Stan: Installation can make or break an automation project. Yet we don’t get to see much in the literature from the people in the front lines of electrical and instrument (E&I) construction. These guys are exceptionally busy and not tuned much into publishing.
Greg: Fortunately we have Tim Green, operations manager for field services at Maverick Technologies making the extra effort to share what has been learned to realize better installations and smoother and safer startups. Tim and I have a common goal at this point in our career to prevent knowledge from being lost and doing as much as we can to advance our profession. We started this conversation with Tim in the November 2014 Control Talk, “What You Need to Know about Starting Up a Plant or System.”
Stan: What do we need to know up-front?
Tim: A design that is rushed or incomplete can cause more problems than is commonly recognized. In an attempt to cut project costs and meet aggressive project schedules, loop sheets and installation details may not be done or may be delayed until the project is completed. The people making these decisions may not know that the people doing the installation are piping and mechanical contractors who, in most cases, do not have any real knowledge of instrumentation. E&I field services can fill in the blanks, but the starting point should not be a blank piece of paper. The better the documentation, the better the focus can be on dealing with the special requirements.
Greg: When I was in E&I construction, I had to teach union pipefitters how to handle, calibrate and install instruments. Fortunately, Monsanto had a rich heritage of developing standards for installation details and loop drawings. I can’t imagine not having this documentation. We had enough challenges without the guesswork. We also had a close relationship with the E&I, mechanical and process design engineers. We could routinely ask questions and could rely upon their presence during checkout, commissioning and startup. Today, different companies may be doing the design and construction, and the design engineers may move on to the next project instead of going to the field to help and to find out what works and doesn’t work. If this is the case, web meetings and phone calls must be used and relationships must be fostered to maximize communication.
Stan: What are the main design considerations that would help in E&I construction?
Tim: A complete design should be issued prior to construction whenever possible. Sometimes design changes after the drawing package has been issued for construction (IFC) are inevitable, but a well-structured change-management system should include full communication to construction as soon as a change is approved. Additional installation charges and the impact to the schedule must be considered, and adjustments should be made accordingly.
Often, drawings such as loop sheets are considered maintenance drawings and are issued after construction is complete. It is important to realize that every drawing produced for the maintenance of a system is also needed by the installation team.
Instrument details should be provided whenever possible to ensure that transmitters are installed according to the manufacturers’ recommendations. An improperly installed instrument will result in years of unreliable service and often make it impossible to reach the stated accuracy of the instrument.
Greg: Smart transmitters and digital positioners today have an extremely low failure rate and nearly negligible drift rate. What do you see as the major reasons why instruments are not performing on startup?
Tim: Instruments and valves not meeting application requirements are mostly due to mistakes made in design and delivery, and damages that occur during construction.
Instrument and valve verification should be performed when the equipment is received on the site. This can be performed by checking the devices against the data sheet as soon as possible. If there are problems, they are caught prior to installation so you reduce the lead time for a replacement and eliminate the rework from installing an improper device. Lead times for ordering instruments are often six to eight weeks. The last thing you want is a startup waiting on instruments.
Stan: What are the main installation considerations?
Tim: In-line instrumentation is often installed by mechanical contractors and is obviously sensitive to location and physical orientation to the process line it is measuring. For this reason, continued diligence during the construction process must be maintained through performing daily walk-downs with someone knowledgeable of instrumentation. An onsite resource should also be available to answer any questions that may come from the electrical contractor. Again, a sustained regiment of daily walk-downs prevents needless rework.
Greg: In my second stint in E&I construction, we didn’t realize open thermowells were installed before the sand blasting of the pipe lines was completed. The thermocouples were dutifully inserted to bottom out in the thermowell for E&I checkout. We did not realize until the temperature trends were too smooth and slow on startup that the bottoms of the thermowells were filled with sand. What can we do to reduce the damage or adverse effects from construction activities?
Tim: Hydrostatic testing must be performed with special attention given to the control valves and instrumentation that is installed in the affected lines. After careful review of all installed items, block valves and bypasses must be used to prevent overpressurization of these devices during testing. Most instruments are not rated for the pressures during hydrostatic testing.
The initial line flush or pneumatic blowdown of newly installed process piping must be performed in a manner that will not push debris into instrumentation and control valves. Disconnect devices as needed to prevent introduction of foreign materials. Debris can not only damage sensing elements, but also can become trapped in the U-tube of Coriolis meters, the impulse lines of pressure transmitters and the trim of sliding-stem valves.
Welding is another way of destroying instrumentation. It is rare to find a bad instrument from the supplier these days. An instrument failure is almost always due to a welder. By the time the instrument is checked, the welder is gone. Welding currents of 60 amps flow through the pipe to the ground clamp. If there is an instrument between the welding rod and the clamp, a circular path between the sensing element and transmitter electronics smokes the circuit board. Any welding that must occur after electronic instrumentation is installed must be done with the ground attachment point as close to the weld as possible.
Stan: Particularly vulnerable to damage from debris are turbine meters and electrodes. There should be strainers upstream of the sensors. Normally, there are strainers on the pump suction. Sensors should not be installed on a pump suction line for this and many other reasons (e.g., greater risk of flashing and cavitation). Strainers should be verified to be in place and to provide enough protection for the sensors used.
Greg: pH electrodes should not be installed until the process fluid is being circulated and startup is imminent. Having pH electrodes just sitting in stagnant process fluid can result in coating of the electrodes. The use of them during water batching is also not advisable because the reading will be meaningless, and the likelihood of deterioration of the glass or reference junction from anything other than actual flowing process fluid is too great.
Stan: During the commissioning of pressure transmitters and differential pressure transmitters, the impulse lines need to have been flushed out and established to the state intended (e.g., dry or full of process fluid). Purge flow rates must be set and verified. Equalizing valves must be closed. Stream tracing must be properly set to prevent freezing, but not cause vaporization in the impulse lines.
Greg: Control valves should be stroke-tested at maximum operating pressures to verify the actuator size is large enough to provide smooth opening and closing of the valve and minimum stick-slip. In an effort to be more competitive in bids, suppliers may choose marginally sized actuators. The result is a tendency for there to be greater stick-slip, particularly for operation near the closed position where seating/sealing friction and valve pressure are both greatest. ISA Standard 75.25.01 defines this stick-slip as resolution and details the small step changes needed to measure the response of a control valve. Steps of 0.5% are needed to verify valve performance, rather than the 25% steps used in previous decades to check if a valve worked. Changing the direction of the steps also reveals the amount of dead band from backlash. Fortunately, today’s smart digital valve positioner (e.g., digital valve controller) can detect changes as small as 0.1% by position feedback measurement. For sliding-stem valves, changes in the valve stem position provide an excellent indication of internal trim movement. For rotary valves, the positioner feedback is an actuator shaft position that may or may not be representative of the actual internal ball or disk rotation due to connections and linkages between the shaft, stem and ball or disk. On-off valves with low leakage should not be used for throttling, and control valves should not be used for isolation. An on-off isolation valve in series and coordinated with the opening and closing of a throttling valve is best from a performance and safety viewpoint. The October 2007 Chemical Processing article “Improve Control Loop Performance” and my 2015 book Tuning and Control Loop Performance - 4th Edition offer further details on this deception and the impact of valve design deficiencies.
Stan: We conclude with a Top 10 E&I Construction “Believe It or Don’t” List
Top 10 E&I Construction “Believe It or Don’t” List
(10) Plant had up-to-date standards for installation details and loop sheets.
(9) Plant had up-to-date process flow diagrams.
(8) Process and mechanical engineers helped develop new installation details.
(7) Project manager refused to start construction until all drawings were complete.
(6) No welding rod was left in any vessel.
(5) Welder asked instrumentation to be removed from all lines being welded.
(4) Plant refused to accept on-off piping valves as throttling valves.
(3) Undersized valve actuators were found before water batching.
(2) ISA standard was used for control valve response testing in pressurized lines just before commissioning.
(1) Plant insisted on the use of integral mounted wireless transmitters to eliminate impulse lines and signal wires.