Source strangulation for an instant water heater?

[Jonathan212]

Gosh, it is really simple. There ain't enough water to keep the heater running and the cold flow fast enough to compensate for the heat.

 

[sysprog]

Gosh, it is really simple. There ain't enough water to keep the heater running and the cold flow big enough to compensate for the heat.

Do you know what the combined flow rate from the 2 taps is when you max both of them after the heater turns off? (a wristwatch and a bucket can tell you that) (then you can try them one at a time)

 

[Jonathan212]

How are you going to use such figures? I'd rather have some figures for pressure loss as a function of pipe size and length in order to predict what will happen with a tank or a thicker pipe or a pump of given specs.

 

[sysprog]

How are you going to use such figures? I'd rather have some figures for pressure loss as a function of pipe size and length in order to predict what will happen with a tank or a thicker pipe or a pump of given specs.

I want to know how much water you get from each pipe -- you said it was simple, and that "there ain't enough water", so I wondered how much deficiency there might be. You still haven't acknowledged that the temperature of the hot water increases when you decrease the flow rate through the heater, and I was wondering whether if you max out the hot, does it still go below 2 LPM if you then cool the mix with cold, or did you first try to reduce the flow through the hot pipe, and then get too slow a flow after adding enough cold to cool the hot.

 

[Jonathan212]

You start with just hot at maximum flow which is rather slow. You wait a little for it to stabilise to its maximum temperature. Then begin to add cold in the mix till you can stand it. Which is fine in the winter, and fine in the summer too most of the time. But occasionally it is not, the heater switches off as you increase the cold water in the mix. Must be because the public water is a little warm when the weather is warm. Another theory is that too many people are having a shower or watering gardens. Either way, the solution is to increase both cold and hot flow.

 

[Jonathan212]

the temperature of the hot water increases when you decrease the flow rate through the heater

The temperature of the hot water increases, the temperature of the mixed water decreases. This is equivalent to pumping more water through the heater while the cold tap is closed.

 

[OmCheeto]

I'd rather have some figures for pressure loss as a function of pipe size and length

Have you looked at the Hose Water Flow - Pressure Loss graph at The Engineering Toolbox?

I digitized it and generated the following equation, as I didn't see one listed:
pressure drop[psi] / 100 ft = 10^(1.848 * (log_10(gpm)) + ( -2.1041 * ln(pipe diameter[inches]) - 1.3211))

It seems to generate numbers consistent with the graph, so I think it's correct.

 

[Jonathan212]

What is the formula at 90 degree turns?

 

[OmCheeto]

What is the formula at corners?

I do not know what your sentence means. Is English your first language?

 

[Jonathan212]

I meant 90 degree turns of the pipe.

 

[JBA]

If I understand your question there are factors for determining the added loss at elbows in terms of equivalent pipe length to be added. According to Crane Fittings the factor for a std 90° elbow is L/D = 20-30 so the maximum equivalent pipe length to be added to your current pipe length is L = 30 x pipe I.D.

 

[OmCheeto]

I meant 90 degree turns of the pipe.

View attachment 247523

A quick google points to an old PF thread: Pressure drop in elbows & bends in a pipe

I'm having trouble finding an answer in the PDF listed in post #2.
Perhaps tomorrow, I'll devote more time to this.

 

[Asymptotic]

No it is not going to give me 37 degrees, it is going to give me burns because the input water is already a little warm and certainly not as cold as when the 37 degrees in the spec was measured. Asymptotic is one of the people who got it:

It is one of the possibilities. I'm a proponent of measuring things then sorting out what is happening based on the results.

@sysprog's "timed bucket" flow measurement method is cheap and easy, and one of the things I'd be doing if investigating a similar problem. How to go about pressure measurement depends on pipe accessibility and similar peripheral issues ("I want to measure pressure at X location, but can't get at the pipe, so what can I do?").

Assuming the shower head is threaded, I might cobble together a tee fitted with a pressure gauge, and ball valve to go between the pipe and shower head. Measure static pressure with the ball valve closed. Measure pressure (and flow rate; temperature too, if you have a way to measure it) with ball valve open for cold water only, hot water only, and at desired mixture.

Caveat: Static pressure will be the only high quality measurement possible in this configuration. With the pressure gauge located nearly at the water outlet, it will read considerably lower values than if it were located just upstream of the heater unit.​

Solving a problem is a lot easier if you've made a bunch of measurements and developed a sense of what really is going on. Going at it the other way isn't generally as effective.

 

[sysprog]

To what @Asymptotic just said, I would add that cooking and weather thermometers are also inexpensive, and that with the right set of measurements, and @Jonathan212's self-identified comfort zones, one could establish sets of linear and non-linear inequality constraints, formulate LP (Linear Programming) problems, and run Simplex on them, and then do the sensitivity analyses to learn how far one can push one or more of the constraints without departing from a feasible region.

 

[sysprog]

The temperature of the hot water increases, the temperature of the mixed water decreases. This is equivalent to pumping more water through the heater while the cold tap is closed.

If you put 3 LPM through the heater the temperature of the water coming out of it will be lower than if you had put 2 LPM through it. The calorimetric output should be about the same. Is there any tap in your place, cold or hot, that doesn't have what you would think of as a slow flow compared to what you've encountered at, say a hotel room sink?

 

[Jonathan212]

Asymptotic, you do not really have an alternative that fits the observations in #40. Can't blame it on heater malfunction, it is almost new. Can't blame it on a clogged filter, the little mesh filter inside the heater is clear. Can't blame it on anything other than low TOTAL flow. Those 6 kW of power end up in the mixed water, that mixed water must be x LPM for the required temperature rise from 23 to 37 degrees caused by 6 kW (exercise: work out x).

If you ain't got x LPM available, forget it, can't have a 37 degree shower. The heater is only trying to protect itself from too low flow through it that would damage it.

 

[Jonathan212]

the factor for a std 90° elbow is L/D = 20-30 so the maximum equivalent pipe length to be added to your current pipe length is L = 30 x pipe I.D.

Actually I have both elbows and turns in the path, what is the formula for turns of a given radius and internal diameter?

 

[Jonathan212]

This link you gave is about hoses, is there another formula for copper pipes maybe ?

https://www.engineeringtoolbox.com/water-pressure-loss-hose-d_1525.html

 

[sysprog]

The average flow rate for an American shower is about 8 LPM, usually with more available. The 37° in the manual was for 3 LPM through the heater only. Asked and not answered: what is the observed flow rate at your place? I don't see a good reason for your being unwilling to do the simplest of measurements, and report the actual numeric values. The heater starts, so the flow rate initially must be over 1.86 LPM. How much over? Does the cold water line alone show a faster flow than the hot water line alone, or are they about even? What are the numbers? Or at least what is your estimate of the percentage compared to a commonly known range of acceptable values, such as that of a normal hotel room sink cold water line?

 

[Jonathan212]

I don't see a good reason for your being unwilling to do the simplest of measurements

How about this for a reason: you do not grasp that the 37 degrees is up to the taste of the user, that the 23 degrees is up to the weather, that the maximum flow available at the balcony varies depending on what the city is consuming, and that the problem does not always occur and cannot be reproduced at will. So taking measurements without a mathematical foundation to use them would be akin to playing.

 

[sysprog]

How about this for a reason: you do not grasp that the 37 degrees is up to the taste of the user, that the 23 degrees is up to the weather, that the maximum flow available at the balcony varies depending on what the city is consuming, and that the problem does not always occur. So taking measurements without a mathematical foundation to use them would be akin to playing.

According to the manual, the "typical" ($\leftarrow$ that word is there due in large part to the variations in input water temperature) warm water temperature of 37 degrees depends on whether you ran 3 LPM through the heater. If your "taste" is for a hotter temperature, you can use a lower flow rate. The mathematical foundation for using such measurements is anything from 4th grade arithmetic, through linear programming, through multivariant linear regression analysis, etc.; however, having a numeric sample of the 'typical' flow rates gives your readers here a better idea of the situation than does use of such subjective words as "slow", or sometimes not as slow. And by the way, where do you live that's not equatorial (inferred by your having a winter and a summer, as distinguished from year-round hot) that has a municipal water supply that runs at 23° in the summertime?

 

[Jonathan212]

You are misapplying the manual: the 37 degrees-3 LPM pair assumes a standard input temperature that is not available here so it is a pointless thing to say or test for.

 

[sysprog]

You are misapplying the manual: the 37 degrees-3 LPM pair assumes a standard input temperature that is not available here so it is a pointless thing to say or test for.

So where do you live (I'm not asking for your hometown) that has a 23° water supply?

 

[Jonathan212]

I do not know if I have 23 degrees water supply, that was an example for the mathematically challenged.

 

[sysprog]

I do not know if I have 23 degrees water supply, that was merely an example.

[edited]

Is that what you meant? You could get a regular thermometer and run the cold water over it and then you'd know what the input temperature to your heater is and you could then tell us.

 

[Asymptotic]

You are misapplying the manual: the 37 degrees-3 LPM pair assumes a standard input temperature that is not available here so it is a pointless thing to say or test for.

Outlet temperature (37°C), flow rate (3.0 LPM), and power (6KW) are given. Inlet temperature can be calculated, and works out to approximately 8°C.

8°C (48.4°F) seems a reasonable value. Several years ago at the plant I measured city water temperature at the mains meter from early January to early August to get a sense of seasonal variation, and found it ranged from about 40°F during a cold snap to 70°F during a heat wave.

 

[Asymptotic]

How about this for a reason: you do not grasp that the 37 degrees is up to the taste of the user

37°C is from the specification table.

So taking measurements without a mathematical foundation to use them would be akin to playing.

This is an unusual perspective. Would Johannes Kepler have developed the math underlying the laws of planetary motion without his and Tycho Brahe's observational data?

 

[sysprog]

Outlet temperature (37°C), flow rate (3.0 LPM), and power (6KW) are given. Inlet temperature can be calculated, and works out to approximately 8°C.

8°C (48.4°F) seems a reasonable value. Several years ago at the plant I measured city water temperature at the mains meter from early January to early August to get a sense of seasonal variation, and found it ranged from about 40°F during a cold snap to 70°F during a heat wave.

In Chicago, the source is Lake Michigan, far enough out and deep enough down that even by the time it gets to residential areas 10 miles away from the filtration plant, even during a heat wave, it's still pretty cold, around 50-60°F.

 

[sysprog]

37°C is from the specification table.
View attachment 247567

So taking measurements without a mathematical foundation to use them would be akin to playing.

This is an unusual perspective. Would Johannes Kepler have developed the math underlying the laws of planetary motion without his and Tycho Brahe's observational data?

Yeah, and what if Galileo had said he wouldn't look in a telescope until Newton came along with a mathematically justified prediction of where each planet would be when (inconvenient, given that Newton was born slightly earlier in the same year as that in which Galileo met his demise), oh and yeah, Newton made some 'preliminary observations' too.

n.b. @Jonathan212 said he had the DH106101 (one column to the left of the one indicated by your arrow), the specs for which are exactly the same, except that the fittings connection is 1/8" on the DH106101, and 1/2" on the 111, and no, to anyone who might suppose otherwise, that doesn't make a whit of difference, except possibly to robustness, and the 111 has a green 'on' lamp, which also has zero effect on performance.

 

[Asymptotic]

n.b. @Jonathan212 said he had the DH106101 (one column to the left of the one indicated by your arrow), the specs for which are exactly the same, except that the fittings connection is 1/8" on the DH106101, and 1/2" on the 111

You're right. My bad :)

 

[Jonathan212]

The math is 100% known for this problem. Nothing to do with Kepler's situation.

But it is not known to any of us here. Yet. We're still missing the copper equation as opposed to hose, the circular turn equation, not sure about the elbow if it is copper rather than what? Once we have everything down, then is the time for measurements, we will then know which set of measurements is sufficient and which is not. And hopefully skip the gauge purchase. Then I would predict what a thicker pipe would do or size a pump.

 

[sysprog]

The math is 100% known for this problem . . . But it is not known to any of us here.

That remark strikes me as extraordinarily presumptuous.

 

[jrmichler]

I just read through this thread. Except for a couple of digressions, the discussion fully covers what is going on with the water heater. The discussion also covers solutions.

While I see no need to calculate pressure drops, it would not hurt to do so. Calculating pressure drops is a good way to better understand the water system. For finding pressure drops in piping systems, the best resource is Cameron Hydraulic Data. The latest edition is quite expensive, but the earlier editions are just as good for water flow through iron and copper pipes. In the U.S., it is readily available by interlibrary loan if one does not want to buy it.

 

[Asymptotic]

I just read through this thread. Except for a couple of digressions, the discussion fully covers what is going on with the water heater. The discussion also covers solutions.

While I see no need to calculate pressure drops, it would not hurt to do so. Calculating pressure drops is a good way to better understand the water system. For finding pressure drops in piping systems, the best resource is Cameron Hydraulic Data. The latest edition is quite expensive, but the earlier editions are just as good for water flow through iron and copper pipes. In the U.S., it is readily available by interlibrary loan if one does not want to buy it.

Several editions of Cameron Hydraulic Data are available to borrow at archive.org, and older editions can be very reasonable to purchase. For example, Abebooks lists 30 volumes ranging from $9.95 to $52.
https://archive.org/search.php?query=Cameron Hydraulic

The Copper Tube Handbook is also a valuable resource.
https://www.copper.org/publications/pub_list/pdf/copper_tube_handbook.pdf

 

[Jonathan212]

Surprise. Three different kinds of copper K, L, M with pressure drops of the order of 45% higher in K compared to M. Looks like the type of copper will have to be discovered too.