Group Actions
Engagement
  • Author Author: KalebTheMaker
  • Date Created: Date Created
  • Views 12713 views
  • Likes 12 likes
  • Comments 58 comments
Related
Recommended

Project - Water Tank Monitoring System

KalebTheMaker

Hi all,

On Bonesnapper Ridge, I have a fairly large system for storing water and I would like to monitor the entire system. This topic started to take off on another thread, so I thought I would make a new post to keep it all in one place and give a little more background.

Wells

There are multiple wells on the property, but I am focused on three of them. These are "horizontal wells" or tapped springs. They are drilled horizontally into the side of the hill and are gravity fed to a set of tanks at the bottom of the property. These wells flow 24/7 365 days a year. The wells flow into the tanks, and then there are overflow pipes that overflow into a culvert. As I said, 24/7. Maybe I could install very small turbines on these to generate a little bit of power?

Tanks

There are two main tank sites on the property. There are three ~2500 gallon tanks at the well site which is at the lower elevation of the property. There are three more ~2500 gallon tanks at the highest elevation of the property. There is also one 2500 gallon tank right in between these two sets of tanks in the middle of the elevation. The bottom tanks are plumbed all the way up to the top tanks and flows through that middle tank. Note: The lower tanks are 100% full until I pump, then it takes about 8 hours to fill up what I pumped out.

Pumping

To get water from the lower tanks to the upper tanks that feed the house/shop, currently, I have to drive down to the lower tanks and fire up a gas-powered honda water pump (currently broken... yay another project!). On a full tank of gas, this can pump something like 1800 gallons. So I would have to pump several tanks of gas worth to fill up the upper tanks. I have to open and close a few valves manually in this process.

Water Usage

I am surprised at how little water I have been using. I don't have laundry up here yet, so that will add a bit, but as of now, it's quite low. Last time I pumped I marked it on the upper tank before I pumped, and then I marked it off after the pump. That's where I came up with the approximate 1800 gallons on a tank of gas. It has been a few months since I have pumped water, and there is still plenty in the upper tanks.

Sensor Data & System Design

My goal at the end of a few projects is to have sensors in each of the tanks ( seven tanks ) and report that back to MQTT or an InfluxDB database here at the house. Build a solar or hydro-powered pumping system at the bottom tanks, so that when the top tanks get below a certain threshold, the pump will automatically turn on and pump water until a threshold is met. 

What data do I need out of the tank sensors?

The nerd in me wants it down to the gallon. But that's going to be more difficult to achieve, and in the system design, I don't really need that. I can probably get away with down to the 100 or 1000 gallon. The resolution will determine how often I pump, and the design of the pump system will also be a factor in how often I can pump. for example, if the pump and battery can pump 100 gallons on a full charge, then I have to know how long it takes to gain a full charge on the battery using either solar or hydro. Then I can set my thresholds in the system for how often to pump. I'll probably end up looking at the state of charge of the pump batteries and determining if it can pump as well.

Maybe for now, I will have to settle for somewhere in the middle and have a minimum resolution of 500 gallons. If it's lower than that great, if it's higher? Well, I will just have to deal with it. 

The order of execution

I want to get the sensors built first, and then the rest of the monitoring system and the pumping system will come at a later date. I would like to get all of it done this summer, but there is a LOT to do here and it's not critical. But the senors are critical to my data-hungry brain, and sore legs from hiking up the hill to check water levels. 

I would love to hear your feedback on this! Some of this topic has been discussed on the thread below, so please check that out. 

/challenges-projects/element14-presents/b/off-grid-makershop/posts/off-grid-maker-shop-general-update

I will add some pictures first thing tomorrow to this post. I'm in a rush to get out of this house this morning. 

Well and Tank Pictures:

Upper Tanks

image

Lower Tanks  

image

Primary Well

image

Secondary Well

image

Tertiary Well

image

The second and third wells are tied together, they do not have as much output as the primary well. 

Elevation Data

Site Elevation (m) Elevation (ft) Barometric Pressure (mb)
House 672.7 2207 1019
Upper Tanks 698 2290 1016
Middle Tank 646.8 2122 1022
Primary Solar Site 628.5 2062 1024
Lower Tanks 622.7 2043 1026
Primary Well 620.6 2055 1025
Secondary Well 00 626.4 2066 1024
Secondary Well 01 630.6 2069 (eyeballed) 1024

NOTE: Updated for the third time. 

For this data I took my Garmin 64s out with barometric pressure turned on. When I got the area the elevation would fluctuate wildly starting low, then gradually going up until it stabilized. This took about three minutes. I waited at each site for appx three minutes until it stabilized. I recorded the barometric pressure for Dave. 

However, The difference between the "Primary Solar Site" and "Lower Tanks" being ~6' does not seem right. The road from the solar site to the wells and tanks goes downhill for quite a ways, then up just a little. There is no line of sight, and I could be deceived, but I really do not think so.  "Lower Tanks" to "Primary Well" seems off as well, data shows ~12', but I think it's more like 5-8'. 

Thanks!

-Kaleb

  • in reply to KalebTheMaker

    How easy is it to get M12 to 1.25"  PVC pipe adapters off-the-shelf ?

    Other threads are available.

    image

  • in reply to KalebTheMaker

    Hi Kaleb,

    If all the tanks are top vented and if they are all connected together without valves or regulators at the bottom, then the water in all tanks will automatically find the same level.

    Take a clear plastic flexible tube, hold in a U-shape and fill it 3/4 full with water. No matter if you raise or lower either end of the tube, the water within the tube at both ends will remain at the same level relative to each other.

    That is what is going on if you connect two tanks with a pipe - fill one with water and the water will rise at the same rate in the other tank such that the water level remains the same in each tank.

    If one tank is slightly higher than the other then it's just like raising one end of the plastic tube above that of the other. The plastic goes up but the water at each end remains at the same level. If you are measuring level relative to the top of the tank e.g. ultrasonic/lidar then you need to keep that in mind.

    The head is the difference in height from where you measure to that water level.

    You aren't measuring the 'sum of all tanks' as such, rather the height of the water which just so happens to be the same across all three tanks. However, knowing that the water will self level, if you know the head and the diameter of each tank connected to the system, you can work  out the volume of water in each.

    The diameter/number of the tanks is irrelevant when measuring the head pressure. 

    Your tanks will be empty when your head is down to 83' (25.3m) but you will still have a 83' (25.3m) head of water sitting in your supply pipe. Your gauge will drop rapidly at this point as your supply pipe drains quickly.

    Your gauge/transducer at the house will need to be able to accommodate that additional pipe head pressure though, then the full tank pressure on top of that.

    Your calculation depends on what sort of transducer you are using.

    3.73bar if you are measuring absolute (relative to a vacuum)

    2.73bar if you are measuring 'gauge' (relative to air pressure)

    Sounds like you are on the right track though.

  • in reply to beacon_dave

    Hi Dave, 

    This has become somewhat of a priority at the moment. At least in the sense that when I woke up this morning there was no water to make coffee! I think I at least need to have a percentage indicator of how much water is in the "system" rather than per tank. 

    I have a friend who is across the valley, he has a really cool old analog pressure gauge that he mounted in the wall of his home which tells him the pressure of his system. It was made for something else, so he only uses about 15% of the gauge, but has it marked with Sharpie. It has worked for him for a very long time, so I thought maybe the simplest answer is the solution in this case. I would love to have each tank monitored, and maybe I'll do that someday, but for now, a percentage will do. 

    And, I think this is the way to do it. I can easily put a pressure transducer in the water main near the house. But, I am trying to understand how to calculate head *correctly* in this type of system. I think what I am reading is that it does not matter what the diameter of the water reservoir is, it's just about the height of the water in the "system". So, I would calculate the height of the water inside the tank, let's call it 8' (2.43m). Then the height of the 1.25" PVC from the tanks to the measuring location would be appx 83' (25.3m) for a total of 91' (27.7m). That would be when the tanks are full, and then 0 when empty. 

    Am I correct that I would measure the sum of the tanks and the PVC?

    According to: https://www.omnicalculator.com/physics/hydrostatic-pressure#fluid-pressure I would be looking at a hydrostatic pressure of 3.73 bar for a full system.

    With that data, I can spec a pressure transducer with a maximum of ~4 bar. 

    Does that sound right to you?

    As a side note, I have adopted MQTT as an intermediate destination for all of my sensor data. I can use simple microcontrollers (esp32 etc) to take the measurement and send it to MQTT. From there, any device can read real-time data. I have also written a daemon that runs on a raspi that captures that real-time data and drops it into an InfluxDB database for time-based historical data. 

    Thanks,

    -Kaleb

  • in reply to beacon_dave

    The sensor appears to go under the code of TL-316 (new style?) or TL-231 (old style?)

    They appear to be available as 3m versions which is closer to your 2.5m tanks but are listed as 24VDC.

    Prices are lower than the DFRobot one but they aren't supplied with a 4-20mA current loop to voltage module.