Dedicated Herms Guide, Problems And Solution Thread

[Mardoo]

Well, your area won't be cut in half but it will be cut down a bit, so I would try to keep it off the sides. Apparently you can fill your coil with sand and then you can hopefully get a tighter bending radius. If you can find a smaller bit of pipe this might help?

What sort of lengths are we talking here?

I believe I can fit roughly 5 meters of coil into the vessel, but I lack the know-how to calculate the length of a vertical spiral, so I'm working from the notion of stacked circles of 1/2" copper tube minus a something-or-othereth of a meter. What I'm talking about is roughly the difference between a 5 meter coil and a 3.25 meter coil.

Pushing the coil all the way out to the edges of the vessel hadn't occurred to me until today and it seemed to have an inherent problem, so I asked. At the moment my question is more a thought problem than anything, and part of my overthinking-my-way-to-simplicity process. (Patented, yep ) Martin, your poofteenth of SFA sounds reasonable (and your advice of build it and learn to brew on it is great) . In truth I think the deciding factor will likely be ease of coiling a 14 cm diameter coil over a 9cm diameter coil. We'll see what my experiments with some spare copper I have show. The 14cm will leave a bit of room between the coil and vessel edge, but not significant. Trevgale I'm guessing you're right about vessel being near the same temp as the contents anyway. And yes Adr_0, I do hope it will be making delicious beer somewhere along the way even if I don't have a longa, stronga coil. I certainly don't have the mathematical and other prowess to get down to the last 1%.

By the way, here's some very interesting advice about coiling copper filled with water, with an appearance by our very own Punkin. The real meat of the post is in the beige section halfway down. I'll report back if I give it a go.

Thanks for the input and suggestions and I'd love to hear any more folks have.

 

[Adr_0]

Looks like the difference between 3m and 5m is about 5min, vs 10min faster for 2m vs 3m. That's knocking a few % off the 5m coil by the way. So somewhere in the 3-4m range you're in the sweet spot. This is time for each 10°C step, so you're looking at saving maybe 30-50min in going from 2m to 3m, but can save maybe another 15-20min if you get to 5m. So basically somewhere in between will be sweet.

Blue - 2m; orange - 3m; red - 5m.


This was assuming 2400W element, 48L mash volume (mass actually - grain and water).

In each case though, the water still boils - for your mash to only get to 65°C (this will be more so when you mash out). You get much better heat transfer with boiling water, so it will only be for a short time - but it's important to make sure you don't overpressure your HEX, i.e. have a little vent/fill point on it.

A larger HEX won't necessarily help - i.e. the same length coil in a 5L HEX will just take longer to ramp, and will boil for longer. It does give you the opportunity to stuff more in though. I think you'll be right - just be conscious of boiling water.

 

[Mardoo]

Wow Adr_0, I really appreciate you plugging my figures into your model. I am much obliged sir. Where's the taking off hat and bowing icon when you need it?

I've already planned for a vent for the HEX, so the boiling isn't an issue pressure-wise, although water supply could be. Looks like I may need to look into that float valve after all

 

[Adr_0]

I don't think it will be too bad, but yes, you should think about it.

Your pump rates through the coil will also help. This was all assuming 5lpm. If you can go higher than this, it will help your ramp times significantly.

 

[Adr_0]

So did an interesting comparison this morning. It's also becoming pretty obvious that PD control is best for HERMS, P is ok for BIAB/1V but you really need that D to make it as aggressive as possible without overshoot.

All of these are 2400W elements.

One guy has a kettle, with 1m coil and a 19lpm pump, into 25L worth of mash. P is 1, D is 125. This is the orange

Another example is a 42L HLT (HEX), 15m of copper coil and still a 19lpm pump. P is 3, D is 150. This is the blue line.

Red line is same as above but 7.2lpm pump, eg LBP.


Pretty well polar opposites of HERMS setups, with very different responses.

The three graphs are the HEX temperatures. Looks like the kettle boils... Good for a cup of tea?

 

[Adr_0]

And here's the 1.8L kettle, 1m copper pipe, 2400W element at 7lpm - blue, vs 19lpm orange, and 42L, 15m, 2400W 7lpm red:


They have each been roughly tuned, but you can see the difference 19lpm vs 7lpm makes.

 

[Adr_0]

I've done a bit of playing around over the last week or so. A few people have said that my calculated 'effectiveness' seemed a bit low as they would get better results than predicted. Turns out one of my factors is meant to be squared, which improved things a bit (added about 80% to some of the numbers).

The amount of heat that gets into your coil from the hot water in the HEX/HLT is the effectiveness. If there is a 30°C difference between the wort in the coil (coming in) and the hot water, the wort mass flow x heat capacity (3700J/kg.°C) you theoretically have a lot of power that can get into the wort, but the coil design (basically length and material) dictates how much of that potential gets in.



The reason I put this up is so that you can see the difference between copper and stainless, and realise as well that if you only have a metre or two, not a lot of power is actually getting into your wort. This means your HEX temp sits fairly high, you have overshoot potential, and your ramp rate will suffer. Seems like the sweet spot for value/space is in the 6-10m range, but obviously if you can get close to 80-90% and have the space/money, then I guess do it.

On overshoot, there seems to be a critical relationship between the kg of mash (i.e. the kg of water + kg of grain, not the weight or volume of your mash tun), the HEX volume and the coil length. If your HEX volume is more than (roughly) 1/3 of your mash kg and your copper coil length in the HEX is (roughly) 2m or less - or your stainless coil in the HEX is 4m or less - your system will inherently overshoot. Using the D will prevent this, but will slow your ramp rate and unfortunately using 'I' on any sort of tank temperature control will, by the nature of it, cause it to overshoot so you don't want to do this anyway. So while PD control of these sort of setups (e.g. 70L/70kg mash, 30L HEX, 2m copper coil) will prevent overshoot, it will do it at the expense of ramp rate so you should look into reducing your HEX volume and/or increasing your coil length if ramp rate is important to you.

You can also increase your element size and/or your ciruclation rate to improve your ramp rate, but will still have inherent overshoot.

If you have a 70kg mash, a 40L HEX and 10-12m of copper coil then you are sweet. Pick the biggest element you can, circulate as high as your mash tun geometry and false bottom will permit, and just use P control (e.g. 1-5), or P (1) with a bit of D (e.g. 25-150).

This is probably old news or uninteresting news for people with 20A circuits or with good setups, but if you want to try to keep on a 10A circuit and your have some options with coil length and HEX volume, hopefully you can squeeze more out without needing to upgrade your circuit. Or go RIMS.

 

[idzy]

When doing research, there were a few important principles that I discovered:
-Total volume of HEX internal and external should be small to decrease thermal mass, increase watt density, and therefore improve ramp times.
-The internal volume of your coil should be as close ratio to your external HEX vessel volume minus coil displacement to improve heat exchange efficiency.
-The more total thermal mass and heating power not directed to the wort in your coil, the more likely you are to overshoot.

 

[Mardoo]

Idzy, can you clarify on your second point. I can't quite envision the equation you're describing.

 

[Benn]

E=MC2 - the something, something... :blink:

 

[MastersBrewery]

Mardoo he is saying the volume of water reqire to fill the hex vessel with the coil installed, should be as close to the amount of fluid the coil itself can hold. Now in most cases the hex vessel would be 2 to 3 times that of the coil in holding volume and that probably fine. But when your coil is holding a litre and it's being heat with 80 litres you will undoubtedly over shoot.

 

[Adr_0]

Hmmm... So in an 80L HEX, how's long would a 1/2" coil need to be to get 40L, ie 40L coil plus 40L water?

 

[TheWiggman]

To keep the engineers happy don't use the term thermal mass unless you are an HVAC technician - the correct term for what's being described is 'heat capacity'.

At risk of sounding like a dad I think we are overcomplicating concepts idzy

When doing research, there were a few important principles that I discovered:
-Total volume of HEX internal and external should be small to decrease thermal mass, increase watt density, and therefore improve ramp times.
Mostly right, but a roundabout way of saying it. The total volume of the HEx fluid should be small to minimise heat input to the HEx fluid. We want heat in the wort/coil (note difference between heat and temperature). Regarding Watt density are you referring to reducing coil volume within the HEx? This is not ideal. Maximise coil volume, minimise diameter.
-The internal volume of your coil should be as close ratio to your external HEX vessel volume minus coil displacement to improve heat exchange efficiency.
As above, volume of the HEx fluid should be minimised. This won't improve efficiency, it will improve effectiveness. The only way to improve efficiency is to improve insulation or use a more efficient element.
-The more total thermal mass and heating power not directed to the wort in your coil, the more likely you are to overshoot.
Truish, but sentence needs to be separated. Higher heat capacity (HEx volume) AND/OR other losses (poor insulation) will result in more energy not directed to wort. The former is more likely to lead to overshoot, the latter no.

Take a step back and think about the point of HERMS (general discussion, not directed at you idzy). Why do brewers use HERMS more than RIMS (where RIMS is wort direct-heated by an element)? Except in the case where we're utilising the HLT, it's because we're minimising the temperature the wort may be exposed to. In RIMS, even with a low density element the temperature on the surface of the element can be higher than the ideal step temperature under the wrong flow conditions. With HERMS, by using a coil as a heat exchanger, we have a buffer between the high surface temperature of the element and the wort. The larger the heat capacity of the HEx fluid, the larger the buffer and the more energy that needs to be put into the system before it results in a change in temperature of the wort.
Generally speaking - minimise heat capacity of the HEx such that scorching is avoided and you're maximising HERMS performance.

Pick the biggest element you can, circulate as high as your mash tun geometry and false bottom will permit, and just use P control (e.g. 1-5), or P (1) with a bit of D (e.g. 25-150).

In addition to minimising HEx volume, this line is all that really needs to be said.

 

[Adr_0]

I would add:
Stuff as much coil as you can into the HEX you have, to a certain level anyway.

This means that the heat being put in from your element goes to the coil/wort rather than the water around the coil. This is done over the surface of the coil, so SA and material matters.

Pump flow helps because you are not directly heating the mash, you rely on transferring heat from hot wort mixing through the mash. So the faster you can circulate (again as long as you avoid compressing your grain bed) this means that the Joules from the element get into the mash faster.

 

[idzy]

Idzy, can you clarify on your second point. I can't quite envision the equation you're describing.

I am not very good with maths, but let me see how I go. There are four variables to consider:

1. Volume of displacement made by coil
2. Volume held by coil
3. Volume of HEX
4. Element size i.e. 2200w

1. Volume of the displacement made by coil
The displacement that the coil creates will give you your internal volume of the HEX - V = π r ^2 h
To determine this you take the Outside Diameter (OD) and the length of the coil (H). In the examples below, I have used 1/2 inch pipe or 12.7mm diameter or 6.35 radius.
For 5 metre coil, the displacement is Pi * (12.7/2) * 5 or V ≈ 633.38ml
For 8 metre coil, the displacement is Pi * (12.7/2) * 8 or V ≈ 1013.41ml
For 10 metre coil, the displacement is Pi * (12.7/2) * 10 or V ≈ 1266.76ml

2. Volume held by coil
Displacement volume is not coil volume, as the coil has walls and those walls don't hold liquid. To calculate, use the same methods as above, however use the Internal Diameter (ID) of the pipe and the length of the coil (H). In the examples below, I have used 1/2 inch pipe with 0.5mm walls, which gives a 11.7mm diameter or 5.85mm radius.
For 5 metre coil, the displacement is Pi * (11.7/2) * 5 or V ≈ 537.56ml
For 8 metre coil, the displacement is Pi * (11.7/2) * 8 or V ≈ 860.10ml
For 10 metre coil, the displacement is Pi * (11.7/2) * 10 or V ≈ 1075.13ml

3. Volume of HEX
Volume of HEX is usually just the size of the pot, but it is best to measure this, due to the displacement of your element (depending on it's size) and the fill height that you will be operating the HEX at. For example, my pot is approximately 11 litres, but my HEX volume is about 10 litres.

Findings
Once you have the above information, you can then model various designs using different size pots, different size elements and different size coils.
The figures that I like to use are:
Heat exchange ratio
How much potential is being used to heat the external volume vs. internal volume.
As I have a 10 litre HEX volume, which when displaced by 8 metres of coil becomes 9 litres.
Therefore my HEX ratio is approximately 9 litres to 1 litre, or 9:1

Watt density / Heat capacity
Given I am running a 2200w kettle element, this means that my watt density (WD) is total watts / volume of HEX or 2200watts / 10 = 220watts per litre.
To determine the watt density the calculation is: watts per litre * (Volume held by coil / 1000ml)
For my HEX, for my 8 metre coil the WD is 189.22 or 220watts * (860.10/1000)
For a 5 metre coil the WD is 118 or 220watts * (537.56/1000)
For a 10 metre coil the WD is 236.52 or 220watts * (1075.13/1000)

Ideally you would stuff as much in as possible, but there are some practical limitations to this.

Mardoo he is saying the volume of water reqire to fill the hex vessel with the coil installed, should be as close to the amount of fluid the coil itself can hold. Now in most cases the hex vessel would be 2 to 3 times that of the coil in holding volume and that probably fine. But when your coil is holding a litre and it's being heat with 80 litres you will undoubtedly over shoot.

This is exactly what I am saying and was a very important discovery. You will overshoot and your watt density will be very low.

Hmmm... So in an 80L HEX, how's long would a 1/2" coil need to be to get 40L, ie 40L coil plus 40L water?

372 metres. I haven't tried it, but doubt it would be possible.

To keep the engineers happy don't use the term thermal mass unless you are an HVAC technician - the correct term for what's being described is 'heat capacity'.

At risk of sounding like a dad I think we are overcomplicating concepts idzy


Take a step back and think about the point of HERMS (general discussion, not directed at you idzy). Why do brewers use HERMS more than RIMS (where RIMS is wort direct-heated by an element)? Except in the case where we're utilising the HLT, it's because we're minimising the temperature the wort may be exposed to. In RIMS, even with a low density element the temperature on the surface of the element can be higher than the ideal step temperature under the wrong flow conditions. With HERMS, by using a coil as a heat exchanger, we have a buffer between the high surface temperature of the element and the wort. The larger the heat capacity of the HEx fluid, the larger the buffer and the more energy that needs to be put into the system before it results in a change in temperature of the wort.
Generally speaking - minimise heat capacity of the HEx such that scorching is avoided and you're maximising HERMS performance.


In addition to minimising HEx volume, this line is all that really needs to be said.

Thank you! I am not an engineer/mathematician and most terms used are probably wrong I have tried to use simple to understand language as it is the way I understand it.

Also good to point out that when watt density is too high, you are back towards a RIMS type problem where you are scorching wort. The point of these calculations is that there is a spectrum between high watt density and low watt density. You want to avoid scorching, but you also don't want to ramp 1c every 4 minutes either, otherwise you will have a very lengthy brew day.

My post is lengthy and probably has loads of errors, please be gracious.

 

[MastersBrewery]

My favourite herms set up was a stainless convoluted counter flow chiller (avaliable from ibrew) with a continuous hot water system in loop, the system had two of these for a 1.5BBL mash tun. Your liquor to hex water ratio would have been pretty even.

 

[TheWiggman]

I would add:
Stuff as much coil as you can into the HEX you have, tomatoes a certain level anyway.

Tomatoes for sure.

Look at your 4 points there though idzy:

1. Volume of displacement made by coil (FYI - displacement is volume)
2. Volume held by coil
3. Volume of HEX
4. Element size i.e. 2200w

If the statement was made 'minimise water volume in heat exchanger' and 'maintain flow to prevent scorching' then you will naturally have a high displacement made by the coil. No maths or anything required. Likewise for Adr_0's statement about cramming as much coil in as you can - by doing this you are minimising the volume in the heat exchanger. Only when you're trying to find ideal power, length, relationship between flow and frictional losses to heat transfer coefficients etc. does the maths need to come into play.
I'm just trying to simplify it.

 

[Adr_0]

Tomatoes for sure.

Look at your 4 points there though idzy:

• Volume of displacement made by coil (FYI - displacement is volume)
• Volume held by coil
• Volume of HEX
• Element size i.e. 2200w

If the statement was made 'minimise water volume in heat exchanger' and 'maintain flow to prevent scorching' then you will naturally have a high displacement made by the coil. No maths or anything required. Likewise for Adr_0's statement about cramming as much coil in as you can - by doing this you are minimising the volume in the heat exchanger. Only when you're trying to find ideal power, length, relationship between flow and frictional losses to heat transfer coefficients etc. does the maths need to come into play.
I'm just trying to simplify it.

What I meant by that was, cram as many tomatoes into your HEX as possible. Add good beef or lamb mince, mushrooms, smoked paprika, thinly sliced zucchini, chicken stock, cayenne pepper, black pepper and salt to taste. Cook and reduce over 1-2hrs in parallel with your mash.

Don't you also-rans love autocorrect? I have some serious issues with mine...

 

[Mardoo]

Sort of the opposite of sous-vide.

 

[idzy]

Tomatoes for sure.

Look at your 4 points there though idzy:

• Volume of displacement made by coil (FYI - displacement is volume)
• Volume held by coil
• Volume of HEX
• Element size i.e. 2200w

If the statement was made 'minimise water volume in heat exchanger' and 'maintain flow to prevent scorching' then you will naturally have a high displacement made by the coil. No maths or anything required. Likewise for Adr_0's statement about cramming as much coil in as you can - by doing this you are minimising the volume in the heat exchanger. Only when you're trying to find ideal power, length, relationship between flow and frictional losses to heat transfer coefficients etc. does the maths need to come into play.
I'm just trying to simplify it.

Not sure of your point or objective here...