A thickener consists of a large (cylindrical) tank (often) with a broad conical section at the base, at the centre of which is an outlet pipe/valve assembly. The stream from this outlet is termed the underflow. Around the rim of the tank is a tray, rather like a gutter, into which clarified liquor (the overflow) flows (as a result of overspill from the rim). A set of slowly rotating mechanical rakes gently scour sediment and direct it to the underflow. Feed is introduced at the centre of the tank (and distributed across it) at some distance below the surface.
How? Overview Cursory consideration of the mass balance for such a device indicates that all solids must leave via the underflow and must not therefore be presented with any opportunity to reach the (overflow at the) rim of the tank. The design must ensure this. At a given feed rate, the vertical flow velocity of liquid (both up and down) in the device will decrease as the area is increased and vice versa. In the upper part of the tank net downward motion of solids is reduced, but in the lower section it is increased – relative to a freely settling particle. What area of thickener will ensure a clarified overflow and how can it be reduced (to save space)?
Feed Conditioning The surface properties of colloids may be exploited to assist separation, by adjusting conditions so that particle-particle interactions are net attractive. Many particles carry electrostatic charge in aqueous media in general the effects of this may be: 1. Suppressed/reduced by addition of electrolyte 2. Eliminated/exploited by adjustment of pH (or pX, of some other surface reactive ion X). 3. Overwhelmed by addition of polymers/polyelectrolytes. (usually termed flocculants)
The size density and strength of the aggregate particles which form are strongly dependent on the conditioning process. For it is necessary to ensure that conditioning agents are dispersed in the feed, and have sufficient time to interact with particle surfaces in appropriate quantities to ensure optimum effect.his treatment neglects the diffusive motion of particles (which is significant at 1, but not 100 m). This estimate of the effect is therefore highly conservative. As a consequence of aggregate porosity or voidage intra-aggregate liquid is transported to the underflow within aggregates. If the integrity of aggregates is maintained, then these will form a loose sediment which will entrap further fluid in the interaggregate spaces. The maximum (volume) concentration of solids is then equivalent to that of an aggregate.
For these reasons, dense aggregates or those which readily suffer consolidation are desirable. Methods of feed conditioning 1 and 2 (above), adjusted to produce compact aggregates (systems of highly non-isometric particles may deviate from this) whilst 3 generally tends to produce a loose fluffy type. It is often necessary to establish the optimum settling properties of the system using experiments guided by theories of colloid stability. How?
Laser diffraction particle sizers usually provide a continuously stirred tank the contents of which are continuously circulated through an optical cell. The flowrate through this can be adjusted (and stopped) so that the hydrodynamic environment of the particles is varied. These instrument sense the area particles present to the laser beam and so aggregation is readily detected from changes in the size distribution. The consequences of changes in the liquid feed composition may, therefore, be investigated. The instruments usually provide an estimate of the volume concentration of particles (including the intra-aggregate fluid) which increases in magnitude above the actual solid volume concentration as aggregation is increasingly facilitated.
The strength of aggregates may be examined by adjustment of the flow velocity in the cell and pipework where they can be made to experience significant fluid shear. Light diffraction data is captured in a matter of seconds allowing process kinetics to be probed with a resolution of ~ 0.5 – 1 minute. The range of concentrations where effects can be accurately determined is limited to those where secondary scatter is negligible; this is determined by trial and error so that a total of 15-30% of the incident laser light is scattered. The level of scatter is indicated by the instrument.
The effects of an appropriate range of concentrations (C), 0<C<Cu, on settling may need to be established (using optimum feed conditioning) by batch settling tests (where Cu is the thickened product concentration). The batch tests provide the response of the solids to gravity via uc as f(C). A downward flux of particles ? through an horizontal plane in the thickener is calculated from uc and C directly (? =ucC). ? rises to a (rounded) maximum and then decays usually passing through a point of inflection and tending to zero (equivalent to a newly formed sediment). The downward component of solid flux ?u due to the (downward) velocity (uu) of fluid in the thickening zone is related to the area of the device and the flux of solids across the inlet into the device.