June 26, 2009 at 3:28 pm #3068
We are entering detail design for our interlock project for a 6 drum Coker. The scope of the project includes the switch valves, vapor valves, isolation valves etc… Delta Valves have been previously installed. My question to this forum is what do others do to prevent valve movement on valves with modern digital Limitorque actuators (MX w/ Modbus)? Intercepting power is not a good option as in our case it would take power from the communication circuit – causing our logic to get hosed up when the position signal is lost. Our current plan is to use the ESD signal through Modbus to prevent the valve from moving out of sequence. For the more severe consequences, the additional hardwired ESD contacts will be used with input from independant process transmitters (ie: drum temp/pressure) through a SIL rated PLC.
There is underlying concern from the operators that without intercepting power the valves are still at risk of moving (voodoo event…..). I’m being reassured by our I&E folks that this is not a concern – but am not getting 100% closure. Unfortunately, we had an incident in the past caused by a valve being closed out of sequence. Investigation summary was that the valve was moved by op’s…but operator did not admit to it. Some tough issues ahead for me…..
Your feedback / wisdom is appreciated.
June 26, 2009 at 3:59 pm #6090
We had simular concerns on a new coker built in 2007. We used Rotork actuators with MODBUS actuation from the DCS. The interlocks are not done through the MODBUS. We take valve postion contacts off the actuator and sent a descrete signal to the valve to the ESD input. This allows the valve to be locked out and can not be operatored from the MODBUS or from the local station. The only way to operate the valve when it is locked out is to engage the clutch and operate the handweel. The clutch has a key lock and chain on it so the operator must involk a lockout/tagout procedure to operate the valve. This requires a management signoff to do.
Our concern was the same. Operators were reluctant to accept valves where the power could not be turned off.. Their concern was that someone could operate the wrong valve locally if the valves were not powered down when not being operated. They were convinced this was acceptable only after we brought in a demonstration valve actuator and show them how it would work.
As a side note we have had some problems with the MODBUS. Occasionally the valve will freeze and can not be operated from the DCS. The only way to get it to work is to cycle power to the valve actuator. We are studing this issue to see how extensive it is.
June 26, 2009 at 4:05 pm #6089
You need to have interlock/PLC signal power and motor operator power separate. We deenergize and lock out the MOV’s at a breaker panel on the structure to protect the open drum from receiving hydrocarbon inadvertently.
I would also recommend not doing a sequence based interlock logic, if that is what you are inferring. I recommend that each valve have a unique relationship with 2 or 3 other valve positions; those valves must be in their proper postion (closed, open, not closed or not open) as a permissive for moving the given valve to the desired position.
Best regards – Mitch
June 26, 2009 at 7:58 pm #6088
Thanks for the response guys. You are helping ease my pain.
We are experiencing issues with our existing Limitorque actuators similar to the Rotork problem mentioned above. So it looks like there is no silver bullet. Limitorque has told us to modify our wiring and also upgrade 14 of our actuators to their latest generation to mitigate our problem. My confidence level is 60%…..
Mitch – regarding sequencing logic…… unfortunately that is the direction we have headed. We have built a gory spreadsheet defining each sequential step with the required valve positions. We tried using a simple permissive table (not step dependent) and it was too complicated with way too many “if” statements that were cycle time dependent. It seemed much simpler to just walk straight through the steps. I’m going to have our team review the spreadsheet and identify the valves that we don’t care to tie into the logic for the next move – that way if we do have a valve with a communication issue, it doesn’t hose the logic for every step. It would only hose the steps where it’s position is critical. I wonder if Foster Wheeler has a standard design/logic we could purchase? Seems like everyone could benefit if there was a standard design.
June 29, 2009 at 8:39 am #6085
For such critical system, all commands should be through hard wired system; Id not recommend using Modbus or any other kind of communication to operate the valves around the coke drums.
Rotork is Ok for all actuators on isolation valves and switch valves, while using “release relay” to disable the main contactor. Thus any move from the motor cannot occur. When the drum switch is done, the operator “release” the valve. Then it will not move until the “release” switch is switched back to his position.
However, no matter what, the human factor will always be there, since everything should be handled from local panel. Even, the release relay could be activated/deactivated from local panel by a local bypass maintenance switch.
August 25, 2009 at 9:45 pm #6020
January 24, 2012 at 7:53 pm #4739
At the Galveston conference on May 2, 2012, a panel of refiners be speaking on Automatic Coker Valve Interlocks, Mitch Moloney leads off with ExxonMobil’s experiences.
1. ExxonMobil leads off the panel with a brief review of their history in developing improved valve operations integrity. Discussion of alternate systems and procedures to reduce risk of incorrect valve line-ups.
LOTO programs, guiding light, trapped key
Cost trends on such systems and how that affects their justification vs alternatives
2. ExxonMobil’s Motor-Actuated design with PLC oversight
Systems on 6 cokers, with 7th in progress
Review of motor actuator types and how they play into reliability of overall system
Review of design defect in some actuators that can cause unwanted valve movement
Review of past incidents at three refineries that caused unwanted valve movements due to not interlocking out power to valves
3. Maintenance responsibilities associated with interlock systems
4. Operations responsibilities
Review of automatic monitoring systems to maintain integrity
Other refiners are lining up to share their lessons learned with the community.
Jschroeder, Craig, Lucky, (posted in this thread) do some of you want to participate? Others reading this, pass on the invitation to your colleagues to contact us.
December 23, 2012 at 8:52 am #4554
this is known as stacking in the valve. cauesd by poor casting when the valve was made. Ive had numerous higher end brands as well as cheapies do this. there is actually a void in the casting which doesnt allow the valve or seal to do their job. fairly certain these parts are in top shape as its most likely a new install. hate to say it but take it back to where you bought it and try diff brand unless youre just in love with this particular style, or cross your fingers if you replace with same model
You must be logged in to reply to this topic.