S10.22 ASSURING TRACEABLE ACCURACY OF TRACE CONCENTRATION GAS STANDARDS

1
S-10.22ASSURING TRACEABLE ACCURACY OF TRACE
CONCENTRATION GAS STANDARDS

• JIM MCKINLEY
• KIN-TEK LABORATORIES, INC
• La Marque, Texas 77568

2
Introduction

• Historic Focus of On-Line Analysis Development
• Getting Comparative Data Quickly
• Increasing Specificity
• Increasing Sensitivity
• Improving Reliability
• Current Focus
• Satisfying Regulatory Requirements
• Satisfying QA Program Requirements

3
Need Demonstrated Accuracy

• Driving Forces
• EPA
• Documented QA Programs
• ISO, TQM, Six Sigma, etc.
• All Require Traceability
• OSHA
• Litigious Society
• How Do You Get Demonstrated Accuracy?

4
Traceability-What is It?

• Different Meanings in Different Contexts
• In Manufacturing
• the ability to unerringly track all of the
  specific materials (identified by supplier and
  lot data), both durable and consumable, that are
  used to create a unique product (identified by
  serial number)
• In Calibration
• the ability to compare data to an accepted
  standard

5
For Our Purposes

• Traceability is
• The property of the result of a measurement or
  the value of a standard whereby it can be related
  to stated references, usually national or
  international standards, through an unbroken
  chain of comparisons all having stated
  uncertainties5
• 5International Vocabulary of Basic and General
  Terms in Metrology

6
Traceability The Source

• In the USA, National Standards are set and
  maintained by NIST - The National Institute for
  Standards and Technology
• NIST maintains vertical traceability to the
  International System of Units (SI)
• They maintain horizontal traceability to other
  national laboratories
• They supply Standard Reference Materials (SRMs)
  for use as Resident Standards
• They maintain the NTRM (National Technical
  Reference Material) program for commercially
  supplied standards

7
Traceability Facts of Life

• By the VIM definition, the best path to
  traceability is comparison to a SRM or NTRM
• SRMs are available at some trace concentration
  level for about 35 compounds
• There are over 500 compounds of current
  industrial or environmental interest
• Clearly, there is a problem with comparison
  directly to an SRM or NTRM

8
Paths to Traceability

• Compare to an SRM
• Approach traceability through physical standards
• Relate mixture concentration to physical
  variables (mass, length, time, etc.)
• Make traceable measurements of those variables
• Then, the mixture concentration is traceable

9
Requirements for Traceability

• From the VIM definition we need
• Concentration related to stated references
  (national or international standards)
• An unbroken chain of comparisons
• All having stated uncertainties

10
Types of Gas Standards

• Static blends
• i.e., a cylinder mixture
• Dynamic blends
• Mixtures of flowing gas streams

11
Static Blends

• Advantages
• Very user-friendly
• Very little ancillary equipment required
• Controlling variable (mass) is readily traceable
• Problems
• Undefined errors
• Mixture stability
• Potential for container to interact with the
  mixture
• Traceable only by comparing to other traceable
  standards

12
Dynamic Blends

• Advantages
• Method avoids undefined error sources
• Applicable to wide range of compounds and
  concentrations
• Problems
• Requires special equipment
• Measuring the trace component (analyte) flow
• Flow measurement traceability for a wide range of
  gases

13
Dynamic Blending Methods
14
Comparison of Methods
15
Direct Gas Blending
16
Direct Blending Advantages

• Simple
• Flow controllers can be calibrated to 1 of
  reading
• Useful over a very wide concentration range
• Rapid Equilibration
• Can be used to dilute a trtaceable, higher
  concentration mixture

17
Direct Blending Problems

• Not well adapted for liquid (or solid) analytes
• Getting traceable flow meter calibration
• Must be calibrated with the actual gas to be
  metered
• Big problem Flow standards for very few gases
• Common practice is to calibrate to N2 and apply a
  correction factor to the response
• Accuracy decreases to 5 or less, and is not
  traceable
• Maximum dilution step is 1041
• Not practical for large numbers of components

18
Are The Requirements Met?

• Concentration related to stated references
  (national or international standards) - Yes
• An unbroken chain of comparisons - Maybe
• All having stated uncertainties - Rarely

19
Permeation Tubes
Permeation Tubes are devices that use
permeation through a membrane to control the
analyte compound flow
f is measured by measuring rate of weight
loss due to permeation
20
Using Permeation Tubes
21
The Concentration Created
Where Ci is the concentration of component i in
the mixture fi is the flow of component i from
the permeation tube measured by weight loss
rate (traceable measurement) F is the dilution
flow (traceable measurement usually available)
Concentration is determined by traceable (or
potentially traceable measurements
22
About Permeation
The permeability K of an analyte compound
through a membrane is the product of its
diffusivity and solubility
23
The Analyte Emission Rate
Where f is the flow rate of analyte
emitted by the tube K(T) is the
permeability of the membrane to the analyte at
temperature T A is the membrane
area ?P is the partial pressure
difference of analyte vapor across the
membrane t is the membrane
thickness
24
Temperature Sensitivity
Where
are constants T is the absolute
temperature inK
25
The Controlling Variables

• Temperature
• Partial pressure difference across the membrane
• Total pressure is not a controlling factor for
  emission from the tube.

26
Permeation Tube Advantages

• Wide range of analyte compounds available (gt500)
• Very stable devices
• Emission unaffected by system backpressure
• Wide range of dilution ratios possible
• Maximum practical single step dilution 1081
• Minimum practical single step dilution 5x10-31

27
Permeation Tube Problems

• Equilibrium device
• Requires days-weeks for initial stabilization and
  emission rate measurement
• Certified tube requires hours to stabilize after
  temperature change
• Very temperature sensitive
• Needs well designed apparatus for effective use

28
Are The Requirements Met?

• Concentration related to stated references
  (national or international standards) - Yes
• An unbroken chain of comparisons - Yes
• All having stated uncertainties - Usually

29
The Diffusion Tube Method

• Diffusion tubes are another way of controlling
  analyte vapor flow for trace concentration
  mixtures
• Similar to a permeation tube, but analyte flow is
  controlled by diffusion across a capillary tube.
• Emission flow at fixed conditions is usually
  measured by measuring rate of weight loss.

30
Diffusion Tubes
Emission flow fi
31
Diffusion Tube Emission
Where E is the emission rate of the
tube D is the diffusion coefficient at
temperature (T) and pressure (P) M is the
molecular weight of the analyte P is the
total pressure T is the temperature
A is the cross sectional area of the diffusion
path L is the diffusion path length
p is the partial pressure of the analyte
vapor Controlling variables
32
Diffusion Coefficient
Where D0 is the diffusion coefficient at
standard temperature and pressure
(STP) T0 is 273.1 K T is the
operating temperature in K P0 is 760 Torr
P is the operating pressure in Torr m
is a constant usually 2, but may be 1.75
33
The Controlling Variables

• Temperature
• Pressure
• Partial pressure of the analyte (insignificant in
  this application)

34
Diffusion Tube Advantages

• Rapid equilibration
• Great for high boiling, semi-volatile compounds
• Good for producing higher emission rates for
  short term use
• Low cost method for one-shot screening or survey
  applications

35
Diffusion Tube Problems

• Very temperature sensitive
• Pressure sensitive - including transient pressure
  changes
• Require well designed apparatus for effective use
• Contamination of the capillary can cause spurious
  changes in emission rate

36
Are The Requirements Met?

• Concentration related to stated references
  (national or international standards) - Yes
• An unbroken chain of comparisons - Yes
• All having stated uncertainties - Usually

37
A Look At Stated Uncertainties

• Each method requires multiple measurements
• The uncertainty of each measurement may be known,
  but what is the uncertainty in the resulting
  concentration?
• Finding the concentration uncertainty requires a
  little calculus

38
A Calculus Lesson
C F(x,y) Where C is the calculated
concentration of the mixture F(x,y) is
the equation used to calculate C from the
measured values of x and y
The question is how much will C change (?C) for
a given change in x and/or y? Dig out the old
calculus book and look up total differentials.
Where ?C is the total error in
concentration are the partial
derivatives of F(x,y) with respect to x and y
are the uncertainties in the
values of x and y
39
The Uncertainty in Concentration
Where U is the uncertainty of mixture
concentration C in percent ?C is the
total error in the calculated concentration
40
Application to Direct Blending
From the 2 gas system described in Slide 15,
For convenience, call
Calculate ?C (this is where you need the old
textbook)
41
Calculate the Uncertainty
Recall that
So, dividing the equation for ?C by the equation
for C gives
If both F (dilution flow) and f (analyte flow)
are measured to 1, then ?F and ?f are each
0.01. Inserting those values gives
42
Comparison of Methods
1Assumes flow meter calibration with actual
analyte gas to 1 of measured value. 2Assumes
analyte flow meter calibrated to a surrogate gas
and the flow value obtained by applying a
flow factor to the measured value. 3Assumes the
emission rate measured in the generating
apparatus at operating conditions. 4Assumes
typical commercial practice of using factory
certified tubes.
43
Conclusions

• Globalization and regulation are driving
  increasing importance of traceability for
  chemical measurements
• For most mixtures, traceability must be obtained
  by comparison to physical standards
• Dynamic blending methods offer the best
  possibilities for traceability
• The best bet techniques are Direct Blending,
  Permeation Tubes, and Diffusion Tubes
• Each can be well quantified and has defined
  uncertainty