Pressure Units in Instrumentation: barg, bara, psig, mmWC and the Errors Between Them

9 min read · updated 2026-09-12

Most pressure "errors" in a plant are not instrument faults. They are a unit conversion done correctly and a reference basis carried across wrongly — and the second one is invisible until something downstream stops making sense.

Converting the unit is not converting the reference

Every pressure measurement has two independent properties: the unit (bar, psi, kPa, mmWC) and the reference basis (gauge, absolute, differential). Changing one does nothing to the other, and confusing them is the single most common pressure mistake in instrumentation.

100 psig converts to 6.895 barg. Both are still gauge — measured relative to whatever the atmosphere is doing outside. Nothing about dividing by 14.5038 turned it into an absolute pressure. To get there you add atmospheric: 6.895 + 1.013 = 7.908 bara. Two separate operations, and skipping the second silently introduces about 101 kPa of error.

Where the basis error actually bites

Gauge pressure is fine for most day-to-day work. The trouble starts wherever physics needs absolute pressure, and there are four places that come up constantly:

• Steam tables. Published in absolute. Feed a 10 barg header into a table expecting 10 bara and the saturation temperature comes out 179.9 °C instead of 184.1 °C — a 4 °C error on temperature-sensitive product. See the saturated steam table.
• Gas law calculations. Density, actual-to-standard flow correction and compressor work all use absolute pressure. At low pressures the error is proportionally enormous — at 0.5 barg, forgetting atmosphere is a factor-of-three mistake.
• Vacuum service. Gauge pressure goes negative; absolute never can. A reading of −0.5 barg is 0.513 bara, and any calculation that lets pressure go below zero has gone wrong.
• Compressible valve sizing. The pressure drop ratio x = ΔP/P₁ needs P₁ absolute — see the gas and steam Cv calculator.

The gauge to absolute converter handles the arithmetic, including the altitude correction that matters if your plant is well above sea level — at 1,500 m, atmospheric pressure is about 84 kPa rather than 101, and using the sea-level figure is its own quiet 17 kPa error.

The units only instrumentation people use

Process engineering lives in bar and kPa. Instrumentation drags in a longer list, because the measurements at the small end need finer units:

Two of these deserve a warning. kg/cm² is not bar — it is 0.98 bar, a 2% difference that looks like instrument drift rather than a unit error, and it appears on a lot of older Indian plant documentation. And mmWC is temperature-dependent in principle, since it is defined by a water column; the 9.80665 Pa figure assumes standard gravity and 4 °C water, which is the convention everyone uses but nobody states.

For the low-range end, useful field shortcuts: 1 mbar ≈ 10.2 mmWC, and 100 mmWC is just under 1 kPa. The pressure unit converter covers the full set.

Differential is a third basis, not a unit

A DP transmitter measures the difference between two pressures, and that difference has no reference basis of its own — atmospheric pressure cancels out on both sides. This is why a DP range can be entirely negative without anything being wrong, as happens with a wet leg on a closed tank, and why "is that gauge or absolute?" is not a meaningful question about a differential measurement.

A habit worth forming

Write the basis every time. Not "10 bar" but "10 barg" or "10 bara". Not "250 mmWC" on a datasheet that might mean differential — say so. It costs one character and eliminates an entire category of error that otherwise surfaces weeks later as a measurement nobody can explain.

And when a pressure reading disagrees with expectation by roughly one atmosphere, or by exactly 2%, stop looking at the transmitter. You have a basis error or a kg/cm² error, and no amount of recalibration will fix either.

Calculators used in this guide: pressure unit converter · gauge ↔ absolute · hydrostatic pressure · saturated steam table