Water Research 36 (2002) 1585–1595

The influence of contact zone configuration on the flow

structure in a dissolved air flotation pilot plant
Ma( ns Lundha,*, Lennart Jonsson
. b
, Jan Dahlquistc
a
Department of Water and Environmental Engineering, Lund University, P.O. Box 118, 221 00 Lund, Sweden
b
Department of Water Resources Engineering, Lund University, P.O. Box 118, 221 00 Lund, Sweden
c
PURAC Sweden, P.O. Box 1146, 221 05 Lund, Sweden
Received 22 December 2000; received in revised form 3 July 2001; accepted 20 July 2001

Abstract

The dissolved air flotation process is used in water and wastewater treatment. Among many parameters the fluid
dynamics determine the capacity of the process. The contact zone is assumed to be important for the removal function,
as it is believed to be the location for the aggregation of bubbles and flocs. This paper presents an experimental study on
the flow structure in a contact zone of a DAF pilot tank and the influence of contact zone configuration. The flow
structure in the contact zone was examined for different horizontal lengths of the zone and for different heights and
inclinations of the shaft wall. The hydraulic surface loading was 11 m/h over the total tank surface area and the recycle
rate was constant at 10% of the main flow. The examined hydraulic surface loading over the contact zone ranged from
40 to 98 m/h. Water velocities in the longitudinal, central section of the tank were measured with an acoustical Doppler
velocimeter in a grid net for the different contact zone configurations, giving an insight into the flow structure. The
result showed that the flow structure in the contact zone was characterised by a turbulent lower region and a plug-flow
higher region. The hydraulic surface loading, a function of the length of the contact zone, seemingly determined the
extension of the turbulent region. A higher hydraulic surface loading decreased the turbulent region while the lower
loading increased it. A hydraulic surface loading of 65 m/h was suggested for design. It was not possible to determine
the turbulent intensity quantitatively due to the measurement method. The height and inclination of the shaft wall did
not seem to have a significant influence on the turbulent region. However, an increased height of the contact zone
enhanced the higher, plug flow region and a recommended height of 0.81 m or higher for the recommended hydraulic
surface loading was suggested when both mixing and plug-flow are desired. The separation zone was characterised by a
stratified flow structure, mainly influenced by the cross-flow velocity that is a function of the distance between the shaft
wall top and the water surface. A cross-flow velocity of 37 m/h or higher resulted in a clearly defined stratification,
believed to be crucial for an efficient separation of flocs. Finally, the extension of the lower, denser plug-flow region in
the separation zone increased when the shaft wall height increased. r 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Dissolved air flotation; Fluid dynamics; Flow profiles; Contact zone; Separation zone; Acoustical Doppler velocimeter

1. Introduction based on removing suspended solids from water by

Dissolved air flotation is a process used in the
treatment of water and wastewater. The technique is
*Corresponding author. Tel.: +46-31-7722133; fax: +46-31-
7722128.
E-mail address: mans.lundh@wet.chalmers.se
(M. Lundh).
utilising the lifting force induced when bubble-floc
aggregates of low density are produced.
The process takes place in two separated zones, each
with a different mechanism. The contact zone provides
for the contact and attachment of bubbles and floc
(aggregation) while the separation zone separates the
generated aggregates from the water phase. Previous
research has established that the contact zone config-

0043-1354/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 3 - 1 3 5 4 ( 0 1 ) 0 0 3 5 7 - 8
1586 M. Lundh et al. / Water Research 36 (2002) 1585–1595
uration could be of great importance for the removal
efficiency, although the lack of quantitative guidelines is
striking [1]. Some researchers have presented ideas about
the performance of the contact zone but few have
actually carried out a systematic study.
Kitchener and Gochin [21] suggested that enough
turbulence must exist in order to provide adequate
collision opportunities between bubbles and flocs but
that regions of high shear should be avoided due to the
risk of breaking the floc and detaching the bubbles. They
also suggested that the flow should be plug-like and that
the contact zone should be separated from the separa-
tion zone. Vrablik [2] emphasised the importance of a
sufficient residence time to achieve the required contact
time. Zabel [3] suggested that the air-enriched recycle
stream should be promptly released into the water
influent in order to minimise the loss of air due to
coalescence of bubbles.
Bratby [4] concluded that an inadequately designed
inlet arrangement caused significant amounts of larger
bubbles to break through the accumulated float layer.
Ettelt [5] tested six different contact zone configurations
and showed that small modifications had a significant
impact on the properties of the float layer. Vrablik [2]
concluded that the velocity of the recycle should be 2–3
times larger than the influent to ensure a sufficient
mixing.
Baeyens et al. [6] reported on an experimental set-up
of a pilot coaxial flotation column and presented
theories on flow conditions using the plug flow disper-
sion model with the P!eclet number. They concluded that
the pilot column behaved nearly as a plug-flow system
within a range of hydraulic surface loading of 5–30 m/h,
and that the injection of bubbles increased the P!eclet
number. Liers et al. [7] developed a model for a coaxial
DAF column and examined the parameters bubble-floc
attachment efficiency and air-to-solids volume ratio in
the aggregates. Shawwa and Smith [8] performed a pulse
tracer response test in the contact zone. Using a
dimensionless P!eclet number for determining the resi-
dence time distribution, the authors showed that the
flow conditions inside the contact zone deviated from
plug-flow. The mixing time decreased with increasing
hydraulic loading in the contact zone (from 30 to 90 m/
h). Park et al. [9] examined the solids removal efficiency
versus water velocity in a coaxial cylinder-type flotation
column. 90% efficiency was achieved for a contact zone
velocity of less than 66 m/h and for a separation zone
velocity of 18 m/h. Moreover, a flow structure was
modelled with the simple-implicit method for pressure-
linked equation.
Studies have shown that the air content influences the
flow structure mainly in the separation zone [10–12] and
to a lesser extent in the contact zone [10]. The studies
did, however, show that the jets, generated from adding
the dispersion, have an impact on the flow structure in
the contact zone. The flow structure in the separation
zone has been modelled [13–17] and measured experi-
mentally in pilot plants [10–12,15], in laboratory set-ups
[18] and in full-scale plants [16,17].
This paper will present the result and analysis of an
experimental study of the flow structure in a tradition-
ally designed rectangular dissolved air flotation (DAF)
pilot plant and the influence of varying the internal
geometry, meaning the variation in height, length and
inclination of the contact zone shaft wall. The flow
structure in the contact zone and in the separation zone
will be discussed and an attempt to quantify design
criteria will be presented.
2. How do hydraulic characteristics affect the
separation?
The study only covers the hydraulic principles and not
the removal of suspended solids. As discussed in the
introduction of this paper, an initial, efficient mixing of
the water and bubbles in the lower parts of the zone
probably enhances the aggregation process. The mixing
should not be too vigorous to avoid the coalescence of
bubbles or the disintegrations of flocs and bubble
aggregates.
One might expect that turbulence and strong motions
would affect the hydraulic behaviour of the separation
zone. Finally, the final vertical transport of the water
and its horizontal cross flow above the shaft into the
separation zone should probably be a distinct forward
and gentle movement with a low turbulence level.
3. Material and method
3.1. General
The purpose of the measurements was to map the
velocity profiles in the contact zone for different
configurations. The measurements were performed in a
pilot plant. An acoustic Doppler velocimeter [19] was
used for measuring the average water velocities in a grid
net, which would then give an insight into the flow
structure. The study was carried out with particle-free
water under the assumption that a solids loading up to
800 mg SS/l would not significantly influence the flow
structure [10,12].
3.2. Pilot tank
The pilot flotation tank (Fig. 1) was rectangular in
shape with a height of 1.35 m, a length of 1.7 m and a
width of 0.7 m. The influent descended behind a baffle
and entered the contact zone through a slot at the
bottom. In the contact zone, the water was brought
 M. Lundh et al. / Water Research 36 (2002) 1585–1595
Valve Sludge Sludge
handle scraper compartment
Level
weir
Water level
Influent Baffle
Shaft
wall
Separation
Contact zone
zone
Effluent
Distribution pipe Needle valve Perforated
for air-saturated outlet pipe
water
Fig. 1. The pilot plant flotation tank.
upward. The water flowed horizontally over the shaft
wall and entered the separation zone. Effluent was
removed at the bottom via two perforated pipes. The
tank arrangement was capable of operating at surface
loadings up to 30 m/h.
The recycle flow was saturated with air in an ejector
that brought water and air together. It was continuously
added to the main flow through three, evenly spaced
needle valves at the bottom of the contact zone. The
released air bubbles were thus mixed with the main flow.
The system could produce about 3 m3/h.
3.3. Acoustic Doppler velocimeter
The principle of the ADV is based on the Doppler
effect. An acoustical pulse that is scattered on a moving
object, experiences a frequency shift that is related to the
speed of the object. The ADV is constructed with a
central transmitter and three receivers located in a circle
around the transmitter (Fig. 2). The receivers are slanted
towards a measuring point approximately 5–6 cm from
the transmitter. The ADV makes it possible to determine
more or less continuously (25 samples per second) three
velocity components in the measuring point.
Lohrmann et al. [20] have reported that the instru-
ment is capable of measuring velocities down to
0.4 mm/s. The maximum was determined at 2.5 m/s.
Other measurements [10,11] showed, however, that in
water with a high concentration of micro-bubbles of
about 40–120 mm, the measured velocity was reduced up
to 70%, depending on the air content and probe
orientation. Furthermore, there were indications that
the measuring point was not located 5 cm from the
probe transmitter as stated by the manufacturer but only
1587
ADV Probe
Transmitter
Receiver
Measuring
point
5 cm
z
x
Fig. 2. The ADV.
about 1 cm. The measurements of this study were hence
assumed to be displaced 4 cm upwards, which is of little
significance to the overall flow structure considering the
size of the contact zone.
3.4. Operating conditions
The hydraulic surface loading over the whole tank
was set at 11 m/h. Nine different configurations regard-
ing height, length and inclination were measured for the
same operational settings, which are displayed in the
vector plots and defined in Table 1.
The required hydraulic surface loading in the contact
zone determined the configuration of the contact zone.
The contact zone was arranged to fit loadings of
approximately 40, 65 and 100 m/h, based on the surface
area above the contact zone. The pilot plant construc-
tion had some limitations and the actual operational
variables are presented in Table 1.
The shaft wall of the pilot plant consisted of two
wallboards giving the possibility to study different
configurations. The dispersion rate was set at 10%,
producing an air content of about 4.5–5.5 ml air/l water
in the contact zone. The saturator pressure was set at
5 bar. As the air content did not seem to have any
immediate influence on the flow structure in the contact
zone it was not analysed specifically. The air content was
taken from Lundh [10] based on the performance of the
saturator system.
The turbulence has not been analysed in this study as
such an analysis is based on the standard deviation of
the velocity signals, which in turn requires adequate
1588 M. Lundh et al. / Water Research 36 (2002) 1585–1595

Table 1
General variables in the DAF tank

Measurement Incoming Recycle Hydraulic Air content Retention Retention time Description
flow rate surface loading in c-zone time in tank in s-zone of c-zone
m3/h % m/h ml/l min min

M4c 9.7 10.3 10.7 5.5 8.0 7.0 Half tilted/

half vertical
M48 10 10 11 4.5 7.8 7.3 Vertical
M49 10 10 11 4.5 7.8 7.1 Vertical
M50 10 10 11 4.5 7.8 6.9 Vertical
M51 10 10 11 4.5 7.8 6.8 Vertical
M53 10 10 11 4.5 7.8 7.0 Vertical
M54 10 10 11 4.5 7.8 6.4 Vertical
M55 10 10 11 4.5 7.8 7.3 Vertical
M56 10 10 11 4.5 7.8 7.1 Tilted

Measurement Contact zone Contact zone Shaft wall Hydraulic Retention Distance Cross flow
length height inclinationa surface loading time in c-zone between water velocity
in c-zoneb level and top
of shaft wallf
cm cm 1 M/h s cm m/h
c d
M4 10/25 89 171 61 57 35 44
M48 24 51 F 65 28 73 22
M49 24 81 F 65 45 43 37
M50 24 101 F 65 56 23 68
M51 24 111 F 65 61 13 121
M53 24 91 F 65 50 33 48
M54 39 91 F 40 82 33 48
M55 16 91 F 98 33 33 48
M56 16/39d 88 151 98/40e 43 36 44
a
Inclination in relation to a vertical line.
b
Calculated for the vertical part of the contact zone.
c
Ejector 1.
d
Length from the baffle to the start of inclination at bottom/length from the baffle to the end of inclination.
e
Loading at the start of inclination at tank bottom/loading at the end of inclination.
f
Water surface at 124 cm above tank bottom.

measurements of the magnitude of the velocity, which is
not possible with the ADV in bubbly water [10].
The water surface was located about 124 cm from the
tank bottom. The hydraulic surface loading of the
flotation tank was calculated as the incoming flow
divided by the tank surface area, which was about 1 m2
excluding the area behind the baffle. In reality it was
measured to be about 1.07 m2, which would reduce the
surface loading of Table 1 by approximately 6%. The
difference was considered insignificant as the parameter
works only as a label and is irrelevant to the flow
structure inside the tank.
The retention time in the tank was calculated as the
water volume divided by the hydraulic loading of the
tank, which is the sum of influent and recycle flow. The
retention time in the contact zone (c-zone) was
calculated as the volume, defined by the baffle and shaft
wall, divided with the hydraulic loading in the tank. The
retention time in the separation zone (s-zone) was
calculated as the difference between the retention time
in the contact zone and the retention time of the whole
tank. The cross-flow velocity was calculated as the
hydraulic loading divided by the area defined by the
water level, the top of the shaft wall and the tank width.
3.5. Method of measurement and the references of
direction
As illustrated in Fig. 3, the tank was defined in an xyz-
coordinate system to give reference points as to where
the measuring point was located. Co-ordinate z gives the
height over the tank bottom, x gives the length starting
from the influent side, and y gives the width, starting
from the right side going to the left. The velocities also
follow the coordinate system, meaning that the velocity
vx refers to the x-direction, etc.
Since it was difficult to determine an actual magnitude
of the water velocity with the ADV in bubbly water,
 M. Lundh et al. / Water Research 36 (2002) 1585–1595
Influent
longitudinal extension
c-z s-z
leftward
lateral
extension backward forward
y rightward
x x
Horizontal cross-section
Fig. 3. The co-ordinate system of the measurements and the directions referred to in the study.
emphasis has been placed on the flow structure rather
than on the exact value of the velocity. The direction
was, however, believed to be correct [10]. The conse-
quence of this approach is that the scale of the vectors is
the same in a plot but could differ between the plots.
Each velocity measurement took approx. 80–120 s
with a sample frequency of 25 Hz. After processing the
data for different parameters, the average velocities were
plotted on a vector plot or on a surface plot showing the
overall flow situation. At the bottom of each vector plot
there is a reference vector defining, by length, the
maximum velocity shown in the plot.
The concept of three-dimensional flows is discussed in
the analysis. The basis is the two-dimensional, vertical,
longitudinal section through the tank. Three-dimen-
sional flows move in the y-direction according to the
applied coordinate system.
4. Flow structures in the contact zone
4.1. General
The evaluation of the mixing properties of the zone
has been restricted to the visual observation of the
velocity profiles. The zone was considered mixed if the
pattern indicated strong turbulence and recirculation of
water. It was not possible to quantify the mixing
intensity by e.g. measurements of the turbulence. The
concept of a plug flow was based on evenly distributed
velocities across the section and directed perpendicular
to the cross section.
4.2. Height
The length of the contact zone was held constant at
24 cm giving a hydraulic surface loading of 65 m/h for all
1589
c-z s-z Effluent
upward
backward forward
downward
z
Vertical cross-section
the cases, while the height of the shaft wall was increased
from 51 to 81, 91, 101 and 111 cm. Other variables are
presented in Table 1.
There was a well-mixed zone in the lower parts of the
contact zone in all cases, more or less independent of the
height of the shaft wall (Fig. 4). The region was located
in the back of the zone. Three-dimensional movements,
defined as horizontal movements in the plane (Fig. 5,
right), also contributed to mixing of the water. As the
height of the shaft was increased (Fig. 4, left to right) the
mixing zone seems to have been relocated higher up. The
mixing does not seem to have been restricted to the
lower region of the zone as vertical water transports
from the top of the zone also occurred. A large part of
the water transport through the contact zone, however,
seems to have occurred up along the shaft wall and
directly out into the separation zone. This is shown as
large velocity vectors along the shaft wall in Fig. 4 or as
an elevated ridge close to and across the shaft wall in the
surface plot of Fig. 5, left.
A more plug-like flow seems to evolve when the height
of the contact zone is increased. The case with a 111 cm
shaft (Fig. 4, M51), giving the retention time of 61 s,
generated a more or less evenly distributed velocity
profile in almost 50% of the zone. If the requirements
were an initial mixing phase with a subsequent homo-
genised flow, a high shaft would be better than a lower
one.
A distinct unidirectional cross flow was generated for
an induced cross-flow velocity of 48 m/h or higher.
Lower velocities generated more complex and uncon-
trolled flow patterns above the contact zone.
4.3. Length
The examined lengths of the contact zone were 16, 24
and 39 cm, resulting in hydraulic surface loadings of 40–
1590 M. Lundh et al. / Water Research 36 (2002) 1585–1595
Fig. 4. The velocity profiles in the central longitudinal and vertical section of the contact zone for an increasing shaft wall height. The
heights are from left to right: 51, 81, 91, 101 and 111 cm.
98 m/h in the contact zone, while the height was held
constant at 91 cm above the tank bottom. The retention
time varied from 33 to 82 s depending on the zone
length.
A mixing zone was observed in the case with the 39-
cm long contact zone (Fig. 6, M54). Water from above
did not seem to mix with the lower parts. There was a
higher region where the water was moved horizontally
but firmly upwards, developing into a relatively homo-
genised flow. The cross flow was horizontally orientated,
indicating a steady and stable flow structure.
There was also a mixing zone in the case of the 24-cm
long contact zone (Fig. 6, M53) with a hydraulic surface
loading of 65 m/h. At 50 cm above the tank bottom, the
flow was uniform and of a plug-flow nature. The
velocities were directed upwards and forwards above
the contact zone.
In the case with the narrow shaft of 16 cm (Fig. 6,
M55), there seemed to be less mixing. The velocity
profile was concentrated distinctly upwards and the
water transport above the contact zone was concen-
trated into a well-defined flow, entering the separation
zone horizontally.
The flow structure above the contact zone was
horizontally orientated for the cases 16 and 39 cm but
also vertically for the 24-cm case.
4.4. Inclination
The first of two designs was a contact zone where the
entire wall was inclined 151 from the vertical. The height
of the wall was 88 cm in total and the hydraulic surface
loading was 98 m/h at the bottom and 40 m/h at the top
of the contact zone. The second inclined design was a
contact zone with the lower wallboard inclined 171 from
the perpendicular to the tank bottom, reaching about
48 cm above the bottom. On top of the inclined board
was a vertical board with a length of 40 cm. The
hydraulic surface loading was 98 m/h at the bottom and
65 m/h in the vertical part of the shaft. The retention
time varied between 33 and 57 s depending on the
design.
The half-inclined, half-vertical shaft wall generated a
strong flow of water upwards along the shaft wall
(Fig. 7, M4), like in the case when the entire wall was
inclined (Fig. 7, M56). There was a downward move-
ment along the baffle in the lower, back regions,
identical for both cases, although more distinct in the
case with the entire shaft wall inclined.
A joint hinge arrangement connected the two wall-
boards constituting the shaft wall, with a possible
influence on the flow structure in the contact zone. This
impact was, however, considered to be small, not
affecting the overall flow characteristics in the contact
zone [10].
Above the contact zone, the half-inclined, half-vertical
case (Fig. 7, M4) resulted in a more distinct forward-
moving flow into the separation zone, encompassing the
whole section part above the contact zone. This was
probably due to the higher hydraulic surface loading at
the top of the zone.
The inclination itself did not seem to significantly alter
the flow structure when compared to the flow structure
found in the vertical shafts.
5. Flow structures in the separation zone
The outflow from the contact zone for the 51-cm high
shaft was not distinct and regular (Fig. 8, M48). There
 M. Lundh et al. / Water Research 36 (2002) 1585–1595 1591

Fig. 5. The water velocities in horizontal sections of a 91-cm high and vertical contact zone. The figures to the left are the vertical
velocities (Vz ) and the vector plots to the right are the horizontal velocities (Vx;y ) in the corresponding surface plot. The z-value is the
level above tank bottom.

was a flow transporting the water from the contact zone
towards the outlet wall and slightly upwards. There were
three large high-velocity arrows just above the shaft
wall, indicating a strong return flow from the separation
zone into the contact zone. The water transport has
become more plug-like just above the bottom of the
separation zone.
For the case with a vertical shaft wall of 81 cm (Fig. 8,
M49) the flow pattern, compared to (Fig. 8, M48), seems
to have been brought up closer to the water surface,
forming a stratified profile from the contact zone to the
outlet wall. There was an indication of a relatively
strong return flow from the outlet wall towards the shaft
wall, which seems to have entered into the contact zone
just above the shaft wall. Underneath the return flow, at
about 70 cm above the tank bottom, the transport of
water has developed into plug-flow.
The flow patterns for the next two cases, with the
shaft wall heights 91 and 101 cm, respectively (Fig. 8,
M53 and M50), were similar to the previous one. The
last case (Fig. 8, M51) did, however, show significant
changes of flow structure. The upper flow, just under-
neath the water surface, was now located higher up in
the tank, at z ¼ 105 cm. There was no observable return
1592 M. Lundh et al. / Water Research 36 (2002) 1585–1595
Fig. 6. A comparison between contact zones with varying lengths. Velocity profiles in the central longitudinal and vertical section in
the contact zone. From left to right the lengths are 16, 24 and 39 cm, corresponding to hydraulic surface loads of 98, 65 and 40 m/h over
the contact zone.
Fig. 7. The velocity profile in the central, longitudinal and
vertical section in a tilted contact zone. The plot to the left is the
contact zone with the tilted shaft wall and the plot to the right
(less measurements done) is the contact zone with the half-tilted
and half-vertical shaft wall.
flow and the water flow underneath the horizontal flow
developed immediately into a plug-flow.
The plot with a narrow shaft with a length of 16 cm
(Fig. 8, M55) indicates an outflow that makes a small
dip just after leaving the contact zone before continuing
towards the outlet wall. After encountering the outlet
wall, the water has returned towards the shaft wall,
generating a return flow. A longer contact zone of 39 cm
(Fig. 8, M54) generated less return flow and a more
concentrated flow underneath the water surface, other-
wise the flow resembles the case with a shaft wall height
of 91 cm and a length 24 cm (Fig. 8, M53).
The cases with inclined wall (Fig. 8, M4 and M56) had
profiles very similar to the case with a 91 cm shaft wall
and a contact zone length of 24 cm.
6. Discussion and conclusion
6.1. Contact zone
The general flow structure in a vertical shaft was
characterised as follows:
The water enters the contact zone through the inlet
slot, follows the tank bottom towards the shaft wall
where it is forced upwards. The vertical velocity
profile along a longitudinal extension has a max-
imum at the shaft wall and a minimum at the baffle.
The flow is transformed higher up where a more
plug-like pattern is developed. Concentrated partial
flows are traced as local elevations in a horizontal
velocity profile. Horizontal movements transport the
water towards the back of the zone, where the water
is further transported down. The velocities above the
zone are low but distinctly directed upwards and
forwards into the separation zone.
Mixing in the contact zone was found in the lower
regions in all the cases, except maybe when using a
 M. Lundh et al. / Water Research 36 (2002) 1585–1595 1593

Fig. 8. Water velocity profiles in the central, longitudinal and vertical section of the separation zone for the increase of shaft wall
height, the shaft length and the inclination of the shaft wall. The positions of the shaft wall, baffle and water level are approximated.

narrow, vertical shaft. If good mixing is required the
hydraulic surface loading should be less than 98 m/h.
A plug-like flow seems to have been enhanced with
increasing height of the contact zone. The case with a
111-cm shaft seems to have generated an evenly
distributed velocity profile in almost 50% of the zone.
For the requirements where an initial mixing phase with
a subsequent homogenised flow, a high shaft would be
preferred to a lower one, meaning a shaft higher than
81 cm giving a retention time of at least 45 s at a
hydraulic surface loading of 65 m/h.
A cross-flow velocity higher than 37 m/h generated a
distinct flow pattern into the separation zone. If the
distance from the water surface to the top of the shaft
wall was increased, the flow structure became irregular,
possibly inducing a risk for accumulation of bubbles just
above the contact zone thus enhancing the probability
for return flows down into the contact zone. It is thus
important to have a firm movement forwards from
above the contact zone into the separation zone.
The inclination of the shaft wall that related to the
range of hydraulic surface loading (40–98 m/h), did not
1594 M. Lundh et al. / Water Research 36 (2002) 1585–1595
Table 2
Design criteria of the contact zone based on the flow structure
in a 1 : 2 rectangular flotation tanka. The value in brackets is a
recommended case
Hydraulic Shaft wall Cross flow Retention
surface loading height velocity time
m/h cm m/h s
40–98 (65) >81 (91) >37 (48) >33 (45)
a
Width : length=1 : 2.
seem to significantly change the flow structure from the
other cases, which implies that the shape of the contact
zone seems of less importance, at least regarding the
inclination of the shaft wall.
The design criteria are summarised in Table 2.
6.2. Separation zone
It is concluded that the investigated contact zone
geometry did not influence the fundamental flow
structure of the separation zone, except for the shaft
wall height. The flow structure was found to be a matter
of a limiting distance between the shaft wall top and the
drainage pipe, and between the top of the shaft wall and
the water surface. If the wall was too low, resulting in a
too large distance between the water surface and the top
edge of the shaft wall, the flow structure would not
develop into a distinct stratified flow. Hence, the cross-
flow velocity seems to have a major significance in the
development of the stratified flow structure. A cross-
flow velocity of 37 m/h or higher is suggested.
Plug-flow started to develop just underneath (about
10 cm) the top of the shaft for a stratified flow. It would
mean that a large distance between the shaft wall top
and the tank bottom, meaning a higher shaft wall would
generate a plug-flow in a larger part of the separation
zone than a lower shaft would do. The design criteria are
given in Table 2.
Acknowledgements
We gratefully acknowledge the support of PURAC
AB, Lund, Sweden, in partly financing the project and
providing the pilot plant. The University of Lund is
acknowledged for providing personnel and laboratory
facilities.
References
[1] Haarhoff J, van Vuuren L. A South African Guide for
Dissolved Air Flotation. Water Research Commission,
P.O. Box 824. Pretoria. 001. Republic of South Africa,
1993.
[2] Vrablik ER. Fundamental Principles of Dissolved-Air
Flotation of Industrial Wastes. In: Proceedings of the
14th Industrial Waste Conference, 1959, Purdue Univer-
sity: IWC, 1959. p. 743–79.
[3] Zabel TF. The advantages of dissolved-air flotation for
water treatment. J Am Water Works Assoc 1985;77(5):
42–6.
[4] Bratby J. Aspects of sludge thickening by dissolved-air
flotation. Water Pollut Control 1978;77(3):421–32.
[5] Ettelt GA. Activated Sludge Thickening by Dissolved
Air Flotation. In: Proceedings of the 19th Industrial
Waste Conference, Purdue University, IWC, 117, 1964.
p. 210–44.
[6] Baeyens J, Mochtar IY, Liers S, De Wit H. Plug-flow
dissolved air flotation. Water Environ Res 1995;67(7):
1027.
[7] Liers S, Baeyens J, Mochtar I. Modelling dissolved air
flotation. Water Environ Res 1996;68:1061.
[8] Shawwa AR, Smith DW. Hydrodynamic characterisation
in dissolved air flotation (DAF) contact zone. Water Sci
Technol 1998;38(6):245–52.
[9] Park K, Cho J-M, Oh J, Chung K. Experimental and
modelling evaluation of upward and downward velocities
in the coaxial flotation column. Water Sci Technol 2001;
43(8):195–201.
[10] Lundh M. Flow Structures in dissolved air flotationFex-
perimental study in a pilot plant. Licentiate thesis,
Department of Water and Environmental Eng., Lund
Institute of Technology, Lund, Sweden, 2000.
[11] .
Lundh M, Jonsson L, Dahlquist J. Experimental studies of
the fluid dynamics in the separation zone in dissolved air
flotation. Water Res 2000;34(1):21–30.
[12] .
Lundh M, Jonsson L, Dahlquist J. The flow structure in
the separation zone of a DAF pilot plant and the relation
to bubble concentration. Water Sci Technol 2001;43(8):1
85–94.
[13] Crossley IA, Rokjer DM, Kim J. Optimising the DAF
Process Utilizing Two Phase 3D CFD Modelling. In:
AWWA Annual Conference Proceedings, Chicago, IL,
1999.
[14] Fawcett NSJ. The hydraulics of flotation tanks; computa-
tional modelling. Dissolved Air Flotation, International
Conference April 1997, London. The Chartered Institution
of Water and Environmental Management. CIWEM
Services Ltd for CIWEM. ISBN No: 1 870752 30 9,
1997. p. 51–72.
[15] Hague J, Ta CT, Biggs MJ, Sattary JA. Small scale model
for CFD validation in DAF application. Water Sci
Technol 2001;43(8):167–74.
[16] O’Neill S, Yeung H, Oddie G. Physical modelling of the
dissolved air flotation process. Dissolved Air Flotation,
International Conference April 1997, London. The char-
tered Institution of Water and Environmental Manage-
ment. CIWEM Services Ltd for CIWEM. ISBN No:
1 870752 30 9, 1997. p. 76–86.
[17] Ta CT, Beckley J, Eades A. A multiphase CFD model
of DAF process. Water Sci Technol 2001;43(8):
145–52.
 M. Lundh et al. / Water Research 36 (2002) 1585–1595 1595

[18] Leppinen DM, Dalziel SB, Linden PF. The hydraulics of a
dissolved air flotation tank. In: Proceedings of the 4th
International IWA Conference on DAF in Water and
Wastewater Treatment, Helsinki, Finland, 11–14 Septem-
ber 2000, 2001.
[19] Nortek AS. ADV Operational Manual, January 2, Nortek
AS Inc., 1996.
[20] Lohrmann A, Cabrera R, Kraus NC. Acoustic-Doppler
velocimeter (ADV) for laboratory use. In: Proceedings of
Fundamental and Advancements in Hydraulic Measure-
ments and Experimentation. ASCE, 1994.
[21] Kitchener JA, Gochin RJ. The mechanism of dissolved air
flotation for potable water: basic analysis and a proposal.
Water Res 1981;15(5):585–90.