We live in a 75-year old stone colonial in the suburbs of Philly. The house is ~4,500 SF, including the finished 3rd floor but not including the finished basement. About 1/2 the SF is the original stone house, with 1-foot thick walls. The other half comes from 2 additions, which do not appear to be well insulated. We've already had an energy consultant come to do as much insulation as feasible with this construction. The windows in the old part of the house are all original and painted shut. About 2/3 of them have storm windows. The house is served by 3 HVAC systems (gas fired furnaces + A/C). An auxiliary hot water heater also provides hydronic radiant floor heating in 2 small floors + hot water heating in the basement. The HVAC systems do not function well -- the equipment isn't terribly efficient, the duct work and returns aren't sized correctly, and the configuration (what unit serves what spaces) doesn't work for us. 2 furnaces are 6 years old, 1 is 11 y.o., and the 3 a/c compressors are 16+ y.o. We want to replace the old windows and the HVAC systems but aren't sure what we should tackle first. Both are huge projects! Also, I can't seem to find any contractors who do Manual D and Manual J calcs. Any suggestions? Thanks in advance!

As for the Manual J calcs keep calling around, somebody in your area has to do them. I find about 1 in 10 of my competitors do a manual J. During the manual J they should also be able to show you the savings in BTU's from new windows vs using the windows that you currently have. If your furnaces are only 6-11 years old they probably arn't that bad, but windows usually have a low ROI(at least in my area).

I know you mentioned insulation but have you done air sealing? You said the windows were painted shut but how about air gaps between light fixtures, wall sockets etc that could be leaking a lot in a 4,500 sq foot home. This often has a quicker payback and doesn't force major renovations such as heating and cooling or windows until the systems start to reach the end of their useful life.

Also have you considered heat pumps instead of AC's since the furnaces are so new? I don't know your climate specifically or your gas or electric rates but these can often help by not having to replace the furnace prematurely and are a cost effective way to heat and cool a home.

If all of the heating equipment uses the same fuel it's possible to do the gross sizing of the equipment on a fuel-use basis, but that doesn't deal with the room-by-room needs.

The 1-foot thick stone walls are probably between R1 & R2 depending on if they're quarried with tight thin mortar (= closer to R2) or rounded field stone with deep inset to  wide mortar joints (=closer to R1), and with an interior plaster & lath might add up to R2.5, give or take.  A fairly tight antique window + triple-track storm is also about R2, and a code-min replacement window would only be about R3, and much more expensive U0.20 window is only R5- the heat load is still going to be dominated by the amount of square footage of exterior wall, and swapping out the windows isn't going to change the heat load by much.  So from bang/buck point of view going after the mechanicals will reap quicker rewards than anything you do with the windows. 

But DO at least treat the single panes that don't have storms.

If the windows that don't have storms are in serviceable shape, a low-E storm window would beat the performance of barely-legal replacement window, at a fraction of the installed cost.  Whether the original windows are worth rebuilding or saving depends- antique windows are often rebuildable to a reasonable tightness without huge expense, and with tight low-E storms performance is pretty decent.  But doing anything beyond low-E storms for the windows without storms comes behind mechanical systems, in your case.

Short of a Manual-J it's possible to come up with reasonable room by room heat load numbers based on the amount of exterior wall using I=B=R methods.

If you assume R2 for the wall + window average for the old section .  The U-factor (=1/R) of R2 is 0.5, which is in units of BTUs per hour per square foot per degree F.  Use +15F (the 99% outside design temp for Philly) as a heating design temp, and 70F as an interior temp.  So for a room with a 12 x 9 exterior wall + window and an interior temp of 70F, at 15F there's a 55F difference, and with  ~110 sq. ft x 55F x U0.5 you get a heat loss of about 3000BTU/hr.

For the other not-too insulated but some walls, figure on a U-factor of 0.1 to 0.15 for the wall. Subtract out the window area for the calc, then add in the window losses. Most double hungs run about 10 square feet, but some antiques are much larger (or smaller), so measure them.  Call it U0.8 if single-pane, U0.5 with storms for calculation purposes.

For a roof/attic with 6-8" of batts or blown use U0.7 

This goes surprising quick if you use a spreadsheet tool, a tape measure and a laptop/notebook and the loads are usually comparable to Manual-J methods with some fudge-factors. What's missing is the air-leakage factors- if a room or the whole house is pretty drafty, either multiply the heat loads by 1.25 or do more air-sealing.

Leaky badly-designed ducts in a leaky house aren't worth re-using, and it may be easier to retrofit a single hydronic boiler as the heating source for all of it.  Ducted air systems inevitably create pressure differences between rooms, and the "return" path is often via the great-outdoors, increasing the heat load while creating drafts.  By going to an all-hydronic solution air-handler driven infiltration losses go away.  The heat emitters in each room can vary from really cheap (fin tube baseboard) to wildly expensive (radiant floors & ceilings), but in-between the european style low-temp convecting panel radiators offer a lot of comfort over baseboard.  You may have to break it up into zones to get reasonable heat balances, but resist the urge to micro-zone it since that would add a lot of complexity & expense to the system for minimal gains in comfort, and you'd probably have to take design measures to keep the boiler from short-cycling itself into an early grave.

Be very careful not to oversize the boiler by more than 25% from a fuel-use based "whole house" heat load estimate even if the I=B=R room by room numbers add up to more, since that increases the short-cycling problem and reduces the as-used efficiency. With a leaky ducted system the actual heat load will be LESS than implied by the furnace efficiency & fuel use, and undersizing the boiler by 10% or more would still have margin on the real heat loads.

To calculate from fuel use the easy math goes like this.

Assuming the furnaces are mid-efficiency single stage and the water heater isn't a condensing version, the average steady state combustion efficiency on all of them is something like 80%, but if the name-plate AFUE on either furnace says 83%, use 83%- we're looking for an upper bound, not high precision. If the gas billing has an average temperature and average use per day on it, that makes the calc easier, otherwise take a mid-winter's bill and note the meter reading days, and download a spreadsheet of base-65 heating degree-days from a weather station near you on degreedays.net that covers the span of the billing period.  Use the spreadsheet to add up the degree-days,  and then run the calculation:

Say it adds up to 978 degree-days for  a January-into-February period, and say you used 517 therms (or ccf)  for that mid-winter period.

That's 517/978= 0.53 therms/HDD

At 100,000BTU/therm thats 53,000 source-fuel-BTUs/HDD

But at 83% efficiency only 83% of that or 0.83 x 53000 = ~44000 BTU/HDD ended up in the house

There are 24 hours in a day, so that's 44,000/24= 1830 BTU per degree hour needed from the heating system.

At a base-temp of 65F (the approximate balance point between heating & cooling for most homes) and a design temp of 15F, that's 65F-15F= 50 heating degrees.

So, the heat load at design temp is about (and usually less than, almost never more than) 50F x 1830= 91,500 BTU/hr.

Your actual numbers may come in a bit higher or a bit lower- these were pulled out of a hat, but probably aren't more than 50% away for a reasonably tight but low-insulation house that size (about 20BTU/square foot). 

Be aware that there are old-schoolers out there who will look at that house and say something like "Lessee, 4500 feet times 35 BTU a foot comz ta 'bout 160K, call it 175K  just to be sure".  While that won't leave you cold, it's GUARANTEED to be oversized, usually by more than 1.5x often by more than 2x. The upside is that you'd be covered all the way down to -50F (a temperature not seen in Philly since the last ice age)! :-)  The down side is that you end up paying for 2x a much radiator and 2x as much boiler for the system.

The heat load simply can't exceed the source-fuel BTUs that kept the place up to temp, so use your billing info to size the boiler, and an I=B=R approach as a sanity check, and for sizing the radiation per room.  You don't have to design the whole system, but if you know the range to expect, and you'd be able to discuss it with some intelligence & foreknowledge with the system designer, and be able to point out when things don't really add up.

While SkyHeating's heat pump option may work, but any time I see:

"The house is served by 3 HVAC systems"

...with yet other zones for aux-radiant I personally want to just start over, preferably with a hydronic heating (if gas-fired), especially in an antique house that is difficult to really air-seal.

Re-working the ducts of a triple system kludge into a system that actually makes sense and works efficiently seems like a very remote possibility, and the total burner & compressor sizings are probably insanely oversized for the actual loads. Retrofitting hydronic heating with 3 large zones with a cast-iron boiler & baseboard (or even a condensing boiler) is likely to be cheaper than any ducted heating solution, with room left over for a more intelligent cooling approach.

The whole house heat load is probably less than the output of the largest of the three gas furnaces, and the cooling should be dealt with separately, since the cooling BTUs are likely to be very different from the heating BTUs. The price of natural gas has come down a lot with the PA & NY shale gas development, electricity not-so much. Only with the HIGHEST efficiency heat pump systems and near-perfect ducts would heating with a heat pump cost less than heating with gas. But I suppose it's worth soliciting proposals.

I'd be surprised if the layout was optimal for both heating & cooling with ductless air source heat pumps, but sometimes in big old stone houses the thermal mass allows you to get away with cooling primarily the upper stories with ductless, and letting convection take care of the first floor. Getting the cooling loads nailed down with any precision is tougher do do, since solar gain & shading factors can make or break the room-by-room loads. It wouldn't surprise me if there were more than 10 tons of compressor on the existing 3 systems for a peak cooling load that is probably no more than 6, but it WOULD surprise me if any one of the systems was reasonably balanced for cooling, given the configuration complaint described.

A zoned 2 stage system with a right-sized single compressor & high-velocity skinny-ducts might work out pretty well as a retrofit, but I wouldn't even try to use the same ducts for heating & cooling in a barely-insulated house- that's almost always going to fail the comfort test on either the heating or cooling end.

SkyHeating – Thanks for your response and suggestions. So for the original question – replace windows vs HVAC, it sounds like NOT windows for sure. Good question about the air sealing. I had to revisit my notes to see what we did. We mainly insulated and air sealed in the attic. The only air sealing we did on lower levels was seal the gaps where pipes, etc. penetrated the exterior wall. When air sealing wall sockets and light fixtures, I assume this is only relevant on the newer exterior walls? The old walls are stone, lathe, and plaster.

We’ve never considered a heat pump – please tell me more. How would we incorporate it into our system? The heat pump replaces the a/c and makes use of the furnace air handler? I’ve heard the heat pump isn’t so great when the temp goes below 40 degrees. Since it definitely goes below 40 in Philly (though not last winter!) that’s not enough heat. Can the existing furnace serve as a backup? Is this what we call a dual fuel system? Does a heat pump cool as well as conventional a/c in extreme heat? Is the primary benefit of a heat pump its efficiency, and therefore operating cost? I’m embarrassed to say I don’t know if electricity or gas is more expensive right now in Philly, which I have read is an important factor in determining whether a heat pump is a good choice. Sorry for all the questions, and thanks for your help.

Dana1 – Wow. Thanks for the detailed answers! Got it on the windows – put storms on the stormless windows (after we get them open) and call it a day.

You’ve offered lots to think about with the HVAC systems. And you hit the nail on the head with several of the assumptions. If we were to replace the furnace and a/c with brand new top-of-the-line equipment, we would still need to zone the sprawling first floor, increase the return on the master bedroom system, take a small bedroom off the master bedroom’s system and put it on the 2nd/3rd floor system, and possibly increase the size of the supply ducts on that same system. And zone the 3rd floor because 98% of the time no one is there. I also wish there were returns in each bedroom because they get stuffy at night when the door is closed or cracked open.

The hydronic heating idea is intriguing…we have those fancy European radiators in the basement and they are quiet and comfortable. If one boiler could serve the whole house, that would be handy since we have to replace the auxiliary hot water heater within a year or two anyways. But do people install new hot water heating systems in houses these days? I always assumed people had hydronic heating because that’s what was in the house when they bought it and they were “stuck with it” (sorry!). Also, did I understand correctly that you think it would be less expensive to run hot water lines to every room than it would be to retrofit the ducts? Very interesting. What would you suggest as a good cooling approach? Unfortunately we already know cooling the upper floors and letting the air waft down doesn’t work – the 1st floor a/c broke last week and it got hot. Whatever the solution (e.g., heat pumps) the problem is we would still need to fix up and use those ducts…right? Oh, just remembered your high velocity duct suggestion. So you are thinking new high velocity ductwork for a/c paired with hydronic heating? Or new forced air heating and cooling throughout using new high velocity ductwork? Thanks again for your insight!

Posted By avalanche85 on 14 Sep 2012 02:31 AM
  I’ve heard the heat pump isn’t so great when the temp goes below 40 degrees. Since it definitely goes below 40 in Philly (though not last winter!) that’s not enough heat. Can the existing furnace serve as a backup? Is this what we call a dual fuel system? Does a heat pump cool as well as conventional a/c in extreme heat? Is the primary benefit of a heat pump its efficiency, and therefore operating cost? I’m embarrassed to say I don’t know if electricity or gas is more expensive right now in Philly, which I have read is an important factor in determining whether a heat pump is a good choice. Sorry for all the questions, and thanks for your help.

Modern day heat pumps (15 or 17 SEER) are good until around 20 degrees or so. If additional heating is needed one can install a heating strip in the air handler and have it kick on when needed.

RHEEM HEAT PUMP INFO

Posted By avalanche85 on 14 Sep 2012 02:39 AM
Dana1 – Wow. Thanks for the detailed answers! Got it on the windows – put storms on the stormless windows (after we get them open) and call it a day.

You’ve offered lots to think about with the HVAC systems. And you hit the nail on the head with several of the assumptions. If we were to replace the furnace and a/c with brand new top-of-the-line equipment, we would still need to zone the sprawling first floor, increase the return on the master bedroom system, take a small bedroom off the master bedroom’s system and put it on the 2nd/3rd floor system, and possibly increase the size of the supply ducts on that same system. And zone the 3rd floor because 98% of the time no one is there. I also wish there were returns in each bedroom because they get stuffy at night when the door is closed or cracked open.

The hydronic heating idea is intriguing…we have those fancy European radiators in the basement and they are quiet and comfortable. If one boiler could serve the whole house, that would be handy since we have to replace the auxiliary hot water heater within a year or two anyways. But do people install new hot water heating systems in houses these days? I always assumed people had hydronic heating because that’s what was in the house when they bought it and they were “stuck with it” (sorry!). Also, did I understand correctly that you think it would be less expensive to run hot water lines to every room than it would be to retrofit the ducts? Very interesting. What would you suggest as a good cooling approach? Unfortunately we already know cooling the upper floors and letting the air waft down doesn’t work – the 1st floor a/c broke last week and it got hot. Whatever the solution (e.g., heat pumps) the problem is we would still need to fix up and use those ducts…right? Oh, just remembered your high velocity duct suggestion. So you are thinking new high velocity ductwork for a/c paired with hydronic heating? Or new forced air heating and cooling throughout using new high velocity ductwork? Thanks again for your insight!

Hydronic heating is DA BOM from a comfort point of view. If you're willing to spend a bit of money on comfort and use low-temp panel radiators rather than fin-tube (or a LOT money on radiant floors & ceilings) it's even better! In New England hydronic heating with fin-tube baseboard is still pretty standard on tract housing.  When people demand central AC but don't want to spend money on separate systems is when it becomes ducted hot-air HVAC systems, and comfort suffers (more so than efficiency) when they do. 

Fin-tube isn't nearly as comfortable as heat emitters that put out a high fraction of their heat via radiation, but it beats the draft-factor of ducted-air.  (The very best and right-sized continuously-variable speed air handler systems can be pretty comfortable though.) Most of the time the HVAC combo units end up being oversized gas furnaces with oversized AC, or sometimes close to right sized for one but not the other.

Retrofitting the plumbing for hydronic heat pretty easy and doesn't need much space.  Expanding undersized ducts and adding returns can be cumbersome in a retrofit. With hydronic heating it's the radiation adds some expense, and boilers (particularly condensing boilers) tend to be a $1-2K cost-adder over condensing furnaces. But the as-used efficiency tends to be better than ducted-air.

In an antique house with low-R walls it's hard to deliver the best balance of heating AND cooling with the same ducts, since the upper floors will generally need more cooling, less heating that the lower floors, etc, yet the ducts deliver the same volumes to the zones in both modes.  If the ducts are designed just for cooling you'll get more satisfactory cooling balance, and you can design the hydronic heating zones for optimal heating balance.  Rare is the case where heating & cooling with the same ducts is perfect, though in a code-min new house it can be OK, but in a R2-wall antique it's damn-near impossible.  Also in a house that's hard to air seal you generally get about a 15% improvement in heating fuel use with hydronic rather than ducts, even with a decent duct design.  But if you can tighten the place up to 2-3 air exchanges per hour at 50 pascals pressure (ACH/50) in a blower-door test and the ducts & air handler are mastic-sealed or taped at every seam & joint those savings with hydronic systems are low "in the noise".

If you combine the cooling system into one set of ducts (even if it's zoned with valves) it might be easier to retrofit with the skinny high-velocity ducts, or not- you know more what your house looks like- you may be able to do it with the same registers, but an all-new (or mostly-new) duct design that meets relevant standards.   But again, COOLING ONLY, heating with the same ducts in THIS house will probably mean you have to make it sub-optimal for either heating or cooling (or both).

At Philly gas & electricity rates a heat pump solution would have to have an average COP of 3 to 3.5 (HSPF 10 - 12) to break even with a condensing gas furnace, and even higher measured against a condensing hydronic system, (due to the aforementioned duct losses and air-handler driven infiltration issues.)  While ductless heat pumps (mini-split/multi-split) can hit those efficiency ranges in your climate, it would be crazy-expensive at what I anticipate your heat loads to be.  There are some ducted heat pumps that might approach that in an ideal house, but that's without factoring in infiltration losses to the system.  One of several efficiency advantages to ductless is that it recirculates only the air in the room/zone where the interior head is installed, and doesn't generate pressure differences between rooms to drive infiltration.  Multi-split ductless systems run about $2.8-3K/ton, and you might have 8-10 tons heating, but only 5-6 tons cooling at your R-values, and the rooms without heads might have serious comfort issues at the winter extremes.  A 120KBTU/hr hydronic boiler (that would probably handle the load) doesn't cost much more than a 60KBTU/hr boiler (that probably wouldn't) and if you're willing to sacrifice a bit of efficiency at the winter extremes (from higher water temps) the radiation required doesn't cost much more either. It's the same amount of plumbing either way.  Even the biggest ductless systems may not handle your heating load, but would probably come close to handling the cooling load. 

A right-sized 2-stage intelligently-ducted AC system would probably come in at half the cost of a ductless solution, and be almost as efficient in the cooling season.  (I'm a big advocate ductless for small-load high-R houses though.)

Got any wintertime gas bills to do a heat load upper bound on?  It's an important number to have in your pocket when dealing with contractors.  Doing Manual-Js on uninsulated stone has big error bars, and the tendency for the contractor is to err to the high side (to avoid the 5 AM call on the coldest night of the year.)  Fuel use against heating degree-days is a MEASUREMENT, and the biggest error factor is the actual vs. nameplate efficiency of the heating system.  The fuel use derived number for ducted air in a leaky house is always higher than reality given the infiltration losses, but even the best-performed Manual-J will often exceed that number.  When in doubt, use the measurement, not the estimate, since the measurement A: has fewer error factors and B: is biased to the high side already.   Upsizing by more than 10% from a fuel-use measured heat load simply isn't warranted, particularly if going from ducted air to hydronic.

"Modern day heat pumps (15 or 17 SEER) are good until around 20 degrees or so. If additional heating is needed one can install a heating strip in the air handler and have it kick on when needed. "RHEEM HEAT PUMP INFO

The newer Misubishi minisplit heat pumps provide heat well below zero.

Almost all mini-splits are good down to +15F (Philadelphia's 99% outside design temp), but it isn't likely to be a very good fit to the application. (Many Fujitsu & Daikin models have an output rating at -20C/-4F.)

On storm window upgrades for the single-panes it's well worth seeking out low-E rather than clear-glass triple-tracks. The ROI will be quicker for the higher-performing low-E storms. (Larsen sells through a number of dealers in the Philly area http://www.larsondoors.com/dealers/ But many manufacturers have low-E options now) see:

http://www.ornl.gov/sci/buildings/2010/Session%20PDFs/22_New.pdf

http://www.low-estormwindows.com/resources/

Lbear and Bob I - Thanks for the info about heat pumps!

Dana1 – Thanks again for the wealth of advice. I’ve been talking to contractors, and it’s not looking good. A plumbing/heating contractor said installing hot water heating throughout the house would be cost prohibitive. Even though he’s a big champion of hydronic heat, he definitely did not recommend it in our house.

Another HVAC contractor walked through our house and agreed that we had a complicated mish mash of equipment and problems. He didn’t think it made sense to replace 6-year old furnaces, which he identified as 90% efficient and of low-ish quality, but agreed it restricts what I can do when I replace my a/c units(s) in the spring. I’ll be talking to another 2 contractors this week, but I’m guessing I’ll be talking about ways to tweak and make-do with what I have rather than rehauling my whole system by the end of the week.

Big picture issues aside, my most pressing issue is my daughter’s room, which drops to the low 50’s in the winter even with the heat set on 70. It is on the same system as the master bedroom, which is partially insulated, partially thick stone walls, and partially heated via hydronic radiant floor heating. As you can imagine, the heat rarely comes on, which is not good for my daughter’s uninsulated, non-stone wall room. The airflow, when it does come on, is practically non-existent because the one return is too small. I thought we could take her room off that system and hook it up to the other 2nd floor system, but the room’s location will make it very tricky. The contractors could open up all the walls and still decide that it can’t be done.

So my new question is, “How can I get auxiliary heat to this room without creating a fire hazard?” I found out a mini-split is out of the question because there is nowhere to put the line and the outdoor unit. Here’s my new far-fetched idea. There is an abandoned in-wall radiator in the attached bathroom (from back in the days of oil heating). Maybe I can re-use the pipes, put in a new radiator, and tap it into the auxiliary hot water heater that does the random floors and basement? I am desperate!

Putting in a radiator to micro-zone in the colder room would work, and it probably doesn't take a lot of radiator to get there either.

Do the I=B=R type heat load calc on the room, and describe the size of the wall radiator. If the pipes were part of a steam heating (as opposed to pumped hot water) it could have enough crud in side that cold be a real PITA if put to pumped water use, with bits of rust & iron to filter out to keep it from chewing up pumps & clogging valves/heat-exchangers, etc. If it was part of a pumped water system, it should be fine (but filter it anyway). A low-temp euro-style radiator could get you there without much risk if the pipes are clean enough.

You have to be a bit careful when using hot water heaters for aux heat, since the low return water temps and turbulence from pumping can cause corrosive condensation in the heat exchangers of non-condensing water heaters.

You might want to poke around and determine whether the partition walls (particularly on upper floors) all have top-plates to the studs, since many homes of that era did not.  If they're open to the attic at the top they're essentially exhaust flues expelling heat in the winter, driving stack effect infiltration elsewhere.

Another thing to consider would be electric radiant cove-heaters, which are mounted up at the crown molding. It's more expensive to operate, but much cheaper to install than plumbing in a radiator, but as aux heat the power use shouldn't be much. Alternatively oil-filled electric baseboard has very low fire-risk too.

Dana1 - Thanks again for the advice. So I see 2 potential solutions, neither perfectly easy: 1) Add an electric radiant cove heater. The ceiling is sloped in half the room, but there are a few stretches of interior wall that might work, if the cove heater can go on an interior wall. Do you know if it requires a dedicated circuit? Unfortunately my electrical panels are completely filled. And unlabeled. I might need to add another subpanel. 2) Install a new radiator in the room - from what you and another contractor are saying about the hot water heaters, probably my best bet would be to replace the two 14-year old hot water heaters now with a single boiler to serve domestic hot water & various heating needs. I have one more contractor to talk to tomorrow. I'll run these ideas by him. Many thanks!

Radiant coves come in a wide variety of wattages and voltages. (eg: http://www.globalindustrial.com/g/hvac/heaters/cove/radiant-cove-heaters ) Calculate the heat load for the room, and know that you get about 3400BTU/hr out of 1000W of heater. The fact that you're getting at least some heat out of the ductwork you can under-size the output of the cove heater by quite a bit (probably by half) relative to the total heat load of the room. But unless it's under 500W (1600BTU/hr) I'd recommend putting it on it's own circuit.

You can use set-back line-voltage thermostats to keep power use to a minimum, since much of the time the heating system probably does keep up, or nearly so. During the recovery from set-back you get a nice warm "sun on your face" feel out of them, without the focused intensity fire-hazard of a plug-in bare-element space heater, and it warms the people & stuff in the room before the air temp comes up- it's just more comfortable.

Sorry for the long delay -- I've been talking to HVAC contractors and plumbers. Just to close the loop here, I wanted to note that unfortunately, given our large, mish-mashy, old house construction, no one thought it would be wise to re-do ductwork or run new water pipes anywhere. Pretty much the prevailing attitude was: you are stuck with what you have, and all you can do is tinker on the sides...maybe add a few returns, swap out the HVAC systems, add a mini split here or there.

The one contractor I liked the most suggested removing a 13-year old system out of the ridiculously tight attic crawl space and installing a Carrier Greenspeed Heat Pump in a former laundry closet of the 2nd floor, and adding several returns to get air flowing into my daughter's room (including one return in her room). In this scenario, the duct work would be cleaned up a bit because of the relocation. At $20K, it seems like an expensive way to get conditioned air into a little room, but it does solve some other problems, like the inaccessibility of that HVAC unit. So I'm wondering if the pricey Carrier Greenspeed is worth the money and will this huge effort really result in a more comfortable room for my daughter?

I don't know if the Carrier is expensive or the duct alterations are. Have him quote you with less fancy options.

My impression was both were expensive. I think I will ask how much the price would go down if I didn't go for the tippy top of the line, but one step down. Would folks recommend one brand over another for reliability? I am partial to Trane products, but I don't actually know how good its heat pumps are.

Dana1--I'm keeping the radiant cove heating idea in my back pocket, in case my daughter's room need supplemental heat after all this is said and done.