I was kinda jumping on the Tankless bandwagon. These are very clear reality checks on that technology coming from you. Definitely needed to hear.

As Dana points out, annual maintenance of all tankless water heaters is a must and not inexpensive. We install xyz tankless water heaters here in the Minneapolis/St.Paul area and service many more including Eternal, Navien, Noritz, Rheem, Rinnai, and Takagi. Almost all tankless water heaters retrofits will require a new gas line if not a new "bigger" gas meter. Furthermore, a typical tankless water heater (199,000 btuh input) will not fill a large tub (over 60 gallons). For this you need standing/stored hot water in a tank for a fast tub filler.

This is also true of most combi-boilers and why we design many of our radiant slab basement systems with a condensing tank/type or storage water heater with a sub-system for space heating. Better performance, longer life and less service.

Thank you Dana, for the "Truth about tankless water heaters".

Great discussion re tankless. I am in a propane/ no ng area. Badger can you g int a bit more detail about how you do a radiant heat system with a condensing tank type heater and then tie in DHW? Dana, as always, thanks for the insights....

Send a pdf or CAD file and we do the rest. We do not design them over the net as each system requires different specs.

Birdman: At the price of propane (and heating oil) in my neighborhood I'm steering people away from the combi-HW/radiant approach, particularly for people with low loads. Even at New England's higher-than average electricity pricing, either better-class mini-splits (cheapest) or the Daikin Altherma (hydronic-output air source heat pumps suitable for radiant slabs, but not-so-cheap) pay for themselves in reasonable time frames. In central-MA(outside design temps from 0-5F) heating with mini-splits (or Altherma) with the 15-18 cent electricity is still less than half the operating cost of heating with $4.00 oil, or $3.00 propane (and the price of electricity is far less volatile to boot, being a regulated market.) But the rationale varies with YOUR actual utility & propane pricing.

Here going with mini-splits is something of a no-brainer type of investment, but you'd need to sharpen your pencil at the upfront cost of the Altherma, and decide if the extra cush-factor radiant floors is worth the upcharge.

Even in this region's current grid-mix going with these modulating variable-refrigerant-volume air-source heat pump technologies is lower-carb too. See: http://www.rmi.org/cms/Download.asp...tPumps.pdf

I'd be interested in how geothermal fares in similar calculations (say in central MA).

Thank you Dana, for the "Truth about tankless water heaters".

Definitely...thank you. I will be reviewing these posts for some time.

Essentially, I will be researching these points for a bit...

Tankless water heaters are pretty much a waste of money unless you have a massive tub to fill. A condensing tank-type water heater and a drainwater heat recovery heat exchanger downstream of the showers can operate at "apparent efficiency" greater than 100%, in a high showering-use envirnment, where an 0.98EF condensing tankless would struggle to break 90% (unless the drainwater heat recovery was used there too.) Even a pretty-good 0.68EF tank can beat a condensing tankless with the benefit of drainwater heat recovery.

but it takes a seriously large tank to fill spa-sized tubs, and higher BTU/hr commercial tankless & ganged-tankless units are growing in popularity to overcome the gpm limitations without having to build a tank-farm in the basement. (I personally don't get either the mega-tub craze or the 6 side-spray monster showers, but they're out there.) They're not a great solution, but they take up less space, and don't run out of water. The main point was that there's no real payback on efficiency with a tankless- they cost more up front, have more maintenance issues, and don't have sufficiently higher efficiency to break even on the PITA factor, let alone fuel savings. Where efficiency is primary, for the money you get more showering performance & efficiency out of drainwater heat exchangers, but they do nothing for tub filling efficiency/performance.

The installed cost of a condensing tank is almost always lower than that of a tankless, the maintenance is lower, and the "personality" quirks far fewer. In low-water-use homes on the gas-grid a standard atmospheric drafted tank might be "right" from the pure economics of it, but in tight homes it's better to have something power-drafted, if not fully direct-vented/sealed combustion.

The smallest stainless steel Polaris runs ~$2700, has a good track record, and would last at least as long as the best tankless. The smaller-burner Vertex runs ~$1700 and has of similar design, but would only last about as long as most standard tanks. Either has far fewer maintenance issues as a tankless.

A condensing Rinnai or Noritz tankless may run $1200-1500 for the basic unit, but requires fatter gas-line plumbing, and usually have custom venting pieces adding to the initial installed price, and more labor-intensive annual maintenance that adds substantially to the 10 year & 20 year costs of ownership, even in case where it's break-even on installed price.

The internals of a condensing tank are just plain dumber and more reliable than a tankless, which of necessity needs a much bigger, and MODULATING burner.

I don't hate 'em, but for most situations a tankless isn't worth the upcharge unless you have no space, or need truly "never-ending" hot water. For as-used efficiency they only beat a condensing tank in either very low volume applications (where standby losses eat into the tank's efficiency), or in truly continuous burn applications, where their very modestly-higher raw combustion efficiency gives them an edge. In most 2-3 person household apps a condensing tank will outperform a tankless on average efficiency, due to the short-cycling losses of the tankless on the more frequent low-volume draws.

As Dana points out, annual maintenance of all tankless water heaters is a must and not inexpensive. We install xyz tankless water heaters here in the Minneapolis/St.Paul area and service many more including Eternal, Navien, Noritz, Rheem, Rinnai, and Takagi. Almost all tankless water heaters retrofits will require a new gas line if not a new "bigger" gas meter. Furthermore, a typical tankless water heater (199,000 btuh input) will not fill a large tub (over 60 gallons). For this you need standing/stored hot water in a tank for a fast tub filler.

This is also true of most combi-boilers and why we design many of our radiant slab basement systems with a condensing tank/type or storage water heater with a sub-system for space heating. Better performance, longer life and less service.

Posted By jonr on 30 Apr 2013 12:13 PM
I'd be interested in how geothermal fares in similar calculations (say in central MA).

Efficiency-wise typical-geo edges out ductless in central MA, but at a gia-NORMOUS uptick in cost.  I've yet to see a quote for under $25K for even a 2-ton GSHP, and typical systems are well north of $30K.  My personal exposure is a very small sample, but consistent with the averages of the rebate subsidized systems in neighboring CT. At a 50 cents a kilowatt hour it might eventually pay to go geo vs. Altherma but not at 15-18 cents (recent central MA delivered residential retail rates.)   To make a rationale for geo vs. mini-splits electricity would have to be even more expensive still to pay off in any reasonable time frame. 

At the current costs & subsidies of photovoltaics & electricty in central MA it's more cost effective to go with the modestly less efficient ductless and spend the difference on net-metered PV. Two tons of ductless is well under $10K, (even when split between 4 heads), and in heating mode can typically  deliver about 30,000 BTU/hr @ +5F (varies a bit with vendor & model.)  Assuming one could even get a GSHP system that could deliver that much heat for $25K (not likely given recent experience), and took the 30% tax credit you're still looking at about a $10K difference in system cost.  That $10K  buys about 2.5kw of grid tied PV (unsubsidized), which in this neighborhood would deliver about 3000kwh/year-  more than  1000kwh greater than the difference in heating power use between a GSHP with an average COP of 3.5 and ductless with an average COP of 2.8 on homes with peak loads in the 30KBTU/hr range. You'd use more annual heat pump power, but reap even more than that difference from the PV at the same up-front money (even without factoring in the PV subsidies).

When the various federal/state/local subsidies are factored in for the solar in central MA it becomes a no-brainer, provided you have reasonable shading factors for the solar, even if your GSHP system designer was a genius and managed to deliver an average COP of 4.5 at that (comparatively low for local GSHP) price point.

I've been waiting in vain to hear of GSHP becoming more cost-competitive in this area (but I've stopped holding my breath.)  You can make a case for GSHP comparing lifecycle costs against oil or propane, but even at $4 oil it's a decade or more proposition before any savings are realized (even on a simple-return basis, let alone NPV), with considerable design risk. That compares to ~3-5 years simple return for ductless,with fewer aspects for the designer/installer to screw up.  With the Altherma it's in-between- somewhat less risk & system cost compared to GSHP, but significantly more than with mini-splits.

Good points, thanks. I'd want to see pretty accurate numbers on temperature swings in closed door rooms before going with ductless ASHPs. But one can always add more heads or aux heat.

Open loop geo + the latest inverter heat pumps have some very impressive COP values and lower upfront costs, but then it has other questions (like "do you already have a well", water quality and where to dispose of the water).

I race  on skis in the local beer-leagues against one of the (until very recently) occupants of a near-net-zero house in Stow MA. I only today became aware that it had been instrumented by the Fraunhofer group, who did this writeup of both their house, and another near net zero house in Townsend MA.  Curiously, the Stow house has more even room temps despite being a 3.5 story (finished basement & attic) heated & cooled with a single ductless head for the whole shebang. 

The folks in Stow sold the place just last week, have moved into a circa 1958 rancher to be closer to work/school, and are already missing the even-ness of the temperatures already! They're talking about point source heating it with a pellet stove after they tighten it up some, but wouldn't be surprised if they went ductless instead, after their Stow experience.  But the high-R has at least as much to do with the comfort of their "old" house as the modulating ductless heating/cooling- it won't be QUITE the same, but after their planned upgrades they should still do OK. (I haven't seen the layout of their "new" digs.)

The occupants in the Townsend house seemed to prefer it cooler (according to the writeup), but also had a bit more temperature variability, despite having one ductless head per floor rather than just one.

A relative's single-story place in WA is not even CLOSE to superinsulated (significantly sub-code, in fact) but does just fine with a single mini-split head in the main living/dining area and it's adjacent rooms, with occasional use of resistance heating for the doored-off master bedroom area at the other end of the house. The bathroom off the main area is somewhat lossier, with a noticeably lower temp, but not to the point of discomfort. The adjacent workroom/office is tighter and is just fine with the door kept open. Previously it had been heated with an electric hot air furnace at much lower comfort at more than 2x the power use, despite deep overnight setbacks (whereas they "set & forget" the mini-split).   It's not anybody's idea "luxuriant comfort", but it's more comfortable (and quieter) than traditional forced hot air heating. The 99% design temp there is in the low 20s, but it's not a stretch to assume that a tighter code min house in a location with a 0F outside design temp would match or exceed it on comfort, given a similar layout. 

I have another relative in that area heating a 2-story with a single head mini-split who seem to like it MOST of the time, though I heard one third hand comment that it doesn't cut it in all rooms all winter. (I'll be getting more details when I visit in July.)

We need to move away from "so and so thought it was fine" to "X degrees under Y conditions" and tested models.

Unfortunately, the closed rooms always move in the wrong direction, temperature wise. Ie, in the winter, when you are keeping the main area as cool as possible, the back room will be cooler. And in the summer, it will be hotter. So you get something like a "-10F (winter), +5F (summer)" spec. Of course the -10F is easier to fix (using supplemental heat). And the summer number might be far higher if there is solar gain.

Perhaps some of the other cold climate mini-split manufacturers will realize that producing a refrigerant to water head would be easy. Altherma needs some competition.

Altherma has plenty of competition, but it's all focused on the European market where hydronic heating is the rule rather than the exception. (eg: http://energimyndigheten.se/sv/Hush...ar1/?tab=2 )

Sanyo even has a GREAT R744-refrigerant (CO2) hydronic output air source heat pump for sale in the EU, (http://www.kcc.se/products.php?cid=26&pid=23 ) but air-handler based heat/coolth distribution reigns supreme in North America, making heat-pump hydronic systems the extreme minority within a minority here.

OK, "Altherma needs better priced competition here in the US". Why it doesn't is an interesting issue. Somewhere I saw of figure of 94% of new US homes are using ducted systems. Of course residential mini-splits were once rare in the US, so things can change.

Posted By jonr on 01 May 2013 11:12 AM
OK, "Altherma needs better priced competition here in the US". Why it doesn't is an interesting issue. Somewhere I saw of figure of 94% of new US homes are using ducted systems.

That's exactly the problem!
 
If only 6% of new homes have hydronic heat, how many of those are even using low-temp radiation suitable for use of heat pump technology?  I suspect that at least half of the 6% are gas-fired, and 2% are propane or oil fired, splitting the final sub-1% between biomass boilers, electric boilers, solar-thermal, and hydronic output heat pumps.  How many players can compete (and how aggressively or profitably) for a piece of the <0.1% slice held by hydronic output heat pump systems?

In Europe heat distribution on new construction since about 1980 has been low-temp (by the <140F definition) hydronic, and air conditioning is all but unyheard of in private residences.  Radiant floors are somewhat rare there, but it's neither rocket science nor outlandishly expensive to install panel radiators sufficient for delivering design-day heat with sub-120F water.

In the US people want air conditioning, and once you're committed to the ducts, it's far cheaper to use them for heat distribution too rather than springing for hydronic heat distribution.  In New England air conditioning hasn't been the rule in new construction until recently, and hydronic heating is still common, but there is a shift occurring toward ducted heating here too, particularly on tract development housing where nickel accounting drives them to the lower cost options.

My house was something of an outlierm built 1923 (when steam was king) with a coal-fired forced hot air system.  It was converted to oil at some point in the 1940s or maybe the 1950s, then to gas in the 1960s.  (I'm really curious as to how the original air handler & furnace looked & worked, but we have no pictures.)  Mi esposa started the conversion to hydronic heat distribution in the 1990s before I was in the picture, and the ducts are now used as the "Hail Mary" hydro-air second stage should the retrofit radiant floors & radiators not keep up during the cold snap of the century, and for the random handfuls of days that the sensible AC load is sufficient to fire up the central AC.  The room to room balance of the ducted system is pretty crummy, and I micro-zoned much of it to deal with dramatically differing heat loss characteristics. There's really no way to run the place as a single zone and achieve perfect balance, (the bedrooms would have huge under/overshoots when controlled by the central T-stat) yet heating solely with the wood stove is pretty comfortable in all rooms except during the deeper mid-winter coolth, as long as the bedroom doors are kept open during the day.  I haven't run any of the hydronic zones since about 25 March this year- she likes the wood stove better than the radiant floors(!), and the bedroom zones weren't calling for heat anyway.  It'll probably be after Thanksgiving before I have to start up the family-room floor zone again (the lossiest room in the house).

How many players can compete (and how aggressively or profitably) for a piece of the <0.1% slice held by hydronic output heat pump systems?

As far as I know, for geo, several and for air source, only one. And of course sales elsewhere (like Europe) make it more attractive. Maybe someone sells a hydronic head that can be used with the same compressor that cold climate mini-splits are using.

And just how many of the ~100,000 GSHP systems installed in the US per year are hydronic output? (Anybody have the hard data handy?)

According to the numbers in Table 3.2 of this document there are only 150 hydronic output models (presumably each size counts as a "model") to about 1500-1700 air-output models:

http://www1.eere.energy.gov/buildings/pdfs/ghp_rd_roadmap2012.pdf

There are about 2 million new homes built per year in the US (with a lot of year to year variation), and even if all of the ~100,000 GSHP systems were going into new construction that would amount to about 5% (but in fact a large fraction, possibly the majority, is retrofits into existing houses), and if the 10/1 ratio of available models is a rough reflection of the air vs. hydronic unit sales volume, that would make the hydronic output versions a 0.5% fraction of the total, 10,000 systems best case.

Divided between 150 models yields a paltry average of 67 units in annual sales per model. Maybe some model & size sells as many as 500 units a year, but I'd be surprised if any particular model broke 1000.

What business decision maker in their right mind is going to go very aggressively at that? One reason (of several) why GSHP systems are so expensive is the extemely low production volumes- they're barely more than beta-prototypes, with prototype-production style engineering & manufacturing costs.

Daikin has a decent world wide market share for variable refrigerant volume air-source heat pump technology behind them, so they can at least dip a toe into this market (especially since the Altherma has been edging out the competition in Europe on tested system efficiency) but it's hard to rationalize the marketing expense of breaking in here for a new player, going shoulder to shoulder in a tiny market. I'd be surprised if the total sales of the Altherma in the US since it's 2009 US debut exceeds 500 units total (divided among nine models/sizes.) Who want's a piece of that action, really? How many vendors would that US market support? Popular mini-split models are FLYING off the shelves by comparison, in an increasingly competitive US market, but are still way behind hot air furnaces & ducted heat pumps in annual installations.

I believe I was told something over 2,000 Daikin Althermas into the US market at last count.

In the US people want air conditioning, and once you're committed to the ducts, it's far cheaper to use them for heat distribution too rather than springing for hydronic heat distribution. 

I and the market agree. It would take a few not too likely things to change that. For example:

1) single story slab on grade house (low cost radiators)
2) low cost AtW heat pump
3) DOAS or fan coil cooling

But the upside could be increased efficiency (over AtA) and better temp control (over a limited number of mini-splits).

It doesn't have to be single story slab on grade as formed concrete slab over basement or main floor slab could easily be used for a lot more construction.