Fully Automated Brewing System Design

[user 4286]

Not sure if this has been asked before, but is there any reason why you're trying to do the whole lot in one go?

Maybe I've been brainwashed by years of Systems Engineering, but I can't help but think that it would be easier to break this down into smaller chunks and get them working first. Of course you need the overall view of the system to know how it all fits together, but once you've got that you should be looking at how you can partition it into smaller subsystems with defined interfaces so that you can build it up in parts.

The actual brewing part is reasonably complicated without adding in everything to get all the right inputs into the right place at the right time. If you could define the interfaces better, then you could build a subsystem that creates wort from grain, water, hops and yeast and get that working. Then build the fermentation subsystem. And then build the rest of the ingredient delivery subsystem to fully automate the process.

Looking at your diagram, at first glance I think the fermentation fridge is too tightly coupled into the rest of the system to partition it up straight away. I would be looking at ways to simplify the interface there and also trying to bring all the 'inputs' over to one side so you can clearly see what needs to go into the 'brewing' subsystem.

 

[Bandito]

Not sure if this has been asked before, but is there any reason why you're trying to do the whole lot in one go?

It has been said before but not asked. So I will actually answer it - The controlling program, Labview, will allow it to be split up. I dont see any advantage to dealing with one little part - I may as well dive in head first and get it all done in one foul swoop. I dont think it will be any more complex this way - in fact I expect it to be less complex doing one big thing taking 2 months, as opposed to 10 small things taking 3 weeks each.
Thats my view anyway, so that how I am going to do it.

Nicely put question.

Oh, and dont get me wrong, it will be split up, but that will happen in the programing stage, not in the design stage.

 

[Jono_w]

I dont think it will be any more complex this way - in fact I expect it to be less complex doing one big thing taking 2 months, as opposed to 10 small things taking 3 weeks each.

Is that the build time ? 2 months?

 

[bigfridge]

You need to either study or stop putting your foot in your mouth.

Just because you are a process designer does not qualify you to design a fully automated brewing system. You need to have some idea of the processes involved in brewing.

I am sure at work you have some kind of guide to let you know what the requirements are for any project you work on. This project sounds like you are flying by the seat of your pants. That may be saying to much as that requires skill and knowledge.

You are working away at a design and to be blunt you have no idea what you are up against. You have simple brewing processes to keep track of that you have shown you do not understand. You then complicate the contraption with dust from milling and humidity from heating water. You want this to be easier to use then a bread machine.

I am not saying it can not be done but you need to get educated on the process and all the dirty parts. Unless you do this you are going to end up with an encrusted pile of stuff not even worth much as scrap because of all the cleaning needed to scrap it.

One more potential trouble to detect. Boilover! That alone will ruin the brew bot in 30 seconds.

Believe it or not I support you on this project. I think it can be done but still question why. It will make a nice conversation piece but I do not see a market for such an expensive system. This entire project reminds me of some films with the inventor that designs and automatic egg fryer. It starts with the chicken and always ends up in a mess. All because he can never see past the idea of the invention. So I return to my suggestion that you look past the idea and learn about the process and all that can go wrong. Then you can take your knowledge of process control and come up with something that may work.

Well said !

I too support people 'having a go' - but I just hate to see someone wasting their money on an unworkable - and possibly dangerous piece of equipment.

I think that it is the overwhelming view of people here that you should get some brewing experience and then try to automate your process.

I asked if you actually indended to build it as I beleive that your design is and will be unworkable. Basic brewing terms like DMS, Boilovers need to be understood first. BTW, you can't prevent boilovers by increasing the headspace of the boiler - you have to reduc ethe energy imput or drop the wort to below the boiling point. Once you have a full coverage of foam the surface tension of the wort causes the bubbles to remain and they rise up like honeycomb. This is why commercial breweries have detectors to sence the rising wort level and cut the energy input.

Once you know how to brew then you need to get some industrial/brewery design experience. Then you will know that the 'stick' that I was referring to was a sticking valve and 'mix proof piping' is a failsafe design principle that makes it impossible for the wrong thing to happen. This can be an automated system where a single valve stem operates two valves at the same time or manual where the head brewer keeps a piece of connecting pipe in his office that is only issued when it is time to do a cleaning run. Google up 'pipe fence' for a start.

Please consult any textbook on brewing (ie one not written by a home brewer) and you will find all the information that you seek.

Best of luck,
Over and out.

 

[Bandito]

Quote: you can't prevent boilovers by increasing the headspace of the boiler - you have to reduce the energy imput or drop the wort to below the boiling point."

Now thats the kind of reply I am after. Will implement in the P&ID phase.

By working together we can achieve a fully automated system that works. Thats my view anyway.

Its not me against the world, its us aginst the world! Surely, if we work together we can achieve a workable system.

YES WE CAN! - Bitch!

 

[LethalCorpse]

No, it's pretty clearly you against the world. Sorry, mate, but noone thinks it's going to work. No one's working with you on this quixotry. You've had a fairly wide range of highly knowledgeable people in this thread, including some pretty talented engineers and brewers. There's a difference between forging ahead to follow your dreams in spite of all the naysayers (or other equally ridiculous Disney movie tagline) and ignoring the sage advice of a whole lot of people who know the field far better than you do.

A fully automated brewery is possible. It's just not feasible. The people in this thread who have many years of experience as brewers, the folks who have studied for ten years to become electrical, mechanical, control system, chemical or other types of engineers, the guys who have BUILT real working breweries - none of them would attempt what you're so insistent upon attempting. That's because they know enough about the process to see where the pitfalls are, and how difficult they would be to overcome. And they've got the expertise to overcome most of them. You know how to make process flow diagrams, this is your only qualification for building this system, and you think you're going to do it where none of those other guys could? That's a pretty hefty serving of ignorance and arrogance right there.

In order to successfully do this, you're going to need to learn everything about how all-grain beer is made, an awful lot about process control, a fair bit of electrical and software engineering, a lot about food science and food/beverage manufacturing processes, a high level of plumbing skills, not a small amount of microbiology and a bunch of other fields. These are not things you learn on the fly. These are things which take four years of university education and another four years of industry experience to properly learn. I suggest you start with the brewing, because you're going to need an awful lot of beer to get through those other degrees. If you build a working brewery, you can then automate bits of it at a time - start with temperature control and work from there. Perhaps one day you'll have automated enough parts to call it an automated brewery. But if you think it's going to happen by poking at PFDs and asking questions like "do I need a kettle" you're certifiable.

 

[Bandito]

Wow, of everyone on this forumn I thought I would get more support from a fellow ocau member! Wow, thats pretty heart breeaking right there! Yep I'm lost for words.

 

[LethalCorpse]

Sorry dude, but you weren't listening to anyone - I had to be blunt. I'm all for big, expansive projects and pushing the boundaries of what you know you can do. But this one just isn't going to happen, at least not until you've learnt a whole lot more about brewing and process control. The best way to learn that is to start brewing - you can do that right now with a manual, gravity fed system.

 

[Bandito]

Still no words. Well, one, shattered.

 

[newguy]

Bandito,

Not sure if you're aware of these electronic liquid level sensors. Just ran across them today and they look pretty good.

 

[OzMick]

Bandito,

Not sure if you're aware of these electronic liquid level sensors. Just ran across them today and they look pretty good.

Don't think they would be the best idea for use with beer, they're not rated for boiling liquids, and beer is reasonably acidic. Differential pressure is best for most volume measurements around a brewery (when you make an assumption of gravity), but might be overly expensive unless you're buying raw pressure sensor modules and doing a lot of circuit work yourself. As for a kettle, a conductivity sensor for overflow or high level might be best, not sure if the foam would trigger it but probably would. Conductivity sensors are easy and dirt cheap to even build an array of (darlington array ICs), in that way can get a crude volume measurement like a very primitive version of that sensor newguy linked to.

Can copy the basic design from http://jaycar.com.au/productView.asp?ID=KG...mp;form=KEYWORD as the parts are probably only a few cents, mostly paying for the instructions and PCB.

 

[technocat]

Wow, of everyone on this forumn I thought I would get more support from a fellow ocau member! Wow, thats pretty heart breeaking right there! Yep I'm lost for words.

I guess LC has put into words what many of us have been thinking. Sorry you didn't see it coming.

BN

 

[bigfridge]

Quote: you can't prevent boilovers by increasing the headspace of the boiler - you have to reduce the energy imput or drop the wort to below the boiling point."

Now thats the kind of reply I am after. Will implement in the P&ID phase.

By working together we can achieve a fully automated system that works. Thats my view anyway.

Its not me against the world, its us aginst the world! Surely, if we work together we can achieve a workable system.

Mate, LC put it clearly that this is not something that is easy to do. You seem to think that this is a 'team effort' but your contribution is that you can draw it up - you need other team members to design it for you. That is a big ask when it is YOUR project for YOUR hobby. There doesn't seem to be much in it for the rest of the team.

Drawing steel buildings is a lot different to the industrial brewing process. While you may get the design finished it will not be workable or reliable. You are assuming that everything will work out 'in theory' but in practise there are sensors to fail, wires will burn out/fall, valves will gum up and refuse to move, grain dust WILL get into your fermenters or bottles/kegs and so on.

There is a fundemental safety issue in your design that you assume that the command that you program into Labview will happen. Any number of events can occur which will result in it not happening eg Labview has a bug where the command is not properly executed, the device or motor does not operate, the valve does not move to the desired position as something has fallen down to block it, a power failure results in the process re-starting at the wrong place in the program - the list is endless. Professional systems ensure that any failure renders the process 'safe', any action is measured to ensure that it has completed before it proceeds to the next step ie you need a sensor.

This is the experience that LC is talking about, the same experience that you argue "no it won't" against when issues are raised.

It would think that you will be lucky to complete any single action (ie software design, procure and install plumbing etc) by your deadline - let alone the whole project.

But don't let any of the above stop you.

 

[newguy]

Don't think they would be the best idea for use with beer, they're not rated for boiling liquids, and beer is reasonably acidic.

I was thinking that it could be used to measure strike and sparge water volume. If you know how much water went into the batch, the volume of wort will always be less than that. But that's just my thinking. I know that with my system if I have x kg of grain and I use y litres of water I'll always end up with ~55l in the kettle. I think a flow sensor would be best but they're really expensive.

As for a kettle, a conductivity sensor for overflow or high level might be best, not sure if the foam would trigger it but probably would. Conductivity sensors are easy and dirt cheap to even build an array of (darlington array ICs), in that way can get a crude volume measurement like a very primitive version of that sensor newguy linked to.

The problem with liquid level sensors on a kettle is that the hops, trub, and other crap in the kettle will foul them quickly.

 

[user 4286]

It has been said before but not asked. So I will actually answer it - The controlling program, Labview, will allow it to be split up. I dont see any advantage to dealing with one little part - I may as well dive in head first and get it all done in one foul swoop. I dont think it will be any more complex this way - in fact I expect it to be less complex doing one big thing taking 2 months, as opposed to 10 small things taking 3 weeks each.
Thats my view anyway, so that how I am going to do it.

Nicely put question.

Oh, and dont get me wrong, it will be split up, but that will happen in the programing stage, not in the design stage.

But... by your own admission you don't know all that much about the actual brewing process.

Why don't you at least define clear interfaces for the brewing part so that when you learn some more about the process you can change that bit without needing to redesign the whole system. You could then simplify the brewing to just a simple infusion mash for now, work out the kinks in the rest of the process, then go full HERMS/RIMS later on.

Anyway, just my thoughts. I think if you build it up in pieces you're less likely to end up with a system that doesn't work at all, and should at least get some stuff built that will be actually usable for brewing.

 

[Bandito]

JFK: "We will do these things, not because they are easy, but because they are hard!"

Sienfeld: "No body wants to build bridges, Its really really hard" - I just finnished detailing a bridge extension in padstow heights NSW, was hard, but not too hard.

Decided a few days ago to get a UPS for the pc - too many blackouts and brownouts here lately. That solves the power issue.

I suppose the grain area will be enclosed mabee with a fan and filter, but with the fan after the filter to avoid grain dust explosions. motors in that area might also need covering up to avoid grain dust explosions in the confined space.

 

[yardy]

JFK: "We will do these things, not because they are easy, but because they are hard!"

Sienfeld: "No body wants to build bridges, Its really really hard" - I just finnished detailing a bridge extension in padstow heights NSW, was hard, but not too hard.

David Daye: If God had intended us to drink beer, He would have given us stomachs.

 

[LethalCorpse]

JFK had 25 billion dollars (in the sixties) and a massive team of military and private aeronautics specialists. Bridges are built with millions of dollars, and large teams of civil engineers, structural engineers, project managers, concrete specialists, governmental liaisons, even metal detailers. And it's still really really hard. What you are seeking to do is much smaller than a commercial brewery, but not much less complex for it.

 

[katzke]

Just went back and read your original post.

You say you have been planning this for a few months but still have questions about the basic brewing process.

I understand your desire to do this and that you see objections as a challenge to overcome.

But it may be time to step back and evaluate the project to see what you have learned so far.

You are trying to do a lot of stuff with one machine. Culturing yeast, brewing, and fermenting are all totally different things.

I can not comment very much on the yeast culturing, as I do not do it. I have looked at the subject and decided it is not for me and dry yeast is all I need now.

Brewing from the dirty part of milling the grain until you start the cooling process is something less problematic. Once you cool the wort then it introduces the challenge of cleanliness. The way most brewers brew, the dust and steam are not a problem because we strip everything down and hose it off, at least I do.

Fermenting is a process all its own. It can be as simple as leave it alone for a few weeks to as complicated as yeast harvesting, dry hopping, and for the mad brewer a true secondary fermentation. Then there is the choice between lager and ale.

I think the idea of making one machine or having one control for the entire process is folly. You do have 3 different things going on. Will the program be able to keep track of multiple yeast cultures and run a brew session at the same time? What about monitoring your fermentatio and brewing at the same time?

As has been pointed out you have many challenges just in the brewing part. If you are going to have a true set it and forget it system how are you going to measure gravities for duplication of recipes?

You answered one question by saying you are going to get a UPS for the computer. How will you handle when the power does go out? The program going to run on even with no power to the system? Will it abort the brew and give you a message letting you know it stopped the process and ask if you want to continue? Lots to think about and no easy answer as it all depends on where in the process the power failed.

I think you are learning from the exercise but it may be time to step back and see where it is going. I would hate to have you compare your self to Edison and find a thousand ways to not build an automated home brew system.

 

[randyrob]

Hey Bandito,

Have you read thru any of the other threads on this site from people that have gone down the same path?

Here are a few that are a decent read:

Zizzle's BrewBot by Zizzle

Automating a Brewery by WortGames

How High Tech is your Brewery by SpillsMostOfIt

Brewing software for automation freaks by zwickel

I have pretty much done what everyone else is suggesting, started of with a "manual AG brewery" and got about 60 batches under my belt and slowly started automating bits and pieces of it until i'm at the stage now where i can set it and forget it but it took me an entire year of tinkering to get to that stage.

I haven't updated my site for a bit but have still been working hard in the background. i've just purchased PDMS and am in the process of modeling up a 3d version of Mark 2 of my Brewery.

Rob.

 

[randyrob]

Another option that would be worth considering is the Beckhoff DIN-rail I/O modules

LINKY

your looking about $5k worth of hw/sw but it should be sufficient enough to run your home brewery.

 

[manticle]

I think the idea is great as part of the enjoyment of building your own brewery is the building and innovation itself.

However I think the point made by several that you should research basic brewing processes is only going to help your design.

There is a guy who used to post here who was automating his brewery. Someone mentioned it a few posts back. I'll see if I can hunt up the thread and link it as it may give some pointers. Definitely familiarise yourself with basic processes though as it will only help you in the long run.

 

[manticle]

Couldn't find the link I was after (think the username was Jonathan) but in searching I came across these which might be useful (as well as users you could PM for advice):

http://www.aussiehomebrewer.com/forum/inde...&hl=Brewbot

http://www.aussiehomebrewer.com/forum//ind...mp;hl=automated

http://www.aussiehomebrewer.com/forum//ind...mp;hl=automated

http://www.aussiehomebrewer.com/forum/inde...mp;hl=automated

 

[Zizzle]

Conductivity sensors are easy and dirt cheap to even build an array of (darlington array ICs), in that way can get a crude volume measurement like a very primitive version of that sensor newguy linked to.

Bingo. It's actually easier than you think. Just have your vessel at GND. Nearly all micro controllers have muxed ADCs. e.g. 8 analog inputs. You need pretty simple signal conditioning circuitry between the probe and the ADC input. And a pull up resistor.

From there software can do the rest. Experiment and see what ADC readings correspond liquid being present. Add some debounce and threshold code and you are done.

On the brewbot, I only use 3 level probes. They have been one of the most reliable parts. Here are the two in the HLT:

But it would be good to do some experiments with a graphite rod in the HLT for level measurement. Graphite being pretty inert and resistive, it should be possible to measure some resistance change due to water level.



But like others have pointed out, I'm not sure if Bandito is at this detail level yet. I wonder if some off-the-shelf gear may help him here.


Not sure if anyone linked it yet, but HERMAN was inspirational for me in some areas.
http://hermanmachine.spaces.live.com/


BTW I think RandyRob is really on the right path with his automation project. Elegant, simple design. Automation in stages.

 

[komodo]

Building bridges big ******* deal - been there done that bought the post card. I did Webb dock bridge.

Spencer Street Railway Station Redev. worked on that one too.

Fed Square - yep

Perth Convention and Exhibition centre - yep again.

Adelaide law courts.

Just a few of the projects I\'ve worked on .
Just cause you\'ve worked as a drafty on projects doesnt mean you know anything about the actual design process and fabrication process of the entire project. Plus your working as part of massive design team who have started working on the project long before you as a drafty have even been concidered.

Personally I think you have NFI. I also think your young and have your head in the cloud as to the scale of this project . 2 months is absolutely ******* dreaming. You\'ve been given bloody good advise by people who\'ve been there done that and your ignoring them.

 

[paulwolf350]

For my 2c. I would like to say that i think it is possible to automate up to the end of the boil, but i dont see how you can automate the fermentation side. There is a lot of non science and "feel" for good and bad fermentation. Its not just a pitch, wait, pump scenario.

and as an aside, I feel your pain. I did three AG batchs, and thought I can automate this, piece of piss, I work with this every day. Well mate I am nearly $2000 down the road and have not automated 1 valve yet. ( got the gear, just havent). I really want you to succeed, (i might get some great ideas, programming or drawings) but in 2 months and $3000, from your original post, I dont think it will happen.

Best of luck, If I can help i will

Paul

 

[paulwolf350]

Couldn't find the link I was after (think the username was Jonathan)

I would like to see that one again too, it was smokin!

Paul

 

[Jono_w]

I would like to see that one again too, it was smokin!

Paul

Here it is Paul..

LINK


Cheers..

 

[thunderchild]

A wise pilot once said that evne though he flew fully automated Jumbo's for a living he loved flying a Cessna - Said it was real flying...

 

[Bandito]

Thanks for all the links guys, that great!

Not sure where this 2 month time frame came from? think someone misread something, the op states that 7 months and about $3000. Got to do a schedule soon to work out milestones.

Komodo, you worked on some nice stuff, you obviously arent scared by extremely complex projects. Its a requirement in our high end architectural sector. So many job applicants get scared off when they see how complex the stuff is - they just want a nice little square house or portal shed to detail. The point is that these sort of things, going to the moon, bridges, architectural nightmare buildings arent done by people that are scared by any amount of complexity - so although it sounds like I am not listening, I really dont see it as any where near the complexity of any of your buildings or quite a few of mine - if that sort of complexity scares people so be it, but I am used to it, so I will just get on with the job.

I saw that pilot thing on discovery channel too