Aerated Grit Chambers Hydraulic Design Equation

Jerzy M. Sawicki1

Abstract: The paper is devoted to the problem of dimensioning of aerated grit chambers. Existing hitherto methods of design of such
Downloaded from ascelibrary.org by SOUTHERN CALIFORNIA UNIVERSITY on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

objects are based on the very large experience of operation but they do not concern the important relation between the intensity of aeration
Q p and the discharge of transverse circulation Q c . For elimination of that fault four concepts 共initially suggested for air-lift pumps兲 of
explanation of the process were analyzed. Three of them were empirically verified and in effect the relation Q c (Q p ) was proposed. In the
next succession a simplified model of transverse circulation was suggested. Results of calculations led to the conclusion that the resultant
bottom velocity of the properly designed aerated grit chamber should not exceed the nonsilting velocity for those fractions which would
be removed in the chamber. Two practical examples of existent aerated grit chambers 共situated in Gdansk and Gdynia兲 were presented
below.
DOI: 10.1061/共ASCE兲0733-9372共2004兲130:9共1050兲
CE Database subject headings: Hydraulic design; Waste disposal; Sedimentation; Water circulation; Bubbles.

Introduction • Intensive transverse circulation can mix added chemicals with

The aerated grit chamber is a specific kind of settling tank, serv-
ing for the selective separation of sand, ash, cullet, seeds, and
other granular suspended particles 共commonly called ‘‘grit’’兲.
Such a device is equipped with an aeration system. Air bubbles,
introduced along one side of the chamber, usually near its bottom,
induce a transverse circulation, which combines with the longitu-
dinal motion of the waste water and forms a characteristic spiral
flow. The aeration system can be replaced by a set of hydraulic
nozzles. In such a case the transverse motion is induced by water-
jets 关see, e.g., Brenner and Diskin 共1991兲兴. However, this kind of
solution is rather seldomly applied. The intensity of circulation
should be high enough to keep the lighter, mainly organic, par-
ticles in suspension, but it should not disturb the sedimentation of
the heavier grit elements.
Practical experience 共e.g., Albrecht 1967; ASCE 1992兲 shows
that aerated grit chambers provide some important advantages in
comparison with conventional solutions. The most important ben-
efits are
• Maintaining of the same efficiency of removal of grit for vari-
able sewage discharge.
• Relatively low content of organic matter in the removed
sludge.
• Preliminary aeration 共‘‘refreshing’’兲 of sewage.
• Low head loss 共much less than in conventional grit chambers,
where troublesome linear weirs are necessary for maintaining
of constant velocity兲.
1
Professor, Technical Univ., Faculty of Hydro- and Environmental
Engineering, ul. G. Narutowicza 11/12, 80-952 Gdańsk, Poland.
Note. Associate Editor: A. Bruce De Vantier. Discussion open until
February 1, 2005. Separate discussions must be submitted for individual
papers. To extend the closing date by one month, a written request must
be filed with the ASCE Managing Editor. The manuscript for this paper
was submitted for review and possible publication on April 2, 2001;
approved on June 10, 2003. This paper is part of the Journal of Envi-
ronmental Engineering, Vol. 130, No. 9, September 1, 2004. ©ASCE,
ISSN 0733-9372/2004/9-1050–1058/$18.00.
wastewater. However, one can enumerate also the following
faults of aerated grit chambers.
• High capital costs 共aeration system兲.
• High operating costs 共growth of consumption of energy and
additional servicing personnel兲.
• Some volatile organic components and odors may be released
from the wastewater.
There are many pieces of technical information which enable
dimensioning of aerated grit chambers. These recommendations
usually have the form of design guidelines or simple physical and
mathematical expressions. However, among the said instructions
there are no reliable relations which express the functional depen-
dence between the discharge of the air Q p , introduced into the
chamber, and the intensity of transverse circulation Q c , which is
a very important disadvantage of existing methods of design. An
empirical relation between the induced transverse fluid flow rate
and the bottom velocity can be found in Brenner and Diskin
共1991兲. However, this expression was derived for the circulation
induced by water-jets and hence cannot be applied for typical
aerated grit chambers.
The lack of the discussed relation is underlined in technical
handbooks 共e.g., ASCE 1992兲 because it creates an important
problem. If someone wants to design a new chamber or to adjust
an existing installation, one must be able to determine the change
of the velocity field, due to the change of the air discharge.
The main goal of this paper is to find a mathematical equation
which describes the qualitative relation between Q p and Q c .
Computational Scheme of the Aerated Grit Chamber
Generally, the perimeter of each chamber can have the shape of a
polygon 共Fig. 1, solid line兲 or can be an oval 共Fig. 1, broken line兲.
The smooth, curvilinear shape of this perimeter was justified by
German investigators 共Pöpel and Hartman 1958兲 to minimize hy-
draulic disturbances of the wastewater velocity field. However,
construction of such oval grit chambers is more difficult and ex-
pensive. Besides, the oval cross section creates some difficulties
also during the chamber design, as its depth H and width B are
changeable parameters: H⫽H(y) and B⫽B(z).

1050 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / SEPTEMBER 2004

J. Environ. Eng. 2004.130:1050-1058.


Downloaded from ascelibrary.org by SOUTHERN CALIFORNIA UNIVERSITY on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.
Fig. 1. Characteristic dimensions of aerated grit chamber cross
section
In this paper it was assumed that the computational shape of
the aerated grit chamber is determined by a schematic polygon
共pentagon兲, shown in Fig. 1. Eventual ‘‘rounding’’ of this shape
can be done individually by the designer. In the other case, the
technical version of the method proposed below could not be
applied and one should make use of its full, i.e., differential,
variant 关see Eqs. 共32兲 and 共34兲, or even the more elaborated sug-
gestion, e.g., Olsen and Skoglund 共1995兲兴.
The aerated grit chamber is fully determined by the following
set of design parameters 共Fig. 1兲: length L, width B, depth H,
depth of the gutter h, volume V, cross section S, distance of the
bottom line from the left wall B L and from the right wall B R
(B L ⫹B R ⫽B), submergence of the aeration system H A , width of
the air-lift zone b B , height of the vertical baffle H B , height of the
gap above (H G ) and under the baffle (H D ), effective detention
time T R , discharge of wastewater Q 共along the chamber兲, dis-
charge of the transverse circulation Q C , air supply Q P 共or Q P1
when related to the unit length兲, mean longitudinal velocity v,
free-sedimentation velocity for critical particles of suspension v si
(i⫽1, . . . ,I c ).
A very important element of the aerated grit chamber is the
vertical baffle. According to practical recommendations 关see, e.g.,
Albrecht 共1967兲; ASCE 共1992兲兴, such a baffle improves the func-
tioning of the chamber so much that it should appear in all de-
signs. However, sewage treatment plants are very often equipped
with conventional aerated grit chambers without the baffle. Such
conventional solutions will be also considered in this paper.
Required Efficiency of Aerated Grit Chambers
Surprisingly, in the available design literature one cannot find
many precisely formulated requirements related to the expected
efficiency of grit removal which should be achieved. Hence it is
reasonable to accept the classical concept 共Imhoff and Imhoff
1979兲, according to which the properly functioning aerated grit
chamber should remove
• 100% of the coarse sand fraction (d⬎0.2 mm); and
• 65–75% of the medium sand fraction (0.1⬍d⬍0.2 mm).
JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / SEPTEMBER 2004 / 1051
J. Environ. Eng. 2004.130:1050-1058.
Moreover, the organic matter content in the removed grit should
not exceed 10% 共percentage by weight兲.
According to the remarks stated above, it is logical to intro-
duce two critical free-sedimentation velocities: v s1 for particles of
diameter d 1 ⫽0.1 mm, and v s2 for d 2 ⫽0.2 mm. For typical grit
density ␳ p ⫽2,600 kg/m3 共Imhoff and Imhoff 1979兲 we have
v s1 ⫽0.007 m/s, v s2 ⫽0.023 m/s (1)
Methodology of Aerated Grit Chamber Design
The aerated grit chamber can be dimensioned with the help of
exact methods or by means of compact technical procedures. The
methods of the first kind are based on equations of mathematical
physics 共differential ones as a rule兲, which describe the inner
structure of physical fields characteristic of the problem consid-
ered. The typical example of such an exact method, applied to
conventional desanders, can be found in an International Associa-
tion for Hydraulic Research journal paper 共Olsen and Skoglund
1995兲. However, procedures of this kind are very time-consuming
共that means expensive兲, so they should be used rather when com-
plex objects are designed that justify the additional cost of labor
and equipment. The simplified variant of an exact method will be
presented below.
The technical methods in turn are based on synthetic balance
equations, combined with practical directions which express the
experience of engineers. We have the following geometrical and
hydraulic formulas for aerated grit chambers:
S⫽BH⫹Bh/2
关 or equivalent relation S⫽S 共 B,H 兲 for more
complex cross sections兴 (2)
V⫽LS (3)
v ⫽Q/S (4)
T R ⫽V/Q (5)
Technical directions, complementary to Eqs. 共2兲–共5兲, can be pre-
sented in the following juxtaposition 共Neighbor 1965; Imhoff and
Imhoff 1979; ‘‘Design’’ 1992兲:
• Depth H⫽2.0– 5.0 m;
• Width B⫽(1 – 5)H 共typical: B⫽2H; formally one can assume
that B/H⬎1, but when B/HⰇ1 the conditions of the transver-
sal circulation worsen and this possibility is not suggested兲;
• Length L⫽(2.5– 5.0)B;
• Minimal detention time T R ⫽120– 300 s 共typical: 180 s兲;
• Air supply Q p1 ⫽0.0045– 0.0125 m3 /sm 共typical: 0.0075
m3/sm, higher intensities for broader and deeper chambers兲;
• Distance of aerators from the bottom H⫺H A ⫽0.6– 1.0 m;
• Width of the air-lift zone b B ⫽1.0 m;
• Circumferential velocity 共about 150 mm under the waste water
free-surface兲
vB⫽0.60– 0.75 m/s
• Mean velocity v ⫽0.20– 0.30 m/s.
The number of relations and directions listed above is less than
the number of unknown parameters of the desander that creates
the necessary conditions of the elastic object design. However, as
was already mentioned, the design method outlined above does
not contain any relation between two crucial parameters: Q p and
Q c . Such a relation will be derived in the next section.

Downloaded from ascelibrary.org by SOUTHERN CALIFORNIA UNIVERSITY on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.
Fig. 2. Scheme of 共a兲 transverse circulation and 共b兲 location of
measurement points
Connection between Discharge of Air and Intensity
of Circulation
Models Considered
The motion of the liquid, induced by the gas bubbles, is a very
important technical factor, not only for aerated grit chambers, but
also for some other objects, as, for example, air-lift pumps or
pneumatic baffles.
In the bibliography there are four concepts explaining the
mechanism of transfer of energy from bubbles into the liquid:
• ‘‘Hydrostatic’’ 共‘‘Pumpen’’ 1984兲;
• ‘‘Empirical’’ 共Hussain and Spedding 1976兲;
• ‘‘Energetic’’ 共Sawicki and Pawłowska 1999兲;
• ‘‘Dynamic’’ 共Sawicki and Pawlowska 1999兲.
Each concept describes in a different way the power output N D
which is passed from the air bubbles to the liquid. It is necessary
to underline that those concepts were elaborated for air-lift
pumps, hence they must be adapted for conditions of aerated grit
chambers.
‘‘Hydrostatic’’ Model
The basis of this method is formed by the assumption that the
column of height H A 共and of the volume of the air-lift zone V A ),
containing the gas–liquid mixture, can be considered as a homo-
geneous substance of mean density ␳ z 共Fig. 2兲
␳ z ⫽ 关 ␳ 共 V A ⫺V P 兲 ⫹␳ A V P 兴 /V A
The total volume of the air bubbles V P , suspended in water at
each moment of time, can be calculated from the equality
V P ⫽Q P T P ⫽Q P H A / v P
where T P ⫽time of air-bubble upflow, and v P ⫽characteristic ve-
locity of the bubble rising which can be calculated, for example,
from the classical relation 共Soo 1966兲
v P ⫽0.687冑gd p
where d p ⫽effective diameter of a bubble
d p ⫽1.49关 ␳Q 2P /g 共 ␳⫺␳ A 兲兴 0.2
Provided that the air bubbles are subject to the isothermal
expansion from the initial pressure p d ⫽p atm⫹␳ z gH A , up to the
terminal pressure p g ⫽p atm , one can obtain the following expres-
sion for the power input, according to the ‘‘hydrostatic’’ concept
(␩ A ⫽efficiency of transfer of energy兲:
N DH ⫽␩ A Q P p atm ln关 1⫹␳ z gH A /p atm兴
1052 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / SEPTEMBER 2004
J. Environ. Eng. 2004.130:1050-1058.
‘‘Experimental’’ Model
The physical analysis of the considered problem, jointly with re-
markable experimental support, presented in Hussain and Sped-
ding 共1976兲, led to the following relation 共definitions of symbols
are given in ‘‘Notation’’兲:
K 2 V A2.75␳ A1.5g K 1V A
N DX ⫽␩ A , m⫽ (11)
p ␳ 共 exp m⫺1 兲
H A1.5Q 1.5 ␳H A Q p
Two coefficients K 1 and K 2 are of empirical character. In the
original paper it was stated that
K 1 ⫽1,050 kg/m2 s, K 2 ⫽133,313 kg/m1.25 s2.5 (12)
However, these parameters were determined for air-lift pumps, so
it seems impossible to adapt this method for grit chambers ana-
lyzed in this paper.
‘‘Energetic’’ Model
The startpoint of this concept was a statement 共Sawicki and
Pawłowska 1999兲 that the terminal velocity of the single air
bubble rising in a perfect fluid ( v pi ) is higher than the same
parameter for the viscous medium ( v pr ⬍ v pi ). It means that the
real terminal energy of a bubble E r differs from its perfect value
E i , and the difference (E i ⫺E r ), multiplied by the number of
bubbles, equals the ‘‘gas–liquid’’ energy transfer. Making use of
this concept one obtains the evident formula
N DE ⫽␩ A Q p ␳ A gH A /␤ (13)
where ␤⫽coefficient of virtual mass; for the spherical bubble ␤
⫽0.5 共Wijngaarden 1976兲.
‘‘Dynamic’’ Method
According to the idea used in this concept, the energy delivered to
the liquid by the air bubbles is equal to the local work performed
by each bubble against the drag force. That kind of explanation
leads to the relation
N DD ⫽3␩ A ␲C D ␳H A v 2p Q P /4d p (14)
In the broad range of dimensions of bubbles we can consider the
drag coefficient C D ⫽0.44; parameters v p and d p , see Eqs. 共8兲
and 共9兲 共Soo 1966兲.
(6)
Intensity of Dissipation of Energy
The relation Q C ⫽Q C (Q P ), so important for the proper designing
(7) of aerated grit chambers, may be obtained from the evident equal-
ity between the input of power 共i.e., the power, delivered by the
air bubbles兲 and the consumption of power
Power InputÄPower Output¿Dissipation
(8)
N D ⫽N U ⫹N V (15)
The function N U describes the output of power and must be
(9) expressed by the different relations for the different systems. For
air-lift pumps we have 共Hussain and Spedding 1976兲
NU air-lift pumps⫽␳gQ W ⌬H (16)
where Q W ⫽pump discharge; and ⌬H⫽elevation head. The dis-
sipation power N V in this case is usually neglected, although in a
more precise attitude it should be taken into account 共Sawicki and
(10) Pawłowska 1999兲.
 r⫽B/H (21)
we can obtain the following expression describing the power of
dissipation of energy 共for the aerated grit chamber兲:
N V ⫽n V V⫽2.56␳Q 3c 共 1⫹r 兲 2 共 1⫹r 2 兲共 r 4 ⫹3r 2 ⫹1 兲 / 共 L 2 H 2 r 5 兲
(22)

Experimental Verification of the Function N D (Chamber

with Baffle)
Downloaded from ascelibrary.org by SOUTHERN CALIFORNIA UNIVERSITY on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

For the purpose of selection of the best model of the process of

Fig. 3. Assumed simplified scheme of transverse velocity field
For aerated grit chambers the elevation head is practically
equal to zero 共the waste water free-surface is strongly foamed by
the outflowing air bubbles, but its mean level across the chamber
is almost horizontal兲, so N U ⫽0 in this case. The main consump-
tion of power is connected with the dissipation of energy, due to
the transverse circulation. According to the entropy balance equa-
tion 共Landau and Lifshitz 1987兲, the intensity of the viscous dis-
sipation 共related to the unit volume of the chamber兲 can be ex-
pressed as follows:
the delivery of power 共function N D ), the parameters of transverse
circulation were measured in the ‘‘old’’ aerated grit chamber in
Gdansk-‘‘East’’ sewage treatment plant, for Q⫽0. The shape of
this desander cross section is identical with the scheme in Fig. 1
for B⫽2.20 m, B L ⫽B R ⫽1.10 m, b B ⫽0.60 m, H⫽H A ⫽1.55 m,
h⫽1.10 m, H B ⫽1.00 m, H G ⫽0.27 m, H D ⫽0.35 m, and L
⫽21.0 m.
Three various aeration systems 共adequate to the existing tech-
nical possibilities兲 were investigated:
• Ceramic diffusers ‘‘Brandol’’ 共12 units along the chamber兲,
AS1;
• Perforated pipes of diameter 50 mm 共804 holes of diameter 2.5
mm along the chamber兲, AS2; and
• Perforated pipes of diameter 50 mm 共300 holes of diameter 6.0
mm along the chamber兲, AS3.
The discharge of air was measured by means of an anemom-

n V ⫽␮ T 冋冉⳵u y ⳵u z
⳵z
⫹
⳵y 冊 冉 冊 冉 冊
2
⫹
2 ⳵u y
3 ⳵y
2
⫹
2 ⳵u z
3 ⳵z
2 eter, installed just behind the compressor. The total amount of the
air was recalculated into the unit air supply:
Q pl ⫽Q p /L
冉 冊册
(23)
2 ⳵u y ⳵u z 2
⫹ ⫹ (17) The intensity of aeration was regulated from Q p1 min
3 ⳵y ⳵z ⫽0.0044 m3 /sm, up to Q p1 max⫽0.0107 m3 /sm 共attention: the air
It was taken into account in this relation that the transverse cir- discharge is related to the unit length of the chamber兲. For each
culation is two-dimensional 共Fig. 1兲 and turbulent. system, five different intensities were established.
In order to calculate the value of n V , described by Eq. 共17兲, The transverse circulation discharge was determined indi-
one should determine the velocity field of wastewater, or 共at least兲 rectly, the mean horizontal velocity was measured in the gap
define some model of this field. Such a model should be realistic above the baffle (u GB , in 10 points, see Fig. 2兲 and under the
and simple enough to be applied in engineering practice. baffle (u DB , in 20 points兲. Subsequently Q C was calculated from
After some considerations and analyses, it was assumed that the relation
the simplified field of velocity of the transverse circulation in the Q C ⫽ 共 u GB H G L⫹u DB H D L 兲 /2 (24)
aerated grit chamber can be described by the single, horizontal
vortex, presented in Fig. 3. Of course, such an attitude can be The velocity was measured by an electromechanical current
considered as an oversimplification, but owing to this we obtain a flow meter 共the propeller diameter⫽80 mm, made in Delft, Hol-
convenient algebraic formula, and in addition, such a model of land兲. This kind of equipment seems to be the most typical for the
the velocity field is quite similar to the measured velocity profiles considered case 关see, e.g., Brenner and Diskin 共1991兲兴. The mea-
共Fig. 5兲. Thus we can define the following estimations: surements were limited to these two regions only 共i.e., gaps above

冏 冏 冏 冏 冏 冏 冏 冏
and under the baffle兲 as the velocity field was so irregular that the
⳵u y 2u H ⳵u z 2u V ⳵u y 4u H ⳵u z 4u V accuracy of measurements in the remaining parts of the real
⬇ ; ⬇ ; ⬇ ; ⬇
⳵y B ⳵z H ⳵z H ⳵y B chamber was found to be too low.
(18)
u V ⬇2Q c / 共 BL 兲 ; u H ⬃2Q c / 共 HL 兲 ; BHL⫽V The obtained results are presented in Fig. 4 in the form of the
experimental points Q C (Q P ). Attention should be drawn to the
The coefficient of turbulent viscosity ␮ T 共its mean value for fact that this type of aeration system exerts rather poor influence
the whole chamber兲 can be computed according to the classical on the investigated function, although in the bibliography 共‘‘De-
Prandtl hypothesis sign’’ 1992兲 medium to coarse bubbles are recommended.

␮ T ⫽0.5␳␹ 2 l 2p 冉冏 冏 冏 冏冊
⳵u y
⳵z
⫹
⳵u z
⳵y
(19)
Experimentally determined points were compared with three
theoretical lines, which were obtained by means of the ‘‘hydro-
static’’ method 关Eq. 共10兲兴, ‘‘energetic’’ method 关Eq. 共13兲兴, and
where l P ⫽mean maximal distance from the wall ‘‘dynamic’’ method 关Eq. 共14兲兴, respectively. All three lines in Fig.
4 describe the idealized case, when the efficiency of the system is
l P ⫽ 共 H⫹B 兲 /4 (20)
maximal, i.e., when ␩ A ⫽1. The expected value of efficiency can
While substituting these approximated relations into Eq. 共17兲 be evaluated on the basis of experimental data for air-lift pumps
together with the auxiliary parameter when the coefficient ␩ A varies between ␩ A max⫽0.65 for minimal

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / SEPTEMBER 2004 / 1053

J. Environ. Eng. 2004.130:1050-1058.

Downloaded from ascelibrary.org by SOUTHERN CALIFORNIA UNIVERSITY on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

Fig. 4. Intensity of circulation Q C versus intensity of aeration Q P

共real object兲

Fig. 5. Measured and calculated velocity fields 关Eq. 共32兲兴

elevation head ⌬H and ␩ A min⫽0.37 for very high elevations

共‘‘Pumpen’’ 1984兲. Provided that the aerated grit chamber can be
treated as a kind of air-lift pump for which ⌬H⫽0, we can sus- By numerical integration of the measured profile of velocity
pect that for this case the object efficiency will reach 60–70%. along the vertical line y⫽B/2, from the bottom (z⫽0) up to the
Analyzing the set of theoretical curves in Fig. 4 we can state zero-velocity point (z⫽z 0 ), or from z⫽z 0 up to the free-surface
that the best coincidence with experimental points is described the z⫽H, the discharge of circulation was obtained
‘‘dynamic’’ method 关Eq. 共14兲兴. By computing the quotient of the
measured and calculated values of N D we obtain:
• For the ‘‘hydrostatic’’ model: ␩ A ⫽0.48;
Q C⫽
1
2 冉冕
0
z0
兩 u y 共 z 兲 兩 Ldz⫹ 冕
z0
H
兩 u y 共 z 兲 兩 Ldz 冊 (27)

• For the ‘‘energetic’’ model: ␩ A ⫽3.27 共physically impossible兲; The set of 10 experimental points Q C (Q P ) is shown in Fig. 6 and
and compared to Eqs. 共10兲, 共13兲, and 共14兲.
• For the ‘‘dynamic’’ model: ␩ A ⫽0.62. In the next succession the theoretical curves ‘‘intensity of cir-
Taking the above-mentioned evaluation into account, it was stated culation versus intensity of aeration’’ have been determined in the
that the best explanation of the function Q C (Q P ) was given by manner already described 共Fig. 4兲 for ␩ A ⫽0.62. As one can see,
the ‘‘dynamic’’ concept. Comparing Eqs. 共14兲 and 共22兲, we can again the best fit gives the ‘‘dynamic’’ method. The truth is that
record the level of coincidence would be better for slightly higher effi-
ciency, about 70%, but the main focus of this paper is not the
Q C ⫽0.74␩ A 关 L 2 H 2 r 5 H A v 2p Q p / 共 R r d p 兲兴 1/3 (25)
experimental determination of the efficiency of the chamber, but
where v p ⫽Eq. 共8兲, d p ⫽Eq. 共9兲 共or equivalent relations兲, r⫽Eq. the choice of the relation between Q P and Q C .
共21兲, ␩ A ⫽0.65, and It is quite important that the same conclusion 关i.e., Eq. 共14兲
gives the best evaluation of the power input兴 was drawn after
R r ⫽ 共 1⫹r 兲 2 共 1⫹r 2 兲共 r 4 ⫹3r 2 ⫹1 兲 (26)
investigations of the air-lift pumps 共Sawicki and Pawlowska
1999兲.
Experimental Verification of the Function N D
(Chamber without Baffle)
Measurements of the intensity of circulation for this case were
performed in a laboratory model of a chamber, also for Q⫽0. The
chamber had a rectangular shape of the following dimensions:
H⫽H A ⫽0.50 m, B⫽0.50 m, and L⫽1.50 m. For the aeration
system, a perforated pipe of diameter 20 mm was used 共100 holes
of diameter 1.5 mm兲. The geometrical scale of the model was not
so large as in the previous case 共the ‘‘old’’ aerated grit chamber in
Gdańsk兲, so it was possible to measure the horizontal components
of the velocity along three vertical lines by means of an electro-
mechanical current flow meter 共diameter of the propeller
⫽35 mm). An exemplary result of measurements, for the air sup-
ply Q P ⫽0.002 m3 /s, is shown in Fig. 5 共full line兲. Particular
attention should be paid to the sudden growth of velocity just
below the sewage free-surface. It is apparently the consequence
Fig. 6. Intensity of circulation Q C versus intensity of aeration Q P
of the low resistance of the air along the surface 共in contrast with
共laboratory model兲
the wall region, where the resistance of flow is much higher兲.

1054 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / SEPTEMBER 2004

J. Environ. Eng. 2004.130:1050-1058.

 Simplified Model of Kinematics of the Aerated Grit
Chamber

The relation Q C (Q P ), formulated above, enables us to determine

the velocity field in the chamber for different supplies of air. In
the most general form, this field is described by the Reynolds
equations 共for the turbulent flow兲. However, taking into account
demands of labor for computational procedures necessary for
solving such a general task 共even if a ready-made commercial
software were applied兲, it is apparent that the use of a simplified
Downloaded from ascelibrary.org by SOUTHERN CALIFORNIA UNIVERSITY on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

version 共or versions兲 of the general model is very expedient.

Characteristic features of the areated grit chambers are:
• Longitudinal component of the sewage velocity vector (u x )
does not depend on its transverse components (u y and u z ); this
statement can be justified by the fact that both the transverse
circulation and longitudinal flow can be induced in the aerated
grit chamber independently each from one another.
• Longitudinal velocity profile can be replaced by its average
value:
ux⬇ūx⫽v⫽Q/S⫽const (28)
• Using Eq. 共28兲 the continuity equation takes the form
⳵uy ⳵uz
⫹ ⫽0 (29)
⳵y ⳵z
according to which the stream-function ␺ can be introduced
subject to the following definitions:
⳵␺ ⳵␺
uy⫽ , uz⫽⫺ (30)
⳵z ⳵y
• The air bubbles, rising upward induce in the chamber a vortex
with a horizontal axis; the intensity of the vortex is practically
constant and equal to 关see Fig. 3 and Eq. 共18兲兴:
⳵u z ⳵u y
共rot u兲 x ⫽ ⫺ ⫽⍀ x ⬇⫺8Q c 共 B 2 ⫹H 2 兲 / 共 LB 2 H 2 兲
⳵y ⳵z
(31)
共although the vorticity ⍀ x can be expressed also in a different
form兲.
• Substituting Eq. 共30兲 into Eq. 共31兲 共or equivalent兲 one obtains
⌬␺⫽8Q c 共 B 2 ⫹H 2 兲 / 共 LB 2 H 2 兲 (32)
Thus, according to the above proposal, the transverse circula-
tion is described by means of the Poisson equation 共32兲. It is a
relatively simple expression and can be solved numerically rela-
tively easily. As a convenient boundary condition of Dirichlet
form, the following obvious relation can be used suit 共⌫ perimeter
of the chamber cross section兲:
␺ 兩 ⌫ ⫽0 (33)
In order to verify the simplified model of the transverse circu-
lation in the aerated grit chamber, Eq. 共32兲 was solved with use of
the boundary condition 共33兲 and Eq. 共25兲. The results of calcula-
tions were compared with profiles of the measured velocity, de-
termined previously. As can be seen in Fig. 5, where the exem-
plary effects of calculations were juxtaposed with the
experimental lines, the level of conformity is quite good. Only
just below the sewage free-surface, the difference is noticeable
since it is a consequence of the negligible flow resistance in this
part of the desander cross section 共i.e., negligible in comparison
with the flow resistance along the rigid wall of the chamber兲.
However, that effect is of local character, and thus it was assumed
finally that the model given by Eq. 共32兲 can be generally ac-
cepted.
Fig. 7. Examples of trajectories of suspended particles 关Eq. 共34兲兴
Motion of Suspended Particles
On having at our disposal an acceptable method of description of
the transverse flow, we can determine the trajectories of indi-
vidual grit particles. These trajectories are described by the fol-
lowing set of equations 共Soo 1966兲:
dvA d 共 u⫺vA 兲 w 1 w 2 g
⫽w 1 g⫹␤w 2 ⫹ 兩 u⫺vA 兩 共 u⫺vA 兲 (34)
dt df v 2s
drA
vA ⫽ (35)
dt
w 1 ⫽ 共 ␳ p ⫺␳ 兲 /␳, w 2 ⫽␳/␳ p (36)
where vA ⫽particle velocity; rA ⫽radius vector of a particle trajec-
tory; ␳ p ⫽density of a particle; g⫽gravity acceleration;
vs⫽particle free-sedimentation velocity.
The system given by Eqs. 共32兲–共36兲 is a set of relations, form-
ing the simplified model of the hydraulic characteristics of aerated
grit chambers. The quantitative analysis of behavior of each par-
ticle leads to the conclusion that two categories of trajectories can
be expected during the motion of suspension in a chamber:
• Given by a closed loop; and
• Given by an open line 共from the initial position, up to the
bottom of a chamber兲.
A series of 64 calculations were carried out 共for four different
parameters of particle, four different intensities of circulation, and
four different initial positions兲. Two example trajectories are
shown in Fig. 7. The number of parameters influencing the mo-
tion of particle is high enough, so it is very advisable to formulate
some compact conditions, which could guarantee the removal of
the coarse grit particles with maintenance of the lighter 共organic兲
particles in suspension.
Such a condition can be derived if we realize that the chamber
cross section can be divided into two parts: one with ‘‘ascending
flow’’ 共near the aeration system兲 and the second with ‘‘descending
flow’’ 共near the opposite wall兲. In the second part each particle
always moves downwards, whereas in the ‘‘ascending’’ zone such
a particle can move upwards only when the driving force is large

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / SEPTEMBER 2004 / 1055

J. Environ. Eng. 2004.130:1050-1058.

 Table 1. Characteristic Parameters of Analyzed Grit Chambers
Gdansk STP Gdynia STP
Parameter Inflow Outflow Chamber Inflow Outflow Chamber
Mean concentration 共mg/L兲:
d⬎0.1 mm 33.42 5.43 53.11 10.86
d⬎0.2 mm 14.85 0.82 37.84 6.74
0.1⬍d⬍0.2 mm 18.57 4.61 15.27 4.12
Mean reduction 共%兲:
d⬎0.1 mm
Downloaded from ascelibrary.org by SOUTHERN CALIFORNIA UNIVERSITY on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

d⬎0.2 mm 83.8 79.6

0.1⬍d⬍0.2 mm 94.5 82.2
75.2 73.1
Organic matter in 5.7 11.3
the removed
grit 共%兲:
Discharge of 1.328 1.898
circulation 共m3/s兲
Mean horizontal 0.063 0.039
velocity 共m/s兲
Maximal 0.126 0.078
velocituy 共m/s兲
Maximal bottom 0.152 0.108
Fig. 8. Aerated grit chamber ‘‘E’’ 共Gdańsk-‘‘East’’兲 cross section velocity 共m/s兲

enough. Especially important is the bottom part of the ‘‘ascend-
ing’’ zone, where the flow is mainly horizontal, with the resultant
velocity
u B ⫽ 冑u 2HM ⫹ v 2 (37)
Each particle which should remain in the suspension must cross
this bottom region, so the velocity u B should be higher than the
nonsilting velocity for smaller particles (d⬍0.1 mm for desand-
ers兲, but should be less than the proper value for bigger particles
(d⬎0.2 mm). Accepting the characteristic nonsilting velocities,
according to the classical investigations of Telford 共Watson and
Burnett 1995兲, we can record
0.076 m/s⬍u B ⬍0.150 m/s (38)
The latter expression is a condition in demand which supple-
ments other prescriptions also presented in this paper, which must
be fulfilled in the properly designed aerated grit chamber.
The shape of the second investigated object 共Gdynia STP兲 is
shown in Fig. 9. The chamber of this sand trap was divided into
two parts: aerated and conventional. The baffle, which divides the
chamber, has a completely different character than the solution
recommended for the aerated desanders, as its upper edge juts out
of the sewage free-surface and the contents of the chamber cannot
circulate around it. The basis of that decision is unknown, as the
designer has not explained it in the desander design.
The object is characterized by the following parameters: Q
⫽0.94 m3 /s 共for one chamber; number of chambers: 2兲, L
⫽21.5 m, Q P1 ⫽0.0019 m3 /sm 共344 holes of diameter 3.0 mm in
the 50 mm pipe兲.

Discussion of the Method Proposed

All considerations presented above were discussed on the basis of

data, obtained during the operation of two real aerated grit cham-
bers, in Gdansk-‘‘East’’ sewage treatment plant 共STP兲 共a ‘‘new’’
one this time兲 and that of Gdynia STP.
The schematic cross section of the first object is shown in Fig.
8. Its main parameters were equal to Q⫽0.52 m3 /s 共for one
chamber; number of chambers: 6兲, L⫽16.0 m, Q p1
⫽0.0028 m3 /sm 共perforated pipes of diameter 50 mm with 1,280
holes of 3.0 mm兲.
The efficiency of that chamber was determined six times from
1998 to 1999 during the first period of its operation. The sewage
samples of the volume 0.15 m3 were taken 12 times a day 共in each
series of measurements兲 and analyzed in the laboratory, according
to the typical standards. The mean concentrations of sand, ob-
Fig. 9. Aerated grit chamber ‘‘D’’ 共Gdynia兲 cross section
tained during these investigations, are presented in Table 1.

1056 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / SEPTEMBER 2004

J. Environ. Eng. 2004.130:1050-1058.

 Investigations of efficiency of the chamber 共carried out in the
identical way with those described previously兲 gave the results
presented in Table 1.
On comparing dimensions and characteristics of both objects
described above with the technical prescriptions, one can see that
the geometrical parameters of the desanders on the whole were
chosen properly. Some discrepancies appear during the analysis
of kinematic variables. Mean velocities are less than recom-
mended values, as for the object in Gdansk v E ⫽0.085 m/s
⬍0.20 m/s, and at the second plant v D ⫽0.074 m/s⬍0.20 m/s.
Downloaded from ascelibrary.org by SOUTHERN CALIFORNIA UNIVERSITY on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.
The functioning aeration intensities also lie below the prescribed
range 0.0045–0.0125 m3/sm 共as Q P1 ⫽0.0028 m3 /sm and Q P2
⫽0.0019 m3 /sm).
In order to assess these discrepancies in the light of the rela-
tion Q C (Q P ), proposed in this paper, the following parameters
were calculated for both investigated objects:
• Intensity of transverse circulation 关Eq. 共25兲兴;
• Mean horizontal velocity 关Eq. 共18兲兴;
• Maximal circumferential velocity; and
• Maximal bottom resultant velocity 关Eq. 共37兲兴.
The obtained results are shown in Table 1. As can be seen
there, the condition expressed by Eq. 共38兲 in both cases is prac-
tically fulfilled, whereas in the chamber ‘‘E’’ the velocity u B
slightly surmounts the upper value recommended. The mean ef-
ficiencies of the considered objects are similar to each other 共83.8
and 79.3%, respectively兲 and generally they can be accepted, al-
though one can suspect that the upper value ␩ G1 E
⫽94.5% is un-
derrated, as u B ⫽0.152 m/s⬍0.150 m/s 关Eq. 共38兲兴. In such an
E
event, the efficiency of the desander ‘‘E’’ could be improved by
diminishing the air supply. But on the other hand, such a conclu-
sion would be in contradiction with the bibliographic recommen-
dations, according to which v B ⫽0.60– 0.75 m/s. However, in the
face of considerations presented in this paper, such a high value
of the circumferential velocity could be obtained for very high
aeration intensities. One can suspect that this discrepancy can be
explained on the grounds of the rise of local velocity, near the
free-surface 共Fig. 5兲. It is possible that the bibliographic prescrip-
tion given by ASCE 共1992兲 was related to the horizontal compo-
nent of the free-surface velocity, which is about three times higher
than the calculated one. In such a case the theoretical velocities
关Eq. 共42兲兴 could be accepted.
In the Gdynia desander ‘‘D’’ the circumferential velocity still
lies in the prescribed range 关Eq. 共38兲兴, but the total efficiency of
that object is apparently less than the first one. However, it is
probably the consequence of the fact that this desander only
partly works as an aerated device, and partly as a conventional
one 共Fig. 9兲. There are no technical possibilities to distinguish the
influence of one part from another, so the efficiency ␩ GD ⫽73.1%
must be related to the total result.
It is interesting to compare the organic matter contents in the
grit, removed from both desanders; for the object ‘‘E’’ this value
is quite low, less than the demanded 10%, whereas for the object
‘‘D’’ it amounts to 11.3%, so it is slightly too high. This differ-
ence can also be correlated with the intensity of aeration, the
circumferential velocity of the desander ‘‘D’’ is lower than that in
desander ‘‘E,’’ so the grit removed from ‘‘D’’ contains more or-
ganic particles 共so much that part of this object was not aerated兲.
It means that the supply of air in the object ‘‘D’’ should be in-
creased 共it unfortunately could not be realized during the investi-
gations without previous reconstruction of the object兲.
JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / SEPTEMBER 2004 / 1057
J. Environ. Eng. 2004.130:1050-1058.
Conclusions
The method of design of aerated grit chambers results from many
years of technical experience. However, each set of prescriptions,
which creates the particular version of this method, is laden with
a very essential fault, viz. the lack of relation between the air
supply Q P and the transverse circulation discharge Q C .
In the technical bibliography one can find practically four con-
cepts, explaining the relation Q C (Q P ), but each of them was
elaborated for the air-lift pumps. In this paper three of those con-
cepts were analyzed, after having adapted to the conditions of
aerated grit chambers and subject to the empirical verification. It
was stated in the conclusion that the best fitting of experimental
and theoretical relation between Q P and Q C is rendered by the
‘‘dynamic’’ method 关Eqs. 共14兲 and 共25兲兴.
The resultant function Q C (Q P ) allows the determination of the
velocity field in a chamber. In this paper, two attitudes were con-
sidered: a ‘‘structural’’ model, based on a simplified model of the
transverse circulation 关Eq. 共32兲兴 and a ‘‘technical’’ one, reduced to
the compact prescriptions given by Eq. 共38兲.
The structural model was solved together with the equation of
trajectory of a particle. The computed velocity field showed
proper conformity with the results of measurements, so this
model can be suggested as a useful tool in technical practice.
The proposal formulated in this paper was compared with
technical situations, two examples of the real aerated grit cham-
bers 共‘‘E’’ in Gdansk and ‘‘D’’ in Gdynia兲. Generally this com-
parison led to the conclusions that both the bibliographic prescrip-
tions and the new relation presented above are consistent with the
courses of real processes. But on the other hand, these new rela-
tions should also be tested on additional data, independent from
the considerations presented above.
Acknowledgments
The writer is anxious to express his gratitude for the PEWIK-
GDYNIA and the BUDIMEX-GDAŃSK for the financial support
of this study.
Notation
The following symbols are used in this paper:
B ⫽ grit chamber width;
b B ⫽ width of air-lift zone;
C D ⫽ drag coefficient;
d ⫽ diameter of suspended particle;
d p ⫽ air bubble diameter;
E i ,E r ⫽ air bubble kinetic energy in ideal and real fluid,
respectively;
g ⫽ gravity acceleration;
H ⫽ grit chamber depth;
H A ⫽ submergence of aeration system;
H B ⫽ height of baffle;
H D ⫽ height of lower gap;
H G ⫽ height of upper gap;
h ⫽ depth of gutter;
K 1 ,K 2 ⫽ empirical constants;
L ⫽ grit chamber length;
l p ⫽ Prandtl’s mixing length;
m ⫽ coefficient;
N D ⫽ power input;
 NU ⫽ power output; ␳A ⫽ air density;
NV ⫽ power of energy dissipation; ␳p ⫽ grit density;
nV ⫽ energy dissipation per unit volume; ␳z ⫽ air–sewage mixture density;
p atm ⫽ atmospheric pressure; ␹ ⫽ von Karman constant; and
pd ⫽ bottom pressure; ␺ ⫽ stream function.
Q ⫽ discharge of waste water;
QC ⫽ transversal circulation discharge;
QP ⫽ air supply; References
Q P1 ⫽ air supply per unit length;
Albrecht, A. E. 共1967兲. ‘‘Aerated grit operation, design and chamber.’’
Rr ⫽ coefficient; Water Sewage Works, 114共9兲, 331–335.
Downloaded from ascelibrary.org by SOUTHERN CALIFORNIA UNIVERSITY on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

r ⫽ coefficient; Brenner, A., and Diskin, M. H. 共1991兲. ‘‘Model study of jet-circulated grit
rA ⫽ radius vector of suspended particle trajectory; chamber.’’ J. Environ. Eng., 117共6兲, 782–787.
S ⫽ chamber cross section; ‘‘Design of municipal wastewater treatment.’’ 共1992兲. ASCE manual and
TP ⫽ time of air bubble upflow; report on Engineering Practice No. 76, Vol. 1, Brattleboro, Vt.
TR ⫽ detention time; Hussain, I. A., and Spedding, R. L. 共1976兲. ‘‘The theory of the gas lift
u ⫽ sewage velocity; pump.’’ Int. J. Multiphase Flow, 7共3兲, 83– 87.
uB ⫽ resultant bottom velocity; Imhoff, K., and Imhoff, K. R. 共1979兲. Taschenbuch der stadtentwasse-
V ⫽ grit chamber cubature; rung, R. Oldenbourg Verlag, Munich 共in German兲.
Landau, L. D., and Lifshitz, E. M. 共1987兲. Fluid mechanics, Pergamon,
VA ⫽ air-lift zone volume;
Elmsford, N.Y.
VP ⫽ total volume of suspended air bubbles;
Neighbor, J. 共1965兲. ‘‘Design and operation criteria for aerated grit cham-
vA ⫽ suspended particle velocity; bers.’’ Water Sewage Works, 117共12兲, 37– 45.
v ⫽ mean longitudinal velocity; Olsen, N. R. B., and Skoglund, M. 共1995兲. ‘‘Three-dimensional numerical
vB ⫽ circumferential velocity; modeling of water and sediment flow in a sand trap.’’ J. Hydraul.
vP ⫽ air bubble velocity; Res., 32共6兲, 833– 844.
vS ⫽ free-sedimentation velocity of suspended particle; Pöpel, F., and Hartman, H. 共1958兲. ‘‘Der neue beluftete sandfang auf der
wi ⫽ parameter; biologischen reinigungsanlage der stadt heilbron.’’ Das Gas und
x,y,z ⫽ Cartesian coordinates; Wasserfach, 22 共in German兲.
␤ ⫽ virtual mass coefficient; Pumpen. technisches handbuch. 共1984兲. VEB Verlag-Technik, Berlin 共in
⌫ ⫽ perimeter of grit chamber cross section; German兲.
Sawicki, J. M., and Pawtowska, A. 共1999兲. ‘‘Energy balance for air lift
⌬H ⫽ elevation head;
pumps.’’ Arch. Hydro-Eng. Env. Mech., 46共4兲, 63–72.
␩A ⫽ aeration system efficiency; Soo, L. 共1966兲. Fluid dynamics of multiphase systems, Blaisdale, London.
␩G ⫽ grit removal efficiency; van Wijngaarden, L. 共1976兲. ‘‘Hydrodynamic interaction between gas
␩o ⫽ organic matter contents in removed grit; bubles and liquid.’’ J. Fluid Mech., 77共4兲, 27– 44.
␮T ⫽ dynamic coefficient of turbulent viscosity; Watson, I., and Burnett, A. D. 共1995兲. Hydrology. An environmental ap-
␳ ⫽ waste water density; proach, Lewis, Boca Raton, Fla.

1058 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / SEPTEMBER 2004

J. Environ. Eng. 2004.130:1050-1058.