Oh.. the soCal house (my bad...)  Fer sher he needs a buffer!

The eKoComfort document http://dsp-psd.pwgsc.gc.ca/collection_2007/cmhc-schl/nh18-22/NH18-22-106-108E.pdf

On page 3:

79-80% raw combustion efficiency (measured) for the tank- they found a pretty good one from a CE point of view.  Forced draft units have higher turbulence on the combustion side of the heat exchanger, which buys you more efficient heat exchange and lower flue losses in standby mode- atmospheric drafted ones tend to top out at slightly lower CE and have measurably bigger standby losses.  You can infer some of that from the difference in how rapidly the forced-draft vs. atmospheric tanks fall off under low use profiles in this summary:

http://www.aceee.org/conf/08whforum/presentations/1a_davis.pdf

...but it's a bit distorted by the fact that the only forced-draft version is also electronic ignition, and all atmospheric units were standing pilot. (I wish they'd tested a standard 82EF tankless in their mix, eh?)

But that's 79-80% CE vs. :

83-86% raw combustion efficiency (measured) for the tankless Rinnai

But that efficiency measurement doesn't account for cycling losses due to burner size and modulation for load.  CE is a limiting factor, but it's not the only factor.  The seasonal efficiency will be less for both, but more dramatically less for the oversized burner.  For estimating real-world losses vs. AFUE type testing you can take 3-10% off of CE for the perfectly sized burner (at which point it'll come seasonally close to AFUE), but 10-15% off CE for a 100% oversized burner (where it'll fall well below it's AFUE numbers), and that's assuming we're comparing apples-to-apples medium-mass burners. 

In the tank vs. tankless scenario, standby losses of tankless systems are extremley low compared to tanks due to the constant heat drain from the buffered water straight out the flue.  Even in a perfectly BTU-output matched tankless vs. tank scenario, even at identical operation combustion efficiencies, the tankless would still win the seasonal efficency game:

Page 4:

"Overall thermal efficiency was calculated for various time intervals during
the project by summing the space and water heating energy outputs from
each system and dividing the total by the energy input. The performance
of the two systems was similar during periods with high loads, with the
overall performance of both systems falling with reduced system loading.
However, the efficiency of the tank-based system fell faster than the
efficiency of the tankless system as the load was reduced. This may
attributed to the constant standby losses from the storage tank, which
has a greater impact on system efficiency when the loads are low."

(emphasis mine,of course)

Standby losses add up- the greater the standby duty cycle, the worse it starts to look.  Most of the standby losses on tanks are flue losses (which really do dissapear), but the heat leak out the bottom is only useful to you during the actual heating season- on the shoulder seasons much of it can be considered a straight loss, where fuel consumption is concerned.

Page 5:

"Based on analysis of the monitored data, and the prevailing gas and
electricity rates during the monitoring project, over one year the
tankless combo system in Unit 25 consumed 3,041 m3 natural gas
(107,352 cu. ft.), at a cost of $1,446. To provide the same loads, the
electric alternative would have used 23,807 kWh of electrical energy
at a cost of $2,324. The gas-fired combo system in Unit 25 saved
$878 per year, or 38 per cent, compared with the electric alternative.
The tank-based combo system in Unit 30 consumed 3,381 m3 natural
gas (119,349 cu. ft.) over the year at a cost of $1,608. To provide the
same loads, the electric alternative would have used 25,368 kWh of
electrical energy at a cost of $2,476. The gas-fired combo system in
Unit 30 saved $868 per year or 35 per cent compared with the electric
alternative."

So in roughly similar heat loads the tankless burned up $1,446 in gas during the same heating season that the tank system ran up $1,608.  That's greater than a 10% difference in annual fuel savings, which is on the order of what I'd expect for the NC shop, but it may be even worse, since the eKoComfort experiment tank's burner may have been better matched to the peak load than 100% oversized the way.  One could argue that the differences in occupancy between the units made a difference (1 person for the tankless combi, 1-3 for the tank-combi), I'm sure it did, but it's hard to argue that the entire 10% delta came from that factor.  Fact is, we just don't know.  For all we know the folks in one unit slept with the windows open, the others didn't.

But we do know that the measured combusion efficiency is consistently higher (and only measured while burning) than the tank, by an average of 4-5%, and that the duty cycle of a perfectly sized bang-bang burner will be twice that of a 2x oversized unit, so give it another 2-3% just on lower cycling losses, then kick in another 2-4% the shoulder season standby losses of the tanks temperature-maintenance burns.   (There's a reason why the EF falls below 50% when you're using less than 60 gallons a day, eh? )   In other measurements (can't find the web-reference right off), but just maintaining  temperature (not using the water) on a typical tank will run 8-12 therms/month.  Assuming you need heat only 25% of the time for the first & last months of the heating season, at least 15 gallons of propane a year goes up the flue when you don't need heat (but you might next week so you leave it running.)  The eKoComfort combi system at least saw the DHW load(!), so it was probably never WORSE than about 55-60% efficient in the shoulder season.

Fuel savings on the order of ~10% is about what I would expect out of going tankless on the heating system for the NC shop.  Even if he only burns 200 gallons/year, at the current ~$2.50/gallon that 20 gallons/year is going to add up to the price of a replacement heater in under 5 years. (And if you think it's gonna STAY $2.50/gal for the next 5 years you're more optimistic than most.)  If he expects to burn over 500 gallons/year (probably does) it's worth giving it a shot.

The NC shop scenario is much worse than the eKoComfort experiment in the shoulder seasons- doubt anybody will be takin' long showers in the shop sink (or at least I HOPE not! :-) )  The Marey 5L really makes sense here, if only for the sink.  If wintertime water temps are too low for the 5L to keep up, running a 50' loop of PEX along the edge of the heated slab to a 5 gallon buffer (insulated or not) on the input side of the point of use heater will surely provide enough tempering for the over the sink heater to keep up unless you're washing the mud off your truck with it (at which point you probably don't care if it's not 100F+ water.)

Gee, I sure am enjoying 'listening' to the conversation. My only problem is that I can't keep track of whether you're talking about my little challenge or the 3500sqft SoCal house.  :-)

I'm getting some good tips regardless.

Dana, I looked at the Marey and it's certainly tough to beat the price - a point of use gas tankless for less than a cheapo electric HW heater...  Coming from a school of hard knocks where there ain't no such thing as a free lunch, I have to wonder... but, as you say, if it doesn't work, you're only down <$200 bucks (well, there is all the effort in installing it, etc). 

Seriously, my main concern about the Marey, however:  It seems as if the unit requires room ventilation to the outside (floor and ceiling recommended) in addition to the combustion air supply/exhaust ventilation.  That now becomes a problem.  Suddenly my tight workshop with 13K heat load is now very leaky with a much higher heat load - hey, maybe this means I can switch to a bigger heater!  :0  Sorry, I had to say it.

Back to my concern, am I reading the installation instructions incorrectly?  Is this additional ventilation required for the Marey?

Tom
BTW, I'm looking at the Marey as my point of use DHW heater, not the space heater.  Sorry for any confusion.

I should have mentioned, I was looking at the Marey as my point of use HW heater, not my space heating source.....

Tom

Dana, that's a pretty good discussion, but on what planet is $250 going to buy a new tankless heater, or am I missing what you are saying?

As loads drop further and further (and here, you might have a point on this) the efficiency delta does grow, but then it always falls back to the fallacy of looking at efficiency percentages vs actual energy use/cost. sure, when the load is ridiculously tiny, the tankless has an edge: so what? that's not much of a difference in actual energy usage anymore. When you're actually using most of your energy, the tank and tankless are very similar, the tank is likely to last longer, have less electrical load in the meantime and be a cheaper initial install to boot, and a cheaper replacement when it does actually fail, and much less of a maintenance issue.

I STILL don't see an arguement for the takagi, rinnai, or any other tankless heater vs a tank in a low load situation in light of that. winning the efficiency battle by a little more than we are at 1k/hr or less, well hey, that's just not thrilling me ;)

Now if a cheap little tankless does the trick, so be it. That would be a worthwhile and cheap experiement for sure.

Rob is right again. Economy-of-scale makes the tank water heater hard to beat, reliable, safe, cheap and prolonged life when used as a dedicated boiler (not of 'open' systems). I installed many of them when working in Albuquerque, NM.

This from the Mod/Con zealot, even I can't believe I said that.

I don't understand the fascination for low efficiency tankless water heaters. The DIY crowd would be better off with tank water heaters for many applications, but good marketing dictates otherwise.

On what planet can you buy a point of use tankless for $250?  Planet 'merica, of course!:

http://3tankless-water-heaters.ecrater.com/product.php?pid=2897078

http://cgi.ebay.com/Marey-LPG-Propane-Gas-5L-Tankless-HOT-Water-Heaters!_W0QQitemZ180326045425QQcmdZViewItem

http://camping-tankless-water-heaters.ecrater.com/product.php?pid=2898958

(Need I go on? google it- $180 is the going price, not a remaindered discount.  If you pay $250 you've paid too much.)

Already did the math for ya on the cost/benefit of the efficiency.  At a mere 10% improvement in annual fuel economy (which you'll almost certainly get, by taking down the standby losses more than an order of magnitude) it's cost-effective at current propane pricing, even if you have to fully replace it every 3-4 years (as opposed to 10 for the tank).  And I'd be surprised if you can't get at least 5 years out of it- maybe 10, since the heat exchanger won't be clogging with scale on the water side, and the number of thermal cycles will be FAR lower than in a kitchen sink/hand washing application.

We agree completely that the bigger-deal more expensive fully modulating whole house beasts like Rinnai or Takagi etc. makes any sense here. They're too much burner, too expensive and too complicated.  And that's not to mention I suspect that at the lowest modulated fire a whole-house 125-200kbtu/h tankless will have lower raw combustion efficiency than when running 30- 80kbtu/h.  A Rinnai running at 17k probably has lower combustion efficiency than the cheap Marey running at 17k 'cuz the turbulence on the fire side of the heat exchanger is starting to go laminar-flow on ya with the bigger heat-exchanger.

The Marey 5L  (and all of the Euro versions not readily available here) is/are the very definition of "cheap little tankless", with the an ideal output range for sub-20kbtu/h heating loads. My primary concern in a heating application is the reliablity & longevity of the ignition system. (I suspect you could retrofit a 2watt 3V "wall wort" DC power supply to supplant the D-cells if the pulse ignition runs continously while firing, which would drain a D-cell pretty fast.)

alright then, our violent agreement is complete ;) that's one cheap heater. thanks for the patience with my denseness Dana!

The burner on the Marey 5L at maximum fire is no more than a typical 30-40gallon tank heater, The only difference is it only runs when you absolutely need the hot water.  The Chinese-English sounding instructions are for an installation where the exhaust is vented directly through the room (typical in tropical regions, I s'pose), but you'd be better off in multiple ways if you vent the exhaust with (DUH!) metal flue suitable for tank or tankless HW heaters.  In warm climates (even not super warm, like southern Japan) it's common to mount tankless HW heaters on the outside of buildings, but for freeze-contro reasons you probably don't want to do that here.

I'd have to search them out again, but there used to be several heat-activated flue dampers on the market that work well with HW heaters.  They use bi-metal technology operated vanes in the exhaust path that cause it to open fully when the flue temp got above ~110F then close again as it cools.  In this case "closed" really means "constricted"- I have one on my antique tankless in the basement (that will be retired soon).  It takes a 5" fully open flue (20 square inches) down to about 2" (3 square inches) when closed to keep exhaust spillage under control on startup.  (I've tested mine with a smoke pencil every year since installation- there's never been detectable spillage even on dead-cold-starts.)  As long as the flue is tall enough it'll draft just fine, but if you find after testing there is some spillage you can add a section on top or use a venturi type flue cap (Empire Syphon vent or Eveco, etc.) to boost drafting performance.

The instruction manual is for all models: http://www.tanklesshq.com/marey-gas-owners-manual.pdf

... so the size of the exhaust & ventilation area is unduly conservative for the smallest version (the biggest Mareys are 3x the burner of the 5L- nearly as big as smaller whole-house Takagis & Rinnais)  The vent spec for 0.06 square meters (90 square inches) can probably be safely cut by 2/3 to ~30 square inches (a 5.5"x 5.5" square, or a 6" round.) Look up what the National Fuel Gas code requires for atmopheric-drafted HW heaters or boilers in terms of square-inches/BTU rating- that's what you'll need for ANY burner you put inside unless you go with a sealed-combustion unit.  There are very low-pressure operated vent flaps available that people in ultra-tight houses use in boiler or laundry areas to provide combustion air and prevent back-drafting of atmospheric burners when exhaust fans or dryers are running (your dryer sucks in a LOT of infiltration air!)  You can probably use one of those and it'll preserve the tightness considerably.  They're cheap- (IIRC sub $50) and you'll need it anyway if the place is that tight.

Check out the ventilation requirements for a typical atmospheric drafted tank heater:

http://www.hotwater.com/lit/im/res_gas/186487-001.pdf

Applying those requirements using a presumed 35000BTU burner should suffice for the Marey.

BTW: Any flues & vent caps etc. should probably be listed stainless flavors, not cheapo vents. Propane exhaust is slightly acidic. If you add a small vent hood for dilution air at the top of the Marey it'll raise the dew point to reduce the liklihood of flue condensation (it may/may-not be needed, but not having the beast in front of me it's hard to tell.)

OK, I'm back.
If I am not mistaken the line of tankless you are talking about is a Chinese brand, I brought some samples of the same units in several years ago (a side bar to that story is they ended up in New Orleans at a friends sons Katrina camp, They had the only hot water showers, The female camp near by was great full for the occasional use of the shower).....
Onward the units at that time were not CSA certifiable, there were a couple issues. They were beautiful in design, shipped with a shower head and a couple D batteries. They had a safety feature..... They locked out after 15 minuets to prevent carbon dioxide poisoning. In Asia as in Mexico many homes are sold with out hot water system, the DYI group fails to understand proper venting in these country's and are regularly killing themselves with poison gas, thus the lock out.
These units are cheep in China, go figure..... about $75. per unit
so the question is has anyone handled the Marey unit? Run time? lock out?
Proper approval for US market, sure is easy to buy but does it meets NSF, CSA certs, That is a bout a $20,000.00 task for the distributor in USA.
To the Takagi, Rinnai, I know the head numbers in a Rinnai are huge, better do the math or she wont work.
Now last thought on the residential water heater argument, again water heaters are not my favorite choice as heat plant but we sell a TACO X Block, these are very cool device, don't know if you all have paid much attention to them but they can transfer 50,000BTU, will modulate based on out door temp, Isolate the domestic H20 from the heat side, completely non ferrous, something to keep in mind, TACO is also designing the X Block for solar applications, coming soon.
Then there is the o'l combi core.
enjoy the day,
Dan

No problema mi amigo- when density meets obtuseness all sorts of good stuff gets hammered out.   

(I find you a very good sounding board for my flights of design fancy- you & Morgan keep me honest by making me document it!)

I'm not surprised that tiny dumbed down point-of-use tankless HW heaters weren't big on  your radar screen- they're just not the 'merican paradigm the way they are in Europe.  In Europe many are small enough to be vented into the room. Some have manual pushbutton igntion strikers, others have standing pilots, but I haven't seen/noticed any with the battery operated pulse ignition like the Marey (but I'm sure they're around.)  But when I read "13K design day heat load" it was the first thing that came to mind (or maybe the second- I first thought maybe a 8-10 gallon RV  tank heater would be about right, then I priced 'em and looked at their specs: Too pricey and not enough burner on most of 'em.)

Dan the Blueridge man writes:

"They were beautiful in design, shipped with a shower head and a couple D batteries. They had a safety feature..... They locked out after 15 minuets to prevent carbon dioxide poisoning."

If Marey is OEMing a standard Chinese unit (and documentation sure has that FEEL), that would be a very good feature to know about before moving ahead on it as a slab heater.  (Thanks Dan!) 

I just shot off an email to Marey to find out more...

A feature like that is most likely something one can disable, but that's not necessarily the right thing to do.  No matter what you'd want a carbon monoxide detector in a small building that has any type of propane burner inside the envelope's pressure boundary but still...

It would be nice if some of the Euro versions were available here, eh?

Gentelmen,
As long as I have your attentions again, here is another bit of carbon monoxide information and the reasons to mandate detectors on any job. I was speaking with one of the upper level folks LAARS the other day. Topic was carbon monoxide deaths. Seems a Baxi HT 330, Mascot HT 330, (Same unit, private label) and a Munchkin all had vent failures.

Results; Baxi I think Main but don't quote me. Cause again as I understand but don't quote me with a longer/maximum coaxial vent application propane fuel, Pop on ignition parted pipe, 4 people died.

Results; Mascot I think Vermont but don't quote me. Cause again as I understand but don't quote me with a longer/maximum coaxial vent application propane fuel, Pop on ignition parted pipe, carbon monoxide detector went off, no harm.

Results; Munchkin I think Colorado but don't quote me. Cause again as I understand but don't quote me with a longer plastic PVC vent application not sure about fuel, Pop on ignition parted pipe, people died.

The feeling at the end of the conversation was that one day soon we may see mandated stainless steel pipe or at least a redesign on required fastening of coaxial pipe industry wide. Something to be aware of, not standard news.

Dan

Ok, so it was annoying me enough that I dusted off my rusty command of Dutch & went surfin'- yep they're out there! Vaillant makes a few models in the sub-10kw range, as does Bosch and half a dozen other vendors you'll never see here.

The only point-of-use type tankless that Bosch sells in the US is the 1000P. http://www.bosch-store.us/images_templ/1000p-specs.pdf It has a mechanical feedback modulation on water temp, but a simple burner front panel slide valve to set the flame level as well. It's LOWEST modulation only goes down to about 30Kbtu (about what a tank HW heater uses), so it's still nearly 2x oversized for the load, the rated thermal efficiency is only 78%, about the same as a atmospheric drafted tank heat exchangers. The propane version is probably actually ~80-82% due to the physics of the higher CO2 content in the exhaust, but they don't spec it separately so don't count on it.

Still, the standby & cycling losses will still be lower than a tank in a heating application- and you'll waste ~4 gallons/month in the shoulder seasons (due to the standing pilot) instead of ~10 gallons with a tank. The price is comparable to a 30-40 gallon tank heater though ($350) (see http://www.bosch-store.us/1000p_aquastar_tankless_water_heaters_gas_tankless_water_heaters_47_prd1.htm )

It'll beat a cheap tank by more than 5% on annual fuel economy, but probably not 10%. It's twice the price, twice the burner, but probably 3x the quality of the Marey- I'd expect that heat exchanger to last quite a bit longer (warranteed for 12 years in water heating apps, but since it's less likely scale up on you it may do better in a heating app.)

And of course like Rinnai, et al, they write all over the documentation "not suitable for space heating" (like THAT ever stopped anybody. :-) )

"Now last thought on the residential water heater argument, again water heaters are not my favorite choice as heat plant but we sell a TACO X Block, these are very cool device, don't know if you all have paid much attention to them but they can transfer 50,000BTU, will modulate based on out door temp, Isolate the domestic H20 from the heat side, completely non ferrous, something to keep in mind, TACO is also designing the X Block for solar applications, coming soon.
"

Dan,
This is one cool device.  Cool price too...  Approximately 4X the cost of my heat source. 

It does make life easy for attaching solar collectors though...  Not sure what changes they're planning for solar application but definitely worth watching.  Thanks for telling us about it.

Tom

that's not a bad item: it does a lot for a combination water heater system.

you don't need it with a dedicated water heater system though, were a basic iValve would do if reset were helpful.

Back to ModCons. I was out on a Buderus 142-60 (twice too big for the load) on loud and inconsistant ignition call . Found 188 equivelent feet of 3" PVC (60 max. allowed). Homeowner said it worked better with the cover off. This installed by a well known "professional" in the area. Also found suspended tube with hardwood floor and 1/2" Thermax under. No design help here.

SS vent will not fix poor combustion or illiteracy. Neglect i.e. no professional, annual, clean and check, will kill people and the ModCon industry.

Factory incentives (not boots and jackets) for boiler-specific certification is the answer. They can't be made idiot proof, they can't even be made plumber-proof, they have to be installed properly and maintained.

I give my good customers a free CO detector and install before I leave as many will not call for service until it gets cold in the house.

I feel better now.

I spoke (wrote) too soon- it seems one of the tiny Bosch units is available from the far off NAFTA land of Canuckistan (or whatever they call that vast wilderness north of the US border. :-) ):

ASTRAVAN DISTRIBUTORS, LTD.
123 Charles Street
North Vancouver, B.C. V7H 1S1
Phone Canada: (604) 929-5488
Phone USA: (206) 860-8448
http://www.astravan.com/

The specs for the Bosch W125 live here:

http://www.astravan.com/W125.pdf

You'd have to call for price, but it should be pretty cheap- the street price is about €200 in the Netherlands, which is about $250USD. see http://www.warmteservice.nl/merk/196/geisers/keukengeiser/bosch.do

Throttled back to the lowest setting the input is ~20K, so at ~80% combustion efficiency it'll deliver ~16K, which is "close enough" of a match to the 13k load to run at/near max efficiency.

On most of these dumber-than-a-box-of-rocks type tankless burners, for a heating app crank the water temp DOWN- it's usually just a flow restrictor, thatzit! Turning the temp up just adds more head for to the pump to overcome, doesn't modulate the flame. On this unit just set the burner setting at the bottom and forget about it- it'll keep up. It's probably a better choice than the Marey for the over-the-sink unit too. If it's only $100 more (shipped), I'd go for it. Bosch has been making the really dumb heaters like this for over half a century (Vaillant has been at for MORE than a century- since 1905 I read on a German website!). It should be pretty reliable- there's just not enough complexity to mess it up. There's probably on the order of a million W125s in service in Europe- heaters like these are as much a commodity there as the 40gallon tank is here, only more so: many homes have two- one for the kitchen, another for the bath/laundry.

Badger, You are exactly right with the "free" CO detector. I have installed a lot of code and inspected systems over the years as well, starting in the mid 1980's in timber frames we built, and feel that this not being required in a world of bureaucratic nonsense requirements is clearly an oversight. Good for you for catching this in your own world early. I am rethinking this in our current business practice.

I would be cautious as installer messing with the installed features on any box of rocks boiler, water heater or otherwise. I hear liability,liability, liability.

Tom, Just showing the X Block as a matter of info, it has a place,

Dan

I am so bored with the on demand discussion. The takagi serves no niche worth serving.

Dana: a thought on the bosch. that unit is built to operate for periodic hand washing, maybe a shower or two. Do you really think it would hold up to continuous duty use?