Closed loop controller I/O Redundancy

nasrollahi

Member
Join Date
Dec 2005
Location
Tehran
Posts
16
Hello everybody,



I am going to select a proper DCS system for LPG plant, According to design philosophy the redundancy for closed loop I/O is not necessary. But according to spec:

" Accuracy Requirements

The accuracy requirements for process controller analogue inputs shall be +/- 0.1 % minimum, and +/- 0.3% for analogue output. For control, analogue-to-digital and/or digital to analogue converter shall be dedicated to each input/output point, otherwise a redundant design shall be provided."



Please advice me, shall above mentioned para. be considered as main reason for redundancy? or, there is better reason for redundancy in close loop I/O’s.



What is the common practice?

Is there any standard in this regard?

Dose it depend on manufacturer standard, normally?

Sasan
 
Last edited:
I think You should not use one closed loop I/O s for any other purpose or any other closed loop.Isn't it?

You have to cosider the above while programming.Then you need not go for redundancy.Am I right? or any thing else?
 
According to IEC 61511 or ISA 84.1, we should use separate system for control and Emergency Shut Down (ESD).

The SIL level of ESD system will be 3, and this will dictate the redundancy and etc…. for ESD system.

But about control system, I need a reference that clearly cut the redundancy matter for controller’s I/O’s !

The requested availability for control system is better than 99.9 % with a MTTR of 4 hrs. (According to Philosophy).

Is it clear now?
 
Hi,

I don't get it, You've selected your I/O for your ESD systems with probably doing a HAZOP which resulted in SIL3.
The requested Availability for the control system is above 99.9% then you should consider each I/O separately and define if its a critical I/O. After this job you have defined how many and which I/O should be redundant.
I hope I'm on the right track, I didn't completely understand your problem.

Rudi
 
Hi rudi,
You got the matter, Continue.
The problem is about the redundancy of control system I/O's.
Normally for monitoring we can use non-redundant I/O.
Is it OK, to use non-redundant I/O's for Closed loop I/O's as well?
Is it necessary to perform the SIL study for Control System (For Non SIS part )to define level of redundancy?
 
I read the spec entirely differently. I see no reference here to emergency shutdown or safety.

I read the spec as not allowing multiplexed inputs or multiplexed outputs.

Some data acquisition systems will use multiplexed inputs, where a single card with 8 inputs has only one A/D converter. Likewise, analog outputs could be multiplexed from a D/A, although I'm not aware of that being too common.

The spec is stating that if the selected system uses a multiplexor on either the analog input or the analog output in a control loop (as opposed to monitored point), then redundancy, is required. DCS systems offer optional analog input/output redundancy, that is, duplicate I/O hardware.

Redunancy has different meaning in different contexts. Here, I read redundancy as "duplicated analog input or analog output if either is used in a control loop, as opposed to a data point used for history, monitoring, or data logging."

Dan
 
Hi Danw,
OK, you are completely right.
But I would like to know. Should A/D or D/A conventors be considered as the main reason for redundancy?
Are the main failure cause for Control loops, A/D or D/A convertor?
I don't think so!
(Sorry for my mistakes during typing!)
 
Last edited:
>Is it OK, to use non-redundant I/O's for Closed loop I/O's as well?

The way I read it, it is OK to not use redundant I/O if the I/O used has an A/D for each input or each output. That's a vendor/ brand/model specific issue. If the I/O uses a multiplexor, then you must use redundant I/O
>Is it necessary to perform the SIL study for Control System (For Non SIS part )to define level of redundancy?

I can't answer this. My background and work is outside of SIL industries and applications.

We use redundancy in limited cases, based on experience (we use lots of redundant power supplies, no redundant I/O, a couple cases cases for redundant controllers, no redundant ethernet comm), but none are safety related control issues, it's all production related.

>But I would like to know. Should A/D or D/A conventors be considered as the main reason for redundancy?

The failure of a multiplexed input can take out multiple points at once. My data acquisition experience shows that some vendors do not do a good job at isolation when multiplexing inputs, leading to ground loop problems. Whoever wrote the spec might have the same aversion (dislike) for multiplexed inputs that I do.

To be honest, it's rare nowadays to get feedback on failed inputs. We keep I/O cards in stock and replace then when they fail. We don't bother to get a failure analysis because it doesn't pay. Our assumption is that our I/O failures are typically from close lightning strikes - lots of summer thunder storms and long wiring runs create a great antenna for inducted pickup.

> Are the main failure cause for Control loops, A/D or D/A convertor?

I don't know. I know some venfor companies track failure modes for in-warranty cases in order to comply with their ISO quality standards, but getting them to admit what the failure modes are near impossible. I've asked, but I've never gotten an answer.

We use lots of redundant power supplies, as those tend to be a weak link in keeping production up and running.

Dan
 
Hi Dan,
Thank you for your attention and explanation.
If we want to find Bottom line:
Consider that you are in my shoes, What do you do?
You select redundant or non-redundant I/O for Closed loop control?
Shall I wait for manufacturer recommendation or I have to dictate my idea?
You know that more I/O is more money.
 
Hi Contr Conn,

Thank you for your comment.

Would you please explain "embargo"! Embargo for What?

I would like to say that here (PLC.net) is not a trading company it is just a place for exchanging idea.

There is no obligation to answer any question.
 
The owner of this forum and posters may get in big trouble for any technical contacts with Iran. This forum located in US and all US laws apply.
 
Dear Contr Conn,
Lets wait for owner of the site to decide. He/She decides. It is better for every body. isn't it?

Good bay
Happy New year in advance
 

Similar Topics

Hi Guys, During PID Tuning by Ziegler Nichols Closed-loop method, In TIA Portal there is a Trend Tool used to determine Ultimate Period; marking...
Replies
9
Views
1,820
So it looks like i am stuck with an old reliance DC drive on this project instead of the powerflex dc w/ 22-COMM-E. That being said i am trying to...
Replies
0
Views
1,184
Hello, I am trying to make an Automatic Coolant Dispensing station for my client. This unit will be stationed next to a CNC Machine to monitor...
Replies
56
Views
14,394
We're implementing a second pump in a piece of equipment that pumps primarily water, or fluids similar to water. This pump is located in a stand...
Replies
23
Views
13,061
I have a project where I need to use a proportional valve to control hydraulic system. Before, I used open loop and now I try closed loop PID with...
Replies
3
Views
3,353
Back
Top Bottom