Electric Recirculating BIAB rig, Temp overshoot problems

[Hpal]

Hi Everyone,

I have nearly commissioned my BIAB recirculating rig. It is 72 Litres with a 4500w element and controlled by an auber 2352 PID. There is a little brown pump recirculating the mash and the pt100 probe in the inlet in a tee piece. Just wanting to know if anyone has temp overshoot problems when recirculating with the probe in this location?

According to all literature this is the optimum probe position. I am seeing a long lag time between when the element is firing and when the pump can pump it up and around and even out the temp, I dont want to pump too hard and suck the element dry under the false bottom damaginf the element. It seems that (on the test run with water) that the heat from the element rise up the kettle and the pump is sucking not so hot liquid, it takes ages for temp to even out with recirculation, that the pid overshoots by 2-3 degrees c. I have resigned to the fact that this is not the optimum probe location for my system, I still have to tune the pid properly though, autotune does not give me the right figures with the slow reacting system. I will be making 400mm long stainless thermowell with a 4mm diameter pt100 probe in the end of it and this will go in the middle of the mash, with the pump recirculating still. Or I can simply use the pid in manual control mode and fire when needed to keep mash temp instead of auto mode and recirculating, either way good beer will be the result! Just wanted to know if anyone else has run into something similar.

Picture of rig attached. Doing maiden run next week to 38L to fill two kegs. Cheers, Hayden

 

[TheWiggman]

The long and the short of it is that you could fix your problem (overshoot) with correct PID settings. There will be other impacts like increased rate of change, but overshoot can be fixed.
Your element is above your suction point so with a LBP I can see this having a bit of an effect. With large batches and a small pump, chances are you're not pulling enough liquid away from the system and with 4000W you're doing quite a bit of heating before the probe has picked it up.
Re: choking flow, max it out. You won't run it dry if the thing's full of liquid.

 

[TheWiggman]

Can't edit above post. (nice looking rig by the way)
Should note that with the 4500W element you'll get some convection currents causing heat to radiate upwards, but this shouldn't have a huge impact. The liquid will eventually flow down through the pump so will be maintainable.
Your sensor is in as good a position as any. Get your PID set right and you'll be laughing. Sorry I can't help with specific setting recommendations, but I think you'll want to reduce your integeral value. If you're going to do an auto-tune you'll want to do it WHEN MASHING, not with a water-only run.

Also, why do you have a ball valve leading into the pump? I can't see why you'd ever want to throttle this. Is it just for the boil?

 

[spog]

I have a similar 2 ball valve setup as Hpal,I put a ball valve between the pot and pump in case a hose bursts or the pump falls apart ,it's a safety concern for me so of the worst happens I can save the brew.
Cheers...spog..

 

[Hpal]

I'm digging up this old thread so I can hopefully gain a bit more knowledge and advice and finally try to get my recirculating system up and running in auto. I gave up ages ago but am determined to get it working. I am going to do something like this as a mash 'bag' and recirculate slowly from the top for hopefully clearer wort and to keep some free space around the edges of the kettle so mash temp can even out.

Changes in the planning;
-magnetic drive pump replaces LBP
-Malt pipe/pot replaces grain bag (grain bag used to not flow well and stuck to the side of the kettle not letting temps even out)
-Try temp probe in kettle wall

Previous problems from memory; (it was a while ago)
-Temp control: Haven't had much success tuning the pid controller (Auber 2352) because the system reacts so slow. Just run it in manual mode for now to maintain temps and run at 80% or so output for the boil.
-Autotune didnt go well. Ran autotune, due to such lag time it overshoots terribly. Will try again with new malt pipe and pump recirculating slowly. Autotune figures generated were way off what I thought would be normal.
-Temp probe location, I tried inline with the pump inlet as pictured, but the pickup is below the element so most of the heat went upwards resulting in massive overshoot. Tried in the side of kettle where dial thermometer is, this may be a good place once it's swimming in wort and not covered in grain bag.
-Mash bag: replaced with a more free flowing bag but it still stuck to the sides and left the hot wort under the bag. Replacing with Malt pipe

Questions;
How much grain can I fit in a 19L stock pot?
PID figures: Could someone with a similar recirculating system please lend me their P,I and D settings so I can give them a try. I know that they wont be optimum but it will get me closer than I am now, which is nowhere.

The main problems are basically system lag, temp stratification and overshoot, and subsequently the PID tuning.

Thanks all, looking forward to getting this out of the too hard basket!

 

[MastersBrewery]

Sounds like you've put some work in! Max grain bill for a 19L pot is around 5.5-6kg though 5 and under will serve you well.
When testing and calibrating the system try and first get a base line without grain.
Your issue sounds a little like stuck sparge so who mills your grain? On a system such as this you need a reasonably course crush compared with biab.

 

[MHB]

The standard number for malt (fully hydrated) 0.7L/Kg it ranges from 0.65-0.8 depending on the grind. Working on 0.7 a 19L pot would hold 13.3Kg (give or take) if you aren't concerned about the risk of some spilling out into the rest of the system (with a LBP I would be).

Lag time has been identified as the problem which makes sense. there are two solutions
Lower the heat (fit a voltage control SSR into the circuit)
Raise the flow (get the wort recirculating a lot faster) - not going to happen with a LBP. There is a very real limit to how fast you can get the wort to permeate the grain bed, go too fast and it will compact solid. Some amount of bypass so not all the wort is trying to go through the grain bed might help.

I think the cheapest and easiest solution would be to either throttle the element back or fit a much smaller element just for the mash cycle then switch to the big one to boil.

 

[MastersBrewery]

The standard number for malt (fully hydrated) 0.7L/Kg it ranges from 0.65-0.8 depending on the grind. Working on 0.7 a 19L pot would hold 13.3Kg (give or take) if you aren't concerned about the risk of some spilling out into the rest of the system (with a LBP I would be).

Lag time has been identified as the problem which makes sense. there are two solutions
Lower the heat (fit a voltage control SSR into the circuit)
Raise the flow (get the wort recirculating a lot faster) - not going to happen with a LBP. There is a very real limit to how fast you can get the wort to permeate the grain bed, go too fast and it will compact solid. Some amount of bypass so not all the wort is trying to go through the grain bed might help.

I think the cheapest and easiest solution would be to either throttle the element back or fit a much smaller element just for the mash cycle then switch to the big one to boil.

MHB you forgot fluidisation any malt pipe acts as a tun in it self and with a 3L/kg ratio around 5kg is it. Example 20l BM.

MB

 

[MastersBrewery]

Here's a link to a recirculating biab setup that has worked well for many others note it shoots a stream of wort over the element continually while also recircing.

 

[hairydog]

I brew with a similar single vessel recirc set up but have a malt pipe mounted inside an old Malleys 50 litre wash tub also known as a copper,
for temp control i have a RTD probe mounted on stainless steel line run to middle of mash (submerged) with small holes drilled and plugged off to act as a sort
of sprinkler and another line running to bottom of vessel to mix wort around element to distribute heat better.The temp controller is an on off control
to a SSR to drive the element and still get a 2 deg swing during mash.
I have got hold of a PID controller with self tuning function which has the ability to output a 4-20ma control signal,bought a 4-20ma input SSR 25A output off ebay $20
to which i am still waiting for and see if i can reduce set point offset.The theory is that the solid state relay pulses the element on a few times per second
to control element at more finer control instead of full heat then off.Not sure if you can adapt this hardware to your set up but will post how my results
turn out when i get SSR and connect it all up.

 

[MHB]

MHB you forgot fluidisation any malt pipe acts as a tun in it self and with a 3L/kg ratio around 5kg is it. Example 20l BM.

MB

Indeed, sorry I looked up the wrong number, as you say it would be about 5kg.
Mark

 

[mr_wibble]

On an non-insulated keggle 50 litre HLT, 2200 watt KK element - my PID settings are Kp = 2, Ki = 0, and Kd = 200
(Arduino PID library)

This never overshoots, but it does crawl to the setpoint for the last few degrees. Setting the target 2-3C over fixes that.

During development I tested this with a 2 litre kettle, and (with completely different PID settings) it preformed flawlessly.
Using those settings with 50 litres it was all over the place. The volume of water, and probably a bunch of other factors really make a huge difference.

Auto-tune completely failed to work for me too.

 

[Hpal]

The standard number for malt (fully hydrated) 0.7L/Kg it ranges from 0.65-0.8 depending on the grind. Working on 0.7 a 19L pot would hold 13.3Kg (give or take) if you aren't concerned about the risk of some spilling out into the rest of the system (with a LBP I would be).

Lag time has been identified as the problem which makes sense. there are two solutions
Lower the heat (fit a voltage control SSR into the circuit)
Raise the flow (get the wort recirculating a lot faster) - not going to happen with a LBP. There is a very real limit to how fast you can get the wort to permeate the grain bed, go too fast and it will compact solid. Some amount of bypass so not all the wort is trying to go through the grain bed might help.

I think the cheapest and easiest solution would be to either throttle the element back or fit a much smaller element just for the mash cycle then switch to the big one to boil.

Thanks Mark,
Unfortunately it's such a slow moving system and that loop is a bitch to tune. It's set-up + loop tuning that will get me there in the end. I've now got a KK mag drive pump with a good flow rate. The malt pipe I'm planning on having some horizontal slots cut just below the lip of the pipe to act as an overflow. So probe placement, wort flow and tuning needs to be my focus.

 

[MHB]

I still think that a relatively short very powerful element will cause you a lot of grief. the wort around the element will heat very quickly (often hot enough to snuff all the enzymes locally) and unless you have a very high flow to move the hot wort away from the element to where the sensor is located you will still get overshots.
It can be sort of like a pulsing action, where slugs of wort way over the set point are followed by under temperature flows in a feed back loop.

the element will be switching very quickly which shortens their life, I really think you should look at reducing the element power.
mark

 

[Hpal]

I brew with a similar single vessel recirc set up but have a malt pipe mounted inside an old Malleys 50 litre wash tub also known as a copper,
for temp control i have a RTD probe mounted on stainless steel line run to middle of mash (submerged) with small holes drilled and plugged off to act as a sort
of sprinkler and another line running to bottom of vessel to mix wort around element to distribute heat better.The temp controller is an on off control
to a SSR to drive the element and still get a 2 deg swing during mash.
I have got hold of a PID controller with self tuning function which has the ability to output a 4-20ma control signal,bought a 4-20ma input SSR 25A output off ebay $20
to which i am still waiting for and see if i can reduce set point offset.The theory is that the solid state relay pulses the element on a few times per second
to control element at more finer control instead of full heat then off.Not sure if you can adapt this hardware to your set up but will post how my results
turn out when i get SSR and connect it all up.

Cheers,
Have you got any photos of the setup? I also have a malleys 50L copper 2400w which I BIAB in.
Here's my control panel, it's got a PID and pt100 probe for the mash but at the moment I just use manual control

 

[TheWiggman]

Hpal,
Check out this post and this and the rest of the thread if you can get your head around it. It gives you a good start on the numbers to enter into a programmable controller.

Regarding "but the pickup is below the element so most of the heat went upwards resulting in massive overshoot", I don't get that. If you have adequate flow rate the liquor will be drawn down which is ideal. I'd wager that tuning or flow is the bigger issue here.

 

[Hpal]

The lowest I could theoretically go is 3600w and I should be able to boil a full volume with that, but the 4500w element is an ULDW element which should be ok, it's quite long. I could maybe set a maximum element output of say 80% on the pid (it's time proportional control) and see if that helps.

 

[MastersBrewery]

Better flow I think is the main issue here, with the top down approach (direction of fluid) there is simply not enough flow through the bed alone hence the grainfather and QldKev have created a second faster path for the wort to travel to enable the stability your seeking,
Also if you troll around ebay you'll find fairly cheap 38L pot to increase batch size. I think there's a link for the pot in this thread of another similar recirc system by takai I believe he used a similar element. Don't be afraid to PM either of OP's of the threads I have linked for any finer pointers.

MB

 

[hairydog]

Hpal,

I will get some photos together and post. Nice control panel, i have put together mY equipment scavenged from equipment

thrown out on-site where i work,like j-boxes controllers,the Malleys boiler was given to me buy wifes grandfather,i am lucky

to be able to get my hands on plenty of 1/2 inch stainless tube and fittings and valves,had to buy camlocks and SSR relays.

 

[hairydog]

Hpal this is the yokogawa YS-170 Controller i have acquired,a bit of overkill for what i am using it for but who
cares when its for free,has a few handy view options.Have to mount this on a plate and wire up with new SSR.