Isa 2008 remote injection montgomery

PROCESS GC ANALYSIS OF A
HAZARDOUS SAMPLE USING A
DISCRETE SAMPLE VALVE
IN A REMOTE LOCATION
Tom G. Montgomery
Siemens Energy & Automation, Inc.
Process Analytics
Presented in Calgary by Bob Farmer, SE&A

Analytical Solutions for Process Control & Compliance
The 53rd Annual Symposium of the Analysis Division
Calgary, Alberta Canada; 20-24 April 2008

Outline
• Introduction and Abstract
• Background and the project requirement
• Solution to project requirements
• Possible issues with solution and investigations
• Result

AD 2008: Analysis Division Symposium Slide 2
Calgary, AB Canada; 20-24 April 2008

Introduction and Abstract
• Practical example of discrete remote sample injection
– Many thanks to Dr. Jimmy Converse

• Owner’s requirement
– Hydrocarbon components in hazardous stream matrix
– The owner made severe constraints regarding sample handling

• Solution
– Description of remote discrete sample injection technique
– Discussion of potential negative impacts

• Investigations
– Discussion of possible negative effects on the measurement


Background
Gas Chromatograph
• GC is a sample extractive technique
– To minimize lag time, a “large” quantity
of sample is removed from process Electronics
and transported to the analyzer at high
flow rates Temperature Oven
– At the analyzer, a small quantity of Controlled
sample is extracted again and Columns Detector
conditioned for presentation to the and Valves
analyzer at low flow rates
– Unused sample is returned to process
or to flare
– In the analyzer a tiny quantity of 3
sample is extracted again and inserted
onto the columns Sample
• Valve
Original extraction and transport
issues Carrier
– Gas
Flow 1: typical 3.5 l/min to ½” tubing
– Flow 2: typical 30-40 ml/min to 1/16” or
1/8” tubing 2
– Sample 3: typical 0.5 µl sample or less
1
Sampling System
Process
Stream

Standard GC Extractive Sampling
Technique
Analyzer Shelter
Unused sample to flare
or return to process

Illustration Ref. Dr. J. Converse

Benefits of Standard Technique

• Ensures fresh sample is present at inlet of GC (reduces
time lag)
– In this application, transport-related lag time = ~30 sec

• No affect on chromatographic method

Analyzer Shelter


Problems of Standard Technique
• Extracts large quantities of sample from process
• Excess sample must be disposed or returned to process
• Failure conditions, including leakage on the transport line must be
accommodated
• Leakage can result in hazards of fire or injury to personnel
– On this project, the stream contained ~2% Hydrofluoric Acid (HF)
– Lethal to humans: STEL <3ppm; Extremely corrosive
Analyzer Shelter


Solution – Remote Discrete
Sampling
Analyzer Shelter

Sample valve
located near
process

Illustration Ref. Dr. J. Converse

Benefits of Remote Sampling
• Eliminates transport of large quantities of
sample
• Minimizes waste
• Reduces issues of leakage away from process
– Reduces human hazards
– Reduces potential for damage


Potential Negative Effects on
Analysis
• Increased cycle time (or lag time)
• Retention time shift
– Effect on valve operating times
– Effect on gating reliability

• Peak effects
– Peak broadening
– Measurement performance


Transport Tubing “Part Of Oven”
Ref BF Main ITC

Transport Tubing must be at
Car. Ref. a controlled temperature: ±?°

Vents 60 Ft “The heated transfer line can
be perceived as a very long,
slender oven.”
(Dr. J. Converse)
KOh C1b

Car. 1
6

6
1

1
C2
5

5
2

2
4

4
3

3
C3
Car. 2

Oven Temp: 60°C ±0.05°C Sample from
Valves and wetted tubing of Hastelloy C Process

Questions To Be Investigated
• How much lag time is introduced?
• How good does the temperature control of
transport tubing have to be?
– Effect on Backflush Valve operation
– Effect on gating of components

• Any other analytical issues?


Test Set
• Maxum GC with divided oven
– Two isothermal halves Detector
– Each temperature
independently controlled
• Application with SV and CV 2 valves
and columns at 60°C (analysis
temperature)
• Coil (60’) of 0.032” ID / 1/16” Divided Oven
OD tubing to simulate process
transport tubing on other side
• Carrier pressure 51 psig and
resulting flow 30 ml/min
• Varied coil temperature from
40° to 80°C in 10°C
increments and plotted
chromatograms
Test Coil Side
AD 2008: Analysis Division Symposium
Application Side (Photos representative of Slide 13
actual test set.)

Lag Time Observed
ITC Without Tubing

ITC With Tubing at 60°C
Ref. Propane:
Elution: ~42sec  151sec
Additional Lag: ~109sec


How Much Shift in BF Time Can
Be Allowed?
Same as previous slide with BF time set out to 240 sec.


Ref. Propane elution = 151sec.

Acceptable retention time shift: <±3 sec.

How Much Retention Shift Is
Acceptable for Gating?
Main Detector – No Tubing

• Adequate separation suggests retention time shifts of ~10 seconds can be
accommodated by normal gating technique.
• Therefore, the BF time requirement will be the primary control. (<±3 sec)

A Quick Test At the ITC
Retention Times:
40°C : 177 (not shown)
60°C : 171.3
80°C : 166.3



Main Detector; Tubing 60°C


Retention Time Effects Summary

• ±20°C shift in temperature of transport tubing
causes retention time changes of >10 sec
• Transport tubing temperature should be
controlled to at least ±5°C to ensure reliable
performance of the analyzer

• (Note: Sample Valve temperature may also vary
±5°C since this is a liquid sample.)


Peak Shape
• Columns used: 0.53 mm ID “megabore” capillary
– Normal response is sharp (Eg; Methane ~5 sec wide
even when retained 2.5 minutes at 40ml/min flow
rate)
– Therefore, negligible peak broadening was expected

• No significant broadening was observed


Measurement Performance
• Due to good peak shape performance,
measurement performance is expected to be
good
• All peaks performed well except C5+ (backflush
peak)
10°C swing in transport tubing
40C 7.46% temperature cause about a relative
1.6% change in the C5+ concentration.
50C 7.57%
60C 7.69% Note: C5+ in this sample is made of
Isopentane through n-Hexane. May be
70C 7.944% a result of backflush of C5s and
possibly C4s at lower temperatures.
80C 8.086%

Summary Of Observations
• Lag Time
– Consistent with predictions from flow calculations
– Acceptable in overall system performance

• Retention Shift
– Tubing must be kept at a constant temperature ±5°C

• Analytical Issues
– Negligible in some cases – but poor performance of
backflush “plus” peak also requires tubing to be kept
at a constant temperature ±5°C


Results
• Remote discrete sampling resolves problems of
hazardous sample handling
• Lag time effects are predictable
• Transport tubing must be kept at a controlled
temperature – but not as tightly controlled as an
analytical oven
• Measurement performance effects are acceptable if
transport tubing temperature is controlled
• Additional benefit: HF neutralization can be
accomplished with a small column instead of a large pot


Other Limitations
• This test was with a liquid sample
– Simpler transport of sample aliquot
– Simpler temperature control requirements on sample
valve (and transport tubing)

• Distance from process to analyzer was not large
(~60ft)
– Lag time increase is acceptable even at relatively low
flow rates of chromatographic columns


Acknowledgements
• Dr. Jimmy Converse (research and publications)
• Steve Trimble (Siemens Process Analytics)


Isa 2008 remote injection montgomery

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Isa 2008 remote injection montgomery

Similar to Isa 2008 remote injection montgomery (20)

Recently uploaded

Recently uploaded (20)

Isa 2008 remote injection montgomery