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

Cooling a House Without Mechanical Air Conditioning

tx_mike7 | Posted in Energy Efficiency and Durability on

Hi all,

I’m looking for a little advise on how to keep a home cool without using a typical air conditioner. Things I’m already considering/pursuing:

High Thermal Mass exterior: ICF
building orientation: long axis +/-15 deg from east/west
window coverings/overhangs: 4′ overhangs on 2nd floor and a covered porch on first floor
Window surface area: no huge windows
ceiling height: 9-10′ ceilings
radiant barrier: in the attic space.  Attic is vented
Earth Cooled Tubes (More for ventilation than the cooling effect but I think it should help)
solar chiney effect: to help pull air through the earth tubes and vented out through the attic
chilled water radiant cooling: in concrete slab on first floor and through copper piping+aluminum panel on ceiling in second floor.

1) Am I missing anything
2) Will it be enough to stay comfortable (not looking for a perfect 72 degrees all the time but reasonably cool)


P.S. – Location is Northeast Texas, east of Dallas

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. exeric | | #1

    This is a laudable goal. I attempted the same thing. I live in a climate similar to yours but probably a little less extreme (northern California) I have a porch on the south side that stops direct sun into the house from that side. I also have a radiant barrier tagged to the underside of the rafters. It helps a lot. I also have loose fill cellulose on the attic floor. Finally, I installed a whole house fan and open the windows at night, which does the most of those three things to keep the house cool.

    Unfortunately, as the climate warms up not only are the average temperatures getting hotter, but the nights are getting commensurately even hotter. This pretty much renders whole house fans ineffective for days at a time. If you don't have cool air coming in at some part of the day high temperatures just roll over into the next with no relief. You have to have some way to actively cool the house, whether it natural like night flushing, or synthetic like a heat pump.

    I had high hopes for doing what you are attempting and somewhat regret some proselytizing I may have done here in the past. It might have been a good approach a few years ago, but at least in Northern California it doesn't work now as I'd hoped. You didn't mention a whole house fan. I can't imagine even a possibility of natural cooling working without that. Yes, I know WHFs don't work in humid climates, like many parts of Texas possess.

    1. exeric | | #2

      I should add that the best thing I've done after coming up short is to install solar panels. I live in a region with high solar insolation and I assume you do also. With that free electricity coming in I haven't felt guilty about increasing my carbon footprint because of the heat pump I've installed. Overall, my utility bill is much cheaper, actually zero, and I use the heat pump both winter and summer. The combination of solar panels and a heat pump is very effective and flexible, especially in areas like Texas and California. I hope you can get a grid tie-in agreement with your local utility.

    2. Expert Member
      MALCOLM TAYLOR | | #4


      "as the climate warms up not only are the average temperatures getting hotter, but the nights are getting commensurately even hotter."

      We are experiencing the same thing here on Vancouver Island. Close to the water we can still rely on late afternoon fog rolling in and cooling the house, but even just a few hundred meters further inland they no longer experience the benefit of that. What happens when the conditions we built to exploit no longer exist? I used to be a bit dismissive of the problems of cooling houses "don't build in areas that are basically uninhabitable". Pretty soon I'll be eating my words.

  2. paul_wiedefeld | | #3

    What’s the motivation? Will this home have electricity?

    1. tx_mike7 | | #6

      The home will have electricity. The motivation is twofold. One, reduce dependance on electricity. Two, to design a home that is built with parts easy to obtain locally (not dependent on parts from China, other supply chain issues), or parts simple to manufacture locally. For instance, passive cooling techniques will not break down or need to be repaired/replaced. Once the home is designed and built, it should not need much else to keep it relatively comfortable.

      1. paul_wiedefeld | | #7

        What will chill the water for the floor? How would thermal mass help you in Texas? Last, I assume you’re paying cash for this?

        1. exeric | | #9

          All really good questions. I don't think the thermal mass of ICF is going to help in a Dallas climate. Even if it did, the carbon contribution to the atmosphere of making the concrete in the ICF will just exacerbate global warming.

        2. tx_mike7 | | #11

          Still trying to decide on what will chill the water and if ground water would be cool enough on its own. Possibly doing a geo loop to dump the heat in the ground outside the house. The thermal mass would help keep the temp constant. We do have cooler nights here in the summers but this will be more for the spring and fall when the temperature is too hot during the day but too cold at night. Why do you ask about cash? Do you think a bank would have a hard time financing it?

          1. paul_wiedefeld | | #15

            I’d double check to see if ground water is allowed to be used like that. I think a bank would want their house to have AC in Texas.

          2. walta100 | | #16

            “Do you think a bank would have a hard time financing it? Tx_mike7”
            It was my experience that the bank and its appraiser saw zero value anything above code minimum windows insulation and HVAC equipment. The appraiser looks at your plans and gives the bank his opinion of what the bank could get if you disappeared and the bank had to complete and sell the house. The bank will do an analysis of the project and require no less than 20% equity on each and everyday of this project. In short if you want anything strange like your floor cooling or above code insulation you will need to fund those systems cost on top of the 20% equity the bank will need to see your down payment/land equity.

            I am not convinced your floor cooling idea is viable in that your local ground temp is unlikely be low enough to transfer much heat. Each time you transfer heat you need at least 15° to move much heat. If you ground temp is say 60° you might get 75° water in the loop, so you might cool the slab to 90°. To my ear the numbers don’t seem to work and run a big pump to move lots of water thru very long pipes.


  3. Expert Member


    You may find this article and discussion useful:

  4. Expert Member
    BILL WICHERS | | #8

    If you’re in an area with relatively low average humidity levels (I think that’s the case in Dallas, but not Houston), I would look into evaporative coolers. You can get whole house size evaporative coolers and they work pretty well, but they depend on low average humidity levels to be most effective. Note that such units basically trade water use for electricity use, so if you’re in an area with high water rates or water shortages, you introduce a new problem. Note also that while air conditioners require you to seal your house up to be lost effective, evaporative coolers are the opposite and require good airflow through the house.


  5. Robert Opaluch | | #10

    Most high performance building designers would suggest designing and building an airtight, well-insulated, well-shaded home that avoids direct solar heat gains in summer, with an ERV. The home won’t cool itself much in summer, but will avoid gaining as much heat to make it easier to keep comfortable, at lower construction and ongoing utility expenses. With a lower cooling load, whatever cooling system you use will be used less, and can be smaller in capacity. The home would be more resilient in a power outage, but would require electrical power to be most comfortable.

    Most high performance home designers would suggest adding more insulation and doing more air sealing instead of focusing on thermal mass in your climate. But you could attempt to capture wintertime solar heat gains or overnight cooling in Spring/Fall and use thermal mass to stabilize your interior temperatures, effectively storing that heating or cooling energy. It just not ideal for your climate and takes a fair amount of calculations and modeling to design it to work effectively.

    Attached are climate charts for Dallas from You can enter your exact location to get more accurate charts for your home’s exact location.

    Natural air cooling of the interior could work for Spring and Fall in Dallas. But for the “hot” period of June 4 thru Sept 18 (shaded in pink in the attached temperature chart) , the outside air rarely gets cool enough to cool the interior effectively.

    If you had substantial thermal mass (say, a concrete slab floor), you could open windows overnight to cool the slab, then keep windows closed during the daytime. But that won’t work unless you can get low temps in the 60’s or lower, which doesn’t happen June-Sept in Dallas. It worked nicely for me in Colorado, when low temps in summer tended to be in the 60’s F or cooler.

    Thermal mass inside the building’s thermal envelope will only stabilize interior temperatures. So interior temperatures might vary daily at plus or minus 5 degrees F instead of much larger variation without any heating or AC system running. If you can cool off the interior overnight with outdoor air temps in the low 60’s F, that could keep the interior cool naturally during the day, although it will still slowly warm up during the day. I assume this is what you are trying to do. It would work in your climate in Spring and Fall, just not in summer when you need it the most.

    ICFs could function as thermal mass, but since there is insulation on both the exterior AND INTERIOR, they won’t be as effective at soaking up interior heat as a concrete, concrete block, brick, or stone walls or floors with insulation only on the exterior. But ICFs would stabilize interior temperatures somewhat.

    Not gaining heat on the interior is likely your first priority. Its easier to control interior temperatures without as much thermal mass, since it takes a fair amount of energy/time to change the temperature of substantial thermal mass. If you can heat interior thermal mass (with solar gains) or cool thermal mass (with cool overnight temps) for free, that’s ideal. For your climate, your summer low temps are too high for cooling. If you have a cheap overnight electrical service charges, it might be advantageous to cool at night and less during the day. But designing a home to work well with thermal mass is not easy to do, even if your climate has sunny cold winters to capture solar gains in thermal mass, and cool overnight temps in summer to cool thermal mass overnight.

    Avoiding solar heat gains
    You seem to already be aware of the importance of avoiding solar heat gains, shading east and west windows, and having longer north and south walls and shorter east and west to minimize unwanted solar exposure in summer. Others may not be as familiar with ways to avoid unwanted solar heat gains. Generally, north facing windows in North America get minimal direct sunlight any time of year. So they would be good for avoiding solar heat gains during the hotter months. South-facing windows get little direct solar gain May thru July (when the midday sun is mostly overhead) but A LOT mid-winter when the sun is lower on the southern horizon (assuming your home doesn’t get shaded by other buildings or trees to your south). East and west-facing windows can get a very large amount of direct solar gain during the hottest months of the year. So minimizing east- and west-facing windows, and window sizes is helpful to avoid summertime solar heat gains. Roof overhangs, covered porches, trees, awnings, solar screens or anything else that will stop direct solar gain during the warmer months are helpful. South-facing windows need shading in late summer too. So more and larger north-facing windows, and fewer, smaller, or only well-shaded west and east facing windows would help keep the cooling load lower. You can calculate the BTUs of solar heat gain, using this web site for data for your location:

    Energy efficient building envelopes
    Of course an airtight, well-insulated building will make it easier to keep your interior temperatures less affected by outdoor temperatures (and humidity). I assume you are already focused on airtightness and more than code minimum levels of insulation. This is likely the best investment to make a home resilient to electrical outages, lower electric bills, and keep utility and building costs reasonably low.

    You may or may not be familiar with energy-recovery ventilators (ERVs). An ERV can bring fresh air into the building that is conditioned by an exhaust air stream. So instead of bringing in outdoor air that’s hot and humid during the summer, the incoming air stream temperature and humidity will be reduced, by dumping that heat and humidity into the exhaust air stream.

    1. exeric | | #12

      "Of course an airtight, well-insulated building will make it easier to keep your interior temperatures less affected by outdoor temperatures (and humidity). I assume you are already focused on airtightness and more than code minimum levels of insulation. This is likely the best investment to make a home resilient to electrical outages, lower electric bills, and keep utility and building costs reasonably low."

      I think what was needed to be said was in this one paragraph. Concrete has a huge up-front cost in embodied carbon. Up-front CO2 costs of carbon intensive materials has disproportionate costs for climate change. See this:

      Txmike can do everything he needs to do cheaply and efficiently, and be a better steward of the planet, without using ICFs. Just use the standard green building practices taught every day at GBA and apply them to his climate. There's no need to reinvent the wheel, or even worse, reinvent the horse and buggy.

      1. Robert Opaluch | | #13

        I would agree with your analysis, although you have to choose a foundation for a single family home. Concrete tends to be the common choice in the US, including GBA articles. It seems that GBA has endorsed concrete foundations, and said little about zero concrete alternatives in recent years. Note these recent GBA articles on foundations:

        The only recent GBA article I found focused on zero concrete construction was this BS&Beer Show article on Permanent Wood Foundations:

        If there is a consensus among high performance building designers and builders on GBA, it would be to use concrete for perimeter foundation walls, or a frost-protected insulated concrete “raft” or thickened edge slab foundation. (See the various construction details in the GBA library.) It also would be to avoid poured concrete full basements and ICFs, though there is a ICF foundation in that GBA library and the first article above. One could use steel or treated wooden piers, or all-treated wood foundation walls to avoid concrete altogether, but I don’t notice that being recommended very often. Maybe more articles on that topic would be helpful to lower our use of concrete.

        1. DC_Contrarian_ | | #14

          There was a good article about going slab-less but I don't have a link.

          Where I am, for the actual foundation concrete is just about the only material local code officials will accept. Wood, even treated wood, is not allowed below grade. You can do helical piers, but those commonly use a grade beam -- an 18x18 concrete wall that sits above grade -- to give lateral stability and give the walls something to sit on, so you end up not using appreciably less concrete than a shallow foundation.

        2. exeric | | #26

          I've been thinking about what you said and have tried to resist arguing with you. To no avail. You missed completely the most important point. Yes, we have problems with using concrete in foundations because it is bad for CO2 in the atmosphere. So, there is "that" problem. But I never said anything about foundations and did not intend to. That is the status quo situation that must be tackled eventually.

          But what you are saying is that ADDING concrete in the form of ICF is no worse than conventional construction. No, that's wrong. That is exactly like saying we can lower gas mileage requirements in ICE vehicles because there are already ICE vehicles on the road getting bad gas mileage.

          OK, so what if ICF is in the GBA library. GBA isn't here to substitute poorly thought-out solutions and it generally doesn't. You should be using your analysis capability to understand that ICF isn't a good solution for residential construction unless it is there for a specific reason - for instance to stand up to tornados or hurricanes. Maybe the OP is actually worried about that so it would be a good solution. He never said that though.

          It wouldn't be important to point this out if there weren't already good solutions that DON'T use the extra concrete that ICF construction imposes over and above standard construction practices. How can you ignore that? That is it in a nutshell. ICF construction creates higher global warming potential than standard construction and does it for no reason. (Unless the OP has neglected telling us something about his reasons for using it)

          1. Robert Opaluch | | #29


            You claimed "what you are saying is that ADDING concrete in the form of ICF is no worse than conventional construction." I never said that, and don’t believe that, so don't claim I said it. I don’t recommend ICFs or full basements. But GBA does in their library. So why don't you argue with them to remove ICFs, or anything else you think is not "good enough" in your opinion. GBA seems to want to be a “big tent” that accepts a lot of “pretty good” gradual improvements that are not as stringent as Passivhaus/PHIUS and PETALS.

            What I said originally was "Most high performance home designers would suggest adding more insulation and doing more air sealing instead of focusing on thermal mass in your climate." So I’m actually saying politely…ICFs won’t work well. Skip the ICFs and focus on air sealing and insulation to lower the cooling load. And maybe focus instead on some of Mike’s other ideas that I didn’t comment on.

            I don’t dispute that concrete is a big contributor to global warming. I actually said “Maybe more articles on that topic [avoid concrete altogether] would be helpful to lower our use of concrete.”

          2. exeric | | #30

            I didn't see the reason for you to respond in comment 13 to me by talking about concrete in foundations. It just diluted what I was saying that was specific to ICF. Sometimes one has to give credit to other individuals to know general information. If one doesn't do that it just adds extraneous information. And then other people cue off that "extra stuff" and the whole conversation goes sideways, which seems to be what happened. I don't see that everyone here does not already know about concrete. GBA does and they have a legitimate reason for including ICF info if only to use it for construction in hurricane and tornado country. Also, we all know there is legacy information everywhere on the internet that is not always up to date, and I just can't fault GBA for that. Yet another case of not giving people credit for knowing general information.

            But I'm glad we agree that ICFs generally aren't a good idea to moderate temperatures when there are so many better ones in the tool kit.

  6. DC_Contrarian_ | | #17

    On a day where heating is required throughout the day, you can reduce the need for heat by reducing heat losses -- ie better insulation -- and capturing more solar. Similarly, on a day when cooling is required your choices are better insulation, and rejecting more solar. It's on days when it's cool at night and hot during the day -- shoulder weather -- that things get interesting.

    I did a simple model to try and illustrate what happens. I assumed a day where it's 60F at midnight and 80F at noon, and for simplicity I assumed that the temperature curve follows a sine wave shape. I tried to model what would happen to the interior temperature if there was no heating or cooling. There are two things that determine that -- the insulation level of the house, and the heat capacity of the house, how much heat it takes to change the interior temperature. I picked numbers that I thought would be reasonable for a 2,000 square foot house. If that house is insulated to a level where to maintain 70F inside at 20F outside requires 25,000 BTU/hr, it would have an overall insulation level of 500 BTU per degree F. I estimated such a house would have internal structure and contents of 25 pounds per square foot, or 50,000 pounds, and that that stuff would have a specific heat of 0.5, so the whole house would have a heat capacity of 25,000 BTU per degree F.

    To model the house, I divided a day into 144 ten minute intervals. For each interval, I calculated the heat flow, by taking the difference between the inside and outside temperature, and multiply that by the insulation level (500BTU/degree). Then I calculated the temperature change, by taking the heat flow and dividing that by the heat capacity of the house.

    I've attached a graph that shows the results. The inside and outside temperature graph have two differences. First, the inside temperature fluctuates much less, while outside goes from 60F to 80F the inside only goes from 65.8 to 74.2, a swing of 8.4 degrees. Second, the time of the maximum and minimum is time shifted, occurring 4:20 after the outside temperature peaks.

    If you've lived in a conventional house in shoulder season this behavior should seem familiar to you.

    1. DC_Contrarian_ | | #18

      Then I did a few changes to see what the impact of varying different assumptions was.

      First was to double the insulation level, keeping heat capacity the same. The basic shape of the curve didn't change, but this change decreased the temperature swing from 8.4 degrees to 4.5 degrees, and delayed the temperature peaks to 5:10 after the outdoor temperature peaks instead of 4:20.

      Second change was to double the heat capacity and keep insulation level the same. This would be analogous to adding 50,000 pounds of material to the interior of the house. This change had exactly the same impact as doubling the insulation level -- reduce temperature swing to the same 4.5 degrees, postpone peaks to the same 5:10.

      The third change was to model an ICF -- keep the insulation level the same as the original model, but double the heat capacity, and add that heat capacity not to the house but to the center of the insulated wall. To simplify calculations I assumed the heat capacity material had no insulating capacity of its own, so that it was a constant temperature. The mock ICF did not damp temperature swings as much as the same heat capacity on the interior of the house -- it was 6.2 degrees, as compared to 4.5 degrees for the same capacity on the interior and 8.4 for the base house. But it also had the longest time shift, of 6:20.

      If anyone is interested I will post a Google link to my spreadsheets.

      1. DC_Contrarian_ | | #22

        I have to say that the result for the ICF surprised me. My intuition is that adding the heat capacity inside the wall rather than inside the house would be more effective, as the name of the game is maximizing temperature swings to maximize heat flows, and the wall center is going to have bigger temperature swings than the house interior. That turns out to be true, but only by about a degree, which turns out not to be enough to matter.

  7. tx_mike7 | | #19

    I wanted to say thank you after every comment but I didn’t want to clutter the discussion so I’ll say it here now. Thank you so much to everyone who took the time to answer my questions! I treasure the knowledge that you have so freely given. I think I am mostly on the right track and will look to make some tweaks as I plan the house. Again, thank you, thank you, thank you!

  8. tx_mike7 | | #20

    Walta100, can you point me to some reference materials regarding your heat movement numbers? This has been one of my larger question marks that I’d like to gain a little more knowledge about. Thanks in advance!

    1. DC_Contrarian_ | | #21

      Mike --

      I strongly recommend that you try to engineer the house before building anything. There are lots of houses built with "alternative" ideas that just don't work, and if an analysis had been done before they were built it would have shown that they were never going to work. The reason I say that is that there's not much quantitative -- numbers-based -- in what you're proposing, and when a conventional house is built the mechanical systems are designed in a very-much numbers-based way.

      To answer your direct question, the unit of heat used in the US is the BTU, British Thermal Unit, it's the amount of heat needed to change the temperature of one pound of water by one degree Fahrenheit. It's used for both heating and cooling and heat flows are customarily measured in BTU/hour. For a system where you're pumping water into a building to cool it, the amount of cooling you're going to get is determined by how much the temperature of the water changes and how much water you're pumping. As an example, let's say your water is at 60F when it comes out of the ground and at 70F after it's been through the cooling loop. Let's say your flow is one gallon per minute. One gallon weighs 8.3 pounds, so that's 8.3 pounds per minute or 498 pounds per hour. With a 10F temperature change that's 4980 BTU per hour.

      Is that good? It depends on what your house needs. There is a process called a Manual J that calculates the heating and cooling requirements of a house. You put in the dimensions of the house, details about construction and the local temperature extremes and it gives you peak loads. There is also similar software from the US Department of Energy called BEopt that allows you to model a house.

      I strongly recommend that you model the house before doing anything. There are people here who can walk you through every step of the process.

      1. tx_mike7 | | #23

        Thanks DC!
        I have heard of the Manuel J but have not used it yet. Any recommendations on what the best software for it is? My background is in engineering so numbers-based is in my wheelhouse. Unfortunately Thermodynamics was not a class I had to take so that will be a bit of a learning curve.

        1. DC_Contrarian_ | | #24

          I'd start with BEopt. People here can help you if you get stuck.

          1. Expert Member
            Dana Dorsett | | #27

            +1 !!

            Manual-J doesn't model thermal mass effects very well. BeOpt isn't perfect at it either, but it's pretty good at figuring out best bang/buck (which is outside the wheelhouse of what ACCA Manual-J is trying to solve.)

          2. DC_Contrarian_ | | #28

            Manual J is about comfort, it's about making sure you're not too hot on the hottest days and not too cold on the coldest days. But the methodology is useful for trying out different changes and seeing how they move the dial.

          3. tx_mike7 | | #33

            I think I’ve got all my inputs into BeOpt and I’ve run the program. I’m not quite sure how to read the outputs and the different graphs. Anyone know what I’m looking for? I was expecting something like, “this house will have a a heat gain of…”. And from there I would be able to figure out how much heat I would need to remove from the house.

          4. DC_Contrarian_ | | #34

            Go to the Output screen. In the graph window click on the down arrow and select "Graph Type" and set it to "HVAC Capacities." This shows your heating and cooling size. It's not as detailed as Manual J but it gives the number you need.

            You also need to look at "Loads not met."

  9. walta100 | | #25

    If you want to use beOpt you should invest the time to watch the training videos

    Do you live in the part of TX where the humidity is so thick you can cut it with a knife?

    Please be sure to avoid HVAC equipment and duct work in the attic or is it required by code in TX? LOL


    1. tx_mike7 | | #31

      Haha! No not required but it does seem like everyone has done it. Thank you for the links. I am working my way through the videos and have started playing with BEopt. Is there a way to add new selections? For example, they have options for 1’, 2’, and 3’ overhangs but I was hoping for a 4’ option.

      1. DC_Contrarian_ | | #32

        If you go into the options screen and pull up overhangs and right click on an option one of the choices is "Option Manager." What you need to do is copy one of the existing options and then edit it, BEopt won't let you edit the system options.

  10. Tim_O | | #35

    I recently was in Germany through the heat wave where temps hit 100*F during one day. As with most homes in Germany, our family's house does not have A/C. It got up to 81* in the bedroom by the end of that day (about 75 downstairs though). And that night it only got down to the mid 70s outside. The strategy they use is to ventilate as much as possible and close off sunny sides of the house with exterior blinds. It really wasn't that bad going through the heat wave.

    With all that said, the humidity in Germany was a lot less than what I am used to here in the Midwest. As it got towards the 90s in Germany, the humidity really dropped, I think it was 15-20% that day. In Dallas, I think you also deal with higher humidity. I find, as long as the indoor humidity stays a little lower, I can tolerate much warmer temperatures. I think your strategies don't have anything to take care of the humidity, and that will be the main problem.

  11. benwolk | | #36

    As a Passive House evangelist, I agree with everything that Robert said above about building an airtight house and insulating well to help buffer the temp swings that you see in a normal house. An airtight house with a good ERV/HRV system will go a long way towards ensuring great IAQ, especially as our outdoor air quality gets worse. I agree with your desire to reduce your reliance on complex mechanical systems and design for resiliency.

    I would suggest planning on your climate getting worse in the future, so in your energy modeling with BeOpt, you might want to use a more extreme climate zone to design your house and test how it performs in both your current climate and an assumed future scenario.

    I would also look into vernacular architecture that used passive strategies to provide cooling without A/C. These probably had greater temp swings than we are currently comfortable with, but the strategies can still be a source of inspiration and integration into our modern comfort expectations.

    Here's some good resources I found:

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |