How to Read House Pressure With a Manometer

Why this matters

House pressures are a few pascals, smaller than the pressure change from lifting a gauge a few feet in the air, and the wind on the far side of a window can swing a reading by as much as the effect you are trying to see. That is why two techs can put a gauge on the same house on the same afternoon and write down numbers with opposite signs. The gauge is rarely the problem. The setup is: an unzeroed instrument, a reference hose flapping in the wind, a door pinched shut on the hose, a reading taken two seconds after the dryer started. This card covers how to set up a house, take readings that mean something, and write them down so the next person can trust them. How a gauge works and what a reference port does belongs to Measuring Pressure and What the Reference Is; the size of a pascal belongs to What a Pascal Is and Why House Pressures Are Tiny.

Before you change anything in the house: turning on exhaust devices can make an atmospheric gas or oil appliance spill flue gas. Wear a personal CO monitor for the whole visit. For any condition with exhaust running, keep atmospheric appliances from firing (furnace thermostat turned down, water heater set to its vacation or pilot setting with the owner's agreement) unless you are deliberately running a spillage screen. If you smell gas at any point, everyone leaves, nobody touches a switch or a phone inside, and you call the gas utility from outside.

What you need

• A digital manometer that reads to 0.1 Pa and has a time-averaging function.
• Two lengths of clear tubing: a long one (25 ft or more) for an outdoor reference and a shorter one for interior zones.
• Painter's tape, and a small container with a perforated lid or a loose wad of foam to shelter the outdoor hose end.
• The log below, on paper or in the job notes.

The steps

1. Write the question on the log before you connect anything. State the space being measured and what it is measured with respect to (WRT), for example "first floor WRT outdoors" or "basement WRT outdoors". The input port goes to the space being measured and the reference port to the comparison. A negative reading then means the measured space is at lower pressure than its reference. Skip it and you end up with a column of numbers whose sign nobody can interpret a month later, which is the most common reason a pressure log is useless.

2. Zero the gauge with both ports in the same air. Hold the manometer in the room, both hoses off or both open to the same still air, and use its zero function. Do it again any time you move the gauge to a different floor. Skip it and an offset of a few tenths of a pascal, normal drift for an instrument sitting in a cold truck, rides on every reading; 0.4 Pa of offset on a 1.5 Pa effect is an error of more than a quarter.

3. Run the reference hose outdoors and shelter its end. Pass it through a door or window and close the door on it gently; check the tubing is not crushed flat at the weatherstrip. Put the outdoor end several feet out from the wall, near the ground, out of direct wind, and away from any exhaust outlet (dryer vent, bath fan hood) and from an outdoor condensing unit's fan. Put the end in the perforated container or under the foam so gusts do not blow straight into it. Note which side of the house the end is on. Skip it and wind blowing across or into the open hose end adds or subtracts velocity pressure, and on the windward side the whole wall is pushed positive, so the "outdoor" pressure you are reading against is not the outdoor pressure.

4. For an interior zone, close the hose under a door. For a room WRT the main body, or a basement WRT the first floor, run the shorter hose under the closed door, with the door resting on the tubing only lightly. Tape the hose to the floor so it is not kicked. Skip it and an open door equalizes the two spaces and you read zero, or a door closed hard on the tubing pinches it and you read whatever pressure was trapped when it closed.

5. Set one house condition at a time, starting from a baseline. The baseline is the house with exterior doors and windows closed, interior doors open, every exhaust device off, the air handler off, and the fireplace damper as you found it. Then add one thing per condition: bath fans, then the range hood on high, then the dryer, then the air handler. Write each condition on its own line. Skip it and when you turn on three things at once you know the total but not which device caused it, which is the question the customer is paying you to answer.

6. Wait, then average. After each change, wait 30 to 60 seconds for the house to settle, then take the gauge's longest time average (or several 10-second averages) and write down both the average and the range you saw. Skip it and a single instant reading in wind can land anywhere in the gust range. If the range is wider than the effect you are measuring, average longer, move the outdoor hose end to a more sheltered spot, or come back on a calmer day.

7. Subtract the baseline to get each device's effect. Stack effect and wind are present in every reading, including the ones with fans running. The effect of a device is its condition's reading minus the reading of the condition before it. Skip it and you will credit a dryer with the stack effect that was already there before you switched it on.

8. Restore the house and write down that you did. Turn off every device you turned on, return the thermostat and the water heater to their original settings, and watch each atmospheric appliance fire normally at its draft hood before you leave. Pull the hoses and close the window or door they passed through. Skip it and the customer finds the water heater cold the next morning, or the dryer still running.

Worked example: a filled-in log

A two-story house with a basement. Outdoor 28 F, indoor 70 F, wind 5 to 10 mph from the west. The outdoor hose end sits in a container on the east side, the sheltered one. Gauge on the first floor, then the basement for the last line.

HOUSE PRESSURE LOG      input = space, ref = outdoors
-----------------------------------------------------
Out 28 F  In 70 F  Wind W 5-10 mph  ref end: E side
Gauge zeroed 1st fl 09:40, again in basement 10:31
WH to vacation 09:45, furnace stat down, CO 0 ppm

#  Space      Condition          Avg Pa   Range
1  1st fl     baseline, all off   -1.6   -1.1/-2.0
2  1st fl     + 2 bath fans       -2.5   -2.0/-3.0
3  1st fl     + hood high         -5.3   -4.8/-5.9
4  1st fl     + dryer             -7.9   -7.3/-8.4
5  1st fl     + air handler       -8.4   -7.9/-8.9
6  basement   same as 4          -10.6  -10.1/-11.2

Restored 10:52: all off, WH on, burner fired and
drafting 10:55, stat back to 68, CO 0 ppm

Reading it, one subtraction at a time:

• Two bath fans: -2.5 minus -1.6 is -0.9 Pa.
• Range hood: -5.3 minus -2.5 is -2.8 Pa.
• Dryer: -7.9 minus -5.3 is -2.6 Pa.
• All exhaust together: -7.9 minus -1.6 is -6.3 Pa.
• Air handler: -8.4 minus -7.9 is -0.5 Pa.

The number that would have fooled you is line 2. The bath fans moved the house 0.9 Pa, and the baseline's own gust range was 0.9 Pa wide (-1.1 to -2.0). A tech taking single instant readings could have written anything from no effect to twice the real one. The averages make it readable; the range column is what tells you how much to trust it.

The air handler line is a clue, not noise. Running the air handler pushed the house 0.5 Pa more negative. Supply duct leaks outside the envelope push conditioned air out of the house and pull it negative; return leaks outside pull outdoor air in and push it positive. A negative shift says supply-side leakage outside the envelope outweighs return-side leakage here. It does not say how much leakage there is, which is a duct leakage test. See related: Duct Leakage Outside the Envelope.

The basement line checks itself against the stack effect. The basement read -10.6 Pa against -7.9 on the first floor under the same condition, 2.7 Pa more negative. At this 42 F indoor-outdoor difference the stack gradient works out to about 1 Pa per meter of height (see related: Stack Effect and the Neutral Pressure Plane for the derivation), and the two gauge positions were about 9 ft, or 2.7 m, apart, so the stack effect alone predicts roughly 2.7 to 2.8 Pa. The reading agrees with the physics, which is a good sign the setup was sound.

What the log does not say. It does not say whether any appliance spills; the water heater was held off during the exhaust conditions on purpose. Whether -6.3 Pa of exhaust depressurization is a problem for this house's combustion appliances is a question for a qualified gas technician, and the log is the evidence you hand them. See related: Combustion Safety After a Partial Conversion; How to Screen for Spillage After You Add an Exhaust Fan.

When the readings cannot be trusted

• The range of any line is wider than the effect you are reporting. Re-take it with a longer average or on a calmer day before you draw a conclusion from it.
• The baseline drifts between the start and the end of the visit. Take a second baseline after restoring the house. If it differs from the first by more than a few tenths of a pascal, wind or temperature changed and the device effects in between carry that uncertainty.
• Water or condensation is in the hose. A droplet in a low loop of tubing holds a pressure of its own. Run the hose without sags and blow it clear if you see moisture.
• The reference end ended up near a running dryer vent or condenser. Move it and redo the lines taken while it was there.
• The gauge was not re-zeroed after moving floors. Any reading taken after the move carries the offset. Re-zero and repeat that line.

References

• See related: Measuring Pressure and What the Reference Is; What a Pascal Is and Why House Pressures Are Tiny
• See related: The House as One Pressure System; Stack Effect and the Neutral Pressure Plane
• See related: Counting Every Exhaust Device in the House; Duct Leakage Outside the Envelope
• Manufacturer documentation for your manometer's zeroing and time-averaging functions