Staying in Control

make a mrking

???A what??? That has to be a typo, but what's left out? "a"? (marking)

haha, marking is!

thanks mention. I hope to make a marking so that can read new threads on time.

its an old topic. just wait for some new viewpoint to comment on.

While central control for a plant sounds attractive, twenty years of experience in industrial controls has taught me that small independently operable pieces of equipment provide the best process reliability. It is much easier to troubleshoot problems or bypass sections when each piece of equipment has the ability to be operated autonomously. Also one must consider the regular need to replace aging equipment while keeping the plant running.

The easiest way to obtain redundancy and process reliability is to use independent units.

That being said, nothing stops us from connecting these to a distributed control system as an oversight and data collection but one should be able to operate each section independently.

Dear marcot,

They are all true what you told but they are also getting old like me, they are being "good old days"!

I have enter the process control area with batch control instruments from foxboro during the years of "77" as a young engineer. On the other hand the controls i have realised are based on CanOpen or Modbus mostly, for the last seven years. Although they have independent control loops on the bus, no one can control independently one of them if the buslines are collapsed, otherwise everything is okay naturally.

I must indicate that i have not seen a crashed control since it is formed by a bussed topology. This is why i'm relaying on the new technologies. Thanks to the bussed controls we have got rid of conventional cabling hotch-potch. This is very important...This is economy...

In general manner individual control loops working via the bus independently and the central station -that i'm using a pc for - is used for rearranging of parameters and diagnostics.

If any part of the system -that is even formed so as we realize at the good old days - is defect then the whole related systems are already stopped regardless of the system is independently formed or bussed! Old styled installations are not make additional gain to the owner of workplace, more initial cost instead.

Oldies are not goldies my dear...

If we want to speak more then we must keep our steps with technology...

BTW what do you think of the wireless instruments?

Best regards

f.h.

G'day fh,

• In general manner individual control loops working via the bus independently and the central station -that i'm using a pc for - is used for rearranging of parameters and diagnostics.

What you are describing is very much along the lines of a Totally Distributed Control network where individual independent loops are controlled locally but supervised from a central control system that monitors and has the ability to override local loops when required.

By the way, Totally Distributed Control is nothing new. Honeywell were doing in back in the 1970s with their TDC‑2000 system that has now been upgraded to the TDC‑3000 system. Basically it consisted of local control loops that operated independently but are supervised from a central computer that could modify, override, reprogram and report on the local stand alone loops.

You can also do very much the same using Programmable Logic Controllers and a centralized supervising system.

As for using a PC running Windows as the central computer I'm not that sure it's the way to go. You can get industrial PCs that are more reliable then the average off the shelf stuff, but Windows, while being popular, is nothing like stable or reliable enough to use in an industrial environment. You certainly wouldn't want it to decide to update itself without telling you and then you have the dreaded blue screen of death to contend with. You can get what is referred to and embedded version of windows which is a cut down version that is permanently loaded into non volatile memory but for me it's just not stable enough.

• BTW what do you think of the wireless instruments?

I wouldn't touch them with a 10 foot barge pole.

While there are claims of safe guards and reliability you are depending on a radio link that is subject to and dependant on factors that are out of your control. It only takes one idiot with something that can broadcast enough broadband RF noise to knock it over.

Another thing that you need to think about is malicious intent where somebody is deliberately trying to sabotage your system. Having radio links all over the place just makes it easier for them to play havoc with your system.

Regards,

masu

Hello masu,

Many thanks for your reply.

I'm already assume that many guys aware of Distributed Control Systems, in this community.

The purpose of the long lines you've pasted is not to teach about DCS to anyone, but all individual controls can work without a central unit and so, not have a logical difference with the installations what we have implemented in old days.

Seems that it would be better if i have made it longer the long lines , adding above mentioned lines.

The central station - a pc or industrial pc or touch scrn pc or plc or whatever will be- and the operating sytem that will be used are all customer's preference, i can only give options those suit the application.

And..about wireless..

I wouldn't touch them with a 10 foot barge pole.

I've also not enough courage to use wless yet. I've not met a person which has a positive reaction, i have talked about wless with. Seemingly, it is not mature and we will wait for a while. On the other hand i wonder, how we know whether wless is mature or not, without implementing?

Best regards,

G'day fh,

One of the things that concerns me with wireless or radio controls is congestion. Basically you are using a serial bus that is open to every Tom, Dick or Harry within radio range. Like any bus the response time can start to blow out if the traffic on it gets too dense and since you don't have control over who, what, where or when people are using the frequency you can never guarantee the response time.

The other thing that concerns me is the possibility of deliberate malicious tampering. Now before somebody pipes up with inscription and a whole hose of other techniques to ensure people cant get at your link yes this is true. However, given enough time and a powerful enough computer you can break almost any code or encryption system. The way many encryption systems work is to use a code that given time could be broken cant be in real time as there aren't any computers capable of doing so.

The only problem with this is Moor's law which in a basic form says that every 18 months or so the power of computers doubles. If you keep doubling the capacity of computers it doesn't take long before what was impossible to solve in real time to become decipherable and there goes your security.

Now this is just my personal view and impressions and as I am not actively involved in control systems any more I could be wrong so if anybody disagrees then I would apreciate your opinion.

Thank you my friend,

As I know the brands those applying wless controls are relaying on the frequency hopping method mostly for the security, i haven't studied much about the new hardwares though. Again as i know the accidental congestions are being solved via this method (true?).

It may be possible even tracking of this f-hops with a powerful computer and a powerful techno-programmer together..??.. In fact the point i can't understand is why anyone want to make thislike action in an industrial environment? This is a different matter i know but i can't hold it to ask.

Supplying security will be even impossible if all possibilities must be counted and maturing of technologies are focused on security subject only, in a growing manner!

We are likely fighting ourselves, this looks me very interesting... Might it be a improvement method?

Regards

Yes, busses are becoming more reliable but they slow down dramatically when loaded. In some process the real time is not so real anymore.

The main problem when having everything going through one or two serial busses is the vulnerability to accidents and the difficulty of removing a specific equipment from the process for maintenance activity. I have worked on some "modern plants" where they will kick you out of the site if you do anything that crashes their network. It does happen and doesn't need a big mistake: Disconnect the wrong cable, update a PLC program with the wrong network address (the default value used by something else) or turn off the control power at a node. All off these may cause a plant crash. IT DOES HAPPEN. And it is very costly...

Dear friend,

I agree fully as an engineer but as a vendor.

I have seen Instruments installed 12 years ago and still working properly under the harsh environment of a boiler plant. I don't think that a networked system can work as much duration as that of i mentioned above, without a crash. Nothing of the DCS systems can be durable so as the conventional system is.

On the other hand, while the count of individual loops are getting more than 3, the networked systems start to be cheaper and confortable! There is nothing to do!

Commercial thinking way is very different from that of technical people and this has no worth if money is on sciene! None of the customers think of a crash 3 months after the time installation is done. Although i have not seen a crashed network except that of caused by energy interruptions, even this is a negative point for DCS. There is an idiom in turkish, i have to apply, that means " best of bads ".

Ah, what we will do is looking for the "best of bads"! This yields us the point what we are discuss. It must not be very hard to say "not so bad" about the ways technology leads us, because DCS is said to be bad with respect to the conventional systems only, and the conventionals are being matter of repairmans day by day.

This is my story my dear, i agree with you though.

stay healthy

f.h.