Control valves limits review

[ww2i]

Control valves operating near limits pose a threat to the continuous operation.
How that should be reviewed, mitigated, and controlled by Operations Engineer?
Share some best practices please.

[ivani1]

This is a very important subject.
I believe just to start with, identification of all such cases is the key:
From DCS / historian:
Valve > 90–95% open for long periods - undersized / flow starved
Valve < 5–10% open - oversized / poor controllability
Frequent saturation - control loop struggling
Ask:
Is this normal operation or only during upsets?
Since when has it been like this?
Is production rate higher than original design?

Now coming to the next step, I would say, it is important to check that valve itself is the reason or it can be the process, so check that:
Check Process:
Review
Upstream pressure available
Downstream pressure requirement
Fouling / plugging in lines or exchangers
Changes in feed composition or rate
Manual bypass cracked open (yes, happens more than people admit)
This avoids blaming the valve when the process is the real problem.

The next thing would be to look into the control performance.
Control Performance:
Look for:
Oscillations
Controller output pegged at 0% or 100%
Long recovery time after disturbances
This tells you:
Is it a capacity problem or a tuning problem?

I think this should provide some leads.

[arcpro]

Interesting, never heard such a thing being considered.
Or if being considered by our Operations /Process, will check with them.

[ww2i]

What are the major hazards here if we do not study that way all these control valves running near limits?
Just need to have that high level understanding.

This is a very important subject.
I believe just to start with, identification of all such cases is the key:
From DCS / historian:
Valve > 90–95% open for long periods - undersized / flow starved
Valve < 5–10% open - oversized / poor controllability
Frequent saturation - control loop struggling
Ask:
Is this normal operation or only during upsets?
Since when has it been like this?
Is production rate higher than original design?

Now coming to the next step, I would say, it is important to check that valve itself is the reason or it can be the process, so check that:
Check Process:
Review
Upstream pressure available
Downstream pressure requirement
Fouling / plugging in lines or exchangers
Changes in feed composition or rate
Manual bypass cracked open (yes, happens more than people admit)
This avoids blaming the valve when the process is the real problem.

The next thing would be to look into the control performance.
Control Performance:
Look for:
Oscillations
Controller output pegged at 0% or 100%
Long recovery time after disturbances
This tells you:
Is it a capacity problem or a tuning problem?

I think this should provide some leads.

[opo21]

Control valves running near their limits (almost fully open or fully closed) are a BIG red flag for reliability and process stability.
An Operations Engineer should treat this as an early warning signal, not just “normal operation.”
For critical valves ask:
If this valve fails now, what happens?
Trip?
Fire?
Environmental release?
Production loss?
This links to:
HAZOP
LOPA
Bow-tie analysis
Operations Engineer should document: “Valve V-101 operating at 98% due to increased throughput. Risk = loss of pressure control → column trip. Temporary mitigation = limit feed to X t/h.”

This has to be applied case-to-case to reach a consequence class.

[neo]

This is interesting, particularly when we are discussing to have this monitored periodically.
Does anybody have this listed as any performance indicator to report on some frequency or just the normal operation engineers' deskwork?

[opo21]

This is required to be done as a monthly exercise at the end of each month through the data feed.
Following parameters if are closely looked into would provide the needed feed for a review:
Valve position (%)
Controller output (%)
PV (process variable)
SP (setpoint)
Your PI system should be able to provide this data.