Activated Sludge

An aeration basin is a farm for bacteria, and the operator runs it by deciding how many to keep.

In 1914, two chemists in Manchester, England, Edward Ardern and William Lockett, were bubbling air through jars of sewage. Cleaning up one jar took about five weeks. Then they tried something that sounds like bad housekeeping: they poured off the clear water, kept the brown muck at the bottom, and added fresh sewage. Each round ran faster, until a jar was done in about a day.

They called the muck activated sludge, and nothing about it had been activated. It was just a crowd of bacteria that nobody had thrown out. Plants today still run on that one idea.

Keep the bacteria, let the water go

A modern plant runs those jars continuously, in a loop. Follow the wastewater through it.

It enters the aeration basin at flow Q0Q_0 carrying BOD, which, as Oxygen Demand showed, is really a measure of food. Blowers keep the tank churning and the bacteria breathing. The resulting brown soup is the mixed liquor, and its solids are measured two ways. MLSS, mixed liquor suspended solids, is everything suspended in it, in mg/L. MLVSS, the volatile part, is what burns off in a lab furnace. That's the organic part, and we use it as a stand-in for the living bacteria. It's usually about three quarters of the MLSS.

The mixed liquor then flows to a quiet tank, the secondary clarifier, where the bacteria clump together and sink. Clean water goes over the weir as effluent. The settled sludge splits two ways: most is pumped back to the head of the basin as return sludge, and a little is taken out for good as waste sludge.

Aeration basin Mixed liquor: V, XA Secondary clarifier Influent Q0, S0 Q0 + QR, XA Effluent Qe, Xe Return sludge QR, Xr Waste sludge Qw, Xw
The activated sludge loop
Water mostly passes straight through. Solids go around and around, and they leave only through the waste line or over the weir.

Here QQ is a flow and XX a solids concentration, with subscripts for the influent (0), the basin (A), the return (R, r), the waste (w), and the effluent (e). These are the symbols of the Handbook's Activated Sludge section, under Wastewater Treatment and Technologies, which prints no diagram of the process. It says "recycle" where most people say "return."

Everything else in this Lesson is bookkeeping on that loop: how much food each bacterium gets, how long the water stays, how long the bacteria stay, and whether they'll settle.

Food per bacterium

Think of the basin as a cattle farm. Hay arrives every day, and a herd stands in the pen. Neither number means much on its own. What matters is hay per animal per day: too little and the herd goes hungry, too much and uneaten hay is everywhere. In the basin, uneaten food means BOD going out with the effluent.

The basin's version is the food-to-microorganism ratio. The food arriving each day is the flow times the influent BOD, S0S_0. The herd is the concentration of bacteria times the basin's volume:

F:M=Q0S0VXA\text{F:M} = \frac{Q_0 S_0}{V X_A}

The Handbook prints this as "Organic loading rate (F:M)," and its design table calls it the food-to-mass ratio. Its XAX_A can be either MLSS or MLVSS, but this ratio is about feeding living organisms, so use MLVSS when a problem gives it.

The top of the F:M fraction is a mass load, and the 8.34 factor also counts what's sitting in a tank: mg/L times MG times 8.34 gives pounds. So in American units, the top and the bottom each carry 8.34, and it cancels as long as the flow is in MGD and the volume is in MG. SI cancels the same way, with m³/day and m³. What's left is mass of BOD per day per mass of bacteria, a rate per day.

The same few numbers give two more measures.

Volumetric BOD loading drops the bacteria and asks how much food the tank takes per unit of volume. It's the Handbook's "Organic loading rate (volumetric)":

Volumetric loading=Q0S0V\text{Volumetric loading} = \frac{Q_0 S_0}{V}

Nothing cancels this time, so the units stay: lb of BOD per day per 1,000 ft³ in American practice, and kg of BOD per day per m³ in SI.

Hydraulic detention time is how long the water stays, the same V/QV/Q you'd use for any tank, as in Sizing Treatment Units. The Handbook calls it hydraulic residence time and writes θ\theta:

θ=VQ0\theta = \frac{V}{Q_0}

The return flow runs through the basin too, but by convention θ\theta leaves it out.

Worked Example Sizing up an aeration basin
1 of 4

A plant treats 6.0 MGD with an influent BOD of 150 mg/L. Its aeration basin holds 200,000 ft³, with an MLVSS of 2,000 mg/L. Find the F:M, the volumetric BOD loading in lb/day per 1,000 ft³, and the detention time.

Two clocks in one tank

Detention time answers a simple question: how much water is in the tank, compared with how much leaves each day? Now ask the same question about the solids.

The basin holds VXAV X_A of them. They have two ways out: waste sludge pumped away on purpose, QwXwQ_w X_w, and the few that escape over the weir, QeXeQ_e X_e. Divide what's there by what leaves each day:

θc=VXAQwXw+QeXe\theta_c = \frac{V X_A}{Q_w X_w + Q_e X_e}

This is the solids retention time, or SRT, also called sludge age. The Handbook calls it solids residence time, and its design table says mean cell residence time. Look at what isn't in the denominator: the return flow. Those solids aren't leaving, just going around.

Which solids go where?

The waste and effluent streams are measured as total suspended solids, so count the basin the same way: θc\theta_c takes MLSS. F:M is about feeding the living bacteria, so it takes MLVSS.

Because the solids keep coming back, the two clocks run at very different speeds. The water is gone by dinner, and the bacteria are still around next week.

Water: θ ≈ 6 hr Solids: θc ≈ 8 days 0 2 4 6 8 10 Time in the plant (days)
Time spent in a typical basin
One tank, two very different stays. The clarifier and return line are what pull them apart.
Worked Example Finding the sludge age
1 of 3

A 1.2 MG aeration basin has an MLSS of 3,000 mg/L. The plant wastes 0.04 MGD of sludge at 8,000 mg/L, and returns 2.4 MGD at the same concentration. Its effluent is 4.0 MGD at 20 mg/L of suspended solids. Find the solids retention time.

Will it settle?

All of this rests on one fragile thing: the bacteria have to settle in the clarifier. If they don't, they go over the weir, the effluent turns cloudy, and the herd you've been keeping walks out the door.

To check, pour a liter of mixed liquor into a tall cylinder, wait 30 minutes, and read how much room the sludge takes up. More solids fill more of the cylinder even if they settle beautifully, so the fair question is how much room each gram takes. That's the sludge volume index, from the Handbook's Activated Sludge section:

SVI=sludge volume after settling (mL/L)×1,000MLSS (mg/L)\text{SVI} = \frac{\text{sludge volume after settling (mL/L)} \times 1{,}000}{\text{MLSS (mg/L)}}

The 1,000 turns milligrams into grams, so SVI comes out in mL/g.

0 200 400 600 800 1,000 mL 250 mL Settles well SVI = 100 mL/g 0 200 400 600 800 1,000 mL 700 mL Bulking SVI = 280 mL/g Both: MLSS = 2,500 mg/L, after 30 minutes
Two sludges with the same solids, 30 minutes into a settling test
At 2,500 mg/L, the left cylinder gives (250)(1,000)/2,500=100(250)(1{,}000)/2{,}500 = 100 mL/g, and the right gives (700)(1,000)/2,500=280(700)(1{,}000)/2{,}500 = 280 mL/g.

A sludge that settles well has an SVI of about 100 mL/g. Above about 150 mL/g, you likely have bulking: often, threadlike bacteria hold the clumps apart, so the sludge stays fluffy and won't pack down. Since SVI is room per gram, high is bad. The Handbook prints only the formula, without the 30 minutes, the benchmarks, or the word bulking, so those are yours to carry.

How much to send back

How hard do the return pumps have to work? Write the balance first. At steady state, the solids going into the clarifier equal the solids coming out, and the incoming flow includes the return, which just came through the basin. The Handbook's Steady-State Mass Balance around Secondary Clarifier reads:

(Q0+QR) XA=QeXe+QRXr+QwXw(Q_0 + Q_R)\,X_A = Q_e X_e + Q_R X_r + Q_w X_w
Secondary clarifier in Q0 + QR, XA over the weir Qe, Xe return QR, Xr waste Qw, Xw
A solids balance around the secondary clarifier

The return carries nearly everything. The solids over the weir and out the waste line usually come to only a few percent, so unless a problem gives them, drop them:

(Q0+QR) XA=QRXr(Q_0 + Q_R)\,X_A = Q_R X_r

Divide through by Q0Q_0 and bring in the Handbook's recycle ratio, R=QR/Q0R = Q_R/Q_0. Solving gives a short form the Handbook doesn't print:

R=XAXr−XAR = \frac{X_A}{X_r - X_A}

The thicker the sludge coming off the clarifier floor, the less of it you have to pump back, and a bulking sludge never gets thick. When a problem gives the waste and effluent terms, keep them and solve the full balance for QRQ_R.

Worked Example A settling test and a return pump
1 of 4

A plant treats 5.0 MGD. Its mixed liquor, with an MLSS of 3,000 mg/L, settles to 270 mL in a 1 L cylinder in 30 minutes. The return sludge is 9,000 mg/L, and the solids in the waste and effluent are small enough to ignore. Find the SVI, the recycle ratio, and the return flow.

Steering with the waste valve

Bacteria keep multiplying as long as there's food. If nothing took them out, the solids would pile up until they swamped the clarifier. So the plant wastes sludge, and that one valve is the operator's main control.

Suppose the operator cuts the waste rate, with the influent unchanged. Fewer solids leave each day, so θc\theta_c climbs. The solids build up in the basin, so MLSS climbs. The same food now feeds more mouths, so F:M falls. Notice what doesn't move: θ\theta, which knows only about the basin's volume and the influent flow.

Before Food: Q0 S0 Waste MLSS lower θc shorter F:M higher the waste rate After cutting Food: Q0 S0 Waste MLSS higher θc longer F:M lower
The same food, with less sludge wasted

The Handbook's design table shows the same trade-off, here reduced to two of its processes: extended aeration pairs a long sludge age with a low F:M. Notice, too, that θ\theta is in hours and θc\theta_c is in days. The table bases its F:M on MLSS, so its numbers run a little lower than an MLVSS-based F:M.

Typical ranges for two activated sludge processes
Processθc\theta_c (days)F:M (per day)θ\theta (hr)
Conventional4–150.2–0.44–8
Extended aeration20–300.05–0.1518–24

Adapted from the FE Reference Handbook, Design and Operational Parameters for Activated-Sludge Treatment of Municipal Wastewater

Checkpoint
1 of 3

An operator doubles the waste sludge rate and holds it there. The influent flow and BOD don't change. Once the plant settles down, what has happened?

A.
θc\theta_c is longer, MLSS is higher, and F:M is lower.
B.
θc\theta_c is shorter, MLSS is lower, and F:M is higher.
C.
θc\theta_c is shorter, MLSS is lower, and F:M is unchanged.
D.
θc\theta_c and θ\theta are both shorter, and MLSS is unchanged.

Recap

  • The clarifier sends the bacteria back, so the solids stay for days while the water passes through in hours. Waste more sludge and θc\theta_c and MLSS fall while F:M rises, and θ\theta doesn't move.
  • F:M=Q0S0/(VXA)\text{F:M} = Q_0 S_0/(V X_A) is food per bacterium per day, using MLVSS. Volumetric loading drops XAX_A, and θ=V/Q0\theta = V/Q_0 is how long the water stays.
  • θc=VXA/(QwXw+QeXe)\theta_c = V X_A/(Q_w X_w + Q_e X_e) is the solids in the basin over the solids leaving each day, using MLSS.
  • SVI is settled volume in mL/L times 1,000, over MLSS in mg/L. Above about 150 mL/g, it signals bulking.
  • A solids balance on the clarifier gives R=QR/Q0=XA/(Xr−XA)R = Q_R/Q_0 = X_A/(X_r - X_A) when the waste and effluent solids are negligible.