Brackish Water Reverse Osmosis: Why Feed Chemistry Sets Your Recovery Limit

A contractor sizes a system for a brackish well running 4,200 mg/L TDS. The datasheet says 99.5% rejection, so he designs for 80% recovery to shrink the reject stream and keep the customer happy about their water bill. Four months later the tail-end elements are scaled solid, differential pressure is climbing, and permeate quality is drifting up. Nothing was wrong with the membrane. The recovery target was set from a salt number when it should have been set from the feed's calcium, sulfate, and silica.

On seawater, salt concentration and osmotic pressure dominate the design. On brackish water — roughly 1,000 to 10,000 mg/L TDS — the salt almost never limits you. What limits you is what precipitates out of the concentrate as recovery climbs. Get it wrong and you are cleaning monthly and replacing elements a year early.

How brackish systems actually behave, what the spec sheets are and are not telling you, and where the recovery ceiling really comes from.

What "Brackish" Actually Means for Design

The water industry draws the lines by total dissolved solids: fresh water is under 1,000 mg/L, brackish runs from about 1,000 to 10,000 mg/L (some references stretch it to 12,000), and seawater sits around 35,000 mg/L. Those numbers are not trivia — they decide which membrane you buy and how hard your pump has to work.

Brackish elements are built for the pressure range brackish feed needs. Most commercial brackish RO runs between 120 and 300 psi, versus 800 to 1,200 psi for seawater membranes. Energy scales with that pressure, which is why you do not want to over-spec. But at the top of the brackish band — say 8,000 to 12,000 mg/L — you have a real decision to make: push a high-rejection brackish element harder at elevated pressure, or step up to seawater elements that are designed for it. That call comes down to feed TDS, temperature, and your energy cost, not a rule of thumb.

Reading the Datasheet: Test Conditions vs. Your Well

Most major manufacturers test standard brackish elements at approximately the same conditions: 2,000 ppm NaCl, around 225 psi (150 psi for low-energy elements), 77°F, pH 7–8, and 15% recovery. Conditions vary by manufacturer and product line — the table below reflects the major differences — but the principles are consistent across brands.

Those rejection figures are stabilized single-salt numbers measured at 15% recovery and 77°F. Your feed is a mix of ions, your groundwater is probably colder or warmer than 25°C, and you are running 75 to 80% recovery. Real-world rejection comes in lower and permeate TDS higher than the sticker. Temperature alone moves flux noticeably: permeate flow shifts by roughly 2 to 3% per °C, so a well that drops to 15°C in winter can lose a meaningful share of its output unless you designed for it. If you are weighing specific elements against each other, the trade-offs are laid out in our side-by-side comparison of FilmTec, Hydranautics, and Toray.

Recovery Is a Chemistry Problem, Not a Salt Problem

Every gallon of permeate you pull leaves its dissolved load behind in a shrinking concentrate stream. The concentration factor is 1 divided by (1 minus recovery): at 75% recovery everything left behind is 4× more concentrated, at 80% it is 5×, at 85% it is approximately 6.7×. Salt does this and so does every scale-forming ion. Your recovery ceiling is whichever scalant hits saturation in the concentrate first.

Calcium carbonate. The most common scale on groundwater brackish. It forms when alkalinity and hardness concentrate past solubility, and it is driven by pH above roughly 7.5 to 8.0. You model it with the Langelier Saturation Index at the concentrate end and control it with acid dosing, antiscalant, or upstream softening.

Sulfate scales. Calcium sulfate is a problem on feeds high in both calcium and sulfate, and unlike carbonate it does not care about pH — you cannot acid your way out of it. Barium and strontium sulfate are worse: they hit saturation at trace concentrations and are among the most stubborn scales to clean, often irreversibly.

Silica. The one people forget. For typical designs, amorphous silica becomes a scaling concern as concentrate silica approaches roughly 120 to 150 mg/L as SiO₂ at neutral pH and ambient temperature — though the real limit depends on pH, temperature, and antiscalant chemistry. High pH and high temperature speed silica polymerization into a gel that is miserable to remove. On a lot of brackish groundwater, including much of the South Texas water chemistry we deal with, silica — not hardness — is what actually caps recovery.

Your recovery setpoint should come from a saturation model of the concentrate stream that accounts for carbonate, sulfate, barium, and silica together. It should not come from the feed TDS number or a generic "brackish does 80%."

What it costs (approximate, small-quantity 2020s pricing):

• 8-inch high-rejection brackish elements: roughly $350–$700 each; low-energy types often $400–$750. Prices vary by brand, volume, and region.

• Small commercial skid (3–20 gpm): about $15,000–$60,000. Mid-size industrial with full pretreatment: roughly $150,000–$800,000.

• Membrane elements are usually only 5–10% of skid CAPEX. Pumps, pretreatment, and cleaning frequency drive the real cost of ownership.

The Scale That Gets Misdiagnosed as Fouling

A system starts losing normalized permeate flow while differential pressure across the array creeps up. That is the textbook fouling signature, so the tech runs a high-pH alkaline clean-in-place — the right move for organics and biofilm — sees a small improvement, and schedules another CIP for next month. The cleanings get more frequent, the elements never fully recover, and a set that should have lasted five years gets pulled at eighteen months.

The tell is where the problem lives. Feed-end (lead) elements foul, because that is where colloids, organics, and biofilm land first. Tail-end elements scale, because the concentrate is most saturated by the time it reaches the last element in the last vessel. Pull and weigh the elements, or read the vessel probe profile: if the lead elements are relatively clean and the tail element is heavy and crusted, you do not have a pretreatment fouling problem — you have a recovery and scaling problem. Alkaline CIP will never fix that. The fix is lowering recovery, correcting antiscalant or acid dosing, or in the case of sulfate scale, accepting that you set recovery too high for that feed. Catching it early is exactly what a normalized performance report is for: normalized flow and salt passage drift before the pressure gauges tell you anything.

Design Decisions That Matter

Set recovery from a concentrate saturation model — LSI for carbonate, plus sulfate, barium, and silica — not from the TDS number or a generic percentage. Choose high-rejection elements when your permeate spec is tight, such as boiler feed or a high-purity process, and low-energy elements when the spec has room and energy cost matters. Before you reach for the CIP cart, figure out whether you are looking at lead-end fouling or tail-end scale, because they read identically on a system gauge and need opposite fixes. And treat the pretreatment targets — SDI at or below 5, turbidity under 1 NTU, free chlorine below 0.1 ppm — as warranty conditions, not suggestions.

Enviro Membranes stocks 8-inch brackish elements from NanoH2O (formerly LG Chem) and DuPont FilmTec, both high-rejection and low-energy, with same-day US shipping out of McAllen. If you are sizing a brackish system or fighting recurring scale on South Texas well water, reach the team at enviromembranes.com or the McAllen office and we will help match the element and the recovery target to your actual feed chemistry. If you are on municipal supply, start with your city's verified water profile — and if you are in oilfield country, Midland's profile shows exactly what Permian Basin municipal water looks like on paper.