The following discussion is part of an occasional series showcasing the ISA Mentor Program, authored by Greg McMillan, industry consultant, author of numerous process control books, 2010 ISA Life Achievement Award recipient, and retired Senior Fellow from Solutia, Inc. (now Eastman Chemical). Greg will be posting questions and responses from the ISA Mentor Program, with contributions from program participants.
See Part 1 here. See Part 2 here. See Part 3 here. See Part 5 here.
Here we continue our series with some measurement and final control element mistakes seen by Greg that are widespread. For more knowledge, see Greg’s 51 tips in the ISA book, 101 Tips for a Successful Automation Career.
Greg McMillan’s Question
What are the biggest mistakes you have seen in automation system design, configuration, calibration, installation, checkout, commissioning, and maintenance? What were the consequences and the fixes and what can be done to prevent future occurrences?
Greg McMillan’s Answer
Six mistakes come to mind that are pervasive, dating back 40 years or more that don’t seem to be widely recognized. My hope is that we can move forward and realize these mistakes are the source of a lot of bad data and poor process control. Artificial intelligence (AI) and machine learning (ML) touted today will be confused. We need to know and fix the root causes.
1. Lack of recognition of the effect of composition, fluid properties, velocity profile and magnitude, and an accuracy in percent of span instead of percent of reading on vortex and differential head flow meter rangeability. Due to excessive error and noise at minimum flow, the actual rangeability is less than ½ what is commonly stated.
Key learning points: Magmeters and Coriolis mass flow meters are not affected as much by process fluid and piping, with an accuracy in percent of reading offering an order of magnitude, better rangeability and accuracy, and much lower installation and maintenance costs.
2. Lack of recognition of the inherent advantages of resistance temperature detectors (RTD), thermowell type and fit, and an overuse of thermocouples in terms of an order of magnitude improvement in drift, precision, linearity, and signal magnitude.
Key learning points: If permitted by process temperature, a spring loaded sheathed RTD in a stepped thermowell with sufficient immersion length offers the best performance. See the ISA book, Advanced Temperature Measurement and Control, for more details.
3. Lack of recognition of the pH electrode response time and equilibration time that can dramatically increase based on electrode design, velocity, fouling, damage, and aging. A measurement electrode 86% response time can increase from 3 to 30 seconds for different glass formulations and shapes and can increase from 3 seconds to 300 seconds or more due to fouling and aging. For cracks or severe dehydration, there may be no response. Reference electrode equilibration time can increase from seconds to days for difficult process applications and the junction potential can be excessive due to contamination.
Key learning points: Use semi-spherical bulbs with glass formulations designed for fast response with a protective shroud and a fluid velocity of at least 1 fps, increased to 5 fps to prevent fouling. The electrodes must be calibrated in buffer solutions with the same ionic strength and solvent as the process. While flowing junction reference electrodes have the fastest equilibration time, least contamination and least potential, solid-state or multiple junction reference electrodes are often the choice to eliminate the need for pressurization and refilling of an electrolyte reservoir. If solid-state reference electrodes are used to prevent contamination, precondition them in a representative process sample for the reference junction to reach equilibrium before installation. This can be done by an automatic retractable probe that is cleaned, rejuvenated, calibrated, and equilibrated before reinsertion. The use of three probes and middle signal selection enables a single probe to be serviced with pH control loop in auto. See the ISA book, Essentials of Modern Measurements and Final Elements for Process Industry, for more details.
4. Lack of recognition of misguidance in control valve specifications by a focus on capacity and leakage with no fields or mention of 86% response time, resolution and dead band requirements, and an increasing emphasis on reducing project cost leads to the common mistake of using on-off valves as throttling valves, causing sustained oscillations in order of magnitude larger than a throttling valve. Furthermore, the internal closure member position does not match the readback differing by 8% or more due to lost motion, resulting in the smartest positioners being clueless as to what is really happening.
Key learning points: The user needs to cite valve response requirements as noted in ISA-75.25.02 Annex A and choose a valve originally designed for throttling service. An on-off valve close coupled and coordinated with the throttling valve can be used to meet leakage requirement and provide isolation. Throttling valves with the best precision have ultra-low friction (ULF) packing and diaphragm actuators sized to give more than 150% of the maximum required thrust or torque. Sliding stem valves should be flow-to-open and have solid rather than caged plugs. Rotary valves should have splined actuator shaft to stem connections, an integral cast ball or disk stem, and minimal seal contact.
5. Lack of recognition that a positioner replaced with a booster on a diaphragm actuator can cause a dangerous instability seen in rotary valves slamming shut. This mistake is the result of the common misguidance that positioners should not be used on fast loops.
Key learning points: For faster response, a positioner should be tuned with more gain and no integral action, and, if necessary, a volume booster put on each positioner output with a bypass needle valve slightly open to stop high frequency oscillations from positioner output going into a small booster volume. See ISA-75.25.02 Annex A for more details.
6. Lack of recognition that attempts to minimize energy use in valves and variable frequency drives (VFDs), which can cause severe nonlinearity and loss of rangeability.
Key learning points: A small ratio of valve pressure drop to system pressure drop or VFD system pressure drop to pressure rise causes a linear inherent flow characteristic in a valve or VFD to distort to a quick opening installed flow characteristic. The rangeability is correspondingly adversely affected by the large flow corresponding to the resolution and deadband in the valve due to lost motion and friction and in VFD response due to input signal card resolution and deadband settings. The excessive slip from older VFD inverter technologies further deteriorates VFD rangeability. See the ISA book, Essentials of Modern Measurements and Final Elements for Process Industry, for more details.
