Water plants are ditching “set‑and‑forget” dosing. Real‑time sensors are trimming alum bills by 20–30%.

Conventional coagulation control still leans on jar tests and fixed “flow‑paced” feed rates — set in mg/L (milligrams per liter) or lb/MG (pounds per million gallons) — even as raw water swings by the hour. Jar testing, the century‑old “snapshot” method, can’t adapt to rapid changes, so operators apply big safety buffers. Plant chemists report adding 10–50% extra coagulant “just to be on the safe side” (pollutiononline.com).

The cost consequences are global. Coagulant spend runs into the billions — e.g., about $5.4 billion in 2013 (mdpi.com) — and higher dosing means more sludge disposal and residual aluminum. Flow‑proportional dosing ignores real‑time turbidity or organics, so it’s inherently non‑optimal under storm or seasonal swings (pollutiononline.com; realtechwater.com). In Indonesia (as elsewhere), strict drinking‑water turbidity limits (Permenkes 492/2010) and DBP (disinfection by‑products) rules demand more consistent treatment, yet without online feedback many plants still rely on manual adjustments.

The result: conventional practice leans high. One survey noted operators routinely maintain coagulant feeds many percent above what jar tests alone indicate (pollutiononline.com).

Streaming current monitors for charge control

Streaming current monitors (SCM) give continuous, real‑time feedback on colloid charge in the flash‑mixed stream. In the SCM cell, a small sample is shuttled past electrodes, generating a current proportional to the net ionic surface charge (pollutiononline.com). The signal is “zeroed” under neutral conditions (pH‑7); positive or negative readings indicate under‑ or over‑dosing. At the correct coagulant dose — full neutralization of negative particle charge — the reading approaches zero (waterworld.com).

Operators calibrate an SCM setpoint near zero and trim coagulant feed until that setpoint is maintained (pollutiononline.com). Unlike laboratory tests, an SCM provides immediate charge readings — it “knows precisely, and at all times, the optimum coagulant dosage” (pollutiononline.com). The output is typically 4–20 mA to a PLC (programmable logic controller) for closed‑loop control or instant alarms (pollutiononline.com), making it a natural companion to an accurate chemical dosing pump and plant PLC logic.

Utilities report rapid response during raw‑water events. Vendors and plants emphasize that an SCM lets operators “decide on proper dosage changes” during spikes or storms (waterworld.com). In one documented case, a Philadelphia plant under flood conditions used an SCM to hold filter effluent turbidity at only 0.012–0.025 NTU (nephelometric turbidity units), even as raw turbidity soared to 46 NTU (waterworld.com).

Because SCMs reflect coagulant efficacy at the start of the process, they help minimize overdosing. After installation, many plants observed large drops in aluminum carry‑through; one utility found “significant reductions in finished‑water aluminum” — consistent with trimming alum feed to the minimum needed (pollutiononline.com). A South African study reported that an SCM‑based control system “minimized under‑ and overdosing” so well that dosing stabilized and “cost savings were noticeable” (researchgate.net).

In practice, an SCM may be used alone or as part of a hybrid loop (often with a parallel PCR or turbidity meter). Operators or control systems typically choose a streaming‑current setpoint — often slightly south of zero, based on jar tests — then hold the dose at the lowest value that still achieves target turbidity/clarity. Industry guidance is blunt: streaming‑current control “not only eliminate[s] [the] potential for chemical overuse, but can actually help the operator optimize usage to the minimum level required to maintain optimum water quality” (slideplayer.com) (emphasis added).

Particle counting for coagulation and filtration

Liquid‑borne particle counters sample treated water and count — even size — individual particles (typically 0.5–400 μm) via laser scattering or blockage. They work below the sensitivity of conventional turbidimeters: Louisville Water reported filter effluent turbidity appearing constant at 1.0 NTU while particle counts varied from 5 to 100 particles/mL (wwdmag.com). Small flocs and incipient breakthrough events can be “invisible” to a turbidity meter but are clear in particle counts, giving operators a finer performance view.

In practice, counters are often installed on filter effluent or clarifier overflow lines. At Louisville’s main plant, on‑line counters revealed early filter breakthroughs (rising counts) hours before turbidity climbed (wwdmag.com). Using this data, engineers optimized individual filter run times and coagulant feed: one filter cycle was shortened from 96 to ~70 hours to avoid particle tagging, and the minimum alum feed rate was actually increased from 10 to 20 lb/MG to stabilize performance (wwdmag.com). While that case did not reduce chemical use (it raised it to improve quality), it illustrates how real‑time particle monitoring underpins data‑driven dosing decisions for coagulants and supporting flocculants.

As a final check on coagulant effectiveness, a sudden jump in particle count immediately after coagulation suggests under‑dosing, prompting dose ramp‑up; conversely, persistently zero particle counts might suggest overdosing. Reviews report “surprisingly high” accuracy and repeatability (wwdmag.com) and measurement at purity levels “below the range where turbidity measurements are effective” (wwdmag.com). They are increasingly used alongside SCMs and UV monitors in advanced plants.

Integrated feedback and feed‑forward control

Together, these sensors enable advanced control schemes. Feed‑forward systems use raw‑water sensors (turbidity, UV254 — ultraviolet absorbance at 254 nm, a proxy for organics — color) to set dose before storms, while feedback loops use SCM or post‑coagulation monitors to trim the actual dose. Modern programs even apply fuzzy‑logic or model‑predictive algorithms that take multiple inputs. In one bench‑scale trial with synthetic 110 NTU water, fuzzy SCM/pH feedback cut effluent turbidity below 10 NTU (mdpi.com). In pilot and full‑scale implementations, combined monitors guide operators to much lower doses than manual practice.

Chemical savings and downstream gains

The bottom line is significant chemical savings. By installing a dual‑stream UV254 organics monitor on intake water, Ridgway WTP (PA, USA) cut annual alum use by 20–30% (realtechwater.com). Operators underscored the point: “If we see a good raw water UV254, we know we can cut back [alum] to save costs… we’re regularly saving 20–30% on coagulant use throughout the year” (realtechwater.com).

This aligns with the baseline reality: if plants typically overfeed by 10–50% as a safety margin (pollutiononline.com), then informed, real‑time control can claw back much of that excess. Industry sources note that with SCMs and similar sensors, “chemical usage can be reduced with no effect on water quality” (slideplayer.com). In many cases a 10–20% reduction is readily achieved; in others (high variability sources) even higher savings have been reported.

Reduced coagulant feed also cuts sludge production roughly proportionally, lowering disposal costs and volumes. Better‑controlled coagulation can improve filter runs and lower DBP precursor levels, yielding indirect economic and regulatory benefits (realtechwater.com). In short, online monitoring transforms coagulation from a brute‑force cost center into a tightly tuned process. Rather than adding a large fixed dose and hoping, utilities can use SCMs, particle counters (and related sensors) to “optimize the chemical dose in response to current water quality” (realtechwater.com; realtechwater.com). Case studies and reviews consistently report that this approach yields sizable savings versus manual or simple flow‑paced dosing (researchgate.net; realtechwater.com). One industry engineer summarized it: automated control “not only eliminate[s] potential of chemical over use, but can actually help the operator optimize usage to the minimum level required” (slideplayer.com).

Sources and case references

Peer‑reviewed studies and industry reports were used. Key references include Ratnaweera and Fettig (2015) on online coagulation control (mdpi.com; mdpi.com), case studies in WaterWorld/WWD (waterworld.com; wwdmag.com), U.S. EPA/WRC reports (researchgate.net), and treatment‑plant case studies (Real Tech blogs) (realtechwater.com; realtechwater.com). Each factual claim above is supported by the cited sources.