Hydrocarbon Processing - April 2022 - 86

Process Controls, Instrumentation and Automation
and they often end up overcontrolling by
making too frequent or excessively large
adjustments. As a result, regulatory control
strategies underperform, not only
due to tuning or design flaws, but due
to operator acceptance. Monitoring this
type of operator behavior can reveal improvement
opportunities.
For the first part of this use case, the
control engineer is concerned about
variability in a reactor feed flow controller
setpoint that operators adjust to manage
process inventory. Reactor operation
is extremely sensitive to feed changes.
Operators have been trained to make infrequent
adjustments no larger than 0.5
units, and the inventory does not need to
be tightly controlled. Starting with only
the reactor feed setpoint signal (FIG. 4A),
the control engineer excludes downtime
periods, calculates sample-to-sample
changes and creates time capsules for
Crew 1/2/3 operating shifts.
A histogram (FIG. 4B) is generated
showing the distribution of setpoint
changes by number (y-axis) and crew (xaxis),
with yellow/red bars flagging undesirable
large changes. The analysis reveals
Crew 2 is performing the best (lowest
number of changes, highest green bars).
Crew 1 behavior is the cause of the undesired
reactor variation, evidenced by the
high yellow and red bars. Action is then
taken to learn from Crew 2's behavior and
improve Crew 1's approach.
For the second part of this use case,
the control engineer starts with Crew
1/2/3 shift capsules, and controller
mode signal data (0=manual, 1=automatic,
2=cascade) in an asset data structure
for the 21 controllers comprising
Unit 1 of the process. Downtime data is
removed and the weekly percent time in
manual is set up as a crew-specific calculation
scaled to each controller (asset).
Value
searches
identify when the
weekly percent time in manual is greater
than 15% and greater than 50% to flag
underutilized controllers. Crew behavior
is displayed in an asset-enabled treemap
(FIG. 5), where each Unit 1 controller is
represented by a green, yellow (manual
> 15% of time) or red (manual > 50% of
time) square.
Unlike Crews 1 and 2, Crew 3 is running
TC109 in automatic mode less
than 50% of the time. TC109 is a recently-commissioned
controller designed to
reduce product quality variability. Action
is taken to further train Crew 3 on
the importance of keeping this controller
in automatic.
FC157 and TC119 are controllers
that all crews leave in manual. Crew
feedback reveals the associated control
valves need resizing to avoid saturation
issues, so maintenance is scheduled.
Crew 3 is running PC117 in manual
mode between 15% and 50% of the
time, resulting in its yellow designation.
Conversations with Crew 3 reveal
PC117 always oscillates following an
equipment procedure only scheduled
during their shift. An action item is given
so the control engineer can address
this oscillatory behavior.
This two-part use case shows how
FIG. 5. Crew 1: Unit 1 controllers percent time in manual mode, greater than 50% in red (A).
Crew 2: Unit 1 controllers percent time in manual mode, greater than 50% in red (B). Crew 3:
Unit 1 controllers percentage time in manual mode (> 15% yellow, > 50% red). Crew 3 is
underutilizing TC109 and PC117.
86 APRIL 2022 | HydrocarbonProcessing.com
advanced analytics can efficiently focus
calculations from the micro to the
macro level, depending on the control
engineer's objectives. Starting from historian
data, actionable results were generated
for both a single controller setpoint
change analysis and for unit-wide
controller mode monitoring.

http://www.HydrocarbonProcessing.com

Hydrocarbon Processing - April 2022

Table of Contents for the Digital Edition of Hydrocarbon Processing - April 2022