Author Topic: WHO USED ALL THE HOT WATER!!! Need a simple temperature data logger ** UPDATE **  (Read 9439 times)

0 Members and 1 Guest are viewing this topic.

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: WHO USED ALL THE HOT WATER!!! Need a simple temperature data logger
« Reply #75 on: August 23, 2023, 01:39:31 am »
Another possibility would be a microphone or vibration sensor to detect if water is flowing in the hot water pipe.

Acknowledgement, apologies and belated thanks to ejeffrey for being the first to present this approach.
« Last Edit: August 23, 2023, 02:19:23 pm by Brumby »
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: WHO USED ALL THE HOT WATER!!! Need a simple temperature data logger
« Reply #76 on: August 23, 2023, 07:44:47 am »
Well, I had decided to purchase an Inkbird IBS-TH2 with remote sensor a few days ago, as it was cheap and ready-to-go.  It arrived today.

I've set it up with the sensor simply taped to the hot water line just before the tempering valve.

Let's see what I get...
« Last Edit: August 23, 2023, 08:00:19 am by Brumby »
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Interesting.....

First data (exported from App, dropped into Excel and charted. Chart chopped up for ease of viewing)

Labelled points:
A. First power up and check.  Body heat used for temperature increase.
B. (Some time later) Took it outside
C. Attached sensor to pipe. Warm pipe caused increase until reaching equilibrium between sensor and pipe
D. First use of hot water - Washing of hands
E. Subsequent uses - hand washing I believe
F. Filling sink for dishwashing
G. Small spike - hot tap turned on for about 2 seconds, if that
... etc. etc.

No showers or drawing of baths during this period.  I am curious to see how they come up.


« Last Edit: August 23, 2023, 02:20:04 pm by Brumby »
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7681
  • Country: ca
Where is the sensor located? On the hot water supply line of a sink?
Scalding guidelines are 49°C (120°F) and others use 60°C (140°F) setpoint to prevent bacteria like Legionella in the tank.
 

Online IanB

  • Super Contributor
  • ***
  • Posts: 12539
  • Country: us
Where is the sensor located?

I've set it up with the sensor simply taped to the hot water line just before the tempering valve.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7681
  • Country: ca
It's the huge time-constant... bigger than I expected for cool down of a sensor on copper or iron pipe. So I thought maybe on a manifold of the sink? Here we use 1/2" copper pipe and plastic braided supply line to the tap.
Consecutive water use, I don't think you can ID the culprit with such slow data. Morning showers I imagine it's a lineup, rush and person after person.
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
It's the huge time-constant... bigger than I expected for cool down of a sensor on copper or iron pipe. So I thought maybe on a manifold of the sink? Here we use 1/2" copper pipe and plastic braided supply line to the tap.
Consecutive water use, I don't think you can ID the culprit with such slow data.
I will ask you two questions:
 1. At what point can you say a particular water usage has begun?
 2. At what point can you say a particular water usage has ended?

Quote
Morning showers I imagine it's a lineup, rush and person after person.
There is usually a minute or two as people dry off and get decent for the public traverse.  It's not as if party B is stripped off ready to jump into the shower as party A reaches to turn the water off.
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Here is a photo of the setup:

Criteria used:
1. No idea what the optimum setup would be, so wasn't fussed about over-thinking an initial configuration - just wanted to get some data to get an idea of the real world
2. Quick and easy
3. Temporary fixture OK for starters
4. Placed on the hot water outlet pipe to gain maximum temperature swing because:
  a. This would allow easier identification of temperature increases
  b. This would produce the fastest temperature drops when water flow ceased because of the highest difference between pipe temperature and ambient than other locations
5. Placed as far as possible from the tank, to reduce bulk heating effects from the stored water
6. Sampling rate set at maximum available (every 10 seconds) to get best resolution possible
7. Transmitter placed on tank as high as possible for best Bluetooth range (it has magnets  :-+ )
« Last Edit: August 24, 2023, 01:21:37 am by Brumby »
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Other considerations for future refinement:

* Place secondary sensor at tank outlet point to see if there is a detectable difference which might indicate flow rate.  (Higher difference implies lower flow rate; lower difference implies higher flow rate.)  I think it doubtful there is enough resolution to achieve this ... but it might be worth investigating

* Fit sensor via a clamp.  While being more permanent, this allows investigation of thermal coupling options.  Thick metal and/or short length provides greater coupling, resulting in the sensor seeing temperatures that are close to the actual pipe temperature.  Thinner metal and/or longer length will reduce that coupling and result in the sensor being more influenced by ambient air temperature.

Maybe something like this...   

In this latter case, while the temperature measured will be slightly less, it will still be notably higher than ambient and a continuing flow of water will maintain this temperature.  When flow stops, the lower coupling will result in the ambient air cooling the sensor more quickly, giving a much more visible indication.  Since absolute accuracy of temperature measurements is of no real importance here, this is not a problem.

* Custom project where:
 - More accurate sensors could be used and configured for best result in targeted temperature range
 - Only recording samples of interest - say 5ºC above ambient
 - Potential for custom software, including computation
« Last Edit: August 24, 2023, 02:24:46 am by Brumby »
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7681
  • Country: ca
I'll ask you once - what temperature delta constitutes the next person and imprisonment?  ;)   One time constant looks like about 7 minutes which may lead to mistakes if the heater turns on temp kicks up possibly and your window is only a few degrees.
I understand you can figure out shower duration by looking at the derivative of pipe temperature. I think you could get one more piece of info by sensing at the shower mixer outlet instead of the hot supply. I'm not sure why you have a mixer right at the tank.

Ideal shower temperature is anywhere between 37°C (98°F) and 40°C (104°F) and no more than 41°C (105°F).
Old showerheads were 5.5gpm or 20.8l/min. US national standard of maximum 2.5 gpm flow rate for shower heads was established through the Energy Policy Act (EPAct) of 1992. California is less at 1.8gpm.

Could I scam your system by taking a shower with master blaster high flow rate, or high temperature? Your graph would not tell you this.

A little math you can check - a 7 minute shower, 50% hot 60°C/15°C rest cold water to get 37.5°C is about 4.73l (10.4l  with high flow head) of hot water. But I like it hot, 42°C would use 5.6l (12.5l) of hot water. Not a big difference unless the cold water is much colder.
Flow.. well it's the big unknown.
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
This data is from the simplest, quickest and easiest install possible.  You have to start somewhere.  There's no point building a Cray to see how well your binary adder works.  Refinement will follow - if it is required.

Even so, this unsophisticated setup has delivered some interesting data.

From what I can estimate, thermal inertia is about 30 seconds for the current setup.  Differentiating hot water events a minute apart has not been difficult with the data collected so far.  Even then, if there are a number of short events close together, then the amount of hot water involved is not significant, so distinguishing each of them is unnecessary.  If there is a need to get so particular, then the security footage is timestamped as well and it would be trivial to match traffic and temperature ... albeit a bit time consuming.

7 minutes is way, way outside any time constant resolution required.  The first 30 seconds after water has stopped flowing is clearly visible and unambiguous.  If water starts flowing again within that 30 seconds, then the inertia will still result in a small, but visible dip unless it was only off for a few of seconds.  In that case, it is highly likely it is the same person ... and in the case of the (rather unlikely) need for full clarification in such a situation, security camera footage is available.

It would be good to get the water flow rate, but it's not clear whether such information is necessary to achieve the desired goal.  That goal, by the way, is not to convict, but to initiate a conversation.

Please keep in mind the purpose of this exercise is not to track every millilitre of water and every millijoule of energy so you can bill each person for the cost.
 

Offline Dundarave

  • Regular Contributor
  • *
  • Posts: 160
  • Country: ca
Have you considered simulating actual showers of different lengths in order to determine typical usage “signatures”?

Once you have several known patterns of actual usage, you ought to be well equipped to interpret accurately (i.e. accurately enough for your purposes) the real-life usage patterns that get recorded.
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Tank heating is done on an off-peak tariff, so the element is switched on at given times, depending on thermostat.

One further bit of data from overnight... tank heating.  Because the sensor is on the outlet pipe, heat conduction from the tank does affect the temperature it reads.  During tank heating, the temperature profile was quite distinctive.  The sensor temperature did rise, but much more slowly than when hot water flows and it levelled off at a lower temperature.  At the end, the temperature drop was again quite different to flow cessation.  The fact of the time of day and the 3 hour duration distinguished this event from any other.
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Have you considered simulating actual showers of different lengths in order to determine typical usage “signatures”?

Once you have several known patterns of actual usage, you ought to be well equipped to interpret accurately (i.e. accurately enough for your purposes) the real-life usage patterns that get recorded.

Absolutely!

The first step is to see if we can get some data which has sufficient distinguishing features to allow such profiling - which we seem to have.

I have already done several "hand washing" exercises to see how they show up.  I have also been making notes of when others have done things of note (See the chart above).  Today I have noted an honest-to-goodness shower event!

Special exercises may come further down the track if there is need for clarification.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7681
  • Country: ca
What are the chances the tank is undersized for the number of people taking showers? How big is it.

Looking at US hot water tank manufacturers, their recommendations seem a bit nutbar:
For 1 to 2 people: 30-40 gallons
For 2 to 3 people: 40-50 gallons
For 3 to 4 people: 50-60 gallons
For 5+ people: 60-80 gallons

I calc 20 US gal (76L) is 8 minutes with a low flow 2.5gpm shower head.
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
What are the chances the tank is undersized for the number of people taking showers?

As has been stated on more than one occasion ... that is not the problem being presented here.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7681
  • Country: ca
To me it looked a little small, that tank diameter...
Datalog should be able to see the recovery time, if the system is working well or not.
I'm not sure how hard your water is, but here it's common for limescale and calcium buildup to impair heating.
Every year or two, we do an acid flush to get the scale out for natural gas heaters, they make a sound like popcorn popping due to local boiling inside.
Electric water heaters the element gets scale buildup and it can't transfer the heat, which eventually causes its demise from the heat.
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
As has been stated on more than one occasion ... that is not the problem being presented here.

Again, I say, the performance of the hot water system is not the problem being presented here.
 

Offline mendip_discovery

  • Super Contributor
  • ***
  • Posts: 1024
  • Country: gb
You might want to monitor the ambient temperature. The speed of cooling may be affected by the ambient temperature.

Can you fit some lagging to the hot water pipe so that you loose as little heat from it when in use.

Have fun logging the data.
Motorcyclist, Nerd, and I work in a Calibration Lab :-)
--
So everyone is clear, Calibration = Taking Measurement against a known source, Verification = Checking Calibration against Specification, Adjustment = Adjusting the unit to be within specifications.
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
You might want to monitor the ambient temperature. The speed of cooling may be affected by the ambient temperature.
It certainly will, but for the current measurement the speed of cooling isn't a big issue - just the fact that cooling can be identified is the important point.
 

Offline AndyBeez

  • Frequent Contributor
  • **
  • Posts: 858
  • Country: nu
Again, I say, the performance of the hot water system is not the problem being presented here.
Yes, they are conducting a surveillance operation against an identifiable but gender unspecified individual who is apparently using too much hot water and eating too much lasagne. So issues about heat capacity, thermal efficiency, shower mixer design and pipe lagging, don't count as this project is just about nailing a pesky perp. Personally, I would just put up a notice telling all gender unspecified users of hot water users to kindly shower 'sans-vapeur' - no steam please as the Stasi are recording your every joule.
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Personally, I would just put up a notice telling all gender unspecified users of hot water users to kindly shower 'sans-vapeur' - no steam please as the Stasi are recording your every joule.

Oh that such a simple action would be effective.

This person would ignore such a sign and probably shower longer to spite.  If another member of the household called them out, they would deny vehemently even if they'd been in the shower for an hour.  The ensuing toe-to-toe would simply escalate as the guilty party would not back down and a third party intervening would have a situation of one word against the other.

Do not underestimate the determination and conviction of an entitled bullshitter.
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
 :palm:

Just as I was looking at some figures, wondering what effect solar heating of the pipework/building might have to the numbers, I received some news.....

The party concerned was a guest.  One that had abused that hospitality in far more ways than just excessive hot water .... and .... they have vacated, finally.


Now I have no need to pursue this any further .... .... but this experiment still has its intrigue.
 

Offline richnormand

  • Supporter
  • ****
  • Posts: 721
  • Country: ca
Repair, Renew, Reuse, Recycle, Rebuild, Reduce, Recover, Repurpose, Restore, Refurbish, Recondition, Renovate
 

Offline BrumbyTopic starter

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Interesting.

That unit has a few more data points than the one I have (32510 vs 30001) and you can select an interval all the way down to 1 second (vs 10 seconds).  It can also use other sensors.  For my use, not a big deal.

On the flip side, it's about 4 times the price.  Also, it does not have any wireless communication, so it's a retrieve and plug in exercise which does not suit my application as well as the one I stumbled across.  It also doesn't seem to have humidity measurement - but that really isn't of much interest for my use case here.

Thanks for the info.  It's always nice to see what else is out there.   :-+
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf