Antiscalant only protects the membranes if the dose in the pipe matches the dose in the model. This is our full chain of custody for that number: water analysis → Proton projection → ml/min conversion → pump setting → graduated-cylinder validation — plus the scale chemistry that explains why we're this careful.
Narrated overview of this module. The article below has the full procedure and the numbers.
The model is only as good as the sample. We pull a representative feed-water sample and get a certified lab report — because the whole scaling calculation hinges on a handful of specific numbers. These are the analytes Proton actually uses:
Watch the gold-flagged ones: barium, strontium, and silica can scale hard at concentrations so low they look like rounding errors on the report. Never let the lab report "< detection limit" get entered as zero without thinking — barium sulfate is one of the least soluble scales there is, and a few hundredths of a mg/L matters. Bad inputs don't make Proton wrong; they make it confidently wrong.
We enter the feed-water analysis into Proton (American Water Chemicals' projection software) along with the system design: feed flow, permeate flow, and recovery. Recovery is the whole story — at 75% recovery, everything the membranes reject gets squeezed into a quarter of the water, so concentrations in the reject stream run roughly 4× the feed. Proton calculates whether that concentrated stream crosses the saturation line for each scale species:
From those saturation levels, Proton recommends a specific AWC antiscalant product and a dose in ppm (mg/L), applied to the feed flow. That ppm number is the handoff — everything after this step is about faithfully reproducing it in the field. It also tells you the maximum safe recovery; if the projection shows a species Proton can't hold, the answer is a design change, not more chemical.
Why does a couple ppm of antiscalant hold back water that's 2–3× past saturation? Because scale doesn't form the instant water is supersaturated — it has to get through a bottleneck called nucleation, and that bottleneck is exactly where the antiscalant attacks.
Nucleation is the hardest step in scaling — and the last moment the process is stoppable. A seed crystal is a template: growth onto it is fast and self-accelerating, and no practical dose of antiscalant will re-dissolve it. Stop the seed and the whole cascade never starts. That's why underdosing "a little" isn't a little problem — dosing below threshold is close to not dosing at all.
Three mechanisms working together: threshold inhibition — the molecule adsorbs onto embryonic crystal clusters and stalls nucleation, so supersaturated water stays dissolved through the vessel; crystal distortion — any crystals that do form grow deformed, soft, and non-adherent; dispersion — charged molecules keep fine particles repelled and swept out with the concentrate.
Same chemistry, different location. Precipitation is solids forming in the bulk water; scale is crystals growing attached to a surface — membranes, brine seals, concentrate piping. Scale is worse: it grows exactly where flow and rejection concentrate the water most (the tail-end elements), and it anchors as it grows.
On an autopsy: a white-to-tan crystalline crust on tail-end element surfaces (carbonate fizzes with acid; sulfate doesn't — a quick field ID). On the panel, before you ever open a vessel: rising differential pressure across the last stage, falling permeate flow, climbing salt passage. Carbonate scale can be cleaned at low pH; barium sulfate is effectively permanent. Prevention isn't a preference — for some scales it's the only option.
Proton speaks in ppm on the feed flow. The pump speaks in ml/min. The conversion has to account for how concentrated the chemical in the day tank actually is:
Worked example — 100 gpm feed, 3.0 ppm dose, neat antiscalant at SG 1.15, no dilution: (100 × 3.785 × 3.0) ÷ (1.15 × 1000 × 1.0) = 0.99 ml/min. Diluted 1 part chemical to 9 parts RO permeate (10% solution): ≈ 9.9 ml/min.
Notice what the example shows: neat antiscalant rates are often tiny — around 1 ml/min. Metering pumps are least accurate at the bottom of their turndown, so we dilute with RO permeate to bring the rate into the pump's comfortable mid-range. Always dilute with permeate or DI, never raw water — hard makeup water can gel some antiscalants in the day tank. Keep dilutions fresh (mix what you'll use in a few days).
Enter your system numbers — get the pump target and what your 60-second drawdown should read.
Illustrative field math — confirm final dose & product selection against the Proton report for the specific project.
Never trust the dial. Pump nameplate output assumes ideal suction, discharge pressure, and fresh diaphragms — reality differs. The drawdown test measures what the pump is actually moving, using nothing but a graduated cylinder and a stopwatch:
Scale is stoppable only before the first seed — so the dose Proton models must be the dose you can prove in the cylinder.
