Tankless WH By the Numbers

Users who are viewing this thread

Jadnashua

Retired Defense Industry Engineer xxx
Messages
32,771
Reaction score
1,191
Points
113
Location
New England
There are two general methods to heat water for domestic use: a tank, or a tank-less system. They each have their advantages and disadvantages. I’m only going to briefly discuss tanks, but spend more time talking about tank-less systems from an energy and performance viewpoint.
Tanks
There are three commonly used tank types: electrically heated, burner based (natural gas, propane, or oil), or indirectly heated from a boiler. The main advantage of tanks is that they are well-known, been around for a long time, fairly simple, and until you run them out, maximum volume of hot water at any volume your piping can provide. The disadvantages are: they are fairly large, especially if you need a lot of hot water quick, the quality of the insulation and the type of tank design will have some standby losses as heat is lost to the room, and they have a limited lifetime. That lifetime can vary considerably depending often, on the luck of the draw, but design can improve things a bit.
Tank-less
The allure of a tankless system is that they promise an unending supply of hot water and have essentially no standby losses. They generally are a lot smaller than a tank, and most of them are designed to hang on a wall. They are usually quite a bit more expensive to install and may require upgrades to the infrastructure, either electrical capacity or gas supply. They are more complicated than a typical tank-type water heater, but their heat exchanger and controls are often designed to be modular and replaceable. While there are parts on a tank type water heater that can be replaced, they often last the life of the tank, which usually rusts out and thus requires replacing the whole thing.
General
With a tank-type, you have a relatively small heat source because you tend to have a long time to recover. If you need lots of water constantly, there are high capacity burners that can aid, but the more common thing is to use a larger tank, especially in a residential situation. That may not be sufficient in a commercial situation, and there they may use both high capacity burners, larger tanks, and maybe multiple units (like say in a restaurant that is washing dishes or a spa that is filling lots of tubs or running showers all day long).
With a tank-less system, you need a big enough heat source to raise the temperature of the water enough to be useful as it passes by. This is much harder to achieve with electricity than with a flame based burner. There are two factors that determine how much you can raise the temperature in a tank-less system: how big the heat source is, and how large the flow of water being used is. It should be fairly obvious that if you have a candle and wave your hand over it, you can barely feel the heat from it. Try that with a blowtorch, and you’d definitely notice! Throw in the speed with which you move your hand, though, and even with a blowtorch, if you moved fast enough, you may not burn yourself, implying you did not get much heat transfer. That example vastly simplifies the issue with a tank-less system: to get usable temperature rise, you either run the water though the thing very slowly, or you need a really big burner.
So, how much heat does it take to warm water? Being backwards from the rest of the world, we use British Themal Units (BTUs) to measure the output of our heat sources and degrees Fahrenheit to measure temperature along with pounds and gallons. A few definitions and conversion factors are needed to go much further:
  • BTU – the energy required to raise one pound of water one degree F.
  • Weight of one gallon of water (at standard temperature – it gets lighter as it is heated and heavier as it gets colder, but for our purposes, we’ll assume it stays the same weight) - 8.345404 pounds…we’ll round that off to 8.35 pounds for our discussion
  • BTUs -> Watts – 1 BTU=0.2930711W (this will be enlightening when comparing a burner to an electrical tank-less)
  • Volume of a gallon of water – 231cuin
(continued in next post)
 

Attachments

  • upload_2015-8-26_3-13-30.png
    upload_2015-8-26_3-13-30.png
    422 bytes · Views: 1,302

Jadnashua

Retired Defense Industry Engineer xxx
Messages
32,771
Reaction score
1,191
Points
113
Location
New England
The specification sheets on many tank-less systems sort of gloss over the real-world situation – not everyone’s cold water supply is the same, especially when you consider some of the northern states and how cold things get in the winter! Also note, you don’t have to live way far north to have quite cold water any time of the year if you have a deep well. For example, in my home in August, my ‘cold’ water was measured at 75F, but in late January, I’ve measure it at 33F! Why this is important will show up shortly.
The www.copper.org people (the people that make, specify, and make recommendation on how to do most anything with copper pipe) in this document: http://pbar.fnal.gov/organizationalchart/Leveling/2004%20water%20cage%20work/Cutubehandbook.pdf have recommendations on how fast you should allow water to flow in a copper pipe based on whether it is cold or hot. For hot water, the maximum velocity is 5fps, and up to 8fps for cold water. So, how much water should be your limit with hot if you hold to that maximum velocity? Well, it depends on the diameter of the pipe. For nominal ½” copper, the ID is 0.569” on type M (the thinnest commonly available, and therefore the cheapest, and most often used). Some people prefer to use type L, and type K isn’t used anywhere near as often but the ID of those two are for K and L respectively: 0.527 and 0.545”.
Now, some calculations on how much water can you run through a ½”, type M copper pipe if you want to hold to the recommendations. At 5fps, you’d have a column of 5*60seconds, or 300’ in one minute. Volume View attachment 30489. Keeping things in the same units, converting 300’ to inches = 300*12 = 3600”. Substituting in the formula, that gives us: 3.14*0.2845*0.2845*3600 =915cuin (rounded up). Now, using the volume of water at 231cuin/gallon take that 915/231 = 3.96g! IOW, while you can get away with it short term, from a design standpoint, trying to get more than rounded off to 4gpm of hot water with a ½” pipe is risking erosion of the pipe, extra noises, and shorter long-term life. Let’s do the same thing with ¾” type M copper pipe. 3.14*0.4055*0.4055*3600 = 1859cuin (rounded up) converted to gallons = 8 gallons, or over twice as much!
So, what does that all mean? It means that if you want to shower with a single shower head meeting the federal standards, ½” pipe works since it is limited to 2.5gpm maximum and not all of that is hot. Throw in a second shower head or multiple body sprays, and ½” pipe probably cannot handle it reliably long-term. This discussion is about heating water, but taking that in context of what may be needed helps make the realities more relevant, or at least I think so!
Now, let’s talk about actually heating that water, and how much can you actually get out of any particular unit. As said earlier, it depends on how much you need to raise the water temperature, and, at least in my case, the difference between summer and winter needs can vary quite a bit. Let’s take my summertime situation where my incoming water is 75F and I want to raise it to a typical 120F. That’s 45-degrees. Let’s say I am maxing out the maximum recommendation on hot water and have ½” copper, that would be 4gpm, and heaters are rated at energy use/hour, we have to multiply that by 60, or 240gallons/hour. Using the conversion above, that gives us 240*8.35pounds, or 2004 pounds of water, or just over one ton! Using the expected temperature rise of 45-degrees * 2004 pounds gives us the need for 90,180 BTU, which is well within the capabilities of many tank-less systems out there. But, now let’s take my wintertime situation. Instead of needing to raise the water only 45-degrees, we now need to raise it 85-degrees, so 85*2004 = 170340BTU, which is close to the biggest tank-less system you can reasonably buy. But, the things are not 100% efficient, so you may be pushing it with just 4gpm in the wintertime! 4gpm might let two showers run at the same time, but also consider that in the wintertime, you need more hot and less cold to temper things to a comfortable output (IOW, you’d need more hot in the winter to get the same showering temp than you would in the summer when the cold water is maybe 40-degrees warmer!).
What does this all mean? As you reach the capacity of the tank-less system, it will do one of two things: it will either output progressively cooler water, or, it will restrict the maximum volume so that it can maintain the same outlet temperature.
Now, let’s see what those outputs in BTU would be for Kwhr. Our summertime electrical need would be 90,180*0.293 = 26,493W (26.5Kw). The wintertime need would be 170340*0.293 = 49,910W (49.9Kw). For practical purposes, whole house electric tank-less systems are not particularly viable unless you are a miser and can use something like a 1gpm showerhead.
All of this can get messier if multiple items are calling for hot water at the same time when using a tank-less system.
Let’s look at a ¾” pipe at that 5fps maximum velocity which we saw earlier was 8gpm. 8gpm*60= 480 gallons/hour, or twice what our ½” pipe can supply safely. That’s 480*8.35= 4008-pounds of water. With rounding, at max flow rate, it would take twice our ½” supply, or 180,360BTU in the summertime and 340,680BTU during the wintertime – that’s 170% of the biggest tank-less system generally available. Now, you can run multiple units, but that adds considerably to the costs.
Some might recommend making the max temperature on the tank-less lower, but you can’t go too far because it will cool off at least a little bit from the outlet of the unit to where you are going to use it, and there are things that really need hot water like the dishwasher. If you’re heating with gas, and you feed your dishwasher with say 110-degree water, unless it is designed to heat the water, it likely will not get things very clean and, if it is capable of heating the water, you’d be using electricity verses gas, which, at least in most places is quite a bit more expensive than heating with gas.
Unless you have soft water, a tank-less system will require periodic de-mineralization, typically at least once a year. Otherwise, the mineral buildup will act like an insulator, preventing all of the heat from getting to the passing water, and in the worst case, start to block the flow of water entirely through the unit. FWIW, in a tank, the overall volume starts to go down as minerals accumulate on the bottom of the tank, and it may slow the recovery rate, but it will still achieve the same outlet temperature and there is no gpm restriction other than because of the piping.
I may add more to this discussion as I see questions from people, but this gives you at least gives you some idea of what you might expect with one.
Summary
A tank-less system can work, but it is not magic. The initial outlay can be quite a bit, especially if you need to rework all of your gas supply lines, maybe all the way to the street with a new meter. Some gas companies base their rate structure on a demand factor…the more you have available, the higher your rate (but the cost/therm may be lower), even though you may only be using that higher volume a relatively short time during the day. The system must be sized properly (and that’s true for a tank-type as well), and you may need to manage how much hot water is being used at any one time as even someone turning on the sink to wash their hands while you are peaking out the tank-less system’s capacity, means everyone may notice a dramatic drop in outlet temperature.
 

Jadnashua

Retired Defense Industry Engineer xxx
Messages
32,771
Reaction score
1,191
Points
113
Location
New England
Earlier, I wrote " Our summertime electrical need would be 90,180*0.293 = 26,493W (26.5Kw). The wintertime need would be 170340*0.293 = 49,910W (49.9Kw)."

But, how does that relate to the size of the wiring required to do this in electric? A couple of things to consider, and I think that it would apply here, code requires a circuit to be loaded at a maximum of 80% of capacity when the appliance could run continuously for 3-hours or more (which might happen if you had a big family where the shower was used serially by all of them then maybe throw in the clothes washer and maybe dishes, etc.).

I know math is not many people's favorite thing, but the equation for power = volts * amps, or, since we have the power consumed and the voltage available (in this case, I'll use 240vac), transposing the formula becomes:
amps = power/volts or for that 26493W, it becomes 26493/240 = 110.39A, or, to account for the 80% rule, that's 110.39/0.8 = 138A (rounded). Taking the wintertime needs, that becomes 170340/240 =710A, or again for the 80% rule: 710/0.8=888A! How many people have a unit capable of that much power? This is the reason why IMHO, an electric, whole-house tank-less water heater is a joke. It can work for a small single point of use where there's not a huge volume required, but it's not really viable for a typical US household unless you have really low expectations.
 

Stuff

Well-Known Member
Messages
1,221
Reaction score
130
Points
63
Location
Pennsylvania
I'm not sure of the point of this other than to show that you can prove anything with numbers.

Let's take a tank approach using the same supposition of 4 gallons per minute. 4 gpm will empty a standard 40 gallon tank in 10 minutes. The recovery rate is about 20 gallons per hour with 85 degree temperature rise on a standard 4.5kw unit. So hot water would run out in about 14 minutes.
 

Jadnashua

Retired Defense Industry Engineer xxx
Messages
32,771
Reaction score
1,191
Points
113
Location
New England
Well, in a tank, it would drop off before you used that 40-gallons. The point is, it really depends on how much water you want at a certain rate. It is really easy to overtax a tankless. Say you want to fill a big tub and you really don't want to wait...a tankless probably won't hack it. Most of the tub/shower valves out there do not have a volume control, so you can't slow the flow to maintain temp. Say you're right at the point where the tankless can keep up, but then someone opens a faucet to wash their hands...your shower immediately becomes cooler. On a tank, that doesn't happen.

So, say you want a shower with 4 body sprays at 1.5gpm, and a showerhead and you live in Minnesota...you'd need probably more than 8gpm in the winter of hot, and you decide you can put in a 1/2" pipe...not a good idea, and you probably can't find a tankless to handle it unless you buy two, especially if someone else in the house wants to use some hot water at the same time. At least with a tank, your shower would be hot until you run out. If a long one is important, you install a bigger tank or maybe hot water recovery system, or a faster recovery tank. Or, you decide you can live with one showerhead! The spec sheets imply wonders of endless hot water, and then specify a warm incoming supply, and it looks good until the heat exchanger starts to get covered in mineral deposits that you didn't plan to deal with on a regular basis, or you realize you really need a water softener that you hadn't planned on.

Knowledge of the realities can help you make a viable decision. A tankless system can work wonderful for some, if this helps some decide if it will work for them, great. If they realize that they will need to buy multiple units, upgrade their utilities to make it work, and find someone on the weekend (or even during the week!) that understands how to diagnose and fix one when it eventually dies verses a tank where almost any plumber can figure out, great.

Knowledge is power...my hope was that this would empower some people to make an intelligent decision as to whether this is a good choice for them.

This is the tutorial section, where hopefully, you can learn something. What you do with it is up to you.
 

Jadnashua

Retired Defense Industry Engineer xxx
Messages
32,771
Reaction score
1,191
Points
113
Location
New England
A few things have changed, at least some, since that report mentioned above was written over 10-years ago...tank type WH now are mandated to be more efficient, and new construction, especially if it wants a green certification, will insulate their water pipe runs, both of which help with standby losses. Gas fired tankless systems are now available with higher efficiency ratings, too. But, the limitation on electrically run tankless systems still exists. THey are best suited for a single point-of-use arrangement, verses whole-house where the flow rate may be more manageable. It would be a rare house that could support a whole-house electric tankless without infrastructure upgrades to the panel, maybe meter, and wiring from the pole. You may have a better chance of being able to run with your NG supply line, but not if it was run with 1/2" pipe...you'd have to go back to find where it was larger, again, unless you are interested in a point of use (i.e., smaller) unit verses a whole house. Given the payback timeframe and the fact that NG pricing is trending down, while electric rates are trending up, one really needs to look at both the economics and operational characteristics of any water heating system you decide to choose.
 

Jadnashua

Retired Defense Industry Engineer xxx
Messages
32,771
Reaction score
1,191
Points
113
Location
New England
"Duplicate (system hiccup) post...don't have permission to delete it, so this is what you get!"
 
Last edited:

GregDIYinWI

New Member
Messages
3
Reaction score
0
Points
1
Location
Madison, WI
I know nothing.

Having said that, an easy fix to offset the higher temp rise in the winter is to have a tank in front of the water heater (be it tank or tankless) to passively warm up the water to the ambient temperature of the house/basement.
 

Gsmith22

Active Member
Messages
234
Reaction score
51
Points
28
Location
Central NJ
You resurrected an old post but I wanted to point out that you need to make sure your passive tank has all the insulation stripped of it to allow the "preheating" to happen. I have a GSHP that pre-heats my buffer tank. Winter and summer, my hot water heater doesn't even run as the GSHP puts 130 degree water into the hot water heater tank "cold" inlet via the buffer tank. Passive tank won't get you that level of pre-heat but it might get you from 50 to 65 (average values assuming basement setting).

I should also point out the original post doesn't mention Heat Pump Hot Water Heaters which are far more common now. While they are electric based, they can run at efficiencies of 300% putting them on par with natural gas hot water heating costs. its what my hot water tank heater is.
 
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