenae of temperature and time on COD reduation.

Figure 2.2 shows nearly all published data. Temperature: and

time are indicated in the graph. The other conditions are pre-sented in the tables 2.1 and 2.2.

The region of reaction times of 60 to 120 minutes represents batch experiments with activated,primary and digested sludges.

The curve (t

=

00 ) indicates that after "maximum" conversion is

attained still a considerable amount of COD reduction can be reached if the reaction is given enough time.

In the first publications of Zimpro (70,74) i t was stated that the only function of the high pressure was to reduce

evapora-tion in the reactor, which is necessary since the reaction

stops if no liquid water phase is present. However, Hurwitz

(79) found that higher pressures also increased the capacity

of an installation. The influence of the oxygen partial

pres-sure was suggested by the experiments of Abel (80) and of

Pepelyaev (81).

In the literature no investigation about the influence of the sludge concentration on the conversion rate could be found.

80 60 COD rec11.lction ( % ) 40 20

c

Tab le 2. 1. temperature (0c) Figure 2. 2.

Survey of data from literature. (): aommeroial installation see also tables 2.1 and 2.2.

Figures in graph are residenae times in min.

Some aommeroial installations represented in fig. 2.2.

location kind of starting temp pressure COD re-sludge COD (kg/m3) (OC) (atm) duction(%)

a

Chicago activated 48 270 125 80 b ·Wausau primary 62 260 115 75 c Wheeling primary 43-95 260 83 65 d Blind primary 41-97 235 53 79 Brook e _{Milwaukee raw/ 36-66 200 34 70} digested primary ref. 82,90 90 88,90 89,90 90,91

Tabie 2.2.

Experimentai conditions of experiments represented in fig.2.2.

residence kind of COD temp. pressure batch/ ref. time (min) sludge (kg/m ) 0 3 ( OC ) (atm) contin.

"' ? ? t i l l 200 ? batch 68,69 180 activated 51-75 260 83 cont. 79 60 primary, 15-100 100-300 ? batch 78 activated 60 primary 62,58, 100-300 ? batch 76 activated 31 digested 60 primary 58 150'-250 ? batch 75 38 activated 48-54 277 122 cont. 79 35 activated 40-49 274 102 cont. 79 20 ? ? 150-225 ? cont. 68 19 activated 28-96 243-272 ? cont. 65 activ/prim. 6 primary 41-107 200-250 32-68 cont. 68 2.4. END PRODUCTS

For the final disposal of the solid end products, these have ·to be separated from the liquid effluent. This process has been studied by Walters and Ettelt (83) and by Teletzke (84,

85) •

The liquid effluent may contain a considerable amount of dis-solved matter and could be returned to the biological

oxi-dation unit.

The chemical composition Teletzke (75).

2 • 5 • COST OF THE PROCESS

of the effluent has been studied by .

The cost of the high temperature version of the wet-air oxi-dation process is determined by the desired COD reduction. Teletzke (24, 25) gives the relative cost as a function of COD reduction.

The absolute cost presented in the literature shows a

consid-erable scatter.

For the Chicago installation Goldstein (82)

reports $23 per ton dry sludge,

which does not include

inter-est on capital invinter-estment.

Five years later,

Dalton (32)

mentions $50

per ton for the

same

installation.

He

also refers to cost of other sludge

treatment processes at the Chicago sewage plant.

2.6.

LITERATURE REVIEWS

In the literature a great many

reviews on wet-air oxidation

have appeared in several languages.

In English,

e.g. refs.

(92)/(97), (111); in German, e.g. refs. (98)/(102); in Polish ,

e.g. refs. (103,

104); in Dutch, ref. (105); in Swedish,ref.

(106). Some additional experimental data can be found in refs.

(107, 108, 109).

2.7.

CONCLUSIONS

From the literature it follows that the wet-air oxidation of

sludge is influenced by the temperature

Quantitative influences are not published.

and the pressure.

Quite clear is the

effect of temperature on the maximum conversion;

however, the

rate by which this is achieved is unknown. Furthermore, no

in-formation was found concerning

(i)

The kinetics of the reaction.

(ii)

The influence of oxygen transfer from the gas phase into

the suspension.

(iii) The dissolution of solid sludge particles,which may have

an effect on the oxidation rate by, e.g., different

sta-bilities of solid and hydrolised, dissolved sludge.

(iv)

The influence of mixing in the reactor.

Without knowledge of or insight into these factors,

a proper

design of a wet-air oxidation process cannot be expected. Only

after realisation of such a design can a fair comparison of

cost with conventional sludge treatment procedures be made.

Chapter 3

THE OXIDA1.ION OF A GLUCOSE

SOLUTION AS A MODEL SLUDGE

In order to obtain insight into the general behaviour of wet-air oxidations, the research project was started with the oxi-dation in a two-phase gas-liquid system of a model sludge for which a solution of glucose was selected, glucose being repre-sentative of the group of carbohydrates.

3.1. PRELIMINARY EXPERIMENTS

Semi-batch experiments were carried out in a one-litre elec-trically heated autoclave which is diagrammatically shown in figure 3.1. For each experiment about half a litre of water was heated in the reactor to the desired reaction temperature. A vertically moving agitator (shaker), placed inside the

reac-oxygen/air spent gas injector cooler

t

sample cooler Figure 3.1. Semi-batoh installation.

tor, provided mixing of the contents of the reactor. 20 to

40 ml of concentrated glucose solution was injected and air or oxygen was fed into the system and passed over or through the

glucose solution. The spent gases were cooled in a condenser

from which the condensed steam flowed back into the reactor. A

dip-pipe enabled samples of the liquid in the reactor to be

drawn.

In order to ensure that the time necessary for mixing the

liq-uid phase homogeneously was short enough not to influence the

outcome of the experiments,a conductivity electrode was placed

in the reactor. Then a concentrated sodium chloride solution

was injected and i t was found that the solution was nearly

completely mixed up in five seconds at the shaker frequency of 144 min-1.

By heating a glucose solution a number of sequential dehydra-tion and polymerisadehydra-tion reacdehydra-tions take place. By these thermal reactions the solution is coloured more or less brown, and ob-tains a sweet smell of caramel (46).These reactions could also

take place during oxidation experiments and oxidation could proceed directly from glucose as well as via thermal reaction products.

The influence of thermal treatment on subsequent oxidation was studied by changing the time between injection of glucose and supply of oxygen 1.which changes the extent of thermal reaction.

The results of these experiments are presented in figure 3.2. Because of the large number of possible reaction products, it was decided not to follow each component individually, .but to

trace the total amount of oxidisable material. The concentra-tion of this oxidisable material is expressed in the chemical oxygen demand (COD).

The experiments were carried out at 178 and 200°c and 50 atm. At the temperature of 178°c air was passed through the solu-tion at. a flow rate of 1 Nm3 /h. · The reactor was filled with

15 COD (kg/m3 ) 10 5 0 (1) non-preheated, 178°c (2) preheated, 178°c · ' - - - . . . < ' _\ -..., _...

,

(3) preheated, 200°c \

'

_{' , ( 4) non-preheated, 200°c} ...

,

...

,

'

_... ... (1) (3) .... _

--

_--

-(4) 10 2 time (min) Figu'I'e 3.2.

500 ml water enabling the shaker to force gas bubbles into the

solution from the gas space above it (see 3.1.3.). The total

-1

pressure was always 50 atm and the shaker frequency 144 min

•

In the experiment belonging to curve (1),

the temperature was

178°c and air was admitted to the reactor immediately after

the injection of glucose.

The curve indicates

that it takes

several minutes before noticeable oxidation takes place.

Ap-parently, active material has first to be formed.

The experiment of curve (2)

was carried out at the same

tem-perature, but now

the glucose solution was heated for 10

mi-nutes

in the absence of air so that only thermal reaction

*)

took place.

Then air was admitted and now,

as follows from

curve (2), the oxidation started nearly immediately.

This could be understood by assuming that the active material

is formed by thermal reaction.

The active material could be a

catalyst for the oxidation of glucose and other thermal

reac-tion products,

but it is also possible that we are dealing

with a consecutive reaction, which would mean that thermal

re-action products are more reactive than glucose.

De Wilt (122)

has shown

that the oxidation of glucose in alkaline solutions

at about 60°c proceeds through enolate ions,

generated from

glucose by thermal reactions, which makes the last mechanism

most likely to occur in our case.

The experiment at 178°c with non-preheated glucose was

re-peated with pure oxygen passed over the solu.tion at 50 atm, a

shaker frequency of 144 min-

1

,and a gas flow rate of 0.6 Nm 3/h.

The reactor was now filled with more than 650 ml water,

so

that no bubbles were forced into the solution

(see 3.1.3.),

resulting in a much lower gas-liquid interfacial area compared

with the above described experiments. The curve obtained

coin-cides nearly completely with curve (1). So, also with pure

ox-ygen, a "lag-phase" is obtained.

*)

By

thermal

reaction we mean

those reactions which take

place without the influence of oxygen.

Curve (3) represents the experiment with non-preheated glucose

at 200°c. Oxygen was passed over the solution at a flow rate

of 0.6 Nm3/h, the reactor being filled with more than 650 ml

water. At this temperature oxidation seems to start almost

immediately. By preheating at this temperature for 10 minutes,

the subsequent oxidation, indicated by curve (4), proceeds

more slowly compared with the non-preheated glucose.

At 200°c thermal reactions will have proceeded further than at

178°c, so, apparently some thermal reaction is necessary for

rapid oxidation; however, too much thermal reaction decreases

the rate of oxidation. This could be explained by assuming

that the generated active material is degraded by further thermal reaction.

When the solution was preheated for over one hour at 200°c,

almost all glucose was converted into a carbon-like product

which was deposited on the reactor wall and the shaker.

From figure 3.2 i t follows also that complete oxidation is not reached at the selected temperatures, but a maximum conversion

is obtained. Since only a slight difference occurred in

maximum conversion for a preheated glucose solution and a

non--preheated one, this maximum conversion can only partly be

in-fluenced by thermal reaction products. Apparently, i t is

mainly determined by oxidation products, which resist further

oxidation under the experimental conditions.

In order to get an impression of the influence of the shaker

frequency on the gas dispersion, model experiments at room

temperature were carried out in a glass vessel of the same di-mensions as the autoclave, the glucose solution being replaced

by hexane, which at room temperature has a viscosity and

sur-face tension near those of water at 200°c.

lu-enced by the liquid level in the reactor. At high levels, when

the upper blade of the shaker was completely submerged in the

liquid as well as the gas-inlet, only a very small fractional

gas hold-up (less than 0.01) was observed, which originated

from the bubbles passing through the liquid.

At lower liquid levels, when the shaker passed through the

gas-liquid interface, a large amount of gas phase was forced

into the solution as bubbles resulting in a gas hold-up of

a-bout 0.11 at a shaker frequency of 144 min-1• At low

frequen-cies all bubbles escaped from the liquid between two shaker

cycles. The observed average bubble diameter {db) seemed to be independent of the shaker frequency and was 2 to 3 mm.However, the average number of bubbles in the liquid increased with the shaker frequency, resulting in an increase of the average

gas--1

liquid interface. At the maximum frequency of 144 min and a

liquid volume of 600 ml, the fractional gas hold-up {£) was

0.11. Consequently the specific surface area of the bubbles,

6£ ab

=

d '

b

was about 270 m-1

When changing the liquid level, i t was observed that for

liq-uid hold-ups larger than about 650 ml the shaker was complete-ly submerged during shaking.

So, by applying a liquid hold-up exceeding 650 ml a relatively low specific surface area must be expected.

The influence of the shaker frequency on the oxidation rate

was determined also at 200°c and 50 atm with an air flow of l

Nm3/h. The liquid hold-up was 500 ml so that according to the

model experiments i t must be expected that air had been forced

into the solution. The results are shown in figure 3.3.

It follows from this graph that the shaker frequency

influ-ences the conversion rate. This indicates that the latter

20

0

_{20 min-l}

c:l

35

II +

60

II

15

x

144

II ( COD

(kg/m3 )

10

5 0 0 1 1 2

time (min)

Figure 3.3.

Effeat of shaker frequency on

o~idation

at

200°c

and

50

atm.

experiments

the conversion rate is also determined by dif

fu-sion.

From the experiments discussed in 3.1.3. it followed that the

oxidation rate of glucose under the conditions of the

experi-ments was lowered by diffusion limitation. so, a direct

deter-mination of the kinetics of the reaction was not possible with

such experiments.

owing to the relatively high solubility of

oxygen above l00°c

(see fig. 3.14), the oxidation of glucose

could be carried out in a homogeneous system.

The results are

shown in figure 3.4.

0.6·

COD

(kg/m3 )

0.2 0

Figure J.4.

200°c

[ob]

=

1 kg/m

3 50

time (sec)

Homogeneous

o~idation

of gZuaose.

In the autoclave

500

ml water of

200°c

was saturated with

oxy-gen to a concentration of 1 kg/m

3

•

The shaker frequency was

144 min-

1

•

Then some glucose solution was injected, resulting

in a initial COD

of 0.6 kg/m

3

•

With regard to complete

oxi-dation, a surplus of oxyg.en of nearly 70% was present.

Although the oxygen consumed by the reaction will be supplied

by mass transfer from the gas phase,

the original surplus of

oxygen present in the liquid will guarantee that diffusion

limitation has only little influence on the observed reaction

rate.

From figure 3.4 it may be seen that the homogeneous oxidation

proceeded so rapidly that after 30 seconds, when the first

sample was taken,

the maximum conversion was already nearly

reached.

For the heterogeneous oxidations which were shown in

figure 3.3,

it took 10 minutes

to reach the same degree of

conversion, while the equilibrium oxygen concentration there

was 1. 4 kg/m

3

•

It is quite clear that homogeneous experiments carried out in

the way described here, are too fast for studying the kinetics

of the oxidation of glucose.

As mentioned in

3.1.2.

complete oxidation was not reached. The

maximum conversion,

defined as the COD reduction after one

hour, was determined as a function of temperature.

Figure

3.5

shows that it increases from

75%

at

170°c

to

90%

at

26o

0

_c.

_{Results are also presented which were obtained from}

a

further oxidation of samples of continuous flow experiments

(see

3.2.).

As follows from the graph, the maximum conversion

data are in good agreement with the results of Yunis

(71),

who

studied the wet-air oxidation of glucose in the presence of

hydrogen peroxide and ferric ions.

This indicates that

prod-ucts resisting wet-air oxidation also resist the action of

hy-drogen peroxide/ferric ions.

100

maximum

COD

reduction

(%) 75 50 25 FiguPe 3.5. T

glucose

• preoxidised glue.

• glucose+ H

_{20 2}

(71)

• sugar

(70)

temperature (

0c)

From the preceeding sections i t followed that complete

oxida-o

tion was not reached even at a temperature of 260 c. It was

found from the continuous flow experiments that the amount of

coo

that could be removed, the "effective"

coo

[c], is a

pa-rameter by which the overall conversion rate per unit volume

can be described with a half order in effective

coo.

For the

semi-batch experiments this results in

-

d£~l

=constant*

[c]~

Integration of eq. (3.1), using the boundary condition

t

=

O,

[c]

=

[c

0] ,

results in

[c

]~

-

[c]~

=constant* t

0

By introducing y =

1s1

this transforms into

[co] 1 - YJ..z

=

constant *t [ c ] J..z 0 (3.1) (3.2) (3.3)

Since this description does not include generation of the ac-·tive matter, i t may only be applied after the "lag-phase".

In figure 3.6 the results of the experiments discussed in

3.1.2. are plotted as 1 -

y~

against time. This results in

straight lines up to 80% of the maximum conversion.

The deviation at higher conversions might be the result of the formation of rather stable oxidation products.

Figure 3.6 also shows that when preheating for 10 minutes in

the absence of oxygen at 200°c, the subsequent oxidation · pro-ceeds more slowly compared with a non-preheated solution.

Finally i t follows from this graph that 10 minutes preheating

at 178°c results in a higher conversion rate compared with

1. 00 l-y15

(1) non-preheated

(2) preheated

time (min)

178°c

Figure 3.6.

( 1) (2)

non-preheated

(2) preheated

10

time

0

(min)

200

c

Results plotted as half-order aonversion rate.

3.2.

CONTINUOUS FLOW EXPERIMENTS

1. 00

0.75

a.so

0.25

o.oo

Figure 3.7 shows the flow sheet of the bench-scale continuous

flow installation.

The reactor consisted of a stainless steel bubble column with

an internal diameter of 0.04m and a length of 1.00 m.

Glucose solution and air

steel electric preheaters

respectively, and fed into

heater : 12

1).

were heated separately in stainless

with capacities of 7.5 and 0.5 kW

the reactor (volume of liquid

pre-sludge

Figure 3.?.

Continuous

fto~

instaiiation.

i--.,_~pen t

gas

air

A: studge pump;

B: preheater;

C: reaator;

D: aooier;

E: separator;

F:

ao~pressor.

Air was dispersed by a multiple orifice distributor,

diagram-matically shown in figure 3.8, and rose in co-current with the

liquid through the reactor.In some experiments the air was fed

into the system just before the sludge preheater so that

glu-sludge

inlet

air

inlet

Figure 3.8.

cose and air were heated simultaneously. This was not the

nor-mal procedure, however, as in that case the locus of oxidation

was not clearly defined.

The gas and liquid phases were removed at the top of the

reac-tor and then cooled rapidly. After pressure expansion, gas and

liquid were separated.

The

temperature in the reactor was

measured by thermocouples in cylindrical wells at the bottom,

half-way,

and at the top of the reactor.

The temperature was

controlled by the preheaters and could be kept constant within

' 0

1 - 2

c.

The total pressure was measured at the top of the

re-actor and was kept constant by a back-pressure valve.

Liquid

phase samples could be drawn half-way and at the bottom of the

reactor.

The oxygen concentration in the spent gas was

meas-ured with an oxygen analyser.

In order to change

the oxygen

pressure,

oxygen or nitrogen could be mixed with the air feed

of the compressor.

The experiments were all carried out at a total pressure of 50

atm and a gas flow rate of 1.7 Nm3/h.

Unless mentioned

other-wise, air was used while the liquid feed rate was l0- 2m3/h,the

feed having a COD of

35 kg/m

3

•

owing to variations in

thee-vaporation with temperature,

the average residence time was

dependent on the temperature.

In figure

3.9 curve (1) represents the influence of the

tem-perature in the bubble column on the COD reduction.

At a temperature of about

2oa

0

c

the COD reduction suddenly

dropped.

Presumably this was caused by a

too far proceeded

thermal reaction in the preheater.

In the semi-batch

experi-ments it was already shown that some thermal reaction was

nec-essary for fast oxidation, but that too much thermal reaction

decreased the oxidation rate (see 3.1.2.).

In accordance with the foregoing, the oxidation will proceed

faster when it takes place during or after the first steps of

the thermal reactions, which means in the sludge preheater. In

100

0 _{air straight into reactor}

+ air via sludge preheater

75 0 COD _~+- (2) reduction

.~:

'.'.!-,...+ (%) 50 0

I

b

I

A " (1) I

-.__.

25

_oh

0 190 210 230 temperature (0

c)

Figure

3.

9.

Effect of temperature on continuous ftow oxidation of gtucose.

order to test this hypothesis, air was mixed with the glucose solution before i t entered the sludge preheater so that glu-cose and air were heated simultaneously.

Also with this arrangement the influence of temperature was determined. The results are given in figure 3.9 by curve (2) which shows that the drop in conversion has disappeared.

Up to though

2os0c the COD reduction is the gas-liquid contact time

somewhat lower now, even is about twice as high as for curve (1). No gas distributor was used, however, which definitely results in a lower interfacial area than in the ex-periments described above. This again confirms that the oxi-dation rate is dependent on the gas-liquid interface, as con-cluded from the semi-batch experiments.

Since the locus of oxidation is only defined when no oxidation takes place in the preheater, the experiments discussed in the

following were always carried out in such a way that air and

glucose were heated separately.

3.2.3. ~n!1Y~n£~_g£_£QQ~22n2~n~=2s!2n_!n_sh~-£~~§_2n§_g~yg~u

E=~!!Y=~

At a temperature of

220°c

the influence of the COD

concentra-tion in the feed was measured.

Results are presented in table

3.1.

It follows from these data that the conversion rate

in-creases only slightly with the COD concentration,while the COD

reduction drops.

This indicates that the conversion order in

the organic material will be somewhere between zero and one.

TabZe 3.1.

InfZuenoe of COD in the feed.

CODf

COD reduction conversion rate

(kg/m

3) (%)

(kg/h)

20.3 72 0.146

29.4 60 0.170

33.9

55

0.186

The oxygen pressure

with the air feed

was changed by mixing oxygen or nitrogen

of the compressor, keeping total pressure

and gas flow constant.

The influence of the oxygen pressure was examined at several

reaction conditions. The results are presented in figure

3.10.

At the temperatures of

207

and

213°c

the liquid flow rate was

5.2 x

10-

3 m3/h, while the COD concentration in the feed was

3 . -2 3

38

kg/m .At the other temperatures these values were

10

m /h

and

37

kg/m

3,

respectively.

From figure

3.10

it can be concluded that the oxygen pressure

affects the conversion rate as could be expected, but that

0. 15 .

conversion

rate

(kg/h)

0. IO ·

o.os

0

Figure 3.10.

oxygen pressure (ata)

Effect of oxygen pressure on continuous flow oxidation of

glucose.

Regarding the influence of concentrations,

it was found

that

-the oxygen flux through -the interface q

0

can be expressed in

terms of oxygen pressure, p, and effective COD, [c]

constant

*

[ c] ?..i

_(3.4)

3.3.

THEORETICAL ANALYSIS OF RESULTS AND DISCUSSION

For a proper analysis of the results,

the mixing state in the

reactor must be known. Therefore, residence time distributions

of the liquid phase were measured,

using sodium chloride as a

tracer.

It was possible to describe the measured distributions as

plug-flow with axial mixing (84, 85). The corresponding Pc§clet

numbers were of the order of 3.5, showing that only a few

mix-ing stages were present. For a definition of the PAclet number

see

3.3.2.

Figure

3.11

shows a measured and a calculated

resi-dence time distribution.

c*

o.s-~~

measurements

calculated

for Pe

=

3. 5

1 2 0 --~~~~--~~~~..__~~~---'

t/t

Figure 5.11.

Measured and aaZauZated residenae time distribution.

In experiments at

190°c,

samples were taken from the liquid in

the reactor,

in which the nitrogen concentration, [cN ], was

2

measured.

The equilibrium concent.ration belonging to the nitrogen

pres-*

sure in the reactor

[cN ] was calculated, using Pray's (78)

solubility data for

2 pure water.

From these two values the overall volumetric mass transfer

co-efficient, K1a, was determined, using the equation

A mean value for Kia of 4.1 x 10- 2 sec- 1 was found. Using this

value for the physical absorption rate of oxygen, and assuming

zero oxygen concentration in the liquid,

the maximum physical

absorption rate of oxygen was calculated to be 50 g/h.

Since the observed conversion rate was 145 g/h,

the oxidation

must take place mainly within the diffusion layer around the

gas bubbles.

3.3.2. Model for the macro kinetics

---From the continuous flow experiments

it followed that the

ab-sorption rate of oxygen is proportional to the square root of

both oxygen pressure and effective

coo.

In the theoretical

a-nalysis presented in 3.3.3.

an expression will be derived for

the constant factor in equation

(3.4).

Introduced in the

ab-sorption rate per unit interfacial area,

q

0 ,

becomes

=

v

2D

0

~[c]p

_(3.5)

With this expression a mathematical model of the conversion

rate obtained in a continuous flow reactor was set up,

includ-ing combined mass transfer and reaction,

convection and

mix-ing.

The model is based upon the following starting-points:

(i)

Equation (3.5)

describes the local oxygen transfer rate

through the gas-liquid interface, using the local values

of [c] and p.

{ii)

The residence time distribution in the liquid phase can

be described as plug flow with axial mixing.

(iii) The gas flows in co-current with the liquid in pure

plug-flow.

{iv)

The radial mixing is so high that in radial direction

the concentration profiles are flat.

(vi) The gas flow, liquid flow and interfacial area are

uni-form throughout the reactor, which means that all

evap-oration takes place at the inlet.

(vii) The reactor is operated in steady state.

A mass balance of COD over the liquid phase between the

cross-sections in the reactor at the heights x and x + ~x results in

the following differential equation

d r,..l d2 [C] L L

~

- u .::..i..£..i.. + E - - - p~[c]~ a= 0

JI, dx dx2 (3.Sa)

where

UR,

.

.

superficial liquid velocity

x

.

.

length co-ordinate in reactor

E : eddy diffusivity taken per unit reactor volume

a

.

• specific gas-liquid interface taken per unit reactor

volume

A mass balance of oxygen over the gas phase results in

u

f!l_o

k

....S:

~

+

p~[c]~

--St: a

=

0

RT dx H (3.6) where R gas constant T absolute temperature

u superficial gas velocity.

g

The boundary conditions for the two simultaneous differential

equations are

x = 0 _p

=

_pf

[cf]

-

[co]

=

E (fil.£1)

UR, dx x = 0

In this the subscript f refers to feed conditions. By

intro-ducing the following dimensionless variables and parameters

*

_[c]/[cf]

y

₌

1T

=

_p/pf

a

₌

x/X

Pe

₌

_E

UR,X

(Peclet nwnber)

2D kRT

Nr

=

_{H U.v,Ug}

o-

a X(number of conversion stages)

M

=

uR.[cf]RT

=

feed rate of oxidisable material

,

ugpf

feed rate of oxygen

the equations (3.Sa) and (3.6) can be transformed into the

di-mensionless equations

dy*

₊

_L

d2~*

_-

~1T~YR~

=

0

(3.7)

do

_{Pe do}

_M

d1T

₊

_N

M~yR~

_{1T ~}

₌

0

( 3.

8)

da

r

while the boundary conditions become

o

₌

0 1T

=

1

l

-

y

*

=

L

_{Pe do}

dy.,

o

₌

1

dy*

_do

₌

0

The two simultaneous equations were solved on an analogue

com-puter.

Some of the results are given in the figures

3.12 and

3.13. These and additional results have already been published

elsewhere (125).

In order to test the model, in figure 3.12 the results of

ex-periments presented in 3.2.3. have also been included.

100 75 conversion (%) 50 25 Figure 3. 12. Effeat of M and N NP is indicated

i~

+ 220°C • 213 y 207 • 186

on aonvePsion of gZuaose foP Pe= 3.5. the gPaph.

while the influence of M was determined by changing the COD

of the feed solution or the oxygen pressure of the feed gas.

The conversion is based upon the maximum conversion and is,

therefore, expressed as a percentage of the reduction of the

effective COD.

In the graph the line for complete utilisation of the oxygen

feed is also drawn, represented by the equation

conversion

=

lOO

(%)

M

It follows from figure 3.12 that there is a reasonably good

agreement between the calculated curves and the measurements.

Another comparison is represented in figure 3.13. This graph

shows several measured local concentrations in the reactor and

the corresponding calculated concentration profile. This also

1.0 y* 0 data points calculated curve for Pe

₌

3.5

0""

Nr

=

1. 25 0.5 M

₌

0.5 0 0 0 0.5 _1.0 a

Figure 3. 13.

Calculated and measured concentration distribution in reactor.

It may be concluded that the mathematical model gives a fair

description of the phenomena in the reactor which influence

the overall conversion.

3.3.3. Model for the micro kinetics

---To understand the square root dependence of the absorption

rate on the concentrations, models for mass transfer with

chemical reaction were examined. From the experiments i t

fol-lowed that the reaction was fast compared with diffusion of

oxygen and, therefore, that the oxidation took place in a thin film near the gas liquid interface.

In this section i t will be demonstrated that a penetration

model of absorption of oxygen, followed by rapid oxidation,

which chemical-kinetically is of first order in organic mate-rial and zero order in oxygen, results in the observed conver-sion orders.