Joe is right except as it concerns heating guys. They are generally "brought up" with a strong work ethic, frugal dispositions and critical diagnostic skills, making it harder for them to fall for the old "100% efficient hack. Determine your available fuels, cost per therm for each and efficiency discount if applicable. ROI is a concern and is a valid part of any conservation strategy.

You'll notice I qualified my 100% eff. comment to "while it's heating" which is far different from 100% efficient. Of course if we wanted the highest efficiency, we'd be talking about a water to water geo here. At 325% (3.25 COP) efficiency or better, nothing else comes close. Of course with the 15-20K installation cost, payback is not so quick.
I agree with Badger about things to be considered when making this decision and would add cost of maintenance in the mix as well.

Didn't mean to jump on you Joe. Resistance electric guys have me all worked up lately.

Of course you Geo guys love to brag too, but you have COP, and who doesn't like that?

hheehee

Of course I like to tease them too about their paltry (imaginary) 100% eff. Unfortunately high propane and fuel oil prices make it attractive again in the sticks.
Curiously a project I'm on right now has a wing of the house (about 1800 squares) already finished (3 stories) and heated by electric baseboard. At $1000, yr to heat, I can save them a lot, but with no duct work available inst. cost is pretty high. Currently looking at ductless, hydronic, wall units (if anyone has a brand to recommend...). Equation really has to do as much with how he hopes to cool the area in the summer as savings alone puts pay back for anything in the 30 year range.
J

High Velocity?

Rob,
thanks for the input.  This was what I was hoping for and the respective low costs of the individual systems has convinced me to keep them independent.

To sanity check, to meet my 13K design heat load, I calculate that it would be running 84% of the time if a single 4500W element electric HW heater was used so derating to 75% means I'd need to have a dual element heater to keep up at the design point.  Make sense?

The smallest DE electric I've found doing a quick search was 30Gal.   Interestingly, Home Depot's price was cheaper for the 40 Gal than the 30.  Any reason (assuming I'm adding extra insulation anyway) that 40 would be significantly different efficiency than a 30?

Finally, since this is a closed loop system, any reason I couldn't charge the system with Glycol rather than water?  This should reduce corrosion and mineral deposits and, perhaps, reduce any effects of being off for long periods of time.  It would also make it easier to pipe in a solar collector later on.  Thoughts?

By copy to Joe, thanks for persisting.  I still may go to propane, however, not for cost reasons (electric HW actually seems 'right sized'), but because I have propane on site anyway (for backup generator) and it would permit me to keep heat in the workshop even when I've lost power as the electric power to keep propane system operating would be very low.

Tom

OK, my thoughts,
I like the little electric water heater 1,000 sq ft shop solution, bronze pump to keep it clean and the simple breaker switch of/on for the season, no drain down.
But not for the guy with the 3,500 sq ft house in California. That is a full blown residence, deserves a look at proper boiler, solar application, storage tanks. The flash water heater will fail in 10 years tops, efficiency in the 84% range, an old iron boiler will do that and last 30 years, with a modest solar compliment your efficiency will increase, and the system can remain mechanically straight forward. Mod con/solar domestic preheat would be my preferred choice on most any residence of size.
Some one slipped the flash water heater Kool aid out last night........
Joe, as to ductless wall units, have a look at the Myson fan convector group. We distribute as do other's, this link will give you an idea of the variety of types they offer. http://www.blueridgecompany.com/hydronic/forced-air/fan-convectors/80
Hi V is great as well but requires a bit more planing, easier in a unfinished frame.
Dan

Boy don't I know THIS movie from beginning to end, backwards & frontwards! ;-)  (This is somewhat less than half the heat-load of my house.)

But as opposed to the cheapest, perhaps the gas/propane water heater with rated output closest to the design-day load.  Derating it as heavily as 75% is a bit conservative in a slab system.  The raw combustion efficiency will be somewhat higher than in DHW/storage mode due to the fact that the water turbulence on the water-side of the heat exchanger will be high for more than 90% of the burn time.  The lower heat exchange efficiency from the higher return water temp is largely offset byt the higher turbulence. (By contrast, when simply heating water most of the burn time takes place in a static tank.)  Ten percent is probably more than adequate.  In static burns starting with colder water the raw combustion efficiency is about 75%, but it'll approach (maybe even exceed 80%) pumping a lot of water through a sufficiently small tank (are there any cheap 15-20 gallon gas fired tanks out there?) I'd assume something like 78% combustion efficiency, which is what the eKoComfort folks in Canada measured for a larger tank HW heater in a heating system feeding an air-handler coil.

The ideal sized burner would then have an input rating of something like 110%-of-load/combustion efficiency, or 13K*1.1/.78= 18,300btu/hr.  Most 30-40 gallon tank heaters are something like twice that, with more than twice the standby losses, but there are a few with 30kbu imputs that won't be overkill (eg. AO Smith model GVCA-30 or GCNH-30 ).  There may be even smaller RV versions (usually propane or propane & electric), but they may be considerably more money.  At 30kBTU input you're still getting nearly 2x your design-day load.  If you can find something under 25k-input it'll run more efficiently, but don't spend anything like 2x for it.

Many whole house tankless HW heaters will modulate down to under 20K.  But at 3x the money, for a gain in efficiency that will pay back at under $100year in your case at current NG prices it's hardly worth it.  But the tiny over the sink models that are ubiquitous in Europe (or at least used to be), can do it!  I'm only aware of one US manufacturer (Marey, in Puerto Rico), but their bottom-of the line sub-$200 version may actually prove to be perfect for this application:

http://stores.gonetankless.com/-strse-5/5L-NG-Gas-Tankless/Detail.bok

http://www.marey.com/english_/home.htm

I'd try to find out more about the actual min specs on the BTU-input & max water flow rates, ( as well as warranty & support). but this might be the sweet-spot on cost & efficiency.  It's max input is 10kilowatts, which is 34000BTU/hr. but the flame can be adusted up/down with a simple front knob.  If it's like most tankless burners, somewhere in the lower to middle-third is typically high point of heat-transfer efficiency on tankless heat exchangers.   If you turn it down to where it just keeps up with the heat load it'll probably give you ~85% total system efficiency with effectively zero standby losses (vs.  70-78% average for a tank with an oversized-for the load 30KBTU input burner, after considerable standby losses are factored in.)  The fact that it's designed for Caribean-type source water temps, odds are it's thermal efficiency in a slab system is very close to that advertised.  The igniton runs off water flow & D-cells- so there's no extra wiring to the heater, just the hydronic pump for the slab.  System cost-wise it's going to run about the same (or possibly even cheaper) as a pretty-good 30gallon tank, yet run at higher efficiency.

There may be futz-factor maintenance issues, but I'd be surprised if they were as big a pain in the butt as annual draining & refilling a tank.

If you go with a tank to,  wrapping the tank in R10+ insulation and keeping the tank thermostat at 110F or lower will make a difference on standby losses. If the building is fairly air-tight, putting it in a closet that's air-sealed from the rest of the building and adding a small external wall vent for supplying combustion air (with a low-pressure atmospheric damper flap) and insulating the pipes will make a difference too, but again, not enough of a difference to spend a lot extra on building the closet.  (The tank & pipe insulation + the vent, is worth it, if you have the isolating room/cabinet already.  Be sure you install it with the necessary fire-safety clearnaces.)

Many, good points are made for both sides of this one. To your points Tom I would certainly be more reluctant to add antifreeze to the electric than a gas fired appliance. You do want to remeber that this derates your heat transfer so it is not always preferred. Most small generators could run a depot electric tank (stab in the dark.....15A 220v with no heavy motor startup drop).
I'm under educated on the wall unit questions and am distracting from Tom's thread so I'll re post the question.
Thanks to all,
J

Dan, thermodynamics doesn't care if a home is a "full blown residence" or a shop. heat load and degrees are what they are either way. So cal can have anywhere from 1500 to 3000 degree days a year. Once corrected for summertime degree days, that is at best a third of the degree days of a real heating climate and more likely something more like a quarter to a sixth. and he's using SIPs, not the typical "2x4 and some plaster" that most SoCal houses are built with. Now he might have window walls or whatever, fine, but those are questions only answerable by a heat load calculation: not by an off the cuff recommendation on a forum based on square footage alone which, as any heating professional should know, is not adequate to figure out the answer to such a question. If solar DHW makes sense for the fellow, great, that's a totally separate discussion that has nothing to do with what his heating heat source needs to be, unless his domestic load is so high it's worth getting a high efficiency heat source just for that (which is pretty freaking unlikely in the land of water rationing). Though if he installed enough solar to heat with as well, that would just further reduce his heating load to be met with backup and make MORE of a case for a cut-rate backup. Fact is, if tom's in a real heating climate, and our SoCal friend is where he says he is, our SoCal friend could have 3 times the "design day" heat load tom has (call it 40kBTUs/hr) or more and STILL be in the exact same boat in terms of yearly energy usage as Tom is: way, way, way away from economical for a mod/con. Now, economics are not the only reason to use a mod/con, but if that is what he is basing his decision on, the idea that "it's a house" means nothing. In fact, I'd say in his case the tank water heater is even more likely a better idea that it is with tom, because his house will have multiple zones, so the tank water heater will add buffer capacity as well. There is absolutely NO reason whatever to use a conventional boiler which will run a minimum of 30 degrees hotter, at an efficiency likely no better than a tank water heater (which will approach 80% if gas, but I would probably go for electric water heater or boiler in his case if the load is low enough to say "damn the fuel economics") in a heating application in this situation, and an installed cost vastly higher. Note that is "more likely", not "definitely". Window area and quality will likely make the final decision clear.

The point that a heat plant is determined on a case by case basis is well made. I think we all have systems we are used to and it's easy to throw our products out there as the best solution for every application. Thanks to this site, however, we all get the advantage of a fresh look at products and different perspectives from time to time.
I'll check out that link Dan.
Joe

Ok, perhaps it is the difference of a proper system, or not. A 3,500 sq foot house is not the same as a 1,000 sq ft shop, or a 10,000 sq ft warehouse. The domestic water alone could support the need for a boiler, suppose this fellow has 3 1/2 baths with a roman tub, all proper under California code. Now even with low flow fixtures he is looking for 6gpm plus. Now take the assumption this is a guest house, not that it is in this case but we are looking at a broad swath of California culture (8th largest economic collective in the world were it a country of its own). A modulating boiler, domestic side arm, preheat system with solar, option to dump in to the heat infrastructure of the house, this is a smart system. It surprises me that there is a population that is increasing that elect to do this type of system with a flash water heater. It is not the design path my company supports. Flash water heaters are fiddly devices, prone to calcification, your local boiler technician may not have access or experience with the parts to maintain. With the mod cons the standardization of boiler cores is worth noting, the Munchlkin, Mascot, Baxi, and others are all using the same boiler core supplier. The modest savings in the cost of a proper system VS a water heater just do not pan out when the investment in the home is so large. If it was all about $ why not place oil filled electric heaters in each room. Nope.
We have only been away from $140.00 barrel oil for 6 months, it will return. I think that with the current global climate/economic issues combined with tax credits the smart systems would have a large (6-10 for instance no math just a what if number) amount of solar collectors for the domestic H20 qualifying the entire domestic mechanical side of the system combined with the labor to install 30% federal tax credit, more state incentives in CA. This is where I think a large focus should be applied regarding low temp hydronic heating.
Now the cost of heating that house in 15 years is affordable.
Enough for now.
Dan

what is a "proper system", one that sells the most boilers? using square footage in your comparison only illustrates that you are not actually interested in the needs of the system: square footage in socal is very different than square footage in blue ridge or here in maine. Continuing to ignore that fact doesn't make a case.

Even a 6 GPM domestic demand is very unlikely to cost justify a mod/con, and for large "dump loads" like spa showers and big tubs flash heaters make a lot of sense. I'm not saying it is impossible to use a mod/con here: I'm saying assuming a mod/con is necessary just because the house is 3500 sq ft is a poor criteria, a poor assumption, as is ignoring the utility of cheap water heaters on low load systems. I suppose if you sell mod/cons though, it makes more sense.

Solar is great. the more load you meet with solar, the less your backup efficiency matters. You only make the case for saving money on the backup MORE compelling by adding solar. You don't make a high efficiency boiler more cost effective by reducing the heat load it has to service. You might make your case for solar better though by saving a "modest" $3-6k on the boiler install to apply to your solar system though, eh? That could cut the payback on a modest solar system in half by itself. Wow! Now what's green?

Please tell me you understand these very basic economic tradeoffs? If not, you tell me what a home with a 50kBTU heat load at a 35 degree design temperature, in a climate with 2000 corrected heating degree days needs in terms of BTUs/yr. Then tell me what a house with the exact same load at a zero degree design temp and 7000 corrected heating degree days needs. then, with a straight face, say the upcharge on a mod/con is worth it, when three service calls on your mascot would nearly pay for a complete water heater replacement, and you're talking about maybe a 15% efficiency difference, which sounds like a lot, but 15% of very little is still very little. This holds true even during the last round of price spikes. But hey, if you bet wrong, the beauty is water heaters are cheap. Replace it if the fuel skyrockets, and you didn't lose that much money!

Your attitude sounds a lot like "hey, they must have the money, so they should be convinced to buy better", whether or not it actually makes any sense. I wouldn't begrudge anyone trying to be green for any reason whatsoever: it's a great goal and we do those systems all the time. But I will never tell a client it makes economic sense to install a mod/con on a load that doesn't justify it based on a fuel cost assumption the client is comfortable with. And I take a pretty dim view of people who purport to "design" by shoehorning equipment into homes based on square footage and housing markets.

But I guess that's what's "proper"? I don't decide what is proper. Math does. I just report the results... just the fact, ma'am..

Rob,
Perhaps you missed this portion of my statement that you quoted me on, recommending the old standard first, then perhaps a mod con.
"an old iron boiler will do that and last 30 years, with a modest solar compliment your efficiency will increase, and the system can remain mechanically straight forward"
I consider this a proper heating system, I do not make assumptions as to wealth. But one can assume a build out in So Ca of about $150.-$200. per sq ft minimum. Is it not safe to budget a boiler in here?
For basic information my company has provided over 25 wall mounted (non condensing type) boiler/radiator systems to Habitat for humanity in the Seattle area to prove to the Seattle housing authority that they could get away from water heater/open loop based systems and be cost effective. In there 2004 build out of High point area phase I, We at BlueRidge Company have provided boilers with flat panel radiator systems in over 350 units. In the current build out phase II of 250 units we provide Custom Myson Radiators exclusive to BRC, (another company won the boiler bid).
This is public income housing at it best, see the link. http://www.seattlehousing.org/redevelopment/high-point/
Water heaters are not my choice for residential heat plant, that all. If you like em and speck em that's good.
My broader point is in my opinion solar is where designers should be looking,
The info to support this is not common place. It needs to be.
That's all,
Dan

You are not doing anyone a favor spec'ing a heat source that costs anywhere from 2 to 10 times as much that doesn't actually deliver any additional efficiency. conventional boilers on low load systems fall into this category. Unless you load them up with extra controls, or buffer tanks, they perform very similarly to tank water heaters in average efficiency. Recommending a conventional boiler to someone in soCal is even more ridiculous than recommending a mod/con. That's just choosing the "well, it's not as efficient as a mod/con, but at least it's more expensive than a water heater" option. Not exactly a "sweet spot".

You might get that I don't think very highly of your stance on what is "proper" here. That's because you're not backing it up with any numbers at all. Just, in effect, saying that "because it's a boiler it's better" doesn't fly. That isn't design. That's marketing.

If you actually do deliver any additional efficiency, such as using a buffer tank and mod/con, great, at least you are optimizing for something: but you again don't do anyone any favors by just specifying it without informing your client as to the economic reality of the decision. The math has the truth. use it, inform, and let your client make the decisions that are right for them. let me solve the math problem I posed for you in the previous post. The "SoCal" system (50kBTU output, which is huge, at a 35 degree design temp, 2000 HDD) needs about 68.5 Million BTUs/winter for heating at 100% efficiency. If we can run your mascot at 95% that's about 72 Million, and a water heater running about 80% in a heating application needs about 85 Million, for a difference of 13 Million BTUs/year. Wow, that's a lot: 140 gallons of natural gas. Except, that's something like a 10+ year installed cost payback even with significant natural gas pricing, with NO service on the boiler at all, which would radically outstrip the outright replacement costs of a water heater.

Cut the load in half for a more realistic figure and you can see how much sense it makes. put in a conventional boiler with some controllers at 85% (hopefully buffered) or a tankless at about that level and you can compute separately for those scenarios too, but since the buffer is likely as expensive as your water heater... well, you can imagine how those numbers turn out.

The very basic fact is that you are masking some the economies of going with good building envelopes and alternative energy by refusing to consider cheap and serviceable heat sources where appropriate. That is a real shame. You could do a lot more for solar by doing a pragmatic and reasonable trade down on a heat source that is no longer required to be "the best" because of load reduction and offset with greener methods. I see a lot of people automatically matching solar with mod/cons. All I can say is, that's another panel array that could have been on the roof: is it doing more good in the mechanical room, to dribble out heat occasionally? maybe, maybe not. but again, the math unmasks the truth, not marketing stances.

I have designed several habitat for humanity houses that use water heaters to economically meet their limited DHW and heating loads and kept them from having to go with forced air. I consider that a victory, not because it's hydronic, but because it's more efficient and healthier than the options it displaced, and cheaper than other options that would not have outperformed it by any significant margin, allowing habitat to do more good in the world. But hey, I guess that's not proper either.

90% of the systems out there should use high efficiency sources. But not all of them.

Buffering isn't gonna buy you much with a cast iron or tankless burner here- the slab itself is buffer enough if the burner is remotely right-sized.  A propane-fired tank HW heater is still 2x+ oversized for the design-day load, takes up space, and has relatively high standby losses.  Those standby losses are considerable when the burner is oversized, since the idle time is high, duty cycle low.

Which is why the concept of using a really DUMBED DOWN not-so-modulating point-of-use tankless heater works here.  The raw combustion efficiency is already higher than the tank's, and it can be adjusted to be perfectly right-sized by cranking the burner flame up a tick at a time until it just keeps up on the coldest day. Then you have a low mass mid-efficiency concrete-buffered perfectly "right-sized" bang-bang burner (fires whenever the pump is on) with miniscule cycling losses and even smaller standby losses for under $200.  So WHAT if it won't last as long as hydronic boiler? Keep spares! You can buy 6-7 of 'em for the price of the smallest cast-iron units out there, the installation is simpler (and far simpler than a modulating whole-house tankless) and tuned-to-load it will operate more efficiently.  The installation is as simple as a tank HW heater but 1/2-2/3 the cost (the Marey 5L is$180 delivered from some internet sources), smaller, and more efficient.  It may/may-not last as long as a 40 gallon tank, but even if it only saves $50/year on efficiency (say 83% operational efficiency instead of 75%) it's likely to be a net win. People pay a lot more to jump from 85% efficiencies to 93%, and here you're paying less for more.

But the difference is in operator futz-factor- it's not "set & forget" the way a tank with an oversized burner would be (at least not the first season.)

I wish there were more vendors to choose from in N. America, but it seems Marey has the lock on that somewhat small market. Perhaps the biggest design issue is sizing the pump for the (unspecified) head of the heat exchanger, but it can't be all that much compared to the head of the PEX- it's designed  to work even with 3rd-world low pressure water systems.  I'd be surprised if you needed to go so far as a primary/secondary setup and multiple pumps, as is commonly done with the bigger more sophisticated tankless hydronic-combi systems. 

They're cheap enough I'm almost tempted to experiment some (as if I don't have enough projects in the queue. :-) )  The configurations are endless, but using an electric tank as the low-loss buffer and letting the tank's thermostat to control the pump you could end up with a gas-fired HW heater with an EF in the high 70s to low 80s no matter what the next  version of the EF test looks like.   This would be comparable to or better than an indirect tied to a mid-efficiency boiler, and there are several ways of turning something like that into a low-cost combi for low load/micro-load situations.  But I digress...  back to the slab heating discussion...

Dana, the burner is NOT remotely right sized for the loads we are talking about. the concrete does provide significant buffer capacity as long as you aren't doing any small zones in that house, but if you do a boiler, you aren't getting a small enough boiler to be "right sized".

for a single zone workshop scenario, the tankless is a fine choice, and if you have access to some awesome cheap tankless solution people have actually used in a heating system in america before, then perhaps I'm interested ;)

I'd say something that's 80-85% efficient with a max input of 34K crankable down to 17K or less is just what the doctor ordered for the 13k design day load shop building here (which IS what we're talking about, or have I had a "senior moment" thinkin' this was a different thread/building-configuration?)  

Sure, in a multi- micro-zoned house with a whole house design day load of 13K, you're absolutely right- buffer it (possibly with an electric HW tank :-) ) but the duty cycle of the burner and standby losses still count at the end of the day, and a low mass 17K-in 80% CE burner will outperform a 30K-in 78% CE burner by measurable amounts- even more so if the 30K burner is buffered by a tank as lossy as a gas HW heate tank.  (Continuously drafting flue losses and the insulation gap at the bottom really eats into system efficiency as the duty cycle drops, and here you're looking at a ~55% duty cycle on design day.)

Tiny tankless point of use heaters are just SO "Keep-It-Simple-Stupid" by design (well evolved and ancient- they've been around since the '20s or '30s) that the reliability has to be higher than all the bigger-deal electronically controlled tankless heaters out there.  With a minimum of feedback (a mechanical overtemp safety- thatzit) and manual flame adjustment only via big fat knobs on the front panel, the weak point is probably how fast scale builds up on the water side of the heat exchanger (which pretty much goes away in a heating loop), and the igntion system (unknown in the Marey case- most of the Euro-zone ones I've used had standing pilots.)  I haven't seen one used in a heating system (yet), but it may be more appropriate & efficient than say, dialing down a Takagi Jr. to self-limit at it's lowest modulation level & buffering it, since you won't be able to take advantage of the modulation function at such low loads.  If I had a low-load shop building like this to experiment on, I'm pretty sure I would.

And worst-case, I'd rip it out and replace it with a tank if it became a pain in the butt- it's only $180, eh? ;-)  But like you, the LAST thing I'd do is buy a mod-con or even a 2-plate cast iron beast for this place!  (Thermal air panels, drainback solar slab, with pellet-stove backup, anyone? All seem well suited to NC's climate, latitude, & vegetation zone.)

BTW: I'd be curious to see/read the designs for your Habitat combi systems, if they exist in electronic format (backchannel via email svp.)  Sounds like they went through a real cost/benefit vetting process to get there.

well we were really arguing over the 3500 socal house.

I don't see any arguement against the tank heater in the workshop which will pick up any heat lost by the tank except some up the stack.. which.. again, by the study you posted yourself... apparently doesn't account for more than a few percent of efficiency compared to a tankless. the tank is about 80% efficient by those studied numbers you posted. I still don't understand why you're into measurements, then ignore the ones you found? for point of use heaters, you know I never bothered to really look at their outputs... but they probably are adequate here. I can't imagine what the problem would be using one... would be fun to see if anyone has before.

I'm not sure I can share my habitat design... they were for (shudder) baseboard, I'm not sure I dare let anyone else have a copy lol... Water heaters AND baseboard? I'd be thrown out of the RPA ;)