Using Ductwork with a Ductless Minisplit
We have really enjoyed our first winter with our MRCOOL DIY Ductless Minisplit, a replacement for our oil-burning furnace (in Zone 5a, CT). When our furnace was removed, so were the ducts, but the duct boots are still in place. We have two minsplit heads, one in a bedroom at the end of the bedroom and bathroom “wing” (about 600 sq. ft.) that blows directly through the open bedroom door down the hallway. That entire wing is perfectly cozy, even during the most extreme days, and is rarely set above 65 degrees. However, at night I would like to enable the resident of that bedroom to close the door, if desired. There is a duct boot directly to the right of the doorway in the path of the minisplit. May I add ductwork in a “T” with an inline fan directing the air through this duct boot and into the two other bedrooms? One run would be a total of 5-6 feet, the other about 8-10 feet.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Following b/c I've been thinking of doing something similar and wondering if there are good recommendations for quiet in-line fans or how to add silencers to the ductwork to reduce room-to-room noise transmission.
To move heat efficiently with a fan requires a decently high temperature difference. Heat a 65F room with any heat load to speak of with 75F air requires moving a metric shit-ton of air (is that the correct units? :-) ) and gia-normous ducts to avoid generating a significant wind-chill effect. An open door & hallway is a big enough "duct" that a ~300 cfm wall-coil head blowing directly at the open doorway can move enough air to sort-of work, but with a normal sized duct the duct velocity & wind chill potential (not to mention the noise) would need to be quite high.
A wall-coil type mini-split's ability to deliver decent amounts of heat at 300cfm depends on an output temp on the order of 110F-120F at the head, more than 40-50F above the average room temp. To move the same amount of heat moving air only 10F above the targeted room's temp takes 4-5x the air volume.
So, with a 1200-1500 cfm air handler (like those used for a 3.5-4-ton central air system) you can do it, but you won't like the result.
What if the jumper vent is within inches of the mini-split head? Then the air should be a lot closer to the 110F-120F coming out of the head.
Something like this:
https://www.hvacquick.com/products/residential/Bathroom-Ventilation/Pressure-Balancing/Tamarack-Return-Air-Pathway-RAP-Pressure-Relief
Located right above or below the mini split head. Combined with a several hundred cfm inline blower pulling air out of the attached room (and through the jumper duct).
Keep dreaming, but don't spend any money on that approach...
Even at a few inches away the dilution with room air is dramatic with a return grille vacuuming up the warm-ish air near the mini-split head. Go ahead and measure the air temp 6" in front of the mini-split rather than directly at the output. It doesn't take much dilution to drop that temp dramatically, and with the nearby return grille drawing in air the dilution factor with other room air will be large.
If one keeps going with improving retrofit kludges it doesn't take long before it's at the same cost as adding another DIY mini-split (not that adding another oversized mini-split is a great solution.)
Ducted mini-splits are really a better (if more expensive) solution to the heat/coolth distribution problem, and can be effective with more reasonably sized ducts. With properly sized (and shorter) ducts the efficiency hit isn't bad either.
With Jennifer_M's duct boots & registers being still in place it's possible (even likely) that a ducted mini-split solution could be effective. The prior oil-burner was likely 3-5x oversized for the design load, so the issue would be whether there is sufficient velocity or throw to the existing registers, but that can be tweaked if needed. With low exit velocity and low throw the location of the return registers begins to matter more, but a decent duct designer could probably make it work without completely ripping them out & starting over.
Heard and curious with a follow-up...
I'm facing a situation where I really wish we had put in a ducted minisplit, but we didn't; and bedrooms get uncomfortably hot in the summer.
Seems like our choice is probably between spending $5-6k+ to tear out the ceiling cassette and put in a ducted unit and trying to kludge some ductwork together which I can do DIY just for the cost of parts. I'm figuring a duct with an inline fan would definitely be quieter, more effective and less disruptive than our current strategy of putting box fans in the room doors? But is there a third option? Is there any market for used minisplit units or is that cost really down the drain if we replace it?
Which ceiling unit do you have?
Some have knockouts on the side for transfer ducts. These are really meant for distributing air though a larger space but can also be used to feed a room.
There isn't much pressure there so you have to take care with the ducting and need a largish door undercut as a return.
@Akos -- I'm pretty sure it's an Mitsubishi SLZ-series 4-way ceiling cassette.. do those have the transfer ducts?
You'll have to check your unit. Some units do have knockouts for ducts, some only for fresh air intake.
Points regarding jumper duct noted.
Regarding heads vs air handlers:
For Mitsubishi, my understanding is that the outdoor units provided with the air handler kits are an older design and perhaps contribute to the efficiency hit:
https://www.greenbuildingadvisor.com/question/mini-split-heating-efficiency-floor-ceiling-mounted-vs-concealed-duct
For Fujitsu, my understanding is that the outdoor units provided with the air handlers do not have pan heaters. Per your comments, the Fujitsu units are best in cold climates due to better thermistor operation at super cold temps. However, only the head-type units come with pan heaters.
For LG Red, air handler type units do have pan heaters. However, these are newer and am unsure on their track record and performance at cold temps.
How is the MR Cool unit performing? I am thinking of going to a Perfect Aire 18k DIY model. This unit seems to be similar to the fourth generation MR Cool that was AHRA tested but not yet on the market; it has an HSPF of 11.6 according to NEEP. HVAC installed units (e.g., Fujitsu and Mitsubishi) have been priced well above reason around me...assuming you can even get an HVAC guy to respond to an estimate
Hi Nick, my MRCOOL DIY was $27k cheaper than the quote I received for a Mitsu hyper-heat. I could love it for that alone, but it has also performed beautifully all winter. I set it and let it run 24/7, dialing it lower as it gets warmer. During the coldest days we needed our fireplace (a Hearthstone Clydesdale insert), but it was just a few awful days in February.
Dana, even when the door is open we need a little fan at the top of the door to move the heat down the hall. Wouldn't the fan in the duct boot do the same thing? Forgive my ignorance!
>...even when the door is open we need a little fan at the top of the door to move the heat down the hall. Wouldn't the fan in the duct boot do the same thing? <
A fan in an open doorway has effectively zero static pressure working against it to limit flow. If you narrow the outgoing pathway to say, a 10" x 10" duct boot or grille (100 square inches) down from a 32" x 78" door (~2500 sq. in.) with a 1.25" x 32" door cut (40 square inches) you have reduced the free path by more than an order of magnitude, which will cause a significant pressure difference across the fan reducing flow, even if the duct runs are blissfully short.
At a 5-10F temperature difference just the open doorway without the fan would be good for 100-200cfm (a WAG made up on the spot, not something I've measured :-) ) of convection flow through the doorway. A 100cfm (free air rating) fan would create higher velocity/throw the outgoing stream, creating more mixing with the air further down the hall, but wouldn't necessarily be the primary driver of total air volumes flowing through the open doorway. It's likely functioning much as as an air mixer, but it's still moving real heat further.
Let's do some napkin math on a "maybe" solution. Assume you get a quiet, efficient 150cfm bath fan to drive your ducts, and assume you have an adequate return path so that it really drives 150cfm. The approximate thermal mass of air by volume is ~0.018 BTU/degree-F per cubic foot, so at a 10F difference (75F air into a 65F room) that's delivering 0.18 BTU per cubic foot. At a rate of 150cubic feet per minute is (x 60 minutes/hr=) 9000 cubic feet per hour, x 0.18 BTU/cubic foot becomes rate of 1620 BTU/hr. With duct losses & leakage or ANY significant backpressure you're probably looking at 1400-1500 BTU/hr, best case. That's enough to heat a small bedroom at typical CT 99% outside design temps, so it's at least theoretically possible, if the delivery air can actually reach as high as 75F, the ducts are short, straight, and air tight, with adequate returns paths. And that's blowing tepid air into a cool room. If it's only 72F air or only achieves 100cfm it fails. But even if it works, an efficient 150cm bath fan runs north of $200.
For an intermittently used guest bedroom there is a superior, cheaper, easier to install solution:
A 400W (~1350 BTU/hr) electric panel radiator can be had WITH a wall-plug line voltage thermostat for under $150 through the big orange box store, or less than $125 for a wi-fi (not manual) wall-plug line voltage thermostat. This allows the occupant to set whatever temp they like for the hours the room is in closed-door use, and it's comfortable. Unless the closed-door occupancy rate is high it won't use as much electricity as a fan blowing a draft of tepid air into the room 24/7 just to keep the room temperature from crashing. Both the big blue & orange stores sell similar Amaze branded 400W panel rads, as well as 600W or larger versions. (Don't go bigger unless you've determined it really needs it), but you'll have to add the cost of a line voltage thermostat (hard wired, or plug-in) to the panel cost. (Amaze also offers a 250W version, but that's only ~850 BTU/hr, which may not be enough to cut it in the cooler winter days depending on the actual heat-loss numbers of the room you're trying to heat.) When the room isn't occupied, turn the panel rad off and leave the door open to keep the room reasonably warm, just as before. If the room is starting off at 63-65F the heat up time to 68F-70F isn't going to be very long with a panel radiator after the door is closed unless it's <15F (a typical 99% condition for coastal CT) outside.
Unless that room is going to be occupied with the door shut a LOT in the dead of winter, that's what I'd do. YMMV.
Thank you, Dana! I might try it, since I'll be doing it myself. The two runs will be short and straight, with solid ductwork, taped and insulated. I'll report back.
"For an intermittently used guest bedroom there is a superior, cheaper, easier to install solution:
A 400W (~1350 BTU/hr) electric panel radiator..."
We have these and they are great, but the minisplit head in question is located in that bedroom. Thus, the need to move the air OUT on the rare times the door is closed.
Seems to me your efforts maybe better spent finding and fixing the air leaks that are pulling the cold air into the cold rooms. Hint the warm air that is leaking out of you house may well be on the other end of the house and that air is being replaced by cold air leaking into the bedroom.
Walta