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

Insulation and air conditioning

kgabbard | Posted in Energy Efficiency and Durability on

After reading this blog I feel much more informed in a general sense about my insulation and heating/cooling needs, but I am still trying to figure out what the return on investment would be for certain applications in my condo.

I live in the DC suburbs in a two level 930sqft end-unit townhouse built in the 1940s. The windows are high-efficiency, installed in the last 4-5 years. Most of my windows are East and South facing. My shared wall is to the West, so I get little or no western exposure. The storm doors are also supposed to help with heat loss (according to the person I bought from). To make my unit more efficient and more pleasant to live in I am researching the return on investment for improving either, or both, the insulation and the air conditioning.

Regarding the wall insulation. (a) There is no insulation currently (b) There is no air space in the walls to retrofit insulation, i.e. blow in insulation (c) there there is no vapor barrier (d) the attic insulation is pitiful looking cellulose. To improve my overall insulation I have come up with the following (which I have ordered in terms of what my research tells me will be diminishing return on investment):
(i) staple an attic entry bag over the stairs and caulk every possible air penetration into the attic I can find.
(ii) blow in new insulation on top of or to replace the existing attic cellulose to bring it up to an R value of at least 20-30 (instead of the highly doubtful R3 that it *may* currently have)
(iii) For the walls, the only option apparent to me is to to tear down the otherwise in good shape plaster, fur out the walls, and install a vapor barrier and rigid board insulation – this is what all other homeowners in this complex have done. Given that I have an end-unit this would be a massive undertaking and would destroy really nice plaster.

Regarding the air- conditioning. The fact that I have an end unit and have an attic makes for easy access for installing a mini-split. The two floors are equal size. The downstairs is effectively an open square layout. The upstairs is split into two bedrooms, a short hallway, and a bath. Currently, I have 3 new-ish PTAC like through wall units – one downstairs, one in each bedroom. For heat there are two baseboard electric resistance heaters on each floor (upstairs is in the master) – I will leave these in place for the handful of exceptionally cold days we get each year. For sizing the mini-split, I believe 18,000 BTUs has to be enough for the space and layout, but given the insulation issues and southern exposure, I am not sure. I especially want to get the sizing right because I want the units to help me dehumidify in the summer, given that where I am is often very humid.

Any recommendations on cost effective ways to counter these issues would be greatly appreciated.

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. iLikeDirt | | #1

    What are the walls made out of? Is this a plastered brick building or something? I ask because adding insulation to mass walls helps a lot more for heating comfort/performance than it does for cooling, and adding that insulation will entail destroying either perfectly good plaster, or a potentially architecturally valuable exterior (if you're even allowed to do that at all). Regardless, your approach of air-sealing and insulating the attic first is the right one. You'll realize a lot of bang for buck, and this will help you in the winter too. Marginal costs for this aren't that high, so it's probably worth it to go much higher than R-20 or R-30. D.C. doesn't have especially high daily temperature fluctuations in the summer so you won't be losing much nighttime cooling.

    How much input do you have over the roofing material itself? The color, material, and ventilation of the roofing material are all big variables that can substantially affect how much heat the attic gains, and for how long it continues after the sun goes down. The best-case scenario is a white metal roof installed over ridge-to-eave purlins. The worst case scenario is dark shingles.

    This might be a case where a DIY-installed radiant barrier in the attic could be effective.

  2. kgabbard | | #2

    I have not broken any wall yet, but I am given to believe that the walls are cinderblock with brick exterior cladding (brick has an acid-wash milky white color, also good for the heat). That said, I cannot do anything to the exterior. I also have no input over the roofing material - but the roof is a light-color clay tile. The attic only has gable ventilation - no ridge vent no soffits.

  3. iLikeDirt | | #3

    Sounds like the attic insulation plan is about all you can realistically do, insulation-wise. Add soffit vents if you're permitted to. Add a radiant barrier under the rafters if you can. Air seal, then pile the insulation deep. Prefer cellulose to fiberglass, which has worse performance during the cooling season.

    I don't recommend adding interior insulation to this wall, for a variety of reasons:
    1. It will be very expensive, and will likely never prove cost-effective. You'll also need to extend electrical boxes and plumbing lines, re-do the trim, handle door and window openings, pay for a full drywall job and paint job, etc. A nightmare.
    2. You will lose usable square footage
    3. It will make the bricks more vulnerable to freeze damage
    4. You will lose a substantial amount of interior durability by making the interior surface drywalled wood framing as opposed to plastered block. Now you have to worry about rot, mold, insects, physical damageā€¦

    I'll let others comment on the HVAC part of your question.

  4. kgabbard | | #4

    Thank you very much for your insight Nate! Will I get any noticeable comfort or savings from putting mylar backed honeycombed shades over the windows in order to keep in warm air during the winter and reflect the heat from solar gain during the summer?

  5. GBA Editor
    Martin Holladay | | #5

    Nate's advice is good. A few more points:

    1. In your climate zone (Zone 4), the 2012 IRC requires that attic insulation have a minimum R-value of R-49. That's a good R-value to aim for.

    2. Your worries about vapor barriers in walls are unfounded. There is no need for a vapor barrier in your wall in your climate zone, and in any case a vapor barrier wouldn't save energy.

    3. Here is a link to an article with attic air sealing tips: Air Sealing an Attic.

  6. kgabbard | | #6

    Hi Martin,
    Thanks for the link - look forward to reviewing. My mentioning of the vapor barrier was based on two concerns - (i) possible issues related to insulating without the barrier leading to condensation problems in the wall and freeze/thaw in the exterior brick (ii) any issues with trying to dehumidify a building with no envelope.

  7. Expert Member
    Dana Dorsett | | #7

    Even with no wall insulation an east facing 930' probably doesn't need 18,000 BTU/hr of cooling unless you're running ducts above the insulation in the attic, but you'll probably need that much compressor for heating, or even more, though there's still a realistic chance it could be less.

    If you have some wintertime power bills as well as a low-cooling load October or May bill (pick the lowest spring and fall kwh use billing periods) you may be able to get a good handle on heat load. Even with pathetic attic insulation the heat losses through the uninsulated walls of a 2-story condo are going to dwarf the losses through the attic.

    Most brick veneers on CMU walls in the 1940s were in fact cavity walls, with an air space between the CMU & brick. If the CMU cores are empty you can get a modest improvement in thermal performance out of filling them with perlite (or for a lot more money, non expanding injection foam). Knowing the as-is wall stackup and the U-factors of the replacement walls it's possible to run an I=B=R or Manual-J load calculation as a sanity check on any power-use load calc.

    It's unlikely you would have sufficient vertical space to get to R49 on the attic floor all the way out to over the top of the structural wall, but there may be enough if you stack some rigid foam in those areas. It's worth getting in there and measuring the height between the roof deck and the top of the exterior walls, and inspecting to see if the masonry cavity vents into the attic (or not) before settling on an insulation plan.

    Freeze/thaw spalling of exterior brick from wintertime interior moisture vapor drives is next to impossible in your climate, and no vapor barrier is necessary. If it's a cavity wall it's absolutely impossible (from vapor drives). Even interior latex paint on wallboard or plaster is sufficiently low permeance that exterior water drives in summer are easily handled by air conditioning. Air leaks are a primary driver of summertime indoor humidity issues, along with interior moisture sources.

  8. kgabbard | | #8

    Thanks Dana for your input! I am looking forward to going up in the attic to take a look. The vertical over two of the three exterior walls is de minimus - there is no overhang and the decking appears to terminate at the fascia board. I will have to go up to see if the CMU core does vent to the attic. Otherwise I will have to wait until the PTAC sleeves are removed. If I determine that the CMU core is both accessible from the attic and presently empty how does one actually get insulation into the CMU cavity - do you top fill with bags and bags of loose perlite?

  9. Expert Member
    Dana Dorsett | | #9

    Empty CMU cores are usually filled by pouring perlite or similar granular insulation, or by injection foam. With the pouring approach there are obstructions around window framing, etc, so you may not be able to get a complete fill without more serious surgery on the interior of the house, but the benefits are modest enough that it's not really "worth it" unless you were planning on replacing the windows. It's really more of a "better than nothing" sort of upgrade.

    When you pull the PTAC sleeves you should be able to determine if it's a cavity wall (likely), with an air gap between the CMU and the brick veneer. If it happens that the brick is tight to the CMU filling the CMU cores could become a problem, since the cores are the drying path and a partial capillary break from exterior rain/dew wetting. Perlite would become a capillary bridge, wicking more moisture toward the interior. With zero roof overhang significant amounts of rain wetting is guaranteed, which is why most mid-century brick clad homes were built with an air space between the brick and the interior layers.

  10. kgabbard | | #10

    I cannot see down into the CMU from the attic - no joy there. If I can determine the CMU configuration when I pull the PTAC sleeves, is there a way to get insulation in there that does not interfere with the brick wetting/drying system? Regarding interior surgery - given the building is of a colonial style I was going to "upgrade" to wainscotting to cover the holes left by the PTACs as well as any other holes I had to make to do other upgrades - so I have some leeway. Is there way to drill and inject insulation into the cavity? It seems like spray foam would interfere with the drying of any water penetration through the brick.

  11. Expert Member
    Dana Dorsett | | #11

    There are multiple vendors of non-expanding injection foams commonly used for insulating CMU walls. (The foam itself is essentially a re-formulation of the urea-formaldehyde foams that went so wrong in the 1970s. See: ) It may be worth soliciting quotes from installers. TryPolymer and Corefill-500 are two vendors that come to mind, but there are others.

    It takes about a 1.5" hole drilled in to the side of the CMU core to provide access for the injection tube. There is some shrinkage of the material as it cures, so it's not a perfect air-seal, but it impedes the convective flows within the cores substantially. The stuff runs about R4.5-R5/inch, but the thermal bridging of the CMU webs and mortar means the net improvement to the whole-wall R is much much smaller. There isn't a huge performance difference between perlite or injection foam, but there's a substantial cost difference.

    Non expanding injection foams are non-wicking but fairly vapor permeable typically ~15 perms @ 1", still ~5 perms at 3", and will never hit approach class-II vapor retardency at CMU core dimensions. It's altogether different stuff from 2lb polyurethane expansion foam.

    If you're stripping it back to the CMU and installing a wainscot, you can probably slip an inch of rigid polyiso under the wainscot and get better bang/buck than you'd get out of injection foam. Even if you're leaving the upper half of the wall with the original plaster, there's no reason not to go with rigid foam behind the wainscot. Heat loss is a temperature difference times U-factor times area type of thing- even an inch of polyiso would cut the heat loss per square foot of the treated portion of the wall by 2/3 or more.

  12. iLikeDirt | | #12

    I'm not sure I can recommend the idea of filling the block cores, even with perlite, because it probably still falls under the category of "will never pay for itself." Honestly, just filling the block cores with sand or mortar--anything cheap and massive--will help during the summer. More mass = lower thermal diffusivity, which is the missing link of summer cooling performance and comfort that's often missed by northerners hailing from cold, wintry climates. Thermal diffusivity measures how fast heat moves through something, taking into account the mass of the material and changing conditions, while R-value only measures the amount of heat that gets through per hour under steady-state conditions. A low-mass-mostly-insulating material (e.g. fiberglass insulation) will be allowing small amounts of heat through all the time, whereas a massive-but-low-R material (e.g. brick) will be allowing a large amount of heat through, but in a "wave" that takes hours to get in. Until it gets all the way through, little to no heat is actually entering because it is moving through the material slowly, not quickly.

    This is why mass materials can perform better than expected during the summer: if the time that it takes the "heat wave" to pass through the wall is greater than the amount of time that the sun is shining and heating the wall up, then the interior doesn't actually gain much if any heat through the wall, and when the sun goes down, the "heat wave" slows down, stops, and eventually reverses direction. The more mass in the wall, the slower the "heat wave" travels through it. The interaction of insulation with mass is a complex one, but by far the best place to put insulation in a mass wall is on the outside--an option unavailable to you. Because of that fact, I wouldn't expect that filling the cores with perlite will do a lot to help you in the summer. It may have a small positive effect in the winter, but it won't be very big because of the effect of thermal bridging through the webs that Dana mentioned. I'd say only do it if it turns out to be very cheap.

    Another thing I just thought of: if you don't have ceiling fans, get some! Mass walls and ceiling fans are a match made in heaven. It's also pretty shocking how cheap effective ceiling fans can be these days. I just recently installed three for $65 a pop.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |