Control of flow in clarifiers

The sewage treatment plant of the sanitary district Obersee (upper lake of Zurich) became operative in 1973. Obersee was about the 20th plant KUSTER+HAGER Engineers had designed. (The Company I was working with for a lifetime). The plant's capacity was 90 l/s dry weather flow and it's biologic capacity was 35,000 pop. equivalents, dominated by a large cider and a large textile company.

Though the activated sludge of the plant showed a surprisingly low index, usually SVI = 80 ml/g, its four secondary clarifiers suffered heavy sludge losses. At the effluent side of the basins, there were constantly seen floc clouds near the saw tooth weirs. At dry weather flow I usually analysed between 30 and 50 mg/l SS suspended solids, and at storm water flow (limited to twice the dry weather flow) the plant lost most of its activated sludge.

At first, I accused the pendulum scraper because there was an obvious relationship between its position, moving-direction and the size of the sludge clouds. For testing the influence of the baffles we had the fins of one baffle removed. The remaining solid baffle immersed now only 0.6 meters. This with a catastrophic consequence.

But then, in 1982, Josef Müller, then supervisor of the plant and his young assistant Erwin Seliner came up with a brilliant idea. They reasoned: After a shortening of the baffles worsened the effluent, then lets prolong them. Without letting know me, they secretly prolonged one of them. They tested this in a simple way by taking a broad plastic foil in which they wrapped heavy iron rods into one end and a wooden beam into the other end. Then they fixed the weighed down foil at the baffle and adjusted its depth by more or less rolling the wooden beam into the foil at its upper end. That way they soon found the visible cloud diminished all the more the deeper they hung the plastic foil.


	Fig. 1: Josef Müller (left) and Erwin Seliner (right) in front of their clarifiers, discoverers of the effect of deep baffles in 1979. This photo was taken 27 years later, in 2006. Josef Müller is now aged 81 years.

That aroused my interests and I developed an easy and inexpensive method to determine speed and location of the flow. At least its horizontal component. Look: Measurement of flow in clarifiers.
With the aid of the two and my laboratory crew, we measured the flow simultaneously in two parallel clarifiers. One with the original finned baffle the other one with the prolonged baffle, 2.0 meters long. There is to mention both clarifiers tested were fed by one and the same activated sludge basin. That guaranteed identical conditions. I didn't investigate the distribution of the suspended solids in the basin then. I measured the flow only.

Fig. 2: The unaltered clarifier with a short baffle behind the inlet, prolonged by fins. Each fin 12 cm broad with 12 cm gaps in between

Surface = 28.0 m x 4,5 m = 126 m²	Activated sludge conc. = 5.1 g/l
Flow rate = 15 l/s,	Sludge index SVI = 80 ml/g
Overflow rate = 0.43 m³/m²·h	Temperature = 15.8 °C

.........

Fig. 3: Improved clarifier. Baffel 2.0 meters deep, prolonged by a plastic sheet.
............Simultaneousely measured under the same load as basin Fig. 2

Conclusions:

As shown in my section "the cause of currents in clarifiers" it is the sinking sludge flocs which emit an impulse onto the water body. The higher the sludge concentration the higher the impulse.

As confirmed with this essay, that impulse causes a current to propagate along the bottom of the clarifier towards the opposite wall. (Fig. 2) There the current is forced to change its direction. So it ascends towards the surface, entraining lots of activated sludge flocs toward the effluent channels. Then the current continues back towards the inlet of the basin, where it dives to the bottom of the tank, uniting with the fresh density current induced by the settling sludge again.

As dozens of further measurements in other clarifiers have confirmed, the return current always flows about halve a meter beneath the surface. The flow at the surface itself is always very slow. That's the reason why this flow remained undiscovered for over 50 years!

Now, as the operators Müller and Seliner of the sewage treatment plant Obersee have discovered, the current may be slowed down significantly by prolonging the baffles at the inlet of the clarifier. I conclude from this: If the sludge is fed to the bottom of the tank, then it can't sink further. So there is no impulse emitted.

As Fig. 3 shows, there is an ideal flow from the bottom of the tank diagonally through the basin towards the effluent channels. That usually leads to a very low concentration of suspended solids of mostly and less than 10 mg SS/l, significantly about 5 mg SS/l.

Hence I conclude: The sludge cloud near the effluent channels of a clarifier is not created by the suction of the effluent as is generally beleved, but by the lift of the reversing current!