wholteza/blog
1
0
Fork 0
Code Issues 1 Pull Requests Packages Projects Releases Wiki Activity

Wip brewery notes

Browse Source
This commit is contained in:
wholteza
2025-02-03 22:28:57 +01:00
parent ecb567a41e
commit 7c0826aab5
1 changed files with 29 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
src/notes/5-measuring-mash-temperatures/note.md Normal file
View File

@@ -0,0 +1,29 @@
# Measuring mash temperatures with custom equipment
For yesterdays brew I wanted to improve something in the brewery, namely having a dedicated thermometer so I don't have to steal the one from the kitchen. I looked at what I had at home to see what I could build for cheap and dug up an old nodemcu, the esp8266 variant and remembered that I set it up with esphome sometime last year. After a quick check at the esphome documentation I found out that esphome supports the dallas ds18b20 temperature sensor which I have one of at home already but i picked up a couple more since in knew I wanted to be able to measure multiple things at the same time.
Picture of equipment
Esphome made the setup very easy. After defining the onewire hub component i could go ahead and connect one sensor at a time to the bus and note down their addresses, I also marked each thermometer so I knew what sensor data would correspond to what physical sensor. I settled for three sensors to start with.
Picture of put together equipment
My esphome instance is running in my local homeassistant (HASS) instance which integrates neatly and makes all sensors available to both systems. I created a new view in HASS and added the sensors showing the current values as well as a histogram of the last 5 minutes thinking that is should be a reasonable timespan to have when comparing temperatures. To my suprise the 5 minute interval in HASS didn't show 5 minutes at all but rather 24h. It was possible to zoom in on the graph to see it in more detail but i wasn't happy with how it turned out. Ideally I wanted this view to be visible on one of my screens in my brewery and it should show what I want to see without me having to interact with it.
Picture of homeassistant histogram
So off i went to Grafana! I set it up a long time ago when I was dabbeling more with smart home things and I needed a dashboard showing all devices that were running low on battery. So HASS was already set up to export all historical sensor data to an InfluxDB database and Grafana has access to that. I created a new dashboard, created a new graph and checked the devices and there they were, brew-temp-1, 2 and 3. I set up a graph showing the last 5 minutes of temperatures as well as three gauges showing the names of the sensors, their current reading and colored them the same colors as their corresponding line in the graph.
Picture of grafana dashboard
Brew day! We usually brew two batches the same day since it allows us to do utilize the time better, not having to clean everything as thorougly in between. This would allow us to do some measurements for the first run and implement changes for the second. We decided that we were going to measure the actual temperature in the middle of the malt bed compared to the water right below the malt pipe in order to see how the temperatures stabilize over time. Our recipes call for a 65c mash for 60 minutes but we had no evidence that were actually hitting that mark.
Picture of mash
We did the first batch as usual, heating up the water to 67c before dumping in the malt pipe along with our 5KG grain bill of mostly pilsner malt and some wheat. Historically we have been going off of the bult in temperature sensor in the bottom of our brewzilla brewsystem, but one of its flaws is that the sensor is located right above the two heating elements which means that it is going to measure the warmest water in the pot. This works for turning the heater on and off while recirculating using a pump while the brewery has no equipment or ingredients in it, but as soon as you add anything it is no longer going do represent the temperature of the whole vessel. We put the first sensor below the malt pipe right above the false bottom of the brewery, then we put the second one in the middle of grains.
Picture of graph of the mash
To our suprise the result was pretty bad. While this method has yielded good beer for the nearly 50 last batches with this brewsystem, the numbers showed us that while the heating is set to 67c our malt bed takes nearly 30 minutes reaching the target temperature. This results in only 30 minutes of most efficient mashing. Our thought was that if we can become more efficient at extracting sugars out of the malt we can use less ingredient than we do to day, making the beer cheaper.
Looking at the graph I estimated that perhaps if we mash in at a highter temperature, say 75c, we would end up at the target 65c immediatealy, giving us 60 minutes of mash time at the right temperature.