Facebook
Google
GitHub
Linkedin
A
In debate about the benefits of Foundation Fieldbus what you don’t hear much talk about is The Process. The process control community seems focussed on technology for technology sake and has forgotten about the processes we are trying to control. Most people seem happy these days if a signal eventually, at some point in time, gets from the transmitter, through the control system, to the final control element.
Quoting Bela Liptak (author of the Instrument Engineers Handbooks that most process control people have used)
“What concerns me in this switch from analog to digital is a global standard of 4-20 is being replaced with eight protocols needing translator cards to talk with each other. Murphy says, "if cards can get mixed up, they will. "I am also concerned that the people who understand the needs of users are not deciding the direction of the process control industry, and instead we are sailing under the direction of salesmen and programmers. This is bad, because you cannot control a process if you don't understand it.”
Found online here:
http://www.isa.org/InTechTemplate.c...Management/ContentDisplay.cfm&ContentID=35438
When early DCS systems replaced single loop controllers and strip chart recorders, they often masked the process variability problems in our processes with slow technology, and poor trending units. After many years of DCS development we had control systems that gave us good control and a good view of the process. Process historians gave us 1 second update rates, and we had the ability to get the high speed data right off the 4-20 with technology like Entech’s toolkit. Analysis of this data allowed us to improve our processes, reduce process variability, and save money. Instead of moving towards systems that could bring higher speed data into the DCS and analyze it there we went backwards….
Foundation Fieldbus.
Foundation Fielbus is a painfully slow technology. In an age of 10 Gigabit Ethernet, and cheap processing power, it is unforgivable that the job of process control is being set back by this slow technology. Our expensive transmitters sit out there with access to all this high speed process data and we can’t get at it because of this slow bus sitting between us and our process.
I have seen numerous Fielbus presentations where the author proudly states “our standard for this project was control in the host system”. Just what does that mean to the ability of this system to do a good job of controlling the process? It means very poor performance because control activity in the controller is not synchronized with activity on the bus. Example: If you have a 500-millisecond macrocycle on the bus, the recommended module scan rate in the controller is 2 seconds. Those with DeltaV should read “Deciding Where to Run Control Function Blocks” in Books Online.
So just how fast should it be? Read Automatic Controller Dynamic Specification from Emerson’s Entech Division available here:
http://www.emersonprocess.com/entechcontrol/publications/
To me it is obvious that many current Fieldbus installations have high levels of installed process variability that the owners are not aware of, it is masked by the technology just like it was with early DCS systems. This problem will cost the plant owner money ever second that the plant is in operation. So much for all the hype about the savings related to Fieldbus installations.
Now all the Fieldbus experts whose careers and company profits depend on you buying Fieldbus will not agree with this. But they are denying the math of Process Control!
Quoting Bela Liptak (author of the Instrument Engineers Handbooks that most process control people have used)
“What concerns me in this switch from analog to digital is a global standard of 4-20 is being replaced with eight protocols needing translator cards to talk with each other. Murphy says, "if cards can get mixed up, they will. "I am also concerned that the people who understand the needs of users are not deciding the direction of the process control industry, and instead we are sailing under the direction of salesmen and programmers. This is bad, because you cannot control a process if you don't understand it.”
Found online here:
http://www.isa.org/InTechTemplate.c...Management/ContentDisplay.cfm&ContentID=35438
When early DCS systems replaced single loop controllers and strip chart recorders, they often masked the process variability problems in our processes with slow technology, and poor trending units. After many years of DCS development we had control systems that gave us good control and a good view of the process. Process historians gave us 1 second update rates, and we had the ability to get the high speed data right off the 4-20 with technology like Entech’s toolkit. Analysis of this data allowed us to improve our processes, reduce process variability, and save money. Instead of moving towards systems that could bring higher speed data into the DCS and analyze it there we went backwards….
Foundation Fieldbus.
Foundation Fielbus is a painfully slow technology. In an age of 10 Gigabit Ethernet, and cheap processing power, it is unforgivable that the job of process control is being set back by this slow technology. Our expensive transmitters sit out there with access to all this high speed process data and we can’t get at it because of this slow bus sitting between us and our process.
I have seen numerous Fielbus presentations where the author proudly states “our standard for this project was control in the host system”. Just what does that mean to the ability of this system to do a good job of controlling the process? It means very poor performance because control activity in the controller is not synchronized with activity on the bus. Example: If you have a 500-millisecond macrocycle on the bus, the recommended module scan rate in the controller is 2 seconds. Those with DeltaV should read “Deciding Where to Run Control Function Blocks” in Books Online.
So just how fast should it be? Read Automatic Controller Dynamic Specification from Emerson’s Entech Division available here:
http://www.emersonprocess.com/entechcontrol/publications/
To me it is obvious that many current Fieldbus installations have high levels of installed process variability that the owners are not aware of, it is masked by the technology just like it was with early DCS systems. This problem will cost the plant owner money ever second that the plant is in operation. So much for all the hype about the savings related to Fieldbus installations.
Now all the Fieldbus experts whose careers and company profits depend on you buying Fieldbus will not agree with this. But they are denying the math of Process Control!
