0 Helpful?

1952 Cape Cod


First-time poster. In the past 2 months, I’ve read a hundred or so GBA articles and comment threads; I’ve been amazed by the expertise and generosity of the community. I’ve also read extensively at buildingscience and finehomebuilding -- I’m new to but very interested in building science. Here’s why:

We just purchased our first free-standing home, a large 1952 Cape on a great piece of property in Maryland (CZ 4). There is too much humidity in the house (inside fountain, which we will remove.) The house is need of air sealing and insulation throughout. (Previous owners did not update air sealing or insulation over 20+ years of ownership.) There are two vented, uninsulated basement crawl spaces. Energy audit/blower door found 4400 sqft of conditioned space and leakage of 7330 cfm50.The HVAC is currently 24 yo oil (no natural gas available in neighborhood) with central air (2 zones, one unit 1994 and one 2015). The smooth front roof is slate, has 2x6 boards as roof deck, and has 5-10 years of life left. The rear roof (2 walkout dormers) is mostly slate; this roof deck has been updated to plywood. Both of these roofs are vented with gable vents – no ridge vent. There is also a rear flat roof over an addition, w/EPDM and poor insulation. Other than these issues, the house is a model of green efficiency!

After reading feedback provided to other posters, I’m including full detail on multiple aspects below.

I have used the search function and read every available article on the below topics. I would greatly appreciate your input on my personal situation. I am happy to post photos if helpful and will post results as we move ahead. I’m interviewing insulation contractors now, which is quite a learning experience.

Can you please confirm or improve my tentative plan, below? Our objectives are to 1) increase comfort and 2) significantly reduce operating expenses.

Thank you.

1. Attic.
Current: There is currently maximum 2” of blown in fiberglass on the floor. Air handler (AC only) is in attic. 20yo ducts are decent shape but likely very leaky and need to be better insulated.
Plan 1: Remove/vacuum blown insulation. Air seal all seams with Great stuff or caulk, especially the top plates and fixtures. Blow in 16-18” of cellulose.
Plan 2: Either majorly seal and bury the ducts in insulation, and/or move these downstairs when HVAC is updated (see below).
Question: What sealant is recommended?

2. Cape kneewalls.
Current: Kneewall space is unconditioned. Floor joists are wide open. Rafters have 6” of available space for insulation. Roof deck looks like original 2x6” boards. Roof is slate and might need replacement within 10 years. (We will likely switch away from slate.) There is 2-3” rockwool insulation in the kneewalls and under floorboards.
Plan 1: Make this conditioned space. Motivation is 1) maximize performance and cost savings and 2) in future likely move HVAC unit into this conditioned space. We do not need this space for storage.
Plan 2: In future, consider new HVAC unit in this space, for an upstairs zone, either sealing/insulating/burying existing ducts in attic, or installing brand new ducts under 2nd floor (so that supply vents would all come from floor).
Question: How specifically – what technologies – to bring this inside the thermal envelope, given the 6” maximum space between rafters and hope to get close to R49?
Question: Any thoughts on the ducts?

3. Slopes.
Current: Rockwool stuffed in first 6” of slopes.
Plan: Install block and seal at the top of the kneewall. Insulate between that block and the attic floor. Install block and seal at the attic floor prior to blowing cellulose.
Question: How specifically – what technologies – to block, seal, and insulate here? Closed cell foam? Dense pack cellulose? The rafters will hold 6”, and there might be a couple more inches until the drywall.

4. Basement
Current: The basement is finished, but there are two vented, uninsulated crawlspaces on either side. Uninsulated ducts are in these vented spaces, (!) The floor joists are not fully accessible.
Plan: Block and seal the grates/vents to outdoors. Seal what is possible at the floor joists. Air seal and insulate the concrete walls. See HVAC, below.
Question: How specifically – what technologies – to block, seal, and insulate here? Seal then fiberglass batts? Rigid foam board?

Current: Oil furnace heats whole house. 2 AC zones. Zone 2 has separate ducts in attic.
Plan 1: I was leaning toward geothermal, contingent on the 30% tax credit returning, but having read GBA on this topic am now thoroughly confused. The house layout would be great for mini splits – it looks like a rancher with a partial second floor.
Question: What is current thinking on geo vs ductless?

6. Other notes/questions:
a. I’ve read Martin and others state that external foam board is the best way to seal and insulate, eg in kneewalls where retro sealing can be tricky. External foam board might be possible on the front slate roof, but that has another 8 years or more of life in it (so we don't to replace it now). On the rear roof (slate w dormers), I don't think external foam board will be possible. For both of these reasons, I’m thinking seal/insulate from the inside.
b. The flat EPDM room will need to be replaced in 10 years. I was thinking of bringing to R49 with external foam board at that time.
c. We’re just updating oil hot water to hybrid heat pump.
d. Is there a specific technology or product to seal recessed lighting?

Thank you again.

Asked by Emerson W
Posted Dec 11, 2017 9:02 PM ET


74 Answers

Select your preferred way to display the comments and click "Save settings" to activate your changes.

I am not a pro, but can start the conversation. Will any of this be DIY, or all contractor work?

I would look at encapsulating the crawl spaces as a good bang for buck. Air sealing the attic(s), knee walls etc to define the envelope boundary would be an important chunk. I would imagine that all of the complications are what makes conditioning the attic appealing, but don't know how to accomplish with skip sheathing. Cost is high, but accomplishes air sealing and insulation.

I'm a mini split convert. Most are heat pumps that will put out heat far beyond previous generation heat pumps. I do not care however for a multi splits complexity and high minimum output. In a 4000+ sqft house the high min modulated output might be appropriate. However, until decent air sealing has been accomplished a single point heating and cooling source may not meet your comfort goals. Namely temperature variations etc. Mini splits can also be purchased in a ducted air handler. Since both units are working(and it being winter) you should have plenty of time to research options.

Pros here can run numbers and have the knowledge to guide you on geothermal. Of the installations I've seen, none were designed well or have a realistic payback. Reducing the load by completing retrofit work is inversely proportional "making" a geothermal system work in my opinion.

As to my only area of expertise; Led retrofit kits can offer fairly decent air sealing compared to incandescent trims. If the can lights are in an area of limited insulation depth, "wafer" LED's can usually be put in place of the existing drywall holes. The largest part is 1.3" thick and the light can normally straddle a ceiling joist if you add lights or move existing locations. The "best" solution is to tear down the ceiling and install airtight electrical boxes. There are lights that look vaguely like a can light and are accommodated by the aforementioned boxes.

Answered by Ken cohenour
Posted Dec 12, 2017 3:34 AM ET
Edited Dec 12, 2017 3:46 AM ET.


I suggest that you post one or two questions at a time. Most GBA readers will ignore posts like yours, because you are asking too much.

My first reaction, as a former roofer, is that you must have been given bad advice about the slate roof. A slate roof is a rare and wonderful asset. Unless the original roofer made major blunders, there is no way that slate installed in 1952 "has 5-10 years of life left." What does that even mean? A slate roof can easily last 150 or 200 years.

Answered by Martin Holladay
Posted Dec 12, 2017 5:11 AM ET


Although you claim to have read up on the issues you ask about, there are articles on GBA that address your questions. Here are some links to get you started:

1. Crawl spaces: Building an Unvented Crawl Space.

2. Attic air sealing: Air Sealing an Attic.

3. Duct sealing: Sealing Ducts: What’s Better, Tape or Mastic?

4. Creating a conditioned attic: Creating a Conditioned Attic.

5. Cape Cods and kneewalls:

“Two Ways to Insulate Attic Kneewalls.”

Insulating a Cape Cod House.

Answered by Martin Holladay
Posted Dec 12, 2017 5:19 AM ET


@Ken, Thank you for your thoughtful reply. If you removed the ceiling to install airtight electrical boxes, would you also insulate from below at that time? Or, wait 10y until we replace the EPDM, and replace boxes/lights and add external foam board from above?

Answered by Emerson W
Posted Dec 12, 2017 7:34 AM ET


@Martin, I want to make it easy for you to help me. Let's please discuss the impact of the roof here. Then, I will create separate posts for he remaining questions.

We have received conflicting guidance on the slate roof. I've posted two photos below. Our inspector said 20+ years left (with a 75 year life expectancy). A slate roofer (who sell roofs) said 5-8 years on the front, and some spots on the back need major repair or possible replacement sooner. The flat EPDM probably has plenty of life, but needs more insulation and is amenable to exterior foam board.

Given that the roof is slate, what approach/products should I use to condition the knee wall area and slopes? How might this impact future roofing repair or replacement? I've read your cautions re: closed cell foam beneath slates and have also just ordered the Slate Roof Bible. Joseph Jenkins has posted here to attach a venting baffle behind the 2x8's, and spray foam under the venting baffle. I just want to make sure I understand.

All of this work, excepting minor sealing and blown in cellulose, will be performed by professional contractors.

front_flat_roof.jpg rear_dormer_roof.jpg IMG_4647.JPG
Answered by Emerson W
Posted Dec 12, 2017 7:57 AM ET
Edited Dec 12, 2017 8:17 AM ET.


I may have overreacted to you comment about the slate roofs. But I have seen many examples of perfectly good slate roofs being ripped off by ignorant roofers (or knowledgeable roofers who are greedy, because used slate is valuable).

Some roofers advise replacing a slate roof simply because they are inexperienced at flashing replacement.

Less expensive types of slate can eventually deteriorate due to freeze/thaw damage -- although Maryland isn't a particularly harsh climate -- and it's possible that your slates are deteriorating. I would start by asking your roofer: Exactly why are you recommending that I remove the existing slate?

Answered by Martin Holladay
Posted Dec 12, 2017 8:36 AM ET


Thanks, Martin. I'm not sure if the photos help you or not, but that is the roof. I also have detailed pics, if you'd like to see them.

Given that the roof is slate and in the stated condition, what approach/products should I use to condition the knee wall area and slopes? Tentative plan is to condition that area, but leave the attic unconditioned. Thus, part of the sloping rafters would be air sealed, but the top attic portion of those same rafters would be open/unsealed. Thank you.

Answered by Emerson W
Posted Dec 12, 2017 9:02 AM ET


Slate is a vapor-permeable type of roofing the dries readily to the exterior. It is usually installed over skip sheathing.

If you want to insulate the rafter bays under a slate roof, here are the rules:

1. Never install spray foam insulation to the underside of the skip sheathing. You don't want to glue the slates to the roof assembly.

2. The best job will include a ventilation space (at least one inch deep) below the skip sheathing, created with site-built ventilation baffles made of plywood or rigid foam.

3. As with any type of retrofit attempt to insulate rafter bays, the basic problem is that it's hard to get enough R-value when you are restricted to the depth of the existing rafters. For a few solutions to this issue, see How to Build an Insulated Cathedral Ceiling.

Answered by Martin Holladay
Posted Dec 12, 2017 11:32 AM ET


Unless you have an infinite budget and can do it all at once, concentrate on the weatherization issues first (beginning with air sealing), and put off what decisions you can on the HVAC. This winter's fuel use on the oil-burner will establish a firm upper-bound on the sizing requirements of the replacement system. It's highly likely that mini-splits sized properly for cooling in the "after building upgrades" scenario will likely more than cover the heat load, but we don't really know that yet. A fuel use heat load calculation wouldn't tell you the room loads, but it would establish the upper limit of the whole-house load. It's possible that the heating ducts COULD be used for cooling, using a right-sized heat pump, at a lower operating cost than heating with oil, at a much lower installed cost than a full-on ductless solution. Right now it's all an open question. If there is a regular oil fill-up service that stamps a K-factor on the billing slips the fuel use heat load is dead-easy, but if not, you'll have to track it, and download weather data for the fuel-use period to get there. See:


With ducts in the attic you may have to make that call before the weatherization issues are taken care of. Converting it to a conditioned sealed attic is quite expensive compared to R49 on the attic floor, and AC ducts buried in insulation in your climate can be a condensation and stained ceiling nightmare.

Dense packing cellulose into rafter bays without exterior insulation is risky in your climate. Slate roofs are inherently vented to the exterior, but the plank decking can't be considered self venting. In zone 4A an unvented roof needs at least 30% of the total R to be either above the roof deck, or air-impermeable insulation applied directly to the underside of the deck. With only 6" to work with (or is it really 5.5" deep milled 2x6 rafters?) you could get away with 1.5"-2" of closed cell foam with 4-4.5" of rock wool or cellulose and still have sufficient dew point control, but the center-cavity R would be well below the IRC 2015 code minimum R49. If there is headroom for some insulated edge strips (see: http://www.finehomebuilding.com/membership/pdf/9750/021250059.pdf ) you can go a bit deeper and improve overall performance with a bit of thermal break over the rafters, but it can be a real PITA to do this working in cramped kneewall attics.

Answered by Dana Dorsett
Posted Dec 12, 2017 12:11 PM ET


Would a slate roof normally have slate laying right on the split sheathing? If there is a barrier on the top of the split sheathing/roof deck, does the "no direct foam" rule still apply?

I looked more closely today, and it appears there tar paper (?) or some other black surface, but not slate, visible through the boards. Note: the 3rd photo I included above depicts the actual split sheathing/roof deck in question.

Thank you.

Answered by Emerson W
Posted Dec 12, 2017 8:28 PM ET


Awesome, thanks. I completely agree re: postponing HVAC decisions. And, your article on sizing was very helpful. I'll ask about K-factor for oil refills.

I'd read recent updates here and from Joe Lstiburek that burying insulated ductwork was OK. But, I want to avoid any "nightmare" risk and get this right from the get go. I don't think we need to condition the attic. So, I see 3 options. What do you recommend?
1. Leave the leaky/partially insulated attic ducts as is. Blow R49 on attic floor. Then, when we update the HVAC in future, bring the air handler downstairs (assuming we have an air handler) into newly conditioned kneewall space, and install new ducts under the second floor, inside the envelope.
2. Same as 1, except: try to airseal/insulate current ductwork in attic. Then, bury newly sealed/insulated ductwork under the R49.
3. Same as 1, except: when we bring the air handler downstairs, do not install new ductwork beneath floor. Instead, continue to use the then updated (i.e., current) ducts in attic.

I'm confused. Martin advises to include a 1" vent under the roof boards/deck. But, you advise close cell foam directly under slate. Can you please help me understand? Today a spray foam contractor told me that vents on the roof boards won't withstand the pressure of spray foam installation, so he recommended against bringing the kneewall attic into the envelope and instead foaming the walls themselves. I assume this is why Martin recommended plywood or rigid foam.

Last, good news: almost all of the rafter bays will hold 7" or more of insulation. Still, getting to R49 will be tough, especially with a 1" vent. But that should be the goal, correct? (Today a second contractor told me R30 is good enough, but this seems to contradict everything I'm reading here.) I will give the FHB article on extending bays to whatever contractor I hire.

Answered by Emerson W
Posted Dec 12, 2017 8:56 PM ET
Edited Dec 12, 2017 10:31 PM ET.


It looks like your roof has solid board sheathing with asphalt felt on top. You're right: That means (in theory) that you could have closed-cell spray foam installed on the underside of the sheathing if you wanted.

In practice, however, I would want to verify that the slate is installed on skip sheathing above the asphalt felt. If that's the way the slate is installed, it's an excellent installation.

However, if the slate is nailed directly to the solid board sheathing (through the asphalt felt), there isn't any air space under the slate, and the spray foam installation would be unwise.

All of this advice assumes that you care about the slate. In my mind, you should. Slate roofs are a valuable element that add to the historic character of our neighborhoods. Every time a slate roof is lost, the entire neighborhood loses something valuable.

Answered by Martin Holladay
Posted Dec 13, 2017 5:24 AM ET
Edited Dec 13, 2017 5:25 AM ET.


It sounds as if you are uncertain about when a roof assembly needs a ventilation channel between the top of the insulation and the underside of the roof sheathing. To achieve more clarity on this issue, I advise you to read these two articles:

How to Build an Insulated Cathedral Ceiling

Site-Built Ventilation Baffles for Roofs

Answered by Martin Holladay
Posted Dec 13, 2017 5:28 AM ET


Hard-piped tight & insulated new ducts in the attic will usually be safe to bury in the attic insulation, leaky ducts, not so much. If you're going to use the attic ducts even for ONE season it's worth making the effort to seal them. Whether the money for new-improved ducts would be better applied to a ductless solution requires more analysis. To be sure, air sealing the attic floor is going to be a lot easier and more reliable over the long term if you can get rid of the dozen or more duct penetrations.

R30 is way better than what you have, but below even IRC 2009 levels. Sounds like you have milled 2x8 rafters (7.25" nominal depth)?? 2" of HFO blown closed cell foam is R14, and you'd have 5.25" for fiber insulation, which would be R19 if cellulose, R22 if rock wool, for a total of R33-R36. If you made that 3" of HFO blown foam (R21), and 4.25" of rock wool (R18) you'd be at R38, which met code under IRC 2009 (and isn't terrible.)

Mind you, 7" of HFO blown foam would give you the full R49, but at north of $5 per square foot (could be north of $7/ft^2) it's probably not financially rational.

Answered by Dana Dorsett
Posted Dec 13, 2017 8:52 AM ET
Edited Dec 13, 2017 8:57 AM ET.


@Dana, thanks. I had four contractors out this morning, so will have follow on questions soon re sealing/insulation. Do you agree with Martin regarding venting under slate, within the rafter bays (assuming there is no skip sheathing above the asphalt felt)? I just want to be sure I understand that you are both telling me the same thing. (I have read Martin's helpful articles.)

RE attic ducts. Im happy to replace/update them if needed, but A) dont know if it is needed (how should I decide??), and B) dont know what I need, materials or sizing. I suppose I can always seal them myself with the tape Martin recommends, to buy some time. It is also possible that the attic ducts are AC only (basement furnance heats everything, but an AC unit was dropped in to attic later), so I could insulate the attic now and then deal with the ducts in the Spring. Either way, given their importance and our expected longevity in the house, Id rather replace the ducts than spend a similar amount on aeroseal. Regardless, I don’t want to was risk condensation in a newly sealed/insulated attic.

Given my situation - not being sure if ducted or ductless will be best long term - what do you recommend?

Answered by Emerson W
Posted Dec 13, 2017 4:12 PM ET
Edited Dec 13, 2017 7:45 PM ET.


If I were you, this would be the order of my decisions:

1. Gather more information on the condition of the slate roof, in order to determine whether the slate really needs to be removed.

2. Look at my bank balance to determine whether I could afford to convert the vented unconditioned attic to an unvented conditioned attic. From a building science standpoint, creating an unvented conditioned attic (in order to bring the ductwork indoors) is always the best approach, unless you decide to abandon the ductwork and switch to ductless minisplits. That said, creating an unvented conditioned attic is expensive.

3. Talk with everybody in my family to determine whether the ductless minisplit option is worth pursuing.

4. Come up with a plan to create an unvented conditioned attic so that a rough budget for the work can be determined. Then, compare the cost of the conditioned attic option with the ductless minisplit option.

Answered by Martin Holladay
Posted Dec 14, 2017 7:02 AM ET
Edited Dec 14, 2017 7:28 AM ET.


@Martin, Thank you very much. These are excellent suggestions, and I will do as you suggest w/roofers. EDIT UPDATE: We have a Bangor slate roof, which has expected life of 80 years, and we are at year 65. The front slates face north and could last that long (this is where the kneewalls are), but the back slates are in rougher shape and might need an overhaul in 5 years. FWIW, at least one slater thought you coukd foam beneath the asphalt felt and not ruin the slates. There is no skip sheathing outside the 2x8 roof boards.

Both my wife and I are OK with ductless mini splits, aesthetically speaking. (Our 4yo son has no opinion.) Thus:

1. Why do you recommend conditioning the attic? Our plan has been to condition the kneewall space and moving any future HVAC equipment there (w/updating and burying the ducts in the attic) - wouldn't this alleviate the need for conditioning the attic? Also, I suspect we can get a higher R value with a blown in attic than in the rafter bays.

2. With the 30% tax credit slated to be reinstated, why would you recommend ductless vs. geo? I've read your articles on geothermal and understand that the installation is the key. But, both have high startup costs, ad it seems to us that geo is more efficient (ie, lower operating costs).

Answered by Emerson W
Posted Dec 14, 2017 7:22 AM ET
Edited Dec 14, 2017 3:12 PM ET.


Don't count the geo-tax-break chickens before they hatch, but also vet the geo contractors extremely thoroughly. The stated COP of the heat pump itself doesn't include the pumping power for the earth loop or the radiation. That additional pumping power adds up, and is in the control of the system designer, some of whom are competent, others, well... Typical solutions will deliver a all-in COP of about 3.5 (=HSPF 12-ish) , even though be$t-ca$e geo solutions can be in the ~4 (HSPF 13.5) range, but hitting the 5s requires hiring the Tooth Fairy and his trusty unicorn for the job (though it can be done with very carefully designed ultra-low temperature radiant floors, etc.).

In your climate it's actually pretty tough for geothermal to beat right-sized better class ductless solutions on annual power use, but it'll beat most ducted air source heat pumps.

An HSPF 13-14 mini-split can be installed for about $4K/ton or less in most areas. Most of MD is in the fewer heating hours edge of the heating zone IV map, and properly sized, ductless solutions can meet or even beat their nameplate HSPF numbers, which are tested & rated at zone IV heating hours assumptions:


In my area (US climate zone 5A, southern New England, on the edge of the zones IV & V for heating hours) the substantial cost difference is often large enough to pay for enough rooftop PV to cover the full heating energy use of heating with ductless (not just the energy use difference!) But geo is also more expensive here than in many areas, thus YMMV.

Answered by Dana Dorsett
Posted Dec 14, 2017 3:03 PM ET


@Dana, thank you. Is there an article you recommend, or why are ductless so much more efficient than ducted?

I realize this decision should be made based on loads, but: I suspect 1-ton dudtless could heat and cool our 900ft second floor? This would simplify the attic ductwork decisions (!), be much more efficient than our current oil/11 SEER AC combo for second floor, and soften the blow a bit of sticking with oil for first floor until the furnace dies. OTOH, when the oil furnace does die, where does that leave us for the first floor? Would we be boxed in?

Answered by Emerson W
Posted Dec 14, 2017 3:27 PM ET


It doesn't take an article to explain why ductless minisplits are more efficient than ducted minisplits. I think I can do it in two sentences.

Pushing air through ducts requires a lot of electricity, because every inch of ductwork, and every duct fitting, is responsible for static pressure that must be overcome by an energy-using fan. That's why heating equipment without ductwork uses less electricity than heating equipment with ductwork.

Answered by Martin Holladay
Posted Dec 14, 2017 3:43 PM ET
Edited Dec 14, 2017 3:57 PM ET.


lol, thanks @Martin. I intended to ask if there is more to it, or simply the ducts. Hard to edit questions on my phone! Do you, Martin, have any thoughts on the question I posed to Dana?

Answered by Emerson W
Posted Dec 14, 2017 3:55 PM ET


Ducted systems have to use a lot more air handler power to overcome duct impedances, and the duct design can bring the efficiency to it's knees with insufficient air flow, and increase the actual loads with air-handler driven air infiltration due to the unavoidable room-to-room pressure differences created.

With ductless systems the air paths are wide open, with essentially zero back pressure (other than the heat exchanger and filter in the head), and the paths are fully within a single room, no room-to-room pressure differences.

This document is feeling a bit dated (but still less than a decade old), but it shows just how much the system design aspect can cripple the as-used seasonal efficeincy:


For instance, the heat pump in the system in CT had a bench-tested COP of 5.1 at low stage, 4.2 at high stage, but averaged 3.7 ( HSPF 12.6) over the third-party monitoring period of two full heating seasons (see Table 2.)

Similarly, the system in VA used a unit that bench tests at a COP of 4.6 (low) to 4.1 (high) and averaged a COP of 3.4 (HSPF 11.6) over a single monitored heating season (Table 3).

The one in WI used a unit rated 4.6(low), 4.0 (high), but averaged 3.4 (HPSF 11.6) for the season.

The description of the geothermal pumps are in Table 1, and the systems are described elsewhere. I've read some fairly blistering critiques of some of the design choices in those systems from geo designers on various web forums, but there nothing that persuades me that these systems are outliers.

There is clearly design risk- every system is a custom system, and you are at the mercy of the local geology/hydrology along with the skills and experience of the designer / contractor. With ductless air source heat pumps there are far fewer ways to screw it up (though there are always a few genius-idiots capable of undoing even the most idiot-proof systems :-) .) Air is air, local weather norms are well documented, so it's primarily a matter of sizing it correctly for the peak and average loads so that it spends most of it's time operating in it's more-efficient zones.

A 1-ton ductless may or may not cover the loads of a 900' second floor. It could take a 3/4 tonner, or even a 1.5 tonner. Just taking a WAG on it using dumb rules of thumb, air conditioning load of a ton per 1000' sorta indicated a 1-ton, and most 1-ton cold climate heat pumps can deliver 16,000 BTU/hr or so at MD style 99% outside design temps, which would be 18 BTU/hr per square foot of space, which would be on the high side for a TIGHT and INSULATED 2x4 framed house at 10-15F outside, 70F inside. A cold climate 3/4 tonner is good for ~12,000 BTU/hr at your design temps, which would be 13 BTU/hr per square foot, which is a realistic ratio for a tight 2x4 framed house, and can also deliver 11-12K of cooling (at max speed, not it's "rated" tested speed), so it's not at all out of the question. But you really have to do the math to know.

Answered by Dana Dorsett
Posted Dec 14, 2017 4:11 PM ET
Edited Dec 14, 2017 4:12 PM ET.


Some more bed time reading about in-situ measurements of ductless system performance:


Fast forward to Table 24 for the fleet average performances, by region. Your winter climate is really somewhere between the Willamette region and Inland Empire for design temperatures and seasonal average temperatures, so figure they would run about a COP 3 (HSPF 10.2) When this testing was done the typical Mitsubishi bench tested at about HSPF 10.5, the typical Fujitsu about 11.5 or 12. It's unknown how many of those systems were under or oversized for their loads, but the overall average isn't very far removed from the test numbers.

Since that time (pre-2012) incremental design improvements have brought the bench tested performance up by 15-25%. While it's true that geothermal heat pumps have also seen incremental improvements since those pre-2010 systems linked to in my prior post, have the geothermal SYSTEM DESIGNERS also improved?

I'm sure some have, and I'm just as sure that some have not, which is why it's important to be skeptical and ask lots of questions when talking to geothermal contractors.

Answered by Dana Dorsett
Posted Dec 14, 2017 4:32 PM ET


@Dana, We are air sealing and insulating, but likely can never get to “TIGHT and INSULATED.”
1. Can a properly-sized ductless unit in the 2nd floor hall heat/cool the four rooms on the second floor?
2. By doing the math, do you mean Manual J calculations?
3. Who should I ask, locally, to help me?
Clearly, there is less design risk with ductless as opposed to geothermal. The best geo firm Ive spoken with has been doing geo since 1981, but the risk of a $30,000 headache is highly aversive.
Thanks again for your generous helpful and expert replies.

EDIT: Our posts crossed in cyberspace. Are you suggesting I should have expect a current COP around 10 near Baltimore, or a current COP around (10 + 10-25%)?

Answered by Emerson W
Posted Dec 14, 2017 4:35 PM ET
Edited Dec 14, 2017 4:44 PM ET.


In order:

1: Don't know need to do the math, but probably not at design temp unless the doors are open and the setpoint cranked way up. The achilles heel will likely be window losses.

2: yes

3: A registered professional engineer or RESNET rater competent with aggressive by-the-book Manual-J could get you there on the load numbers for both the current and "after building upgrades" version of the house. With the room load numbers the path to heating & coolling solutions may become clearer.

Answered by Dana Dorsett
Posted Dec 14, 2017 4:48 PM ET


@Dana, Sizing issues aside, are you suggesting that I should expect COP around 10, or around (10 + 10-25%)? If a ductless solution is my best option (ie, if it will perform as well as GSHP), at least for the 2nd floor, this would be great to know. Other alternatives will be GSHP or ducted ASHP.

Answered by Emerson W
Posted Dec 14, 2017 5:59 PM ET


Don't confuse COP (coefficient of performance) with HSPF (heating season performance factor). They're different. Dana wrote, "COP 3 (HSPF 10.2)."

Answered by Martin Holladay
Posted Dec 15, 2017 6:47 AM ET


@Martin, thanks for correcting me. If the ductless performed @ COP of 10, this would be a no brainer. Martin, what is your take on ductless in my situation?

Answered by Emerson W
Posted Dec 15, 2017 7:26 AM ET


In general, I'm a fan of ductless minisplits. Abandoning your attic ductwork and installing minisplits may be less expensive that creating a conditioned attic or moving your ductwork to a new location inside your home's thermal envelope.

That said, some people don't like the look of minisplits, and some minisplit systems introduce the problem of room-to-room temperature variations. (Both of these problems can be addressed by using ducted minisplits instead of ductless minisplits.)

For more information on this question, see Rules of Thumb for Ductless Minisplits.

Answered by Martin Holladay
Posted Dec 15, 2017 8:21 AM ET


Ductless is definitely better than ducted heat pumps or AC with ducts in the attic. It's not always going to come in at less power use than a well designed geo system in your climate- it'll be comparable or even a little bit higher, but the upfront cost will be half (or less.)

eg: Couple of years ago I was advising on a ~3200' house at the warm edge of zone 5A in MA in a location with a 99% outside design temperature of +13F, which is comparable to MD type outside design temps. Even though it was sub-optimally oversized for the actual loads, four tons of ductless multi-split (six zones) came in at an installed prices of about USD$15K, for units rated ~HSPF 9.5 BTU/watt-hour. (9500 BTU Four tons of geo would have been in the neighborhood of $35-40K. With MA and utility efficiency subsidies they ended up paying around $11K (? hard to remember the details with precision), but that was less than half the quoted cost of a condensing propane + 3 ton split-AC proposal for the same house (that would have had no subsidy.)

With an up-front difference of $20K+ between ductless & geothermal they could install install 5500-6000 watts (DC) of rooftop PV, which pretty much covers their heating & cooling power use (all of it!), but not 100% of their power use. (I recently heard they intend to install PV next year now that the abutting neighbor removed some trees.) They'll still have a power bill, just not a heating & cooling bill. YMMV.

Geothermal in my area is more expensive than some other locations. I recently corresponded with a guy in a more remote part of upstate NY who was quoted ~$25K for 4 tons of geothermal (before any subsidies or tax rebates), which seemed suspiciously inexpensive, but the local reputation of the installer/designer was quite good, and there weren't many ductless installers within an hour's drive of his place, so he's going for it. If the 30% federal tax credit gets reinstated (a big "if"), after NY state rebates he'll be into it for less than $12K, but with just the state incentives and no tax credit it'll be on the order of $18-19K out of pocket. Subsidies can skew the markets and affect these decisions by quite a bit.

Answered by Dana Dorsett
Posted Dec 15, 2017 3:14 PM ET


Dana: Dandelion, a Google offshoot, has been pushing ground source heat pumps in upstate New York, for $20,000 net after state incentives. As far as I can tell, they offer a one size fits all retrofit for oil or propane fired hot air systems. The cost includes domestic hot water. They claim to have a more efficient drilling system that helps keep cost down.

Answered by stephen sheehy
Posted Dec 15, 2017 4:01 PM ET


Stephen: I've yet to find a single person who has made contact with Dandelion. Do YOU know of any?

Have they even installed more than ONE system in NY since their launch last summer? They commissioned their very first system on 26 September, (nearly a full quarter after launch, to the founders of an anti-fracking anti gas pipeline organization), which isn't exactly a rapid roll out to the masses:


They are primarily a financial & marketing company, and sub out all of the geo work to local contractors. Their "special sauce" is their use of big data to be able to figure out which houses are most likely to be energy pigs heated by expensive oil or propane, targeting their marketing & sales extremely narrowly. They have access to cheaper capital than the average geo installer, and are able to offer better financing to their clients by not having to get third party financing. I'm not sure how much that affects the total cost to the customer or the project compared to other financing paths, but it probably streamlines it a bit.

It looks like they should be able to make some headway in Rhinebeck, after convincing the town council to allow installation of the ground loops on public land in front of homes before contracting with the homeowners for a system:


The Rhinebeck case brings a certain economy of scale to the process to bring the installation costs, since it's not a separate custom bidding process, and it keeps the equipment and crews busy without having to move it around, reducing the per-system cost even if only a large fraction of them buy-in within a few years. It's sort of an "If you drill it, they will come." kind of marketing hope, sort of like stringing the cable on the street before any customers have signed up for cable services. Eventually most of the abutting houses will probably bite- it's just one oil price spike away to make it a complete no brainer even for short-termers.

They are currently targeting only Albany, Columbia, Dutchess, Fulton, Greene, Montgomery, Rensselaer, Saratoga, Schenectady, Schoharie, and Ulster counties. The rest of the state (and country) takes a much bigger data crunch, I s'pose, but targeting counties with less developed gas grids in a cool climate with a lot of older-stock house is certainly going to have a higher density of high heating cost homes than the US average.

Their "more efficient" sonic drilling system isn't new tech- it's been around for decades in numerous applications (including utility-scale PV array foundations) not just geothermal, and not always an optimal or cheaper way to install geothermal in NY, according to various contractors who actually DO install geothermal in NY.

Answered by Dana Dorsett
Posted Dec 15, 2017 4:49 PM ET


Thanks all. My takeaway re GSHP vs mini splits is: both could be good systems for my situation (and a huge improvement over oil/SEER 11 AC). The best option depends on the up front costs, which are driven by incentives. So, I will watch and wait and learn. I have two HVAC guys coming by next week to discuss - both from good firms - and I’ve already spoken at length with one very well-regarded geothermal installer rep. Of course, we’ll tighten and insulate first.

Relatedy I now have air sealing/insulating estimates in hand and would like to disuss these here. I will post questions shortly.

Answered by Emerson W
Posted Dec 15, 2017 9:51 PM ET


Thank you for all of your expert responses. I had five for air sealing/insulating contractors visit and was able to ask intelligent questions and identify several red flags (including the R value doesn;t matter w/foam sales pitch). I am now reviewing quotes. Questions:

1. Knee wall space. Under what circumstances, if any, is it recommended to NOT move the thermal boundary to the roofline? The two most knowledgeable contractors - very experienced, BPI, RESNET, energy auditors, etc - BOTH recommended leaving the thermal boundary at the kneewall in this instance. Rationale offered by both was the heating/cooling costs of additional 400ft2 of conditioned space - there is a large, long kneewall attic, accessed by a full-sized door and which is manageable to work in.

The advantages to not moving the boundary seem to me: 1) much lower up front cost and 2) higher R value - because we could insulate the kneewall floor area to R-49 or R-60, whereas the rafter bays are only 7.25" deep. I'm not sure of the disadvantage, but everything I've read says, "Move the thermal boundary to the roofline."

2. Slopes. If we do not move the thermal boundary to the roofline, how should I insulate the slopes (7.25" rafter bays)? This was discussed above (w/somewhat conflicting guidance - or maybe I'm just dense!), but we were assuming that we would be moving thermal boundary to roof line. We will block/seal above the kneewall and also from the top unconditioned attic.

3. Basement crawl spaces (currently unfinished). If walls are air sealed first, are fiberglass bats OK? One of the two best contractors prefers fiberglass bats. He has been doing insulation since 1986 and has a great reputation, so I assume they know how to install batts properly. But every other quote was for spray foam or rigid foam board.

Thank you.

Answered by Emerson W
Posted Dec 17, 2017 11:37 AM ET
Edited Dec 17, 2017 3:11 PM ET.


Insulating at the roof line results in more conditioned floor area, but less insulated surface area (= less heat loss, all else being equal.) Did the contractors do ANY math on that?

Fiberglass batts alone on basement/crawlspace walls is risky, even if it's air tight. An inch of foam-board or closed cell spray foam between the foundation and any fiber insulation is prudent. At that point it's generally cheaper to go with an all-foam solution.

To hit R49 in the 2x8 rafter bays with a 1" vent gap simple isn't possible. But 7" of HFO blown foam directly on the underside of the roof deck would get you there. I defer to the roofing as to whether the slate roof details make an unvented solution problematic.

Answered by Dana Dorsett
Posted Dec 17, 2017 3:42 PM ET


@Dana, thanks. No, they did not do any math on it. Hmmph. To be clear, move the knee wall boundary to the roof line, 100%, even if we can't get the R value as high?

Answered by Emerson W
Posted Dec 17, 2017 4:00 PM ET


It's easier to air seal if the insulation is at the roof deck. With ~20-40% less total area the R-value can be lower and still hit the same or better performance marks, but do the math!

Answered by Dana Dorsett
Posted Dec 17, 2017 6:15 PM ET


@Dana, how "do I do the math" to make an evidence-based decision? Numbers I know are surface areas of roofline, floor, and kneewall, and likely R-values post sealing/insulation. I also know baseline leakage, but that will change. Thank you.

Answered by Emerson W
Posted Dec 17, 2017 10:28 PM ET


You need to do a heat loss calculation, using either paper-and-pencil methods or software. Here are some links to help you:

How to Perform a Heat-Loss Calculation — Part 1

How to Perform a Heat-Loss Calculation — Part 2

Calculating Cooling Loads

Who Can Perform My Load Calculations?

Answered by Martin Holladay
Posted Dec 18, 2017 6:52 AM ET


Calculating the U-factors of the roof vs. joists + kneewalls on the kneewall -attic requires doing careful estimates of the actual framing factors, which will be higher than those in generic tables due to the shortness of the spans. But a parallel path calc on the kneewall, attic floor, and roof using the correct framing fractions and cavity fill Rs will get you there.


Answered by Dana Dorsett
Posted Dec 18, 2017 8:43 AM ET


@Dana, I've been playing around with parallel paths but frankly am not sure how to apply the formulas. The rafters would are in all areas in the same proportion, which I suspect roughly makes them a wash.

1. Floor is 500sqft (50x10), would be up to R60 (blown cellulose)
2. Kneewall is 350sqft (50x7), would be R13 to R19 (current rockwool - then cover w additional 1-2" of rigid foam)
3. Roof is 600sqft (50X12), currently fully vented - would be max R32 or so (w a vent and drywall or foam board cover). I believe this would require closed cell foam (over a baffle or if asphalt felt would be OK) - but all estimates listed open cell foam, w/rationale to identify leaks in roof. Thus, either the contractors are wrong, and I should be using closed cell on roofline, or the R value would be lower than R32.
4. Spray foaming ceiling is considerably most expensive option - 3-5k additional costs. This $ could be used to enhance other areas of the envelope, eg drill and fill other external walls on this cape second floor.

Answered by Emerson W
Posted Dec 18, 2017 11:07 PM ET


To see a step-by-step demonstration of the parallel-path method of U-factor calculation, see this article: Three Code-Approved Tricks for Reducing Insulation Thickness.

Answered by Martin Holladay
Posted Dec 19, 2017 6:22 AM ET


@Martin, @Dana, et al.

I've met with 6 (!!) insulation contractors re: this project. I've read GBA (simply an endless library - it's all here - amazing) and think I have a winner.

What are your thoughts on this approach?

BASEMENT - currently vented, unconditioned crawlspaces
1. Spray foam all rim joists
2. Cover grates and seal vents
3. 2" rigid foam board on crawlspace concrete walls
4. Cover crawlspace concrete floor w/drain mat and vapor barrier liner.
Q: Is the floor cover necessary?

KNEE WALL ATTICS - currently unconditioned, poorly insulated
1. Move thermal boundary to roofline
2. Cut and cobble 7.25" rafter bays w/6" rigid foam board (>R24) w/reflective barrier, leaving 1.25" open beneath board sheathing and slate roof. Seal w/closed cell foam.
3. Cover cut/cobbled rafter bays w/ 2" of thermax, to increase R value to >R36 and address thermal bridging. This will also add ignition barrier, semi-finish space, and protect my 4yo son from spray foam.
4. Continue this approach through slopes to attic.

TOP ATTIC - currently vented, poorly insulated
1. Move aside or vacuum remove 3" old cellulose insulation
2. Remove whole house fan and air seal all penetrations
3. Seal or replace ducts, consider burying
4. Blow R60+ loose cellulose
Q: Is there any reason I shouldn't blow the cellulose myself, eg, purchasing cellulose from Home Depot?

Thank you.

Once we finish this work, future updates will include 1) add efficient HVAC equipment, and 2) consider add wall insulation (currently R3 on 1st floor stone walls and R11 on second floor aluminum siding).

Answered by Emerson W
Posted Dec 20, 2017 9:36 PM ET


Q. "Is the floor cover [over the concrete basement floor] necessary?"

A. A layer of polyethylene over or under the basement floor is necessary. If you have a dirt floor, you know you need the polyethylene. If you have a basement slab, there may be uncertainty about whether the contractors remembered to install polyethylene under the slab. You can drill a hole through the concrete to check. If there is no poly under the slab, it's a good idea to install poly over the slab.

Q. "Is there any reason I shouldn't blow the cellulose myself, eg, purchasing cellulose from Home Depot?"

A. No reason. Here is a link to a relevant article: Borrowing a Cellulose Blower From a Big Box Store.

Answered by Martin Holladay
Posted Dec 21, 2017 4:58 AM ET


Having watched the evolution of my thought, what is your assessment of my planned approach? If there are no red flags/obvious deficiencies, I am excited to move ahead.
Thank you

Answered by Emerson W
Posted Dec 21, 2017 6:00 AM ET


I think you can move ahead. Remember, air sealing work is at least as important as adding R-value -- perhaps more important. Remember, too, that air sealing work usually comes first, before adding insulation.

Answered by Martin Holladay
Posted Dec 21, 2017 6:38 AM ET


While open blown cellulose is cheap and easy with a rental blower, it isn't always cheaper than hiring a contractor to do it.

If you are going to be obsessive about the air sealing details, you might do that part DIY rather than leaving it up to the contractor, unless the contractor is also obsessive, and has both a blower door and infra-red camera for chasing down the leaks. (Some do, most don't.)

Answered by Dana Dorsett
Posted Dec 21, 2017 5:36 PM ET


Thanks @Dana. I believe that the contractor wants to do a good job - he "gets it," although I expect I will have to stay on him about blower door and infrared throughout the project. (He agreed to do pre-post blower door. My compromise is to work from home during the 3-day job, and oversee.)

My joining GBA and your and Martin's helpful replies, in particular, have really boosted my confidence that we have a great plan. And, the contractor is clearly motivated to do a good job. So, we will see.

I'll post updates re: project in this thread in January. Many here on GBA might be surprised by the brand/franchise whom I've selected (and in whom I had the most confidence).

Answered by Emerson W
Posted Dec 21, 2017 6:09 PM ET


@Dana, quick follow-up question, re: obsessive about air sealing. I don't have a blower door or infrared camera, but I do want this to be done right. It's an expensive project (>$12k) that will hopefully make a dramatic and permanent improvement in the comfort and our house. What else can I do, other than stay on the contractor, observe closely, and request blower door and infrared review for OA, as often as I can?

Answered by Emerson W
Posted Dec 22, 2017 9:54 PM ET


Your list of things to do is good. I suppose if you have doubts about your insulation contractor, another approach would be to hire a different contractor -- one who advertised themselves as a "home performance contractor" or "weatherization specialist" and who specifically markets their building science expertise and approach.

Answered by Martin Holladay
Posted Dec 23, 2017 5:39 AM ET

Other Questions in Plans Review

Radon is 9.3 pCi/L even after radon mitigation — Is this OK?

In GBA Pro help | Asked by Hotandhumid | Jan 22, 18

Kitchen Hood

In Building Code Questions | Asked by TracyPTM | Jan 23, 18

Half inch plywood battens for rain screen?

In Green building techniques | Asked by Cragged | Jan 22, 18

Composite Foundation Walls?

In General questions | Asked by Michael | Jan 23, 18

Where can I get a forced air heating wall vents that can be 100% sealed shut?

In Green products and materials | Asked by Bryce Nesbitt | Jan 22, 18
Register for a free account and join the conversation

Get a free account and join the conversation!
Become a GBA PRO!