Pre-heating water for tankless system in cold climate

Users who are viewing this thread

Goglu

New Member
Messages
8
Reaction score
0
Points
1
Location
Montreal, Canada
I purchased an EcoSmart 27kw system for a small home (2 bedroom 2 bathroom; 1 bath + 1 shower). Its winter in Canada up here, and its minus 20 degree C (-4F); so you can imagine water temperature coming in is very low (2-3C or 35-37F; I have yet to mesure it).

My tankless heater is working great for bathroom tap, kitchen sink, showers, but I can`t get it hot enough for the bath. After fiddling around with the valve and debating to change my pressure-balance valve to a thermostatic valve, I realized my problem wasn`t pressure, it was GPM demand for the tub faucet which probably doesn`t allow enough time to the water to heat up enough.

I read quite a bit on the topic from the web and all I come across is recirculating system and/or putting a tank to avoid cold phase between demand. Now, I am pretty sure, none of that will help me increase my hot gpm need.

I had an idea and tried to get information from the manufacturer or any manufacturer of tankless to see why I could not use a loop to bring hot water back to the cold water inlet of the tankless so the water temperature at the entry would be warmer, which, as per the manufacturer spec sheet, should increase hot water gpm. Obviously, I would need to put a check-valve and possibly a temperature valve.

I am looking at professional opinions to tell me why this should not be attempted? Since I cannot find anything on the topic, I am thinking there maybe something bad (wrong) in doing something like this... I still think this might be a solution.

Looking forward to hearing your thoughts and professional opinion.

Thanks
 

Dana

In the trades
Messages
7,889
Reaction score
509
Points
113
Location
01609
How are you intending to make the water flow in a loop from the output side (= lower pressure) to the input side (= higher pressure)? It would need to have a pump, and the pump speed would have to modulate with the flow rate.

The fundamental problem is that 27kw isn't enough to fill a tub at your tub-filling flow rate. Looping the water back through the water heater to achieve a higher temperature doesn't really fix the gpm problem. But adjusting the flow to the bathtub to a slower rate with a ball-valve would allow the water heater to deliver higher temperatures, and that's probably the cheapest and easiest fix.
 

Jadnashua

Retired Defense Industry Engineer xxx
Messages
32,770
Reaction score
1,190
Points
113
Location
New England
Assuming the unit is in heated space and you have the room for it, a large storage tank of water might help things out by letting the incoming water acclimate to the room temperature prior to it reaching the tankless unit. That tank couldn't have any insulation or it would take forever to warm up.

You might also consider trying adding a bunch of fin-tube radiators inline of the tankless that should be able to pick up some room heat along the way. Neither would be particularly efficient. You'd need to run some numbers...there probably wouldn't be enough heat gain at the small delta T, but a sufficiently large tank should, over time between major water uses, warm up. It will cool the area off, which may or may not be an issue, and it could end up with a lot of condensation, so a drip pan may be required.

BOttom line, IMHO, a tankless system isn't the greatest choice where winter water temps get low.

Expensive, but will work, would be to install a second or even third tankless system. Some are designed for that, and the units communicate with each other to decide how many need to turn on to meet the demand.

If you heat with a boiler, an indirect WH is often the best choice.
 

Goglu

New Member
Messages
8
Reaction score
0
Points
1
Location
Montreal, Canada
How are you intending to make the water flow in a loop from the output side (= lower pressure) to the input side (= higher pressure)? It would need to have a pump, and the pump speed would have to modulate with the flow rate.

The fundamental problem is that 27kw isn't enough to fill a tub at your tub-filling flow rate. Looping the water back through the water heater to achieve a higher temperature doesn't really fix the gpm problem. But adjusting the flow to the bathtub to a slower rate with a ball-valve would allow the water heater to deliver higher temperatures, and that's probably the cheapest and easiest fix.

Thanks for the reply Dana

Yes, it came to mind that the flow against would be the challenge. I’ll need to brush up on the Bernoulli president principal as I was hoping I could increase pressure by changing the size of the pipe for that feed. Just a little warm water could help warming up the water.
The manufacturer confirms at the same pressure, with warmer water, the GPM would increase.

I agree limiting the flow at the faucet would probably be cheaper but access is limited and this would increase time to fill the tub (a definite thought I am considering as it would most likely be easier and cheaper). But if I slow the hot water flow, do I have to slow the flow at the cold side as well?? Pressure-balanced valve seem to be my chalenge also. Might have to consider good old cold and hot separate handles.
 
Last edited:

Goglu

New Member
Messages
8
Reaction score
0
Points
1
Location
Montreal, Canada
Assuming the unit is in heated space and you have the room for it, a large storage tank of water might help things out by letting the incoming water acclimate to the room temperature prior to it reaching the tankless unit. That tank couldn't have any insulation or it would take forever to warm up.

You might also consider trying adding a bunch of fin-tube radiators inline of the tankless that should be able to pick up some room heat along the way. Neither would be particularly efficient. You'd need to run some numbers...there probably wouldn't be enough heat gain at the small delta T, but a sufficiently large tank should, over time between major water uses, warm up. It will cool the area off, which may or may not be an issue, and it could end up with a lot of condensation, so a drip pan may be required.

BOttom line, IMHO, a tankless system isn't the greatest choice where winter water temps get low.

Expensive, but will work, would be to install a second or even third tankless system. Some are designed for that, and the units communicate with each other to decide how many need to turn on to meet the demand.

If you heat with a boiler, an indirect WH is often the best choice.


Hi there,

I like the idea and had consider creating an extended line (creating a radiator like system) prior to entering the unit but I read somewhere that even 100’ of 1/2 piping would only contain around 2 gallons of water.

The holding tank could be a great idea if I had the space but that’s the reason why we went with the tankless (that unit is small) and unfortunately the hole to the crawl space is only a large enough for climbing down, not large enough for a standard diameter water heater tank. Any idea who makes narrow tanks? I have lots of head space but no way to get large 20 some inches wide tanks down there.
 

Reach4

Well-Known Member
Messages
38,798
Reaction score
4,413
Points
113
Location
IL
The holding tank could be a great idea if I had the space but that’s the reason why we went with the tankless (that unit is small) and unfortunately the hole to the crawl space is only a large enough for climbing down, not large enough for a standard diameter water heater tank.
You could put in a bigger trap door, if that was your only problem.
 

Dana

In the trades
Messages
7,889
Reaction score
509
Points
113
Location
01609
That's right 100' of half inch is good for about 2 gallons, about 5% of a tub fill, less than a minute's worth of 2.5gpm showering time.

A fin tube baseboard convector isn't a very effective heat exchanger at the ~30F temperature difference you would have between the incoming water temp and the room temp near the floor (35F water, 65F floor). The amount of heat it picks up will be at MOST about about 1 BTU per minute (60 BTU/hr) per running foot of baseboard. So maybe with 100' of baseboard it would muster ~100 BTU per minute, but the total heat needed (preheat + tankless) is at least a couple thousand BTU per minute for a tub fill, 1000-1500 BTU/minute for a 2-2.5 gpm shower. (The arithmetic on that: 2.5 gpm = 21 lbs/minute x 70F temperature rise= 1470 BTU/minute. A 5 gpm tub fill is twice that.) It would be hard to measure the difference.

If you have sufficient vertical distance on a drain downstream of the shower you can preheat the incoming water with a drainwater heat exchanger, which is great for shower capacity, but does NOTHING for tub filling, since the drain has to be flowing at the same time as the water is flowing through the heater. It's still probably going to be worth installing one if it doesn't involve demolishing finished walls to get at it.

power-pipe-dana-2.jpg


The current performance leader on those is EcoDrain's V1000 series, kicking Renewability's PowerPipe series from that perch after 10+ years of market dominance. (A 3" diameter V1000 beats a 4" PowerPipe at any given length on return efficiency.)

The shortest V1000 is 36", and would take about a 48" tall crawlspace to fit, but there are 30" tall PowerPipes that would make it in 40" of overhead.

If throttling back the hot flow to get an adequate temperature, yes the cold flow would have to slow as well. No amount of recirculation of the output back to the input overcomes the fact that rate of heat going into the water doesn't change- it's still 27kw. That 27kw is the same heat rate as 92,124 BTU/hr, or (/60=) 1535 BTU/minute. At a 110F tub filling temp and 35F incoming water temperature you have a 75F temperature difference, so the fastest flow rate it will muster is 1535BTU/75F= 20.5 lbs/minute, which at 8.34 lbs/gallon is (20.5/8.34=) ~2.5 gpm, not more. That's true even if you spin the water enough times through the tankless to bring it to boiling temperature.
 
Last edited by a moderator:

Goglu

New Member
Messages
8
Reaction score
0
Points
1
Location
Montreal, Canada
How are you intending to make the water flow in a loop from the output side (= lower pressure) to the input side (= higher pressure)? It would need to have a pump, and the pump speed would have to modulate with the flow rate.

The fundamental problem is that 27kw isn't enough to fill a tub at your tub-filling flow rate. Looping the water back through the water heater to achieve a higher temperature doesn't really fix the gpm problem. But adjusting the flow to the bathtub to a slower rate with a ball-valve would allow the water heater to deliver higher temperatures, and that's probably the cheapest and easiest fix.

Just reading on thermal expansion. Since pressure would remain (more or less) the same from the inlet to the outlet of the tankless heater (in the spec of the manufacturer; would probably require proof testing), I’m wondering if I could bet on that element of physics to push back a little water back toward the main so I can have a minimalistic amount of hot & cold mixing at the inlet. If pressure increases with increased temperatures, could that actually work?

Since pressure increases with larger pipes (but flow rate drops) maybe it’s worth going 1/4" inches larger for that loop??

I guess I’m trying to limit the amount of addition I need to do to this system (power or equipment).
 

Goglu

New Member
Messages
8
Reaction score
0
Points
1
Location
Montreal, Canada
That's right 100' of half inch is good for about 2 gallons, about 5% of a tub fill, less than a minute's worth of 2.5gpm showering time.

A fin tube baseboard convector isn't a very effective heat exchanger at the ~30F temperature difference you would have between the incoming water temp and the room temp near the floor (35F water, 65F floor). The amount of heat it picks up will be at MOST about about 1 BTU per minute (60 BTU/hr) per running foot of baseboard. So maybe with 100' of baseboard it would muster ~100 BTU per minute, but the total heat needed (preheat + tankless) is at least a couple thousand BTU per minute for a tub fill, 1000-1500 BTU/minute for a 2-2.5 gpm shower. (The arithmetic on that: 2.5 gpm = 21 lbs/minute x 70F temperature rise= 1470 BTU/minute. A 5 gpm tub fill is twice that.) It would be hard to measure the difference.

If you have sufficient vertical distance on a drain downstream of the shower you can preheat the incoming water with a drainwater heat exchanger, which is great for shower capacity, but does NOTHING for tub filling, since the drain has to be flowing at the same time as the water is flowing through the heater. It's still probably going to be worth installing one if it doesn't involve demolishing finished walls to get at it.

ee-piste-1-recup-eaux-grises-diag-lrg.jpg


The current performance leader on those is EcoDrain's V1000 series, kicking Renewability's PowerPipe series from that perch after 10+ years of market dominance. (A 3" diameter V1000 beats a 4" PowerPipe at any given length on return efficiency.)

The shortest V1000 is 36", and would take about a 48" tall crawlspace to fit, but there are 30" tall PowerPipes that would make it in 40" of overhead.

If throttling back the hot flow to get an adequate temperature, yes the cold flow would have to slow as well. No amount of recirculation of the output back to the input overcomes the fact that rate of heat going into the water doesn't change- it's still 27kw. That 27kw is the same heat rate as 92,124 BTU/hr, or (/60=) 1535 BTU/minute. At a 110F tub filling temp and 35F incoming water temperature you have a 75F temperature difference, so the fastest flow rate it will muster is 1535BTU/75F= 20.5 lbs/minute, which at 8.34 lbs/gallon is (20.5/8.34=) ~2.5 gpm, not more. That's true even if you spin the water enough times through the tankless to bring it to boiling temperature.

Thanks for the info; i didn’t know this existed to be honest. For the btu calculation, you lost me a little. But any improvements like this should have been thought at the stud level of the renovation. Most of the drain is above the crawl space and not easily accessible anymore.

One thing I have though is one of those garage heater down in the crawl space. I try to keep it low around 6-10 C (43/50F). I can for sure coil pipes or squeeze in a smalle tempered tank (well tank style) to help pre-warm the water.

It becomes a cost trade-off now.
 

Dana

In the trades
Messages
7,889
Reaction score
509
Points
113
Location
01609
Just reading on thermal expansion. Since pressure would remain (more or less) the same from the inlet to the outlet of the tankless heater (in the spec of the manufacturer; would probably require proof testing), I’m wondering if I could bet on that element of physics to push back a little water back toward the main so I can have a minimalistic amount of hot & cold mixing at the inlet. If pressure increases with increased temperatures, could that actually work?

Since pressure increases with larger pipes (but flow rate drops) maybe it’s worth going 1/4" inches larger for that loop??

I guess I’m trying to limit the amount of addition I need to do to this system (power or equipment).

It reads as if you're looking for some magic to overcome the laws of thermodyamics. Can't help you that, but good luck, eh? ;)

No matter how you pressurize back-spin or mix the water along it's way, 27kw is still 27kw, not more, and that's going to limit you to less than 10 litres/minute at bathtub flow. To get the heat out you need to put the heat in somehow. To fill a tub in reasonable amount of time is going to take something like 50kw, and the 23kw difference between 50kw and 27 kwh is still a pretty substantial heat rate, not something that easily tweaked around the edges. It takes setting something on fire, drawing a lot more electricity, or storing that energy in a tank big enough to fill the tub.
 

Goglu

New Member
Messages
8
Reaction score
0
Points
1
Location
Montreal, Canada
It reads as if you're looking for some magic to overcome the laws of thermodyamics. Can't help you that, but good luck, eh? ;)

No matter how you pressurize back-spin or mix the water along it's way, 27kw is still 27kw, not more, and that's going to limit you to less than 10 litres/minute at bathtub flow. To get the heat out you need to put the heat in somehow. To fill a tub in reasonable amount of time is going to take something like 50kw, and the 23kw difference between 50kw and 27 kwh is still a pretty substantial heat rate, not something that easily tweaked around the edges. It takes setting something on fire, drawing a lot more electricity, or storing that energy in a tank big enough to fill the tub.

So, I guess I was looking for some magic. Now, since I don’t believe in magic for say, I went through the "Tinker” approach (sorry, I got young kids) and came up with a home-made solution.

Now bare with me, I didn’t fix the gpm issue at the source as I was looking to do . I instead just managed to slow the water down to be able to get water hot enough to get a warm bath.

The down side is that it will take a little longer to fill the tub (and as the seasons moves along, this might be something I can revert back to the original valve assembly. It might have been easier to change the valve casing to one with adjustable flow valves but in a inaccessible (from the back) finished wall, it was what I came up with.

I inverted a rubber washer that I had to shave the edges off and glued it inside the hot water access within the valve assembly.

Thanks for everyone’s suggestions above.
 

Attachments

  • 450B5F2F-3D38-41A2-8640-8F9A207E7EDE.jpeg
    450B5F2F-3D38-41A2-8640-8F9A207E7EDE.jpeg
    84.4 KB · Views: 219

Jadnashua

Retired Defense Industry Engineer xxx
Messages
32,770
Reaction score
1,190
Points
113
Location
New England
WHen it comes time to remodel, consider one of the valves that has a volume control versus the much more common single handle valve that only runs at one volume. Also, a rough-in valve that has in-line shutoffs might be able to throttle the max flow, but you might want to adjust it season to season so you'd improve your flow once the incoming water warms up.

One option that may also help would be a thermostatically controlled valve. THose tend to have a volume control, and then, you set your desired temp, and turn the volume down if it can't keep up.
 

Goglu

New Member
Messages
8
Reaction score
0
Points
1
Location
Montreal, Canada
WHen it comes time to remodel, consider one of the valves that has a volume control versus the much more common single handle valve that only runs at one volume. Also, a rough-in valve that has in-line shutoffs might be able to throttle the max flow, but you might want to adjust it season to season so you'd improve your flow once the incoming water warms up.

One option that may also help would be a thermostatically controlled valve. THose tend to have a volume control, and then, you set your desired temp, and turn the volume down if it can't keep up.

Thanks. I have learnt my lesson. Now, about the thermostatic option; I called Delta technical support to brainstorm with them and they could not confirm or not if this would make a difference in my situation (tankless with very cold incoming water). They did suggest as you did with the rough-in Control valve. The price is very high in comparison to the moen I had purchased; a lot more.
 
Top
Hey, wait a minute.

This is awkward, but...

It looks like you're using an ad blocker. We get it, but (1) terrylove.com can't live without ads, and (2) ad blockers can cause issues with videos and comments. If you'd like to support the site, please allow ads.

If any particular ad is your REASON for blocking ads, please let us know. We might be able to do something about it. Thanks.
I've Disabled AdBlock    No Thanks