Creating an Automatic Keg Washer

Eno Sarris, October 25, 2016

Walking into a brewery before they've christened their system offers you one last glimpse of that glinting steel before it gets thrashed to death in the name of good beer. It's a beautiful sight, all that clean, expensive metal. Just look at it up there! 

"Looks expensive" I said to Federation Brewing co-founder Aram Cretan. He rolled his eyes. 

We ran threw the various items that stand in between them and opening this fall. At one point, he trailed off, and I tried to make light of it. "Another ten thousand dollars," was my attempt at gallows humor. 

"Everything's another ten thousand dollars," he laughed through the pain. 

But not one little thing. One hugely important thing might have cost them another ten thousand dollars -- almost literally -- but didn't because Cretan showed some creativity. He built his own DIY keg washer. 

Here's what the thing of (shabby chic) beauty looks like: 

The basics of keg washing are that you have to purge the CO2 if your caustic doesn't work in that environment, so you need to pull that out of the keg first and replace it with regular air. Then you throw your caustic cleaner in there, then you'll probably want to rinse twice -- we have some gray water and some fresh water in this case -- and then you have to fill it up again with CO2 so that the remaining space after your beer goes in is populated with CO2 and not air. 

Here's a basic run-through from Cretan himself: 

0) Hook up the keg, let it spew out its nasty. (Known as de-ulling, if you're curious.)
1) Purge CO2 with compressed air (caustic cleaners don't work well in CO2 environments.)
2) "Dirty rinse" from the "grey water" tank, to get the worst of the mess out.
3) Caustic cycle. The first few seconds worth go down the drain, then it switches to recirc. Helps keep the caustic cleaner longer. Caustic reservoir has a heating element in it run off a thermostat set to 130.
4) Rinse with clean water. This drains into the "grey water" tank for reuse, my contribution to forestalling the inevitable Mad Max-style water wars.
5) Recirc with an acid-based sanitizer.
6) Re-pressurize with CO2 to 10 psi, ready to fill with beer from the brite again.

If you're an amateur electrical engineer, you might recognize this as a simple problem, solved easily by simply starting and stopping different pumps for different amounts of time. Which is what Cretan did, with an arduino board shoved into his old college filing cabinet. 

The arduino board is in blue, and you can program that with some simple javascript. The other wires go to timers and pumps, and then there's a fan to keep the thing from overheating. It looks complicated, but this part isn't impossible with a little help from arduino, hobby boards, and homebrewing forums. 

Add in some line, and a few plastic kegs, and you're fine. Anyone can pick up plastic kegs now, now that they've killed people and the industry has decided they shouldn't be used for beer, for the most part. At one point, they were a cheap and more readily found alternative for an industry that couldn't provide enough metal kegs for burgeoning demand, but that's a story for another time. 

Back to this Keg Washer, here it is in live action. 

If you want to do this yourself, you'll have to read a few different accounts and cobble them together. Here's a decent one that went all PVC but seems to be on the same level. Here's another non automated approach, this time with the costs associated with each part.

Here's an arduino keggerator (the kegduino) with some insight into how to use the arduino board for these types of things. Cretan offered a bit more help: "Tutorials from Adafruit and Make plus a little bit of (but really not much) coding knowledge were really all I needed. Platforms like Arduino and the communities that have sprung up around them make this sort of prosumer hardware hacking really danged easy if you're not intimidated and just take things a step at a time."

Sum those hard materials up and you're south of $500, even if you splurge and get the $30 arduino board. Sure, there are some costs associated with spending your time learning how this works, but you do the calculus.

Is that time worth ten thousand dollars yet?