The thin-film-composite RO membrane rejection math the tool uses, the temperature-correction-factor curve that turns nominal GPD into temperature-corrected GPH, the pressure-correction factor for inlet conditions below the 60 psi reference, the deionization-resin polish thresholds that separate residential-acceptable from commercial-spec pole water, the verdict-precedence ladder that names the failing stage rather than the failing reading, and the boundary cases where the diagnostic catches the mechanical failure that a one-meter reading would have written off as a chemistry problem. The Bluegrass commercial-route decision frame, ported to software.
What the Pure-Water TDS Diagnostic does, in five points:
The tool exists so the operator does not have to carry the TFC membrane performance curve and the TCF temperature chart and the DI resin capacity tables in their head at the truck at the second account of the morning, on a commercial property where the meter reads eighteen ppm at the pole and the resin canister was replaced six weeks ago. It does the math, and it returns the trade decision in a form the operator can act on in the time between cracking the meter case and choosing which component goes back to the shop.
There was a Tuesday in October of about year four of running the Louisville commercial book when the rig started reading fourteen ppm at the pole at the second account of the morning. The resin canister had been replaced six weeks earlier — well inside the service interval for our feed-water hardness — and my first call was that the canister was a bad lot, which happens occasionally and which the supplier will swap if you call within ninety days. I called, I got the swap, I installed the new canister, and the rig read seven ppm at the pole for two routes and then climbed back to fifteen ppm by the end of week three.
The mistake was diagnostic, not mechanical. The resin canister was not failing because it was a bad lot. It was failing because the RO membrane in front of it had been slowly degrading for six months and was no longer doing the bulk-removal work it was sized for. Every canister we put in was burning through its capacity in three weeks instead of the four-to-six months we had calibrated against, because each canister was being asked to polish a much heavier dissolved-solids load than the rig's stack was designed to deliver. We spent roughly six hundred dollars in resin canisters chasing a two-hundred-dollar membrane problem, and the customer on the second account spent two visits with visible spotting on the dark-glass storefront before we caught the actual failure mode.
The Pure-Water TDS Diagnostic is the tool we built to put the three TDS readings — source, post-RO, post-DI — in front of the operator at the truck before the wrong component goes back to the shop. It does not tell you whether the rig is broken. What it does is convert the readings the operator can take with a handheld meter in roughly two minutes at the curb into the trade decision the operator actually has to make: which component has failed or degraded, and what intervention does the failure name?
This piece is the methodology behind that decision. The thin-film-composite membrane rejection math the tool uses, the temperature-correction-factor curve that turns nominal GPD into operating GPH, the pressure-correction adjustment for inlet conditions below the lab reference, the deionization-resin polish thresholds that separate residential-acceptable from commercial-spec pole water, the verdict-precedence ladder that names the failing stage rather than the failing reading, and the boundary cases where the diagnostic catches the mechanical failure that a one-meter reading would have written off as a chemistry problem. The companion encyclopedia piece is Jan Davenport's pure-water-systems reference, which covers the equipment side of the RO/DI stack in long-form prose. This piece is the methodology behind the tool's component-by-component diagnosis.
The resin reading is downstream of the membrane reading, both literally and diagnostically. This is the single most important rule the tool puts in front of the operator, and it is the rule the shop that wrote this piece learned the hard way over the year-four episode described above. The post-DI reading at the pole is the number that matters for the glass — it is what determines whether the dissolved solids on the panel will dry visible or invisible. But the post-DI reading is not the diagnostic. The diagnostic is the relationship between the post-DI reading and the post-RO reading, because a high post-DI reading can be caused by a failing resin canister, a failing membrane, a clogged pre-filter, or a mechanical contamination event — and the remediation for each is different.
An operator who reads only the post-DI number and replaces only the resin canister will get the rig back to spec in some cases and will burn through resin canisters chasing a different failure in others. The diagnostic that catches it is the one this tool runs: read the rejection number first, the resin number second.
The water-fed-pole rig produces three TDS readings at three points in the stack, and each reading is the output of one stage of the purification process. The tool reads all three and uses the relationship between them to localize the failure:
Source TDS (inlet). The feed-water profile, measured at the inlet to the pre-filter stack. For a Louisville municipal supply this typically runs 120-180 ppm; for the Bluegrass horse-country karst-aquifer wells it runs 200-280 ppm; for the harder limestone-karst wells in Bourbon and Anderson counties it can reach 350-400 ppm. The source reading is the baseline against which the rest of the math runs. It is not a diagnostic in itself — the source is what it is, and the rig has to deal with whatever the feed water delivers — but it is the calibration input for the rejection ratio.
Post-RO TDS. Measured between the RO membrane housing and the deionization canister. The bulk of the dissolved-solids removal happens at the membrane stage; a healthy TFC residential membrane fed 220 ppm source water at reference conditions should produce roughly 5-10 ppm post-RO. This is the reading that tells you whether the membrane is doing the work it was sized for.
Post-DI TDS (at pole). Measured at the working end of the pole, after the DI resin has polished the post-RO water. This is the reading that matters for the glass — the trade standard for residential pole work is ≤ 10 ppm, and the standard for high-end commercial is ≤ 1 ppm. The post-DI reading is the customer-facing output. But the diagnostic is the relationship between the post-DI reading and the post-RO reading, not the post-DI reading alone.
The pattern across the three readings is what names the failing stage. High source → high post-RO → low post-DI is a healthy rig on hard feed water. Low source → low post-RO → high post-DI is a resin failure. Moderate source → moderate post-RO → low post-DI with a low rejection ratio is a membrane failure that the resin has so far been able to mask. Each pattern has a different remediation, and reading any one number in isolation cannot distinguish them.
The dominant diagnostic for membrane condition is the rejection ratio:
rejection = 1 − (TDS_post-RO / TDS_inlet)
A healthy thin-film-composite residential membrane rejects 95-98 percent of dissolved solids at the lab-reference conditions of 60 psi inlet, 77°F feed water, 500 ppm TDS, and 15 percent recovery. The Filmtec residential TFC literature quotes 96-98 percent rejection on the 100 and 150 GPD residential lines. The Dow and GE/Suez residential bulletins give 95-97 percent on equivalent constructions. The trade replacement threshold is 88 percent — below that, the membrane is no longer doing the bulk-removal work and the resin canister is being asked to absorb a load it was not sized for.
The 88-percent threshold is not arbitrary. It corresponds to roughly twice the post-RO TDS that a healthy membrane would produce on the same feed water. If a healthy membrane on 220 ppm feed produces 7 ppm post-RO (96.8 percent rejection), a failing membrane at the 88-percent threshold produces 26 ppm post-RO on the same feed — almost four times the dissolved-solids load reaching the resin canister. At that rate of throughput, a resin canister rated for six months of service will exhaust in roughly six weeks. The shop that does not catch the membrane failure will spend the difference in resin canisters before the failure becomes visible at the post-DI reading.
The MEMBRANE-FATIGUE band, between 88 and 95 percent rejection, is the scheduled-replacement situation. The rig still produces acceptable pole water for routine residential work, but the margin is shrinking and the resin is absorbing more than its sized load. The right move is a scheduled membrane replacement in the next regular-service window — typically within four weeks, before the readings drift into immediate-replacement territory. The tool flags MEMBRANE-FATIGUE specifically so the shop can plan the replacement rather than scramble to it after a customer call.
The second piece of math the tool needs is the temperature correction for membrane productivity. Manufacturer GPD ratings are quoted at 77°F feed water, and real-world inlet temperatures almost never match. A Louisville municipal supply in February runs 48-52°F at the meter; a Bluegrass well in October runs 58-62°F; a residential garage rig with the bladder tank in unconditioned space in winter can be feeding 42°F water to the membrane housing.
TFC membrane productivity drops with feed temperature because the viscosity of cold water is higher, the diffusion rate of water through the membrane is lower, and the net driving pressure relative to osmotic pressure is reduced. The drop is not linear — it is roughly 1.2 percent per °F below 77°F at the top of the curve and steepens to 2 percent per °F at the colder end. The tool's piecewise TCF approximation follows the manufacturer published curves:
A 150 GPD membrane fed 50°F well water produces roughly 60 percent of its nominal GPH — 3.75 GPH against a nominal 6.25 GPH — and a measured production of 3.5 GPH at those conditions is healthy, not failed. The TCF is the input the tool uses to distinguish a cold-feed production reduction from a membrane-condition production reduction. Without it, every winter route in the Ohio Valley would read as MEMBRANE-FATIGUE.
The third math input is the pressure correction. RO membrane productivity scales roughly linearly with the net driving pressure above the osmotic-pressure floor of the feed water. For residential-spec inlet water under 500 ppm TDS, the osmotic-pressure floor is roughly 5 psi, so the productivity correction is:
PCF = (inlet pressure − 5 psi) / (60 psi − 5 psi)
The 60 psi reference is the standard residential-RO inlet condition. Real-world rigs typically run 45-65 psi from a municipal supply with no booster pump, 55-75 psi with a permeate pump, and 65-85 psi with a dedicated pressure-boost pump. Below 50 psi the membrane is operating near the lower edge of its productive range; below 40 psi it is below useful operation and the rig will produce slow output even with a healthy membrane.
The pressure correction is what the tool uses to distinguish a low-pressure production reduction from a clogged-pre-filter or membrane-failure production reduction. A rig fed at 45 psi will produce roughly 73 percent of its temperature-corrected nominal GPH; if the operator reads 4.5 GPH against a 6.25 GPH nominal at 77°F feed, the rig is healthy at those pressure conditions, not membrane-fatigued.
The fourth math input is the resin polish standard. A healthy DI resin canister produces near-zero post-DI TDS regardless of post-RO input — the resin's job is to remove the last fraction of dissolved solids that the membrane could not, and a 2-cubic-foot mixed-bed nuclear-grade canister has enough capacity to polish hundreds of gallons of post-RO water to commercial-spec quality before exhausting.
The trade thresholds the tool uses:
Commercial-spec polish (≤ 1 ppm): The standard for high-end commercial work — coated-glass office buildings, hospitality, architectural-grade residential. Below this threshold the dissolved solids on the panel are below the visibility limit on full-sun and dark-glass surfaces, and the squeegee-and-detail step becomes truly optional.
Residential-acceptable polish (≤ 10 ppm): The standard for routine residential pole work. Between 1 and 10 ppm the dissolved solids are visible under specific lighting on specific glass types (dark-tinted automotive glass, low-iron coated panels at the right sun angle) but invisible on the routine residential mix. The IWCA water-fed-pole technique reference uses 10 ppm as the residential threshold, and the tool follows that convention.
Spotting-likely band (10-20 ppm): The post-DI TDS at which the dissolved solids on the panel begin to dry visibly on the routine residential mix. The customer will see it within hours on dark-glass storefronts and within days on residential.
Resin-exhausted band (> 20 ppm): The canister is no longer doing useful polish work. The rig is effectively running on the post-RO water only, and the customer will notice on the first route.
The 10 ppm residential threshold is the trigger for the RESIN-SPENT verdict, paired with a healthy membrane. If the rejection ratio is below 88 percent and the post-DI is above 10 ppm, the verdict is MEMBRANE-FAILED rather than RESIN-SPENT — the resin is responding to a membrane failure, not to its own exhaustion.
The fifth piece of the diagnostic ladder is the LINE-CONTAMINATION override, which fires above everything else when the post-DI reading exceeds the post-RO reading. This is not a component failure — it is a mechanical or instrumentation problem, and no component replacement will fix it.
A DI resin canister is a polishing stage. The chemistry is straightforward: the resin's ion-exchange sites bind to dissolved cations and anions and replace them with H⁺ and OH⁻ that combine into water. The resin can be exhausted (the binding sites are saturated, no more ion exchange, post-DI ≈ post-RO), but it cannot add dissolved solids to water passing through it. A post-DI reading higher than the post-RO reading is physically impossible if the system is intact.
The reading indicates one of three things:
A cross-threaded fitting or O-ring leak downstream of the resin canister. Air ingress or contamination from the fitting environment is being introduced between the canister outlet and the meter probe. Inspect every quick-connect, every gasket, every threaded fitting from the canister outlet to the pole hose.
A cracked canister housing or compromised seal. The resin canister itself is leaking around the bypass, allowing some fraction of the pre-resin (post-RO) water to mix with the post-resin water and skewing the reading. Inspect the canister O-rings for compression marks or grit, and check the housing for hairline cracks (most common on freeze-thaw-stressed canisters that wintered in unconditioned space).
A meter probe fault. The handheld TDS meter probe has been dropped, fouled, or had its conductivity cell contaminated. Rinse the probe with the pole-water output, dry it, and re-test. If the reading persists with a clean probe, the contamination is downstream and not at the meter.
The LINE-CONTAMINATION verdict is precedence-first because the most expensive mistake on this reading is to replace the resin or the membrane in response to it. Both components are healthy and the problem is mechanical, not chemical. The fix is usually a fifty-cent O-ring or a re-tightened fitting; the wrong call is a two-hundred-dollar resin canister thrown at a fitting that needed a quarter-turn.
The tool's verdicts are ordered by precedence rather than by severity. The order matters because some failures mask others, and the wrong remediation order will leave the rig out of service longer than necessary:
The precedence is what makes the diagnostic useful. An operator who reads only the post-DI number and replaces only the resin canister will be correct in maybe half the cases and will burn budget on the other half. The diagnostic that catches all six failure modes is the one that reads all three TDS numbers, applies the temperature and pressure corrections, and tests them against the verdict ladder in precedence order.
The diagnosis-and-remediation paragraphs the tool surfaces are not generic recommendations. Each is keyed to a specific verdict and names the specific intervention. The order they appear matches the order an operator would consider them at the truck:
The tool is calibrated against Bluegrass and Ohio Valley commercial-route data, with rejection thresholds anchored to Filmtec, Dow, and GE/Suez residential TFC membrane bulletins, temperature correction curves anchored to published manufacturer TCF charts, and post-DI thresholds anchored to the IWCA water-fed-pole technique reference. The calibration holds well for the 100-300 GPD residential and light-commercial membrane range with standard 2-cubic-foot mixed-bed nuclear-grade DI canisters; it is conservative at the boundary cases.
The math assumes a properly-sized booster pump and a stable inlet pressure. Rigs running on raw municipal pressure without a booster pump will see more variation in production than the tool's PCF accounts for, and the diagnostic should be re-run at multiple inlet-pressure readings across a route before any membrane verdict is committed.
The math does not handle nitrate or organic-contaminant breakthrough, which are separate diagnostic categories and not captured by a TDS-only reading. A rig with a healthy TDS profile but a visible-residue problem on the glass may have an organic-contaminant breakthrough that requires the carbon-pre-filter swap regardless of the TDS reading. The tool's PRE-FILTER-CLOGGED verdict catches the flow-rate side of pre-filter exhaustion but does not catch the carbon-chemistry-exhaustion side; both should be on the operator's maintenance schedule regardless.
The math also does not handle silicate or boron breakthrough at very high source-water TDS, which can produce a high post-RO reading even on a healthy membrane. Rigs operating on source water above 600 ppm should add a silicate-specific test to the diagnostic routine, since silicate is a known TFC membrane bypass species.
The diagnostic is the relationship between the readings, not any one reading in isolation. The tool exists so the operator can see that relationship at the truck before the wrong component goes back to the shop. The shop that wrote this piece learned the lesson the expensive way — six hundred dollars in resin canisters thrown at a two-hundred-dollar membrane problem — and the diagnostic in front of you is the one we have run since the year-four episode. Read the rejection first, the resin second, the pressure third, the temperature fourth. The math will tell you which component to replace, and the math will tell you what not to throw money at.
Wade Marler runs a four-truck residential and commercial window-cleaning operation out of Louisville, Kentucky, with a twenty-one-year working book across the Louisville Ohio River corridor, the Northern Kentucky exurban commercial route, and the Bluegrass horse-country karst-aquifer residential market. He covers the Kentucky and Ohio Valley beat for this site, with a particular focus on commercial route operations, water-fed-pole rig diagnostics, and the production-cleaning calendar discipline that Derby Week imposes on the Louisville-area trade.