DIY Installation of a Ducted Heat-Pump System
Hello,
We have a house north of Seattle currently heated by all electric baseboard heating. 2400 sq. ft. and 80k btu’s of baseboard glory. The Manual J first floor total heat load is 18k btu and the second 12k. My calcs and a paid online service were within 1k btus on both. I got a quote for two ducted mini split systems, one for the first floor and one for the second. It came in at $14k before tax, and I wasn’t impressed by the company or their bid (i.e. I didn’t get the feeling I would get a quality install). So being a mechanical engineer I decided I would design and install the ducts for both systems myself and just have an installer install the two air handlers, fab the transition ducts from the AH’s to the duct I installed, install the outdoor unit, and charge the refrigerant. I know Fujitsu’s are the favorite on here but I can’t find anyone local and reputable that installs their equipment, but there are no shortage of Mitsubishi installers. So I spec’d a system with a Mitsubishi PEAD-A18AA7 for the first floor and a PEAD-A15AA7 for the second. I read Martin Holladay’s article on “Getting the Right Minisplit” noting that individual minisplits are more efficient than a multisplit system. This time around I got two quotes from a reputable company to install these AH’s with either a common compressor or alternatively individual compressors. The quote for the common compressor was $15k (with no ducts!) or $23k for individual compressors, both before taxes! Prices like this make me solidly consider just doing the whole thing myself. I will get no warranty if I do it myself, but I could replace the entire system a couple of times over for these prices. Am I out of whack on this?? I’m going to see about a quote from another HVAC company, but…
Anyone in the Seattle area have an installer they like?
I haven’t finished the second floor layout yet, but the first floor is attached.
Thank you
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
I’m in a similar boat with my roof/Insulation.
If you’re certain you understand the process of installing a mechanical system then
I’d say do everything up to the refrigerant piping yourself. The money you’d spend on joining tools(flaring or torches)/leak checking equipment/ a vacuum pump wouldn’t be substantially less than the 4-8hrs labor it would take a savy service company to connect your systems. And you’d get a professional startup. Also, major component failures on ductless within warranty periods are rare, but if an inverter board or a compressor did blow up on you, you’d probably still be able to warranty the part via the serial number of the Equipment. It’s be worth a shot.
I’d price it out though. It will take you a long time to pull this off and you may screw something up royally. Load calcs and good duct drawings are a good start but making sure that you get it in right is an art that many people who do it for a living struggle to master. that’s not because their stupid, but because there are a lot of variables that need to be accounted for and properly responded to. I’d even call around until you find a contractor to chat you up about the process. Sometimes that sort of interaction brings out contractor insecurity but some guys will see an opportunity because they know that when you need it serviced or screw up you’ll probably call them or refer them to others.
Take your time and plan out as much as you can. That’s my Advice. Good luck.
I was in the same boat (also a Mech E by education), and decided to almost-DIY with an HVAC tech friend of mine. It's not rocket science, and the Mitsubishi install guide is very clear on each step. You probably spent 10x more time learning about this, considering all the details, and over-thinking it than an HVAC contractor that is under the gun to get as much billable works done as quickly as possible. If it's simple, why not? If it's not, hire an HVAC engineering company to review your design for piece of mind.
Frankly, I think we did a much better job than many "professional" installers that wouldn't have paid the same attention to detail (ran the first floor lines concealed in the walls, proper crimp terminals, caulking and spray foam weather sealing, good brackets, etc). Sure, it took much longer, but I figured we'd come out well-ahead $ overall even if there was some failure that needed component replacement. You could do all the tedious work, and have a tech moonlight and do the final line connection, vac-purge-vac, and commission the system. Or work side by side so it's done to your satisfaction... The tools aren't $10k expensive either. Follow all the instructions to the letter, don't hack it with shortcuts, and it'll be fine (I'm sure there are lots of folks that might not agree with me, but I've seen very poor "professional" installs, and I was confident in my ability to follow instructions and ask questions/get help when needed).
Here's how ours finished: https://www.greenbuildingadvisor.com/question/mini-split-indoor-wall-locations-1-5-story-cape
Hi Guys,
Andris, sounds like we think alike. The more I think about it the more I'm inclined to just take the plunge. I read a few other threads on here about people DIY'ing ductless minisplit installs and I'm sure I can handle it. I priced out a system on Ecomfort and for a PEAD 12k and 18k each with individual condensers, line sets, mounting hardware, etc I'm looking at $7.2k. For a system with one common MXZ-3C30NA2 condenser I'm looking at $6.2k. An equivalent Fujitsu individual mini split system is like $5k.
The Fujitsu's are looking like an enticing option based on the specs and price, only down side to me is the controls look like they are straight out of the 1980's... Is there a combination of parts with Fujitsu that allows you to use a wired thermostat and also be able to control it from your phone? It looks like UTY-RVNUM and the wifi interface module are mutually exclusive. I would like to be able to change the set temp on the phone and it reflect on the wired thermostat. This is one area where the Mitsubishi's have a leg up as far as I can tell. The MHK1 thermostat can be controlled by a Redlink internet gateway.
Also I have one area where duct size is limited, it works out on paper with the PEAD's, will have to see if it works for the RLFCD's lower static pressure.
Thanks,
Cody
Other thread about DIY's
https://www.greenbuildingadvisor.com/question/dare-to-diy-a-mini-split-install
>"Also I have one area where duct size is limited, it works out on paper with the PEAD's, will have to see if it works for the RLFCD's lower static pressure."
You may want to compare all of the above to Fujitsu's beefier xxRLGX cassettes too:
https://portal.fujitsugeneral.com/files/catalog/files/(D&T)%20ARU12-48RGLX.pdf
Would be interesting to compare to the cost/complexity of a hydronic/fan coil based system (like Chiltrix or Arctic). Maybe it's not legitimate, but I feel more comfortable working with water than refrigerant (I've done a little with both).
In much of the US the "water" would have to include anti-freeze to do it with a Chiltrix. (And it's not as if there isn't any R410A in the Chiltrix, even though it all stays outside the house.)
Ducted delivery doesn't break down into micro-zones as easily as hydronic coils, but ducted split system heat pumps are a well understood well trodden path. A single mini-duct cassette/air-handler is almost always a less expensive solution than a half-dozen hydro-coils.
I say go for it. I literally just finished DIY installing two ducted Fuji units, you can read all about it here: https://www.greenbuildingadvisor.com/question/ducted-mini-split-duct-design-and-distribution-questions
I paid $3k for the units and maybe another $4k for ducting, insulation, line sets, miscellaneous parts and some sheet metal tools. That cost also included a helper @$20/hr for 5 days which made things go faster (he could seal/insulate while I ran more ducting) and easier (it's super annoying to fish thermostat wires by yourself) As suggested above I also paid a few hundred bucks for an HVAC friend to pressure test, pull the vacuum and charge the lines. You'd easily spend that much on gauges, vacuum pump, nitrogen tank, etc. It took me ~10-11 full days effort to complete start to finish, and I probably have one more day of finish/trim work to complete sometime.
I didn't even get quotes because I enjoy learning/doing things, but I think it would have cost me double easy. Also I don't think most contractors would take the time and care I did with my own system including all the detailed planning, careful sealing/insulation, running ducts out of the way so they're not annoying later, returns up high via wall stacks, etc. Sheet metal itself is pretty easy to work with... hit up YouTube and watch some videos by grayfurnaceman and T&N Services. If you are smart and meticulous I'm sure you can do an excellent job.
One last thing I'll add is get into your basement, attic, wherever this is going and really plan it out. Take measurements and update your drawing with actual locations of joists. Cut little holes in places if needed to see where studs land, where stuff upstairs lines up downstairs. Note where water lines, wires, etc are, and what might have to be moved. I made a lovely CAD duct plan too but it only gets you so far in the real world.
I am going through something similar with my downstairs after paying about $10k for a professional to put a 1-ton Mitsubishi in my enclosed attic to heat my upstairs. And that $10k was with me pulling the electric. Nice outfit, but they used flex duct - albeit pulled straight - and undersized the return filter relative to what I’ve since learned on here is ideal.
Re: Fujitsu - I had a good conversation with Heating Works LLC whom I found on Fujitsu’s site: https://contractors.fujitsugeneral.com/listings/138637.htm, but haven’t engaged him for anything yet, and he said for his normal 1:1 ductless head installs he rarely charges less than $5k.
I have no doubt these guys deserve every penny, and I don’t think they’re getting insanely wealthy on the installs, but the tinkerer in me kicks in the same way as you when I see that this gear is like $1600 online, plus lineset, electrical, and incidentals.
If anyone in the Seattle are wants to start something like a GBA reader/homeowner minisplit-install co-op where we share tools and tips, I’d be game for that, and I already have a vacuum pump sitting in my shed from another project that I’d be happy to loan out if it’s powerful enough for these jobs.
Hi Max,
Heck I'd be good with $10k for two one to one's. I tried to take on as much of this project as I could so that from the contractor's perspective this ducted install would be on par with a ductless install. But $16k for a two to one and $23k for a one to one, uh uh.
That's interesting that you talked to Heating Works. I looked them up and all of the others in my area on Fujitsu's website. It looked like a guy working out of his garage or something since I couldn't find any internet presence, so I dismissed them as an option.
I'm pretty much convinced I'm going to do it myself at this point. I'll probably hire someone to come out and do the leak test\charge the lines. I have access to an AC vacuum pump, but I don't know what I'm doing with refrigerant.
Also excitingly for me I think I found an option to be able to have a wifi connection to the heat pump concurrent with a wired thermostat. This manual states that you can have two thermostats in a master-slave relationship connected to one heat pump.
https://hvacdirect.com/hvac/pdf/uty-rvnum-design-technical-manual.pdf
This manual for an Intesis product that Fujitsu might distribute? seems to be able to take the second position, connect to wifi, and still have the temperature control close loop around the sensor in the wired remote. Hopefully that also means when you change the target temp over wifi the wired thermostat reflects that change.
https://www.ecomfort.com/manuals/fujitsu-bd71a33cd2d6fd737ee57c61f15cafd2.pdf
Any of you guys that know and love these Fujitsu's have any experience with this Intesis wifi interface?
I ran one of Fujitsu’s hard wired thermostats to get a more central temp reading compared to the heads. For WiFi I used a Sensibo controller which works great, but needs to be line of sight to the unit and needs power. It can do programs, adjust modes fan etc, can also set upper and lower bounds based on its built in thermostat. Works with ifttt for quick widgets on iPhones.
For a data point, we were in about $8k for four Mitsubishi 1:1 mini splits and misc hardware and plus roughly four Saturdays of labor working together with our friend. He did the connections, and I followed his directions. Check reseller reviews if you are buying online. Our local HVAC warehouse was better if you have a license.
I almost went the DIY route myself on two ductless Mitsubishi systems. If it wasn't for the State and power utility rebates that I would only receive having it fully professionally installed by one of their approved contractors I would have done it myself.
The labor markup they are getting for these in our area are incredible. Pricing out a FH12NA unit with all materials needed based on online prices (and I'm sure a high volume Mitsu Diamond Dealer is getting better pricing than I'm seeing), I was being billed nearly 360-400 dollars per hour in just labor. Comparably a service call would have cost $120 per hour.
Someone said before " I don’t think they’re getting insanely wealthy on the installs", I can tell you high volume dealers, like all of the Diamond Dealers, are. I own a business and I understand "business costs", labor overhead, etc but this right now is a matter of what the market will bear and it's willing to bear a lot and they are doing extremely well with these.
On top of it all their quality of installation was mediocre, the most glaring issue was that both units were installed mildly undercharged because the line set lengths exceeded 25ft (26.5ft and 27.5ft), I was getting gurgling noises from the line-set under high load. Once they finally agreed to come out (which took alot of pulling teeth) to check it they added the correct amount and the issue was resolved.
I ordered a complete Fujitsu setup this evening, 18RLF for the first floor and 12RLF for the second, along with all of the line sets, covers, thermostats, wifi adapters, etc. Shipped $6.2k. Not bad considering the alternatives. Turns out several other engineers that work at my firm had the same idea as me and have installed their own systems. So I can borrow all of the equipment (micron vacuum gauge, manifold, etc). Bonus!
I'm working on finalizing the duct design. I currently have a friction rate of 0.08 using unadjusted equivalent fitting lengths. Is 570 FPM the end of the world for the main trunk or should I really increase the size to get it to ~450 FPM that John Semmelhack advises in his NAPHC presentation? What about the branch ducts, is 600 FPM too fast?
Thanks
The Fujitsu order showed up yesterday. I got the outdoor units installed (prioritizing beating the rainy season on the outdoor work here in Seattle). I've got about 10% of the first floor duct work done. Should be able to hammer it out in a couple of weekends, time permitting.
Hi Guys,
I've got 2/3rds of the first floor duct work in and I have a question. Where do you get a 16x25 filter grille for a 2" thick filter? I found the filter frames below but I need a complete filter grille. If you need thicker than the standard 1" is this something that you have to field fabricate?
https://www.aafintl.com/en/commercial/browse-products/commercial/frames-and-latches/pf-1-pureframe
Thanks
You can often modify an existing filter grill to take a deeper filter.
Price 630FF will take a 2" filter: https://www.priceindustries.com/grilles/products/600-supply-return
Is the 16x25 for 1.5 tonner? If you can, go a bit bigger. You will be a be around .07/.08 static with a new filter. With a bigger filter, if you have the room, you will be able to extend the replacement cycle. Bigger filters are often not much more expensive. 2" is the sweet spot for performance and price as well.
See this from John Proctor, he recommends 2.4sft/400CFM:
https://www.proctoreng.com/dnld/Article_ResidentialACFilters.pdf
For reference, John Semmelhack is at 2.7sf/400CFM here:
https://www.phius.org/NAPHC2018/Think%20Little%20Slide%20Deck%20-%20NAPHC%202018.pdf
I am putting in a 18RLFCD as well and have a 25x30 filter for 3.7 SF/400CFM (I hate this unit, but this seems to be what others are referencing). I have ducted returns though and should end up below .2" static for the total system supplying 1800SF.
*Check out the baffled seal John Semmelhack is using. Anyone know what that product is?
Hi Matt,
Yes this is for the 1.5 ton. I'm planning on using the same filter setup for both so I only ever have to stock one type of filter. I'll take another look but I'm not sure if I can fit much larger of a filter in the only space I have to work with for the 1st floor. I read both of those articles as well as Allison Bailes on here in which he recommends a minimum of 2 sq. ft./400 cfm and thought at ~2.7 sq. ft. for the 554 cfm of the 18RLFCD I wasn't too far off the mark.
https://www.greenbuildingadvisor.com/article/ensure-low-pressure-drop-across-high-merv-filter
I too want to know what gasketed filter grille JohnS pictures in his presentation?
When you modify an existing 1" deep filter grille I assume you cut out the back flange to allow for a deeper filter. What or how do you construct a deeper flange for the 2" filter?
I'll see if I can find a distributor for Price Industries. Thanks
I'm pretty sure that gasket is a surface applied profile. Similar to the link below.
https://www.trimlok.com/rubber-extrusion/epdm-seal/ribbed
It's your static budget, so you can choose how to spend it. Big filters are generally cost effective.
You are at 200fpm with your 16x25. Check these charts:
Merv 8
https://www.aafintl.com/en/commercial/browse-products/commercial/pleated-filters/megapleat-m8
Merv13 - You would be at about .1" static to start.
https://www.aafintl.com/en/commercial/browse-products/commercial/pleated-filters/prepleat-m13
Aaron- I think you are right. When I looked at before, I thought it was setup to edge seal on the filter. This is actually setup to face seal, which make more sense.
Honeywell make a deep filter (FC40R) that fits into a 1" filter frame. Not cheap though, so if you can get something to take a 2" standard filter, it would be better in the long run.
Maybe a diy filter frame out of some sheet metal L shapes at an angle inside the return plenum just behind the grill. If you have an 8" return, you can probably squeeze in a even a 4" filter.
Half of my duct order for the first floor with all of the fittings ended up getting delayed so I finished up the second floor install instead this last weekend. That was a pretty good amount of work, probably four full weekend days, plus tinkering in the evenings during the week. I'm glad I decided to do it myself because the layout of my house forced some creative duct routing, and I'm sure no one would have taken the time to do a clean job like I am.
I had custom transitions bent up by a local HVAC shop and they did an awesome job. The second floor uses metal for all of the main trunks and flex for the room connections just to help reduce sound transfer between the bedrooms.
Regarding the filter for the 1st floor, per Matt's advice I increased the filter size to a 24" x 24". In order to do so I won't be able to use a filter grill. I will have to build the filter housing at an angle below a bench where the return grill will be mounted. Probably isn't clear but I'll post a pic once I finish the 1st floor. I purchased a 24" x 24" x 2" filter frame off of McMaster for this.
https://www.mcmaster.com/catalog/125/732
Hey Cody, couple questions for you, as I'm planning a ducted ductless mini-split install of my own.
Were you able to pull all the required permits yourself since the work was on your own residence?
Did you have someone come out and test/charge the system for you, if so which company?
What local company did you use to build your custom transitions?