Lookin for an Innovative Idea???

[BillyB]

Hi guys,
This is a really broad Question.I work in a plant here with over 60 different PLC's across different systems(mainly SLC's)and non Networked.The DCS is the main Control. We maintain our software on RSMACC.Modifications are made to these systems on aregular basis. Apart from Networking the PLC's so as to allow Uploads from a central point, what sort of innovative idea can you suggest. Just something general. Maybe a money saver, maybe something to just make life easier. Theres money to spend. I'm just wondering what the gradual progression might be for a plant like this. They are looking for some Innovative ideas and I'm stumped. It dosn't have to cost anything either (new software/ Equipment etc). Just something that adds value in its implementation.

regards
BilyB
P.S Not looking to upgrade to Logix

 

[PeterW]

If your not looking to network the areas, then what we used to do before the ability to program over ethernet became more acceptable, was to selecet an area whithin each area where a number of programming lead extensions could be run.

For example we used to have 8 PLC's accessable from the batching platform, these were the only ones required when fault finding on that batching system, other areas could be in switchrooms, where all the PLC's within the area were again accessable at one point.

 

[JesperMP]

There are two things that I would contemplate for a large plant.

1. Networking from everything in the plant to a higher level HMI/SCADA/DCS.
And the network itself should be potent and scalable which more or less implies ethernet.
In you case it would mean that you should exchange the SLCs for 5/05s.
The benefits would be that you can operate everything from everywhere (in principle), and you can log data from everything, and you can quickly get online with a machine in order to troubleshoot, reducing downtime.

2. Networking from the PLCs to lower level i/o.
In stead of pulling all i/o back and forth to centrally located control panels, you split the CPU from its i/o.
By placing the i/o locally you can dramatically cut down on complexity. This again reduces initial and long time costs.
I am using Siemens ET200S on Profibus.
In your case I would suggest AB Flex i/o on DeviceNet.

You are in luck that you work at a place where the management is so forward-looking.

 

[BillyB]

not neccessarilly looking for a networking solution

Thanks guys,
Just a little more info. Its a pharma plant so we have everthing from utilities(RO, Clean Steam) to Autoclaves to UF Skids etc(just talking PLC control!). All these systems are located in their respective rooms in their respective panels. Most are tied in to the DCS via KF2 boxes(which are under utilised!). So any information required can be transferred to the DCS and interrogated there. I suppose I am not neccessarilly looking for a networking suggestion.
I guess, what would be the ideal plant setup if the sky was the limit.

regards and thanks
BillyB

 

[JesperMP]

KF2 boxes !

arrrrggghhh !!!

You must have a very very low update speed if you say they are "under utilised".
And then they are prohibitively expensive.

Anyway, I think that the "ideal plant" does have some kind of Networking.

 

[PeterW]

I guess, what would be the ideal plant setup if the sky was the limit.

Wireless Ethernet

 

[SNK]

I work in a Pharma-Plant as well. Most systems are independent of one-and another, but the Massive CIP/SIP systems are all Control-logix platforms with ControlNet as the main bus. We don't utilize Ethernet for plant control, just tying information back to the mainframe.
You have to REALLY watch what you log in Pharma business. CFR21-Part 11 is really critical. If you plan to be FDA regulated, you had better know your stuff before automating your facility. Your situation is actually quite common in the Pharmaceutical Industry, just for that reason.
Product Demands change, and lines change also, that is why generally the lines are built non-networkable.
Hope this helps..
BTW, our plant has over 200 PLC's, about 30% are networked.

 

[JesperMP]

Product Demands change, and lines change also, that is why generally the lines are built non-networkable.

I dont understand the above at all. To my opinion this is a contradiction. Networking increases flexibility.

I dont work in the pharma business, but it must be obvious that product tracking is a huge advantage.
If FDA approval is mandatory, then I would go for software solutions that are FDA approved.
BillyB says that he uses RSMACC which I think is a change control software from Rockwell. So maybe he is very close to actually having an uptodate system. Only the networking is arcaic.

 

[SNK]

Strictly speaking, in the Pharma business, it really depends on whether your company is Brand or Generic. With today's business, a line may only be configured for 1 product. The FDA will not allow switching between products without approvals. Products fail launches, and products have a lifecycle before they are turned over to generics. This is why most pharma companies are building "Re-Contructable" lines. This way, they can change requirements as needed.
Product Tracking is great, but as I explained earlier plants have to watch what is tracked. (FOR THE FDA's case). Every bit of information cannot be destroyed, or written over. (In a DAQ system or FIS system).
Networking increases flexibility, but is generally not favoured by most Pharma plants, due to network security and integrity of software.

 

[JesperMP]

Hi SNK.

I have heard about FDA requirements before.

You argue that "Product Demands change, and lines change also," and this is a problem with regard to FDA because if data is stored then it must be according to FDA rules. In that respect it is better not to store any data at all (it doesnt make sense, but hey its the law).
I guess that is also why BillyB is not so hot about networking.

But maybe this is more a historic problem than a current problem.
Most of the big vendors (Siemens, Rockwell Automation, others) have software products that have FDA approvals.

 

[SNK]

Yes, you are right, some products have FDA approvals.
The point that I was trying to make was that one must understand the consequences of one's actions BEFORE you make the changes. The FDA has been KNOWN to shut down facilities in order to sort out these kind of things. We have Change Control Management practices in place, which means ANY change to network, I/O, General Process improvements....need to be traced for the FDA/HPBD depts.
Before you connect all of the KF2 modules, and start messaging information to another PLC, and reporting data.......know what you are doing....

 

[DDV]

Wireless Ethernet

Wireless Ethernet…and Siemens Plc’s.

 

[PeterW]

BillyB

This ain't your place is it??

http://www.jobserve.com/WC17091C0EF2619E4.job

 

[BillyB]

Hears a thought...

Right guys. Heres one ides thats being thrown about here.Its simple but I like it.
A gentleman observed that there was no Isolator on one of the panels and that it "may" be dangerous going into the panel enclosure with live equipment and steam pipes nearby.Aswell as installing isolation, it might be nice to cut room for a DH+ adaptor port on the door of the enclosure so as to avoid ever having to open the panel for communicating with the plc.Just walk up to the panel door and plug in directly into the panel door.Simple and safe.
We could do this in everycase where there is a safety risk with piping or small enclosures.
Sorry to bring it down to a basic level, but thats the kinda idea I'm after.

regards
BillyB

 

[krk]

In the bad old days before network access to the PLC was common , we used to fit these units to the door/side of all our panels.

As well as providing some physical protection for your data port, you can also add a power socket for your laptop....
.
.