Air-source heat pump for existing home: Trane XV18 or Bryant 280A or something else?
Sorry this is a long one, feel free to skip to the end if you don’t want the background. We have a fairly large home (4600 sq ft above grade, plus 1800 sq ft of finished basement, about 400 sq ft of unfinished basement on two ends for two utility rooms (one HVAC, one electric) that we bought as 2nd owner about 18mo ago. Climate zone 5A. Built 2000 (sold in 01), by a local builder with a good reputation but still production builder (not full custom). So this is not what I would call a high performance house, but we are doing what we can to make it more so where cost effective.
I know that’s not quite the point of the site but from reading the Mechanical forum here it seems more educated of most, particularly doing things that make sense vs the way they have always been done).
We have done a quick air sealing job which is free from MassSave and its about 4 ACH-50 right now. Believed to be R-13 in the walls including finished basement (but not confirmed) and known R-30 attic insulation, all fiberglas bats. Windows are original from 2000, dual-pane aluminum clad wood frame (no idea on energy ratings, the manufacturer, Malta, is out of business).
We just installed a 11.7 kW DC / 9.75 kW AC PV array (would have preferred to go a little higher on the DC but this was fixed ratio given the panels we were using and the fact that we were using microinverters, AC limited for net metering cap exempt). We plan to add cellulose blown-in in the attic to take it to R-50 or R-60, but only after we do HVAC work because the upstairs units are in an unconditioned attic (see more below).
I have not run a manual J myself but at least one contractor I am working with is running one (and I plan to myself once I figure out all the input data).
My general view is that if I can see some path to break-even in 10 years or close I am willing to make the investment, or if its an inexpensive increment to get higher performance. Plan to stay here at least 20 years but who knows, and planning for modern HVAC gear to last that long itself seems like a gamble.
We are in an area with no natural gas, so we are on oil heating (hydronic-air) right now. We have a propane tank for cooking but I don’t see how touching that makes any sense for heating (we get killed per gallon). The heat works fine but I would prefer to be on something cleaner plus we over-sized our PV array with the intent of converting at least some of our heating to ASAP.
I looked into GSHP but today our house is designed in 3 zones each with their own air handler and it just seemed like it got too expensive to install a 3-zone GSHP system, particularly since the incentives in my state (MA) while significant require you to install GSHP for 100% of the building heating load (no ASHP or supplement allowed), and everyone I talked to was proposing closed loop GSHP, which didn’t seem to have crazy-different COP’s than ASHP (obviously more consistent temperature of loop vs outside air, but not enough to justify the cost even after rebates and ITC).
We have a new incentive program from MassSave that gives $1k/ton for ASHP that offsets oil, plus if I complete the whole house on modulating ASHP that is on a specific list (a subset of the NEEP list) I can get about $5k of pre-minted APS credits (basically renewable energy credits). Given our climate my focus here is all about heating, any improvement on cooling is a bonus but not meaningful cost wise (we spend under $1k/yr cooling, although comfort increase of modulating is interesting).
Right now the current HVAC situation is as follows:
DHW: Glass filled 50 gallon direct oil fired Therma-flow unit the previous owners installed only 3 years ago. I don’t care for this from an efficiency perspective but given its age its not my focus and its hard to argue with the recovery speed. Probably will replace with an 80 gallon heat pump unit when it dies.
Boiler: Burnham PV75WC from 2000. Rated at 166MBH IBR and 191MBH DOE. Has electronic honeywell aquastat with outdoor reset although I honestly thing its oversized and I could kill the outdoor reset and always run it at like 150/130 deg and it would be fine. Right now we setback all our thermostats during the day and at night (would not if we had modulating), so the recovery speed is nice but the boiler itself still runs on very short firing cycles. Taco 007 circulators – if we don’t touch the rest of the heat I plan to switch these out for at least 007e’s if not VT2218’s (not sure how much benefit I will get from the delta-T without zone valves though, but I think it will help in reducing piping loss to the 2nd floor units since the 007’s are just fixed speed).
1st floor: 3.5 ton 13-SEER R-22 outdoor Concord (Lennox) condenser that the previous owners installed only 2 years ago (2016 date code). Must have been one of the last R-22 units sold. Original FirstCo hydronic air handler (PSC motor, single speed, 48k BTU sized).
Finished basement: Baseboard electric heat that we have never turned on (for all I care it could be removed). The 1st floor duct work was cut in here when this was finished. Unfortunately a lot of it directly to the rigid trunk, which I think would negatively impact pressures (although far vents on 1st floor still have plenty of airflow). All of those registers are closed, and the basement is comfortable during primary heating and cooling seasons just due to leakage with the registers closed, it can get cold during the shoulder seasons where the runtime on the first floor unit is low, but we are ok with that. In an ideal world this would be zoned, but that seems difficult after the fact with how it was cut into the trunk.
Master Bedroom: This is somewhat large (about 800 sq ft including bathroom and closet) had its own air handler and 2 ton AC unit. AC unit is an original 10-SEER from when the house was built and I believe has a refrigerant leak. We were not getting freeze-ups but were getting very bad performance out of it. Air handler in unconditioned attic. Since this was just one room and there were some great rebates to be had, we put a Fujitsu AOU15RLS3H+ASU15RLS3Y ductless in this spring. It has integrated controls with the hydro-air system as backup heat but these don’t really work great and the unit performs well down to pretty low temps so we plan on not really using these (was required for one of the rebates). It seems to perform fine and we are happy with it.
Remainder of upstairs: 3 bedrooms on a FirstCo air handler and 2.5T 10-SEER R22 outdoor unit. Unconditioned attic. We don’t currently use this space much so we have been keeping it at safety temps of 50F (heat)/82F (cooling) but we plan to use it more in the future. Replacing this plus the new rebates is what got me down the path of thinking about upgrading it.
We are consuming about 1100 gallons of oil a year including DHW, my guess is about 100-150 of that is DHW, so that puts about 1000 gallons across the 3 zones, with the upstairs getting minimal use today, so lets say 650 gallons first floor, 300 gallons MBR, 100 gallons upstairs today. MBR will go to 0 next year, and I expect upstairs to increase to closer to 400 gallons if we were heating it to livable temps.
Our electricity is pretty expensive here (21.5c/kWh today but there is already a rate increase on the table, lets say 22.5c/kWh going forward).
Here is where my thoughts are:
Upstairs: Its tempting to go cheap here and just meet the requirements of the rebate (9 HSPF) but from what I have seen so far there is real payback to be had on modulating units in heating mode if only due to their better HSPF/COP’s but obviously additional comfort benefits. I ran some degree-day analysis using prior year’s temperatures (could not find daily averages anywhere) and COP data that I could get from various units. Plan is to 1) up the insulation both attic generally and around duct work and 2) build a rigid foam insulation cap for the unit itself post-install. Fully conditioning my attic would be really quite expensive given its size.
1st floor/Basement: I was not originally going to touch this due to the newer condenser but it seems like most of our runtime savings would be in this zone, and it would be required to get the APS credits. This is also the zone I would care the most about the additional comfort modulating would bring.
I’ve had a few dealers come out. Most of them I have eliminated in not really knowing what they are doing. One does seem to (started a Manual J/D, but has not shared it yet, believes both systems are oversized today and the correct size is 2T for upstairs and 3T for first floor) and is a Trane dealer but will also sell Bryant. All have been more than I really want to spend – cheapest quote I have gotten is about $12k per zone for a 16-seer single stage unit – scope involves flush and pressure test existing refrigerant lines, install new air handler with hydronic coil and condenser, plumb it all in, remove old stuff, some slight ductwork adjustments to fit the sizing of the new units but no actual duct changes. (Existing ductwork had some reasonable thought into it, all insulated rigid rectangular trunks with short flex runs to the registers, returns in every bedroom, etc)
Trane XV18 (or American Standard equiv) – ignoring XV20 as it only has cooling improvements (heating performance seems to be identical)
Pro’s: I think the Trane brand is higher quality, factory hydronic option (fits in the air handler in place of the heat strips), Nexia seems to be better than Evolution thermostat (not thrilled about either).
Cons: Low heating capacity at low temp like most ducted heat pumps, not currently on the APS list for the state credit (I believe due to a data error)
Bryant 280A (or Carrier equiv)
Pro’s: Great low temp heat output, and best COP’s.
Con’s: Questionable reliability (early units only?), native hydronic option discontinued (but relay kit available for external coil), weaker thermostat.
The only other thing I have seen that is similar to the Bryant in its low-temp heat output (other than a low static pressure system on a mini-split VRF condenser, which I think would be too difficult to connect to our duct work) is the Rheem RP20 but this seems pretty obscure. I looked at the Lennox XP20/XP25 but these really seem focused on warmer regions (pretty weak low ambient heat capacity), same as the Amana/Daikin AVZC20/DZ20VC.
- Does keeping the oil boiler and hydronic as backup heat make sense? My breakeven COP between oil and electric is about 2.8 (includes oil efficiency, ductwork loss would be the same for both, includes the blower electricity on heat pump but not on oil, but counting electric strip heat as “1.0” also doesn’t include the blower). So in that sense, it clearly makes sense. But on the Bryant, since it produces enough heat at low temp, it’s unclear if I would ever really use the oil very much. Install cost is probably slightly cheaper on hydronic is already there and heat strip would require a bunch of electrical work that I believe would far exceed the hydronic coil + labor, but the maintenance on the oil is not cheap (annual cleaning/service plan) and if the boiler ever breaks I’m in for big money. My current plan is to do the hydronic at install, and if I loose my boiler, do the work to change to heat strips. I like the idea of keeping the boiler as it lets me adjust my switchover point as both oil and electric prices move around. We have some over-sizing on our PV array to cover this, but I like doing the math at market electricity price as if we are fully heating on ASHP we will blow through that in no time.
- I think the Bryant makes the most sense spec-wise, but is there anything else I should be looking at? Has the reliability improved? How painful is the thermostat? The Chiltrix stuff looks interesting particularly given that we already have hydronic air handlers, but it just seems too out there.
- Is that type of pricing ($12k for single stage 16-seer, $14k for Trane XV18, both 24k BTU) fair given that I am in an expensive labor market and it includes the hydronic work? It seems higher than what I expected, but the 3 quotes I have gotten have all been in the same ballpark. Still trying to find another dealer who is affordable and knows what they are doing.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
If you go on http://www.energyswitchma.gov/#/ you can usually go 100% renewable on a 1-3 year contract and still beat the Eversource / National Grid base rates, which will change the COP at which the heat pump is more cost effective than the fossil burner.
I've given up on trusting ANY HVAC contractor to get the load numbers right. Hire a competent third party such as a certified professional engineer or RESNET rater to run the room by room, zone by zone Manual-J, and use the fuel-use history to sanity check those numbers. HVAC contractors are human, and are far more likely to put conservative rather than aggressive assumptions into the load calculation tools, which is the exact opposite of the instructions explicitly stated in Manual-J. Oversizing by 2x or more is almost a given when you let the contractors run the numbers, whereas to get the best efficiency out of a heat pump with limited modulation you'd want to oversize by no more than 1.2x, or even UNDERSIZE it by 10-15% and include enough auxiliary heat strip to cover the absolute peak loads.
Whether or not 5 tons of ducted heat pump makes sense or not depends. For Mitsubishi or Fujitsu cold climate heat pumps with any ducts fully inside the thermal envelope of the that would probably be sub-optimally oversized, but maybe not for the models in question. A lot depends on your 99% outside design temperature and the actual loads.
If you have some K-factors or exact dates and quantities from mid-winter fill-ups and a ZIP code it's possible to get a reasonably accurate ball-park on the load numbers using these methods:
There is no way to separate distribution losses from ducts in attics or hot water use, but it estabilishes a fairly firm upper bound on the whole house load.
You don't say where in MA you're located, so it's hard to take a stab at it based on your 1100 gallons/year number.
There isn't enough information here to make a call as to whether keeping the oil burner as back up makes sense.
1) We are right next to Framingham which has a 6F 99% design temp. We have a town-wide power supply agreement at 10.8c/kWh which is very advantageous in winter compared to basic service (particularly if we move heating load to it). Thanks for the site, I hadn't seen that one before, but looks like the cheapest for our zip is 10.2c/kWh which doesn't make sense to switch to for 6mo assuming prices go back up for the winter like they have the last few years. This is National Grid territory.
2) Degree day calculation - I had already done the same math for my savings calculation but not done the part to come up with the design load. I took one year's of fuel usage and aligned it with 1 year of degree-day data (aligned to those fills) and that is how I am doing my break-even calculation on the heat pump (so its for a sample year, Feb 2018-2019, not an average year). And then for every day I looked at the temperature, calculated the COP based on the data I had for the equipment (my Trane dealer got me good data for that air handler and condenser , and Bryant publishes theirs publicly, others are harder and I only have the 47F/17F numbers and a 5F number if they are on the NEEP list), and then assigned either an electric consumption or oil consumption value based on the MBH per degree day times the cost per MBH at that temp. I'm assuming 86% efficiency on the oil boiler (this is off the annual test, not sure if that is accurate enough or not, factory rating may have been like 84%) and then 15% average duct+water loop loss across the two floors (assuming 30% on the upstairs and like 5% on the downstairs). Applying same duct losses to the heat pump, so it doesn't really impact the COP.
For the actual calculation - I didn’t subtract the DHW but I’ll call that a hand-waive with the fact that we are under-heating part of the house right now and I came up with 75.6MBH. 15MBH is covered by the Fujitsu in the MBR, leaving 60MBH for the rest of the house which seems to well-align with the Trane dealer who did the math at 3T+2T (he did have an engineering degree of some sort, FWIW). I promise I really didn’t massage the numbers to get there, it actually came out that way.
3) I forgot to list it in my pro/con's above but the modulation limit is one of my concerns with the Bryant - the Bryant supports 40-100% but the Trane supports 25-100%. However do you think that duct work not designed for a modulating system can make use of the 25-40% I would be loosing?
I think you should consider abandoning the attic ductwork and replacing them with on the wall mini split units.
If you choose to keep the duct work in the attic consider making a conditioned attic by sealing all the vents, removing the insulation on the attic floor and adding insulation to the rafter.
My problem with ductless units is we have a hallway bathroom, which would then remain unheated. I guess I could add an all-electric heating solution for this, but that seems really inelegant.
I've been told $15k+ to convert to a conditioned attic based on the size of the roof. We just did a roof and it was 54 roofing squares.
I had considered these, but neither one seemed that appealing.
>"My problem with ductless units is we have a hallway bathroom, which would then remain unheated."
If there are no exterior walls the heat loss of that bathroom is negligible- less than the throw-off of one sitting human even when it's +6F outside (Framingham's 99% outside design temp). Even if it has one small wall and a window and a heat load worth addressing, an electric radiant floor is a great thing to have in a bathroom. An electric heated towel rack is a nice feature too.
Mini-duct cassettes work well for splitting output between multiple low-load rooms.
Running the fuel use load numbers on an annualized basis has a huge error, exaggerating the load numbers, due to the very low efficiency during the shoulder seasons and summer. Wintertime only fuel use minimizes all errors, including hot water use (much smaller fraction of the total in winter), standby loss (lower due to a much bigger duty cycle) and solar gain (due to fewer daylight hours.)
Got it. Reason for annual is that I do not have really precise dates on the winter fills, because the oil dealer prints the tickets a few days before they actually fill, so the date on the ticket is wrong. So I was worried that would induce much more error (3-5 days off on a monthly fill is significant) but I will try an align it for a few fills across the winter.
I do have a good Fujitsu dealer, and I think it would actually be less expensive, even with 3 heads, so may look down that path if I can get wife approval for it. Individual temp control is a nice bonus.
A 4-ton Mitsubishi MXZ-8C48NAHZ can deliver 54,000 BTU/hr at +5F, which is probably going to be enough to cover the load of a Y2K vintage 4600' house + 2200' of mostly-below-grade basement, if any ducts are inside of conditioned space, with up to 8 zones:
If using any of their full-sized air handlers are used (not necessarily recommended if the ducts are outside of conditioned space) there are heat-strip options to cover any shortfall at low temps.
Mitsubishi's regional training and design center in Southborough is pretty close to you, maybe next door if you're on the southeast side of Framingham. Finding competent Mitsubishi installers isn't that hard in your neighborhood. (But I still wouldn't let any of THEM run the load numbers.)
I still have a hard time believing 75,600 BTU/hr is the real heat load at +6F unless it leaks air like a sieve or wasn't even build to the code minimums 19 years ago. Even with air handler driven infiltration from unbalanced ducts and the direct duct losses to the attic that's high.
Unlike Carrier, Trane, et al the compressors used in the -NAHZ Mitsubishi units are designed and optimized for capacity & efficiency at low temperatures, (like the cold climate Fujitsus) and despite comparable (or better) HSPF numbers from the US vendors you'd probably do better on both efficiency and capacity with a 3.5 or 4 ton Mitsubishi than a 5 ton Trane or Carrier.
If you went completely ductless you could use the fossil burner as your "Hail Mary" backup initially, then mothball or remove it when you figure out it's not really needed.
You will have better luck getting advice on this web forum if you make your questions shorter and more focused. Few readers are going to read such a long question.
If you have 4 or 5 related questions, it's often best to ask one question at a time, and then post a second question later in a new Q&A thread.
You may be interested in reading this relevant article: "Ducted Air-Source Heat Pumps from American Manufacturers."
As unappealing as the bumps on the wall are they start looking better when you consider keeping the ductwork in an unconditioned attic represents no less than 25% penalty in operating costs.
This is in reply to both you and Dana, since I guess I went too deep on the threads there.
Yes, I think I am reasonably convinced now to do ductless, at least for the upstairs and then size a reasonably sized system for the downstairs. I am still getting some Carrier/Bryant quotes as those at least have reasonable heat output at low outdoor ambient which the Trane really does not. I saw AqueCoil makes one that fits the Mitsubishi air handlers, so thats an option with the boiler instead of the heat strips, not sure if its worth the hassle but would mean no new electrical run.
With increased rebates for ductless, and a Fujitsu Elite installer that I know that has very reasonable costs, it seems to come out at almost half the out of pocket of a ducted system, gain the lack of loss in the attic, and individual room control.
Honestly the biggest issue I have now is having to have a condensate pump on the wall as well. Even the ductless style ones still look pretty ugly, and I don't think they can fit one of the integrated ones into the 7k/9k Fujitsu's.
The rebates require keeping the legacy system around on integrated controls (this horribly hacky 24V converter that just sets the ductless unit to a fixed target temperature when called - we will basically set this to a safety temp and do the primary control on the units themselves for normal mode), so we will do that to start, but I think with H series Fujitsu's we will never use it.
Then we can some-day rid ourselves of the boiler.
Maybe when we swap out the DHW, which after looking at some runtime numbers on my Sense for the DHW burner and the boiler burner, and adjusting by the nameplate GPH on the oil, seems like a much more significant amount of my oil burn than I expected (maybe like 30% instead of 10%).
Before diving into a multi-split solution, take the time consider all of the issues discussed in this podcast:
Oversizing of multi-splits is rampant.
Please post a copy of the rebate form that requires you to keep the legacy system that would be something I had not seen before.
Be careful when shopping for mini splits many dealers play a bait and switch game. That offer to quote minis but they plan a head in every room, the system is often 5 times larger than needed. Big surprise when they offer a conventional system for much less money.
There are a lot of people on this site with minis and I read about very few condensate pump issues.
My Rheem 17 SEER heat pump heats my house by itself below 8°. If you look for a few key features a convention system can work well. Variable speed compressor, electronic expansion valves one indoor and another outdoors and finally the thermostat indoor unit and outdoor unit must digitally communicate with each other.
The dealer we are using is someone we have used before and I don't expect any nonsense out of him.
CWV: For the bedrooms, we want a head in every room but are planning two 7k's and one 9k on a Fujitsu AOU24RLXFZH. We want a head in every room because we want to be able to have it operate with the doors closed. I don't think there will be problems with the condensate pump, I just think it makes the unit even more ugly.
The Fujitsu system with 3 heads is much less expensive than a modulating conventional system which I was getting $12k+ for quotes on.
The 7k is the smallest head they make. Minimum heat output on the condenser is 6.8k, but given that all units are similarly sized and all rooms are similarly sized (one is larger with an attached bath and that is where we would use the 9k) we would expect then to be all running basically at the same time except for solar load impact.
For the first floor we will do some sort of ducted system, since that ductwork is in conditioned space anyways.
The rebate form is here https://www.masssave.com/en/saving/residential-rebates/electric-heating-and-cooling/ - we would be looking at this one: Displacing Oil or Propane Air Source Heat Pump Non-Ducted $1600/ton - and it mandates the integrated controls tying in the legacy system in for "dual fuel" style operation.
>"For the bedrooms, we want a head in every room but are planning two 7k's and one 9k on a Fujitsu AOU24RLXFZH. We want a head in every room because we want to be able to have it operate with the doors closed. "
That's classic bad thinking regarding multi-splits. Run the actual load number for those bedrooms before throwing a 2 ton cold climate compressor at what's likely to be less than 3/4 tons of total design load. A single 7K is good for 8100 BTU/hr in heating which might cover the combined load of all three bedrooms.
A 7K mini-duct cassette might work efficiently as one zone of an AOU24RLXFZH as long as the other heads/cassettes were appropriately sized for their loads.
A few years ago I was involved in a project where two AOU24RLXF (not-"H") multisplits heated and cooled a sub-code 3000' house at an outside design temp of +12F, and it was oversized for the actual loads enough to cut into efficiency. Yes, there is individual zone control, but even when it was into negative digits outdoors the multi-split serving bedroom heads and the compressor were cycling.
The other AOU24RLXF serving the bigger and less well insulated spaces with a 12K and pair of 7K head and a better matched load does a lot better, except during the shoulder seasons when the 7K in the low-load basement office causes more cycling when they don't turn it off.
Run a loadcalc.net load calculation on each of those three bedrooms. If it doesn't add up to at least 15,000 BTU/hr (and I'm pretty sure it doesn't with the windows closed) find a different solution. At max speed the AOU24RLXFZH delivers 25,500 BTU/hr @ +5F, and a 15K load would already be way sub-optimal at a 1.7x oversize factor. If the combined load is closer to 9K (likely) you're in completely ridiculous oversizing territory and it will never come close to meeting it's HSPF or SEER numbers.
Cheap (and not too equipment-specific) rules of thumb on oversizing factors.
Keep it under 1.25 x for multi-splits...
..under 1.5 x for single zone mini-splits.
Staying within those bounds will still be reasonably efficient and comfortable. It's better to undersize a bit (even if it means you then need an electric blanket during cold snaps) than going much over those factors.
Ok. I did plan to listen to the podcast on the way home.
I would love to just put it on a smaller unit, but they don't make 3-zone multi-split's smaller than 24k that I have seen, and all the other ones I saw (LG RED, Mitsu) seemed 7k+ heating minimum just like this.
I assume by cassette you mean putting it in the hallway? The hallway is open to the first floor so I felt like it would not be that effective at heating the rooms.
Slim duct into the two adjacent rooms seems interesting but then I am back to the unconditioned attic issue.
A mini duct cassette can be mounted below ceiling level in a closet between a pair of rooms that it serves, or with longer soffited duct runs to rooms not directly adjacent. The ducts don't need to go up into the attic, just through the walls or transom above the closet door but the return paths also have to be figured out.
There are additional construction costs to installing ducted zones, but the heat distribution and load/capacity balance can be better optimized.
Without room by room load numbers you're shooting in the dark and unlikely to come up with an optimal solution. Do NOT let the HVAC contractor run those numbers (very few do it properly). Hire a P.E. or RESNET rater to do it- somebody who makes their living on the accuracy of their numbers, not installing equipment. You can also use online freebie tools like loadcalc or coolcalc (all of which will oversize compared to a pro tool used by a true pro, but not by 3x) to get a handle on it.
FYI the zoned mini split systems do not operate at as low a speed as one compressor one head systems will. That translates into less cycling on and off and higher efficiency. Plus when something breaks you only loose 1/3 of your heating, I think you will find the up charge is surprisingly small.