What your membrane actually produces at your water temperature and pressure, and how long the DI resin lasts.
| Feed temperature | Correction | Output from a 75 GPD membrane |
|---|---|---|
| 10 °C, 50 °F | 0.584 | 48.5 GPD |
| 15 °C, 59 °F | 0.703 | 58.4 GPD |
| 20 °C, 68 °F | 0.841 | 69.8 GPD |
| 25 °C, 77 °F | 1.000 | 83.0 GPD |
| 30 °C, 86 °F | 1.157 | 96.0 GPD |
Membrane ratings assume 77 °F feed water at 60 psi. Real tap water in winter runs 10 to 15 °C, which cuts output by 30 to 40 percent on its own. A membrane producing half its rating in January is working correctly.
TCF = exp(3020 × (1/298 − 1/(273+15))) = 0.7034 · output = 75 × 0.703 × 1.106 = 58.4 GPD
Membrane GPD ratings are measured under standard test conditions, 77 °F feed water at 60 psi with low TDS. Real conditions are almost always worse, so the rating is a ceiling rather than an expectation.
Temperature is the largest factor and the one people miss. Cold water is more viscous and passes the membrane more slowly. DuPont publish the correction as TCF = exp[2640 × (1/298 − 1/(273+T))] for water at or above 25 °C and the same form with 3020 below it. At 10 °C that gives 0.584, so a 75 GPD membrane makes 44 GPD before anything else is counted.
Pressure drives the rest. What matters is net driving pressure, the applied pressure minus the osmotic pressure of the feed, which runs about 0.0117 psi per ppm of TDS. Below 50 psi both output and rejection fall off, and a booster pump usually pays for itself in DI resin alone.
Rejection is what fraction of dissolved solids the membrane refuses. Standard thin film membranes are specified at 98 percent, with 96 percent as the minimum before replacement. Feed water at 250 ppm through a 98 percent membrane leaves 5 ppm for the DI resin to handle.
DI resin capacity follows from that. A litre of mixed bed resin holds roughly 556 grains as calcium carbonate, and resin life in gallons is capacity divided by the TDS reaching it. Halving the TDS out of the membrane doubles the resin life, which is the argument for replacing a tired membrane promptly.
Carbon dioxide passes straight through an RO membrane as a dissolved gas and then becomes carbonic acid on the resin, using it up while the TDS meter reads zero. Resin that dies unusually fast on low TDS water usually means CO₂.
Waste water is not a fault. The concentrate carries away what the membrane rejects, and restricting it to improve the ratio fouls the membrane. Four parts waste to one part product is normal, and a lower ratio shortens membrane life.
Check TDS both after the membrane and after the DI. The first tells you the membrane's condition, the second tells you when to change resin. A single inline meter after the DI hides a failing membrane until the resin is gone.
Sources: DuPont FilmTec technical manual