Secondary Clarifier Design: Why Overflow Rate Isn't Enough

Secondary clarifier design is one of the few places in a treatment plant where a tank sized correctly by one criterion can still fail completely. Most sizing exercises begin, and too many end, with surface overflow rate: divide peak flow by the available tank area, compare the result against a guideline table, move on. The tank that results will clarify well at design flow and then lose its blanket the first time the sludge settles poorly or the influent doubles. The reason is structural. A final clarifier is not a hydraulic device with a solids problem attached to it. It is a solids separation device operating under a hydraulic constraint, and the criterion that actually governs its capacity is usually the one checked last.

Two jobs, one tank

A final clarifier has to do two different things at the same time, and they are governed by different physics. Clarification is the separation of solids from the rising liquid to produce a low effluent suspended solids concentration. Thickening is the downward conveyance of settled solids to the floor so they can be withdrawn as return activated sludge. Jeyanayagam, writing in the Florida Water Resources Journal, puts the split bluntly: clarification concerns less than two percent of the solids entering the tank, while thickening handles more than ninety-eight percent.

That asymmetry matters because the two functions fail differently. A clarification failure shows up immediately as a rise in effluent TSS. A thickening failure shows up first as a rising sludge blanket, and only later, when the blanket reaches the weirs, as an effluent violation. By the time the permit parameter moves, the tank has been failing for hours or days. Worse, solids retained in a deep blanket are solids removed from the aeration basin, so the biological process loses inventory precisely when it is under load. Long-term blanket storage also creates conditions for denitrification and secondary phosphorus release in a tank designed for neither.

What surface overflow rate actually tells you

Overflow rate is a velocity. As floc settles, displaced water rises, and the upward velocity of that water is Q divided by the clarifier surface area. Any particle whose settling velocity exceeds the overflow rate is retained; any particle slower than it is carried to the weir. That is the entire physical content of the criterion, and it is a real constraint, but it is a constraint on the clarification function only.

Published guideline values reflect this. The SunCam course notes compiled by Voutchkov tabulate Metcalf and Eddy's recommendation for settling following air activated sludge at 16 to 32 m3/m2.day at average flow and 40 to 50 m3/m2.day at peak hourly flow, with extended aeration considerably lower at 8 to 16 m3/m2.day average. Ten State Standards caps peak hourly overflow rate at 49 m3/m2.day for conventional air activated sludge. Weir loading is typically limited to 124 m3/day per metre of weir length. Meet all three and you have a tank that will clarify a well-settling sludge at the flow you assumed.

What overflow rate cannot tell you is whether the solids arriving at the tank can physically get to the bottom of it. That question depends on the mixed liquor concentration, the return rate, and the settling character of the sludge, none of which appear in Q over A.

Solids loading rate and the limiting flux

The solids loading rate is the mass of solids applied per unit area per unit time. It is calculated from the combined influent and return flow multiplied by the mixed liquor suspended solids concentration, divided by the clarifier surface area. Note what this includes that overflow rate does not: the return activated sludge flow, and the MLSS the process is carrying.

There is a maximum rate at which solids can be conveyed downward through a settling suspension. Jeyanayagam calls this the limiting flux; exceed it and a rising blanket is the inevitable consequence. Most design engineers, he notes, keep maximum solids loading in the range of 25 to 35 lb/d/ft2, with rates up to 50 lb/d/ft2 encountered only in plants combining low SVI, well-designed tanks and effective solids removal. The guideline tables agree in metric terms: Metcalf and Eddy's 4 to 6 kg/m2.h at average flow for air activated sludge, and Ten State Standards' peak hourly ceiling of 10 kg/m2.h.

A secondary clarifier rarely fails because water moved through it too quickly. It fails because solids could not move down through it quickly enough.

The International Water Association model describes the clarifier solids profile in four zones: a clear water zone, a separation zone, a sludge storage zone and a thickening and removal zone at the floor. When applied loading exceeds the limiting flux, the storage zone expands to absorb the excess. Continued expansion pushes the sludge interface toward the effluent weir, at which point a thickening failure has converted itself into a clarification failure. The tank has not changed. The solids flux through it has.

The Vesilind relationship

Zone settling, Type III settling, where flocculated particles descend as a blanket with fixed relative positions, is the predominant mechanism in final clarifiers and is what governs design. Its velocity falls as concentration rises, and the standard description is the Vesilind equation, which expresses zone settling velocity as an initial velocity multiplied by an exponential decay in solids concentration. Multiplying that velocity by the concentration gives the gravity solids flux, and plotting flux against concentration produces the characteristic curve that rises, peaks and falls away.

The two Vesilind constants are properties of the sludge, not of the tank. They are obtained from a series of settling column tests at different concentrations. Where column data is unavailable, several published correlations estimate them from sludge volume index; the Maine DEP summary of state point practice cites the Daigger relationships alongside the Hartel and Popel and Ozinsky and Ekama formulations, noting that all four give broadly similar curves but that measured settling data is decisively better for day-to-day process control. That caveat is worth taking seriously: SVI correlations were fitted to data collected under a specific test protocol, and a plant using a two-litre settleometer rather than the Standard Methods graduated cylinder is not measuring quite the same quantity.

State point analysis: reading both constraints on one chart

State point analysis is the graphical device that puts the two constraints on the same axes. Plot the gravity flux curve. Draw the overflow rate operating line from the origin with slope Q over A. Draw the underflow rate operating line with negative slope equal to the return flow divided by area, starting from the total solids loading rate on the vertical axis. Where those two lines intersect is the state point, and the concentration at that intersection is the aeration basin MLSS.

The chart is then read twice. The position of the state point relative to the flux curve describes the clarification condition: inside the curve is underloaded, on the curve is critically loaded, outside it is overloaded with solids carryover. The position of the underflow line relative to the descending limb of the curve describes the thickening condition: wholly beneath it means no appreciable blanket, tangent means the blanket is at limiting flux, intersecting it means net transfer of solids from the aeration basin into the clarifier.

The practical consequence for design is that state point analysis lets you establish site-specific overflow and solids loading criteria rather than borrowing a table. That is what allows a designer to justify a smaller safety factor with evidence instead of habit. The practical consequence for operation is more immediate. A good-settling sludge produces a larger area beneath the flux curve, which means the state point has more room to move before either constraint binds. Everything that improves settleability widens the operating envelope of a tank whose dimensions are already fixed.

Settleability is a design input, not just an operating outcome

Sludge volume index remains the most commonly used settleability measure, and the working thresholds are well established: below roughly 100 mL/g is good, above roughly 175 mL/g is poor. The mechanism behind those numbers is bioflocculation. Floc-forming organisms produce exocellular polymers that bind cells into small, shear-susceptible aggregates; filamentous organisms provide a reinforcing network that lets floc grow larger, up to about 2 mm, and settle faster. Too few filaments gives pin floc and a turbid effluent. Too many gives bulking. The design question is not how to eliminate filaments but how to avoid designing a system that selects for the wrong balance.

Settleability also propagates directly into the return sludge hydraulics. A clarifier mass balance gives the required RAS flow as influent flow times MLSS, divided by the difference between RAS solids and MLSS. Jeyanayagam works a case where RAS solids falling from 8,000 to 7,000 mg/L forces the return rate up from 60 to 70 percent of influent flow simply to move the same mass of solids, and notes that the sensitivity worsens as operating MLSS rises. A plant whose RAS pumps were sized for good sludge has no answer when the sludge deteriorates.

Several design choices influence settleability before the sludge ever reaches the clarifier:

• Aeration intensity. Diffused air input above roughly 90 scfm per 1,000 ft3 of tank volume is likely to shear floc; for mechanical aeration the corresponding limit is about 3.5 HP per 1,000 ft3.
• Mixed liquor conveyance. Turbulence in channels and pumps breaks floc that the aeration basin worked to build. If mixed liquor must be pumped, pump selection is a settleability decision.
• Selected MLSS. Concentrations much below 1,000 mg/L settle poorly; above about 6,000 mg/L, mixing and oxygen transfer become limiting. A higher MLSS shrinks the reactor and enlarges the clarifier, so the optimum is the one that minimises the cost of the pair, not of either alone.
• Feed hydraulics. Energy-dissipating inlets, a deep flocculating centre well and internal baffles that suppress density currents all raise solids capture without changing tank area.
• Flow splitting. Uneven distribution between parallel clarifiers cannot be averaged away; the good performance of an underloaded tank does not compensate for the poor performance of an overloaded one.
• Sludge withdrawal. A rapid removal mechanism keeps the thickening zone thin and limits the residence time over which denitrification and secondary phosphorus release can occur.

Depth, and the variable the loading rates ignore

Neither overflow rate nor solids loading rate contains a depth term, which is why both can be satisfied by a tank that still fails under wet weather. Depth buys storage volume for the blanket to expand into during transient loading without reaching the weirs. Voutchkov's review cites full-scale work by Albertson finding the maximum hydraulic overflow rate a primary clarifier can process to be proportional to sidewater depth, and studies by Parker and others showing deeper secondary clarifiers better able to absorb hydraulic surges. Where wet weather peaking factors exceed about 2.5, a sidewater depth of 4.3 to 5 metres is recommended.

The blanket allowances that follow from this are specific. Conventional activated sludge clarifiers should hold a 0.3 to 0.6 metre blanket under average dry weather flow; biological nutrient removal plants should stay below about 0.5 metres. During transient flows, plants on separated sewers may allow the blanket to rise to about 1 metre, and combined-sewer plants with high peaking factors up to about 1.8 metres, but in all cases a buffer of at least 1 metre should remain between the blanket and the surface.

There is an alternative to buying depth, and it is worth understanding because it explains a common misreading of field data. Reducing the total solids inventory achieves much the same protection. Voutchkov's worked example takes a system at 2,500 mg/L MLSS generating a 1.8 metre transient blanket; cutting MLSS to 1,500 mg/L and the overall inventory by 40 percent reduces the transient blanket to roughly 1.1 metres under otherwise identical conditions. This is why shallow clarifiers at low solids inventory often appear to outperform deeper tanks at comparable or higher surface loading, and why comparing clarifier depths without normalising for solids inventory leads to the false conclusion that depth does not help.

A defensible sizing sequence

The order in which the checks are made matters, because each one constrains the next:

• Fix the process first. Select MLSS and SRT from the biological requirement, then recognise that this choice has already set the solids load the clarifier must handle.
• Size on solids loading before overflow rate. Compute required area at peak solids loading using a realistic design SVI, not the best SVI the plant has ever achieved.
• Check overflow rate at peak hourly flow, not average. Average-flow compliance tells you almost nothing about the wet weather condition that actually causes violations.
• Take the larger of the two areas, then check weir loading against it.
• Select depth from the peaking factor and the blanket allowance, not from the area already chosen.
• Size RAS pumping for degraded sludge. Work the mass balance at a poor RAS solids concentration and confirm the turndown range covers both ends.
• Build the state point chart for the design case and for at least one deterioration case, and confirm the underflow line still clears the descending limb of the flux curve.

Applied to an existing plant, this sequence is diagnostic rather than dimensional. The Maine DEP case study makes the point well: a 55 MGD facility losing solids at an SVI near 375 mL/g was restored to stability first by raising RAS flow, moving the underflow line back beneath the flux curve, and only then by correcting the settleability itself through wasting and dissolved oxygen control, ending near 122 mL/g. The tanks never changed. The state point did.

The judgement to carry away is that clarifier capacity is a property of the tank and the sludge together, and it is never fully described by a single loading rate. Overflow rate protects the clarification function; solids loading rate and the limiting flux protect the thickening function; depth protects both against the transients neither rate captures; and settleability sets how much of the resulting envelope you actually get to use. A design that satisfies one and assumes the rest will hold up only until the day the sludge stops cooperating.

AIRFIN fabricates primary and secondary sedimentation equipment for STP, ETP and water treatment installations, and can advise on clarifier sizing, blanket behaviour and retrofit options where an existing tank is loading-limited rather than hydraulically limited.