GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Which way does the air flow within the house?

Jeff Watson | Posted in General Questions on

Might be an odd question but I have been trying to understand this for the longest.

With all the windows closed in my single-story house + basement, how can I figure out the direction of natural air flow. I assumed that due to stack effect, I would be able to locate those positive & negative air pressure differences to identify air infiltration/exfiltration.

My (failed) attempt at solving this was to buy a (differential) manometer and take air pressure readings, comparing different areas/rooms of the house. The result: inconclusive as my manometer gives readings that are seemingly all over the place and inconsistent.

A more practical approach was figuring out air flow when the furnace is on. I figured this would help explain temperature variations room to room. For example, there is one room where when I close the door, I can feel air rushing into the room at the jambs…which makes me believe this room is under overall negative pressure & sucking up air from other rooms.

So I guess my question is – how does one accurately measure a house to determine air flow & pressure differences?

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. GBA Editor
    Martin Holladay | | #1

    Q. "Which way does the air flow within the house?"

    A. Air flow requires two conditions: there must be a difference in pressure between two spaces, and there must be a hole. If these conditions are met, air will always flow from high-pressure areas to low-pressure areas.

    Q. "How does one accurately measure a house to determine air flow & pressure differences?"

    A. With a manometer.

    Note that many factors affect air flow in a house: the leakiness of the thermal boundary, wind direction and speed, the delta-T between indoors and outdoors, the location of leaks in the thermal envelope, the operation of appliances with fans (including bathroom exhaust fans, range hood fans, clothes dryers, and furnaces), the operation of combustion appliances (including wood stoves), the leakiness of forced-air ductwork, and whether the ductwork is inside or outside the home's thermal boundary. A change in any one of these factors can change the air flow direction or air flow rate.

  2. Jeff Watson | | #2

    Thank you for your answer, Mr. Holladay. So I guess that leads me to my next question...what is the most practical approach in measuring air pressure differences within a house? If someone were to say, "when the air handler is on, why does all the return-air seem to flow into this room," how would you approach that in a methodical way? What factors would you attempt to minimize or rule out first?

  3. GBA Editor
    Martin Holladay | | #3

    It's hard to give a tutorial in residential pressure diagnostics in a few paragraphs. If you are serious about these issues, I suggest that you take a course through BPI or RESNET. You might want to purchase and study Residential Energy by John Krigger.

    A manometer is used to determine whether a space is pressurized or depressurized with respect to another space or with respect to the outdoors. Once you figure out these pressure differences, it's up to you to use your noggin to figure out what's going on.

    In most cases, this type of diagnostic work requires a blower door and a Duct Blaster.

    Q. "When the air handler is on, why does all the return-air seem to flow into this room?"

    A. Return-air grilles should be sucking, not blowing. If your return air system is delivering air to one of your rooms, something is indeed wrong.

  4. Jesse Smith | | #4

    It sounds like you're mainly trying to diagnose pressure imbalances caused by operating your furnace blower motor. You can do this by running the blower motor (ideally on the highest fan speed setting, which is generally cooling), then placing the hose from the manometer beneath closed doors to the various rooms in your house. It's very common to see very high pressure differences caused by the blower motor. For instance, I was in a house a couple of days ago which had a small utility room containing the air handler which was depressurized to over 40 Pa.

    In principle there should be no pressure difference across the room. IMO 1-2 Pa is friendly. Over 4 or so starts looking actionable if owners are likely to operate the house with the door closed. Pressure imbalances can significantly increase air infiltration/exfiltration and are therefore not desirable.

    Some gross generalizations - Air handling equipment in basements tend to depressurize basements/positively pressurize upstairs rooms because the return ducts are typically panned joist bays, which are significantly leakier than rigid ductwork. Don't jump to the conclusion that sealing those leaks is a good idea. Often the majority of return air is basement leakage. While this can tend to cause spillage at combustion appliances, it's also the only functional return air. Therefore sealing basement return leaks will reduce net system airflow in potentially destructive ways (eg. increase risk of freezing AC coil and furnace overheating). Often the best solution for these systems is to add a single, central return from the bottom of the furnace return drop to an open area in the floor above.

    Air handling equipment in attics generally uses central returns with dedicated room supplies. Bedrooms tend to be positively pressurized, but the house tends to be weakly depressurized because of the higher quantity of supply ductwork (and thus leakage) that occurs outside conditioned space. Once you've made the attic ducts tight, deal with pressure imbalances on a room-by-room basis using transfer grilles, jump ducts, or the really cheap fix of leaving the door open.

  5. GBA Editor
    Martin Holladay | | #5

    Thanks very much for your detailed, helpful answer.

  6. Flitch Plate | | #6

    Wouldn't it make sense to turn everything off, close all openings, and take a relatively neutral weather day and describe (measure) the natural force pressure patterns, first. Get that baseline of what the house “wants” to do in terms of positive and negative pressures, convection pathways, and ambient temperatures. Characterize your site specific weather patterns and prevailing winds (means, ranges and frequencies).

    Jesse’s description of how mechanical systems affect pressure is exactly the same as how natural pressures work in your house. The basement and lower levels are naturally negative while pressures increase with altitude inside the envelope. You mentioned stack effect, which has been amply discussed on this site (search the site). External loads by even the lowest speeds of bulk air pressures against or across the house increase positive pressures to windward and depressurize the house to leeward; the windward and leeward roof planes are negatively pressurized.

    By this basic pattern of forces, you can see the tendency of air to move through a building and also the effect of openings through various locations in the envelopes pressure boundary.

    Jessies point was that pressure wants to balance and will do so if it can. Room to room openings will allow that. A mechanical room 40 Pa below the pressure outside the door is a door too tight, without openings (external or internal envelope) to allow pressure balancing. Obviously this air handler will not work as it’s supposed to and envelope damage is likely.

    If we leave the doors and windows open in a house, pressures will balance naturally and dynamically according to conditions: sack effect, wind and temperature/humidity, day and night outdoor air flow patterns and temperatures, will all cause the breezes to blow, stack effect to be harnessed. Generally they blow in at the lower levels and out at the upper levels; although, here is the rub; most often conveniently ignored in on-line discussions about pressure: the in-flow of the air will be on the external higher pressure sides of the envelope and the out-flow will be on the negative pressure or leeward sides. So it’s multidimensional. Even with all openings shut, this windward side in-flow and leeward side out-flow will be in effect via infiltration/exfiltration pathways in the envelope and every time someone opens a door, window or vent.

    There is no such thing as a neutral pressure building; only neutral pressure moments. Buildings have site specific patterns (climate and location) as well as design feature patterns of pressurization and internal air flow pathways.

    Obviously, if you extract pressure from one location, nature’s famous operating principle of filling vacuums and moving cold air in behind and underneath more buoyant and warmer air will come into effect and the exhausted or forced-out air will be replaced from the shortest pathway if and wherever there is access (infiltration/exfiltration, from above, below or sideways, through leaky vents, open windows, around door jambs, through vent stacks and down chimneys).

    Before designing or resolving mechanically handled air systems or problems, the building and site natural flow patterns and pressures should be understood and optimized. There are debates on this site about whether we want a substantially positive or negative net pressure in a house. At the low pressures and rates of flow needed to refresh air in the increasingly tighter houses these days, and the use of natural solar heating and natural cooling designs, its gets progressively more difficult for the integrated fresh air/heating-cooling systems to overmatch the natural forces in and around the house; especially in the off-seasons or transition seasons.

    During peak heating and cooling seasons the mechanicals and artificial pressure making devices are turned up high enough to overmatch the natural pressures and create their own flows. Generally that means over pressurization of houses by accident or design; especially with forced air. It also means that centralized and concentrated heating/cooling/pressure making sources will create greater pressure differentiations, more air flow and dynamics and so less comfort, in the envelope.

  7. Jeff Watson | | #7

    Martin -
    I agree, I probably could use a course or some literature on the topic. I'm just a curious homeowner who likes to understand this environment we call "home." But the second question, I think you misunderstood me. The cold-air returns do suck. I was saying after closing the door to one particular room, I can feel a lot of air trying to rush into the room at the jambs, the bottom; almost like a poorly-weatherized exterior door.

    Jesse -
    Yes, that's pretty much what I'm trying to do; figure out the pressure imbalances. 'Cause I figure that could then help me better approach proper ventilation and probably explain temperature variation (to some degree). I bought a cheap Pyle differential manometer, but could not get consistent readings when I placed one hose under the door and the other in the "hallway". About the only consistency I had was if I stuffed one hose down a supply register and the other, out in the open towards the middle of the room. Just not sure if the manometer is sensitive enough or my approach is flawed.

    My house is single-story, so just a basement + main floor. All ducts are rigid square ducting that terminates up into the floor. The ducts are not sealed or insulated, although I did use some foil tape right at the air handler/plenum/filter as I figure those leaks probably have a big impact on losses. In hindsight I have been recently getting some short-cycling (furnace turns off before set-temp & turns right back on) but I figured I was using a too-restrictive filter, reducing air flow. You are right that some basement air is getting into the equation as when I'm down there, I can just "feel" air moving around me. Just makes me even more curious in figuring out the pressure differences.

    Fitch -
    Yes, that makes a lot of sense; I need to establish a baseline. As Martin did originally state, there are many factors that could affect the pressure differences in the house and I erred by seeking to minimze these factors as opposed to allowing them into the equation for better understanding. I understand stack effect, and I always figured if air rushes in as you open a window that this particular area is depressurized (within the envelope assuming the air coming in isn't a strong wind), and that I must therefore have a corresponding volume of air somewhere that is pressurized & being forced out of the envelope.

    To all of you, thnks a lot for the information; it really gives me something to go off & do more targeted research on. Great forum.

  8. GBA Editor
    Martin Holladay | | #8

    You wrote, "After closing the door to one particular room, I can feel a lot of air trying to rush into the room at the jambs."

    That is a symptom of a poorly designed return-air system. Ideally, every room in the house should have both a supply register also and either a return air register or a return-air path (for example, a transfer grille or a jumper duct).

    If the entire house has only one or two return air grilles -- and especially if the return-air grille is located in a room with a door -- the return system can be starved for air.

    For more information on this issue -- actually, a related issue (rooms with a supply register but no return air grille) -- see Return-Air Problems.

  9. Jeff Watson | | #9

    Kind of brings me back to figuring out the pressure differences :)

    But yes, every room with X supplies does have X returns, except for the kitchen(2 supplies) & bathroom(1 supply); these have no returns.

    I've read that those do not commonly have returns....but I guess I've always felt that was contradictory.

  10. Jesse Smith | | #10

    Jeff - It could be that your equipment isn't sensitive enough. 4 Pa isn't a lot of pressure. Running a smoke pencil (incense anyone?) will put you in the ballpark. Open the door to the point where the pressure equalizes to determine the required square inches of transfer grille.

    Pressure and temperature can tell you a lot about HVAC, but I don't think they can tell you how much to ventilate. In a nutshell, room-to-room variations in temperature occur when the required btu's differ from the actual btu's. Dealing with rooms under pressure may help alleviate comfort problems, but it may not.

    If your furnace is cycling on high limit, then you have a serious problem which you should address asap. Here's a fun magic trick: If the furnace is an upflow, hold a tissue over a supply duct. Then go to the basement and remove the blower cabinet door. Temporarily clamp the door switch. This will cause the blower cabinet to act like a big return duct. Take your tissue back to the supply duct. I would be willing to bet that you significantly increased airflow, which demonstrates return side restriction.

    Total external static pressure, pressure drop across filter and coil are more accurate tests. But this is a little less risky (no holes in your heat exchanger or evap coil). The best return fix is the one I discussed in my first post. Return side restrictions are most often caused by improper ducts around the handling equipment and restrictive filters.

    Bathrooms and kitchens shouldn't have returns. A 1" opening at the bottom of a door is usually good for ~50cfm, which is pretty typical supply air for a 5"-6" diameter duct connected to an unsealed basement duct system and a PSC motor.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |