RO Membrane Cleaning Chemicals: Three Makers, Three Different Envelopes

Most CIP programs run on a recipe somebody wrote down once. Caustic for the organics, citric for the iron, flush in between. The recipe travels between brands fine. The pH and temperature ceiling attached to it does not, and that's where a cleaning turns into damage.

Here's the specific trap. The NanoH2O (formerly LG) BW 400 R G2 data sheet lists a cleaning pH range of 1 to 13, and the only temperature on the page is a 45 C maximum operating temperature. Read those two together and a pH 12 caustic at 35 C looks comfortably inside spec. Technical Service Bulletin 113 is the document that caps pH 12 at 25 C (77 F) on a brackish element, and nothing on the data sheet tells you it exists.

What the three makers put in the tank

Published formulations, by weight, from each maker's current cleaning document.

The FilmTec temperatures are recommended conditions for those specific solutions, and they don't always match the compatibility envelope in the next section. The alkaline pair line up at 35 C. The acid doesn't: Table 3 puts HCl at 25 C where the compatibility table allows 45 C down to pH 10.5. Mix to the solution recommendation.

Now look across the chelant column. Hydranautics' default is sodium tripolyphosphate with EDTA at pH 10.0, not caustic with EDTA at pH 12. Lift the familiar caustic-plus-EDTA recipe out of a FilmTec manual, run it on a CPA or ESPA element, and you're two full pH units above the cleaner Hydranautics wrote for it. NanoH2O splits by grade too: its caustic range for a BW R G2 runs to pH 13, a plain BW R to 12.

A 1,000 gallon batch weighs roughly 8,350 lb, so 0.1% NaOH is 8.3 lb of pure caustic, about 1.3 gallons of the 50% liquid, and 2.0% citric is 167 lb of anhydrous powder. Acid is the one to check twice, since 31.4% muriatic (20 degrees Baume) and pure HCl are a factor of three apart and a bulletin saying "0.5% HCl" doesn't always say which it means. Compare acids by target pH instead. Acid into water, never the reverse, face shield on, and keep hydrosulfite clear of the acid step, because dithionite and acid make sulfur dioxide.

The temperature ceiling is where the brands separate

Read the DuPont row carefully, because it covers the FilmTec BWRO, SWRO, low energy, HSRO and NF90 families. The NF270 side of the same table is tighter, its 45 C band reaching only pH 3 to 10. Quote the wrong column and you get a correct number attached to the wrong element.

Line the first two up and the gap is plain. A pH 12 caustic warmed to 35 C is inside FilmTec's published envelope and 10 C outside NanoH2O's for a brackish element. Both are defensible calls by the people who built the respective membranes, and both are company positions about their own product rather than properties of polyamide, so neither carries across to somebody else's element.

Hydranautics is organised differently again, per element family, so its CPA and ESPA rows aren't interchangeable. It also attaches a clock the other two don't: at the limit pH you get 60 minutes under 40 C, half that over it. The long soak and the limit pH are separate modes. NanoH2O's extended soak window is 1 to 12 hours, standard recirculation 45 minutes.

There's a floor too. Below about 20 C (68 F) the kinetics get slow enough to waste the clean, and sodium lauryl sulfate can drop out of a cold batch, putting your surfactant on the lead element instead of in solution.

Matching the chemical to the foulant, and the order

You can narrow down what's on the element without an autopsy. Where the pressure drop climbed tells you most of it: lead vessels load up with colloids, organics and biology, tail vessels are where scale drops out because the concentrate is most saturated there. Feed source narrows it again, well water pointing at iron and hardness scale, softened city water at organics and biofilm. Then pull the end caps at both ends of the rack and look. There's more on catching fouling earlier in a separate piece.

All three run high pH before low pH on mixed fouling; DuPont puts it plainly, recommending acid first only where you know the deposit is calcium carbonate or iron oxide and nothing else. Hydranautics writes the sequence out: flush with permeate, high pH CIP, flush until brine-side pH is below 8.5, low pH CIP, acid flush. That flush to 8.5 earns its step. Drop acid onto a caustic-wet element and you get local neutralization, a heat release, and precipitation of whatever the caustic just lifted off.

Acid first on an organically fouled element can do the opposite of cleaning it. The organics coagulate and set, and the alkaline wash that should have lifted them off then has to reach through a harder layer.

If you run reverse-direction cleaning, NanoH2O puts the forward clean first in every case and wants scale gone before any reverse flow.

Flow and pressure, where the chemistry stops mattering

Every cleaning flow rate in these documents is per pressure vessel. Not per element.

• DuPont: 30 to 45 gpm per pressure vessel on 8-inch elements, 35 to 45 gpm for fullfit and 440 ft² elements
• Hydranautics: 36 to 48 gpm per pressure tube for standard 8-inch, 40 to 53 gpm for LD elements on the 34 mil spacer, with inlet pressure not to exceed 60 psi
• NanoH2O: introduce at 20 gpm per 8-inch vessel, recirculate at 40 gpm for 45 minutes, and on vessels of five elements or more, never let vessel pressure drop exceed 60 psi during cleaning, or 40 psi in the reverse direction. Below five elements TSB 113 tells you to call NanoH2O rather than assume the 60 psi

Those two 60 psi figures are different quantities: Hydranautics caps the pressure you feed the tube, NanoH2O the differential across it. DuPont's ceiling is 50 psi per multi-element vessel, so a mixed rack takes the tightest of the three.

Read 40 gpm as a per-element number on a six-element vessel and you're asking the CIP pump for 240 gpm, past every pressure drop limit in the table before the caustic reaches temperature. That damage is mechanical: telescoping, crushed feed spacer, glue lines letting go.

All three want chlorine-free makeup and flush water. DuPont and Hydranautics accept permeate or DI; TSB 113 says RO permeate. Only Hydranautics puts numbers on the rest, with hardness under 30 ppm as CaCO3 and iron under 0.05 ppm in its Table 4. The BW 400 R G2 data sheet caps free chlorine below 0.1 ppm, and that's a ceiling on a material with no real tolerance for it rather than a dose to aim at. Mix a caustic batch with chlorinated city water and you've run an oxidation event and called it maintenance.

Cleaning on a number that was never a cleaning trigger

DuPont's cleaning manual spends a line heading off a specific misuse of its own figures. The 15 psi per element and 50 psi per vessel numbers are mechanical limits, and the manual says flatly that "these recommended pressure drops should NOT be used as a cleaning criteria." Its trigger is a 15% rise in pressure drop, and its normalized-data bullet puts the band at 10 to 15%. NanoH2O words the same trigger more carefully: normalized pressure drop up 15% since startup or the last cleaning. Normalized is the part to hold onto, because raw gauge differential moves with feed flow and with temperature, so two readings taken at different throughput aren't comparable until they've been corrected.

Run it on a real rack. A vessel whose normalized baseline is 18 psi is due at about 21. Waiting for 50 means nearly three times baseline, by which point the deposit has had months to consolidate, and that's the whole reason the trigger is set low. Early cleanings give flux back; late ones give a fraction back and leave the rest. Rejection is a poor trigger too, since it moves after flux and pressure drop have turned.

What goes on the skid door

Pick the element family first, then print that family's own pH and temperature table and tape it inside the skid door. The envelope doesn't travel between brands, and a mixed rack takes the tightest number in it. On the alkaline end FilmTec publishes the wider window, 35 C at pH 12 against NanoH2O's 25 C on a brackish element, so a plant that leans on warm caustic for heavy organic loading should write that into the element spec rather than find out at the next autopsy. The envelope is one input into that choice and not usually the deciding one; the performance side is a separate head-to-head on the BW 400 R G2 and the BW30 PRO-400. What standardizing on one family buys you is narrower: one table on the door instead of three, one set of numbers for whoever mixes the batch at 6am. Good part of why we keep NanoH2O 8040s on the shelf. The step-by-step CIP procedure is written up separately.

Enviro Membranes stocks genuine NanoH2O (formerly LG) RO elements in McAllen, Texas, with same-day US shipping on in-stock items. Prices are published, with 5% off at 5 to 9 units and 10% off at 10+ per model; see current models and prices at enviromembranes.com/shop. Send us your water analysis and we'll tell you which membrane fits. Reach us at enviromembranes.com.