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

Badly executed heat, what to do now? (To mini or not to mini)

KeithH | Posted in Mechanicals on

Hi all,

I’m performing some repairs on my house and discovered that most of the central heating duct work was poorly executed and needs rework Or does it?

I have a fundamental question to decide that I think the GBA community could help me with:
– Should I repair the central heating system
– Should I largely abandon it for mini splits?

House Basics:
– 1976 split level (~6 elevations)
– Climate 5b
– Low slope black epdm roof (relevant to cooling)
– Furnace: 95% 2 stage ECM variable. 75k/115k btu. Fan rated 1100 to 2000 cfm.
– Two older a/c mini splits are around but don’t have to be part of the plan. SEER 10 and 15k btus and need servicing. Both are located in spaces without central returns.
– 3200 sf interior sf + 400 sf unventilated indirectly conditioned crawl space
– Above average tightness: an older blower result of ~0.3 nACH.
– Wall: generally R19+5
– Roof: Mostly R-30, 25% R22, 15% R46
– Windows are mixed 70s double, 80s double, and recent low-e argon double.

– Cost of installation is a primary consideration at this point. Mini splits require paid labor, fixing ducts may not.
– House currently is not well cooled. WIth a flat black roof and a zero overhang roof line, cooling comfort is a problem. I’d love to say new windows and more insulation will address that but I don’t think it is true. We prefer lower design temps (heat and cool).
– Daylight basement is usable but too cold in the winter (no slab insulation), two upper level additions made in the late 80s are much too hot in the summer to use without cooling.

The summary of the ducting defects is none of the returns are right/work/exist and all of the ducts are unsealed, unscrewed, falling off, and a few that don’t exist/work are needed. They are 95% inside the envelope (there is one 4″ round that goes through the garage ceiling).

Fixing the supplies is doable, though it likely means some drywall removal. Fixing the returns is going to be a challenge due to furnace location and truss directions.

Manuals J and D:
– I paid to have a J and D done. The D is still being tweaked but I have the J results.
– I also spent an hour doing a more back of the napkin J using The results are similar.

Heating: both say 57k
Cooling: pro 41k , load-calc 44k

This isn’t really a question about whether my manual J is accurate (more about what to do) but if the numbers seem crazy, please speak up. I’m going to ask another question about a few specific manual J things.

Heating: I am not surprised that the J says the furnace is oversized. It brings the house to temp quickly and doesn’t struggle during design weather.

Cooling: Existing cooling is inadequate. We do know that we like minis so more/better is an option. I think 43k could be low with the black roof.

Why not repair the ducts? Valid choice, although it leaves me needing to figure out the cooling, possibly central or possible minis later. And maybe the right answer is do both ducts and separate cooling. So the question might be more subtly stated as exactly how much to fix the ducting.

Why not improve the envelope? I will likely be able to reduce both loads over time with DER-style work but some areas are tough. The daylight basement and crawl space have no slab insulation and would be very expensive to retrofit or be unacceptable height wise (slab to truss is 91″ in basement). All of the spaces added by a previous owner in the late 80s are substandard for insulation and glazing coatings but to DER these spaces might be a lot of $. Anything north of 30k and I’d rather re-roof and add PV.

Opinions? Love some opinions based in experience. While I love the knowledge fountain that GBA is, it isn’t a replacement for experience. (just as experience isn’t a replacement for knowledge)

Extra credit: Phased approach allowing a slow injection of cash rather than $20k+ right now.

Sorry this is so long.

GBA Prime

Join the leading community of building science experts

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


  1. GBA Editor
    Martin Holladay | | #1

    I think that this is the kind of situation where it's hard to give advice without a site visit. Even with a site visit, there are many variables -- so your own instincts and desires are going to play a role in your final decision.

    If I understand your current situation, you don't have whole-house cooling at the moment. So that even if you manage to repair all of the defects in your current ductwork, you won't have any cooling unless you install a split-system air conditioner.

    If I'm correct about that, then that fact may tip the decision in favor of ductless minisplits. But really, this situation can't be resolved by advisors acting at a distance.

  2. jinmtvt | | #2

    Any excuses are good to justify mini splits for me!!

    How old is the furnace? oil ? how long is your house ?
    What is your current heating budget ( that might say more than calculations )
    and your local electricity cost?

    In any way, never forget that the savings using minis start when you install them,
    so if you are going to end up installing some in the future, might be worth to start doing so now.

    There is always the question of DIY also, if you have the required skills/experience,
    that could favor the budget.

  3. KeithH | | #3


    Yes, of course. Asking about complicated situations on a forum is a lot like asking for medical advice from afar.

    No, I don't have central a/c. I used to believe I could avoid it with insulation and windows and such but am now thinking that would require all the windows to be retrofitted and to add awnings to at least a half dozen.

    Unfortunately, local HVAC resources haven't been any help. None have ever suggested doing a Manual D or J. One suggested the furnace could be oversized. Several have suggested that the current ducting and house design would not support central a/c well but haven't proposed solutions.

    I currently have a energy rater available (who did the D/J). Even he, a very energy oriented guy, reacted that I could not heat the house with mini splits: it gets too cold in the winter. So I'm not finding a local asset who has the combination of knowledge and experience to come up with a plan. I've tried at least 8 hvac guys over the 5 years we've owned the house.

    Furnace is natural gas (no oil here), ~8 years old. It has at least a decade of efficient heat production left in it.. However, I suspect the short cycling is contributing to comfort issues.

    The bills aren't too useful. They are relatively low but we ask for much less than design load heating and I believe the house has a lot of solar gain on mild days. I'd guess gas, which includes domestic water and the stove, is ~$140/month in the winter. Electricity can easily reach that using a single mini during the summer.

    I've used some online calculators regarding a/c SEER. Upgrading will never be about saving money. The ROI is negative.


  4. Expert Member
    Dana Dorsett | | #4

    It will (almost) always be far cheaper to fix the ducts than to add a mini-split. If it's all DIY labor (and you heavily discount your labor) it's a no-brainer- fix the ducts and buy some time to figure the rest out. No rocket scientists are wasting their careers applying duct mastic and re-hooking up ducts, adding insulation to the ducts that are outside the thermal & pressure envelope of the house. It's a PITA job, but it's not an endless job. (Been there, done that on ducts that were original equipment in 1923, and yes, it made a difference.)

    There is very little efficiency loss from short-cycling a hot air gas furnace, but a 2-stage with an ECM drive blower should deliver at least 5 minutes/burn anyway, unless it's RIDICULOUSLY oversized for the load. You should be able to force the thing to run exclusively in the low-stage since it's low-stage is higher than your Manual-J number anyway. You could set up the thermostat for a bit more hysteresis to lengthen the burns- there a lots of possible tweaks to inhibit short-cycling.

    Both natural gas and electricity have a delivered price. Look at a gas bill, divide the $ by the therms (or ccf, or decatherms, or whatever.) Look at an electric bill, divide the $ by the kilowatt-hours, then report back.

    In a zone 5B location a highest-efficiency cold-climate mini-split will deliver an average COP of about 3.3, which means averaged over a season you'll be getting about 11 KBTU of heat into the house (even though the HSPF number might be 12.5 or something, implying 12.5KBTU/kwh.) One MMBTU= 1,000,000 BTU, so it takes about 1,000,000/11,000= 91 kwh to deliver 1 MMBTU of heat into the house. Multiply your $/kwh number x 91 to get $/MMBTU.

    A condensing gas furnace delivers about 95,000 BTU per therm (or ccf), so it takes 1,000,000/95,000= 10.5 therms to deliver 1 MMBTU of heat to the house. Multiply your $/therm x 10.5 to get the $/MMBTU number.

    Compare those numbers- in much of the US the will be within 15% of one another, which isn't much incentive to change, and the mini-split might even be more expensive to run- it just depends.

    Heating a house with a design load greater than 50K solely with mini-splits in a zone 5B climate would take about 4-5 tons of mini-split. If you're lucky you can get it down to $3K/ton, $4K/ton is more likely, making that a fairly pricy way to go. If you replace the existing mini-splits with a cold-climate heat pump mini-split the up-charge for a heat-pump version isn't outrageous, and would give you options for heating those zones. The rest of the money might be spent reducing the loads on a least-cost basis. Since you have the load calculations, you can probably figure out where the lower hanging fruit lies.

    Is there any foundation wall insulation in the basement or crawl space?

    If the clear glass double-panes are otherwise in good shape, a tight low-E storm window will bring the net performance to about U0.3-ish. (Harvey Tru-Channel with low-E glazing would be the tightest, the Larson Silver or Gold series sold through box stores aren't bad, but pass on the low-end Bronze series.)

    Shading east and west facing windows with roll-down exterior blinds can be huge for cutting the cooling load (especially the west facing windows, followed by awning shades on the south windows. In many 5B locations it's dry enough & cool enough to do much of your sensible cooling via night-time ventilation, taking care to shade the windows and keep them closed during the day.

    A jet-black low-slope roof is a bad idea in any climate, but there are mop-on "cool roof" products that would keep the roof temps to the mere "toasty" level rather than "hell on earth". Look for CRRC rated products with an aged solar reflective index (SRI) no lower than 70 (which for mop-on products is pretty much all of them).

    Or you could shade the roof. The best shade for a flat roof is solar panels. It's a pretty expensive "shade" right now, but the cost is continuing to crash. I'ts already under $4/watt in my neighborhood (Massachusetts), under $3/watt in parts of Texas- even banging on $2 there. In less than 5 years most analysts (including Citi Group and the Sanford Bernstein investment bank, not just industry cheerleaders) believe the average cost in the US will be under $1.50/watt for grid tied solar, and it could easily be under a buck a watt. At or before it hits the buck-a-watt point it's worth adding that "roof shade" even if you have to finance it, since the net ROI of the low-cost power it produces is pretty good, with or without any tax/other subsidy.

  5. KeithH | | #5


    Thorough as always. I owe you a pitcher at your favorite pub.

    DIY: Yes, I waste my talents working on duct work but the labor market in my area is challenging. The best labor is not available for small projects but only for new home construction or 6 figure remodels. I have found some great independent guys but it's hard to cover every trade and HVAC is the toughest. I'm getting a bid Thursday so we'll see.

    Energy costs:
    Therms: ~$1.36 delivered
    Kwh: ~$0.115 delivered

    Consumption (monthly avg a few years ago)
    Kwh: ~775
    Therms: 56

    Notable power consumers: electric dryer, well pump, mini split. 90% LED lighting at this point so the low hanging fruit is gone.

    COP and other data: I'm going to need to pour over your numbers a bit to fully understand them. Thank you for the info.

    Cost comparison: At this point, I think I'm committed to sealing and repairing the supply ducts and the existing returns. So that's just a fixed cost. The question is whether to run new supplies and returns to the three areas (260 sf, 380 sf, and 120 sf) that don't have existing or working returns and supplies. These rooms are prime candidates for either two separate mini splits or a multi head (you have previous indicated your preferences there). If I went multi head, it would be very tempting to drop one large head in the common space to deliver heat pump heat. The 120 sf is probably just going to be an unconditioned heat vampire attached to the house.

    The three spaces added on in late 80s are the low hanging fruit and also the areas not conditioned property. They are each 6k to 8k loads. That's part of the quandry. If I pull these loads from the whole house to address with minis (or not condition one of them), then the furnace starts to be grossly oversized (~35k load, 77k 1st stage burner). But getting proper returns into these areas is a pretty difficult task. To improve the envelope in those areas such that they don't need separate heating is possible but those three areas have a lot of glazing and solar exposure. It is likely that I cannot improve the envelope enough to omit active cooling.

    Envelope improvements:

    Basement slab: a bridge too far. I lost the battle to bash it out and the head height situation is terrible.

    Crawl space slab: I have yet to figure out a durable interior crawl space slab option for insulation.

    Basement walls: Insulated

    Crawl space walls: Low hanging fruit. Plan to insulate with thermax poly iso. Savings: 3k btus?? Manual J modeled this as unconditioned space.

    Room over the garage added in the 80s: Low hanging fruit. R-22, perhaps 1/8 of total square footage and 1/4 of the roof. Savings: ~3k btus?

    Passive cooling: unfortunately, both of us have developed a certain amount of seasonal allergies in the decade we have been in this area. Immunotherapy is great but no substitute for closing the windows in April/May and July/August. For the bedrooms at least, we need active cooling for at least part of the year. Using a high COP mini split in April/May doesn't have a lot of power consumption related to it. August is another matter.

    Absolutely. On the 5 year plan is a new epdm roof with ~R-20 topside insulation (5" rockboard probably) with a solar array. I'm hoping I can get 6 kw or more up there. However, that's not a project for this year. Roof, insulation, and solar is probably north of 30k. This plan for solar is part of my interest in having at least one device that can produce heat using electricity.

    I'll need to do some more analysis based on the info you provided. Thanks for the thorough reply.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |