HEATED, SMART & MODULAR SAMPLE HANDLING
SYSTEM: ENABLER FOR CLOSED-LOOP PROCESS
CONTROL
Jamie Canton
Analyzer Specialist
Lanxess
1265 Vidal Street South
Sarnia, Ontario N7T 7M2
Steve Doe
Analytical Market Manager
Parker Hannifin Corporation
2651 Alabama Highway 21 North
Jacksonville, Alabama 36265
KEYWORDS
Sample Handling Systems (SHS), Modular, Integrated, Heat, Flow Indication, Calibration,
Validation, Distributed Control System (DCS), Closed Loop Control
ABSTRACT
Accurate moisture measurement and control is necessary to maintain profitability in many
chemical and petrochemical processes. While on-line moisture analyzers provide plant
operators with moisture level and trending visibility through a Distributed Control System
(DCS), many process engineers and control room plant operators remain skeptical over the
accuracy of such data. This is because sample handling systems (SHS) are historically prone
to problems that compromise the analyzer’s data accuracy.
This paper will outline the rationale and steps for deploying a modular sample system solution
in a moisture-sensitive process from concept through commissioning and operational
performance. The system utilizes compact SP76(3) modular flow control components in a
heated enclosure, with a novel proportionally controlled conductive heater for both the sample
system and enclosure purge. Key tools enabling operators to interface with the system
include an automated validation function and sample bypass and analyzer flow indication
available over the DCS. These tools enable the process engineer to assign more weight to
the analyzer’s results in the application program. Lessons learned from this project and a
resulting strategic plan for deploying the technology in other areas, including closed loop
process control applications, will also be discussed.
INTRODUCTION
Born of a strategic realignment within Bayer AG in 2004, Lanxess produces Performance
Chemicals, Chemical Intermediates, Engineering Plastics and Performance Rubber in 50
plant locations worldwide. The Sarnia, Ontario plant has produced Performance Rubber
products, including Butyl Rubber, for over 60 years. Accurate moisture measurement and
control are key process variables in the production of quality Butyl Rubber. Legacy moisture
analysis systems at the Sarnia plant incorporated conventional sample systems fabricated
with large volume 3/8” and ½” carbon steel pipe that increased the sample delivery lag time to
the analyzer. Such high volume sample systems require very high flow fast loops to deliver
representative process parts-per-million (PPM) moisture conditions to the analyzer. The high
flow rates led to high steam trace energy requirements necessary to prevent moisture
absorption to the sample transport pipe wall. Steam is readily available and a cost effective
heat trace method at the Sarnia plant, but steam trap or other steam system failures
commonly led to SHS and analyzer downtime.
While plant operators were provided with DCS visibility of the moisture analyzer’s data output,
they were essentially blind to the health and operation of the accompanying SHS. As the
analyzer reported moisture levels approaching out-of-specification conditions, the plant
operator’s typical initial reaction was to not believe the analyzer and report the drift to the
Analyzer Maintenance Department. A technician would subsequently be dispatched to
physically inspect the analyzer and its SHS. A manual validation routine would be done by
cycling stream switching valves to a reference gas and reading local flow indicators to
determine if sample and reference gas flows were indeed introduced to the analyzer. Callouts
for analyzer technicians during evening and night shifts were common and the turnaround
time for such SHS health inspections were as long as 6 hours. Such a time lag that proved to
be extremely costly on several occasions, when the moisture levels actually exceeded
specification resulting in poor product quality and ultimately extended plant downtime.
NEW SAMPLE HANDLING SYSTEM
WHAT DOES “GOOD” LOOK LIKE?
The Sarnia plant, in the year 2000, participated in a Bayer initiative to deploy an analyzer and
SHS downtime measurement system to benchmark performance throughout the Corporation.
A DCS sub-routine was programmed internally, providing plant operators with visibility of an
analyzer’s status (on-line or off-line). However, status change authority was granted only to
the analyzer technicians and all status changes required inputs that documented the cause
behind the downtime. The launch of this program started with reporting latitude that fostered
gray areas in terms of Corporate-wide plant comparisons, but the Sarnia plant engaged the
program as a cornerstone for the plant to focus efforts on eliminating out-of-service conditions
for their nine process gas chromatographs (GC’s). The success of the GC initiative in late
2003 led to serious efforts toward establishing future state analyzer and SHS requirements
necessary to support a Butyl Rubber plant which was prone to downtime due to excessive
moisture conditions. The analyzer project team adopted a strong customer service
perspective in their approach to answering a critical question: “How can confidence be instilled
in the process engineers and plant operators (the customers) to trust and act upon the
analyzer’s data (the deliverable product)?” The answer was simple: include the “customers”
on the project team. This resulted in a list of key system deliverables as follows:
1. Incorporate a “mini-process” methodology -- complete with information delivery -- over the
DCS and operator action assignments based on conditions reported.
2. Reduced sample volume & delivery lag time as well as energy requirements for heat
trace.
3. Remote validation functionality to operators in the control room.
4. Remote flow indication to operators in the control room.
New Sample System/Sensor Initiative(1,2,4) (NeSSI™) technology was identified as a potential
alternative platform for SHS fabrication early in the project cycle. The attractive features of
this technology included:
1. Proven, robust, automated double block and bleed (DBB) stream switching capabilities.
2. A technology roadmap that outlined additional functionalities such as smart sensors
and on-board analysis.
3. Reduced sample volume that would result in faster sample transport times.
4. Compact system size that enabled additional sample systems and analyzers to fit into
existing shelters.
5. The ability to develop standardized system designs for order as a single part number
from a manufacturer, which solved internal manpower fabrication capacity issues.
The features and capabilities of various manufacturers of NeSSI™(1,2,4) system platforms were
researched, and the project team chose the Parker IntraFlow™ system for their first
NeSSI™™(1,2,4) project. The project team saw advantages with a field-proven air-actuated
(DBB) stream switching system, which was seen as a critical requirement to deliver the
remote validation function to plant operators. Additionally, the robust stream switching
functionality provided confidence to enable multi-stream analysis on a single analyzer, a
practice previously not allowed at the Sarnia site. A remote flow indication device, which held
promise to provide key analyzer flow information to the operators, was also in development at
the time.
DESIGNING AND BUILDING THE NEW SAMPLE HANDLING SYSTEM
The SHS manufacturer was selected, and the system design for the Butyl Rubber PPM
moisture analyzer SHS was undertaken in early 2005. The overall system design from a
sample conditioning standpoint is relatively simple, as shown in Figure 1.
= Membrane Separator
FT
= Flow Transmitter
FIG. 1: BUTYL RUBBER MOISTURE SAMPLE SYSTEM DESIGN
The system features two membrane separators in series to protect the moisture analyzer from
flooding due to heat trace failures. Such flooding had previously proven to be very costly in
downtime, as purging dry-down was a lengthy process. The membrane separators are
configured in an in-line fashion, without traditional bypass legs, which will be addressed
momentarily. Plant Nitrogen is dried in an on-board desiccant dryer and is used as a standard
reference gas for calibration and validation functions. A normally closed (NC) air operated
DBB switching valve can be called upon to direct sample flow to the analyzer, while an
adjacent normally open (NO) air operated DBB valve directs the standard reference to the
analyzer. The NO and NC switching valves are controlled by solenoid-operated pilot valves
on-board the SHS, and their orientation provides a power or instrument air fail-safe condition
for reference gas flow to the analyzer.
Bypass and analyzer flow indication is accomplished with 1 psig differential pressure (dP)
sensors measuring pressure drop across restricted orifice plates embedded between the
sensors and the substrate blocks upon which they’re mounted. The sensor, manufactured by
Honeywell Sensotec and shown in Figure 2, incorporates a robust micro-machined silicon
structure that changes resistance (ohms) when force is applied. A wheatstone bridge -arranged to measure diaphragm movement from pressure changes up and downstream of the
restricted orifice -- provides 1% accuracy over a wide-temperature range. The sensor’s
transmitter sends a 4-20mA output to the DCS system to provide operators visibility of flow
conditions. The dP sensor can be calibrated to provide actual flow rates (within 10%
accuracy), but the project team did not pursue this capability as it was felt that relative flow
indication provided the necessary information required by the plant operators.
FIG 2: 1PSID SENSOR FLOW INDICATORS
The membrane separators configured in series without their respective bypass legs would
appear to be a practice to avoid, as moisture slugs tend to quickly flood and plug the
membrane separator, preventing flow to the analyzer. However, the addition of the analyzer
flow sensor provides a unique method to alert plant operators of flooding and plugging
conditions, as a rapid decrease in the analyzer’s flow rate will cause an alarm condition on the
DCS. Once detected, corrective action can immediately be taken on the condition, yet the
analyzer remains protected from flooding. The system can be quickly purged by virtue of
manual valves that can direct reference gas through the entire system. The second
membrane separator is simply a fail-safe device to protect the analyzer. The membrane
separators also provide excellent filtration, as excessive moisture in the system tends to wash
particulates from the tube walls.
Representative sample delivery in PPM moisture analyzer applications is susceptible to
temperature influences throughout the system, as moisture absorbs to cool stainless steel
surfaces and levels will equilibrate throughout the system. Electric traced sample transport
lines and an enclosed, heated SHS, maintaining a constant 50°C sample temperature
throughout the system, was therefore specified. In the case of the first NeSSI™™(1,2,4) based
system, the manufacturer’s “pegboard” backplane design feature supported use of a new,
“smart” conductive heater provided by Intertec Instrumentation, Limited. The Class 1, Division
1 conductive heater block is clamped between the backside of the pegboard and the top of the
enclosure-mounting rail system, as shown in Figures 3a and 3b. This simple method allows
the inflexible rigid conduit (necessary for electrical area classification compliance) to be
installed independent of the more flexible tube and power runs required on top of the
system.
FIG 3a: HEATER INSTALLSTION
FIG 3b: PEGBOARD INSTALLATION
The system also required an instrument air purge as a Health and Safety measure to protect
technicians from potential hydrocarbon exposure upon opening the enclosure door when
servicing the SHS. This type requirement is common, but it also creates a conflict in SHS
heat management, as purge air is supplied through unheated and un-insulated tubing. Winter
conditions in Eastern Canada therefore set the stage for cold purge air working against the
efforts of the SHS heater. A simple purge air pre-heater was installed to resolve this conflict,
by milling a serpentine slot on the backside of the pegboard to encase a ¼” air supply line, as
shown in Figure 4.
FIG 5: PURGE SYSTEM PRE-HEATER
Computational Fluid Dynamic (CFD) software calculations indicated the pre-heater supply a
10x volumetric enclosure purge rate per hour at a constant 46°C temperature with a heater
set point of 50°C, 60psig air supply pressure at a worst case ambient temperature of -40°C,
as outlined in Figure 5.
FIG 5: CFD PURGE AIR PRE-HEATER RESULTS
The “smart” features of the heater include a proportional controller that receives a temperature
signal from a thermistor strategically positioned on the pegboard. The combination of the
smart controller and conductive heater mounted directly to the pegboard provides a very
stable sample temperature, even when the enclosure door is opened on a cold winter day in
Sarnia. The basis for the conductive heater efficiency is the direct heating of pegboard
(essentially making it a hot plate), which in turn conductively heats the fluid-handling substrate
blocks and flow-control components, a much more efficient method than convectively heating
the air inside the enclosure. Figure 6 outlines the responsiveness provided by the heater; as
the set point is incrementally increased, the sample temperature similarly increases. Note the
heater and energy percentage output curves provide an interesting signature of the
proportional controller’s performance.
FIG 6: INCREMENTAL SETPOINT AND SAMPLE TEMPERATURE RESPONSIVENESS
FIELD RESULTS
The system was successfully commissioned at the Sarnia site on June 24, 2005, with
analyzer and bypass flow-rate information delivered to the plant control room operator's DSC
terminal along with the ability to remotely actuate the stream switching valves to validate the
analyzer’s calibration. Figure 7 details the first hour’s analyzer and flow conditions.
FIG 7: DSC SCREEN SHOT OF 1st HOUR’S SYSTEM PERFORMANCE
The yellow sinusoidal curve on the top of the chart indicates the bypass flow signal and the
variations are indicative of inlet pressure fluctuations. The most interesting line is the bottom
blue line, which is the analyzer flow indicator. The signal initializes with indication of nitrogen
reference gas flow through the cell for 9 minutes 25 seconds, followed by a 5 second
interruption in flow as the switching valves (actuated from the control room) shut off the
reference gas and open the sample flow. The sample flow signal is a slightly different value,
revealing the specific gravity difference between the sample and the Nitrogen reference gas,
and the associated influence on the dP sensor. The signature repeats itself as 9 minute, 25
second switching intervals continue throughout the hour-long snapshot.
Training the control room plant operators on the new tools was a straightforward endeavor as
they had ownership in the project due to their early involvement. The validation functionality
sends an “OK to use analyzer data” signal through the DCS, which has eliminated costly lab
sample validations. Analyzer technician time checking systems that were properly functioning
has also been significantly reduced. A flag is sent to both the operator and analyzer
technician requesting corrective action in cases where the validation does not pass.
Understanding the signal limits and signature based on normal events has provided
benchmarks of normal operating conditions. Abnormal conditions similarly present signatures
that are easily detected as “different,” and a library of such condition signatures has been
cataloged that provides operators with associated corrective actions. Confidence in the
analyzer and sample system increased significantly and the analyzer’s data plays a much
more significant role in the process engineer’s application program.
Since startup, the system, as shown in Figure 8, has posted a 100% uptime record and the
analyzer’s data is regarded as absolute and conclusive. Process instrumentation has not yet
been installed to run closed loop on this application. The system proved its worth when an
unexpected moisture event was detected by control room plant operators. Following protocol,
a validation cycle was initiated from the DCS, including confirmation of flow to the analyzer.
The analyzer subsequently reported the same uncharacteristic moisture condition that
prompted corrective actions on the process. Management, process engineering and control
room plant operators all agree that such an event prior to the installation of the new tools
would have resulted in a costly plant shutdown. The key enabling tool that overcame the
shutdown condition was the analyzer’s sample flow indicator over the DCS. The dP flow
inference technique provides a robust, cost-effective solution that process engineers had
requested of the analyzer department for many years.
FIG 8: NEW MODULAR SHS INSTALLED ON SITE
CONCLUSION
The project design, implementation and commissioning cycle was a collaborative effort
involving several suppliers integrating several new products and technologies into the system.
The overall system functionality expectations were achieved, but several developmental
issues that required attention throughout the project cycle prolonged the project. The project
team and suppliers persevered through these challenges to see a successful conclusion and
determined the following lessons learned:
1. Plant Operations’ involvement in the early stages of the project proved to be invaluable.
Their early input and buy-in created an atmosphere of teamwork that set the stage for a
win-win conclusion.
2. Radical departure from legacy mindsets (breaking free of “the way we’ve always done
it”), requires open-mindedness to determine the vision and single-mindedness to work
through the inevitable hurdles.
3. Selling management on the future state vision of SHS capabilities, then fulfilling it with
a successful implementation, easily prevailed over initial budgetary concerns. Since the
initial project’s conclusion, the value provided by confidence in the analyzer’s data and
prevention of a plant shutdown have resulted in a strategic implementation plan for
installing such systems throughout the Sarnia site.
REFERENCES
Paper:
1. Gunnel, J.J., Dubois, R., “NeSSI™ Systems: The Approach to a Best Practice;” Center for
Process Analytical Chemistry/University of Washington, Spring Meeting, NeSSI™
Workshop, May 5, 2005, Best Western University Tower, Seattle, Washington.
2. Gunnel, J.J., Van Vuuren, P.; “Process Analytical Systems: A Vision for the Future;”
Fourteenth International Forum Process Analytical Chemistry (Process Analysis &
Control).
3. ISA Draft 76.00.03, “Modular Component Interfaces for Surface-Mount Fluid Distribution
Components-Part 2: Elastomeric Seals.”
4. Van Vuuren, Peter; Dubois, Rob; Koch, Mel: Center for Process Analytical
Chemistry/University of Washington, “New Sampling System/Sensor Initiative” Information
Package,” Instrument, Systems and Automation Society-2000 Conference & Exhibition,
August 20-24, 2000, Morial Convention Center, New Orleans, Louisiana.