Segment loading consideration:Control valves

[aaditya]

Hi all,
Just wanted to know has any one faced any problem if the number of control valves on a FF segment is more than four.What i presume here is that 4 is the optimum number.Please share, if there may be some problem..

[PranavJ]

Aaditya,
I dont think it matters whether you put 4 or 5 valves the segment. what matters is how many VCR you are using and that should not exceed. You will find a segment calculator tool on P+F or MTL site. you can download that excel macro file and try to make a demo segment with 5 control valves. It will let u know whether it is feasible or not.

Pranav Jani
Applications Engineer
Honeywell
Abu Dhabi

[dhobart]

aaditya and Pranav,

As much as I might have agreed with Pranav three or four years ago, VCRs are not really the major issue here. For example, the Honeywell C-300 and the DeltaV (Rev 9) now have 84 VCRs per channel which is fairly comfortable and probably far more than the average user will require. Obviously, the VCR count involves all devices on the segment, not just the valves, and, how the control strategy is implemented, but generally, the VCR count is an old challenge that has gone away.

The primary purpose of the various segment design tools (P+F, MTL, Emerson, INtools, etc.) is to check voltages. I am not aware that they check the VCR count. However, your engineering tool will check the VCR count and warn you if you exceed the allowable.

Now, the real concern about how many valves to put on a segment is risk. I suggest you download the Systems Engineering Guidelines from the Foundation website. It has a very good discussion of this design considerations. Even here, when the guidelines were put together, there was still the concern of a lost segment; what happens if this segment fails? The next question must be; what is the likelihood of this segment failing? To a large extent, that will depend on how robust a segment you have designed. With the physical layers tools available today, segments can be extremely robust.

But, you still must answer the question, what happens if this segment fails? No matter how unlikely, the segment is a single point of failure for all the devices on it, no matter how robustly you build that segment. Experience has shown that Foundation fieldbus segments, properly designed, are fairly bullet-proof, but what are the consequences of a failure? The designer must balance the risk against the cost.

David Hobart
Sterling Valley Associates
Automation Engineering and Project Managent

[tklatt]

Hello all,
I agree with David, that today the number of required VCR's is not the problem. Also from the physical point of view today it's no problem to connect 4, 5 or more control valves to one segment. The problem is the risk and, especially in case you are not using the "control in the field" functionality of fieldbus, the cycle time.

Thomas Klatt

Manager International Sales Support PA/
Business Development Manager Fieldbus Systems
GAM/GPP

Pepperl+Fuchs GmbH
Koenigsberger Allee 87
68307 Mannheim
Germany
Phone: +49 621 776-2130
Fax: +49 621 776-1140
E-Mail: tklatt@de.pepperl-fuchs.com
www.pepperl-fuchs.com

PROTECTING YOUR PROCESS

[Mike ONeill]

Glad someone mentioned the risk factor about having all FF-H1 power and communications via one cable. Reliable as they undoubtedly are for most applications, one cable is still a risk that should not be ignored, particularly if Control in the Field is being considered. FF-H1 is the only major bus on plant that isn't duplicated. The FF Middle East End User Meeting in Bahrain 2006 placed redundancy for FF-H1 #3 on their list of desirable developments.

MooreHawke have just released TRUNKSAFE as a fault-tolerant physical layer system which provides a real increase in security for the physical layer. It still uses 2 power conditioners and 2 H1 cards just like everything else, but there is a special automatic termination feature in the field device coupler that allows it to accept 2 segment cables and have both of them active all the time. Now if there is any single fault (H1 card, power conditioner or cable open or short-circuit), the device coupler disconnects that side and auto-terminates the other side, so the segment continues normally. This now offers a real opportunity to use CITF with confidence. The risk to the loop is still the possibility of that individual valve failing, which is the same whatever you do. (To remove that, we have to go into complex device duplication plus 1oo2 or 2oo3 voting software; a much greater expense and complication that may be justified in some exceptional circumstances, but rarely just for control.)

We've calculated MTTF for segments based on data by Dr D J Smith in his books on industrial reliability:
30 years - simplex segment (one power conditioner, one cable)
50 years - duplex segment (two power conditioners, one cable)
345 years - fault-tolerant segment (two power conditioners, two cables)

The best thing is, TRUNKSAFE works with all DCS systems and needs no special software or configuration within that system. It is a really effective and economical solution to the problem you raise, and you only need to use it where you consider that risk needs the mitigation - TRUNKSAFE merges with our normal power conditioner & device coupler packages.