fermenter temperature sensors
i am in the process of commissioning a small brewery with 2 fermenters and 1 brite beer tank. one of the vessels has a 12" thermocouple with a micromega controller. the TC is threaded into the thermowell (although the housing cap was cross-threaded and probably needs to be replaced).
the other two have been retrofitted with ranco controllers. the the ranco controllers come with about a 2" thermistor and 8' lead hard-wired into the controller. these thermistor were just shoved into the thermowells with some insulation backing (on the side for the fv and on the bottom for the bbt).
first of all, is this a reasonablably accurate method for temperature sensing? i have a friend that is a controls engineer and is going to wire up a scavenged PLC for controlling all of the tanks. should we invest in two more 12" thermocouples or try to scavenge the thermistors off the ranco controllers or buy new thermistors?
on another note, what type of temperature differential exists between the tank thermowell and the actual beer inside? is this something that bears investigation?
Thermistors can be fairly accurate, thermocouples tend to be more accurate, and RTDs are extremely accurate. If the controller allows, I would ditch the thermistors and use the same technology all around. I assume you have either type J or K thermocouples (they come in dozens of configurations). K is most common in brewery applications, although I've seen places outfit in type J. In the US, thermocouple type may be determined by the insulation color on each lead; ie red/yellow. Thermocouples are dead cheap. Just buy some cable and twist the bare leads together. Stick the twisted pair in the well and terminate at the controller. Make sure you have polarity correct when you terminate at the controller and also make sure you do not splice the thermocouple wire at any point. Each run must be a continuous length (or use a special splicing junction). Fermenters and conditioning tanks tend to change temperature very slowly and the well will indicate actual temperature accurately. I've found that mash tuns are more difficult both because temperature changes are more rapid and because it is difficult to obtain a representative temperature sample throughout the enitre mash piece. Placement of the thermowell is important, as is mixing the piece. But once you've used a particular tun a few times, you will adjust temperatures up or down based on results. Consistency is key, absolute temperatures are not. For better response time performance, I use themocouple couplant gel available wherever you get the thermocouple wire. Hope this helps. Good luck! Cheers!
no thermocouple sheath?
so are you saying that i don't need the stainless steel sheath and housing for the thermocouple? those entire setups cost about $60-$100. just shove the thermocouple into the thermowell and maybe back with some insulation.
Thermistors generally have a stainless sheath for this type of application, however thermocouples don't need anything other than being bared, twisted up and shoved in the well. Again, you may wish to grease up the thermowell with some couplant to ensure rapid response time. It really is that easy. Cheers!
i would use RTDs as they are best for temperatures from 0-100degC. thermocouples can have an inherent error of ~3degC in this range, and there is no way to correct it (unless done through software).
A 3 degree error - really? I've never read anything about that. Do you know what the source of the error is?
Thermocouples are wonderfully simple and pretty rugged. I've never used an RTD, so no experience to draw on there. Yes, you can just stick the twisted or soldered pair into the thermowell - maybe so that the tip is in contact w/ the metal interior of the well, and you're off to the races.
The greasy paste that was mentioned enhances thermal conductivity between the thermowell wall, and the TC - but as mentioned, if temp changes are not rapid, you're probably okay without.
I wouldn't ever do a thermwell without "thermal mastic." All it takes is cooling the fermenter and not having the sensor able to sense the actual beer temperature (because it can only sense the air around it in the thermwell) until the beer is actually a couple degrees below what your yeast likes, then you've got trouble. Exactly right that slow temp changes aren't so critical, but glycol cooling can be a fast change.
that sounds like really good advice. which brand do you recommend and how do you apply it and install?
Wish I knew- My refrig guy gave me an unmarked tub of the stuff. Looks like creamy peanut butter, taste likely different. I'm really curious if this could be easily homemade. Probably mineral oil with something to thicken it up. Seems like playdough with extra mineral oil to make oozy would work. Maybe even axle grease would work. Anything is better than dead air.
The goop that I see in mine is more like the texture of petroleum jelly, filled with what looks like little metal flakes - hopefully not lead or mercury! I think the idea is that metal in this form is a good conductor of heat. I'm just wondering how it doesn't make the contact w/ the TC at the base of the wire pair - versus the "business" end.
You'll find it wherever you buy your thermocouple wire. Squirt it in the well, push in your twisted pair until it touches the well end, and tighten your gland fitting. Couldn't be simpler. The "hard" part is programming your controller for the type of sensor (type K, etc) you have installed. Cheers!
Thermocouple color code guide
Came across this, a nice reference for determining Thermocouple Types based on wire color: