War. Res. Vol. 26. No. 5. pp. 625-628. 1992 Pnnted in Great Britain. All rights reserved
0043-1354/92 $5.00 + 0.00 Copyright ~ 1992 Pergamon Press pk
FLOW REGULATION BY AUTOMATICALLY CONTROLLED OVERFLOW WEIRS L. CSI~PAI~) and F. KASTANEK Hauptstrasse 25. A-2M0 M6dling, Austria
(First receired March 1991; accepted in rerised form October 1991)
Abstract--ln larger sewage plants, motor-regulated overflow weirs are sometimes used to achieve a uniform flow distribution into the secondary settling tanks, backed up by simultaneous control of the immersion depth of the upstream aerating units. It is demonstrated that, if these overflow weirs are arranged in series on one side of the feed channel, they are unable, with a fixed setting, to ensure equal flow distribution when there is a variable inflow. If flowmeters are mounted in the discharge pipes, with feedback controls to the motor-regulated weirs, it is then possible to achieve equal flow distribution. Key wortgv--equalization of flow distribution, non-uniformity of overflow height, motor-regulated weir controlled by flowmeter
INTRODUCTION
The studies which have already appeared (lmhoff, 1950-1985; Stalzer and v d Emde. 1972; Escritt and Haworth. 1984) are concerned only with the following devices: Means of waste water flow division: (a) Division of the channel (using a baffle); (b) Division by regulated gate valves; (c) Division at the overflow; (d) Division by high inlet losses relative to overflow height. All these methods of division, with simultaneous regulation of surface-aerators according to oxygen content, are only able to operate, if at all. to an inadequate degree, because the raised head of water or overflow height with the types of dividing techniques referred to above depends primarily on the quantity of feed water, not on the oxygen content. It is proposed that this deficiency should be overcome by regulating wastewater distribution by means of mobile overflow weirs, with feed flow from the side. As the horizontal or vertical rotors have an optimum operating position, at which the introduction of oxygen is most favourable, the task is to maintain the water level in the feed channel, i.e. upstream in the aeration tank, at a given height with a widely varying inflow, and thereafter effect equal distribution of the overflow volume. The basic arrangement comprised a feed channel with two oppositely placed weir gates, which, according to the oxygen demand of the aeration tank, controlled by oxygen probes for regulating the immersion depth, are raised or lowered, thereby automatically maintaining the optimum water level. The quantity of water discharge on both sides is required to be the same, and, with the tops of the 625
weirs at the same height, it follows that the overflow heights are also the same, as determined by the water level contour along the oppositely placed overflow weirs. However, with overflows on one side only, the configuration of the water level critically affects the equal distribution of the flow volume at any moment, and therefore the maintenance of the optimum water level in the feed channel. With overflows on one side, flow measurements using water-level gauges are generally difficult because of the changing position of the water level (see Fig. I) (ATV-Arbeitsgruppe, 1987). and the problem becomes .
t
.[
Fig. I. Water level contour along a distributing weir under tranquil flow conditions.
V
'T" I
Originol wo~er level
:
hz
I
Weir sill
Fig. 2. Flow reaching weir via upstream baffle.
r T~--
FI
.__
i
I
I
ii ,
I
L.f.'~
C~ 7
~ection B-B:
:.
c~ 7
;lotion A-A:
"~r
"
F--] I !
. . . . . .
'!tl---_-_J
:ig. 3. Distribution system of M6dling sewage works.
I I
I
I Floating scum
i
'H--~,,T
11
!
F
.~
~ection C-C:
~ ~ A
~ ' ~ A
]
B
-7
¢
i,
?
Flow regulation by overflow weirs
more acute when babies are fitted in front of the overflows to catch floating scum (Fahner et al., 1990), as in the present case (see Fig. 2). The problem is considered below in the case of a one-sided weir arrangement with three overflows placed in series. A practical example---the distribution system of M6dling sewage works (see Fig. 3)--is examined, to see how the feed water is distributed over three side weirs situated one immediately behind the other, and whether it is permanently possible to maintain an accurate 3-fold division of the incoming wastewater without the use of motor-regulated gates controlled by flow meters. SITUATION The input to the distribution system takes place via an inlet channel carrying volumes of sewage (including return sludge) which can vary between 0.24 and 2.7 m)/s. The incoming sewage is discharged to the side over three distributing weirs arranged in series and each 4.0 m long (see Fig. 3). The crest of the weir is parallel to the bottom of the channel. The distributing weirs in question are straight, as there is no appreciable change in channel width. The width of the feed channel is 1.6 m. and the maximum height of the moveable weir crest above the bottom of the feed channel is 0.78 m. Because of the complicated spatial configuration of the flow, only an approximation to the hydraulic flow calculations is possible. The overflow volumes can be calculated using Polcni's formula:
Q where Q = the overflow volume in unit time (ml/s)
L = the length of the weir (m) h,, =, the mean overflow height (m). The overflow coefficient is equal to 95% of that for a weir lying at right angles to the flow, i.e. where/** =,/J x 0.95 (/~ - 0 . 6 3 ) , we obtain a value o f t ~ * - 0 . 6 0 . The overflow height along the distributing weir is not constant, and a uniform increase in the overflow height occurs. For the purposes of calculation, a mean overflow height, h = - ( h t + 82)/2 is used. A lateral deflection is technically rational, under tranquil or suberitical flow conditions. A modified Poleni formula is applied, in accordance with Schmidt's method of calculation, the Froude number of the feed water should be
Q~b
Fr = ~
< 0.75,
in order to be able to discount a hydraulic jump along the weir. In the above expression:
b = the width of the water level (m), and A = the flow section (m:). According to Schmidt. the following relationship exists between the overflow height at the beginning h s and end h 2 of the distributing weir: h I=
h 2 - C(I.I v~/2g - I.I v~/2g)
where r I = the mean velocity at the start of the weir (m/s) t,2 = the mean velocity at the end of the weir (m/s) w = the height of the weir crest above the bottom of the feed channel (m) and = an experimentally determined correction factor.
627
030
I
°2°
t +~
0.15
010 0 05 -o
. 06
[
[
I
1
i
!
07
O0
09
~0
11
~2
Fig. 4. Schmidt's correction factor ~ (Preissler. 1980).
The correction factor ~ is a function of h,,/(h, + w) (see Fig. 4). If there is no flow under the distributing weir. the value is to be taken as 0.65 where h=/(h= + w) > 0.3, or as 0.875, if there is underflow.
CALCULATION
As the velocity is a function of the overflow height, it can only be calculated iteratively. The three distributing weirs (each 4.0 m long) are arranged in series. and can therefore be regarded as a single, 12.0 m long weir. For Q = 0.240 m3/s, the calculation is very simple. With a weir length of 12 m, the overflow height works out at: h t = 0.048 m;
h, ffi0.053 m. Case I Given data: w = 0.38 m,
L = 12.0 m,
Q = 0.240 mJ/s.
Calculated values: h~ = 0.048 m,
h 2 -- 0.053 m,
Fr = 0.03.
The difficultiesarose when calculating the overflow height for Q = 2.7 m3/s, as this volume of water gives rise to the danger of shooting, or supercritical flow, as a result of which a poor distribution of the feed water is unavoidable. With shooting or supercritical flow, a reduction in the volume of water to be carried off laterally is likely. A calculation is therefore made of the water volume remaining to be carried off with the given dimensions, when Fr--0.75. This water volume, Q, was found to be 1.27m3/s (the value found for Fr-- I was Q -- 1.65 m3/s). Case 2 Given data: w=0.38m,
L=12.0m,
Fr--0.75.
628
L. CS~PAt and F. KA$'rANEK
Calculated values: Q = 1.27m'/s,
h~=0.105m,
h:=0.201 m.
The crest of the weir would have to be raised in order to guarantee the maximum discharge water volume of 2.7 m3/s. The weir crest height was investigated with a view to ensuring tranquil flow. The height w was found to be 0.63 m. Case 3
Given data: L=12.0m,
Q=2.7m3/s,
Fr=0.75.
Calculated values: w=0.63m,
hl=0.173m,
h,=0.336m.
The distribution of the feed water volume to the three 4.0 m weirs is calculated as follows. The total weir length of 12.0 m is reduced to 8.0 and then 4.0 m, and the differences in each case are calculated. This produces the following values:
Distributing weir I: Distributing weir 2: Distributing weir 3:
Case 2: Q = 1.27 m'/s
Case 3: Q = 2.7 m~/s
Q, = 0.245 m~/s Q2 = 0.425 m~/s Q; = 0.6 m~/s
Qi = 0.490 m~/s Q: = 0.915 m~/s
Q~ = 1.295 m~/s
In view of the small differences between the overflow heights, no calculation was made for case I (Q = 0.24 m~/s), as such calculations do not properly reflect actual conditions.
is the accurate positioning of the weir gate, as a difference of just a few centimetres can have a significant impact on the volume of water passing over. This accurate positioning of the gates, or rather the weir crests, is not simple, and their control by feedback of overflow height measurements in a heavily contaminated medium---even though this presents a theoretically possible solution--is unreliable and beset by problems. Response to the amount actually flowing over the weirs can be most reliably effected by metering the flow in the discharge channels. Besides measuring and recording the genuine throughput, this also provides a means of automatic control. It involves the transmission of pulses to the motorregulated gates, causing then to be lowered or raised when the disparity between the individual metered flow rates exceeds, say, 10%. Apart from accurate distribution, a freely selectable program also makes it possible to choose the height of the water level in the feed channel, or ultimately in the aeration tank. This method of control can therefore be used to ensure that operation takes place at the optimum position for the surface aerators, as well as equalizing the sewage flow distribution with weirs arranged in series. The problem posed in the Introduction is therefore solved, and the solution leads to simplified sewage plant design, which is an advantage not to be ignored, given the increasing shortage of available space.
REFERENCES C O N C L U S I O N ANt) S O L U T I O N
Accurate division into three of the feed flows, without adapting the heights of the weir crests to the actual discharge conditions, is not possible. To ensure tranquil flow conditions when Q > 1.27 m3/s, the weir crest has to be raised to more than w = 0.63 m. At the same time, to ensure the same water level (and the same immersion depth for the surface aerators) at all discharge conditions the three weir crests are mobile gates, automatically lowered by increasing discharge and raised by a falling one. A precondition for this, with discharge-regulated automatic overflow control,
ATV-Arbcitsgruppe (1987) Quantitative Durchflussmessung (Quantitative flow-measurement). KA 11/87, p. 1207. Escritt L. B. and Haworth W. D. 0984) Sewerage and Sewerage Treatment. Wiley, New York. Fahner H., Peter G. and Se.},boldW. (1990) Problematik der Entlastungsmessung an Uberlaufbauwerken (Problems of discharge measurement on overflow structures}. KA 10/90, pp. 1175-1188. Imhoff K. (19.50-1985) Taschenbuch der Stadtentwiisserung (Urban sewerage manual). OIdenburg, Germany. Preissler B. ([980) Technische Hydromechanik (Technical Hydromechanics), Vol. i, p. 518. VEB f/Jr Bauwescn, Berlin. Stalzer W. and v d Erode W. (1972) Division of wastewater flow. War. Res. 6, 371-373.