Winston Water Engineering Version history
Design sheetsSheets Material compatibilityMaterials Ask a questionAsk SupportSupport Sign inSign in
Home  ›  Version history
Releases

What changed, and what comes next.

Every release of the design calculator, what went into it, and what is being weighed for the next one. Buy the workbook once and every version after it is free — your original download link serves the newest build, so there is nothing to repurchase and nothing to chase. And the file keeps outgrowing the site — future releases carry more unit operations and reference material than the free pages do.

Get the workbook SGD 19.90 · workbook + data sheets · every future version included
v1.7 Current
Released 14 October 2026
28Design sheets
2 109Live formulas
35Tabs
SIUnits
scroll for older releases
  1. v1.7
    14 October 2026
    • T-1200 chemical dosing. A dose is set in milligrams per litre of water and a metering pump is bought in litres per hour of solution. Between the two sit a strength and a specific gravity, and almost every dosing error in practice is one of them applied upside down — at 12 % that mistake is wrong by a factor of eight. The sheet carries the chain through once, in both L/h and the mL/min a pump dial is actually marked in.
    • The turndown check, which sizing calculations usually skip. A diaphragm metering pump controls down to roughly a tenth of its rated capacity. Chosen only on the peak dose it frequently cannot deliver the low-demand dose at all, and the first anyone knows of it is a residual that will not hold overnight. The sheet asks whether the solution flow at the minimum dose is still above the pump minimum, and puts a pass mark on the answer rather than a sentence of advice.
    • The day tank, on usable volume. The suction lip, the low-level cut-out and the ullage above high level are not available to the process, and on a small day tank they are a third of it. Sizing storage on the geometric volume is how a tank that should last a fortnight runs dry in nine days. Bund volume comes out with it.
    • T-1210 polymer preparation unit. Sized from the process duty rather than from the solution draw: dry powder demand, sacks a day, and the hopper refill interval an operator actually plans around. It also separates make-up strength from injection strength — the unit makes 0.25–0.5 % and the line dilutes to 0.05–0.1 % — which is the distinction behind most plants that believe they are using twice the polymer they should.
    • Maturation is not mixing time. Dry polymer disperses in seconds and hydrates in thirty to sixty minutes, and the ageing chamber exists for the second of those. Undersize it and the unit still makes solution at the right strength — it simply does not work, because the chains have not uncoiled, and the symptom is a dose that has to keep climbing.
    • Open on the website too. Chemical dosing has its own calculator page, free to subscribers alongside the design sheets. The file carries the same arithmetic with the design envelopes, the notes and a printed calculation record — and the two are checked against each other on every build, so the page and the file cannot drift apart.
  2. v1.6
    4 October 2026
    • MD-630 membrane degasifier. A hydrophobic hollow fibre holds the water back and lets the gas through: the pore stays gas-filled, so dissolved carbon dioxide crosses it and leaves without a bubble ever forming. No tower, no packing, and no air to re-saturate what you have just stripped. It is the unit that sits between RO and EDI on almost every ultrapure train, because the resin otherwise spends its current splitting water to deal with the CO2.
    • The contactors. Parallel trains come from the hydraulic rating per module, modules in series from a counter-current transfer-unit requirement — every decade of removal costs 2.3 of them, whatever the flow. Shell-side transfer goes as the square root of flow, so splitting into one more parallel train buys about 40 % more transfer units per module as well as a quarter of the pressure drop.
    • The vacuum pump is sized by steam, not by gas. The gas side leaves saturated with water vapour at the water temperature, and at 200 mbar and 25 °C that vapour is 31.7 mbar of the 200 — a sixth of the suction volume, and more than a quarter at 35 °C. A pump chosen on the kilograms of CO2 removed comes out roughly a third of the size it needs to be, so the sheet selects it on actual cubic metres an hour at the suction pressure and splits the load between vapour, sweep gas, in-leakage and the CO2 itself — which is under half a per cent of the volume.
    • Why oxygen needs sweep gas and carbon dioxide does not. Nothing can be stripped below equilibrium with the gas on the far side of the membrane. Carbon dioxide is four parts in ten thousand of air, so whatever leaks into the shell brings almost none with it and the floor sits near zero. Air is a fifth oxygen: the same leak sets an oxygen floor of tens of ppb however hard the pump works. The sheet reports that floor, and it is what turns “add nitrogen” from an optimisation into the only way to ppb.
    • Checked against a vendor sizing report. Six cases — two arrangements across 25, 30 and 35 °C — reproduced to the printed removal efficiency, and suction volume within half a per cent.
    • The contactor is free on this site as an animated unit operation at MD-630, with its own section on the design sheet page. The design calculation is in the workbook, which is where new unit operations have gone since v1.4.
  3. v1.5
    19 September 2026
    • CT-710 chlorine contact tank. Disinfection is not a residual, it is a residual held for a time — and the time that matters is T10, the time the fastest tenth of the water gets, not volume over flow. An unbaffled tank gives back about a tenth of what it was sized on. The sheet works CT from the baffling factor, checks it against the USEPA regression behind the SWTR tables — temperature, pH and residual all move it — and reports what the tank would need to be at each baffling class.
    • pH quietly halves your disinfectant. Only hypochlorous acid works properly; the hypochlorite ion is perhaps eighty times weaker, and the split sits on a pKa near 7.5. Above pH 8 most of a perfectly good measured residual is doing very little, and the CT you need roughly doubles between pH 7 and pH 9. The sheet shows the split directly.
    • D-1300 thermal drying, convective and conductive. It sizes the heating surface or the drying air depending on which machine you are buying, and puts the sticky phase between forty and sixty per cent dry solids where you can see it — that band, not the heat duty, is what decides the machine. It also compares the energy in the dried product against the fuel burned to make it. From undigested sludge the granulate can carry more energy out than the dryer put in, which is the whole argument for drying; from digested sludge it usually cannot, and the sheet takes the calorific value as an input so you can see which side of the line you are on.
    • D-1310 solar drying. The same water taken out by sunlight: about a thirtieth of the energy and perhaps a hundred times the area. It sizes the greenhouse from your irradiation, gives the retention time in weeks, and makes the point that a house sized on annual average irradiation works for eight months and overflows for the other four.
    • The sludge group is now dewatering and drying, with a new icon, and all three new unit operations are free on this site as animated cutaways — CT-710, D-1300 and D-1310. What ships in the workbook is the design calculation, which is where new unit operations have gone since v1.4.
    • The design sheet page now lists the workbook-only sheets in their proper place rather than leaving them out — you can see what exists and where the sizing lives.
  4. v1.4
    12 September 2026
    • CM-560 ceramic membrane filtration, in the workbook. Someone pointed out that a site covering microfiltration through to reverse osmosis had nothing on ceramic membranes, and they were right. It is a separate sheet rather than another format on MF-570, because the design driver is genuinely different.
    • The loop is the plant. A hollow-fibre plant is essentially dead-end and the feed pump is the whole energy bill. Ceramic is crossflow: the recirculation that keeps the channels scoured carries thirty to sixty times the permeate flow, and on a typical duty it takes around ninety per cent of the connected power. The sheet puts that in front of you instead of leaving it in a vendor’s quotation.
    • Two dimensions give you the loop. The channel bore area falls out of the membrane area as d/4L — whatever the channel count in the monolith — so the recirculation flow follows from the channel diameter and the element length alone.
    • Reynolds number, not velocity. Scouring is a turbulence effect, and a wide channel at a respectable-looking velocity can quietly drop into laminar flow. The sheet flags it, because when it happens the cake stops being swept off whatever the TMP does.
    • This one is not on the site, and that is the point from here on. The free pages stay free and nothing has been taken off them — but new unit operations now go into the workbook and stay there. At this release the site carried twenty-one design sheets and the file twenty-two — a gap the subscription has since closed, and every sheet is now open to subscribers here. Ceramic already appears on the site as a module format on MF-570 and UF-580; what the workbook adds is the sizing.
    • System recovery on MF-570 and UF-580 is now a water balance, not a flux ratio. Both sheets used to report net flux divided by gross flux. That figure is recovery multiplied by the fraction of the day spent filtering, so it charged the plant for the minutes lost to backwashing as though they were water lost — about 1.8 points low. Recovery is now net permeate over raw water, counting backwash and CEB and nothing else. Net flux is still there beside it, which is where the throughput question belongs. MF-570 at its defaults reads 94.9 % instead of 93.6 %.
    • Both sheets now report net permeate delivered. The daily mean, which is lower than the permeate rate while the plant is filtering — 111.8 against 120 m³/h on MF-570 — because the wash water and the idle minutes both come out of it. That is the number a plant gets judged on, and it is worth seeing beside the rate you sized for. MF-570 also gains CEB interval and CEB volume inputs, so the workbook now models the chemically enhanced backwash the site page always did.
    • The equipment data sheets and the free preview pick the new sheet up automatically — the data sheet file is now 22 forms across 24 tabs.
  5. v1.3
    6 September 2026
    • Two new design sheets, and a new group. D-1200 decanter centrifuge and D-1210 recessed plate filter press — the first sheets on the site for what happens to the solids after everything upstream has separated them. Both have their own animated unit operation, so at this release the site carried twenty-one of each.
    • The decanter is sized on sigma, not on volume. Ambler’s equivalent settling area, and the cut point d₅₀ that falls out of it — the particle size below which solids leave with the centrate. Comparing that cut point against the floc the polymer has actually built is the clarification check, and it is why an under-dosed machine cannot be rescued with more G-force.
    • Residence time is the limit nobody quotes. Everyone quotes G-force. Below about twenty seconds in the pool, solids never reach the bowl wall however hard it spins, so the sheet reports the hydraulic limit that criterion sets and flags the feed against it.
    • The press is sized on the cake filtration equation. Constant-pressure filtration with a compressible cake, so the specific resistance rises with the pressure you apply — which is why doubling the pump pressure never halves the cycle. The sheet splits the cycle into the cake term, the cloth term and the handling time, and tells you which one you are bound by.
    • The preview now carries two working design sheets instead of one. H-100 and the new D-1200, both fully live, with every count in the file derived rather than typed.
    • The dashboard no longer assumes a flow in m³/h. A filter press is a batch machine quoted per day; the dashboard now normalises it instead of breaking on it.
  6. v1.2
    5 September 2026
    • F-600 is a new design sheet — oxidation and filtration for well water. It was already on the site as an animation with no sheet behind it; now it has both, and the workbook has a nineteenth design sheet.
    • Oxidant demand comes from the balanced reactions, not a rule of thumb. Chlorine at 0.635 mg per mg of iron and 1.291 per mg of manganese, permanganate at 0.943 and 1.918, dissolved oxygen at 0.143 and 0.291. The sheet reports it as the theoretical minimum, because chlorine also has to meet the ammonia and organic demand of the water.
    • The run length is set by the solids, not the metals. Every mg/L of iron becomes 1.913 mg/L of ferric hydroxide and every mg/L of manganese 1.582 of manganese dioxide, and that is what blinds the bed.
    • The governing row is the manganese pH margin. Iron oxidises almost anywhere from pH 6.2; manganese will not move until the pH clears the medium’s threshold — 7.0–7.5 on a catalytic surface, 8.5–9.5 without one — and no amount of contact time, oxidant or bed depth substitutes for it. Below the threshold the sheet says so and the manganese passes straight through.
    • Media are typed in rather than picked from a list. Loading rate, bulk density, backwash rate and the manganese pH threshold are all inputs, with greensand, GreensandPlus, pyrolusite, Birm and plain aeration given in the notes — so a medium the list would never have carried can still be costed.
    • Each filter medium is now drawn as the plant it actually is. Greensand and pyrolusite get a chemical dosing pot; Birm gets a venturi aspirator, because it runs on dissolved oxygen and has no dosing equipment at all; aeration gets a packed tower with a blower, and the dual-media bed below it is drawn as anthracite over sand. Choosing a medium used to change every number and none of the picture.
    • Fixed: the oxidant dose label had its line drawn through it, which split the figure from its unit and made it read as a dose with no units. It now sits clear, and names the reagent — mg/L of permanganate is not interchangeable with mg/L of chlorine.
    • Fixed: a dashboard bar divided by zero. A governing result with a lower bound of zero and no upper bound scaled its bar to nothing, and the tab showed #DIV/0!. The scale can no longer be zero, whatever envelope a sheet declares.
  7. v1.1
    2 September 2026
    • RO-600 takes a concentrate recycle flow returned to the high-pressure pump suction. Enter it and the sheet reports the system recovery, the make-up actually drawn from source, the concentrate going to drain, and the blended feed TDS the array ends up seeing.
    • The recovery input is now "recovery per pass", which is what the array does in one pass. With recycle the system recovery sits above it; without recycle the two are equal.
    • The recycle is capped where the salt balance stops closing. Return too much and the loop concentration runs away, so the sheet limits it and says why.
    • Set the recycle flow to zero and every number matches v1.0 exactly.
    • Three drawings were wrong and are now right. The DAF recycle pump was drawn facing against its own flow; the pump sheet drew its suction from the liquid surface rather than from the tank outlet, which is the arrangement that destroys NPSH; and the RO array stretched the drawing far past the vessels it actually shows once the duty flow was raised.
    • Axis units on every chart. The pump and system curves, the transmembrane pressure cycle and the carbon breakthrough curve were plotted against unlabelled axes.
    • Labels no longer sit on top of lines. Nine sheets had a label crossed by a pipe, a dimension or a plotted curve. Every figure is now checked for it before release.
    • The impeller diameter on the mixing sheet was rendering as a currency symbol rather than a subscript. Corrected, along with one other symbol on the lamella sheet.
  8. v1.0
    1 September 2026
    • Eighteen unit-operation design sheets — pipework and pumps, four routes to solids separation, media and carbon filtration, ion exchange, MF through RO, electrodeionisation, UV, heat transfer, mixing and neutralisation.
    • Every equation printed beside its result, in the symbols used on that sheet. Nothing is hidden, so the working can be checked rather than trusted.
    • Design envelope on every result — PASS, CHECK or FAIL against the band industry practice accepts, so a number out of range announces itself.
    • Design Basis — set plant flow, temperature and tariff once and every sheet follows, with a per-sheet override for when one unit runs at a different duty.
    • Dashboard — all eighteen units on one page, live, each with its governing figures and a status.
    • Live readout on every design sheet: a scaled bar for each checked result that moves as the inputs change.
    • Schematic on every sheet, showing the arrangement and the symbols.
    • Summary, equipment list, calculation record, and a pump-and-system curve charted from your own duty point.

How an update reaches you

When a new version is published, the download link in your original receipt starts serving it. The link does not expire and does not change, so there is no new purchase, no new link, and nothing to ask me for. If a release is worth knowing about, you also get an email.

The version number and its date are on the Start Here tab of the workbook itself, with the same history you are reading here — so you can always tell which build you have open.

Included with the workbook

Equipment data sheets. A second Excel file, delivered with the workbook and covered by the same one-time price: one vendor enquiry form per unit operation, 28 in all. Each carries the same process parameters and the same design envelopes as the matching calculation sheet, then the rest of what a supplier needs to quote — materials, nozzle schedule, driver and utilities, instrumentation, codes and inspection, vendor scope. The calculator tells you how big it has to be; the data sheet is what you send out once you know.

The file and the site now carry the same sheets. Until recently some unit operations lived only in the workbook. Since the subscription opened, a subscriber gets all 28 of them in the browser too, and the workbook is the Excel copy of the same work — offline, yours, and fit for a design submission. From v1.4 a new unit operation goes into the workbook and stays there. The site has twenty-one design sheets; the file has twenty-two, and the gap widens with every release. Buying once covers all of it, including every version still to come.

Being weighed for a future release

Nothing here is promised, and the order is not fixed. What people actually ask for moves up.

  • More unit operations — sludge dewatering, degasification, chlorine contact tanks, and anything else that comes up often enough to earn a sheet.
  • Sensitivity tables — vary one input across a range and see the governing result move, without editing the sheet.
  • Chart-drawn schematics that redraw as inputs change. Built and evaluated; shelved for now because the drawing quality was worse than the current figures.
  • Costing — deliberately last. Wrong costs damage trust more than absent ones.
© 2026 Winston Water Engineering. All rights reserved.The equations themselves belong to the engineering literature and to anyone who needs them. The sheets, calculators, drawings and workbook that present them here do not — please link to a page rather than republish it.Privacy