Where to start with a PLC controlled brewery?

Homebrew Talk

Help Support Homebrew Talk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

JKoravos

Well-Known Member
Joined
Jul 24, 2009
Messages
1,206
Reaction score
80
Location
Chelmsford, MA
I'm mainly thinking of the control unit here. I have some experience using and programming PLCs, but I don't have a good handle on what's available out there (for cheap ;)), especially when analog control is involved.

The brewery I'm designing in my head would probably have 15-20 temperature/pressure sensors, which would control about as many valves, pumps, motors, etc.

Data logging is also a plus.


Anyone have an idea where to begin looking?
 
I've seen several on eBay. I have almost no experience programming them but I have several Inst. Eng. and Elect. Eng. friends. My understanding is that you need one point for input and one point for output for each device, so I think that means you would need 30 - 40 points. I really don't know if that's accurate but when I was looking, I told my friends what I was doing: pump, 2 control valves, 2 temp. controllers. They told me to get one with 12 - 16 points in case you want to add on. I haven't done it but I have seen PLCs on eBay that would handle that (12 - 16 points) for $125 - $250. Hope this helps - Dwain
 
Yeah, I did a quick troll through the ebay listings. The problems is that there are so damn many up there. I have a reasonably good idea what I need and I could spec it myself, but I'm always looking for advice from people who know more than me.

I guess I'm looking for recommendations on hardware people have used and know will work will for this application.
 
A question is how cheap and how much software one has to write to make it function. The Opto 22 hardware is fairly common on Ebay and with careful research and incremental purchase you could add points as you go. Belimo proportional control ball valves are another item that shows up regularly on ebay these days. For flow meters the Malema line are also found on ebay for a reasonable price in the 1 gpm flow range.
I have put together a PLC controlled system based on the old opto 22 G1 series hardware and monitor temperatures, flows, pressures, to control liquid levels, liquid flows, gas flows to burners, pumps, conveyors, malt mill, exhaust system. With the current configuration there are about 30 analog points and 32 digital points in use. The easy part was the construction of the hardware, for me the difficult part has been learing Java and writing the control software and functioning sequence control program. The software is an ongoing effort with learning how to incorporate SQL database functions beyond point configuration tables to develop and implement recipe formulation like other recipe formulation programs. The current goal is to be able to formulate recipe, pour ingredients into hopper, attach fermenter to CFC connection, press auto run button on touch screen and walk away to do other brewing activities while system takes care of brewing process.
After 15 years of brewing there is not many brewing challenges left to try any more so the automation was the last challenge to try.
 
A question is how cheap and how much software one has to write to make it function. The Opto 22 hardware is fairly common on Ebay and with careful research and incremental purchase you could add points as you go. Belimo proportional control ball valves are another item that shows up regularly on ebay these days. For flow meters the Malema line are also found on ebay for a reasonable price in the 1 gpm flow range.
I have put together a PLC controlled system based on the old opto 22 G1 series hardware and monitor temperatures, flows, pressures, to control liquid levels, liquid flows, gas flows to burners, pumps, conveyors, malt mill, exhaust system. With the current configuration there are about 30 analog points and 32 digital points in use. The easy part was the construction of the hardware, for me the difficult part has been learing Java and writing the control software and functioning sequence control program. The software is an ongoing effort with learning how to incorporate SQL database functions beyond point configuration tables to develop and implement recipe formulation like other recipe formulation programs. The current goal is to be able to formulate recipe, pour ingredients into hopper, attach fermenter to CFC connection, press auto run button on touch screen and walk away to do other brewing activities while system takes care of brewing process.
After 15 years of brewing there is not many brewing challenges left to try any more so the automation was the last challenge to try.


Someone on another forum linked me to your setup. It's impressive, to say the least. It's certainly in the realm of what I aspire to create for my brewery. I love process design, control, automation, etc. I've been dreaming about doing this for almost as long as I've been brewing. It's a side effect of having a degree in chemical engineering, I'm sure. I finally have the resources to start the project...so here I go...:D
 
If all goes well and the tech gets the touch screen pc ready for me next week, I hope to tune PID loops and do a test run next weekend when I am home on a R&R run from Arizona. Everthing works correctly during simulations so I hope all that is needed is final tuning and malt for a test run. If this works as planned then I will be looking for another challenge to design and build to keep busy during spare time in Arizona.
 
I am a system integrator and the company I work for manufacturers their own IO modules and PLCs. The problem is market price is something like 2k retail for the PLC and IO module. comes with 40 digital inputs, 40 digital outputs, 40 analog inputs and 8 analog outputs.
You have to add your own isolation and relays though. The good thing is the programming software is free, which is where every one else wants to drill you.

The big thing is digital is cheap, you pay out the nose for analog, specially outputs. Programming software is also the other industry 'got yah'. I would look on ebay for any kind of IO brick that supports MODBUS and then try to find a cheap PLC with software that supports modbus as well. This way you can build your system on an industrial standard.

Allen Bradly has a good portion of the market though and uses their own proprietary com protocol, their software is mega bucks too. but old equipment is plentiful. If you buy a copy of RsLogix off of ebay, AB WILL come after you so be careful.

good luck man and let us know how it works out. Kladue has a VAST knowledge of process controls equipment and instrumentation. Mine is limited to the water industry, hoping to change that soon... However, if you have any questions I might be able to help.
 
I guess the next question is what do want the brewing system to do?, do you have an existing system to r&d the control approaches. I put together a prototype SRIMS system 6 years ago and refined the operation and design before begining the current system hardware and automation design and build. Do you have an existing system or one to use for a design example that you would automate all the manual functions?.
You could look for the current generation of Opto hardware and use their software packages to build the control program, or try and build your own from scratch in one of the various programming languages available. Even with purchasing the control hardware and pumps on Ebay this system cost has gone well past 2K and is still going as next project design is underway, automated measuring of malts from user recipe input on touch screen.
 
I have a system but it's not suitable for prototyping. I'm going to be building the new system from the ground up. It's going to be an all electric, steam injection, continuously stirred mash system. Fortunately, I've already built up a lot of the hardware for that system which should keep the brewing related hardware costs down. Although, I fully expect to spend several thousand dollars by the time everything is said and done.

At least in the first rev, I'm not going to be going as far as automated malt weighing (though I could see that as a future addition). I'd weigh out the grains, load them in the hopper and from there I'd like it to be virtually touch free to pitching, possibly minus hop/whirlfloc/etc. additions. Although, it's not that tough to imagine a system that automatically adds boil additions on a timer.

Also, I remembered that I have some hardware and software that I've been carrying around with me for probably 6-7 years (I jumped the gun a bit on that purchase!). It's an old ISA DAQ card with 8 diff. analog inputs and 40 digital I/O. My original plan (circa 2002) was to use a data acquisition and control software package to run the system via PC. I also have a copy of LabTech Control for windows that I can use as an interface. I haven't tried to load up the software since I got it, I think it's got a pass key, which I don't have. I got it from a company I used to work for that was going out of business.

I still maybe be able to use that hardware/software combo for this, I just hadn't even thought about that stuff for years.
 
My approach to the malt weigh system will be to use differing diameter tubes that hold a known quantity and dispense once per rotation. A simple gear motor and switch for revolution counting should suffice. The weighing method involves load cells, amplifier, and output scaling hardware that is not cheap or simple to setup and calibrate. In my career I have installed and worked with various weighing systems and opted to keep the measuring system design simple and digital. Now that the new system is ready I have realized that the first control panel has to move to make room for the malt measuring system so the malt can be directed into the hopper. The alternative is to build another hopper and use a flexible conveyor to connect the equipment.
 
How many grain varieties can your system dispense?

Also, how's the recipe integration working out? Seems like a fairly time consuming programming project.
 
The current design is 4-20Lb hoppers and 4-2 Lb hoppers for the prototype unit. That would keep initial point count at 8 DI-8 DO which will fit on a 16 position opto 22 G1 series IO board. As to the programming, I am currently learning the nuances of putting together a MySQL relational database app for the recipe portion of the program.
 
At the time I started aquiring hardware the Opto 22 G1 hardware was cheap and plentiful on ebay. The advantage of the modular system is the ability to swap modules, and the support for industry standard sensors and control devices. The down side is the B1-B2 brain boards only save the configuration information and are not smart enough for much else so all the control has to happen elsewhere. The hardware communications is by RS 422 @38.4K which was not much of a problem as I found a smart ethernet interface unit that was configurable for TCP/IP to interface with the laptop, and now the touch screen panel pc. I have taken the initial programming that my son put together and have been learing Java and SQL as I go to build the graphics and develop the sequence control program for the system. All of the system components have been run in a manual mode to test operation and performance and some PID loop tuning was done but tables have been flushed during software construction and will have to be rebuilt during loop testing.
 
I admire the effort and progress made with the brewtroller system but I am afraid that the platform is not suited for analog control methods. It works well with the limited inputs and simple control strategies for traditional 3 vessel brewing systems, but when you need to use industry standard analog control hardware, you need a more robust hardware platform to start from.
 
What's your dosing mechanism?
If every thing goes as planned my final design should have 6 mini silos with a 50 kg capacity each.
Hopefully the silos are first grain in first out designs.
The PLC program selects the required silos based on the beer style I am brewing, one by one.
Each silo has a dump valve and is connected through a feeding tube to the weighing hopper.
Similar to this setup
100_5574.jpg


From the weighing hopper the grains go to the mill and then to the mash tun.

Cheers,
ClaudiusB
 
Looks like you will be finished long before I will, construction project supervision commitment through mid 2010 in Arizona, with 3-4 days a month at home it sure puts a crimp on brewery hardware construction.
 
Looks like you will be finished long before I will, construction project supervision commitment through mid 2010 in Arizona, with 3-4 days a month at home it sure puts a crimp on brewery hardware construction.

All my projects are on hold.
Both of us will cross the finish line at the same time.
I know the sh** you are going through, used to travel a lot.

I was an automotive guy until last month.:mug:


Cheers,
ClaudiusB
 
If every thing goes as planned my final design should have 6 mini silos with a 50 kg capacity each.
Hopefully the silos are first grain in first out designs.
The PLC program selects the required silos based on the beer style I am brewing, one by one.
Each silo has a dump valve and is connected through a feeding tube to the weighing hopper.
Similar to this setup
100_5574.jpg


From the weighing hopper the grains go to the mill and then to the mash tun.

Cheers,
ClaudiusB


Impressive looking setup, CB. Do you do any research on the hopper design? Solids flow in conical hoppers can be a tricky thing. If you don't get the hopper angle right you'll get dead spots along the walls.
 
Impressive looking setup, CB.
Thanks JKoravos
Do you do any research on the hopper design?

Currently I am reading:
MECHANICAL PROPERTIES OF GRANULAR AGRICULTURAL
MATERIALS CONSIDERED IN SILOS DESIGN BY
Manuel Moya, Francisco Ayuga, Manuel Guaita,and Pedro Jose Aguado


Solids flow in conical hoppers can be a tricky thing. If you don't get the hopper angle right you'll get dead spots along the walls.
I am trying to understand these fancy terms, angle of internal friction, grain to wall friction coefficient, dilatancy angle, Poisson’s ratio, specific weight, modulus of elasticity.
If not, I have to get the answers trough experimentation like a third grader:D

Cheers,
ClaudiusB
 
Thanks JKoravos


Currently I am reading:
MECHANICAL PROPERTIES OF GRANULAR AGRICULTURAL
MATERIALS CONSIDERED IN SILOS DESIGN BY
Manuel Moya, Francisco Ayuga, Manuel Guaita,and Pedro Jose Aguado



I am trying to understand these fancy terms, angle of internal friction, grain to wall friction coefficient, dilatancy angle, Poisson’s ratio, specific weight, modulus of elasticity.
If not, I have to get the answers trough experimentation like a third grader:D

Cheers,
ClaudiusB


Here's a good place to start exploring...if you haven't already:

Powder and Bulk
 
I mapped out my preliminary control structure. I identified 9 analog inputs, 3 digital inputs and 22 digital outputs. Plus a few more digi outs that I'm still up in the air about.

If I can get the hardware I have working it should be able to handle the load.
 
I guess the next question is what do want the brewing system to do?, do you have an existing system to r&d the control approaches. I put together a prototype SRIMS system 6 years ago and refined the operation and design before begining the current system hardware and automation design and build. Do you have an existing system or one to use for a design example that you would automate all the manual functions?.
You could look for the current generation of Opto hardware and use their software packages to build the control program, or try and build your own from scratch in one of the various programming languages available. Even with purchasing the control hardware and pumps on Ebay this system cost has gone well past 2K and is still going as next project design is underway, automated measuring of malts from user recipe input on touch screen.

I'm new to home brew talk, but saw this post and thought an attached photo of my HERMs / PLC system might provide a starting point.
 
I'm new to home brew talk, but saw this post and thought an attached photo of my HERMs / PLC system might provide a starting point.

DSCN3313.jpg


DSCN3391.jpg
 
Nice looking panel build, looks like strictly on/off operation though. The challenge for PLC builds comes when you go to analog process control. I went quite a bit farther with software to implement a system that handled the brewing process from recipe to process control, you could pour grain in one end and get chilled wort in the fermenter at other end without touching the process. With a brewing system that was more complex than necessary to just make beer, it was not practical for sales to the hobbyist, and it was a great challenge to build and program.
 
Nice looking panel build, looks like strictly on/off operation though. The challenge for PLC builds comes when you go to analog process control. I went quite a bit farther with software to implement a system that handled the brewing process from recipe to process control, you could pour grain in one end and get chilled wort in the fermenter at other end without touching the process. With a brewing system that was more complex than necessary to just make beer, it was not practical for sales to the hobbyist, and it was a great challenge to build and program.

On/Off is the primary control Process, for temperature maintenance. But, it also.
1. calculate mash water volume based on grain weight & Grain to water ratio.
2. calculate strike temperature based on ambient temperature and the grain and water factors.
3. keeps time and sets temperature for each stage; Mash, Sparge & Boil.
4. starts boil heat up half way through the Sparge. Full boil usually starts at the end of the Sparge.
5. sounds a signal for each boil addition like hops, Irish moss, yeast nutrition or other ingredient at the proper time.
6. Each brew requires grain weight & temp, mash temp & time, sparge temp & time, boil temp & time and each boil addition item & time.
7. The boil is not continuous heat; but a pulse heat starting at 80% and adjusting up or down to keep the boil rolling.
8. complex mashing is also available, but have only did a 1 step 130F rest once and didn't seem to see any advantage or improvement in the beer.
9. all temperatures for the HLT, MLT, BK and HERMS exit are displayed. Still have my old kettle thermometers in my HLT & MLT but very seldom use them

I did have my pumps controlled by the PLC at first, but found manual was better. Things are never to the second when changing from mash to sparge to boil.
 
Not bad with software so far, do you measure liquid level in various kettles, or just manually control valves with a visual measurement. The fun part with the software is making all the process variables accessible during run time for calibration. It took me about 2,000 hours of programming and testing to get it right for my system.
When you get to automated valves look at Belimo SS valves with 24V actuators. Not unreasonably expensive if you spend time looking for bargains. With both binary and analog control actuators you can do many interesting things.
 
Not bad with software so far, do you measure liquid level in various kettles, or just manually control valves with a visual measurement. The fun part with the software is making all the process variables accessible during run time for calibration. It took me about 2,000 hours of programming and testing to get it right for my system.
When you get to automated valves look at Belimo SS valves with 24V actuators. Not unreasonably expensive if you spend time looking for bargains. With both binary and analog control actuators you can do many interesting things.

Visual on liquid level using a graduated sight gauge on my HLT.

My system is a 24 VDC system. I did considered automating 6 to 8 valves and finding an accurate way to measure volume with flow rates, but the project became too expensive for me, a retired computer engineer.

In the winter of 2015/16 I converted my old 3 half barrel gas system to an automated HERMS electric system for $150 in plumbing and $650 for a PLC, pt100's, relays, case, 5500 watt elements and other components. I did spend some time trying to see if I could control the temperature using gas, but the cost and additional complexity required me to convert to electric.

I'm satisfied with my increased inefficiency (80-90%) and performance. Also I'm less likely to forget an additive or miss a time point. Constant viewing of temperature and time increases my level of confidence in brewing.

If I look at changes to my original design, I would use 1/2" ID Hi temp food grade hose instead of 1/2" OD stainless steel tubing. Some savings in cost, but much easier to plumb, and the higher flow volume would reduce heat-up times like rest to mash to sparge. I would also look at a different PLC. My Easy ELC-26DC-DA-R (Siemens) works good for analog and most digital; However, on math you can not have a math result over + or - 32,767. This means that many calculations must be preformed in 2 or 3 step calculations. The results are correct, but the programming is annoying. I would also like a larger display for information, a 16 x 4 line is too small.
 
I built a system around the Opto 22 generation 1 hardware, modular design with plug in modules. The generation 1 hardware was cheap and modules easy to acquire from Ebay at the time of construction, 2008. Upside was the availability of input modules to handle 4-20MA, 0-5V, PT100 and thermocouple sensors, downside was the modules and brain boards were limited to IO only. All the calculation and control had to be done in the software application, then sent to boards to make things happen.
With wet level transmitters, flow meters, proportional valves, and mass flow controllers for gas flow to burners, the software became fairly complex quickly. With a flash boiler for water heating/steam generation the system can reach water temp targets in about 45 seconds from a cold start, and raise wort temperatures with direct steam injection for step mashing by maintaining wort target temperatures after steam/wort mixer. Boiler control requires both gas and water flow control to maintain temps, and variable swapping for PID loop control to fit various operating scenarios.
With a background in process control installation, design was based on previously built manually controlled system, and software programming was self taught. Programming started with Java and MySQL, then ended in C# with MySQL application built for touch screen control.
 
Back
Top