Water Distillation Project

@davidmcsweeney

Obviously in Australia, we have access to much better quality materials and tools, and you can purchase activated charcoal rather than making it yourself.  It's the idea created by "Prof. Josh Kern" that makes it a great off-grid project 🙂

There are more video's and information at: charcoal-biochar-water-treatment

Cheers.

@frogandtoad

Looks very interesting. Thanks again

You've already made a good start with the system engineering you've done so I will only offer some comments:

1/ re batteries.  I would use two batteries, one lightly loaded and hopefully more reliable feeding the electronics and another independent one for power components.  This way a discharged battery with near zero volts will not disable the electronics needed to charge it.  The battery used for electronics could be on a constant float charge fed by a small solar panel and would last forever.  I used this for my boat battery for years and it worked very well. (and I expect we get a lot less sunlight than you do.)

The circuit diagram you've picked up for battery charging uses an Arduino circuit board for charge control but there is no mention of software for the Arduino.  For a lead-acid battery a commonly used algorithm is: charge at constant current until the voltage rises to a preset level then charge at constant voltage until the current falls to a preset level then change to a constant low level current for the float charge.  Other batteries such as LiPo will require a different algorithm.    

2/ re solenoid valve.  A DIY job sounds like an invitation to pesky problems and fiddling. Take a look at Adafruit product code 996 and 997 for a ready to go 12v solenoid valve which I hope is in the size range you need.

3/ re borehole water.  Where I live deep well water water is normally quite hard.  OK for drinking and cooking but requires a lot of soap for washing. But it quickly leaves a deposit of lime in any plumbing carrying it so if yours is similar you may have to think about an acid flush.  Your mention "flecks of iron" makes me think you well water is also quite hard and the flecks you see are precipitating out when the pressure falls.

That's enough for the moment but I'll probably think of more to expound about later.

Foxy

Another thought

4/  re plastic.   I've found that plastics are sensitive to sunlight in spite of the "last forever"  mythology.  The transparent acrylics I've used, plexiglass, perspex tend to develop crazing (tiny surface cracks) after a few years use in sunlight.  Transparent polycarbonate (lexan) develops a frosted surface after a few years in sunlight.  Both still pass light but it is diffused and (I expect) weakened in the transmission.

5/ re collecting roof rainwater.  I hope you don't have a lot of birds roosting on your roof.

@foxy

Hi foxy,

Thanks for your two posts and excellent insights. I've been so busy lately that I have been away from the forum for a bit.

I'll go through your points as best as I can.

1) Two batteries - yes, I have already been thinking about this solution and will most likely use this in my final design.

Here is a copy of the code (written by the author of the design):

#define charge 4 // pin 4 controls relay
#define b_volt A0 // battery voltage read by A0

float bat; // stores battery voltage

float low = 11.9; // cut-in voltage (begins charging if bat < this)
float high = 14.5; // cut-out voltage (stops charging if bat > this)
float gain = 3.69; // adjusts voltage divider gain for accurate voltage

void setup() {
// put your setup code here, to run once:
Serial.begin(9600); // open serial port and set baud rate

pinMode(charge, OUTPUT); // sets charge pin as output
}

void loop() {
// measures battery voltage and does basic filtering to smooth out noise
// each reading is averaged into 49 previous averages

bat = ((analogRead(b_volt) / 1023.0 * 5 * gain) + (bat * 49)) / 50;

Serial.println(bat); // display batery voltage

if (bat < low) // if battery is below cut-in voltage, begin charging
{
digitalWrite(charge, LOW); // system charging when relay is off
Serial.println("Charging"); // tell me this
}

if (bat > high) // if battery voltage is above cut-out voltage, stop charging
{
digitalWrite(charge, HIGH); // system not charging when the relay is on
Serial.println("Not Charging"); // tell me this
}

}

I've run it and it seems to work fine.

2) Solenoid valve - I've had a look at the Adafruit products and both look good. I think the plastic one is better suited to my project because of the gasket arrangement inside.

3) Bore water - our water comes from a deep (60 metre) underground reservoir and is actually very potable and "soft" except for the flakes of iron in it. Because Western Australia is very sandy generally, I guess this has the effect of being a wonderful natural filtration system.

4) Plastic - yes, I am currently looking into various materials from which to build the unit. My first thought was to build it of Perspex/acrylic but have since looked at Aluminium with an Acrylic lid (which could easily be replaced if necessary - according to the supplier, clear Acrylic is UV resistant). I have talked to a company that supplies and fabricates Aluminium and would like to get a smaller prototype constructed to test my theories first. I've also considered Fibreglass but have yet to talk to anyone about this although I have found a company that deals with this as well. 

5) Oddly enough, we have so many trees around that I rarely see birds on our roof.

I've ordered a weatherproof ultrasonic sensor so my first test is to see how effective this is in measuring water height. The literature says it has a 0.5 cm resolution. I am hoping to have only 10 or 20 mm of water in the unit at one time to speed up the evaporation process. It should be here in about a week so I can test my theories then. From there, I will design a prototype storage unit and see what sort of evaporation rates I can get at this time of year (in the heat of summer - we are currently having 35-40 C days most of the time).

 So I'm taking small steps in between everything else I am doing. As with all things, there's always more to discover as you go along.

Cheers,

David

@davidmcsweeney

Re batteries.   Take a look at https://batteryuniversity.com/learn/    for information on batteries.  The various battery chemistry technologies need different algorithms for charging.  Seems to me if you use lead-acid batteries unattended and don't have the charging right you run a risk of decomposing the water so they run dry. Good luck on this.

Interesting your deep well water is reasonably soft -- I can only guess the sand you mention is silica as this is pretty insoluble.  Where I live the countryside was under a mile or so of glacier ice 15000 years ago which ground up some of the limestone bedrock to powder so the soil is saturated with it and at any appreciable depth the limestone is somewhat soluble which makes for very hard water when you pump it up.    Any idea what the "iron" is?  Metallic iron oxidizes in water so what you see is the oxide (ferrous hydroxide, ferric hydroxide,  hematite, magnetite, limonite to name a few) just curious. 

@foxy

Thanks Foxy. I realize I have a lot yet to learn about the different batteries and their charging protocols. I have a bit of time tomorrow morning and plan to spend some time reading that website you recommended. Rather than re-inventing the wheel, I may just buy a commercial system depending on what batteries I end up using.

Over the past couple of weeks, I've done a bit more on this project (and others). I built the power supply from the ATX computer power supply as per the DroneBot Workshop video. Just waiting for delivery of the little meter to install in that. 

Also, I sourced a Weatherproof Ultrasonic sensor JSN-SR04T and ran some tests with it against my HC-SR04 (robot eyes) sensor (which I know works). I tested them both with a horizontal target and a vertical target (pan with/without water) to see how they went at measured distances. I found that the JSN sensor was consistently 2 to 3 cm short in its measurements whereas the HC was spot on 99% of the time. Another limitation of the JST is that it has a minimum measuring distance of 25cm whereas the HC can go right down to 2 or 3 cm. I'm thinking of using a better (well, more expensive anyway) waterproof sensor which is available from my supplier. I'll get one and try it next.

Of course, absolute distances are not critical in my application as we can calibrate it to the empty / with water state but I'd still like it to be a little more accurate.

@dronebot-workshop

@foxy

@frogandtoad

Well, this post could be to anyone really.

I've made a lot of progress in recent weeks. On the non-electronic side, I built a small test rig to test the evaporation process (basically a box lined with aluminium foil with a perspex lid). I am going to test various lid angles to see if:

a) the process works, and

b) what is the most efficient angle if it does work.

I'll be setting that up properly over the weekend so will report my results after that.

On the electronic side, I've combined a number of ideas from Bill's videos as follows:

• 128x64 OLED screen to display necessary data (I may upgrade this to a 128x128 module if necessary. Ran some test code and now have that working just fine
• Tiny RTC to keep track of accurate date and time information. That works fine too and can now display the correct date and time (regularly updated) on the OLED screen. Initially I was going to update the display every 1 minute but have reduced that to 10 seconds for the time being.
• BME280 module for temperature, humidity and atmospheric pressure (and calculated height above sea-level) data. This is also successfully displayed on the OLED. Temperature and humidity affect water evaporation so I thought it may be useful to display this data.

So with that all working, my next step was to add the measurements from the JSN-SR04T Waterproof Ultrasonic sensor. (The issues that I previously had with the inaccuracy of the module have been resolved.) I've got test code that I run that works fine. As soon as I added it to the main program I am writing it returned only zero values from the JSN-SR04T. So I went through and commented out every piece of code unrelated to the JSN-SR04T, got it working and started adding code back in one line at a time to find the cause; and here it is:

display.begin(SSD1306_SWITCHCAPVCC, SCREEN_ADDRESS);

As soon as I run this line (which I gather is necessary for initializing the OLED screen and worked previously) the JSN-SR04T stops working properly and returns only zeros (0). The question is WHY?

If you or anyone can help with some advice, I would really appreciate it. If I can manage to obtain the source code for the library, I'll go through it and look myself.

This is weird!!!

Cheers,

David

@dronebot-workshop

@foxy

@frogandtoad

Well, it appears that the solution to my previous post was easier than I thought.

I went down to see a mate who is into Arduino. I had the  JSN-SR04T ultrasonic sensor (trigger and echo) connected to digital pins 3 and 4 on the Arduino board. He suggested moving them which we did and it now works. We moved them to D5 and D6. I had thought of trying that but hadn't yet done it. Still unsure as to why it was a problem but now I can proceed with writing the rest of the program. He also gave me an ESP-32 board to try out. Very nice.

😎 😎 😎