Heating Upstairs Versus Downstairs – Part II

This is a continuation of the previous blog post about where heat is used, first floor versus second floor, in two story buildings with an open staircase, like most homes.

I finished tests on the high-performance new home that I mentioned in the first post. Specifically, this newer set of tests were with the interior room doors CLOSED inside the house. Three upstairs bedroom doors were closed, and one downstairs guest room door was closed. This house has a single heat pump in the open downstairs living area, and five electric resistance baseboard heaters:  one in each of three upstairs bedrooms (with doors), one in the downstairs guest room (with a door), and one in the downstairs entrance (no door between this area and the open living area).  Previously, with all interior doors open, I had found that 82% of the heat load is downstairs, when the five electric heaters are on a medium temperature setting, and 100% of the heat load is downstairs when the five electric heaters are on a low temperature setting. Again, these results were “open-door”.

With all interior doors closed, 77% of the heat load is downstairs, with the electric heaters on a medium temperature setting.  When the electric heater temperature setting is low, we again see 100% of the heat load downstairs – no heat is consumed upstairs!  A sample upstairs bedroom air temperature dropped about 7 degrees, between the test with the heater set on “medium” to the test with the heater set on “low”:  From a comfortable 69 degrees on “medium” to a cooler 62 degrees on “low”.  But 62 degrees is still pretty comfortable for a bedroom, and certainly is fine for unoccupied hours, during which time 100% of the heating is being carried by the downstairs system.

In summary, EVEN when interior doors are closed, most (77%) of the heating load of this two-story house is being carried by the downstairs heating delivery system, when upstairs and downstairs air temperatures are comfortably at their desired levels.  And ALL of the heating load is being carried by the downstairs system when the upstairs temperature is 7 degrees lower.

These results are generally consistent with what I have now seen in three other buildings. I hope to test a fifth building in January.

To be clear about what we’re seeing here:  I’m not saying that the heat LOSSES are predominantly downstairs. The heat losses are likely very much still happening in each room, much as we might expect – heat is being conducted out of upstairs walls, upstairs ceiling/roof, upstairs windows, etc. (Other than infiltration, which is likely concentrated downstairs, due to the stack effect.)  But what is happening is that the loads are being internally TRANSFERRED from upstairs to downstairs, due to cold air falling downstairs, from the upstairs. In addition, heat from downstairs is being delivered upstairs due to warm air rising.

Cold air becomes a load on a heating system when it hits a specific heater. Imagine a building with hot water baseboard radiators. With cold air falling from upstairs to the downstairs, down the stairs, this cold air never “sees” an upstairs radiator. Instead, the radiator seeing the largest load is the one in the room at the bottom of the stairs, where the cold air comes tumbling down. I’m guessing that some cold air, traveling at floor level, can even bypass this downstairs room and reach other rooms and other heaters downstairs. So, low spots in a building are likely real problems, such as sunken living rooms.

This “rising-warmth and falling-cold” phenomenon does not affect the overall heating load in a building. So, when we size the overall heating system, our OVERALL calculations are not affected.  The boiler or furnace or heat pump size does not change. BUT, it DOES affect the heating DISTRIBUTION system – where the heat is delivered. We have probably been always under-sizing our downstairs heating distribution systems and over-sizing our upstairs heating distribution systems.

Another important thing to be aware of is that this phenomenon (heat rising, cold falling) happens 24/7, not just when the heating system is on. It probably varies in intensity when the heating is on (stronger due to more warm air in a building when the heating is hot, weaker between heating cycles). But it is likely still happening 24/7.

It’s also important to note that we don’t yet know the impact of a variety of other factors, such as outdoor air temperature, the temperature of hot water radiation or forced air heating system warm-air temperatures, the type of heating system, the location of heating distribution, the size of door undercuts, and more.

I’m just starting to wrap my mind about the implications of 80% of a two-story building’s heating load ending up downstairs:

  • For new buildings, we absolutely must plan on separate temperature controls for upstairs and downstairs, as I mentioned in the first blog post. The biggest takeaway.
  • For buildings with a single temperature control, this explains overheating upstairs and uncomfortably-cool downstairs.
  • For buildings with a single temperature control, there are opportunities to save energy and improve comfort through rebalancing. Though rebalancing may not be easy. I can even imagine strategies such as removing fins from upstairs fin-tube radiation, or otherwise reducing heat output.
  • As I mentioned in the previous post, we can consider highly-affordable heating strategies for both new buildings and existing buildings, such as heat pumps downstairs (maybe even a single heat pump, just in the living room), and low-cost electric resistance heat in upstairs rooms. This will need to be combined with educational strategies, encouraging building occupants to keep the temperature in rooms with electric heaters as low as is comfortable, especially when not occupied, and keeping interior doors open when possible.

I’d like to also issue a couple big cautions:

  1. We need to be VERY careful not to over-optimistically respond to this data. For example, I would NOT yet try to design buildings without heat upstairs. You risk ending up with a building that is cold upstairs.
  1. I speculate that high-performance buildings may be more at risk for comfort issues relating to this phenomenon. In Green Building Illustrated, I wrote in Chapter 18 that “Green buildings may be even more vulnerable to certain kinds of failures than conventional buildings.”  I believe this may be the case with this “rising-heat/falling-cold” phenomenon.  Let’s take a hypothetical old house with 50,000 Btu/hr design heat loss upstairs and similarly 50,000 Btu/hr design heat loss downstairs. Now let’s ASSUME that the “rising-heat/falling-cold” phenomenon is CONSTANT in magnitude, just an assumption, pure speculation, and that its magnitude is 10,000 Btu/hr in upstairs load being transferred downstairs by falling cold air. Happening 24/7 all winter, this small but constant transfer of heat could well add up to the large imbalance in delivered heat downstairs, versus upstairs.  But the COMFORT impact at design conditions is small, in this old house.  Let’s say the downstairs heating distribution system is already over-sized by 20%, and so is 60,000 Btu/hr: It is readily able to handle the additional 10,000 Btu/hr of cold that is tumbling down the stairs, on top of its own 50,000 Btu/hr downstairs heat loss load.  NOW, how about a high-performance house, with 20,000 Btu/hr heat loss each, upstairs and downstairs?  If the SAME 10,000 Btu/hr of the upstairs load is tumbling downstairs as cool air, the downstairs heating distribution system is going to be challenged in mid-winter: If the downstairs system is 20% over-sized, so is 25,000 Btu/hr in capacity, it’s going to struggle to meet its 20,000 Btu/hr downstairs load PLUS the 10,000 Btu/hr that’s tumbling down from upstairs. It’s going to get cold downstairs!

Now, remember the “wet hand” test from the first blog post in this series?  Well, this morning I took three work colleagues to the top of the stairs of our two-story office building, and I asked them to each dip a hand in water, and then hold it above their heads near the ceiling of the stairway.  I asked them what they felt:  Warm air rising, no surprise.  Then I asked them to put their wet hands down near the top step at floor level. I don’t think it’s an understatement to say that they were stunned at how much cold air was CASCADING down the stairs. Try it!



2 thoughts on “Heating Upstairs Versus Downstairs – Part II

  1. Both posts were interesting. I will have to try the wet hand experiment this weekend at my house.

    It might be interesting to run some CFD models (computational fluid dynamics) to see if they substantiate the magnitude of the effect you are describing. It would also give you some cool animations (pardon the pun) of the phenomena for your blog.


  2. Let me know how the wet hand experiment goes. I agree that a good CFD model would tell us a lot. It would be helpful to understand parametric impacts (effect of heating temperatures, effect of heating distribution locations, effect of ceiling heights, and more). It would also allow us to look at related issues, such as interaction with the attic and basement. Ian


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s