What a Pascal Is and Why House Pressures Are Tiny

Why this matters

An HVAC tech is used to reading half an inch of water column across a furnace and three and a half inches at a gas valve. Hand that tech a house pressure reading of 4 Pa and it sounds like nothing, a rounding error, a number you would never act on. That instinct is wrong in a way that matters: 4 Pa is enough to pull a few hundred cubic feet of air a minute through a typical house's leaks, and enough to overpower the draft of a natural-draft water heater. Getting comfortable with the unit, and with how small house pressures really are, is what lets every trade read the numbers a blower door tech or an energy auditor hands them.

The unit

A pascal (Pa) is one newton of force spread over one square meter. It is a very small pressure. The units the trades already use are all much bigger:

  PRESSURE CONVERSION STRIP (approximate)

  1 Pa         = 0.004 in WC
  5 Pa         = 0.020 in WC
  25 Pa        = 0.100 in WC
  50 Pa        = 0.200 in WC   blower door test point
  125 Pa       = 0.500 in WC   typical duct static
  249 Pa       = 1.000 in WC
  872 Pa       = 3.5 in WC     gas manifold, nat gas
  6,895 Pa     = 1 psi = 27.7 in WC
  101,325 Pa   = 1 atmosphere = about 407 in WC

  1 mm WC = 9.8 Pa   1 mbar = 1 hPa = 100 Pa

  typical house to outdoors:  1 to 10 Pa
  = 0.004 to 0.040 in WC

The conversion that matters is one inch of water column to about 249 Pa. Everything else follows from it. To go from inches of water column to pascals, multiply by 249. To go the other way, divide by 249, or multiply by 0.004. NIST's guide to SI units (SP 811) lists the exact factors, which differ slightly with the water temperature they assume; 249 is close enough for house work.

Check the strip once by hand. 0.02 in WC times 249 is 4.98, call it 5 Pa. One atmosphere, 101,325 Pa, divided by 249 is about 407 in WC, which is the familiar 34-ish feet of water: 407 divided by 12 is about 33.9 feet. The strip holds together.

How small that really is

A house pressure of 5 Pa is about one twenty-thousandth of the atmosphere pressing on both sides of the wall. Barometric pressure swings by far more than that when a weather front comes through, which is why house pressures are always measured as a difference between two places at the same moment, never as an absolute reading. Which two places, and which side the gauge calls positive, is the subject of Measuring Pressure and What the Reference Is; this card owns the units only.

Some feel for it:

• On a door. A standard 30 by 80 inch interior door is 2,400 square inches, about 1.55 square meters. At 5 Pa across it, the force is about 7.7 newtons, a little under 2 pounds. That is enough to swing a free-hanging door shut, hold a bedroom door against its latch, or make it whistle at the undercut.
• Through one small hole. A 1-square-inch hole with 5 Pa across it, treated as a sharp-edged orifice, passes roughly 2 to 3 cfm. Trivial alone. A house has hundreds of such holes, plus a few big ones.
• On a gauge. A dial gauge whose smallest division is 0.01 in WC is reading in steps of about 2.5 Pa. It cannot resolve a house pressure. House and room pressures are read on a digital manometer that resolves a tenth of a pascal.

Why a few pascals moves real air

The flow through a house's leaks depends on the total leakage, not on any one hole. A blower door test measures that leakage as CFM50: the airflow that holds the house 50 Pa below outdoors. Below 50 Pa, flow falls off with pressure, but not in proportion. With a typical leak exponent of about 0.65, flow at any pressure is CFM50 times the pressure divided by 50, raised to 0.65. The relationship and its reverse, pressure from a known exhaust flow, are laid out in Kitchen Hood Makeup Air.

For a house that tests at 2,000 CFM50:

Look at the shape of it. A tenth of the test pressure, 5 Pa, moves nearly a quarter of the test flow. One pascal still moves about 160 cfm, more than a bath fan. House leakage is generous at low pressure, so the house never has to get very negative to feed its fans, and in a leaky house that keeps every number small.

The flip side is the tight house. The same 450 cfm through a house that tests at 1,000 CFM50 takes about 15 Pa: 450 over 1,000 is 0.45, raised to about 1.54 is 0.29, and 50 times 0.29 is about 14.6 Pa. Halving the leakage roughly tripled the pressure the same fans produce, because the relationship is curved. See related: Counting Every Exhaust Device in the House.

Why a few pascals beats a natural draft

A natural-draft appliance, the kind with a draft hood or an open draft diverter, moves its flue gas by buoyancy alone: hot gas in the vent is lighter than the air around it and rises. The pressure that buoyancy creates at the draft hood, once the vent is warm, is only a few pascals. At a cold start, before the vent warms, it is close to nothing.

So the comparison that decides spillage is a few pascals of draft against a few pascals of house depressurization, the same order of magnitude. That is why house pressures that look like rounding errors on a duct gauge are a life-safety number in a house with a natural-draft water heater or furnace. Combustion safety standards set worst-case depressurization limits by appliance type; Kitchen Hood Makeup Air cites the BPI figures, and the test itself belongs to a qualified combustion-safety tech. See related: How a House Pulls Flue Gas Back Down a Vent, and How Draft Actually Works.

Sealed-combustion appliances, which take combustion air from outdoors through a pipe and vent to outdoors in a sealed system, most of them under fan pressure, are largely immune to these small house pressures. That is one reason a tight house and a natural-draft appliance are a pairing every trade should flag.

Worked example: reading a number someone hands you

An energy auditor's report says: "Mechanical room, worst case, -0.012 in WC with reference to outdoors." The plumber about to replace the water heater with another natural-draft unit wants to know whether that is a problem.

Convert. 0.012 times 249 is 2.99, about 3.0 Pa, with the mechanical room below outdoors.

What it means in flow. The house tests at 2,000 CFM50, the same house as the table. At 3 Pa, 3 over 50 is 0.06; raised to 0.65 that is about 0.161; times 2,000 is about 320 cfm moving through the leaks under worst-case conditions. Reverse check: 320 over 2,000 is 0.16, raised to 1.54 is about 0.060, and 50 times 0.060 is 3.0 Pa. It closes.

What to do with it. 3 Pa negative in the room where a natural-draft appliance lives is the same order as that appliance's draft. Whether it passes is set by the combustion-safety standard the auditor worked to, not by this card, and the auditor's report should say. The plumber's useful move is to read the limit off the report, compare, and if it fails, raise sealed-combustion or power-vented options with the customer before ordering a like-for-like replacement.

The failure mode: units dropped. The same number written as "-3" with no unit is unusable. As pascals it is a borderline worst case. As inches of water column it would be 747 Pa, impossible in a house. As millimeters of water column it would be about 29 Pa, a severe problem. A metric manometer left on mbar would display 0.03 for the same 3 Pa, and someone reading "0.03" as inches of water column would think it was 7.5 Pa. Write the unit, every time, and set the manometer to Pa for house work.

Pressure splits across layers, and the tight one takes most of it

A space between the house and outdoors, an attic, a crawlspace, a garage, sits behind two layers in series: house to space, and space to outdoors. Whatever total pressure the house has to outdoors divides between those two layers, and the tighter layer carries more of it, because the same air has to squeeze through both and the smaller opening needs the larger push.

That is how a blower door tech reads where the boundary is without opening a wall. With the house held at -50 Pa to outdoors:

• Attic reads about -45 Pa to outdoors, so about 5 Pa across the ceiling and 45 across the roof vents. The attic is behaving as part of the house, which in a vented attic means big ceiling bypasses.
• Attic reads about -3 Pa to outdoors, so about 47 Pa across the ceiling. The ceiling is the tight layer and the attic is outside it, as a vented attic should be.

The two readings add to the house total: 45 plus 5, and 3 plus 47, are both 50. If they do not add up, one of the hoses or one of the reference sides is wrong. The same logic is how a kneewall attic that "should be outside" turns out to be connected to the house through open floor joists. See related: Where the Air Barrier Actually Is in a House, and Blower Door Numbers Explained for Other Trades.

Pressure is not flow. In series, the same air passes through both layers. A big pressure across a layer says that layer is the tighter of the two, not that air is rushing through it; the leakier layer is simply the one with little pressure across it. Reading a big number as a big leak sends people to seal the layer that was already doing its job.

What changes the reading

• Weather. Stack effect adds pressure in cold weather and wind adds it in gusts. A reading taken on a mild, calm day understates what the room sees in January. See related: Stack Effect and the Neutral Pressure Plane.
• Fans and doors. A worst-case reading means the exhaust fans on and interior doors set to make the appliance room most negative. A reading with nothing running is a baseline, not a worst case.
• Fluctuation. Wind makes house pressure jump around, so manometers used for house work average over several seconds. A single glance at a flickering display is not a reading.

References

• See related: Measuring Pressure and What the Reference Is (reference side, sign convention)
• See related: Kitchen Hood Makeup Air (flow-pressure relationship, combustion safety limits)
• See related: How Draft Actually Works; How to Read House Pressure With a Manometer
• NIST, Guide for the Use of the International System of Units (SP 811), pressure conversion factors