thermowell

[Demy]

For my BEER I usually use a pid combined with a pt100 temperature probe (economic type). I have a small mash tun with a mixing system and I usually insert the temperature probe (with a long stainless steel casing) from the lid. I decided to insert a thermowell as I already have a side hole, so as a test I built one with a 0.5mm thick piece of stainless steel sheet that adheres very tightly to the probe. The problem is that it distorts the temperature reading. I don't think it depends on the material because in the past I have tried with copper and I have often had the problem, the probe is very adherent (it looks like a jack) so I would exclude this too. Have any of you who use the thermowell ever checked the accuracy of the reading?

[Yummyrum]

I have never used one as I have always been dubious about them . I prefer to have the thermometer probe directly in contact with the vapour or liquid and the prob insulated from boiler or still head with teflon tape .

I’ve always figured that if the thermowell is in physical contact with the boiler or still head wall , it would skew the readings and it would appear from your experience that this is the case .

I can see no real need for a thermowell other than the ability to replace the thermometer without having to switch off or empty the boiler .Probably an advantage in industry but not for us .

[Demy]

I have never used one as I have always been dubious about them . I prefer to have the thermometer probe directly in contact with the vapour or liquid and the prob insulated from boiler or still head with teflon tape .

I’ve always figured that if the thermowell is in physical contact with the boiler or still head wall , it would skew the readings and it would appear from your experience that this is the case .

I can see no real need for a thermowell other than the ability to replace the thermometer without having to switch off or empty the boiler .Probably an advantage in industry but not for us .

Yes, yummy I have always done as you do, but I have read of many homebrewers that use them and they are widely marketed that I have wanted to experiment even though I have had the same disappointment in the past. Before saying this, today I did some tests side by side with various thermometers, in addition to this I removed the same probe from the thermowell and dipped it directly into the liquid and immediately it aligned with all the other thermometers (analog and digital). The material I used is identical to many commercial wells also in thickness, so serious doubts have arisen. I had a deviation of 2 ° C around 50 ° C, while between 60 and 78 ° C the difference increased to 3-3.5 ° C. I think it's huge in a beer mash.

[NZChris]

Are you using thermal paste?

Have you stuffed insulation into the thermowell behind the thermocouple?

Have you put insulation around the well?

[seabass]

I love thermowells. I put them in my kettle, mashtun, boilers, and fermenters. I haven't used them in vapor, so I can't comment on that. I just calibrate them to another thermometer I have that I put directly into the liquid. With thermal paste it's pretty accurate.

Not sure how well they will work in vapor, but if you calibrate one to the abv coming off the still, you should be able to get the temp accurate. At least that's the method I'll use when I install my copper thermowells in my column.

[NZChris]

Discrepancies due to thermowells can also be influenced by ambient temperature and air movement. If I'm controlling a temperature with a PID, I also have a thermometer that I trust in the mash/milk/whatever, and adjust the set point of the PID to force it to control to my desired temperature as read by the thermometer.

[Yummyrum]

by NZChris » Tue Feb 09, 2021 4:52 pm

Discrepancies due to thermowells can also be influenced by ambient temperature and air movement.

I’m sure this is a thing . If the thermowell is physically connected to the wall of a boiler or still , its temp will be effected by the heat being sucked our from the mass of metal its joined to .

If they could be thermally insulated from the body of the still or boiler they would be more accurate . A teflon gasket either side and a securing nut . Hmmm sounds like the next product from the China pump it out machine .

[NZChris]

Think about the stuff you were taught at school, like 'heat rises'. Angle a thermowell uphill and it will be more accurate in a hot vessel than if it points downhill.

[Demy]

Are you using thermal paste?

Have you stuffed insulation into the thermowell behind the thermocouple?

Have you put insulation around the well?

No thermal paste because I find cleaning very messy. the well is welded to the saucepan ... do you think this is the problem? I used the same system in my fermenter (same self-made and welded well) and it is accurate but in the wort it gives me this problem. Clarification: we are talking about mash, not about measuring vapor.

[seabass]

In my experiments, the thermal paste made a huge difference even with a very tight fitting thermowell. I got a lot more bouncing around and it was slower to register temp changes without the paste. It's really not that messy. It wipes off easily with alcohol.

But I had to adjust mine by 0.8-1C while in the thermowell. Each sensor was different. They are consistent at 25c and 100c though.

[Demy]

In my experiments, the thermal paste made a huge difference even with a very tight fitting thermowell. I got a lot more bouncing around and it was slower to register temp changes without the paste. It's really not that messy. It wipes off easily with alcohol.

But I had to adjust mine by 0.8-1C while in the thermowell. Each sensor was different. They are consistent at 25c and 100c though.

Thank you for reporting your experience. So you too have had similar problems ... I try to understand, some argue that even without pasta they have very low error values in the order of 0.2 ° C ... I would be fine with an error of 1 ° C but in my experiment they are many more. I hope others report their experience to clarify.

[seabass]

But if the error is consistent, it's pretty easy to work around. I also have much thicker walled thermowells than you even if they are copper. I wouldn't be surprised if thermal paste brought you closer.

[Demy]

You could take it into account in the steps or correct the instrument but it is very unpleasant as a solution. Beyond this I had a non-linear error so not perfectly predictable ... I would like to approach it without pasta and then decide whether to use it.A curiosity, are your thermowells isolated from the pot ?

[seabass]

Mine are not isolated at all. They're soldered to the pot. I think part of the issue is the length. With a relatively short thermowell that is welded or soldered to the pot, you will get some deviation from the outside of the pot.
Mine are 50mm long and the probes are 40mm long. The one in mt mash tun has reflectix insulation covering it and the rest of the sides of the pot.


I don't get any deviation on my fermenters where the thermowell is 6 -8 inches into the beer. I think it is because any variation from the outside is overwhelmed by the liquid.

[Demy]

I had a similar suspicion, even though the sensitive part of the probe is the tip. My pt100 has its own stainless steel casing (it is very long, I immersed it from the lid), I know how it is done inside because I disassembled it by removing the actual sensor (one about 3cm), but the well I built it is as short as the internal sensor because I thought it is the one to take the reading. I have another spare sensor without casing and I tried the one inserted in the thermowell and it gives the same error as the one with stainless steel casing. I will have to do further tests to understand.

[Demy]

To better understand the first test I did it with this probe (similar to mine)

sonda.png (26.51 KiB) Viewed 1875 times

Then I thought that maybe the long stainless steel casing could be the problem then I used a spare sensor (similar to the one inside the probes)

sensore.png (15.02 KiB) Viewed 1875 times

Both gave the same error (compared to other thermometers). So I removed the probes from the thermowell and dipped it directly into the liquid and they lined up with the control thermometers. I used 2 alcohol thermometers and 1 digital at the same time to check. From here the post was born ..

[seabass]

Out of curiosity, how long is the thermowell?

[Demy]

It is as long as the sensor inside the stainless steel tube, about 3 cm. These probes are long but the sensor inside is short.The second probe I posted is about 3cm and is the internal one.

[Demy]

Ok, since I have a hard head, today I tried to understand why my thermowell is not working well. I started redoing a slightly longer one than the previous one. This is about 8cm long, I looked around for some material as it is an experiment. I found an old electrical element and started with that.

It looked like stainless steel but it was chromed so I thought I'd find out. It was a bit difficult to straighten it up but I managed. I tried to shape its tip to match the shape of the temperature probe (for maximum contact and elimination this variant).I tried to minimize the thickness of the welds so as not to interfere.

The probe fits well

[Demy]

To better understand this is the probe I use (it's long because I used it from the lid), I ordered a shorter one.

This is the internal sensor that these probes generally have (PT100 type)

This is the shape of the probe head and the thermowell

[Demy]

Having remade the probe I started some tests, nothing scientific is clear but a simple practical test. I filled my little mash pot with water, placed some thermometers to compare the measurements and started heating. In this first phase the PT100 has no well.

I read the temperatures step by step taking some photos, sorry for the poor quality but I had to make do, it's not easy to frame everything. I started at 38 ° C, then around 45 ° C. Sebrano lined up (except the large alcohol thermometer which is less accurate) here the thermometers

here the PID

then around 45 ° C

[Demy]

I continued up to 50 ° C and they look and the thermometers seem to line up. I repeat again my probe does not have the thermowell. At this point I inserted the thermowell and monitored the temperatures as usual. To be precise, I did not use any thermal paste to better test the "naked and raw".

To my surprise, this time it worked out well. I'm not looking for extreme precision but there we are. I had about 0.3-0,5 ° C delay in reading compared to that little digital thermometer.

[Demy]

Seen and there are, I decide to complete the thermowell with scrap pieces ... I think they will be quite definitive. I already have a welded nut in the pot so I improvised something on the fly. Keep in mind this is experimental ... be kind to me ...

At this point I have to test the probe again because the pieces I added could create problems (they have a certain mass) so the only way is to try.

As you can see, it worked well for an arranged piece. I ordered a PT100 like that but shorter, this works but could get in the way. What was wrong with the first thermowell I built remains unknown. Any opinion is welcome.

[SaltyStaves]

I use Glycerine inside my thermowell with a PT100.

[Demy]

I use Glycerine inside my thermowell with a PT100.

Seems like an easier option to clean. For now, given the results, I don't think I'm using anything.

[NZChris]

I don't clean my thermocouples, I leave them in their wells permanently.

If I often need to use the same controller elsewhere, I get another thermocouple rather than going to the trouble of moving them every time I need to. It used to be worthwhile moving them several decades ago when such things were quite expensive, but they are much cheaper now. In the early days of my distilling/fermenting adventures, I had several refurbished thermocouples from an electrician's rubbish bin, but only one display, and used a rotary switch to select which thermocouple to read from.

[Demy]

I don't clean my thermocouples, I leave them in their wells permanently.

If I often need to use the same controller elsewhere, I get another thermocouple rather than going to the trouble of moving them every time I need to. It used to be worthwhile moving them several decades ago when such things were quite expensive, but they are much cheaper now. In the early days of my distilling/fermenting adventures, I had several refurbished thermocouples from an electrician's rubbish bin, but only one display, and used a rotary switch to select which thermocouple to read from.

Yes, I see. My beer mash tun is small so I can put them in my tub and wash them comfortably, so disconnecting the thermocouple is convenient, as it's pretty accurate I don't put anything. I believe that each of us chooses the most suitable path. I am left with the question of what went wrong with the first thermowell ... I was tempted to weld the latter to the vessel just to prove your theory, but it would have taken me longer.