HVAC concept selection
I am considering HVAC options for a house in Michigan (climate zone 5). I would appreciate advice from this forum. Am I forgetting some important issues? Is my assessment of the advantages and drawbacks of various options faulty? The house is 2300 ft^2 above grade plus a finished basement (including bedrooms). I calculate a design heating load of about 30k btu/hr using Marc Rosenbaum’s spreadsheet. Regardless of which heating and cooling option I choose, I will be installing an HRV or ERV.
Baseline system – Gas forced air furnace and central AC. This is the standard in my geographic area. The reasons for looking at other options in a high performance house (i.e. unavailability of appropriately sized hardware) are described in numerous articles on GBA.
Option 1 – Mini-split heat pumps for all heating and cooling. I initially discarded this option because natural gas is available pretty cheap. Dana’s comments in another thread caused me to revisit my assessment. I now consider this a viable alternative, although not without some issues. I think I would need four indoor units including at least one ducted unit in order to get good distribution. Even then, I may need resistance heaters for some spaces some of the time. The operating cost would be competitive with the baseline most of the year, substantially better during air conditioning season, and substantially worse during the coldest part of the winter. I haven’t found anyone else in my area doing this. The couple HVAC contractors I have talked to think I am nuts, as does most everyone else in Michigan. My wife can hesitantly accept the aesthetics of the wall units.
Option 2a – Gas force air furnace plus mini-splits for AC. For air conditioning, I would need just two wall units. Basements in Michigan typically don’t need AC. The upstairs unit would probably be effective cooling part of the main floor space underneath it. I would also use the mini-splits for heating for the portion of the year in which they are efficient then switch to all gas furnace for the coldest couple months. I am concerned about the capital cost of having two separate systems.
Option 2b – Hydronic floor heating plus mini-splits for AC. I had previously ruled out hydronic floor heating based on several columns on GBA. The capital cost concern with 2a is even more of a concern with 2b. I am revisiting this option based on comments from multiple people (not on GBA) about how much they love their radiant heating systems. If I install floor heating only in bathrooms and a few strategic other spots, do I mitigate the drawbacks of hydronic without losing the advantages? I am envisioning 600-800 ft^2 of heated floor. To what extent would the system cost scale with the heated floor area? A system like Warmboard does not have the high thermal inertia. Can I reduce the boiler cost by using a second water heater as the heat source or by using an indirect water heater for DHW?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Reid,
It sounds like you understand all of the issues. All you have to do is make a decision.
There's a lot to be said in favor of the option you called "Baseline": a conventional natural-gas fired furnace and a split-system air conditioner, both with forced-air distribution. These systems are affordable and local contractors can maintain them. Keep your ducts inside your conditioned space, and make sure that you buy right-sized equipment (or, in the case of the furnace, as small a system as you can find -- hopefully a modulating furnace), and you should be fine.
A hydronic floor system while very nice, is going to cost more than 2x what a heat pump solution would be (even with 4 mini-splits), and WAY more than a condensing gas furnace heating solution. Using something like an HTP Versa for both hot water a space heating keeps the cost of the boiler down HW heater and buffer tank down, and allows you to micro-zone with impunity, but the high cost of low temp floor radiation doesn't change. Low temp panel radiators are substantially cheaper, but not as cushy as heated floors, and still more expensive than a condensing gas furnace solution or a mini-split solution.
Everybody thinks mini-split solutions are nuts until they've seen them working at temps way sub-zero. When heating oil prices were very high over the past 5-6 years they became very popular almost overnight in New England in areas off the gas-grid.
At 30K heat load (at what outdoor design temp?) seems quite high for a high-performance house that size in zone 5, but may be about right for code-min houses that size.
You may want to read the following discussion and the very last response .
https://www.greenbuildingadvisor.com/community/forum/passivhaus/44561/skipping-radiant-floor-heat-passive-house
Then this recently released report , keep i in mind that your house may well be code compliant as opposed to something that could be considered a pretty good house . Please feel free to open that report through this link .
https://www.linkedin.com/grp/home?gid=1110797&sort=recent&trk=my_groups-tile-flipgrp
A recent project we designed the heating system for in Ann Arbor can be seen here as a reference to a pretty good house
Usually, it doesn't bother me if people think I am nuts. There are two times that it does bother me i) if they are right, or ii) if it makes them unwilling to installing what I specify.
Martin, your comments appear to be directed to the option I labeled Baseline as opposed to the option I labeled Option 1. That option gives me the lowest cost, lower energy bills than a code house, and adequate comfort. In the spirit of high performance house, I am willing to spend some extra to increase comfort or further reduce the energy bills. Of course, it depends on how much extra, how much it impacts the energy bills, and whether the house would actually be more comfortable.
Dana, the design temp is 0F. I hope you are correct that my heating load is lower than 30K. I don't think that would change the design of the heating system much. For conventional systems, I will already be at minimum size. For the mini-split system, the number of units is dictated by distribution and each unit is already minimum size. In addition to thinking I was nuts using mini-splits for heating, the HVAC contractor also thought I was nuts when I said my heating load was 30K. He thinks it is way higher. I am much more inclined to believe you or believe my calculation than to believe him. (I am continuing to look for alternate HVAC contractors.)
Where in Michigan are you located Reid ? There are several quality contractors that get it and would not have a problem installing right sized equipment . The fact that one would not entertain your heat loss as fact should immediately disqualify him from consideration . You will need a well designed system and specifications .
Richard, the house is to be built in Linden, MI which is near Flint.
Reid,
You're right -- I meant to write "Baseline." I've corrected my answer.
There is no doubt the you can use ductless minisplits and ducted minisplits to heat and cool this house. If you go that route, I think that it would be silly to also install a furnace or a boiler. The minisplits can do everything -- if you want them to.
Only 41 miles to Ann Arbor . If you'd like the names of a couple of contractors that could do whatever you end up doing properly you may contact me . [email protected] . There may also be a homeowner or 2 in Michigan that can tell you how their experience with designing , installing and living with radiant has been .
My experience tells me that there is between 200 -300 CDD in that area
Richard, I read through the CARB Building America report you link to above. It is interesting that the mini-split systems that did really poorly in Jan and Feb were being used to supplement hydronic systems whereas the ones that were at least close to expected performance were the sole or primary heating source. That challenges my expectations about my options. I was thinking that the 2a and 2b options, although they would be more costly, would get better efficiency by only using the mini-split in the situations in which they are most efficient. I don't know how the controls of the systems in the report were coordinated. I can imagine that getting the systems to cooperate in a close to optimal manner would be difficult.
I was surprised to see the COP of one of the systems actually fall as the outdoor temperature got warmer. That illustrates the danger of too much oversizing that Dana has mentioned in other threads. The report illustrates some other ways people can end up with low performance from a mini-split, like frequent thermostat set-backs. Fortunately, I do have a lot of control over occupant behavior. Educating myself in forums like this would reduce the likelihood of me botching it up, but I am not immune from mistakes. The report mentions that positioning the units high on the wall, although great for air conditioning, reduces COP in heating mode due to warmer inlet air temperatures. My plan was to install the units high on the walls under the tallest ceilings in the house.
I would say go for a normal furnace. HVAC expert needed for this and the mini splits. If your are so inclined I would also suggestion a EPA wood stove. Its carbon nuetral and is occupant controlled. If you want less work a pellet stove would work, but fuel could cost more.
I have encountered many delays getting this building project started. That has given me time to add some more HVAC options and flesh out the baseline case some more. I had a manual J done which predicted 32K design heat load and 18K design cooling load. I think the load is probably actually going to be less than that. I would appreciate comments on these revised choices:
Baseline - Traditional forced air using the brand that my HVAC contractor is used to. The furnace would likely be a condensing two-stage 40Kbtu unit with a variable speed ECM blower. I haven't been able to find the output at low setting but I suspect it is more than my design load. A 24Kbtu inverter driven heat pump is available. Due to seasonal variation and solar gain variation, I expect to need a 3 zone duct system.
Dettson - This company from Canada offers products they call a Right-Size System. Condensing, modulating furnaces are available in 15Kbtu increments down to 15Kbtu. They modulate to 40% of rated capacity. An 18Kbtu inverter driven heat pump is available. They have a ducting system designed to work with these components. I am in the process of getting a price quote for the components. My HVAC contractor would need to install it and has no experience with these products. I don't think that installing them would be a technical challenge, but the contractor might shy away from them for business reasons.
Option 3a - Hydronic with heat pump. Upgrade the water heater to something like a Phoenix and use an air handler to distribute the heat. I could probably eliminate one zone of ductwork by using fan coils in the basement. That would give me the option to not heat that area when it isn't occupied. (Bedrooms and rec room for my college age kids are located there.) I have attached a diagram of what I envision. No cooling would be available via the fan coils, but that is ok for the basement.
Option 3b - Hydronic with chiller. Like option 3a except use a Chiltrix chiller instead of the heat pump. I have also attached a schematic for this. The advertised COPs of this product are outstanding, but I am not sure I believe them. Since the hydronic fluid would go outside, it would need anti-freeze, necessitating segregating the HVAC water system from the potable water system. Many air handlers don't work with chillers, but First Co makes some that do. I could use a fan coil or two upstairs and eliminate any need for zoning the duct system.
I have no idea how my HVAC contractor would respond to these last two options. They might run away screaming.
Reid , Have you ever heard of radiant ceilings ? Very nice option for heat that does a fantastic job and is much less costly than floors . Floors will still be at a temp comparable to floor install , lower water temps are also acheivable due to the fact that in this scenario there is no Unheated space with loops below . More output per sq ft , less resistance since drywall is the only thing between the tubing and the space . I don't know if you looked at the report I attached above from a home in Ann Arbor . That gentleman is right on track to use 700.00 dollars per year to heat , make hot water , cook and do laundry in a similar climate as yours .
Don't be afraid to use ducted system and the water heater as the source for a coil in air handler . I do suggest letting only the most talented design that system , that rules out lots of option unfortunately . I do have access to a few of those if you wish . A Phoenix light duty is more than adequate for your home at the provided loads , my feeling is they are even a bit exaggerated . As far as cooling goes you could also use water in a coil , probably be able to cool using an average water temp of 60* and control the outdoor unit so that it comes on only when it can run a full cycle and raech max efficiency . Not all that complicated or expensive either .
Aleways remember that your water heater should store at 140* or above and you should mix down to 120* , very good protection from many nasty things in potable water systems . Also increase the storage capacity and you'd swear you have a tank twice as large as you do , holds the burner out longer between calls also . An ODR mixing valve for the space heating operation will also allow on eto use that and vary water temps instead of using VAV or the like . Let's face it , 18,400 cu ft is always 18,400 cu ft and all the air needs to be conditioned all the time . No gimmicks , no crazy math that we find out is flawed later after many have been installed . Just Mother Nature's natural Environmentally friendly refrigerant with a bit of help with nasty refrigerant that is minimally inside where you are . NOTHING is as efficient as these systems , . As they say , pay now or pay forever .
To run with a hydonric chiller for heating in climate zone 5 would require going with anti-freeze, and the system really has to be designed, not hacked. If you're on the gas-grid gas-fired heating would probably have a lower marginal operating cost than a chiller based system, but if done right the chiller system could be cheaper than propane or oil.
The stated COPs in the specs are for the chiller-only, and not the distribution pumps & blowers, making direct comparisons with a mini-split more complicated. A mini-split has only one power input, and though the as-installed-as-operated efficiency will vary from the bench-tested HSPF test, it's still a pre-engineered "system in a can", with very little design risk.
The Chiltrix does include a pump in the system, so I think the COP includes pump power. And Chiltrix also has reasonably priced fan coils with very low power ecm fans for much lower prices than the absurdly high price I paid for a Jaga ecm fan coil. It approaches the system in a box advantages of a mini split.
Rich, something you mentioned that I would like to understand more: radiant ceilings being cheaper than radiant floors. I would have thought the costs would be pretty similar. What is behind that difference?
Labor is less intensive Charlie , resistance is less so you can widen the spacing and get the same result without the furniture , carpet problems , cabinetry . Most usually choose to stay with the tight spacing and use lower water temps however . Higher allowable surface temps also can lessen the requirement for mixing / temp control devices . Alot of the expense is removed due to these factors . Very responsive and very comfortable . Been being done in Commercial spaces for decades now , bet not many realize that . Example , the radiant boards I use directly under hardwood cost 4.25 sf , the channeled , graphite coated ones I use under tile and in ceiling applications cost 3.60 sf . That one line item equates to 1,300.00 in a 2000 sf home , the labor would also cost about 500.00 less , lose the boiler and 850.00 worth of now unneccesary controls and the same house costs about 7,500.00 dollars less than the guy next door who had his supplier design it , needless to say later he will have to have someone more than likely fix issues .
The water heater that is in every house as long as it is capable can easily do both jobs in most cases . That eliminates the cost of boilers also , If only some would realize that this is no longer your Uncle's radiant the entire home performance and energy efficiency sectors could be even more helpful .
Richard, wouldn't a radiant ceiling system significantly increase heat loss through the attic?
Charlie, your statement that Chiltrix approaches the system in a box really only applies if one sticks to the Chiltrix supplied fan coils or something very similar. The more I attempt to figure out how to control the Chiller-Air Handler-Water Heater concept, the less I like that concept. The chiller and the air handler both want to be in charge and have other components respond to commands. The other components are not necessarily set up to respond to those commands the way that makes the system work as intended. If I pursue the air handler idea, I think I will stick with the type of heat source and cold source that the designers of the air handler had in mind.
Reid , is that a loaded question ? Most homes now have attics inside the thermal envelope and those that don't suffer heat loss also . Will you suffer any more heat loss upward with radiant than another medium ? Radiant also requires insulation on the opposite side of the panel than the space to be heated . Answer is , basically , not any more than any other heating system . Radiant heats objects and surfaces upon contact , not air per say .
Rich, my question was not intended as some sort of gotcha question. My house will have an unconditioned, vented attic. With most types of heating system, the delta T through the attic will be between the indoor setpoint of about 70 and the outdoor temperature. With a radiant ceiling, the ceiling would be maintained at a higher temperature than the indoor setpoint, so the delta T would be greater. Of course, this could be offset by increasing the amount of insulation. The same issue would come up with radiant floors on the lowest level of the building.
Reid and Rich,
The nice thing is that if you want to mitigate the loss from a hot ceiling to the attic, a little extra cellulose is cheaper than extra sub-slab insulation would be.
Rich, I agree that using third party fan coils with Chiltrix could be tricky, and it would be best to get ones made to work together.
Reid ,
The assembly that I recommend and use myself is of a sort that damn near eliminates thermal bridging between the attic and living area . Please look at the following that gets attached to the bottom of the framing and has tubing inserted right into it . Full contact between the product and the drywall and outstanding output at a higher rate than other products due to the unique transfer properties of the graphite . I believe and have witnessed that this product outperforms anything on the market . Look specifically at the SunFoam Complete product , also works great for radiant walls , a hardly talked about opportunity that is also possibly better for very tight , well constructed homes , as yours will be , I think .
http://www.sunboardpanel.com/Products.html
http://www.sunboardpanel.com/Sunboard_Panels_-_Heat_Output_parameters_v1.pdf
I have a test house in my programs that I can move anywhere on the globe that is very similar to the homes discussed here , maybe even a little worse , which is good for demonstration here .
This house is presently in Thunder Bay Canada based on a discussion here awhile back . I changed the design to 0* w/ 22 mph wind and the fluid temp required to heat the home to 70* is 92* with surface temps (drywall) of no nigher than 78* . Interesting to say the least .