Posted By ICFHybrid on 01 May 2013 11:49 PM
I believe I was told something over 2,000 Daikin Althermas into the US market at last count.

Count me as surprised! That would make it north of 500 units/year then!

Divided by 9 models, is the same order of magnitude sales per model/size guesstimated for hydronic GSHP, a handful of 10s per year per model/size.  But since many of the components are shared with their mini-splits etc, it's not quite the same prototype-ish production costs as with GSHP.

I wonder how many Quaternity's they sold in the same period?

I'd be interested in the cost per ton of WtA (higher volume) vs WtW (low volume) geo heat pumps from companies like ClimateMaster. Once you have most of the pieces in place, I'm not sure that a different design makes much difference in production cost.

FWIW: Some of Marc Rosenbaums near-Net Zero houses are heated/cooled with a single-head 2-ton Daikin, and use electric radiant panel radiators to adjust-up the doored off rooms as-needed for comfort.  There's a writeup of one of the those developments that tracks both the radiant-panel and minisplit power use (and many other aspects of the house) located here, (see figure 11), but the upshot is that the combined net COP is still > 2. 

While you can increase the average efficiency by quite a bit with more ductless heads or a DOAS/hydronic or GSHP approach, but at what cost? In particular, at what cost relative to the cost of the PV required to cover the efficiency difference?  Putting the money into the building envelope rather than the mechanical systems buys comfort while lowering the loads.  At some point occupant behavior trumps the actual heating/cooling loads or the raw efficiency of the systems on the amount of energy used, as many PassiveHouse and Net Zero Energy dwellers have discovered.

Marc's houses are built to about 1/2 the R-values of what it would take to hit PassiveHouse spec in that climate, and are Net Zero Energy for some families, not so much for others. The roof-R is barely over IRC 2012 spec, but the windows/walls/foundation are ~2x code. They're also pretty tight- PassiveHouse tight for most. (The leakiest house in the development tested <250cfm/50.)  From a "best practices" & planning point of view I find this approach pretty compelling, taking pretty much a lowest-cost approach to achieving the performance without sacrificing the comfort & aesthetic issues.  Marc designed the mechanical systems for the  Concord Riverwalk houses, which are by all reports pretty nice houses, with 1-2 ductless heads per unit.  For the kind of budget we're talking for the Chicago town house you could go quite a bit nicer still- the up-charge for the high-efficiency building envelope just isn't all that much, especially if you're taking some of the costs out of the highest-possible-efficiency mechanical systems.

BTW: Marc did some datalogging of room by room temps his own house that did NOT have the radiant panels for balancing, heated only a 1-ton Fujitsu on the first floor. His room to room deltas are lower than mine were prior to micro-zoning, and lower than when I'm point-source heating with the wood stove. His whole-wall R-values are about 3x mine, and his climate is modestly warmer in winter but not by a huge amount. (Design temps in the low-mid teens, rather than low to mid single-digits.)

Looking at Marc's data, I conclude that he definitely needs either a separate head or supplemental heat in closed rooms for good comfort. Of course good is quite subjective. I like his use of setback, which isn't very practical with a hydronic slab.

I completely agree that the upfront costs on a COP=5 system might make a COP=2 system the better choice. And that the ROI should always be limited by what PV solar would cost.

Posted By jonr on 02 May 2013 12:08 PM
Looking at Marc's data, I conclude that he definitely needs either a separate head or supplemental heat in closed rooms for good comfort. Of course good is quite subjective. I like his use of setback, which isn't very practical with a hydronic slab.

I completely agree that the upfront costs on a COP=5 system might make a COP=2 system the better choice. And that the ROI should always be limited by what PV solar would cost.

Marc says as much in the article that he'd put a separate head in the other bedroom if he actually used it, and bits of resistance-heating elsewhere to balance it all out if necessary. 

That house was neither his design, nor was it designed with point-source heating in mind. When he first moved in the heating plant was a pretty-good Buderus oil-fired boiler & baseboards.  It's clearly possible to design a new house more mindfully toward multiple-point source heating, and have fewer remote-room temperature issues, whereas in retrofits you're more limited in how the layout works and just how much you can reduce the load of a remote room (short of gutting it and pretty much starting over on the layout, as my  friend did on his DER.)

The ROI relative to PV isn't always a simple function.  Where it's simply net-metered at retail it's a pretty easy calculation, but in places like Austin TX where the PV's power is compensated via a VOST (= Value Of Solar Tariff) formula, based on the real-time value of the power to the utility & the stranded costs it can be a bit squishier.  The deal struck in Austin is pretty good, whereas a similar approach in nearby San Antonio is pretty daunting for the PV owner, termed  a "solar killer" by some local industry insiders.  Local subsidies and conditions WILL vary, and it's worth running the numbers more than once before locking into a decision in a fluid & rapidly evolving market.  In most places net-metering is still at full residential retail for the site-used power for grid attached PV, but it's important to weigh the details.

CPS Energy just abruptly announced it will terminate net metering

Things like this make the value of solar nearly impossible to calculate. To be really safe, add in the cost of batteries.

It's interesting to look at where geo would make sense - for example, clean water and discharge are all ready available for open loop and electric rates are high. In that case, air source HPs probably aren't the best choice.

There will always be a place for GSHP, but that place isn't necessarily going to be high-R new construction. Heat pumps (all types) now comprise something like 40% of the residential heating market in Sweden, according to an RMI document:

http://www.rmi.org/PDF_heat_pumps_an_alternative_to_oil_heat_in_the_northeast

I'm not sure what the market share for ASHP is in Sweden, but my understanding is its a mostly GSHP show, averaging over 25,000 installations/annum (about 1/4 of the annual total installations in Europe in recent years) in a country with a population a small fraction the size of the US. The hydronic ASHP market share of space heating in Sweden appears to be 33% to GSHP's 67% share in 2007, up from a 25%/75% split in 2006 , looking at tables 2 & 3 in this document, (that I have yet to fully read):

http://www.ehpa.org/fileadmin/red/Heat_Pump_Statistics/European_Heat_Pump_Statistics_2007.pdf

It's not clear of ASHP has continued to gain share there since.

The groundwater temps in Sweden aren't all that different from US climate zones 5-7, which presents some performance issues in the cold-dirt north:

http://www.gns.cri.nz/content/download/6905/37729/file/Swedish%20Ground%20Source%20Heat%20Pump%20Case%20Study%20%282010%29.pdf

Swedish electricity prices are higher than US pricing but not the highest in Europe, at about 20 Euro-cents per kwh, (about 26 US cents at today's exchange rate) and probably not the total driving factor behind GSHP over ASHP sales there.

(edited to add)

Sweden is an interesting case, since as I understand it, their codes related to a building's energy use is only the result, not the method. Unlike the IRC and local codes there are no prescribed maximum U-factors or minimum R-values, no minimum efficiency of mechanical systems, etc, only parameters for how much energy the building uses, leaving it up to the builders, engineers & architects to meet the spec, which must be verified & remediated if necessary after the fact.  This gives the engineers & architects the freedom to come up with the right mix of building envelope & mechanicals, but also puts the onus on them (as well as on the builders to properly implement the design).  If it's more economic to spend more on the mechanical system than the building envelope (or conversely), that's up them. 

Sweden has gone from a nearly all heating-oil fueled heating in the 1970s to an overwhelmingly heat pump + district heating mix in only 40 years. With their newer somewhat unique building energy codes, the trends there are worth watching, since it's likely to tip towards a lowest-cost mix of solutions for their climate, which is relevant to the colder half of the US and most of Canada.  It's certainly relevant to a Chicago climate, which would be comparable the warmer edge of Sweden, say, Gävle or Uppsala, from a wintertime outdoor temperature point of view, but not from a cooling season perspective.