Building House. Budget Gone. Still Need Heat

[JonBodd]

I put up a modular cape, 1200 sq ft first floor factory rated at 29000 BTU/h, 800 sq ft upstairs to be finished later but would probably lose another 20k BTU/h. 52' of hydronic baseboard on the first floor. There is NO chimney and no NG available. This is NE CT. Just me living there minimal DHW consumption.

I'm leaning towards doing a propane system with a low-end combi boiler. I'm thinking either a Navien NCB-180 (http://us.navien.com/Product/Page1/Details/NCB-180?Category=NCB series) or a Dunkirk DWB CCB-150 (http://www.dunkirk.com/product-details.asp?id=309).

The price points of the 2 are similar. The Navien is a full Mod/Con and seems pretty popular, the Dunkirk can modulate somewhat but is not condensing. I can't find much about the Dunkirk, but it seems to be a rebranded Biasi Riva.

My main concerns are ease of install and reliability as I will be tackling this myself. The Dunkirk looks simpler and apparently doesn't even require primary/secondary loops. I've read through a lot of Dana's excellent posts on the Navien products and I think I could handle that as well, although I'm still a little unsure of how much tweaking will be required to get it condense with my baseboard.

The Navien has a lot more features, better efficiency, and a better warranty, but I'm still considering the Dunkirk becuase I really want something simple and reliable.

Any input on these 2 boilers, or other ideas for a heating system are greatly appreciated. Also, thanks to all the experts who contribute their knowledge to these forums, it's been very informative reading. If I could afford to hire a skilled tradesman I would do it in a second, I am a firm supporter of the trades but I just can't afford it this time.

 

[Dana]

A "factory rated" 29K design load for 1200' of conditioned space is an awfully high load/area ratio for new code-min construction. I recently ran the numbers on a ~1700' 2-story 1920s bungalow in suburban Boston with some retrofit insulation & windows that came in at ~23-24K @ +10F for the whole shebang, and that's at insulation levels below current code min.

Your lower outside design temps are maybe 5-8F lower than that other house, but it's almost inconceivable that the heat load for the 1200' of lower floor is pushing 30K, and it's probably less than 20K. Do yourself a favor and run an I=B=R heat load calc using fairly aggressive U-factors & air leakage assumptions, assuming an outside design temp no lower than 0F.

At the recent years' cost of propane & electricity in this region it's probably better to go with a mini-split heat pump solution than a hydronic propane-burner. A single 3/4 ton Fujitsu -9RLS3H might cover your actual load for the first floor, if it's a reasonably open layout. It delivers 15,400 BTU/hr @ +5F outdoor temps into a 70F room. (The 1- ton delivers 16,800 BTU/hr @ +5F, the 1.25 tonner puts out 18,400 BTU/hr @ +5F). Assuming a mostly-DIY installation, internet pricing on the 3/4 ton is about $16-1700, the 1.25 ton runs about $1900. Add another few hundred for refrigerant lines and wall-mounting bracket and paying a technician for a few hours of system pump-down/refrigerant charge/ test, you'd be into it for about $2500, and the operational cost would be less than half what you'd pay for propane.

 

[Sponsor]

 

[JonBodd]

Thanks Dana, interesting suggestion to go with a heat pump, I wasn't familiar with the split design. At one point I was considering a geothermal heat pump but I had ruled it out because I didn't want to run ductwork. I have free access to a backhoe and could install a horizontal loop system pretty easily.

Any other thoughts? Seems like oil is on the way out.

 

[Dana]

A DIY geothermal project is pretty risky, and the hardware is quite a bit more expensive than modulating mini-split heat pumps. A geothermal project done perfectly would use about 75% of the annual power of a mini-split., but would take decades to make up the difference in up-front costs. To get that kind of efficiency out of hydronic geothermal also requires low/very-low temperature radiation. With the exception of a slab-on-grade radiant slab, the radiation costs alone would exceed the hardware cost of a mini-split. There is great deal of design risk there too. Unless you have a lot of experience with geo it's probably not the right solution here. (There are many geo systems in New England running at lower efficiency than best-in-class mini-splits.)

By contrast, ductless mini-splits are a "system in a can", pre-engineered with pretty good performance, and fewer things needed for a DIYer (or pro) to get right in order for it to deliver. The first order if business is the getting the sizing right (as it would be with geothermal as well). In your climate the other things to get right are:

1) Mount the outdoor unit where it won't get snow clogged or destroyed by a roof avalanche, etc, which usually means bracket mounting it on the exterior wall above the snow depth line, protected by roof overhangs (under a gable rake is preferable to eaves.) Sometimes mounting it under a deck can work (I've seen that even at ski-resort condos), but wind-driven snow and drifting snow can still be a problem under decks.

2) Only use units with automatic pan heaters on the outdoor unit. Pan heaters to keep defrost melt-water from building up is also important (the Fujitsu xxxxxxH versions have that as standard equipment, but it's a standard additional part for the Mitsubishi -FHxxNA units as well.)

Ductless mini-split is point-source heating, much like a pellet stove or wood stove, so for conventionally insulated houses with a lot of doored-off areas it's less than ideal, and would require some amount of resistance heating in doored off rooms to maintain temperature balance (or to meet code.) But with more open floor plans leaving the doors open is usually sufficient to avoid using the much less efficient resistance heaters. DIY electric baseboards or cove heaters are pretty cheap backup.

The guy with the Boston area bungalow with the ~23K heat load is currently heating with oil, and has opted for a ductless solution, using three 3/4 ton Mitsubishi units. While a 3/4 ton in the upper floor and a 1-ton on the main floor would meet his actual loads, for distribution reasons he's going for a pair of 3/4 tonners on the main floor. I had recommended a 1.5 ton Fujitsu ducted solution for his first floor, but it's cheaper to go with with a pair of 3/4 ton Mitsubishi units there. (Quotes for full turn-key installation are coming in at about $10K for three units.)

In the MA/RI/CT area there is more technical support & distributor support for Mitsubishi than Fujitsu, which is something to consider. Efficiency is about the same between comparable units but the heating output at +5F is a bit lower, eg: The 3/4 ton MSZ-FH09NA delivers 10,900 BTU/hr @ +5F, to the 15,100 BTU/hr output of the AOU-9RLS3H. Your heat load for the 1200' modular is probably under 18,000 BTU/hr @ +5F, but it's probably not quite as low as 10,900 BTU/hr. You'd probably have to bump up to the somewhat more expensive -FH15NA to be covered in their lineup, but it would be a pretty good choice. Like the 1.25 ton -15RLS3H it's about $2K for the basic unit, but you'd have to pay a bit extra for the pan heater to keep the fan from getting dinged up during extended cold/very-cold weather.

If you size it right and use a "set and forget" strategy on the temperature rather than using setbacks, it would deliver a seasonal average coefficient of performance of a bit better than 3.0, but not as high as 3.5 in this climate. So even if you assumed electricity to be 25 cents/kwh in winter (like it was for some of us in southern New England this winter), that means it delivers at least 3412 x 3.0= ~10,000 BTU per kwh, or (1,000,000 / 10,000=) 100 kwh/MMBTU, at a cost of $0.25 x 100= $25/MMBTU.

Propane was down sharply this year, but still averaged about $3/gallon. A gallon of propane contains about 91,600 BTU of source energy, and burned in a 95% efficiency boiler delivers ~87,000 BTU of heat into the house. That's 1,000,000/87,000 = 11.5 gallons per MMBTU, which at this year's $3 price comes in at $34.5/MMBTU, plus the electricity used for pumping & running the boiler, call it $38-40/MMBTU (unless you spring for some awesome-efficiency ECM drive pumps and have a perfect hydronic design.)

At last year's ~$4/gallon propane price spike that's about $46/MMBTU + electricity. It's down to about $2.80 now, but it would have to average under $2 to be cheaper than 25 cent electricity in a mini-split, (an average price not seen in CT since before 2005.) The actual average price of electricity in CT this winter was about 22 cents, and it's likely to be going back down.

 

[JonBodd]

Thanks for the input Dana, you really know your stuff! Unfortunately my house is pretty much the opposite of an open layout, so I am definitely concerned with heat distribution. For DHW, would you recommend an electric tank heater?

You were also right about the heat load. The rescheck for the conditioned space came in with a UA of 223, so around 15.6 kbtu/h assuming 0 F design day, right? The only thing I can figure for the 29k value is that it's based on the 52' of baseboard assuming 170 average water temp. Rather than sizing the baseboard for the load they just put a piece under every window, so it's pretty oversized.

I'm still considering propane because I have an opportunity to buy a 500 gallon tank for cheap, and for tank owners I've gotten quotes as low as $1.19/gal, but I also know it's real cheap right now.

 

[Tom Sawyer]

The last thing you want to do is go with propane. Electric strip heat would be cheaper to run and way cheaper to install. Look into a ducted mini-split.

 

[Dana]

Thanks for the input Dana, you really know your stuff! Unfortunately my house is pretty much the opposite of an open layout, so I am definitely concerned with heat distribution. For DHW, would you recommend an electric tank heater?

You were also right about the heat load. The rescheck for the conditioned space came in with a UA of 223, so around 15.6 kbtu/h assuming 0 F design day, right? The only thing I can figure for the 29k value is that it's based on the 52' of baseboard assuming 170 average water temp. Rather than sizing the baseboard for the load they just put a piece under every window, so it's pretty oversized.

I'm still considering propane because I have an opportunity to buy a 500 gallon tank for cheap, and for tank owners I've gotten quotes as low as $1.19/gal, but I also know it's real cheap right now.

DON'T get sucked into propane on summertime pricing on a low-price year ! The average price this past winter was more that 2x that $1.19 sucker-bait pricing, and in recent years it's been more than 3x that much(!). Since 2008 the average heating season cost of propane in CT has been running about $3/gallon.

A Fujitsu 18RLFCD ducted mini-split would cover your ~15-16K load, and would have a seasonal COP of about 3 in your climate. It has fully specified output down to -5F, and keeps cranking (at unspecified but still significant levels) below that. It's far better to install ducted systems in new construction than to try to retrofit them later, so find a local installer who is up to the task quote it sooner than later, unless you think you're capable of doing the duct design yourself. The hardware itself is about $2200 -2300 (internet pricing), plus mounting brackets, refrigerant lines, ducts, so with a mostly DIY installation and professional refrigerant-charge/test you can probably get there for about $3500. I don't have the full extended temperature tables on that unit, but it'll probably deliver 16K even at -5F, but can modulate down to 3.1K @ 47F, which means it'll lope along at low-speed/high efficiency most of the time, at whisper-quiet air velocities.

Whatever you do, keep the ducts completely inside the insulation & pressure boundary of the house or it won't necessarily hit the capacity & efficiency numbers. If you have a full basement or crawlspace, an insulated sealed foundation with the duct & air-handler cassette under the floor is the right way to go.

With ~16K load and 52' of pre-installed baseboard that's ~300 BTU/hr per running foot, which means at 0F you'd need ~135-140F average water temp to heat the place, which for a propane-burner means you'd probably be in the condensing zone, but so what? The fuel cost is insane. If the gas-grid ever reaches out to you it's worth doing the math, but for now the 1.5 ton Fujitsu would be a far more reasonable solution.

 

[JonBodd]

Thanks for the wisdom Dana. I've talked to someone who installs the Mitsubishi units and asked them to take a look at my situation. Like you mentioned it seems there aren't as many Fujitsu installers or dealers around. If you know any place local that might sell to a homeowner let me know!

Also, I'll probably go with an electric tank heater for DHW. I'm thinking just a 40 gallon Rheem from Home Cheapo, unless there are better options out there you would recommend.

And not to get into the weeds, but it seems like the larger capacity heat pumps are less efficient than the smaller ones, and ones with multiple heads also seem less efficient. Is this right, or am I just not understanding the numbers? If so, is it better to go with multiple smaller units? I'm still trying to learn the basics about these systems, thanks!

 

[Dana]

The Mitsubishi SEZ/SUZ ducted mini-splits all crap out on capacity & efficiency by +5F, and are 25% less efficient that Mitsubishi's better wall-coil type miniplits. But their MSZ-FH09NA wall units are great! A pair of them would handle your load ($6-7K installed, in my neighborhood). They test north of 13 in an HSPF test, and have a large modulation range compared to most.

The larger mini-splits tend to be lower efficiency at the tested modulation level & 47F temperature for HSPF, but at lower modulation levels they're still pretty good. At minimum modulation at +20F the efficiency of any of them is higher than the higher modulated output test at +47F. Modest oversizing is actually GOOD for efficiency, as long at the minimum output is lower than your shoulder season loads. The multi-splits lose out a lot on efficiency due to fact that the minimum modulated load of the compressor is usually more than half the modulated output of the smaller heads/cassettes, which means the individual heads operate at a much lower efficiency than they would if matched with a dedicated properly sized compressor. Efficiency-wise it's best to oversize the heads by no more than 1.5x the design load for it's capacity a the 99% design load.

ALL 2- heat Mitsubishi multi-splits are too oversized for your load. The MXZ-2C20NAHZ has a minimum output @ 47F that's about half your heat load at 0F. Even with a pair of FH09 heads it only musters an HSPF less than 10. The heads can modulate down to a higher-efficiency level but the compressor can't go there with them- it's just too big. The compressor's max output isn't ridiculously oversized (22KBTU/hr @ +5F), but it's modulation range is only about 3:1, which limits it's operating efficiency. With a 99.7th percentile load of 15.6K @ 0F, your average mid-winter load is ~9-10K, and the heads would spend more time cycling on/off than modulating.

A dedicated FH09 single mini-split can deliver 10,900 BTU/hr @ +5F (and still puts out ~7500 BTU/hr @ -13F), yet it can throttle back as low as 1600BTU/hr @ +47F. Modulating at mid-speed at +15F it has a COP of about 3, but running low speed at 40F it'll have COP of about 5.

Another strategy: A cheap ~$800 5-10kw electric boiler and a Taco-007 hooked up to the baseboards as a single would be enough for the place to meet code. If you then put ONE, FH09 in the room with the biggest load it would cover your average mid-winter load for the whole house just fine, and you'd only have to run the electric boiler during cold snaps as your "hail Mary" backstop. It's really not much different than heating with a wood-stove or pellet stove- the remote doored-off room would run cooler than the room with the stove, but if you crank up the stove you can still make yourself comfortable MOST of the time. If you had a more open floor plan it might make sense to go with the FH15, which can deliver 18K @ +5F, but it's min capacity at +47F is 5K, which is well over the heat load of any one room, if it's as walled off as you imply. That's not going to be an acceptable solution for everyone, but it is for many.