Cities Are Flushing Away 2–8% of Treated Water. A Clarifier and a Pinch of Polymer Can Get It Back.

Water treatment plants routinely clean filters by “backwashing” (a short reversal of flow to lift and expel trapped solids). That rinse is just 0.5–2% of plant flow in typical systems (EPA design guidance), yet global estimates reach 8% of production (Wolska et al., 2023). It’s low-strength waste—often ~0.1% total suspended solids (TSS)—and it’s recoverable (EPA).

Wolska and Urbanska‑Kozłowska report that backwashing can consume 2–8% of a plant’s treated water and that re‑using it “can lead to a decrease in the cost of the water being used,” with potential savings of €150–250K per year in fees (source; source). Regulators back the trend: the U.S. Filter Backwash Recycle Rule encourages recirculation because spent backwash is often richer in pathogens (Giardia/Cryptosporidium) and metals than raw feed (WesTech case).

At most conventional plants, backwash originates from dual‑media beds of sand and anthracite; modernized trains add ultrafiltration (UF) membranes upstream of reverse osmosis (RO) polishing. Utilities looking to cut waste are pairing those filters with compact clarifiers and polymer flocculants to settle and reuse the wash water. As an equipment anchor in those trains, sand media such as sand/silica filtration remains the workhorse pretreatment.

Equalization basin and design flows

Designs start by assuming total backwash ≈0.5–2% of plant flow (EPA). Because backwash volume may surge rapidly, an equalization/surge basin is required upstream of treatment. From there, the water is pumped (or gravity‑fed) to a clarifier or thickener.

As a sizing gut check, a 10 MLD (million liters per day) plant produces backwash on the order of ≈1–2 m³/min at peak, meaning ≥600 m³/day to treat. EPA guidance assumes initial solids ~0.1% rising to ~1% in the thickener—a 10× concentration of solids and a 90% volume reduction in settling (EPA). In practice, operators might aim for clarifier underflow ~2–5% solids.

Clarifier/thickener parameters and polymer

A sedimentation tank (clarifier/thickener) receives the flocculated backwash; “clarifier” refers to a gravity settling basin, while a “thickener” is designed to concentrate sludge underflow. Ferric‑lime clarifiers or lamella (“SuperSettler”) units sized for slow flow—0.5–2 m/hr overflow—typically yield effluent ~<5 NTU (NTU, Nephelometric Turbidity Unit) when paired with proper flocculation.

Retention time of roughly 2–4 hours plus a flocculator is common. In Napa’s design (21 MGD WTP), a single basin handled full surges and achieved <2 NTU effluent without polymer (WesTech case). Bench tests there used only 0.4 mg/L polymer to reach ~0.7 NTU (source), consistent with typical feeds of a cationic polyacrylamide at ~0.2–1.0 mg/L. Plants commonly meter polymer with a chemical dosing pump and standard flocculant formulations.

Lamella settlers and fill‑and‑decant options

Options include fill‑and‑decant clarifiers with floating decanters (as in Napa) and lamella settlers with continuous overflow. Mechanical flocculators or static inlets provide mixing. Design surface loading (settle velocity) might be ~0.5–1.5 m/hr (2–6 gpm/ft²) for these dilute flows. Plants seeking compact footprints often standardize on a lamella settler sized for surges or coupled with equalization.

Importantly, no polymer or coagulant should carry into the recirculated water; the decanted “clear” should show minimal coagulant residual. If needed, dechlorination (e.g., activated carbon) can remove chlorine from filter backwash to protect seeded biofilters later—an application well‑served by activated carbon media.

Supernatant return to plant head

After settling, the clean supernatant (clarifier overflow) is returned to the plant’s influent. EPA explicitly recommends pumping thickener overflow back to the “head of the treatment plant” (EPA). Napa pumps clarified filter wash (<2 NTU) directly to the flocculation inlet; downstream coagulant/floc steps remove any small remaining turbidity (WesTech case).

Typically, reuse is capped at 5–10% of flow to avoid perturbing plant balances; Napa recycles ~8–10% flow (source). Systems with very high reuse (Napa did ~98% overall reclamation) require zero‑liquid‑discharge setups, but even 5–10% yields big savings (source). For safe filtration, turbidity is often held <1–2 NTU (source) and microbial counts minimized; depending on raw feed, operators may blend clarified backwash slowly or add extra disinfection at the head. One study found that up to 10% blending of backwash with raw caused no deterioration in UF permeate and even improved organic removal (Gora & Walsh), aligning with how many utilities operate ultrafiltration trains.

Water‑footprint and cost impacts

Reusing supernatant drastically cuts make‑up water needs. The Napa upgrade (21 MGD WTP) reclaimed 98% of all plant feed (WesTech). Returning even 5% of flow reduces raw‑water intake by that fraction. A 10 MLD plant recovering 1 MLD costs about 1 MLD less abstraction (≈1,000 L/s less pumping). Economically, Wolska et al. (2023) estimate reuse can cut environmental costs by €150,000–€250,000/year for moderate‑size plants (source).

More broadly, recycling backwash effluent lowers net demand by the returned fraction. If backwash is 2% of flow and 90% of that is cleared and reused, raw intake drops ~1.8%. Large plants may realize >5% savings. An EPA example shows a clarifier can eliminate 90% of backwash waste volume (EPA). In Napa’s case, after settling and mechanical processing, 96–98% of water in the backwash stream was ultimately recovered (WesTech).

Sludge thickening and dewatering train

The settled solids from the clarifier go to sludge handling. After gravity thickening, sludge solids typically reach ~1–3% (Napa reported ~3%) (WesTech). This wet mud is then concentrated via centrifuge, belt/filter press, or drying beds. With conditioning (polymer and/or lime), filter presses can produce ~40–70% solids for lime and alum sludges (EPA). A bowl centrifuge might similarly achieve 20–30%.

Net effect: from 0.1% original solids to ~20% final cake, sludge volume drops ~200–300×. For example, the Cibinong WTP in Java generated ~394 m³/day backwash/sludge (at ~0.1% solids), which thickening would reduce to ~39 m³ of 1% sludge, and a press (40% solids) would yield only ~1 m³ of cake per day. The cited study projected ~394 m³/d → 1–2 m³/d cake (EPA; EPA). Dewatering polishes the cake and produces a clean filtrate (often recycled or discharged).

Disposal routes and local regulations

The dried sludge cake is normally hauled to landfill or used beneficially. Drinking‑water sludge is typically classified as non‑hazardous (largely iron/aluminum/silica), so it can often go to municipal landfills (Hidalgo et al.). In some jurisdictions it may even be sold into cement or brick manufacturing, where alum/silica content is an asset (review).

Indonesian regulation (PP 16/2005) mandates that WTP wastewaters (including sludge) must be treated before discharge; in practice this means sludge should not simply be dumped into rivers (Adityosulindro et al.). In Indonesia, some WTPs use sludge drying beds or composting. Without reuse, alum‑laden WTP sludge is “typically considered a waste and disposed to landfill” (review; Hidalgo et al.).

Performance targets and trade‑offs

A backwash recovery system can cut raw‑water needs by the processed volume (~5–10% of flow) and greatly reduce discharged effluent. Savings come at the cost of capital (clarifier, pumps, polymer feed) and O&M (polymer, sludge handling). However, sludge handling costs drop because water is removed on‑site. One pilot study found that blends of 10% backwash had no adverse effect on permeate and allowed smaller coagulant doses (Gora & Walsh), suggesting cost‑neutral or positive operation.

Clear performance targets emerge: reclaimed‑water percentage; residual turbidity <0.5–2 NTU; sludge cake dryness 15–30%. Empirical projects report >90% backwash recovery, treated water turbidity <1–2 NTU (WesTech), and sludge cake solids ~20–40% (EPA). In summary, a properly sized clarifier system with flocculation can enable >90% of filter wash water to return to service (EPA), translating into roughly a (feed‑flow)×90% water‑footprint reduction for each volume of backwash treated. Modern claims (vendor data) even suggest >99% water‑recovery is achievable (WesTech), although real projects often aim for 90–98%.

Equipment anchors in the recovery train

For the settling step, utilities typically install a dedicated clarifier sized for 2–4 hours detention and 0.5–2 m/hr overflow; compact retrofits favor a lamella settler. Polymer addition is controlled with a dosing pump and standard flocculants, while front‑end filter trains continue to rely on sand media and, where applicable, membrane units such as UF pretreatment. If residual disinfectant removal is required before recycle, activated carbon is a common dechlorination medium.

Sources: Wolska & Urbanska‑Kozłowska, Water 15 (2023) (source; source); City of Napa WTP case (WesTech, 2011) (source; source); EPA (1993) (source; source); Gora & Walsh (2011) (source); Adityosulindro et al., Lingkungan Tropis 7(2) (2013) (source; source); sludge reuse reviews (source; source); vendor recovery claim (>99%) (source).