See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/254854008

Scaling Up Bubble Column Reactors

Article · January 2000

CITATIONS READS

21 538

1 author:

M.I. Urseanu
DSM Resins
23 PUBLICATIONS   1,017 CITATIONS   

SEE PROFILE

All content following this page was uploaded by M.I. Urseanu on 21 August 2014.

The user has requested enhancement of the downloaded file.

 ChapterChapter 1 Introduction Introduction

ChapterChapter 1

Introduction Introduction

Whyy bubble column reactors?

Thee extensive and sustained research in the field of bubble column reactors is there
becausee of the broad range of potential applications in the chemical, petrochemical and
biochemicall industries.
Bubblee column reactors are simple vertical cylindrical vessels with intense contact
betweenn the two phases. The gas phase is dispersed into the liquid phase using specific gas
distributorss at the bottom of the column. The gas phase is in motion with respect to the
continuouss phase. The continuous phase (liquid or slurry) may be operated in a batch manner
orr it may also circulate in co-current or counter-current manner to the gas flow. A simplified
representationn of a bubble column reactor is shown in Fig. I. I.

Gass phase outlet

G,, > 0

Liquidd phase outlet L,

>0 0

Solidd catalyst particles

Internall heat-exchange (eventually) )
tubess (eventually)

Gass distributor
Liquidd phase inlet Ln
>0 0

Gass phase inlet

G0>0 0

Fig.. 1.1. Schematic drawing of a bubble column reactor.

Sincee bubble columns offer several advantages in comparison to other kinds of multiphase
reactorss (see panel 1.1 and panel 1.2), they are subjected to an increasing use in industrial

11 1
ChapterChapter 1 Introduction Introductio

practicee (as absorbers, fermenters, strippers, coal liquifiers and chemical reactors for gas-
liquidd and gas-liquid-solid reactions).
Duee to the several practical applications of bubble column reactors, these reactors have
beenn the subject of detailed research studies and analyses. This research has pointed out
severall aspects related to the advantages of bubble column reactors:
Panell 1.1 Advantages offered by bubble column reactors
efficient contact between the phases, the gas and the liquid, and eventually the third
phase,, the solid catalyst
high liquid hold up, recommended for reactions taking place in the liquid phase (as the
casee of bubble columns)
reasonable inter-phase mass transfer rates at low energy input
limitation of pressure drop
easy temperature control
little maintenance due to the simple construction
lack of moving parts
high adaptability for a specific process
no serious erosion and plugging problems due to the catalyst
»» low costs of construction and operation
Thee disadvantages of bubble column reactors should also be mentioned:
Panell 1.2 Disadvantages of bubble column reactors
considerable degree of backmixing in both the liquid and the gas phase
short gas phase residence time
higher pressure drop with respect to packed columns
rapid decreasing of interfacial area above values of the aspect ratio greater than, say 12,
duee to the increased rate of coalescence
Severall types of processes of major importance in industry take place in bubble columns,
somee of them being outlined in Panel 1.3. For more examples see Mashelkar (1970), Shah et
al.. (1982), Deckwer and Schumpe (1993).
Panell 1.3 Practical examples of reactions taking place in bubble columns and slurry reactors
oxidation reactions (e.g. oxidation of cyclohexane to adipic acid, partial oxidation of
ethylenee to acetaldehyde, oxidation of n-parrafins to sec-alcohols)
hydrogenation reactions (e.g. saturation of fatty acids, hydrogenation of glucose to
sorbitol) )
chlorination reactions (production of aliphatic and aromatic chlorinated compounds)
hydrotreating and conversion of petroleum residues
fermentation (production of ethanol and mammalian cells)
biological waste water treatment
oxidesulfurization of coal
oxichlorination of ethylene to dichlorethane
Fischer-Tropsch synthesis
methanol synthesis
polymerisation of olefins

12 2
ChapterChapter 1 Introduction Introduction

Bubblee column reactors are particularly suited for slow reactions taking place in the liquid
phase.. The main resistance to the mass transfer is located in the liquid phase. The gas-liquid
contactt achieved is reflected by the parameter (3 which is the ratio of the liquid phase volume
too the volume of the "film" diffusion layer. High values of (3 are obtained with bubble column
reactorss (see Fig. 1.2) and are suitable for reactions which demand high "bulk" liquid volume.

Fig.. 1.2. Bubble column selection criteria (Krishna and Sie, 1994).

Becausee of the high degree of mixing, there is no need of additional internal mixing
devices.. In some cases continuous and efficient temperature control is easily achieved by
meanss of internal heat exchangers, both for cooling or heating the system. In this case, also
thee backmixing phenomena is reduced; this aspect has positive advantages for some
applications.. Furthermore, bubble column reactors may be used when the fluids carry solid
impurities,, that would plug packed columns. In this case, the fine solid catalyst is kept in
suspension,, such as in the Fischer-Tropsch synthesis.

Scale-upp methodology
Scale-upp basically means the conversion from laboratory scale experiments to successful
commerciall size design. The proper scale-up and design is accomplished by developing a
reactorr model. Typical of laboratory work, experiments are carried out in columns of
relativelyy small diameters, ranging from 0.1 to 0.5 m, whereas in industry 1 to 10 m diameter
reactorss are commonly encountered (Fig. 1.3).
Changess in reactor diameter, height and superficial gas velocity will be accompanied by
changess in flow regime, phase hold up characteristics, axial and radial mixing and transport
coefficients.. There is therefore a scale up "problem".
Emergingg examples of processes using this reactor technology, such as Fischer-Tropsch
andd methanol synthesis, ask the chemical engineer in this field to be able to deal with the
followingg extreme features: high pressure operation (typically 30 to 50 bar), together with
highh superficial gas velocity (typically higher than 0.1 m/s), use of high catalyst
concentrationss in the liquid phase, high reaction exothermicity requiring heat removal by
meanss of cooling tubes inserted in the reactor, large dispersion heights (typically 20-30 m)
andd large reactor diameters (up to 10 m) because of high gas throughputs, often operated in
parallel. .

13 3
ChapterChapter 1 Introduction Introduction

Fig.. 1.3. Comparison between laboratory and industrial bubble column size.

Successfull scale-up methodology starts with proper understanding of physical phenomena

andd the inter-relations between parameters and it ends with the appropriate quantification of
thee various variables and especially sets of variables associated to reactor performances (i.e.
conversion,, selectivity, productivity).

Objectivee of the thesis

Despitee their simplicity in construction, operation and maintenance, gas-liquid bubble
columnn reactors are notoriously difficult to scale up. The hydrodynamics is strongly affected
byy column geometry, system properties, pressure, presence of surface active agents, etc. The
objectivee of this thesis is to develop a systematic scale up strategy which takes all such effects
intoo account. The key, distinguishing, feature of this approach is that the overall column
hydrodynamicss is described in terms of single bubble rise velocities and bubble-bubble
interactions.. Once the bubble hydrodynamics is properly described as a function of scale and
systemm properties then the column hydrodynamics is fully determined. The "key" to our scale
upp strategy is to link the various hydrodynamic phenomena (gas hold up, circulation velocity,
mixing). .

14 4
ChapterChapter 1 Introduction Introductio

Examplee of an industrial scale-up problem

AA typical industrial scale-up problem is now outlined (see Fig. 1.4), to give an idea of the
importancee of proper and adequate scale-up.
Thee industrial process considered takes place in a huge reactor: 6 m diameter and 35 m
height.. The column operates at 0,2 m/s superficial gas velocity and it contains a highly viscous
liquid.. In order to fully describe the behaviour of such industrial reactors it is required to
extrapolatee some parameters such as the gas hold up and the liquid phase backmixing.

Industriall process
** Column diameter 6 m
** Column height 35 m
** Superficial gas velocity 0.2 m/s
** High viscosity liquid

Scalee up data we need

** Gas holdup
** Liquid phase residence lime
** Liquid phase backmixing

Fig.. 1.4. Scale up problem.

Inn order to obtain the scale-up information for this process, an extensive campaign of
measurementss started in our laboratory (details are reported in this thesis in Chapters 2, 3, 4, 5
andd 6). The investigation of scale effects on hydrodynamics and mixing phenomena was
carriedd out by using a wide range of bubble columns of different diameters: 0.051, 0.1, 0.15,
0.174,, 0.19, 0.38 and 0.63 m. The influence of scale (i.e. column diameter) was carefully
investigatedd on single bubble rise behaviour, bubble-bubble interaction, gas hold up, liquid
circulationn and liquid phase backmixing. The influence of physical properties of the liquid
phasee on the column performance was also taken into account, by experimenting extreme
systems:: air - water and air - Tellus oil (a mixture of hydrocarbons 75 times more viscous
thann water). We developed a consistent data bank on which to develop our scale up model and
strategy. .
Consider,, for example, the gas hold up measurements, in columns of 0.1, 0.19 and 0.38 m
forr air - Tellus oil system. The results (Fig. 1.5) demonstrate quite clearly the significant
influencee of column diameter.
Ass we can see in the detail in Fig. 1.6 the laboratory range of experiments is far removed
fromm the industrial range. For a commercial scale reactor operating at a superficial gas
velocityy of 0.2 m/s, various extrapolation schemes lead to completely different answers. What
wayy should be chosen when performing this extrapolation? A linear, an exponential or a
powerr law one? Or even another one?
Lett us examine first what the published literature has to say about this extrapolation
problem.. The easiest and most experimented field in bubble columns is in fact the gas hold
up.. Some of the literature correlations are plotted in Fig. 1.7 together with our experimental
pointss (details given later in Chapter 4).

15 5
ChapterChapter 1 Introduction Introduction

DTT = 0.19 m

\\
A
^ oo o o D = 0.38 m
AA TT
A D
DDTT = 0.10m . Do o o n
0.155 - A
, aD^ o ' 'O O
AnnC;c c ' '

air r-- Tellus oil

'71 1== 0.075 Pa s
3
Pi i == 862 kg/m
0.055 - (7 7 == 0.028 N/m

0.11 0.15 0.2 0.25 0.3

UUGG I [m/s]

Fig.. 1.5. Influence of scale on the gas hold up.

0.22 r

0.255 r

Fig.. 1.6. Extrapolation problem: what is the gas hold up for a 6 m column operated at 0.2
m/ss superficial gas velocity?

16 6
ChapterChapter J Introduction Introduction

0.4 4 UUGG = 0.2 m/s

OO experimental data

0.3 3
Zehner(1989) )

££
GG 0.2
Wilkinsonn etal. (1992)
rr , 0.2

Akitaa and Yoshida (1973)

0.16 6
Riquartss and Pilhofer (1978)
Westermeyerr (1992)
0.122 -
0.2 2 0.44 0.6 0.8 8 11
DT/[m] ]

Fig.. 1.7. What does previous research say about scale dependence on gas hold up.

Singlee bubble
risee velocity

Liquidd phase Bubble-bubble e

backmixing g interaction n

Axiall liquid velocity distribution Gass hold up

Centre-linee liquid velocity Bubblee swarm velocity

Fig.. 1.8. Strategy for scaling up bubble column reactors.

17 7
ChapterChapter 1 Introduction Introductio

Thee literature survey revealed a striking fact: the majority of the researchers did not take
intoo account the influence of the column diameter on the gas hold up! Moreover, the
differencee between the predicted values is rather large. Furthermore, we have to keep in mind
thatt all these correlations were developed for columns smaller than 1 m in diameter and most
off the time by using just one column.
AA careful scale up strategy is clearly required. A schematic representation of this is shown
inn Fig 1.8. In this thesis each of the steps outlined in Fig. 1.8 has been thoroughly
investigated. .

Inn Chapter 2 we study the rise velocity of single gas bubbles in a liquid in order to determine
thee scale effects.

Bubble-bubblee interactions are studied in Chapter 3 and a model is developed to determine

thee rise velocity of bubble swarms.

Thee model developed for bubble swarm velocities is used to predict the gas hold up. Chapter
44 compares the predictions against experimental data for a variety of systems.

Thee rising gas bubbles carry liquid upwards and create liquid circulations. Chapter 5 studies
thee scale dependence of the liquid circulation velocities measured in three different columns.

Chapterr 6 presents experimental data on the axial dispersion coefficient of the liquid
measuredd in three different columns. A simple model is developed for scale up purposes.

Finallyy in Chapter 7, we show how the information developed in this thesis can be
incorporatedd into a CFD code (developed in a companion thesis of van Baten, 2000) to predict
thee performance of commercial scale reactors.

18 8

View publication stats