I installed HaaS on PI4/2GB, spent 2 weeks setting up the system, and currently it is working with Zigbee temp and wall socket switches/monitors and Thermopro 433Mhz wireless temp sensors.
Now, one of the next project would be to understand what my monthly water usage looks like. We are living in a remote area, basically, the only utility connection we have is electricity. Water is coming from 200ft below ground via a well/pump (240V) and there are water filtration systems (Evolve, see attached photo at the bottom) attached to the pressurized tank. These filters do not display usage information.
What would to affordable (I don’t want to spend >$50 on the bill of materials) to do this?
From what I am reading, it seems that I need to get a hall effect water flow sensor (preferably a metal one I guess), and then connect to something to measure pulses and do a calibration. I have unused PI Zero WH, which I can repurpose. And after that push the info to HA whatever means available.
Anyone has done a similar project? If so, could you please share your setup/ideas/observations?
Sorry, can’t help you with a water flow sensor, but I just attached a PZEM-004T to an esp flashed with Tasmota to check power usage for my hot water. If you measure how long the pump runs for, I would guess this would give you a good enough measure of how much water you are using.
I agree that a smart socket with current sensor, such as the TPLink HS110 for around £15, is a quick and cheap way to measure overall usage. You could probably even get rough calibration on how much water is pumped per minute of operation.
For more accurate usage, then yes, you’d need to add a flow meter with a counter. A weimos or ESP8266 connected to them (the meter comes with a wire attached that simply closes a circuit every revolution. Once added, you need to do some calibration tests to work out exactly how many revs in a 100 litres or whatever)
Beware some of the cheaper models of these meters, which are not well made. I’m not sure I’d trust them in my main supply.
I’m not sure that with a pressure pump, you’d get a linear relationship between runtime and volume pumped. It seems to me that as the pressure rises, the volume moved per unit of time would decrease. But I’d love to be proven wrong on this. It would be a simple solution!
measure the power utilization of the pump and using the pump’s specs - how much water it pumps per minute, calculate the usage.
attach hall effect sensor in piping: water going from the pump to the pressure tank and calculate actual usage.
The challenge with the first approach is the pump is 240V 60Hz and I don’t know which power utilization sensors would be suitable for such an approach.
The challenge with the second approach is that it requires additional components as well as more effort going into it - like putting things together, writing code, etc.
Let me start with the first one and see how much can I progress.
Thanks for the ideas.
True, this requires some work and you need to (want to) like it. There are plenty of examples using this (similar) hall/pulse sensor with a ESP32/8266 device so this will get support… but you do need (!) to do a bit too and not everyone wants to go this path with chip-boards and soldering (which I fully understand)
I think you’re absolutely right. Higher pressure means more effort to move the same amount of water.
Also, unless the non-return valve is very good, there might be a warm-up time involved where it’s sucking air or building pressure.
But - OP’s on a budget, and the only water meters with rotation sensors I’ve seen that I would trust not to split or fall apart cost more than his total budget, so I think accuracy costs money.
