Looking to upgrade mill wide systems looking for opinions

[sparkytex]

hello everyone,

Our mill Is looking at the possibility of upgrading our current automation system. We currently have 50 PLC's through out the mill. Most of our PLC's are Allen Bradley PLC 5 processors that communicate over Ethernet IP and or DH+, or both. We have a few stand alone PLC 5's in smaller area's of the mill. We have quite a few Allen Bradley SLC 500 series processors that communicate solely through DH+ or are stand alones. We also have a few AB Compact Logix as well that are stand alone PLC's. Unfortunately we also have some very old Mitsubishi F series processors for smaller applications throughout the mill.

Currently I have no ability to communicate with the Mitsubishi processors as I do not even know the software required and how to use it. Obviously upgrading these PLC's to a newer AB PLC would be beneficial to us.

I was going to suggest two different upgrades and I would appreciate your opinions please.

1.) This suggestion is more cost effective.

*Upgrade all of the PLC 5 processors to a PLC 5/40E Ethernet processor (the ones that aren't already 5/40E)
* Upgrade all of the SLC 5/04 processors to a SLC 5/05 Ethernet processor
* Upgrade all Mitsubishi PLC's to a SLC 5/05 Ethernet processors
* Get rid of all DH+ networks and use Ethernet IP for mill wide system comms.

2.) This suggestion is a little more pricey and will require more work.

* Upgrade all PLC 5 processors to 1756 L6 Control Logix processors
* Upgrade all SLC 500 processors to 1769 L35E Compact Logix processors
* Upgrade all Mitsusbishi processors to 1769 L35E Compact Logix processors
* Get rid of all DH+ networks and use Ethernet IP for mill wide system comms.

Obviously suggestion one is a lot more of a cost effect route, but then we are sticking with older technology using the PLC 5 and SLC 500 series processors. Suggestion two would be an expensive and timely route, but would ensure we are up to date with AB's most current PLC product line. I know that RSLogix 5000 is great software and would be beneficial for our shift electricians for trouble shooting. I also feel getting rid of the DH+ is a must since it is a very old system and constantly causes problems for us impeding production on a regular basis.

Let me know your thoughts and suggestions.

Thanks, Tex

 

[Luke_Mages]

I like option 2 but using all compact logix. I haven't priced a 5/05 in a while but if its not more expensive than compact by now it will be soon.
Also, would point io help lower cost? Do you have multiple plcs currently running one line?

 

[dginbuffalo]

I think I would not use option 1. It may be cheaper and easier, but that technology is getting older and older.

On option 2- are you sure you would even need to go from PLC5 to Controllogix and SLC-500 to L35E Compactlogix? Compacts are pretty powerful today- you may be able to substitute instead of a Controllogix. And even the L35E's swapping from SLC-500 may be able to be downgraded to 5370 L24 with built in I/O or L30 processor.

Or even a 1 larger system that handles the job that a few of the older PLC's did.
I'm pretty sure there are some conversion utilites that would make the move from Logix 500 to 5000 easier also.

IMO- If i was really upgrading, it would be to the latest and greatest if possible.

 

[sparkytex]

Some of the PLC 5/40E's run massive programs for our digesters, Bleach Plant, Acid Plant, Boiler burner management systems, Evaporators and so on. Some have 6 to 10 - 16 slot RIO racks full of Analog and Discrete I/O. Would the Compact Logix be able to handle this? A lot of our Ethernet processors communicate to each other over our Ethernet network as well. I believe it would be quite difficult to utilize a few PLC's to do the job of many older ones as far as programming and convenience.

 

[Luke_Mages]

Some of the PLC 5/40E's run massive programs for our digesters, Bleach Plant, Acid Plant, Boiler burner management systems, Evaporators and so on. Some have 6 to 10 - 16 slot RIO racks full of Analog and Discrete I/O. Would the Compact Logix be able to handle this? A lot of our Ethernet processors communicate to each other over our Ethernet network as well. I believe it would be quite difficult to utilize a few PLC's to do the job of many older ones as far as programming and convenience.

do you have the ra architect builder? That would tell you definitively.
I just quoted a system that has 160di, 80do,40ai and 10ao all using a single compact logix processor.

 

[sparkytex]

do you have the ra architect builder? That would tell you definitively.
I just quoted a system that has 160di, 80do,40ai and 10ao all using a single compact logix processor.

I do not have the architect builder unfortunately. I can tell you this, out digester PLC 5/40E has 592 Discrete channels and 170 Analog channels. A few other PLC's are of this size as well.

 

[Peter Nachtwey]

I would chose option 2. Those PLC5/40E will cost an arm and a leg and are obsolete. So are the SLC5/05. I bet you can buy 2-4 compact logix for every PLC5/40E and each compact logix can do twice as much at least. In the end I bet option 2 will be cheaper hardware wise. I do agree that option 2 will require more effort to translate the old PLC code to RS5000 but I bet it will be worth it in the end.
As far as combining PLCs. I think this is more of a locality issue. RS5000 can import the old PLC code from different PLCs as different tasks so that logically they can remain different or separate.

 

[Luke_Mages]

I threw this together...600 digital and 180 analog. Its all compact logix, 4 banks I think, estimates pricing at about 40k

 

[dginbuffalo]

I threw this together...600 digital and 180 analog. Its all compact logix, 4 banks I think, estimates pricing at about 40k

I see you used IAB for that? I was just messing around trying to do the same thing

I can see where doing the full upgrade would be pretty expensive in buying all the I/O modules vs. just buying "newer" PLC5 and SLC processors. Maybe you could hodgepodge it with newer processors talking to the old RIO if that's even possible. Then as time and $$ allow, replace RIO racks with newer 1769-RIO. Just a thought.

 

[rkukl]

I too would go with option #2. Have you looked at this for you PLC 5 Chassis? I've used this on a small project and it saved lots of time.
http://literature.rockwellautomation.com/idc/groups/literature/documents/pp/migrat-pp003_-en-e.pdf

 

[ASF]

I would also be looking at option 2. It might be more expensive upfront, but it'd be massively rewarding in the long term. It'd also ensure that all your PLC's use the same programming platform. Plus, networking is a whooooole lot easier on the Logix series of processors than the SLC series.

For the larger stuff I'd be going for the Control Logix platform. Without spending a bit of time going through fine detail I couldn't say whether the Compact Logix would handle the whole lot or not - but the Control Logix is more or less infinitely expandable (well, expandable enough that it probably stops mattering). If you have so many remote I/O racks that your ethernet card runs out of connection resources, slot in another ethernet card. If your CPU fills up, slot in another CPU card. If your CPU is running too slow for some high speed process control, add in a high speed CPU. Have a safety application? Throw in a safety CPU and safety partner. If you wanted you could have a 13 slot rack with 5 CPU's and 8 ethernet cards, and distribute out to everything from there.

The only other point I'd make is that the compact/control logix parts you spec'd are not the latest any more.

Control Logix is up to the L7 series, e.g. 1756-L71.
Compact Logix 5370 controllers are the latest:
- The 1769-L1 series (e.g. 1769-L16ER-BB1B) has onboard power supply and ethernet, and has 16 digital inputs and 16 digital outputs onboard. Then it uses 1734 Point I/O expansion cards. For small setups it's extremely cost effective
- The 1769-L2 series has the same as the L1 series, but also have onboard analog in/out and high speed counter. It uses the usual 1769 series compact I/O expansion cards. Good for slightly larger applications
- The 1769-L3 series is the equivalent of the 1769-L32E/L35E. It's a standalone CPU for which you need to buy a power supply and 1769 series I/O expansion cards. This is for larger systems, right up to the point where you start eyeing off the control logix.

 

[Paullys50]

Option #1: Will be more expensive in the long run as you'll pay twice for migration costs to eventually get to a "modern" PAC you only "maintain" what you have. You won't evolve it to improve overall operations.

Option #2: Huge project, but if you do it right it will sell itself.

You need to do a "proof-of-concept" system. Take a smaller yet important system and migrate to a ControlLogix solution. You didn't mention SCADA but do that too. Dump Ignition on there and now you have data that can sell the value to upper management. Get all the stake-holders (operators/operations managers) of that system on board, plan out what you want/need to do. Implement and enhance it so the operators and operations management are praising it. Review with upper management letting the stakeholders sell it for you.

Then you can justify capital investments into that large of a project!!!

For IO, instead of rkukl's suggestion as that migration is expensive too, and clunky replace the PLC5 rack IO with PointIO. Chances are you can create the same IO structure on PointIO and have it fit in the same footprint of the PLC5 IO. I did this a few years ago and worked extremely well. I will recommend that over the migration solution any day. Especially if your control panels aren't deep enough, their solution simply isn't and option.

 

[sparkytex]

Thank you all for your opinions and consideration. I will be taking all I have learned from this and writing the pros and cons of each suggested route. It will then be up to management to make a decision.

Just FYI - I personally would 100% go with route 2 as its a true upgrade and the PLC 5's are becoming increasingly outdated.

Thanks, TEX

 

[danny.michael]

The PLC5's are more than "increasingly outdated". The 5/40E was discontinued in December 2014. The only PLC5's that are still available from Allen Bradley are the 5/40C (ControlNet 1.25 at that) and the PLC5/80's.

 

[rdrast]

We've done several of that type of upgrade here, and once again, the only real answer is option 2. The question really boils down to, do you want to try to keep the PLC5 / SLC I/O.

Again, we've done it both ways, and without a doubt, I'd opt to replace everything, racks and I/O included. It works to make the 5 and 500 racks remote I/O racks, but it ends up much cleaner in the long run to convert over.

In many cases, the 5 and 500 existing racks can be redone using Point I/O very cost effectively.

If you currently have very distributed control (one line with multiple PLC CPU's), you should consider consolidating all control into one large ControlLogix CPU, and converting the smaller nodes to pure remote I/O.