Getting the most out of PID Control - Modelling HLT/1V, RIMS and HERMS

[Adr_0]

Oh man, I spent countless hours on Simulink for my thesis. I haven't used it since. Would you be willing to share your models?

Sure - it will be good to have someone check my numbers. It will also be good to set the variables to your system and see how close it gets.

1. 36l, typically 3l/kg liquor:grist. 2kW element. For a 4.5kg about 15l including losses.
2. Not during normal operation. Accidentally close a valve however...
3. 2 l I think. Can double check.
4. I haven't actually timed but this will vary depending on controller settings obviously. No to overshoot. Will get back to you on next brew day provided my wife hasn't arranged other activities during mash time.

Ok, so when you say you could accidentally close a valve, you circulate water through your HEX I assume? Does that water go back to the HLT to cool off a bit?

 

[TheWiggman]

Basic process is -

1. Fill MT to mash-in volume, set temp on controller and turn pump on.
2. Dough in: turn off pump, controller and HEx outlet valve.
3. Wait 5 mins, turn on pump and just crack outlet valve. Set controller to first step temp (i.e. 52°C)
4. Gradually open valve every few mins until full flow is achieved
5. Go through mash cycles
6. FLY SPARGE - reconnect pipework so
   HLT > Pump > HEx > MT
7. MT > Kettle

For the sparge I open the drain valve for the MT so it's trickling into the kettle, aiming for a 45 min - 1h sparge. I then open the HEx discharge valve to match that flow which as you can imagine, is hit and miss. It's a balancing act for about the first 15 mins. Now if I'm dumb enough to accidentally leave the discharge valve closed then there is no flow past the thermowell and the controller will simply leave the heating element on.

 

[Adr_0]

Ok, think I understand. I was curious if the WATER in your HEX boils. I can understand if you have the wort return closed this would overheat

 

[mofox1]

Will get back to you on next brew day provided my wife hasn't arranged other activities during mash time.

Sounds more entertaining than I assume was intended...

 

[Adr_0]

Sounds more entertaining than I assume was intended...

Make sure a protein rest is done...

I've done a few examples - using referenced and calculated values for heat transfer - and seem to be getting only 2-25W being picked up by the coil. That is examples with a few degrees temperature difference, 1 or 2m coil lengths and up to about 40-45°C (based on boiling water). This was based on 5lpm wort flow through the coil.

So I guess what this suggests is that once the HEX ramps up to temp, most of the heat goes into the water around the coil rather than the coil itself, causing the water to continue heating (to boiling?) until the tube set point is reached. It's hard to say what the best control would be, but theoretically you should be able to match the coil output to the demand from the coil/mash. But at only a few watts (even with a 2m coil) it's not a lot of power going in, so the ramp rate would be slow.

If the water in the HEX is boiling, this would really help the heat transfer and power, but the catch is that it would subsequently cause a lot of overshoot because your heat source is now at 100°C but your mash is at 65 or 76°C - with no way to cool it.

I would just like to get a better idea of what goes on so that I can model it and try and get the best tuning/configuration out of it.

 

[trevgale]

I think the heat transfer into the coil is better than what the model is showing. I have a 2L HEX with a ~2m long 12mm diameter stainless coil. I haven't compared the temperature difference but I would not expect it to be greater than 20 deg C. I only get boiling in the HEX if I accidentally shut a valve in line with the coil or if I heat to over 85 deg C for cleaning.

Hope this helps.

 

[verysupple]

:huh: "gluten" = "the"?

Kids been playing with your phone?

 

[trevgale]

Yeah, shortcut setting on iPad. I think my wife might be messing with me. Either that or my 1yr old is a genius.

 

[Adr_0]

This look a bit more realistic?

 

[verysupple]

Yeah, shortcut setting on iPad. I think my wife might be messing with me. Either that or my 1yr old is a genius.

Ha, I figured it was someone playing funny buggers.

And I should also say great work, Adr_0! I love this kind of stuff - someone with some know-how applying what they know and sharing. Keep it up.

 

[fcmcg]

I'm keen to hear from HERMS brewers, a few questions:
-what is the size of your mash tun (mash volume actually) and element?
-does your HEX water boil, assuming you control your outlet temp?
-what is the volume of water in your HEX?
-how long does it take to make a 10°C step, and do you get any overshoot?

1.Mash tun is a 100 litre keg...usuallu dough in with about 15 kg grain , thereabouts and 45 litres strike water ..2400watt element
2.Doesn't boil , usually...unless its been ramping for a while and the water level gets a bit low..( i must put a lid on it)
3.HEX is a 10 litre pot
4. 10 degree step usually takes 17 to 25 min depending on ambient temp , actuall mash volume.No overshoot

I'm usuing some old shimaden SR44 pids that came out of Generous Motors...
I don't even know how to program them as i have no manuals and they are at least 10 yrs old

 

[trevgale]

1.Mash tun is a 50 litre pot, dough in with about 11 kg grain , thereabouts and 27 litres strike water ..3000watt element
2.Doesn't boil under normal operation starts to boil if trying to heat I excess of 85 deg C for cleaning
3.HEX has a 2 litre volume
4. 10 degree step usually takes around 8-10 min.No overshoot

I'm using and Auber ramp soak PID.

 

[Adr_0]

Ok, one more go tonight and I might have a bit more of a play later in the week. This five days in a row of not drinking is making me do crazy things...

This is a bit better:
- Pulls power out of the HLT better (almost like the coil EXCHANGES heat with the HEX/HLT now... amazing)
- Less sensitive to coil sizing
- Seems to give about 10-12°C warmer in HEX/HLT which is what most people see


Tuning-wise, it looks like it's pretty flexible with a variety of gains and integral times, which is sweet.

However if the HEX is more than about 20-30% of the mash tun size, there is some nasty overshoot.

 

[trevgale]

Those results look like they match pretty well with what I see on my HERMS. We I was researching prior to setting up my system there were discussions about overshoot and poor response with larger HEX sizes. The consensus seemed to be the smaller the volume the better.

 

[Adr_0]

Those results look like they match pretty well with what I see on my HERMS. We I was researching prior to setting up my system there were discussions about overshoot and poor response with larger HEX sizes. The consensus seemed to be the smaller the volume the better.

Yeah, bingo. I was confident of that, but wanted to understand how coil length and P vs PI control would impact the response.

I will put another comparison up in a few days. If people aren't convinced big HEX = bad response after all the threads and these response curves (coming in a few days), they deserve the overshoot headaches.

 

[trevgale]

Yeah, bingo. I was confident of that, but wanted to understand how coil length and P vs PI control would impact the response.

I will put another comparison up in a few days. If people aren't convinced big HEX = bad response after all the threads and these response curves (coming in a few days), they deserve the overshoot headaches.

It's always good when the modeling matches with the observed results. As an engineer I know it can be difficult to achieve sometimes.

 

[mofox1]

You've convinced me to take some measurements on my next brew day...

Mine is the "big hex" scenario.... So big you might call it a HLT. h34r:

Usually 70L water in there, my initial tests with 60L could ramp the HLT by about a degree every 45 or 50 sec, but that was without recirc thru the coil. Other (separate) tests showed no discernable difference between the coil out temp and the HLT, even with a 10 degree difference between the mash and HLT temps. But I haven't looked at overall system response wrt to overshoot...

For the greater good, I must brew!

So, mash in at 55, get everything stable, and ramp to 65... I can do that. Should probably also check what values the autotune set.

 

[Adr_0]

Sweet! I'm doing a brew this weekend too, to try a different set of PI values.

The 'effectiveness' I was using was 2.5% which got pretty good results. I just realised I made a slight surface are a boo-boo on my original numbers, and I'm getting calculated values of 0.5% for a 2m copper coil, 2.3% for 10m copper and 3.6% for 16m copper. My original calculated effectiveness values were more like 0.005%. Sweet. Stainless numbers are about 80% of these.

This means more power, but at the expense of time constant (time to get to this temperature). Curious to see the impact of the balance...

 

[Adr_0]

I think if you have a largely volume HEX and a long coil, your HEX temps will only be a couple of degrees above the coil outlet temp.

I would say that Integral control would cause some overshoot in the HEX temp, but would say this would help the initial ramp rate and would hopefully even out by the time you got close to your setpoint.

What is not yet clear is the impact that a time constant of 7-20s William have on overshoot. It might not be that significant by the time it gets to the mash.

If your coil length is small on a big HEX, the HEX temperature will sit pretty highly normally, and will be exacerbated when using Integral control.

Anyway, I'll have a look at it all and see what comes out of it.

 

[krausenhaus]

Hey Adr_0, I haven't had a chance for a good read of this but can't you model dead time with a transport delay block?