Hello everyone,
I work as Instrument Process Control Technician and I have and will always be admiring the products of Arduino. They are east to use and very simple to learn on the other hand. I have recently read an article here about a guy who wanted to replace Allen Bradley PLCs with punch Arduinos. Well that is possible, but to be honest, if we were to upgrade a system, it must be better and also it has can provide the most of the functions the older system had, if we compare an Arduino with an early model of an Allen Bradley PLC we would see the big difference and capabilities an industrial PLC like an Allen Bradley SLC 500 would have over Arduino boards. Anyway, let’s talk about my idea to make Arduino be used in Non-Critical Industrial Automation.
INTRODUCTION
Well, in our power plant we have General Electric GE Turbines coming in one package with their Mark VI control system. For any engine that sucks air to make good mixture of fuel and oxygen for efficient combustion, atmospheric air is not really clean all the time. So we use static air filters to prevent any solid particles and dust from entering the engine. A 200 MW turbine power generator will have filters aligned in the compressor suction side, allowing clean air to the engine at all times. A simpler example is your car’s intake filter for example.
So this power turbine drives a huge generator and we are relying on that to power the rest of the plant facilities running 24/7. Filters in the suction side of the turbine filter air intake, let’s say we had dusty weather for a long period of time, what could possibly happen? Dust will accumulate and cover all the filters. If filters are clogged or blocked, we will get a relatively low differential pressure before the filter and after the filter, causing the responsible Shutdown differential pressure switches to read low and send a trip signal to the control system and then the control system will safely trip and stop the power turbine preventing a possible disaster ( think of it like a vacuum cleaner and you accidentally blocked the suction side for a while it is running) (we are talking about huge thrust power delivered from a shaft weighing about 15 Tons with like 14 stages of air compression). To fix this problem, someone would say we can trip the turbine several times to manually clean the filters. Another might say we can replace filters for more reliability. I would say we can’t trip the turbine, we don’t want a blackout. And we won’t waste money on buying filters every 2 months. The real fix to this problem is down below.
About APU
Well GE has installed a simple system for that matter called an APU, Air Pulsation Unit. The only purpose of the system is to continuously clean air filters from dust accumulation 24/7. It simply uses compressed air (100 psi) on the opposite hallow side of each filter controlled by a solenoid air valve. It’s only a timer circuit with 32 AC power outputs to the solenoids controlling the cleaning sequence. The system is composed of:
- 3 timing cards with fuzzy logic shift registers and de-multiplexers each card having 32 AC power outputs to air solenoids.
- Timing relay.
- Card switching standalone relays to select the card active at a time.
The real function is just to send the signal to solenoids at a sequence one every time. If card one finished its job, the last signal from card one is sent to the relay, relay switches contacts and gives the order to the second cards to start. If the second card finishes its last signal is sent to the second relay allowing the last third card to start. If the last card finishes it sends its last output again to the first relay and the first card starts again and the sequence is repeated. That’s its only function. Then after 8 years of startup of the power plant, we had issues in that system and we discovered that cards are the fault. Ok let’s install new ones (one costs about 700 USD) and we need 9 cards for 3 turbine generators (that’s 2100 USD). Simple function, but for such complex system that costs a lot.
The Idea”
So I figured why not use an Arduino for that system. I thought we could use Arduino for that because it costs less and it can do even more complex functions which makes it the best for the job. I’ve fixed all the things needed for the whole system upgrade it definitely won’t cost more than 400 USD to replace the 9 cards with Arduino, but still I haven’t confront my Supervisor and Forman about it. I must be sure and have answers for all their questions.
That was it, please share your points of view about this idea and feel free in your replays. It is an open subject and if you even think that u need the schematics for the system, I would love to give for free. If it wasn’t used in my case, I would like to see it work anywhere, who knows.
Thank you for your time. 
Welcome to the forum
There have been many times I wish to introduce new and innovative ideas to my industrial control systems like you, but there are times that either presenting the idea of an alternative system to other powers that be who have investment in such projects or for the simple reason of your own liability by way of introducing non-conforming, non-complaint or non-standardised systems into an already complex and expensive investment would surely be inviting disaster for oneself.
I am embarking on yet another project using Allen Bradley CompactLogix PLC with FactoryTalk SCADA and yet I could offer the same functionality using suitable ruggedised Arduino hardware into standard industrial I/O modules, together with a functional SCADA (my own developments) system for a lot lest cost. In this case, the control system is non-critical, so I'd have little problem with doing this, but would not even consider it if were in any way a critical control system, where other systems depend on it or the cost of down time or failures is not an acceptable option.
I might suggest such ideas as you present might be better played at the workbench at home where a simple failure allows us to sleep at night, and keeping the status-quo with existing systems.
But on the other hand, if you are feeling confident, then mock up a demo system using an Arduino Mega and an industrial grade relay card on DIN rail and program it to do the required functionality.
Document its complete functionality and how it will cope with faults and error conditions.
Maybe introduce the concept of future enhancement or other features like you mention, but keep it simple at the first instance.
Take the demo and present it to the powers that be, together with a bunch of fresh home made muffins and good quality coffee from the north western slopes of Kilimanjaro, Africa and you may have them in awe.
Until you do this, you may never know where such things might lead
Paul
Interesting. Is the engine shut down for filter cleaning? Where does the dirt go when it is blown off ? I would guess that the specification of the "box" would demand high cost design and verification requirements. I would be asking for the minimum design standard the customer wants before going too far. Technically I'm sure that an Arduino could do it but satisfying and demonstrating to the customer is likely to be far more expensive than then simply buying replacements of the original board. Best of luck though. I wish you well
rockwallaby:
Welcome to the forum
There have been many times I wish to introduce new and innovative ideas to my industrial control systems like you, but there are times that either presenting the idea of an alternative system to other powers that be who have investment in such projects or for the simple reason of your own liability by way of introducing non-conforming, non-complaint or non-standardised systems into an already complex and expensive investment would surely be inviting disaster for oneself.
I am embarking on yet another project using Allen Bradley CompactLogix PLC with FactoryTalk SCADA and yet I could offer the same functionality using suitable ruggedised Arduino hardware into standard industrial I/O modules, together with a functional SCADA (my own developments) system for a lot lest cost. In this case, the control system is non-critical, so I'd have little problem with doing this, but would not even consider it if were in any way a critical control system, where other systems depend on it or the cost of down time or failures is not an acceptable option.
I might suggest such ideas as you present might be better played at the workbench at home where a simple failure allows us to sleep at night, and keeping the status-quo with existing systems.
But on the other hand, if you are feeling confident, then mock up a demo system using an Arduino Mega and an industrial grade relay card on DIN rail and program it to do the required functionality.
Document its complete functionality and how it will cope with faults and error conditions.
Maybe introduce the concept of future enhancement or other features like you mention, but keep it simple at the first instance.
Take the demo and present it to the powers that be, together with a bunch of fresh home made muffins and good quality coffee from the north western slopes of Kilimanjaro, Africa and you may have them in awe.
Until you do this, you may never know where such things might lead
Paul
Thank you for your reply, really appreciate your feedback about the subject. I'm really into innovative ideas and making things better for everyone, yet simple and inexpensive. As I mentioned earlier in the subject, the operation of the APU is not that critical and if faults occurred in the system it won't cause, disrupt or lead to disaster. And to be honest, I don't really think we would replace anything in critical industries without any certifications, but of course we can play little under that at our local non-critical operations or even at home for personal needs. well, I have thought a lot about replacing expensive devices with just simple and easy boards like an arduino, but the thing is I happen to think about all the faults and errors that would face that device and ask myself, what would an arduino do in that case? and I relate to the existing device and finally say " an arduino won't be the perfect replacement for that". I would love to try and make myself standalone systems at my home for trial like a small layered control loop with SCADA as you proposed and see the limitations of such systems using arduino products. It would really help me in my real job, like in troubleshooting and diagnostics.
So, thank you again rockwallaby for your kind attention to the subject. I will be providing more info about the real circuitry of the original APU and the replacement very soon. my next project is going to be about Control Valve Positioners using arduino I have the idea and i will talk about it right after APU.
Best regards..
Gerihatrick:
Interesting. Is the engine shut down for filter cleaning? Where does the dirt go when it is blown off ? I would guess that the specification of the "box" would demand high cost design and verification requirements. I would be asking for the minimum design standard the customer wants before going too far. Technically I'm sure that an Arduino could do it but satisfying and demonstrating to the customer is likely to be far more expensive than then simply buying replacements of the original board. Best of luck though. I wish you well
Thank you Gerihatrick for your reply, to answer your questions:
Is the engine shut down for filter cleaning?
No, the filter cleaning system is separated from the control system of the turbine, but it can cause shutdown in only one scenario: if the filter cleaning system has been shut off for a long time and whether was dusty all the time, there are 3 main differential pressure switches connected ( Low Set point ) the low side of each one is connected to the intake, first switch activates at 6 "H2o and it sends a logic signal "Low" to the control system alarming that either there is no filter cleaning for a long period of time or the filters have already worn out and need replacement. the other two activate at 8 "H2o and one is enough to shutdown the turbine if activated. This will happen only if operators were too lazy to follow instructions and daily checklists for major equipment.
Where does the dirt go when it is blown off ?
There is a helical conveyor right underneath the filters connected to an AC motor that can be operated manually. all dust and particles are collected inside a dust collector tank on the ground for ease of access to be cleaned if it got filled up with dust over time.
Thanks again for your kind reply,
Regards..