Help needed…  Friend is getting NG down his street and is looking to convert his appliances over from LP(current fuel) to NG. Right now he is using an outdoor wood boiler to heat the following loads connected in series….

1. DHW in the winter via a pipe-in-pipe sidearm heat exchanger external to a propane fired atmospheric, tank HWH.. the DHW circulation through the HX is driven only by convection).
2. House heating via a coil installed in the plenum above his propane furnace.
3. Well insulated shop via a small unit heater (coil/fan combination).

Combined space design heat load is around 100 MBH…. DWH would be in addition and of course very “peaky”… due to family size and usage.

Converting the LP clothes drier and stove (cooking) won’t be a problem should be easy and relatively cheap parts. Converting the furnace isn’t a high priority as they haven’t used the burner in some years and have been fine relying on the heat from the slab coil fed by the outdoor wood boiler loop.

The current LP tank HWH is only about 2 years old and apparently a conversion (to NG) isn’t possible or too expensive so he was considering a new NG water heater. I thought it would be an ideal time to try to come up with a way to install a water heater and connect it in such a way as to allow it to essentially work as backup for the woodboiler, if they should run out of wood, were out of town, or just not feel like loading the stove, etc. If the water heater could be setup to do this, it would provide a natural gas backup to the wood supplied heat to the house and shop (further reducing any felt need for converting the current propane furnace).

The simplest way I could come up with was to disconnect the propane from the existing water heater, but leave it piped as is… then install a NG fired on-demand water heater in a side-stream or “sidecar” arrangement. On-demand water heater would take from somewhere near the bottom of the tank and returning via one of the top connections. A stainless-steel or potable rated circ pump would be set to circulate the water between the tank and the on-demand heater. My idea is that this sort of setup would allow the wood boiler continue to heat as it always has, but during high DHW usage periods or if the wood boiler is running out of wood or out of wood… the temperature of the water in the existing tank water heater falls below an adjustable setpoint and the circ pump would turn on and the on-demand heater would kick in.

Incremental cost is an on demand water heater(associated venting, gas piping, and electrica), potable rated circ pump (and associated electrical and basic controls - aqua stat / temp sensor), and some piping to connect the new water heater and circ pump.

My questions are:

1. Does this concept sound feasible?

2. Is there a better way to do this for same or cheaper cost?

3. Can the existing “convection driven side-arm HX” transfer heat “backwards” to heat the wood boiler loop ? If so will it be enough to heat the house or shop or both? Would adding a pump to force that loop make up the difference?

4. Existing wood boiler loop runs around 180F so I would need an on-demand heater that can produces that right?

5. I’ve heard that some on-demand heaters don’t like warm/hot incoming water. Seems like in the arrangement proposed, temps could be as high as 150 or 160F entering the on-demand heater. Do I need to look at certain specific brands/models?

Any tips or advice would be appreciated.

you do not want to constantly run your on demand at high temperatures. you have two issues here, DHW and backup heat.

-As for the DHW, it would be pretty hard to beat the price of a new sealed combustion NG water heater. and unless he's really good with wood year round, you do want sealed combustion. the natural gas utility may even offer rebates or discounts for these kinds of conversions. All that said, if he really keeps up on the wood, it may not be cost effective to switch out this water heater... what's his current propane bill?

-As for backup heating, the real question is how often does he fail to heat with wood? That will determine what makes the most sense for backup heat. a cheap on demand can be an ok choice IF you can run lower water temps and IF you don't need to use it much. if his heating system actually REQUIRES 180, it won't. however hydroair can usually be run for 140 or lower, depending on how it was sized etc. so chances are good: heat load calculation is required to determine for real. It may make more sense to back up with a cheap electric boiler if higher temps are required and it's not necessary most of the time... and he has the installed electrical capacity to run it.

You say he has a 100kBTU load. is this house uninsulated, or over 5000 square feet? If not I suspect your load calc may be high.

it's POSSIBLE to use the DHW heat exchanger in reverse IF it's got the rated capacity you need at the water temperatures we are dealing with. but that's unlikely with DHW hardware and higher heating system water temps.

Hey NRTRob, chopping wood not only warms you twice: http://newsfeed.time.com/2013/08/26/chopping-wood-a-manlier-feel-than-sports/

absolutely. one of the only chores in life I love is splitting and stacking wood.

Just because the wood boiler loop runs at 180F doesn't meet the on-demand HW heater needs to. It's a matter of what temp the coil or radiation (and thus thermal store) temps need to be to deliver the heat, and it's unlikely (though remotely possible) that it needs to be 180F. Unlike a wood boiler, the on-demand can be set up to modulate.

Running a condensing on-demand at 180F out would be a waste unless the return water was 120F or less, and it may degrade the unit or compromise the plastic venting to run it with a higher return temp than 130F. But there's little harm in running a 82-86% efficiency on-demand at 180F output. Rather than high output temp, the bigger problems to attend to when tankless water heaters in space heating application are over-pumping (higher flow than the water heater tolerates over the long term) and over-firing (running it only at or near it's maximum firing range.) A 4gpm flow and a 50F delta-T gets you the 100KBTU/hr you're looking for and the water heater would be happy. But while 8 gpm @ 25F delta-T would deliver the 100K of heat, a constant 8 gpm wreaks significant wear & tear on the unit- you'd be going though flow sensors like crazy.

The tankless running my own kludgey combi heating system is set up at ~2 gpm flow through the tankless and ~158F out, serving a 48 gallon buffer tank with an internal potable water coil, with the a single aquastat set to kick on when the bottom third of the tank hits 122F, off at 130F. With all heating zones running, including the "hail mary" hydro-air coil second stage, the return water to the tankless can drops to ~117F, and when someone is taking an endless shower it can drop as low as 105F (it would drop lower were it not for a drainwater heat exchanger feeding hot water coil & shower), but that's about it. With the burner on and pump running the water at the top of the tank headed out to the radiation is ~15-20F hotter than the return water being drawn to the water heater. It was a hack, but the math was done up-front on both the space heating and hot water loads. It keeps up with both, and operates within the anticipated margin of error. YMMV.

In my system the water heater (Takagi KD20) is plumbed as a boiler, running only heating-system water through the heat exchanger, not potable, using a standard Taco circulation pump on the "boiler" loop. The heat exchanger for the potable hot water is inside the buffer tank (Ergomax E44). Both the water heater and buffer were acquired at well below wholesale from surplus equipment vendors. At full retail I probably would have done something else. I consider the water heater a temporary solution (if it makes it through 10 seasons I'll be thrilled) but it seems to be hanging in there with fairly minimal maintenance with 4 full heating seasons behind it. But neither the flow rate nor the firing rate never exceed ~1/3 the unit's max ratings.

I could back off to 140F out in the summers if I wanted to, bumping up the flow to ~3gpm (so the thing runs stably when the aquastat is nearly satsified) but there's no efficiency to be gained by that, and inevitably sometime around the beginning of February the hot water performance during space heating calls would be iffy. It can take the higher delta-T just fine- they're designed to handle 80F or higher deltas (35F in, 115F out), and I'm never exceeding ~55F.

Bottom line, sure, you could do it, but do both the hydronic design math and the financial math before diving in. Whether it's the best way to back up the wood boiler and provide hot water is an open question. Gas fired on-demand hot water heaters make better modulating boilers than hot water heaters, IMHO, but the vendors may tell you otherwise- some void or severely shorten the warranty when used in space heating applications, but if set up conservatively there's no reason to believe they won't have a reasonable service life. (I'll report back when mine croaks, if I don't replace it before then for other reasons.)

keep in mind dana that on demands are not typically expected to run anywhere nearly as much as a boiler. that's why the MFGs nix their warrantees... in space heat, they are going to run at least an order of magnitude more gas through them a year than they would in a DHW only application. If they test by expected cycles... those two applications are nothing alike.

There are a couple of aspects to it beyond the raw volume of gas run through it, not the least of which is the total number of burn/temperature cycles, and both average & peak firing rate. A typical DHW setup sees an order of magnitude or more burn cycles than a buffered heating system, and if designed to a realistic somewhat modest heat load the space heating app sees much lower peak firing rates, and much lower thermal stresses overall, even at higher than DHW output temps.

Of course it won't last as long as a boiler, but if you baby it by-design it's not going to crap out nearly as quickly as the shortened (or voided) warranties imply. There are plenty of examples of abusive over-pumped and over-fired radiant slab systems based on these things being promoted over the web, and I understand why the manufacturers won't sign up for the standard warranty in a space heating application. I'm not complaining about the warranty issue, only running it up the flagpole as a warning for those who think they want to walk this path. Just because I did it that way doesn't mean I'm advocating as the real solution here- far from it. Rather, I'm trying to point out the potential pitfalls of using an on-demand as the backup, and how to not trash the thing in short years. Rinnai will let you use them as space heaters, but only in conjunction with their own air-handlers & controls, and not all models apply. Noritz will only warranty space heating apps with their ASME rated units. Takagi & Navien are a bit more liberal, and often the units chosen by web-vendors selling junky abusive pre-packaged systems.

No matter which manufacturer or model, setting it up to run 180F out @ 100 KBTU/hr just because that's the way the wood boiler was set up would be a mistake. Like any other heating system, starting out with a more careful calculation of the real heat load and water temp requirements is the first priority. Then if taking this approach, setting the flows & temperatures conservatively such that it'll actually modulate over a range at reasonable delta-Ts & flows gives the on-demand a fighting chance of survival, delivering reasonable efficiency for what it is.

And that's the long answer to:

"1. Does this concept sound feasible? "

and

"4. Existing wood boiler loop runs around 180F so I would need an on-demand heater that can produces that right? "

The answer to 3 is probably not, but we don't have enough information to make that call. On #2 is also insufficient information.

On #5, it depends- some early Takagi units had that issue, later versions (TK-JR or TK3 or later) not so much. Don't know about the others- it's a moving target. I can believe many would have controls that interpreted 150F at the input as a fault condition.

yes, I was using gas volume as a proxy for run time/cycles etc. I do not believe you are correct with your DHW analysis vs buffered heating system though. even a small domestic demand is usually above min mod for an on demand and total DHW usage isn't even within an order of magnitude of space heat in a mild climate on a bufferred tank, by cycles or by btus.

Total DHW BTUs are well-within  an order of magnitude of space heating BTU in any US climate (save maybe central AK). On a US-wide basis space heating BTUs are only about 2.5x the annual DHW BTU load, according to D.O.E. data, as summarized in this pie-chart. (Source: 2010 Buildings Energy Data Book, Table 2.1.6) The EIA puts it at about 2.35x and the ratio has been steadily shifting toward a smaller multiplier over time, partly due to better weatherization & codes, partly due to populations moving south. In climate zones 5 & 6 the multiplier will run about 3.5-4x, but still nowhere near 10x.  In climate zones 2 - 3 the space heating consumes only 1-2x the BTUs applied to heating hot water, in zone 1 it's at rough parity, with hot water being 1-1.5x ahead of space heating.

On a cycle per day basis, monitored water use profiles in a 2003-04 CA Building America study he average was over 20 draws per day, and in homes with clothes washers designed with intermittent draw filling it it was over 30x/day. (Couldn't find the original data online anymore, only  a few refrenced tables included in a HomeEnergy mag article.)  With the 48 gallon buffer my system the on-demand typically cycles 3 sometimes 4 or even 5 times per day from  mid-May to mid-September, about 30x/day during the middle of the heating season. (Other families with more people and different bathing schedules might see more cycles than us in DHW only mode- we tend toward serial-showering in the AM, with the buffer not fully recovering between showers.) During periods when primarily only the lossiest radiant zone is calling for heat there are times when the zone calls don't overlap sufficiently it'll 2-4x/hour for awhile, but those conditions never last all day & night.  On average it may be on par with or somewhat more than the total annual cycles of a tankless DHW application, but nothing like an order of magnitude more cycles, despite the micro-zoning.  I haven't put a datalogger on it, but I've monitored the system behavior under different load conditions. Put some error bars on it if you like, but I'm not under-shooting by half, might even be overestimating the true summertime averages (not that it would skew the annual numbers much.)

your pie chart includes climates I would term even more than "mild". I guess I should be clearer: I am referring to places that are more heating than cooling only. when their usage is not diluted by the southernmost states, it's not even close. but you're right... order of magnitude is an exaggeration on a BTU basis.

I don't think you can fairly use your rube goldberg bufferred contraption as representative of heating and DHW demands. that eliminates a LOT of space heating cycles during shoulder seasons when apparent load is below min mod. In a more typical application I will stand by my statement that on cycling at least you are well over an order of magnitude increase for a modulating on demand heater in a space heat application. But I will acknowledge that is not a scientific analysis, for sure.

Without knowing exactly what we're dealing with on the wood-boiler store, I'd hazard it could be controlled/designed for significantly fewer cycles than my kludge, given the thermal mass of most wood-boiler buffers (way more than 48 gallons of water, I'd hazard.)

If you use the EIA data for all of New England, the space heating to hot water BTUs are about 4/1, (3.7/1 for MA) not exactly a mild climate, and still well below an order of magnitude. See the US, NE, and MA pie charts:

http://www.eia.gov/consumption/residential/reports/2009/state_briefs/pdf/ma.pdf

It's not that different in WI (climate zones 6 & 7):

http://www.eia.gov/consumption/residential/reports/2009/state_briefs/pdf/wi.pdf (also about 3.7/1, just like MA, albeit with colder incoming water temps and colder winter outdoor temps.)

I suspect there are families of 8 that burning more fuel for domestic hot water in a tankless than I burn annually for all of it in the Rube Goldberg contraption (and I'll bet they won't have to make a warranty claim either), but while my house be near the left side 1-sigma mark, our total energy use is still in the middle zone, not super-outliers by any means. With the drainwater heat recovery in the system my heating/hot-water ratio is probably higher than the MA average, but the overall space-heating + water-heating energy use somewhat lower than the MA average.