Automating a water butt


I’m pretty sure that I don’t have any regular readers, but if I did, and let’s pretend that you are one of them, the question that would be on your lips would be:

Don’t you normally write about languages or parsing Latin?

Yes. Yes. Yes. Mainly. I also have other preoccupations/hobbies/interests and if I can’t write about them here, where can I write about them?

One of these preoccupations is our Garden, and in it half way down on the left hidden by a small yew tree is a water butt.

But I get ahead of myself.

The garden is 13 metres long by 6.6 metres wide, shaded at the far end due to being slightly below ground level and surrounded by other houses.

There is an ash tree at the bottom left, and at the far end tucked against by (the falling down fence) a silver birch (dead centre) and slightly towards the house a eucalyptus tree.

I planted the eucalyptus.

The garden has been neglected for years. I can’t even tell you why, as I like gardens, gardening and being in the open.

Due to various reasons that I’d rather not detail here (but Brexit, pandemic and an idiot surely feature), I became an ex-automation engineer.

I now have time, but more importantly the inclination and motivation to korak po korak (step by step) transform our garden into what it should be.

Today, with assistance from SWP, I placed the new water butt into its new home.

But again, I’m getting ahead of myself.

You are of course wondering why I should want to automate a water butt, what that really means and maybe even how you could do the same.

So, chill and consider the situation I found myself in last summer, when we could still travel, and did, for several weeks.

The garden as a whole has been neglected, but small areas have always been loved and nurtured, and within one of these I had planted a tomato plant and extra geraniums.

While most of our garden is shaded either some or all of the time, luckily there is also an area which receives a fair dose of summer sun, and it is in this area that my tomato plant lived and (briefly) thrived.

The soil in what shall be known as the arid zone dries quickly.

The sun beats down.

Time passes.

No rain falls, and no water is watered.

After two weeks of no water, the tomato plant was more or less history, and the geraniums didn’t thrive as they could have done. Think brown leaves.

And now we find ourselves in December, driven inside because of the rain, with my non-automated water butt slowly filling.

This begs the question as to what I need to do to automate it, or encourage it to dispense water to the needy. Admittedly, at first blush, it is straight-forward :-

  • check moisture content around plant
  • turn on valve
  • turn off valve after a set time.

Of course, anyone who has ever done, anything oft this nature, soon realises that there is more to the problem domain than meets the eye.

Consider for example, as an illustration the humble teapot, and the steps required to make a cup of tea. Again this looks straight-forward, and in days gone by was used as an illustration as to the complexity off simple tasks.

So we have for starters:-

  • Pour water into kettle
  • Boil kettle
  • Put tea bags into a teapot
  • Pour water into teapot.

which is what n00bs tend to come up with as their first attempt.


If we wonder a little bit more what is appertaining, the following questions might bubble to the surface …

… You are going to pour water into your kettle are you? How will you fill the kettle. How much water? Are you going to fill the kettle to the top? If so, how are you going to detect the kettle is full? How will you get the water into the kettle?

  • Decide on how much water is needed
  • Move kettle to water source
  • Turn on water source
  • Observer closely and turn off water source when kettle has reached the required level.
  • Etc.

And as it was with the teapot example, so it is with water butt automation.

Simple on the surface, but with hidden depths which become more intricate for any other than the most basic solution.

What am I going to do?

I am going to massively over-engineer the solution, partly for fun, partly because I can, but mainly as an experiment for the later greenhouse automation project.

This is what my system needs to be able to do:

  • provide differing amounts of water to four different areas
  • monitor the water content of four different areas
  • call for help if a fault is detected
  • run using solar power
  • monitor the water level off the butt
  • make intelligent decisions depending on the water level.
  • allow control from an external controller if installed..

Being wise to such things, and not wanting to bother with a pump, my water butt is balanced on thick pieces of wood about a foot (30cm to you heathens) above the ground. Balanced does not do the construction justice. It is carefully placed within a wooden structure built for the purpose. I used dowels, sleepers and a brace drill from e-bay with a 22mm 😉 bit for this purpose.

I didn’t want a pump because …?

Why do you think?

Why shouldn’t we use a pump here?

No peaking until you have guessed the answer.

I knew that you would know.

I don’t want to use a pump because of the power consumption required.

This is the same reason that I won’t be using a PLC. These are usual in the automation world where reliability is crucial. The current drain from a PLC will be orders of magnitude higher than the system I will build. Now normally that wouldn’t matter, but I want to power my system using solar power and low current which hits power consumption (P=IV) matters.

I’m going for a professional although not properly industrial design – remembering, in the real world, industrial mostly means PLC.

My first step will be to decide on a manifold and appropriate low power consumption electric valves.

I’ll keep you up to date, as and when this project develops further.

Besos and Baci,


This entry was posted in Automation, Garden, Water butt. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *