Water heater efficiency, temp high vs low

[Coolwhip]

In general, per BTU, do NG condensing tankless water heaters operate more efficiently at higher or lower temperatures? Would it be better to run it less often at a higher temp or more often at a lower temp?

Basically wondering for radiant floor heating, in an open system, would it be better to set the water heater higher and use a mixing valve, or set it lower and forego the mixing valve?

 

[Dana]

The entering water temperature is the primary determinant of combustion efficiency. The next most important factor is firing rate- keep it as low as possible. With a condensing boiler condensing only starts at about 127F and lower, but with some tankless water heaters it has to be 120F or lower (due to higher excess combustion air dilution). The lowest possible temperature and firing rate that actually meets the load is the recipe for longevity. Lower temperatures also result in lower lime accretion on "open" systems (but those are a bad idea anyway.)

When using a tankless as a boiler it's important to not over-fire or over pump the tankless or it won't last very long. Many on-line kits are just crazy on both accounts. Design your system- don't hack it. A flow of 1.5-3 gpm through the tankless would be a safe range for most. Higher than that will wear out the flow sensors, or even erode the internal plumbing. If it needs more heat to meet design conditions, raise the output temperature. These things are designed to do really well with low flow and high delta-Ts, and if you over pump them the delta-T drops. A flow of 2 GPM at a delta-T of 30F is 30,000 BTU/hr which is more than enough heat for most reasonably tight homes in GA.

The pumping head of tankless water heaters is pretty high, and in many /most radiant slabs it will have to be plumbed primary/secondary ,and sometimes it needs a thermally massive hydraulic separator (rather than closely spaced tees) to keep it from short cycling.

Open radiant heating systems have pathogen risks due to stagnation at tepid temperatures, and are disallowed in some states for that reason.

FWIW: I've been heating my home with a tankless water heater as a boiler for the past 10 years without any real problems with the tankless. My temperature requirements are higher than most radiant slabs, and it needs a buffering thermal mass due to micro-zoning. It's a closed system, with the tankless heating up a 48 gallon "reverse indirect" buffer tank, from which all zones (including the radiant zone) sip heat a diverse flow rates. The tankless and primary loop are set up to run 2.5-3gpm at a delta-T that varies from 35F-50F for normal space heating just maintaining the buffer, but can be as high as 60F when someone is taking a long shower while all zones are calling for heat.

Using a 40 gallon electric water heater as a buffer tank can be a cheaper way to go, if you're doing something else for hot water.

 

[Sponsor]

 

[Coolwhip]

Thanks, I think I understood most of that. I was really wondering whether it would be better to set a larger allowed delta-T before the pump kicks back up again, and have the water heater temp set higher(which might require the use of a mixing valve, splitting the return from the radiant heating loops between the heater and the mixing valve)....or if it would be better to set a more narrow delta-T on the pump and keep the hot water heater turned down.

From what you said...it seems like it would be the latter as far as efficiency goes, even though(as far as I understand it) that might mean the boiler is on more often; and forgo a mixing valve.

"Open radiant heating systems have pathogen risks due to stagnation at tepid temperatures, and are disallowed in some states for that reason."

I have worried about that as well, so I guess I just need to make sure I don't go too low in pursuit of cost savings....do you have any idea whats the lowest safe temperature setting I could use? If it matters it's a 3 zone system, largest loop is around 300ft, ~750ft total, but I doubt I will often(probably never) have more than 2 zones going at the same time....along with a groundfos alpha 2 pump, possibly using the autoadapt feature(we will see how that goes).

I plan on just making sure the system is either ON or OFF, and possibly run a good bit of water through it at the end of the heating season when I shut it down(to quickly cool the system off)...and use other means of heating in the fall when temperatures fluctuate...rather then turning the system on then off then on then off and allowing water temperatures to repeatedly transverse through the "danger zone".

And I have considered closed and indirect systems, as I plan on adding some hydronic cooling in the future(not to the floor)...but cost is a concern and the simplicity of an open system is very attractive to me. It's a small house and considering its layout its just the right thing for my situation.

I also want the(small) amount of free cooling I will receive in the summer as hot water use will still draw new water through the tubing loops...don't worry, I've checked and at these water temps condensation won't be a concern and my subfloor/radiant system is designed in such a way that even if a small amount were to occur it wouldn't be a problem.

But I do have a spare hot water heater, and maybe I will give indirect heating another look, although I was hoping to use it in conjunction with a air-to-water heat pump to provide hydronic cooling before the next summer came....maybe I'll put that project off and instead use the water tank and rather than spend my time/money working on hydronic cooling as my next project do solar hot water first(to assist the tankless)...it's just with the houses small size and how cheap gas is here I would get a better ROI on hydronic cooling than solar heating, but if pathogen formation is a bigger threat than I was aware of I don't want to put my family's safety at risk.

 

[Dana]

Thanks, I think I understood most of that. I was really wondering whether it would be better to set a larger allowed delta-T before the pump kicks back up again, and have the water heater temp set higher(which might require the use of a mixing valve, splitting the return from the radiant heating loops between the heater and the mixing valve)....or if it would be better to set a more narrow delta-T on the pump and keep the hot water heater turned down.

Describe the type of radiant floor (slab? suspended tube? suspended tube with convection fins? staple-up? staple up with plates?), and how many BTU/hr per square foot the floor needs to deliver. The particulars make a huge difference as to what the water temperatures need to be.

From what you said...it seems like it would be the latter as far as efficiency goes, even though(as far as I understand it) that might mean the boiler is on more often. And that would forgo that shouldn't require a mixing valve.

With a mixing valve the flow through the tankless can become too low to keep the flame going- this really has to be designed, not hacked. A high mass buffer tank can take high temp water out of the tankless at any arbirtary flow rate and still deliver cooler water to the radiation, if necessary.

"Open radiant heating systems have pathogen risks due to stagnation at tepid temperatures, and are disallowed in some states for that reason."

I have worried about that as well, so I guess I just need to make sure I don't go too low in pursuit of cost savings....what do you think a good temperature setting would be? If it matters it's a 3 zone system, largest loop is just under 300ft, ~750ft total, I doubt I will often(probably never) have more than 2 zones going at the same time....along with a groundfos alpha 2 pump, possibly using the autoadapt feature(we will see how that goes).

It's better to install a plate type heat exchanger (rated for potable water) to isolate the potable water from the heating system loop and get rid of the problem than it is to run at some high temperature in the hopes that nothing bad will happen. Unless the water in the heating loops are reliably below 75F in the summer and over 125F in the winter you're just asking for bad things to happen. Letting even one zone stagnate all summer can contaminate the rest of the system if something starts growing, so in the summer every inch of pipe has to have flow at some minimum duty cycle to stay reasonably clean. (The minimum duty cycle in summer to avoid problems related to stagnation is enshrined in code in Massachusetts, where the pathogen risk isn't even as high as as it is in Georgia due to the lower incoming water temps.)

Open systems are more trouble than they're worth, especially in locations with warm enough summers (and water temperatures) to require. substantial air conditioning. You can maybe get away with it most of the time in, say, Edmonton Alberta, but it's riskier south of the 48th parallel, riskier still south of the Mason-Dixon.

Seriously, isolating the heating system costs you at most one pump, the heat exchanger and an expansion tank for the heating side, and it allows you to set the flows on the heating side independently of the flow on the potable side. This works just fine with a tank"less", where it's labeled "Tank":

But whether this exact configuration works with the temperature requirements of your radiation remains to be seen. You have to do at least some of the basic design math, starting with:

Heat loss calculations, broken down zone-by-zone (Manual-J or I=B=R method works, but not the typical crummy online calculators.)

Available square footage of active radiant floor, to determine the floor temperature needed to support the calculated load.

The radiant floor type, and floor stackup, which will determine the water temperature needed to deliver the calculated floor temperature or BTU/hr per square foot.

While it's a great pump, the minimum flow rate of the Alpha 2 isn't as low as the Taco VR1816 or Aquamotion AM55 "Einstein" (which I had to use in "Night Setback" mode to avoid over pumping one of my micro-zones) and may end up over pumping either the tankless or the system, depending on how you have it configured. Don't count on "Autoadapt" avoid having to actually design something (at least to the napkin-math level.)

With the Alpha 2 in Autoadapt mode if there's less than 4 feet of pumping head @ 2gpm some smaller zone the flow rate is going to be potentially higher than what you want or need.

^^Alpha 2 pump curves

^^ Aquamotion AM55 pump curves

^^VR1816 pump curves