1. Finding and Evaluating Valve Condition:
A part of any Safe and Efficient Operation

2. Presenter – Steve Obermann
A Chemical Engineer with emphasis in process automation.
Expertise in advanced process control designs for refinery and petrochemical process units, process modeling, inferential properties, real time optimization and cost/benefit estimation.
Metso develops and markets pre-packaged industrial software for the processing and manufacturing industries which maximizes productivity, efficiency and reduces waste.

3. Agenda Stuff happens..... Knowing enables avoiding Detection Examples
Questions & Answers

4. Common mechanical valve damage
Bearing, shaft, sealing elements

5. Condition detection Sooner is better, but more involved COST TIME
Obvious Failure
COST
Acceptable operating state
Low-grade and production losses
Unexpected shutdowns
Detection Triangle
Process Measurements
Smart Device level
“Involved” = visual inspection of metric history; must combine outside information to gauge importance/priority
TIME

6. Finding Performance Deterioration
Clues exist throughout the process
Process Plant
Sensor
Controlled Variables
Controllers
Stiction
Manipulated Variable
Sensor Fail
Excessive Noise
Oscillation
Temperature
Pressure
Composition
Flow
Disturbances
Action
Feedback
Sizing
Hysteresis
Tuning
Interactions
Constraints
Non-Linearity
Poor Dynamics

7. Existing Control Performance
INTERNAL
Observations from our evaluation services
10%-35% of control loops are in manual
? Related ?
30% of control valves have problems

8. How is it done...? Gather condition information
Automatic monitoring and analysis via control loop data
Collection via OPC from any DCS /PLC system
Analysis of analog and smart devices
ENG
HIS
PCS
Online
Smart device information analysis
Full life time data available in the devices
Local communicator download
HART or Foundation Fieldbus to DCS collector
Valve leakage analysis with portable device
Valve testing with portable device
Includes experts visual inspection
Offline

9. How is it done...? Smart device evaluation
Examine history on parameters like......
SteadyState deviation
Dynamic deviation
Stiction force
Load for opening
SteadyState Load
Spool Valve position
Look for noticeable change in pattern
Knowledge of the metric and its connection to the mechanics of the device.
SteadyState deviation % (control error, in steady situation)...position target(controller CO and actual VP)
Dynamic deviation % (control error in transients)
Stiction (friction pressure/force needed to change direction)
Load for opening
SteadyState Load min (pressure in actuator in steady situation)
Spool Valve position %

10. Actuator damage example:
Actuator Leak
Actuator membrane has failed, leakage increasing!
Problem
Leak is Developing in the actuator
Positioner compensates for the small leak
Control System Diagnostics
Once the actuator leak becomes more than the
positioner can compensate the process will be affected by a poor response of the valve

11. Condition detected COST TIME Free to prioritize when to repair.
Obvious Failure
COST
Acceptable operating state
Low-grade and production losses
Unexpected shutdowns
Detection Triangle
Process Measurements
Smart Device level
TIME
Free to prioritize when to repair.
Know what materials are needed.

12. Seat damage example:
Broken seat in a ball valve

13. Condition detected COST TIME Catastrophic failure?
Obvious Failure
COST
Acceptable operating state
Low-grade and production losses
Unexpected shutdowns
Detection Triangle
Process Measurements
Smart Device level
TIME
Catastrophic failure?
Steady State load changes over time?
Observable process effects?

14. How is it done...? Process Measurement Diagnostics
Assess performance in time blocks.
Automatic diagnosis based on metric combinations and heuristics.
Embed knowledge of the type of equipment and process
Temperature, level, flow, etc.....
Ball, globe, butterfly, etc...
Fail open/closed, positioner?
Heat exchanger or fired heater?
Connections to economic drivers are possible.
All equipment types analog or digital; positioner or not.

15. EOP – drop leg level control An unplanned shutdown was prevented by the analysis
Automatic valve report to area manager’s mailbox
Diagnosis: Oscillation due to valve
Field Inspection / Actions
The actuator barely connected to the valve (two bolts holding loosely)
The actuator was fixed properly with locking glue since pipe vibrates heavily
An unplanned shutdown was prevented
Extraction with O2 and peroxide- unstable feed situation
Automatic report on bleaching area

16. Condition detected COST TIME Expensive failure avoided.
Obvious Failure
COST
Acceptable operating state
Low-grade and production losses
Unexpected shutdowns
Detection Triangle
Process Measurements
Smart Device level
TIME
Expensive failure avoided.
Trivial repair.
Observable process effects?

17. Filter Reject Tank level control Extension of valve life
1 hour trend before
1 hour trend after
Automatic valve report to area manager’s mailbox
Diagnosis: level control oscillating, and valve moving far too much (13% up and down 5 times/hour)
Field Inspection
Replaced an aged positioner with a new model.
Level control oscillation settled down and amount of valve movement was reduced dramatically
Less movements  less wearing  increased valve lifetime

18. Condition detected COST TIME Schedule corrective measures.
Obvious Failure
COST
Acceptable operating state
Low-grade and production losses
Unexpected shutdowns
Detection Triangle
Process Measurements
Smart Device level
TIME
Schedule corrective measures.
Observable process effects-oscillation cycle
Disruptive to downstream elements

19. A Colorant Dilution flow in Bleaching Predicting maintenance needs
Biggest payback opportunity
Diagnosis: Flow cycling occasionally
Severe wear on valve under these conditions.
Further analysis:
Valve start to sticking due to chemical build-up
Change operating procedure: Started to use condition assessments to schedule line flush
Find right maintenance moment instead of fixed maintenance intervals.

20. Condition detected COST TIME
Obvious Failure
COST
Acceptable operating state
Low-grade and production losses
Unexpected shutdowns
Detection Triangle
Process Measurements
Smart Device level
TIME
Avoided material re-processing, wasted additive.
Opportunity to prioritze against other needs.
Observable process effects- periodic high freq. cycling
Enabled condition based maintenance.

21. Detected issues Pulp mill case examples Valve amounts Pulp mill 1
(1100 valves)
Pulp mill 2
(390 valves)
Typical
World Class
Excessive valve travel
4 %
9 %
5 – 15 %
< 5 %
Stiction
6 %
5 – 10 %
Oversized
20 %
12 %
Undersized
Loops in Manual mode
21 %
10 – 25 %
< 10 %

22. Valve condition analysis
Production plant needs:
Availability
Performance
Quality
Safety
Reliability
Control system
Shutdowns
Evaluate & Diagnose Valve:
Stiction,
Oversize, undersize
Incorrect Characterization
Hysteresis,…
Production
Plans
Device
management
Daily maintenance

23. Benefits Focuses maintenance actions to devices needing it
Improved process performance
Detecting valves causing process cycling
Reduced maintenance costs
Detecting valves requiring maintenance (valves with hysteresis and stiction)
Preventing excessive wear (valve travel and reversals)
Detection of process bottlenecks
Devices operating at limit
Prevent Expensive Consequences
Failure and accident / release
Loss of production
Measureable Benefits

24. Thank You
Questions?