I don't have that problem, we work with the system. I don't see the buffer tank as saving you any money.

My Takgai saves me money in two ways. It only heats the water I use and only to the temp I need.
I don't need mixing valves, I have the water set to between 104 and 110.

I should add, the beauty of the  on demand gas system is that it cuts out all the extra switches, flow meters and extras that are needed in a larger demanding system.

If you are goin to add in lots of extra controls and gizmo's to get to the "universal North America hot water on demand system" you should just sell the Takgai to someone who will use it to it's potential and put in a mod/con.

(regarding http://www.taunton.com/finehomebuilding/how-to/articles/add-a-tank-and-pump-to-a-tankless-water-heater-to-save-energy.aspx) Holy cow that article is silly. By the time you add a buffer, pump, and tankless up, you might as well have just gotten a high capacity tank water heater.

Paul I keep getting new things from your plans here, are they evolving as we go? If you're going to have a "buffer tank", then go with dana's idea of using the ergomax as a combo "instant" hot water heater, and buffer tank. It might be a bit pricier than something you might cobble together, but it will be more efficient and it will work better, and it will eliminate the "start up delay" the heat exchanger itself has.

If you are married to your current heat exchanger, then you are talking about two tanks if you want to do buffering, and to eliminate the startup delay. You can't "buffer" through a HE, not effectively or easily at least, and you can't eliminate the heat exchanger "startup delay" with the buffer tank on the closed side of the system.

You're really trying to do a lot of stuff with all the wrong equipment. I don't think you're doing yourselves any favors here. You could really simplify your life by talking about a material return.

Rob, I can see your point. I thought I had mentioned the buffer tank earlier, but that was a different forum.
What's wrong with buffering through the HE? All the circulator activity?
If I downgraded to a tank heater, I'd still need a primary and secondary circulator for the hydronics and another circulator between the tank and the HE.
An ergomax would cut out one circulator, true. I'll look into that option a little bit more.

I guess my mind just isn't accepting all you are trying to do with the hardware you have. I'm sorry if I confuse the issue further for yo.

HX's do carry an efficiency penalty. you're looking at transferring twice across the HE if you are trying to draw heat out of it, and charge the tank up through it. whatever you use to do that is likely to be complicated, and certainly silly and cycling a lot more than you need to.

the pump you need on the DHW side is more than $200 by itself, as well, with its associated hardware The relays and/or temperature controls need to be added in. What happens if you get a domestic call while you are heating? Is the takagi hotter than the domestic to transfer out or is the buffer hotter than the takagi to transfer in? How much additional energy are you wasting with the unnecessary pumping that you're trying to save by buffering? If you can't draw the heat back out of the domestic side buffer on a heating demand, it's not much of a buffer.. half as much as it could be. I suppose you could then use a tank twice as large..

the ergomax eliminates bronze high flow pumping and a serious amount of control logic here: Keep the tank hot. End of complication. If this is the situation you find yourself in, it's pretty hard to argue with the simplicity, even if you lose some money reselling hardware you already own.

If you're heating a slab, the thermal mass of the slab is effectively your buffer, but if it's a staple-up with multiple zones buffering WILL make it run more efficiently. How much more depends on the system, but unbuffered it'll probably run at the EF rating or slighly lower.  The Takagi TK2 rated 78% in an AFUE test despite scoring 80% in the EF test.  But in a buffered low radiation temp system should be able to hit low mid-'80s performance no problem, since it's raw combustion efficiency is ~85%.  AFUE tests boilers, not systems, and not in a very realistic fashion, and are not designed to tease out the best performance of the device.  The in-situ real performance can fall on either side of the AFUE (usually lower in high-temp systems), but never higher than it's raw steady-state combustion efficiency (but close!)

The buffer tank saves you money by (nearly) eliminating cycling losses, at the cost of ~1/4-1/2% in tank-standby losses.  Even on a low-mass boiler like your Takagi they add up- FAST.  It's far more efficient to have one 40 minute burn of varying low modulation than to have four 3 to 5 minute burns at higher fire in the same time period.  In systems chopped up in to  handful of micro-zones there's no averaging it out- it WILL short cycle, even if it's at varying modulation levels.  But with a central buffer that all gets smoothed out into longer, lower-fire burns.  If you're planning a system with DHW heat exchangers anyway, a low-head buffer-tank type heat exchanger will save you on electricity as well, since you need less pump than you might need for driving a higher-head type of exchanger + boiler/tankless.

The cleverness of heatpro's approach is that you set the temp of the aquastat of the buffer/heat exchanger to the temp the radiation needs, set the DHW output with a tempering valve, program the output temp of the tankless to 10-15 degrees hotter than you set the tank and you're done.

The required controls are deceptively simple:  The aquastat on lower 1/3 point the buffer is the master of the boiler/tankless, which remains agnostic of the state of zone or DHW calls/flows.  The bottom of the tank is colder when all zones are running (or someone is taking a 20 minute shower), and the boiler modulates up to still deliver the same temp output water.  If the output temp and primary loop pumps are properly sized, when all zones turn off,  it backs off fairly rapidly (with return water no longer being injected to the bottom of the tank the return to the boiler quickly starts warming up) to the lower/lowest modulation and heats up the tank until the aquastat's call is satisfied (if you use the Ergomax you can run boiler pump directly off the aquastat, which a line-voltage device with 16-amp contacts.)  Since all heating water is being drawn from the top of the tank near where the boiler output is being inserted (with some designed in turbulence & mixing), the heating water temp remains somewhere near the aquastat temp even when the bottom of the tank is filling with much colder water from the zone returns.   The DHW temp remains whatever you set the tempering valve to, even if the temp at the top of the tank is varying several degrees.  It's a KISS (Keep It Simple Stupid) approach, with decent efficiency and not a lot of complication.  There's no outdoor reset, no mixing valves, flow meters- nada, just zone pumps (or if it's small enough, one pump plus solenoid operated zone valves.)

The downside to the approach is you never run it lower than you DHW temp requirements, so you lose a bit of potential heating efficiency on the shoulder seasons, but in most cases you'd be talking something less than a 5% difference in seasonal fuel consumption.

But if all you're doing is heating DHW with it, the heatexchanger/buffer-tank will only save you the "cold water sandwich" and a small amount of maintenance- the total efficiency will be about the same for a long more complication & expense.

Okay Rob, that logic makes sense. What about this as an alternative (it sounds as thought it may have been what you thought I meant earlier):

-tankless runs on closed loop, high temp (mixing temp down for radiant loops)
-tankless keeps a small buffer tank hot ON THE CLOSED LOOP
-buffer tank is immediately adjacent to HE (relatively massive tube-in-tube 27ft. coil, insulated, from Tesmar) which keeps the HE nearly as hot as the liquid in the buffer tank
-HE now acts as "instant" and "endless" hot water coil, supported by Takagi once buffer tank gets below a certain temp

ok, I see what you are doing. sorry again for any confusion/denseness on my part, you are only getting part of my full attention this time of year.

that's not too bad. I'd never do it as a default, to be clear for future readers, but to utilize what you have got, I imagine it would be appropriate.

Dana, I need to digest all that information slowly.... I gather in a first run-through that buffers pay off in low mass radiant systems (mine is radiant ceiling) and that using stratification enables you to eliminate extra equipment, but i'm gonna have to draw that out to understand...

Rob, you've been following me fine---I just changed gears in that last post to ask if having the buffer IN the closed loop sounded more reasonable to you than buffering THRU the heat exchanger.

I understand and GREATLY appreciate that you're trying to help me utilize what I've got, not design the ideal system from scratch.

Three follow-up questions on that then:

-Am I right in thinking a 40 gallon buffer in this case would be very much oversized?
-So am I possibly correct in thinking that keeping the buffer very close to the HE, all of them well insulated, will keep the HE nice and hot for its next use?
-And if so, what info. do I need to determine whether the HE's heat transfer can or can't keep up with the Takagi to provide as much gpm of hot water as the Takagi would have by itself? (HE is 27 ft. of coaxial 1/2" and 1-1/4" copper, coiled and in contact with itself, and insulated). This by the way if exactly the way a local guy does it, except he uses a flat plate HE and no buffer tank. If I assume it works for him, then the question is whether my HE can also deliver "endless hot water" with the right temp settings in the Takagi, etc....

Thanks all for taking the time. I am really appreciative of it. Once it's all said and done, I'll be sure to post how things turned out, hopefully with some numbers to support.

Okay, here's a picture of what I'm thinking about at this point. (Option x and y as opposed to different options a and b from the beginning of this thread)

Option x has the buffer tank on the DHW side and the Takagi is buffered by it via the HE. 3 circs required. kind of clunky, right?

Option y has the buffer tank in the closed loop, buffering the Takagi. It sits immediately next to the HE, to help keep the HE warm (possible?). This option depends on two things:

-most importantly that the HE can keep up with the Takagi's output (5.5 gpm at 60 deg F rise, for instance) when DHW is called for

-also that heat conduction through the pipes from the buffer tank to the heat exchanger can actually keep the HE hot.

Option y, if it isn't totally flawed, would be my preference, for one because it requires one less large, expensive, brass circ.

Any thoughts?

you probably want a small recirc pump between buffer and HE to keep the HE hot in option Y. But I'm guessing, having never used that configuration myself.

You have already purchased the equipment. The down side to having the tankless on the DHW side is scale build up on the HE in the water heater. If you place the unit on the closed loop side you will still have a heat exchanger that will be subject to build up. The controller for the unit will flash a code when the internal exchanger has a build up. It is a fairly easy process to circulate a cleaner such as white vinigar through the unit and the secondary exchanger. If you put the unit on the closed side you will have to rely on your own judgement as to when the exchanger needs to be cleaned. You will also have to have the proper valves and boiler drains installed to be able to accomplish the process. I would place the unit on the DHW side and rely on the internal sensors to notify me when the cleaning is necessary. This will also eliminate the need for a buffer tank (unless you want to eliminate the time delay for unit start) and will take the guess work out of maintenance scheduling.

Well, I realized a major flaw in "option Y" is that every time I want to circulate water from the buffer tank to the heat exchanger, it would have to also run through the Takagi, which could short cycle the takagi and defeats the purpose of the buffer. So I'd need third circulator anyway, as you point out Rob, so option Y is really no better than option X.

Thanks for that logic, Geofreak. Makes sense. But on the other hand, wouldn't I notice when the heat exchanger needed cleaning when my hot water temp starts dropping? If that was the case, all I'd need would be a little temp gauge. I called Takagi and learned that the unit flashes its code only when the exchanger is actually blocked with scale enough that the unit can't function. So it seems to me it would be burning more and more, getting less and less efficient, well before it actually let me know there was a scale buildup. If that's true, I'd need to be doing guesswork anyway to try and keep the Takagi clean.

not true. you can have a circuit that includes on the buffer, HE, and a pump, you don't have to circulate through the takagi.

I think we agree, Rob----What I was saying was that I drew option y incorrectly by putting buffer, HE and takagi all on the same circuit. So when I was excited about that option only requiring 2 pumps, I was wrong. It would need 3 to actually use the buffer tank as a buffer.

Sorry, I was thinking of the Rinnai unit and their codes.