Hey all,

I have a question for those who are experienced in radiant heat. I'm looking at an energy retrofit situation for my small house in climate zone 5. Bathroom and kitchen remodel will allow me to put radiant heating in the new floors, but in the rest of the (small) house the existing floors will remain. Since I'm not planning to have ductwork, I would like to explore options for baseboard or panels to heat in the rest of the house.

This is a SMALL house. It's super-insulated, and super-tight. I've got my ach less than 2 and my insulation is R-60 attic (unconditioned), R-30 walls, R-10 foundation. It's less than a thousand square feet.

Is it optimal for a heating load as small as mine to use a condensing boiler for the zones (I'm thinking there will be 5 or 6, including kitchen and bathroom floor) or off of a condensing gas water heater, such as the vertex? I've been looking for attractive baseboard or panel units, but I don't know much about them because I just got started.

For my ac, I plan to use a few minisplits. The floor plan is very open, so I think it will be quite possible to use one split serving more than one room since it's a small, open area.

Thank you.

if you have a floorplan (open, limited number of rooms) that allows inexpensive minisplit install and you're superinsulated, then going for hydronics/radiant is probably unnecessary. Just use the minisplits to heat as well as provide AC. it's not quite as ideal but it is less of a detriment to comfort in that kind of a situation.

in superinsulated homes, hydronics are great for floorplans that are not conducive to single point heat sources... multiple bedrooms, hallways, etc.

If you do go hydronic, a water heater with heat exchanger and panel rads/radiant ceiling is probably the most cost effective route to comfort. You can upgrade to a condensing water heater but it's PROBABLY not cost effective to do so on a very small heat load.

Posted By NRT.Rob on 19 Sep 2012 09:11 AM
if you have a floorplan (open, limited number of rooms) that allows inexpensive minisplit install and you're superinsulated, then going for hydronics/radiant is probably unnecessary. Just use the minisplits to heat as well as provide AC. it's not quite as ideal but it is less of a detriment to comfort in that kind of a situation.

in superinsulated homes, hydronics are great for floorplans that are not conducive to single point heat sources... multiple bedrooms, hallways, etc.

If you do go hydronic, a water heater with heat exchanger and panel rads/radiant ceiling is probably the most cost effective route to comfort. You can upgrade to a condensing water heater but it's PROBABLY not cost effective to do so on a very small heat load.

Do you think the minisplits would be able to keep up? Here where I live, it occasionally will get down to -10F. Will that call for seperate electric resistance backup heat then? Or is there electric resistance backup built into the minisplits?

I am open to whatever kind of water heater/combination system is ideal because I will be upgrading my water heater at the same time. I have been looking at the vertex and a couple of on demand combi units.

upgrade to a condensing water heater

Do condensing water heaters/boilers suffer from decreased life due to increased corrosion of the heat exchanger?

Do you think the minisplits would be able to keep up? Here where I live, it occasionally will get down to -10F

I wouldn't bother with the hydronic radiant in a small super-insulated place. You can use electric resistance on timers for the areas you want to be barefoot friendly and a minisplit for the rest of the heating/cooling needs. The modulating function of the minisplits allow you to increase the overall system capacity without too much of an efficiency hit. Take the money saved and put it in some PV panels.

You might want to spend the money on the better minisplits for low temp operation, but in general they are rated for quite low temperatures and finding out if they will keep up should be fairly straightforward.

My thinking was that I would use the minisplits for cooling and heating when the temp is moderate then switch to hydronic heating when it gets really cold (usually from late novemeber to early march). I have natural gas where I am and I will be upgrading to a natural gas water heater, so I figured I'd take care of both of them at once.

minisplits can run easily down to zero or below with COPs that, depending on where you are, should keep competitive with natural gas. Where are you and what are your rates for electricity and natural gas?

Posted By strategery on 19 Sep 2012 10:02 AM

Posted By NRT.Rob on 19 Sep 2012 09:11 AM
if you have a floorplan (open, limited number of rooms) that allows inexpensive minisplit install and you're superinsulated, then going for hydronics/radiant is probably unnecessary. Just use the minisplits to heat as well as provide AC. it's not quite as ideal but it is less of a detriment to comfort in that kind of a situation.

in superinsulated homes, hydronics are great for floorplans that are not conducive to single point heat sources... multiple bedrooms, hallways, etc.

If you do go hydronic, a water heater with heat exchanger and panel rads/radiant ceiling is probably the most cost effective route to comfort. You can upgrade to a condensing water heater but it's PROBABLY not cost effective to do so on a very small heat load.

Do you think the minisplits would be able to keep up? Here where I live, it occasionally will get down to -10F. Will that call for seperate electric resistance backup heat then? Or is there electric resistance backup built into the minisplits?

I am open to whatever kind of water heater/combination system is ideal because I will be upgrading my water heater at the same time. I have been looking at the vertex and a couple of on demand combi units.

The Mitsubishi H2i series output at -25C/-13F is rated to deliver slightly over 70% of the nominal heating rating. (The nominal rating is good down to +5F.)   It's pretty simple math to see if it will keep up at -10F.   Daikin & Fujitsu don't spec an ouput at -10F, but offer -20C/-4F specs for many models- you can presume the derating is at least somewhat similar to the Mitsubishi units. The others don't shut down at low temp, they just don't guarantee or specify the output at temps any lower than -20C, whereas Mitsubishi gives a spec for -25C/-13F.  (The MUZ-FE18NA 1.5 ton has a ~20KBTU/hr nominal heating output, and delivers ~14KBTU/hr @ -13F.)

Many of the locations in eastern ID monitored in this document regularly see negative double-digits, but they weren't really looking at output at those temps, mostly average seasonal efficiency, etc.

SFAIK there are no mini-split vendors that use resistance heating for anything other than standby-heat on the compressor to keep the mechancal component dimensions within operating tolerances on startup. (This can be a ~100W order of magnitude load at low outdoor temps.)

In superinsulated homes floor plans are pretty irrelevant for mini-split use, even less so for SMALL superinsulated homes- you can use the ERV/HRV  design to provide significant temperature balancing.  The room with the mini-split head gets supply-only ducting, and the room(s) without get exhaust only, drawing their ventilation air via jump ducts.  This is commonly done in Net Zero and PassiveHouse homes in New England.  One ductless head per floor seems to be pretty common, primarily to deal with cooling load differences between floors.

On a 3-family deep energy retrofit I've been advising on locally (middle of US climate zone 5, 99% outside design temp of +5F, but it hits negative double digits here at least once every 25 years, and negative single digits at least every 5)  they went with one mini-split head per unit, with temperature balancing to the bedrooms & bath via jump ducts and exhaust-only ERV ducting in the doored off rooms, supply-only ERV in the more open kitchen/dining/living room areas.  The doored-off bedrooms  have ~10-12 square feet ~U0.20 window each, with ~R40 (whole-wall) walls. This place is tighter  (under 600cfm/50 for the whole building) and better-insulated than many NetZero homes in climate zone 5 that utilize the same balancing strategy with ERVs & mini-splits.

"Do condensing water heaters/boilers suffer from decreased life due to increased corrosion of the heat exchanger?"

No, they are made to condense and feature sealed combustion burners; a must for tight houses. For the money, a condensing storage type water heater does double duty with a plate heat exchanger and outdoor reset. If you have natural gas, electricity makes little sense anywhere as NG is likely to remain plentiful and affordable for decades. Electricity, not so much. If you are super-insulating with good windows, the matter is mute but one must consider the output of light fixtures.

Since the ROI of super-insulating, GSHP, PV and hydronic solar panels with such small loads, is in the decades, fuel sources become more academic. My mini-split is great but the radiant floors rule with R-30 walls and R-50 hot roof.

For small loads systems costs also factor into the upfront cost, and you pay a premium for separate heating & cooling systems.  If I'm reading it correctly the R30 walls/R60 attic and <2ACH/50 are already bought & paid for. 

The heat load is almost certainly low enough to be handled with a 1 or 1.5 ton ductless, just as it is for Marc Rosenbaum in a similarly insulated similar sized house in zone 5, but Marc's house DOESN'T benefit from an open floor plan.

New England has some of the highest electricity prices in the lower 48, but heating with a better-class ductless is pretty much a wash against condensing natural gas here.   Average COP on ductless is closing in on 3.0 in a zone-5 climate (see the data for Eastern Idaho), and with 1.25x oversizing would likely hit 3,  but even if one assumes a COP of only 2.5 it's still comparable on operating costs with a condensing gas water heater based radiant  floor system at current buck-a-therm-delivered MA residential retail gas pricing and ~15cent electricity.  Going for a mod-con for the extra few percent efficiency would pay off approximately never.

Electricity pricing has been cut by lower natural gas prices too, but is also being cut by utility scale wind, which has a comparable levelized cost of energy to natural gas in thermal plants without the price volatility of gas-fired plants.  Both gas & wind are on a big upswing due to the lower operating cost. If gas prices rise back to $4/mmbtu wind at a 30% capacity factor will be even cheaper than combined-cycle gas plants.  I see the future pricing of electricity and natural gas becoming even MORE coupled over the next decade or so, not less, as gas takes a larger share of the grid baseload.   Residential retail pricing of gas in New England may fall even further if the proposed NY southern-tier gas pipeline project gets built, but regional gas-fired electricity will fall along with it.

Bottom line, it's probably not worth the cost-adder of building in radiant on the modifications of this house as a hedge against electricity price increases if a ductless fills the bill (and I think it can.)

"Do condensing water heaters/boilers suffer from decreased life due to increased corrosion of the heat exchanger?"

After some research, I'll say "it depends". Condensation creates acid on a heat exchanger. This would cause lots of corrosion, but at least the condensing portion of the heat exchanger should be made of stainless steel, which is corrosion resistant. How resistant depends on the grade of stainless steel, the thickness, etc. So condensing or non-condensing, when it fails depends on the quality of the unit.

Wind is not making for cheaper electricity anywhere I know of. In fact the local utilities have asked to take off the mandate in order lower their burden, along with the solar tax, sorry, initiative with comparable results. Just not here yet.

I would have no problem with your numbers but the comfort level is not the same, living with both here in Minnesota. I do not have triple pane or R-60 and admit to some thermal bridging, but still rely on the radiant floors even now with 46°F nights.

We can talk ROI or comfort, but rarely both at once. Zoned heating of bathrooms with a low-voltage matt is a good idea...but I don't like the future of electricity given the rabid hatred of all things coal and they are not cheap. The use of NG to make electricity will put pressure on this commodity like any other, but we have our friends up north to make more :-).

The marginal cost of wind is insanely low, making wind operators a $0 bidder in the day-ahead market. Grid operators already invested in fossil plant base generators have issues with increasing fraction of wind to the point that cuts into the capacity factor on their base generating capacity, but that's related to the business model that went into the financing the fossil plant investment, not the cost of wind.

Zero bidders (whether nukes or wind) drive the average cost of electricity down, since they get paid whatever next-highest bidder gets for their grid capacity. As wind hits about 20% or more of all kwh generated (as it has in IA and parts of MN), it's putting low capacity-factor peak generators back on their heels pretty hard when it's windy during peak hours, and it's starting to affect capacity factor on base generators too. With a different business model for the grid operator it doesn't have to stop at 20%, and in the more deregulated markets it won't. The federal subsidy may make the difference at the margins, but it only affects the speed with which the resource gets built. The fundamental costs are within a competitive economic range now, where they weren't 25 years ago, and the cost trend still shows a real downward slope with time.

The levelized cost of wind power (that includes financing operating and maintenance over it's lifecycle, unsubsidized) at a ~30% capacity factor (typical of TX or upper midwestern output- some are getting better than 30%) currently installed in the US is between $60-80/MWH (6-8 cents/kwh) and that cost has fallen dramatically over the past 5 years as a result with bigger-cheaper turbines and better control technology. For new projects using latest-greatest larger scale technology it's now down to the 3-6 cent/kwh range, all costs levelized, with no carbon taxes or production credits to monkey with the numbers.

See: http://eetd.lbl.gov/EA/EMP/reports/wind-energy-costs-2-2012.pdf (That link seems to be working only intermittently today...)

That is cheaper than the levelized cost for any new gas or coal fired plants running at ~35% thermal efficiency even at high capacity factors, but more expensive than combined cycle gas at 50%+ efficiency at current fuel prices. But the marginal cost of power for gas plants is primarily a function of fuel cost- they can't be zero-bidders at any fuel price, and the price of the fuel carries a volatility risk to be paid for. Power companies regularly make long term contracts for gas that are HIGHER than the spot-market lows as a hedge against that risk.

Projections for 2015 using fairly conservative assumptions are that the LCOE of onshore wind will be roughly comparable to but slightly more expensive than combined cycle gas, with all subsidies or potential carbon taxes removed.

http://www.ucsusa.org/assets/documents/clean_energy/Appendix-Key-Assumptions-Levelized-Costs.pdf (this one does work consistently)

In places with cheap existing large scale hydro the locals costs are much lower than new-wind or new-gas, but there's not much else out there that's cheaper. Even without carbon taxes the clean-up costs for getting the sulfur & other crud out of coal emissions would make it more expensive than new-wind or new-gas generation. But even in raw $/MMBTU numbers the current price coal is higher than it is for gas too, and that's not likely to change given the pace of gas drilling in shale formations. Love it or hate it or plain don't give a shi..nola, coal is just flat-out more expensive than natural gas. And there are more reasons to hate coal more than love it- it's dirty stuff with has bigger environmental impact than gas starting from where you dig it up to the acid emissions & ash of burning it. And at the comparatively low thermal efficiency its a real dog. You can't blame anti-coal activism & regulation for the slide in the coal industry if the economics just aren't there against gas even without the environmental regulations. Cheaper/comparable cost fuel and higher powerplant efficiency for gas is what's killing king-coal, not rabid coal-haters. (New-wind is also cheaper than new-coal by a good margin.)

In New England almost all new generation built in the past decade has been combined cycle gas, and the price of electricity rises & falls with natural gas prices, and I don't expect that to change much in the next decade even with the development of (currently even more expensive) offshore wind. Already NG accounts for more than a third of the power generated in MA (and growing), with another quarter or so coming from base-generator nukes. If the pipeline to the Marcellus shale gets built the cost of gas to local power plants will come down (currently a good fraction of regional gas comes in on LNG tankers at a much higher price) and the price of both gas and electricity should drop a bit. The MA renewable portfolio standard mandate for 2020 is 22%, but the state is on track to meet that target. The price of electricity has fallen (to a hair under 15 cents/kwh, down from 18-20 cents/kwh) since the RPS was engraved in granite, primarily due to the drop in natural gas pricing. If the price of gas ends up doubling by 2020 the renewable fraction of the grid-source will be substantially cheaper. Coal is all but done here, mostly due to local air-pollution regulation reasons. With no fuel cost-incentives to keep it alive I don't expect a return no matter how future Federal policy plays out.

Wind isn't the answer to all grid issues, but cost isn't the reason. I liken wind development somewhat to the federal push for hydro between 1935-1960 with the BPA & TVA et al- at a scale big enough it's cost-competitive. But that didn't work out nearly as well with nukes, which made even bigger promises and had less to show for the ROI in the end. (It went from "too cheap to meter" to "too expensive to matter" in only a half-century.) Yeah, I know, fusion power is only 20 years away, just as it has been for the last 60 years, and it's all the enviro-lobby and chicken-little NIMBY's fault...no wait, maybe the WPPS bond default may have had something to do with it? Whatever- it's still toast.

We've been getting nights in the mid 40s too, and I haven't even turned on the heat yet, go figure! :-) But when I do, the radiant zones will be the first-used, leaving doors open to let them heat the rest of the house.

I'm not the smartest tool in the drawer so bear with me on this.

So are we saying here that the Mitsubishi system is sufficient for ALL of my heating and cooling with no backup and the cost is not that much more than radiant heat?

Surely the comfort of that system isn't as nice as radiant. I bet it's better than a central furnace though, even a variable speed unit.

Do you have a basement, a usable one? I am with the radiant types on this one (almost always am). I renovated a 800 ft2 house a number of years ago that had electric baseboards and we put in an overpour with tubing in the previously unusable basement and used a small gas water heater (Integra, not available anymore) and monitored the main floor electric usage. It dropped by almost 90% and there were no insulation changes at that time. The average MAIN floor temp increased enough to limit the use of electrics to quite cold weather.

If you cannot do that, the condensing tank with some panel rads would be my choice. Until someone comes out with an affordable A/W heat pump for very small loads, this is is the best you will get IMHO.

Surely the comfort of that system isn't as nice as radiant.

You haven't talked much about what your definition of "comfort" is other than to wonder if a minisplit can "keep up" during the sharp cold spells. You could easily oversize the system to account for the slight derating that takes place at those very cold temps, or add some electric resistance wire under the bathroom or kitchen floor that would not only boost interior heat, but give you some comfy floor surfaces at pre-arranged times.

As for cost, since you opened with considering both a radiant system and a minisplit system, it would appear that with just a minisplit, you would be spending much, much less, not more.

Posted By strategery on 21 Sep 2012 01:39 AM
I'm not the smartest tool in the drawer so bear with me on this.

So are we saying here that the Mitsubishi system is sufficient for ALL of my heating and cooling with no backup and the cost is not that much more than radiant heat?

Surely the comfort of that system isn't as nice as radiant. I bet it's better than a central furnace though, even a variable speed unit.

That depends entirely on what your actual heat load is at design temp.  But there are now MANY houses well over 1000 square feet being heated & cooled with 1-2 heads of ductless.  Typically one head per floor works in a superinsulated house with ~U0.20 windows and a not-huge glazing fraction.

If your heat load at design temp is under 20,000BTU/hr  (and I suspect it is, with good margin) you have several ductless heat pump options.

The cost savings are primarily from having just one system rather than two.  A 2-ton single-head minisplit might set you back $5K, installed. A 2-head version might be as high as $6.5K.  Sometimes it's cheaper to use a pair of 3/4 ton units than a 2-head multi-split but the standby power use goes up, with 2 controller boards and  two compressors using internal standby-heaters to keep the compressor within startup tolerance rather than one.

Comfort is quite good compared to most ducted-air systems, but it's more "warm light summer breeze" effect when cranking full-out rather than a "cozy-toes"  factor you get with radiant floor.  Most ductless heads are quieter than your refrigerator when running at low to mid speed, which is where it would be running most of the time.

First & foremost in any heating app is to get a handle on just how big the heat load is at the 99% design temp. From that all else flows.  In a house that tests under 2ACH/50 a simple-minded I=B=R approach works fine.

Say your outside design temp is -10F, your interior design temp is +70F, for a design delta of 80F degrees.

With R30 walls, the U-factor (heat loss per square foot of area per degree F) is 1/R, or 0.033.  For an 80F delta that's (0.033 x 80=) 2.65BTU per square foot of exterior wall.

For an R60 attic it's half that per square foot of upper-floor ceiling, 1/60 x 80F or 1.33 BTU/ft of ceiling.

Look up the U-factors for your windows & doors, multiply those by 80F.

Ignore the heat loss through the R10 under the slab.

Give yourself a 15% margin for air leakage, ERV/HRV ventilation, random bits of thermal bridging etc.
 
Subtract 250BTU/hr for every sleeping human.

Subtract another 200BTU/hr for the refrigerator.

Add it all up, and that'll be pretty close to what Manual-J would deliver.

For a 20 x 25' 2-story (1000 square feet) 20' tall with a 15% window to floor area ratio (150 square feet) and U0.25 windows & two 36" x 80" (20square foot each) U0.25 insulated doors (R4)  it works out to about like this:

Windows: 150' x U0.25 x 80F= 3770 BTU/hr
 
Doors: 20' x 2 x U0.25 x 80F= 800 BTU/hr

Walls: 20' x (20 + 20 + 25 + 25)= 1800 square feet gross, less 150' of window and 40' of door yields 1610' of exterior area. 
1610 x U0.033 x 80F= 4250 BTU/hr

Ceiling/attic: 20' x 25' x U0.017 x 80F = 680 BTU/hr

subtotal:  3770 + 800 + 4250 + 680 = 9500 BTU/hr

Multiply by 1.15 for fudge-factors, and you're at 11,000 BTU/hr.

Assuming 2 humans, one refrigerator, subract 500 for the peops, 200 for the refer and you're at 10,300 BTU/hr.

A 1-ton Mitsubishi H2i wouldn't quite cut it at -10F (it almost does),  but a 1.5 ton could heat it with margin, as would a pair of 3/4-ton units.

When the heat load is that low, consider that just three 1500W portable space heaters could actually meet the design condition load, if for some reason you had a catastrophic failure on the ductless and were waiting for repairs on design-day.  If you run into load-balance issues with some rooms too cool when it's really cold out judicious use of radiant cove heaters mounted in lieu of a section of crown molding can quickly make up the difference when those rooms are occupied.

So, do a spreadsheet calculation on the actual house/doors/windows/occupants in question-  the heat load is probably in the 10-12K range, and almost certainly under 20K with the windows closed.  Yer on yer own if you insist on sleeping with the windows open when it's -10F out.

As a sanity check, here's an example of a 1200'/1400'  house only moderately better-insulated & tighter than yours,  heated & cooled with a  single-head 1.5 ton Mitsubishi:

http://uphillhouse.wordpress.com/20...ed-part-1/

http://uphillhouse.wordpress.com/about-2/

Manual-J came in at ~12.4K.  It's located in Cambridge NY (Climate Zone 6) where it hits negative double-digits most years (if not this record-warm winter). The 99% design temp is probably only in single negative-digits, but not far from -10F.

Radiant for comfort and economy. Heat pump for economy. As Dana points out, it all starts with a proper heat load. Once the heat load is secured, the experienced designer can advise with some confidence on the predictable results of the various options.

We are just finishing a total mechanical remodel on a 1955 brick bungalow in Minneapolis. Radiant slab in the basement, garage and sub-floor aluminum panels for the existing oak floor driven by an IBC condensing with indirect for domestic hot water. Cooling will be a properly sized AC unit at low (noise) velocity ducted to each room.

Perfect comfort and the driveway will melt snow too!