My new Electric Build: Bryn Mawr Brewery

After about 8 months of planning and building I've finally successfully built my own electric brewery in the basement and brewed my first batch! A huge thanks and shout out to everyone on this forum, many of whom have unknowingly provided a lot of help and input, as well as to Kal at theelectricbrewery.com for the inspiration!

Since my build ended up being quite different from how many others have gone about it, let me first start with my criteria:

1. I wanted to move to the basement because our kitchen was proving to be to small for all-grain batch sparging and because I have never desired to brew outdoors or in the garage.

2. The build needed to be fairly compact as I really don't have much space available even in the basement. I decided to go with a three vessel system with pumps because I don't have much ceiling height for a tiered system and because I don't like the idea of vessels higher than I can see into without climbing up on a stool or ladder.

3. I REALLY liked and was inspired by theelectricbrewery.com and decided to base my build on Kal's system. I was hoping to increase my efficiency and get some consistency, BOTH of which were just terrible in my all-grain batches. I liked the idea of a HERMS system because I felt that would really help with my efficiency and consistency issues and because it would help me create much clearer beers.

4. I ALSO knew I couldn't afford to emmulate Kal's build very closely due to budget limitations. I needed to use what I already had and what was available to me. THUS, since I already had 2 15gal stainless steel kettles and 1 10gal cooler MLT it was clear that these needed to be the basis for my build.

5. Coming from the Lighting Controls industry I knew I wanted to use as much equipment from that industry as I could. For one thing I'm extremely comfortable with that equipment, and I can also source a lot of it for little to no cost. Thus I moved forward with a design utilizing relay controls instead of SSR's and a touchscreen control interface instead of the typical control panel I've seen.

6. I saw no need for a system which is portable. I was more interested in building a system in to my house for my own use. If we unexpectedly need to move then the only thing I'll need to leave behind is some 10 guage wiring. Everything else can come with me and be re-installed in a new basement if absolutely neccessary.

So that was how I set about designing my build. To be honest a majority of the cost was in "finishing" my little corner of the basement. I have a very small footprint inside of what is otherwise the laundry room. This worked out really well when it came to ventilation as I was able to tap off of my existing dryer vent and not drill a new big hole in the foundation. The room also got a new floor (epoxy), a new utility sink, and some new drywall and framing. I had the added benefit of my wife being very happy with getting a finished laundry room. I have yet to install a planned false ceiling, but will get to that in the next couple of months.

I was lucky enough to have a co-worker who used to work as an electrician and he offered lots of free wiring work for me and helped me with a lot of the electrical decisions such as how thick the wiring from point A to point B needed to be etc. I had a 50amp outlet installed next to my existing main electrical panelboard (conveniently enough this is in the laundry room), and ran a 6 gauge, 4 conductor "extension cord" above the ceiling and over to my 50amp spa panel. From there I feed into another small panel which has two 30amp breakers for my 5500w elements: one in the HLT and one in the BK. This was to be honest a rather poor design and if I could go back and just forgo the SPA panel in lieu of two 30amp GFCI breakers I would do it in order to keep things more simple.

From the 30amp breakers I go straight into my lighting control cabinet which has up to 8 30amp relays, only four of which I need to use. These relays are made for switching up to 30amps of 277v circuits multiple times a day with lighting loads which carry a huge current inrush. They are warranteed for 5 years and have a 20 year expected lifecycle. I felt comfortable using these instead of SSR's because know they are well made and rarely fail. I can also source them at no cost if I have any unexpected failures.

My main worry about using a relay based system instead of an SSR based system was temperature control. An SSR switches much more quickly and more often than a mechanical relay can, and thus offer more precise control. (it's why they were designed into Kal's system in the first place, and I probably wouldn't personally advise anyone else to go my direction because no one else would save anything in cost or have easy access to these relays).

But so far my system has proven capable of holding temperatures steady very precisely. I've over-shot my temp a couple of times, but I think I've got the system calibrated accurately now. I will say this: my system at this point seems to be fairly slow to adjust/respond when I want it to. For example, Mashout will take probably 15 minutes or so unless I go to manual mode and set it for 90%. But then I need to sit and watch my temps so that I don't overshoot! This may get better over time, but it's the biggest drag for me so far.

Here are some pics. More to come and I'd he happy to entertain questions or see comments.

Thanks for sharing!

geoffey said:

But so far my system has proven capable of holding temperatures steady very precisely. I've over-shot my temp a couple of times, but I think I've got the system calibrated accurately now. I will say this: my system at this point seems to be fairly slow to adjust/respond when I want it to. For example, Mashout will take probably 15 minutes or so unless I go to manual mode and set it for 90%. But then I need to sit and watch my temps so that I don't overshoot! This may get better over time, but it's the biggest drag for me so far.

Click to expand...

Your PID will learn your system over time and become more accurate. Many PIDs have a learning mode where you can force it to spend about ~15 minutes to learn to speed up this process.

Once trained it should be nearly impossible to overshoot since the PID will be in control. It should also be nearly impossible to do it faster yourself by going to manual mode. Once the PID understands your system behaviour it will crank the heat at 100% and then back off by cycling the heat as it nears the target temperature so that you don't overshoot.

I say "nearly" simply because environmental differences between batch to batch can have have a slight effect. Things like the rate at which you recirculate (if you don't always do the same), ambient temp, or (the biggest one): The amount of grain. The more grain you have, the more heat it takes.

Got any pictures of the relays you used?

These relays are made for switching up to 30amps of 277v circuits multiple times a day with lighting loads which carry a huge current inrush. They are warranteed for 5 years and have a 20 year expected lifecycle.

Click to expand...

What's the expected duty cycle for such a warranty? They may be meant to be switched off many times/day but what about many times/minute or even a few times per second? If you can, I would set your PID to "relay output" so that it doesn't switch on/off too fast.

Best of luck!

Kal

A note on my PID settings:

I have them set up for relay output, and have changed the cycle rate from 2 seconds to 40 seconds on the HLT and to 10 seconds on the BK. This means the relays will cycle a maximum of one time every 40 seconds on the HLT and when I'm boiling once every 10 seconds. I found it necessary to set the BK PID to 10 seconds just in order to control the boil effectively. When set even to just as high as 20 seconds and at 85% in manual mode I found that the system wouldn't sustain the boil when the element was off for the few seconds.

The relays I'm using wouldn't last with a 2 second cycle rate which is the default for the Auber PID I'm using. But a 10 second cycle won't affect the lifespan of the contacts unless I were using the system on a regular daily basis doing multiple batches.

The 10 second cycle rate on the boil kettle is the one that would cause most people problems. On my setup the boil element goes on/off every second to maintain a nice boil at 85% duty cycle. Any longer and I'd get a virgorous boil followed by nothing, followed by vigorous boil, then nothing again, and so on.

Sounds like you matched the element size with the batch size fairly well if you can maintain a rolling boil instead of constant stopping/starting.

You may get a bit of over/under shoot with a 40 second HLT time too as you're noticing. I've noticed that when maintaining temp my HLT element often only comes on for a split second every 10-20 seconds. It basically pulses the HLT.

Kal

A few notes after having 4 batches completed on this system:

1. I realized that my HERMS coil was much to short at approx 22 feet this was primarily what was making my system so slow to respond. I went out and purchased a new 50 foot cooper coil and the system has responded much better now.

2. The auto tune on my PIDs doesn't work for my system very well at all. I'm using Auber SYL-2362 PIDs in relay/contactor mode. It just results every time in a very very slow responding system. I've resorted to manually adjusting the P, I, and D settings in order to dial things in. The system is now faster to respond but it consistently overshoots by 2-3 degrees. Very frustrating, but as long as I am aware of it I can work around it. Not sure if this is because the controller lacks fuzzy logic or what, but its my biggest disappointment with this system so far.

3. Temperature probe placement is something I wish I could do over. My temp probe in the MLT is drilled into the side of my Cooler, instead of being installed on the outlet with the ball valve. This makes it difficult when doing a 5 gallon batch of 1.040 beer when my mash volume of water is below the level of the thermometer.

4. My two pumps were sourced from Greatbreweh and have proven a real pain to prime. I'm thinking I need to add a bleeder valve on the output to help with priming issues.