Three documents, two unit systems, one van
The product came in on a pallet at half seven. It is a two-part cementitious tanking slurry, the pails are labelled in kilograms, and the printed sheet folded into the box gives a minimum substrate temperature in Celsius, a water addition in litres per unit, and a soundness test pressure for the finished tanking in kilopascals. Everything in the van reads the other way: a dial thermometer in Fahrenheit, a test pump with a nought-to-three-hundred psi gauge on it, and a bathroom scale a previous job left in the site office. The client's inspector arrives at eleven and will want the substrate temperature recorded and the test held.
Nothing here is unusual or difficult. What makes it go wrong is that it looks like three multiplications, so it gets done in somebody's head on the tailgate, and by the time the inspector asks how a figure was arrived at there is no record of which number was original and which derived. The failure is almost never the arithmetic. It is the loss of provenance — a converted figure on a ticket in the same handwriting as a measured one, with nothing to say which was which.
There is a document problem underneath the unit problem too. The safety data sheet in the box has a format that is legally fixed: sixteen sections in a mandated order, physical and chemical properties in Section 9, required in the United States by OSHA's Hazard Communication Standard at 29 CFR 1910.1200 Appendix D and in Europe by Annex II of the REACH Regulation, both tracking the UN's Globally Harmonized System. The technical data sheet has no such format. Nobody governs its layout, its units, or whether this revision uses the same ones the last did — which is why the application temperature sits on the unregulated document and the flash point on the regulated one, and why people so often read the wrong page.
Sort the sheet before you convert any of it
Go down the page once with a pen and put every figure into one of three groups, because they behave differently under conversion and mixing them up is the whole of the risk. The first group is measured quantities: a temperature, a pressure, a mass, a coverage rate, a pot life. These convert. The second is intervals and differences: a tolerance band, a permitted rise, a margin above dew point, a temperature drop across something. These convert by the ratio only, never by the ratio plus the offset. The third group is designations, and these must not be converted at all.
The third group produces the expensive mistakes, because a designation looks exactly like a measurement. DN 15 is not fifteen millimetres of anything; ISO 6708 defines DN as a dimensionless designator, and its inch-system counterpart NPS, dimensioned in ASME B36.10M, is equally a label rather than a size. M12 is a thread system, not a diameter, and it does not mate with a 1/2-13 UNC nut whatever the calculator says. A concrete class of C25/30 to EN 206 is a pair of characteristic strengths on two specimen shapes, and pushing 25 MPa through a factor to get 3,626 psi does not give you a strength specifiable to a plant testing cylinders to ASTM C39/C39M. Rebar is the trap in miniature: a 12 mm bar is not a #4 bar, which is 12.7 mm.
IEEE/ASTM SI 10, the American national standard for metric practice, has a pair of terms worth borrowing even if you never open the document. A soft conversion is arithmetic on the number while the physical article stays as it was. A hard conversion changes the article so it lands on a round figure in the new system. A 100 mm block soft-converted is a 3.937 in block; hard-converted it is a 4 in block, made in a different mould, and the tested performance on the sheet belongs to whichever one was tested. The product does not change because your tape does.
The output of this pass is a short list, not a full translation. Convert only what you will act on today: the figure you have to hit, the one you have to prove, the one you have to buy. The rest stays in the unit it was printed in, which is safer than translating it into a column somebody will later read as original.
| Kind of figure | Examples off a typical sheet | What conversion does to it |
|---|---|---|
| Measured quantity | Minimum substrate temperature, test pressure, unit mass, coverage per unit, pot life | Converts normally. Record the original figure alongside the converted one. |
| Interval or difference | Tolerance band, permitted temperature rise, margin above dew point, temperature drop | Converts by the ratio only. Applying a scale offset to a difference is the classic error. |
| Designation | DN and NPS pipe sizes, M and UNC threads, EN 206 concrete classes, rebar bar marks | Must not be converted. The number is a label for a product, not a dimension of it. |
| Ratio by mass or by volume | Water-to-powder ratio, part A to part B, dilution rate quoted as a proportion | Carries across unchanged, because both sides of the ratio are in the same unit. |
| Rate against a unit of packaging | Litres of water per bag, square metres per pail, kilograms per square metre at a stated thickness | Does not carry across. The packaging changed size; rebuild the rate from the ratio. |
A minimum application temperature is a point, not a quantity
Temperature is the one quantity on the sheet whose scales do not share a zero, and that single fact accounts for most of the errors people make with it. Celsius and Fahrenheit are offset from one another, so a reading converts through both a ratio and a shift: five-ninths of the difference from thirty-two going one way, nine-fifths plus thirty-two coming back. Nothing else on the page works like this. Mass, pressure, length and volume all share a zero across the two systems, which is why they behave like simple multiplications and why temperature quietly does not.
The threshold on this sheet is a minimum substrate temperature of five degrees Celsius, which is forty-one Fahrenheit, and it is worth having a few anchors in your head rather than deriving them every time. Zero is thirty-two. Ten is fifty. Twenty is sixty-eight. Twenty-five is seventy-seven. Minus forty is minus forty on both scales, which is the check that catches a factor applied in the wrong direction. If your converted number is not near one of those anchors when the original is, the mistake is in the arithmetic rather than in the reading.
Read the clause around the number as carefully as the number itself, because manufacturers rarely mean the air. A minimum substrate temperature is a property of the wall, and the wall lags the air by hours — a north-facing concrete elevation on a bright cold morning sits several degrees behind the thermometer hanging in the shade of the scaffold. Many sheets go further and require the temperature to be rising rather than falling, which no single reading can demonstrate. That means two readings and a time between them, and it means the eleven o'clock inspection needs a nine o'clock number to sit against.
Instrument error is the next thing to be honest about. A dial thermometer graduated at two degrees Fahrenheit cannot resolve a one degree Celsius step, so a threshold that is a clean number in Celsius lands between graduations in Fahrenheit and gets rounded by eye, in whichever direction the person holding it would prefer. An infrared thermometer reads a surface, and what it reports depends on the emissivity setting and on the spot size at your standing distance: a reading from the top of a ladder averages a cold shaded band in with a warm sunlit one. Where the decision turns on a degree, use a contact probe on the substrate and let it settle.
Then decide which way to round, before you see the number rather than after. ASTM E29 governs how many digits a conformance decision is entitled to use, and its logic applies without the document in front of you: a stated limit is not made more permissive because your conversion produced a decimal. Five degrees Celsius is forty-one Fahrenheit exactly, and a substrate at forty point six is below the threshold however you round it.
The minimum substrate temperature off the sheet, in the unit your thermometer is actually graduated in — the number that decides whether this morning's reading lets the work start.
The temperature value to convert.
The unit your starting value is in.
Converted temperature
21.11 °C
They open the calculator with your figures already in it
Temperature Conversion Calculator: 21.11 °C — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
What this calculation does not cover
- This converts a point on the scale, not a difference between two points. A 20-degree Fahrenheit rise is 11.1 °C; type 20 here and you get −6.7 °C, because the 32-degree offset belongs to where the scale starts and not to the interval. Anything built on a temperature difference — delta-T across a coil, heat loss through an assembly, degree-days — converts at 5/9 or 9/5 alone, with no offset.
- Air temperature and surface temperature are different numbers, and the application limits in the FAQ above mean the surface. A slab on grade or a north-facing wall sits below the shade air reading and lags it by hours, so a coating or adhesive can go on at a compliant forecast temperature onto a substrate still under its minimum, where it never properly cures. Shoot the substrate with an infrared thermometer rather than converting the forecast.
Anything with a delta in front of it loses the thirty-two
Immediately below the minimum temperature, most coating and tanking sheets carry a second temperature condition of a completely different type: the substrate must be some margin above the dew point, commonly three degrees Celsius, and ISO 8502-4 is the guidance that tells you how to establish whether you have it. That three is not a temperature. It is a difference between two temperatures, and differences do not carry the offset. Three degrees Celsius of margin is five point four degrees Fahrenheit. Running it through the point conversion instead gives thirty-seven point four, which would have you refusing to work on a substrate thirty-four degrees clear of condensation.
The same distinction governs every band on the sheet. A pot life quoted at twenty degrees plus or minus two degrees Celsius is a point and a band: the point becomes sixty-eight Fahrenheit, the band becomes three point six, and the range is sixty-four point four to seventy-one point six. A permitted rate of rise during cure, a maximum drop across a joint, an allowable differential between mix water and ambient — all differences, all converting at nine-fifths flat. The test that never fails is to ask whether the number would still mean something if the scale's zero moved. A wall at five degrees would not; a gap of three degrees would. Dew point itself is a point in exactly the sense the word suggests, so it takes the offset — which is why a psychrometric reading to ASTM E337 and the margin above it need two different arithmetics on one line of the log.
Test pressure: find out which zero it was measured from
The soundness test on this sheet is quoted in kilopascals and the pump has a psi gauge. The factor is not the difficulty: one pound per square inch is 6.894757 kilopascals, so a hundred kilopascals is 14.50 psi, to a precision far beyond what any site gauge delivers. Bar turns up on European equipment at least as often and is easier still, being exactly a hundred kilopascals. Megapascals appear on structural sheets and hydraulic equipment, at 145.04 psi each.
The difficulty is the datum. Pressure is only a number once you have said what it is measured from, and the two candidates differ by one atmosphere — 101.325 kilopascals, or 14.696 psi. Nearly every site gauge reads gauge pressure: zero when open to the air, reporting the amount by which the system exceeds the atmosphere. Absolute pressure counts from vacuum. On a high-pressure hydrostatic test the difference is a rounding error nobody notices. On a low-pressure air test it is the entire measurement — fifty kilopascals gauge is a modest 7.25 psi of air in a pipe, while fifty kilopascals absolute is a partial vacuum, and those are not adjacent readings but opposite operations.
SI practice, in the BIPM's SI Brochure and in the American guidance that follows it in NIST Special Publication 811, is explicit that you may not solve this by decorating the unit symbol. A datum belongs to the quantity, not the unit, so kPa(g) and kPa(a) are not admissible symbols in the way the trade habitually uses psig and psia. In practice that means a metric sheet is less likely to tell you which datum it means, because the convention that carried the information has been removed from the unit. Where the sheet is silent and the number is small, ask. Where you cannot ask, note the ambiguity on the record rather than resolving it silently in the direction that suits the programme.
Head is the third form the same quantity takes, and it appears on any sheet written for people who work with water. Metres head and feet head are pressures expressed as a column of water, converted through the fluid's density: a metre of water is about 9.81 kilopascals, and a psi about 2.31 feet of water. A pump curve in metres head, a static lift in feet and a working pressure in kilopascals are three units of one quantity, and the system is only checkable once they share one.
Be clear, too, about what the sheet is entitled to specify. A manufacturer can tell you the maximum pressure its product withstands, which is a product property. The pressure at which the installation must be tested, for how long, with what medium and to what acceptance criterion is a code requirement — the International Plumbing Code and the International Residential Code each set out tests and inspections for plumbing work, ASME B31.9 governs building services piping, and BS EN 806-4 covers the metric side. The code sets the test; the data sheet tells you whether the product survives it. Converting the manufacturer's figure and testing to that runs a test nobody asked for.
Kilopascals off the sheet into the psi your test gauge is graduated in — and back the other way when the completion paperwork wants the figure in the unit the specification used.
The pressure value to convert.
The unit your starting value is in.
Converted pressure
206.8 kPa
They open the calculator with your figures already in it
Pressure Conversion Calculator (PSI ↔ kPa): 207 kPa — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
Add the equipment this sizes
This result is a specification — 206.8 kPa — not a quantity. Put the thing it sizes into your project: how many, what you call it, and your supplier’s price.
What this calculation does not cover
- The conversion changes the unit, not the reference datum. Almost every dial on a compressor, pressure tank or tyre reads gauge pressure, so 30 PSI converts to 206.84 kPa gauge, which is about 308 kPa absolute once the 101.325 kPa of atmosphere at sea level is added. A spec sheet quoting kPa(a), bar(a) or psia is describing a different quantity, and this page will not shift a reading between the two scales.
- Only PSI and kilopascals are offered, so a figure quoted in any other pressure unit has to be brought into one of those two first. One bar is 100 kPa and one MPa is 1000 kPa, both exact; a gas appliance figure in inches of water gauge is a far smaller unit, at roughly 249 Pa or 0.036 PSI per inch, and reading such a figure as PSI overstates it by a factor of about 28.
- Pressure is not head, and no fluid property enters this arithmetic at all. The same factor is applied whether the reading came from air, water or hydraulic oil, whereas converting pressure to a height of liquid depends on density, so one PSI is about 2.31 feet of water column only for water at ordinary temperatures. Use the PSI to feet of head calculator rather than reading this result as a lift.
- A converted figure says nothing about whether the pipe, hose, fitting or vessel is rated for it. Pressure ratings are also derated as temperature rises, sharply so for plastic pipe, and this page has no material or temperature input, so a result inside a component's cold rating can still exceed what that component may carry hot.
The gauge decides what you are able to prove
Having the number in psi is not the same as being able to demonstrate it. Pressure gauge accuracy is graded as a percentage of span rather than of reading — ASME B40.100 sets out the North American grades and EN 837-1 the accuracy classes for Bourdon tube gauges elsewhere — and for the lower grades the permitted error is larger at the ends of the dial than through the middle. A nought-to-three-hundred psi gauge carries an error measured against three hundred wherever the needle sits, so a test at fifteen psi is judged by an instrument whose uncertainty is a large fraction of the reading.
The guidance that accompanies those grades is to choose a gauge whose full scale puts the working pressure between a quarter and three-quarters of the dial. A test near a hundred kilopascals therefore wants a full scale around thirty psi, not three hundred. That is a second gauge in the van rather than a piece of arithmetic, and the order of events is worth being blunt about: no conversion factor improves an instrument. If the only gauge on site cannot resolve the figure, the honest options are to fetch one that can, or to record that the test was witnessed on an instrument unsuited to the pressure.
The same reasoning covers the rest of the van. A torque figure in newton metres converts at 0.7376 to pound force feet, but a click wrench carries its own tolerance under ISO 6789, stated against the setting, so a wrench near the bottom of its range is a poor instrument however carefully the conversion was done. Weighing instruments are governed by OIML R 76 internationally and NIST Handbook 44 in the United States, and the site-office bathroom scale meets neither. Convert, then look at what will read the result, and let the weaker of the two decide how many digits you may write down.
Bag weight converts; everything hung off the bag does not
The pails are in kilograms and the merchant's replacement stock will be in pounds, so the first conversion is the easy one: a kilogram is 2.204623 pounds by international agreement, making a 25 kg unit 55.1 lb and a 20 kg unit 44.1 lb. The other way, a 50 lb bag is 22.68 kg, and the 94 lb sack of Portland cement that American practice is built around is 42.64 kg. None of those masses is a property of the material. They are packaging conventions, the metric sizes owing a good deal to manual handling guidance — the Manual Handling Operations Regulations 1992 in the United Kingdom, the revised NIOSH lifting equation and ISO 11228-1 elsewhere — none of which OSHA turns into a numeric lifting limit of its own.
The trouble starts a line below, where every useful figure is quoted against the bag rather than the material. Coverage of 1.2 square metres per 25 kg unit at twenty millimetres is a statement about the material — roughly 0.048 square metres per kilogram — wearing a unit of packaging as a disguise. Buy the same product in a 50 lb bag and the coverage is not 1.2 square metres but 22.68 over 25 of it: 1.09 square metres, or 11.7 square feet. Order by bags off an unconverted coverage figure and you are nine per cent short, which on a tanking job leaves a wall open overnight at the exact junction the detail exists to protect.
Water addition behaves the same way and hurts faster. Five litres per 25 kg unit is a ratio of 0.2 by mass wearing packaging units, since a litre of water is near enough a kilogram. The ratio is the durable figure: dimensionless, indifferent to unit system, applicable to any bag size. Carry the litres-per-bag figure to a smaller bag unchanged and the mix goes wet, which for a cementitious product means lost strength and, on a tanking slurry, a coating that will not hold its head. Batching in US gallons, a gallon of water weighs about 8.34 lb, so a 50 lb bag at that 0.2 ratio wants 10 lb of water — 1.2 gallons, reached through the ratio and never through the litres figure.
Two more mass traps arrive whenever something is being lifted. A tonne is a thousand kilograms and 2,204.6 lb; a US short ton is 2,000 lb, or 907.2 kg; a long ton is 2,240 lb. A pallet described as one and a half tonnes is 3,307 lb, not three thousand, and the ten per cent sits directly on a telehandler load chart. And mass is not force: a load in kilograms per square metre becomes a pressure only after multiplying by gravity, at which point 250 kg per square metre is 2.45 kilopascals, or 51.2 pounds per square foot. A floor loading and a bag weight look like the same kind of number and are not.
Unit mass across the two systems — the figure every coverage rate, yield and water addition on the sheet is secretly quoted against.
The weight value to convert.
The unit your starting value is in.
Converted weight
22.68 kg
They open the calculator with your figures already in it
Weight Unit Conversion Calculator: 22.68 kg — shown in imperial, US market. The link sets both, so the result they see is the one on your screen.
Estimated cost — your price
This site holds no price list for this material — local prices vary too much to publish honestly. Enter your supplier's price and the result is costed with it.
Punctuation and capital letters that change the number
European sheets are commonly typeset with a comma as the decimal marker, which ISO 80000-1 permits alongside the point, and the same standard asks for a thin space rather than a comma to group thousands. A sheet to that convention reads 1,5 mm for one and a half millimetres and 1 500 kg for fifteen hundred kilograms; read with Anglo-American habits, both misreadings are an order of magnitude or worse. The comma is the more dangerous because it produces a plausible number: 0,45 read as forty-five is absurd and gets caught, while 1,5 read as fifteen is a thickness somebody will happily apply.
Unit symbols carry their own traps, and the case of a letter is load-bearing. MPa and mPa differ by nine orders of magnitude. Capital T is tesla, lower-case t is the tonne. Capital K alone is kelvin, a scale with an absolute zero that takes no degree sign, so a sheet reading K rather than °C is not a typographical variant. kN is kilonewtons and kn is knots. Kg is not a legal symbol at all — the kilogram is kg — and a sheet that writes Kg is one where somebody has been retyping figures, which is reason enough to check the rest of it.
Both points argue for the same habit: transcribe the figure and its unit together, exactly as printed, before touching either. Separated from its symbol, a number becomes a bare quantity in whatever system the reader works in, which is the entire subject of this article. Photograph the relevant panel with a phone and keep the image with the job record — it settles which revision was on site, and it is the only evidence that survives a pail going in the skip.
Write it down once, in both units, with the direction of the rounding
All of this is undone by one bad habit: converting a number that has already been converted. Each pass rounds, each rounding is invisible in the result, and the third figure in a chain is out by an amount nobody can reconstruct because the intermediate steps were never written down. IEEE/ASTM SI 10 is firm about that and about its sibling rule, that a converted value should carry no more significant digits than the original had. A sheet's five degrees Celsius written as 41.0000 degrees Fahrenheit claims a precision the manufacturer never offered, and somebody downstream will treat the claim as real.
Round in the direction that keeps you inside the requirement, and say on the record which direction that was. A minimum converts upward: a hundred kilopascals of test pressure is 14.50 psi, and setting the pump to fourteen is a test that was not performed. A maximum converts downward by the same logic. Where a figure is a target rather than a limit, round to the resolution of the instrument that will read it and no further, because digits beyond that resolution are a fiction the paperwork will inherit.
The record wants both numbers on one line — the practice that metric-practice guidance calls dual dimensioning, and that the trade has always done informally on drawings. Original figure as printed, unit as printed, converted figure, factor used, direction of rounding. One extra column on a site sheet, and it answers in advance the only question an inspector or a warranty claim ever asks about a converted number. It is also where ambiguity gets recorded rather than resolved: if the sheet gave a pressure without a datum, the record says so, and treating it as gauge pressure becomes a named person's decision on a named date rather than an assumption on a tailgate.
- Separate the technical data sheet from the safety data sheet, and confirm which revision of each is actually in the box.
- Mark every figure on the technical sheet as a measured quantity, an interval, a designation, or a ratio, and leave the designations alone.
- Convert only the figures you will act on today: what you must hit, what you must prove, what you must buy.
- For temperatures, apply the offset to points and the ratio alone to margins, bands and rises.
- For pressures, establish the datum before applying any factor, and check the gauge you own puts the target in the middle of its dial.
- For anything quoted per bag or per pail, rebuild it as a rate per kilogram or a dimensionless ratio, then re-apply it to the packaging you actually bought.
- Write both figures on the record with the factor and the direction of rounding, and photograph the printed panel the originals came from.
- Convert once, from the original, every time. Never from a number that has already been through the arithmetic.
| Quantity | Metric | Inch-pound | What it catches |
|---|---|---|---|
| Temperature point | 0 °C / 5 °C / 20 °C | 32 °F / 41 °F / 68 °F | A factor applied without the offset, or with it applied twice. |
| Temperature interval | 1 °C of margin | 1.8 °F of margin | The offset wrongly applied to a difference or a tolerance band. |
| Both scales agree | −40 °C | −40 °F | A conversion run in the wrong direction. |
| Pressure | 100 kPa (1 bar) | 14.50 psi | A factor of ten from mis-set units on a calculator. |
| One atmosphere | 101.325 kPa | 14.696 psi | A gauge reading compared against an absolute figure. |
| Head of water | 1 m ≈ 9.81 kPa | 1 psi ≈ 2.31 ft | A pump curve and a working pressure compared in different forms. |
| Mass | 1 kg / 25 kg | 2.205 lb / 55.1 lb | A coverage or water figure carried across on the wrong bag size. |
| Large mass | 1 tonne | 2,204.6 lb | A tonne read as a US short ton on a lifting chart. |
| Areal load | 1 kPa | 20.89 psf | Mass per unit area used as a pressure without gravity applied. |
Before the first pail is opened
Four figures decide whether this morning's work is defensible: the temperature you must beat, the margin above dew point, the pressure you must hold, and the mass every rate on the sheet is quoted against. Settle them on paper, in both units, before anything is mixed.
- The revision number of the technical data sheet in the box — Units and thresholds change between revisions and the sheet's layout is governed by nobody. The safety data sheet's sixteen sections are mandated; the technical sheet's are not.
- Minimum substrate temperature, as a point, in the unit your probe reads — Substrate, not air, and check whether the sheet also requires the temperature to be rising. That needs two readings and a time between them.
- The margin above dew point, converted as a difference — Three degrees Celsius of margin is 5.4 °F, not 37.4 °F. ISO 8502-4 is the guidance on establishing whether you have it; the psychrometry sits alongside under ASTM E337.
- Test pressure, its datum, and the gauge that will read it — Gauge or absolute changes a low-pressure figure completely. Pick a gauge whose full scale puts the target between a quarter and three-quarters of the dial.
- Unit mass, and every rate rebuilt off it rather than off the bag — Coverage per kilogram and water as a ratio by mass both survive a change of packaging. Litres per bag and square metres per bag do not.
- One record line per converted figure — Original as printed, unit as printed, converted value, factor, direction of rounding. It is the only answer to the question an inspector will ask.
Opens the calculators above on one screen with the dimensions from this article already filled in. Quantities only — this site publishes no price list, because local prices vary too much to publish honestly.