Dedicated Herms Guide, Problems And Solution Thread

[Mardoo]

Would you want a wider-than-taller mash bed to deal with the increased flow rate? (E.G drawing from a greater surface area so better able to deal with a higher flow rate?)

 

[Adr_0]

Would you want a wider-than-taller mash bed to deal with the increased flow rate? (E.G drawing from a greater surface area so better able to deal with a higher flow rate?)

Flippin A you do. Better for temp distribution too.

 

[Tex N Oz]

And yes, you might want as much coil as you can get your hands on and a big pump to back it up.

I can't see how a heat exchangers efficiency becomes greater with increased flow rates. Increase the velocity and you pass the apex of exchange rate for a given pressure drop due to laminar flow and then your efficiency has dropped out. This may work ok if your goal is to reduce the dt of your supply and return, but that's because you are working the edge of the curve of your heat exchanger's efficiency envelope.
It's my advice to pump the same volume at a lower velocity through a larger diameter heat exchanger with less thermal mass and greater conductivity.

 

[Adr_0]

I can't see how a heat exchangers efficiency becomes greater with increased flow rates. Increase the velocity and you pass the apex of exchange rate for a given pressure drop due to laminar flow and then your efficiency has dropped out. This may work ok if your goal is to reduce the dt of your supply and return, but that's because you are working the edge of the curve of your heat exchanger's efficiency envelope.
It's my advice to pump the same volume at a lower velocity through a larger diameter heat exchanger with less thermal mass and greater conductivity.

You've touched on a correct principle (why industrial heat exchangers have banks of tubes) and process gain (separate to controller gain) goes down with increasing flowrates. Unfortunately these factors are more than offset by the increased turnover in the mash (our mash ramp time directly hinges on this) and the increased power (heat power, not electrical) in the heat exchanger.

So overall, higher flowrate = better ramp times. We did mention a large pump will be required.

Trust me - I'm a vegan.

 

[TheWiggman]

Quite the opposite tex n oz, lower diameter is better for conductivity except in one particular case.
In these systems the flow shouldn't be laminar. I mean that practically, typical setups for home brew have a Re > 15000. The higher flow the better as this increases turbulence and thus the convection coefficient between the tube and liquid inside it.

 

[Tex N Oz]

You've touched on a correct principle (why industrial heat exchangers have banks of tubes) and process gain (separate to controller gain) goes down with increasing flowrates. Unfortunately these factors are more than offset by the increased turnover in the mash (our mash ramp time directly hinges on this) and the increased power (heat power, not electrical) in the heat exchanger.

So overall, higher flowrate = better ramp times. We did mention a large pump will be required.

Trust me - I'm a vegan.

I'm getting that it's all a matter of rapid, yet controllable heat exchange. It's kind of boggling my brain as I've never worked on this scale in either size or time.
I'm thinking a different approach all together. I've got a 5 kW steam boiler, heaps of small steam valves and I've got a few Innotech and Siemens controllers laying around. Real proportional controllers.
Funny thing is one of the Innotech controllers is from the Rio Tinto plant there in Gladstone. Smoked a $2 capacitor but since I'm not qualified to solder in a new one, they bought a new controller and I brought this one home for a quick rebuild. Anyway, I might blow the dust off those and see if I could weld up a steam injection mixing chamber. The problems I can see already are I'll have to balance the pump flow rate with the throughput of my filter bed and I'll have to find a valve with the right control authority for such a small mass of steam. Process control won't be an issue at all. I can't see I'd ever need more than 2 kg of steam for a 23 litre batch or double for a 50.
Not a problem since I'll have 5 kg @ 127°C.
Now I've got visions of calandria driven boil kettle and glass filter beds. GEEZUS!!

I think the best place to start is designing a mash tun with HUGE filter bed surface area. Never be low on flow..

Don't trust me.... if it moves I'll eat it..

 

[droid]

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

1.8L 7lpm 1.8L 19lpm 42L 7lpm.jpg

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

ok so here I am, initially thinking "should I even open this pic because what the hell will it all mean" I opened it and everything seems understandable, maybe i'm not so silly after-all
... until I tried to close the window. yep I just kept on clicking on your close window icon instead of mine...uhah some people you just cant reach

 

[Mardoo]

Tex just build yourself a steam jacketed mash tun. If you can't build the jacket just wrap a pot in 3/4" copper pipe.

 

[Adr_0]

If you can get it superheated (which 127°C would be, I guess) it would be a sweet oak barrel steriliser on the side.

I did steam injection once with my bro, maybe 12-13 years ago. Worked really well, not sure how even the mash temp was though, so would definitely put a recirc pump in there.

 

[Tex N Oz]

Tex just build yourself a steam jacketed mash tun. If you can't build the jacket just wrap a pot in 3/4" copper pipe.

Well my theoretical design is just like a RIMS but with a massive flow rate (as high as I can get through the filter bed) and instead of a heating element, it will direct inject steam through a sintered stainless pipe into the wort..
This way I could make step adjustments extremely fast and accurate or very slow and accurate with never a chance of scorching the wort at 127°C. I've thought about the whole tun wrap or copper tubing but that adds another process. In order to modulate the temperature of steam, you have to reduce its pressure (to adjust temperature) and the condensate has to be sent back to the boiler via condensate return tanks and pumps.
IF I add a steam boil kettle then it's inevitable I'll add those and instead of direct steam injection, I'll have a steam HEX in the RIMS cylinder.
I keep going back to steam... then avoid its complications and go to electric or gas heating in my design, then back again to steam..... all professional breweries use steam so why couldn't I?
I originally took on the boiler to make a pasteurising machine and autoclave. Now it's looking like a multi-purpose heat source for all my processes. If I keep my boiler stand-alone and my condensate return traps and tanks, etc, stand-alone, it will be just that.

If you can get it superheated (which 127°C would be, I guess) it would be a sweet oak barrel steriliser on the side.

I did steam injection once with my bro, maybe 12-13 years ago. Worked really well, not sure how even the mash temp was though, so would definitely put a recirc pump in there.

Great minds think alike.. A steam autoclave was already on the menu..

 

[TheWiggman]

HURRY UP AND START A NEW THREAD WITH PICS

 

[Mardoo]

I'd love to go steam but $$ and my engineering abilities are lacking. I contemplated a reverse HERMS, circulating the hot water around the mash tun but finally decided to go with tried and true. I'll let the professionals build my steam jacketed tun, one day.

 

[Moad]

Would a 10L HEX vessel with 12m of coil (assuming I can fit it in) and a 2400w (3600w might be better) element be sufficient for ramping and step mashes with up to 90L mash? Will I need to directly heat the MT as well for ramping?

I am also under the impressions I will need a bigger pump than the March 809? I use these on my current rig and get a decent flow rate.

I know there are lots of variables with the coil and hex size but any advice would be great.

 

[QldKev]

Would a 10L HEX vessel with 12m of coil (assuming I can fit it in) and a 2400w (3600w might be better) element be sufficient for ramping and step mashes with up to 90L mash? Will I need to directly heat the MT as well for ramping?

I am also under the impressions I will need a bigger pump than the March 809? I use these on my current rig and get a decent flow rate.

I know there are lots of variables with the coil and hex size but any advice would be great.

I run a 112L (final vol) batch on my 3V. It's a 82L mash tun and 140L kettle (pre boil is very full). I run a March 809 and it flows enough for me. The pumps I brew with at work flow about the same as a March 809, and as you can imagine the brewery is a lot bigger than our systems. You need adequate flow, not monstrous flow. Your proposed HX vol and coil will be heaps / overkill for the system. I use the herm-it coil for mine which from memory is about 3m long in 2.5L water. I know people run 3600w into this system, and read once on here someone was using 4800w, but they never posted the outcome. Remember the shortest coil in the min amount of water that can do the job is best. 2400w will not be enough power to ramp a mash of that size in a decent timeframe. 3600w is about the min and would be a slow ramp that you would need to allow the time in the mash schedule, 4800w would be better but still on the slow side. On my 3V I did also run an 2400w internal rims element in my mash tun, but it has died and I don't plan on replacing it. For mine I'm looking at changing the house kettle element (approx 1850w) in my current herms and going to use possibly 4000w (2 x 2000w elements). I don't like to run 2400w elements as I use extension cords to get power to where I brew and I find the plugs overheat in them, and fuse the plug into the socket. I'm even possibly looking at going 3600w which will give me a slower ramp rate but makes the setup neat and clean. I'm thinking one of these elements from 5star, and wiring it as 2 x 1800w. But my mash is smaller than your 90L one.

I have a calculator on my website qldkev.net where you can plug in the figures and work out heating times. The perfect ramp rate is 1degree per minute.
So for your 90L mash
2400w would give you 0.38c per min
3600w would give you 0.59c per min
4800w would give you 0.77c per min
6000w would give you 1.00c per min
7200w would give you 1.11c per min

If you could get 6000w/7200w that would be great, but I think you could get away with 4800w and allow the ramp times into the mash schedule. Remember you are only ramping the wort through the heat exchanger, so in the real world the rise in temp of the wort exiting the herms will most likely rise in temp faster, and the mash bed will lag behind. The slower the flow through the HX the quicker you will see the outlet of the herms raise, but the more the mash bed will lag behind.

For coil length and water bath size, I think somewhere around 6m would be heaps.

 

[RelaxedBrewer]

I have a calculator on my website qldkev.net where you can plug in the figures and work out heating times. The perfect ramp rate is 1degree per minute.
So for your 90L mash
2400w would give you 0.38c per min
3600w would give you 0.59c per min
4800w would give you 0.77c per min
6000w would give you 1.00c per min
7200w would give you 1.11c per min

Hi Kev,

Thanks for posting that, useful stuff. I assume your calculator is based off your HX system. I am trying to get my head around how the flow, volume, coil length/diameter and element wattage all work together.

If I was to change the parameters would you be able to get a lower watt coil to heat faster? Say if we increased the volume of water in the HX and the length of the coil? Or are we limited by the actual heat output from the element and its ability to keep up with the heat transfer.

My large system has a 130L mash tun so I max out at about 100L. Looking at your calculator I should be able to use a 2400w element to accurately control a single infusion mash but not step. If I want to step I would need 6000 -7200w which is totally impractical for me.

 

[QldKev]

Hi Kev,

Thanks for posting that, useful stuff. I assume your calculator is based off your HX system. I am trying to get my head around how the flow, volume, coil length/diameter and element wattage all work together.

If I was to change the parameters would you be able to get a lower watt coil to heat faster? Say if we increased the volume of water in the HX and the length of the coil? Or are we limited by the actual heat output from the element and its ability to keep up with the heat transfer.

My large system has a 130L mash tun so I max out at about 100L. Looking at your calculator I should be able to use a 2400w element to accurately control a single infusion mash but not step. If I want to step I would need 6000 -7200w which is totally impractical for me.

The calc is based on the energy required to heat water. It is not dependent on any specific system. It will work for any system. I plugged in 90L for the figures I quoted, as you mentioned it was Moad's mash size. Then I just used 10 degrees as the amount of temp change and divided the time taken by the 10 for degrees per minute. The same calc can be used to size heating elements in a HLT etc. It is the basic principle of working out how much energy we need to get into the wort to heat all the wort and mash bed, and not just the wort exiting the herms coil.

You want enough pipe surface area in the herms to be able exchange all the heat the elements are producing, otherwise the herms will boil. Increasing pipe length or diameter will give more surface area for heat transfer. So increasing pipe via length and diameter beyond the potential for the heaters is just going to introduce potential overshooting due to the increased volume of the water bath. A decent pid will smooth this out, but I would design the system to minimize it. Although I don't see a lot of double helix coils for a herms, if I was building one that's what I would use. I would just keep a little bit of spacing on the inner coil.

Flow through the herms will also effect the amount of time in contact with the heating. The slower the flow, the longer it has to transfer the heat. But a slower flow will mean the mash bed lags behind more. The faster the flow the more uniform the heat is in the mash bed. But we are limited to flow potential by false bottom designs. I'm probably lucky on my 3V that my false bottom allows me to run the pump with fully open valves. You could have a massive pump on the system, but without a false bottom to allow that flow it is a waste of time.

When ramping, we aim to keep the heating elements on full the entire time until the entire mash bed is virtually up to temp without over heating any of the wort. After all it is the element(s) that are producing the heat. If you think of a swimming pool and you had a really slow flow of hot water into it, it would take a very long time to heat up.

We tend to aim for a ramp rate of 1c per minute. But you don't have to achieve it. You would simply update your brew schedule to allow for the lower ramp rates. Obviously we are still mashing during the ramp times. So you would update the schedule to suit.

eg. just as an idea we could compare a schedule for a 1c per min Vs 0.5 per min (temps rounded to make maths easier)

System 1 - 1c per min
62 for 35mins
ramp to 72c over 10mins
72 for 20mins
ramp to 77c mash out over 5mins
hold mash out 5mins

total mash time 75mins (inc mash out)

System 2 - 0.5c per min
62 for 30mins (remove 5min from here for extra ramp up to 72 time)
ramp to 72c over 20mins
72 for 10mins (remove 10min as we allow time for ramp from 62 and also to 77 mash out)
ramp to 77c mash out over 10mins
hold mash out 5mins

total mash time 75mins (inc mash out)


You can see we start ramping temps earlier, but hit the next step later, so the enzyme activity is in theory averaged out. (even though activity at lower temps is technically slower) Then you would taste the beer to see if the desired beer profile was achieved and tweak the schedule to suit. My current 3V's rims element is dead, so with just the current herms I can only get about 0.4c per minute. My current mash spends a lot of time ramping as I start my mash at 55c, and I'm not in a hurry to update the herms at the moment.

I remember one guy on here was trying constant stepped mashes. He had a schedule something like, mash in at 55c and ramp to 77c over 75mins. You could even tweak it to 2 simple steps, like mash in at 62c and ramp to 72c over 60mins, and then a faster ramp to mash out or even skip mash out. You don't need a lot of power in the herms for this style.

 

[Moad]

Thanks Kev,

I'm a bit confused, you say the proposed 10L/12m/2400w would be overkill but then say the herm-it coil you use is sufficient. Do you only use it to hold temps and not ramp or are you OK with the slower ramp times?

Your volumes are pretty much what I am after so would be very interested to get some more info on your setup. I am still making my way through this thread to hopefully not repeating questions already answered before.

I had no idea what was involved in designing a HEX/HERMS

 

[RelaxedBrewer]

Once again thanks for a couple of fantastic posts Kev, really helpful.

 

[Adr_0]

Kev, that advice is bloody unreal.

For those interested, assuming the top of your coil is 0.5m above the suction of the pump:


These were using the pump curves from the March website, assuming 50Hz. Sure enough the 815 is a little over the top.

The 809-PL-HS/SS (which is probably the most popular beer pump) will do the job comfortably. That's 13.5lpm with a 4m coil, ~11lpm with an 8m coil and still 10lpm with a 12m coil. Ample either way.

A 10L HEX and 90L mash will be absolutely fine. But yeah assuming the mash kg x 65 rule, you need 5800W assuming 100% effectiveness of your coil to get 1°C/min, but realistically a coil will only be 40-60% effective.

I will try and do some simulations over the weekend sometime.

 

[idzy]

I can ramp with reasonable times with my HEX 2x2400w (one OTS) for a 400 litre MLT.