I'm currently planning on having a ~80 gallon buffer tank connected to the desuperheater of our geo-exchange unit.  As well, the buffer tank will contain a heat exchanger connected to a set of solar thermal panels.  I'm told by my contractor that both the geo (mostly winter) and solar (mostly summer) are sized & configured to cover about 75% of our DHW load to avoid overheating.

Currently the plan is to put a residential tankless natural gas water heater in line after the buffer tank to provide final heating.  An anti-scald valve will ensure the entering temp to the tankless is at least 10 degrees below the desired hot water temp, so the tankless unit doesn't "pulse" on and off.

My question is whether a small tank (maybe 40 gallons or smaller) natural gas heater would be a better idea than the tankless unit.  My goals are energy efficiency (least amount of NG burned) and comfort/water savings (avoid running cold water while waiting for hot to arrive). Any suggestions on the best approach?

My thought has always been try the simplest which usually turns to be the cheapest method first. Take metal water tank elevatte paint it black run that line into a gas water heater. Theory here is that if you set the thermo at say 100 the tank thermo might not kick in at all but if it does it only has heat 10 degrees of water not 70.
My thoughts

A gas fired tank has a significant standby loss, whether it's heating the incoming water 10F or heating it 80F.  Running it in the mode you describe you'd spend 80-100therms/year to support the standby loss, and it's effective average efficiency will be less than 10% due to the high standby loss (you'd literally be better off with an electric tank, which would be ~30% thermally efficient assuming a fossil-fired generator and standard grid distribution losses.)  A tankless has effectively zero standby loss, and will run better than 50% average efficiency operated as described (but still nowhere near it's EF rating- more on that later.)

Since your stated goal is to use less fuel, if it's not already too late, insulating the HW distribution lines everywhere to at least R4 (5/8" wall closed cell foam pipe insulation.)  That drastically reduces the amount of tepid water purged, since it provides ~20-30 minutes of standby time between draws where the standing water in the distribution plumbing remains hot enough to be useful:



Abandoned heat in distribution plumbing accounts for something on the order of 15% of the total fuel used to heat water, and it can be cut to under 5% if the plumbing is well insulated.  ($100 in pipe insulation is a much better investment than $2500 in gas-fired tankless installation, and will save as much or more fuel/electricity in your case.)

Rather than using a mixing valve to provide a delta-T for a gas-fired tankless, it may be more efficient to mix it to the exact desired output temp (say 115-125F) and run it through an electric tankless programmed to the same or 2 degrees cooler output temp.  Unlike gas-fired tanklesses, electrics can modulate to extremely low levels, and won't have pulse issues related to minimum fire that gas units do.  If the solar/geo will reliably be doing more than 50% of the heavy-lifting under worst-case conditions, the amount of electricty used by the electric tankless will minimal, and you can probably run the whole house on a smaller point-of-use type electric tankless.  Even though it's relatively source-fuel inefficient compared to gas fired tankless, it's duty cycle will be quite low.  You'd have to do some careful modeling to know exactly how it compared to a guaranteed 10F load to every draw using a gas-fired tankless, but 30% source-fuel efficiency at a 25% the duty cycle is lower footprint than ~70-75% fuel-efficiency at 100%.  Maintenance will be lower, and temperature regulation (particularly at very low flow and low delta-Ts) will be higher.

If you DO end up going with gas-fired tankless for capacity reasons, look very closely at the specifications for minimum modulation levels.  Under 15KBTU/hr in is good, but under 10K is much better when looking at the ultra-low delta-Ts you're planning on.  Also, it is more efficient  to run it boiler-style in a circulation loop to support a minimum temp to the solar/geo storage tank than it is to run it in series & low delta-T.  By setting it's output temp and circulation pump flow such that it's somewhere between 1/4-1/3 of full-fire, and setting the hysteresis on the tank controls to guarantee it runs at least 5 minutes for every fire you're running it near the sweet-spot on it's combustion efficiency, and losing far less in cycling losses.  (Every ignition cycle & flue purge throws away a fixed amount of BTUs- fewer are definitely better.) Assuming a storage tank setpoint of ~125F that would guarantee a standard-efficiency tankless would run at a true ~82-84% efficiency, and a condensing tankless somewhere around 90% (not more, due to the high incoming water temp.  It has to be under 100F on the input to get to 95%:



The expense of a condensing vs. standard efficiency burner is extremely difficult to rationalize if 75% of the heat is already being supplied by high-efficiency geo/solar.  Even a standard efficiency tankless seems like a lot. But if you go that route, supporting the  thermal mass of the storage will be far more efficient than short-cycling it at low delta-T in a series configuration.  And by keeping the storage above 120F you'll limit the legionella hazard (which is low, if the geo or solar regularly raises it above 140F, but non-zero if it stagnates at under 120F for weeks.)

Also, if it's not already planned in, for less money than the installed cost of a gas fired tankless, drainwater heat recovery heat exchangers will cut fuel use by ~15-25% on average (assuming 3-4 people taking showers, not baths per day.)  By pre-heating the incoming water stream with the warm water going down the shower drains it can even allow one to downsize the solar thermal array by 15-25%.

Outstanding response as always Dana1.  To answer some of your questions:

• Yes, I've asked my builder about insulating the hot water distribution lines.  Hopefully we'll be able to do this.
• Yes, based on your other posts about DWHR, I'm planning to install a good-sized Powerpipe unit (handling both upper showers).
• The tankless NG unit currently spec'd is an Bradford-White Everhot which modulates down to 15K BTU/h
  

Regarding the idea of using a boiler-style loop to maintain heat in the buffer tank, rather than adding a heater in series afterwards: Wouldn't the higher temp of the buffer tank reduce the efficiency of the desuperheater and solar thermal heat exchange?  My understanding is that these renewable systems are better at warming cold water than making already warm water hotter.

Some further background in case it helps: The DHW load for the home will be 6-adults, well really 4 adults and 2 kids, but they grow up fast :)  We're planning a 4-ton Nordic liquid-to-water GSHP and 2 panels of 24 evacuated tubes each for solar thermal.  The GSHP will mostly run in the winter heating via radiant in-floor and we won't have A/C. The solar thermal will be angled to output best in spring, summer, and fall.  My goal is to have these two renewable systems complement each other over the entire year.  In 5-10 years, we hope to achieve Net-Zero by adding a bunch of solar PV on the roof.  We will have a natural gas line for the stove, fireplace, and BBQ but we'll try to minimize their use.

I like the idea of a tankless electric unit to do the final heating.  My builder has experience with them as they used it in a Net-Zero home.  However, they'll likely say the unit would have to handle cloudy days in spring/fall when geo/solar weren't doing much, so it would have be sized for the full 6-person DHW load.  This might mean a very powerful electric tankless unit, including upgraded electric service. (?) I'll definitely ask my builder for cost estimates though.

Thanks again for your detailed response,
Guy

It's true that the solar & geo work more effectively & efficiently at lower temp, but using a standard backup method for active solar space heating apps.

Installed cost-wise an electric tankless might not be as bad as you think. A 4-ton geo unit on-site means you already have a pretty fat transformer & lines planned for the house. Electric tankless units aren't ultra expensive- the installed cost (including panel upgrade) is likely to be less than what it takes to install a gas-fired tankless, once venting, plumbing & electrical for the latter are factored in. Sizing a tankless the primary factor it meeting the peak load- the average is less relevant. Even on cloudy low heating/cooling load days the a 48 tube evac tube array will be putting out something, so it doesn't have to necessarily support 3 simultaneous showers with 40F incoming water. If you're at all conservative about volume used during those worst-case periods its likely you wouldn't drop below 65F in the solar tank, but that's a bit hard to gauge ahead of time. Some habits may have to adjust during times of low solar/geo input. A 24kw unit would be 100A of 240V but would be enough to support a continuous single shower with cold water input, two (or three, with DWHR) when the solar/geo storage is above 80F. A 12kw unit would be fine for a single full-flow shower at a time with storage at 80F, but probably would come up short for a family of 6 from time to time- you'd have to be checking the storage temp before stepping into the shower sometimes. But there's a good chance a 20-24kw unit would be sufficient, and would leave you at least SOMETHING to shower with in a worst-case scenario where the desuperheater & solar were both off line for an extended period.

I applaud your Net-Zero ambitions- it looks like you have the rigth builder on board!