Transcript Presentation - AL/MS Section of AWWA

LABORATORY MANAGEMENT
and QUALITY ASSURANCE
Introduction
“The analytical laboratory provides qualitative and
quantitative data for use in decision-making. To be valuable,
the data must accurately describe the characteristics and
concentrations of constituents in the samples submitted to the
laboratory. In many cases, because they lead to faulty
interpretations, approximate or incorrect results are worse
than no result at all.”
–
HANDBOOK FOR ANALYTICAL QUALITY CONTROL IN WATER AND WASTEWATER
LABORATORIES, EPA 1979
Quality Assurance - Defined
“Quality Assurance (QA) is a set of operating principles that,
if strictly followed during sample collection and analysis, will
produce data of known and defensible quality.”
“The Accuracy of the analytical result can be stated with a
high level of confidence.”
–
STANDARD METHODS, 18th EDITION, 1992
Outline
• Laboratory Management
• Introduction to Quality Assurance Concepts
Laboratory Management
• Who should be involved in laboratory management
and quality assurance?
Laboratory Management
• Everyone involved with the lab:
–
–
–
–
–
Person sampling
Person running the test
Person washing the glassware
Person doing maintenance on the instruments
Person interpreting the results
Laboratory Management
• Quality Assurance Program
–
–
–
–
–
–
–
–
–
–
–
Staff Organization and Responsibilities
Sample Control and Documentation
SOP for Analytical Methods & Procedures
Analyst Training Requirements
Equipment Preventative Maintenance
Calibration Procedures
Corrective Actions
Internal Quality Control Activities
Performance Audits
Data Assessment for Bias and Precision
Data Validation and Reporting
Laboratory Management
• Keys to Quality Assurance Program:
–
–
–
–
–
Documentation
Communication
Training
Cross-Training
Updating
Sample Control and Documentation
• A record keeping system (paper trail, chain of
custody) should track samples before, during, and
after analysis.
• Everyone involved needs to understand and utilize
the system.
Sample Control and Documentation
• Efficiently process information through lab system
while minimizing actual time spent recording data
• Keep it simple!
– Collect only the information you need
Suggested Information - Field
Hayfield Site
Influent
HS IN 1
Code
Site
Conditions
Comments:
Date 04-15-02 8am
Collected By Jim S.
Sunny, 75F
pH adjusted to <2 with
nitric acid
Grab sample
Suggested Information - Lab
•
•
•
•
Date of analysis
Laboratory technicians performing the analysis
Results (including units)
Analytical comments: based on need to know
– Dilutions
– Interferences encountered
SOP for Analytical Procedures
• Describes method in enough detail that an
experienced analyst could obtain acceptable results.
SOP for Cleanliness
• Labware cleaning procedures should be documented
and all persons involved should be trained.
Routine Cleaning Procedure
• Rinse glassware with tap water.
• Clean glassware with a solution of water and
laboratory detergent.
• Rinse the glassware with an acidic solution
– 1.0 N HCl
– 6N HNO3 for regulatory reporting of heavy metals
• Rinse glassware at least 3X with DI water.
Routine Cleaning Procedure (cont.)
• Glassware should be stored in a manner that
prevents contamination from dust particles.
• Prior to analysis, rinse the glassware with sample to
prevent contamination or dilution.
SOP for Instrumentation Maintenance
• Preventative maintenance is the key to optimal
instrument performance.
– Follow any maintenance program and guidelines
suggested by the instrument manufacturer.
– Instrument manual
• Reduces instrument downtime
• Service Contracts with Manufacturers
Analyst Training
• Sample Logging and Preservation
• Method SOPs
• Measuring
– Use of Volumetric glassware
(pipettes, graduated glassware)
• Weighing / Use and care of
Analytical Balance
• Washing and Care of Glassware
• Operation of Analytical
Instrumentation
• Data Handling and Reporting
• Quality Control Activities
• Safety
QUALITY ASSURANCE
CONCEPTS
Quality Assurance
Quality Control
• Certification of Analyst
Competence
• Recovery of Known Additions
• Analysis of Standards
• Analysis of Reagent Blanks
• Calibration with Standards
• Analysis of Duplicates
• Maintenance of Control Charts
Quality Assessment
• Performance Evaluation Samples
• Performance Audits
Quality Assurance
Quality Control
• Certification of Analyst
Competence
• Recovery of Known Additions
• Analysis of Standards
• Analysis of Reagent Blanks
• Calibration with Standards
• Analysis of Duplicates
• Maintenance of Control Charts
Quality Assessment
• Performance Evaluation Samples
• Performance Audits
Certification of Analyst Competence
• Demonstration of acceptable precision and accuracy
for each analyst
• Minimum of four replicate analyses on a known
standard
– Look for acceptable accuracy and precision
– Acceptable limits vary per analytical method
• ‘Demonstration of Capability’
What is Accuracy?
• Accuracy is the nearness of a test result to the true
value.
What is Precision?
• Precision is how closely repeated measurements
agree with each other.
• Although good precision suggests good accuracy,
precise results can be inaccurate.
Imprecise and inaccurate
Accurate but imprecise
Precise but inaccurate
Precise and accurate
Quality Assurance
Quality Control
• Certification of Analyst
Competence
• Recovery of Known Additions
• Analysis of Standards
• Analysis of Reagent Blanks
• Calibration with Standards
• Analysis of Duplicates
• Maintenance of Control Charts
Quality Assessment
• Performance Evaluation Samples
• Performance Audits
Standards
• What is a standard?
– Solution containing a known amount of a
specific substance
– Example – 1.00mg/L iron standard
Standards
• How are standards used?
– Instrument calibration
– Instrument verification/accuracy check
– Analyst training
Standards
• Analysis of Known Standard Solutions – Am I
running the test correctly?
– Verifies instrument, technique, and reagents
Standards
• Analysis of Known Standard Solutions –
– How often?
– Daily, every Sample ‘Batch’?
• National Institute of Standards and Technology
– “NIST”
Standards
• Recovery of Known Additions –
– Is my sample compatible with the test?
– Identifies interferences and percent recovery
• Standard Addition
• ‘Spiked sample’
= 1.00 mg/L
Correct??
1.20 mg/L 1.39 mg/L 1.58 mg/L
1.20 mg/L
33
50 mg/L Iron
Standard
1.40 mg/L
1.60 mg/L
1.39 mg/L
1.58
mg/L
1.20
1.40 mg/L
1.60
mg/L
1.20
34
X 100
100==98.7
99 %
%
100
%
Calibration with Standards
• Some instruments have built-in calibration curves,
not necessary to calibrate
• Instrument without preprogrammed calibration
curves
– Prepare curve daily - OR
– Whenever a new lot of reagents is prepared
Calibrations
mg/L
ABS
pH Calibration Curve
+180
mV
0
-180
4
7
pH
10
Standards
• “It’s what I always get”
• “It meets the permit limit”
• “I did”:
– what the manual said
– what tech support said
– what you told me
• “It’s the same number the
City of ____ gets”
• “I got what I expected”
• “I’ve run standards”
• “It’s a XXX brand
instrument, the best!”
• “After 20 years you get a
feel for it”
• “I’m a chemist”
• “It’s the same answer the
lab got”
Quality Assurance
Quality Control
• Certification of Analyst
Competence
• Recovery of Known Additions
• Analysis of Standards
• Analysis of Reagent Blanks
• Calibration with Standards
• Analysis of Duplicates
• Maintenance of Control Charts
Quality Assessment
• Performance Evaluation Samples
• Performance Audits
Reagent Blanks
• Some reagents contribute color to a sample
– Quantifies amount of reagent contribution to color
formation
– Monitors of purity of reagents
• On each new lot of reagents
• 5% of samples (Standard Methods)
Reagent Blanks
Reagent Blanks
Quality Assurance
Quality Control
• Certification of Analyst
Competence
• Recovery of Known Additions
• Analysis of Standards
• Analysis of Reagent Blanks
• Calibration with Standards
• Analysis of Duplicates
• Maintenance of Control Charts
Quality Assessment
• Performance Evaluation Samples
• Performance Audits
Analysis of Duplicates
• Assesses precision
• 5% of sample need to be Duplicates
– (Standard Methods)
Quality Assurance
Quality Control
• Certification of Analyst
Competence
• Recovery of Known Additions
• Analysis of Standards
• Analysis of Reagent Blanks
• Calibration with Standards
• Analysis of Duplicates
• Maintenance of Control Charts
Quality Assessment
• Performance Evaluation Samples
• Performance Audits
What is a Control Chart?
• Quality control (QC) measuring device that visually
represents the QC data
• Information in a control chart can aid in
determining:
– Probable source of measurement variability
– Whether or not a process is in statistical control
How do Control Charts Work?
• If the chart displays other than random variation
around the expected result, it suggests a problem
with the measurement process.
– Control limits are plotted on the chart, to assess whether
this has happened. The measurement results are
expected to remain within these limits.
Normal Distribution
(Standard Deviation around the Mean)
-3s
-2s
-1s
MEAN
+1s
+2s
+3s
Confidence Limits
68%
-3s
-2s
-1s
10.00
+1s
+2s
+3s
Confidence Limits
95%
-3s
-2s
-1s
10.00
+1s
+2s
+3s
Confidence Limits
99%
-3s
-2s
-1s
10.00
+1s
+2s
+3s
Control Charts
• A control chart is essentially a normal distribution
flipped on its side
• A control chart is a plot of:
– Test units on the vertical scale
– Sequence of time on the horizontal scale
Control Chart
+3s
+2s
+1s
Mean
-1s
-2s
-3s
Control Chart
+3s
+2s
Upper Warning Limit
+1s
Mean
-1s
Lower Warning Limit
-2s
-3s
Control Chart
+3s
Upper Control Limit
+2s
+1s
Mean
-1s
-2s
Lower Control Limit
-3s
How do Control Charts Work?
• Warning Limits
– Set at ±2s
– Standard Methods suggests:
• If 2 of 3 points are outside warning limits, analyze another
sample. If it is within warning limits, continue. If it is outside
warning limits, stop and troubleshoot.
How do Control Charts Work?
• Control Limits
– Set at ±3s
– Standard Methods suggests:
• If any point is outside control limits, analyze another sample.
If it is within control limits, continue. If it is outside control
limits, stop and troubleshoot.
How do Control Charts Work?
• A standard is measured regularly, and the results are
plotted on the control chart.
• Control chart is a graph of concentration versus
time.
Control Chart
Iron Standard, FerroVer Procedure
UC L
UW L
+3s
+2s
+1s
Mean
-1s
-2s
LW L
LC L
Time
-3s
Constructing a Control Chart
• A control chart can be constructed in a variety of
ways:
– Graph paper
– Spreadsheet problem, such as Excel
Constructing a Control Chart
• Analyze 10-15 replicates of a standard.
• Determine the mean and standard deviation.
– Calculate ±2s and ±3s
• Construct the control chart around the mean value
– Use ±2s as the warning limits
– Use ±3s as the control limits
Example – Iron Standard Replicates
Sample
mg/L Iron
8
0.986
1
1.003
9
1.014
2
1.010
10
1.005
3
0.995
11
0.990
4
1.007
12
1.000
5
0.993
13
0.982
6
1.018
14
1.000
7
1.000
15
0.997
Example – Iron Standard Replicates
• Calculate:
–
–
–
–
Mean
Standard Deviation (±1s)
±2s
±3s
Example – Iron Standard Replicates
• Calculate:
–
–
–
–
Mean
Standard Deviation (±1s)
±2s
±3s
1.000
±0.010 (0.990-1.010)
±0.020 (0.980-1.020)
±0.030 (0.970-1.030)
Control Chart
Iron Standard, FerroVer Procedure
1.03 mg/L
1.02 mg/L
UC L
UW L
+3s
+2s
+1s
Mean
1.00 mg/L
-1s
0.98 mg/L
0.97 mg/L
Time
-2s
LW L
LC L
-3s
Constructing a Control Chart
First, set up a
spreadsheet
with columns for
UWL, LWL, UCL,
LCL, and sample
results
Constructing a Control Chart
Fill in values for
UWL, LWL, UCL,
LCL, and sample
results
Control Chart
Iron Standard, FerroVer Procedure
1.03 mg/L
1.02 mg/L
UC L
UW L
+3s
+2s
+1s
Mean
1.00 mg/L
-1s
0.98 mg/L
0.97 mg/L
Time
-2s
LW L
LC L
-3s
Constructing a Control Chart
Fill in values for
UWL, LWL, UCL,
LCL, and sample
results
Constructing a Control Chart
Highlight data
and create a
graph
Constructing a Control Chart
Format graph as
necessary
mg/L Iron
Iron Control Chart
1.05
1.05
1.03
1.03
1.01
1.01
UWL
LWL
UCL
0.99
0.99
0.97
0.97
0.95
0.95
1
2
3
Sample
4
5
LCL
mg/L iron
Example Control Charts
• Control Analysis Results – Week 1
Sample
mg/L Iron
Thurs
0.988
Mon
1.003
Fri
0.992
Tues
0.995
Sat
0.992
Wed
1.006
Sun
1.004
Example Control Charts
mg/L Iron
Iron Control Chart - Week 1
1.05
1.05
1.03
1.03
1.01
1.01
UWL
LWL
UCL
0.99
0.99
0.97
0.97
0.95
0.95
1
2
3
4
Sample
5
6
7
LCL
mg/L iron
Week 1 results
display normal,
random
variation
between the
UWL and LWL.
Example Control Charts
• Control Analysis Results – Week 2
Sample
mg/L Iron
Thurs
0.993
Mon
1.008
Fri
0.989
Tues
1.000
Sat
0.988
Wed
0.996
Sun
0.983
Example Control Charts
mg/L Iron
Iron Control Chart - Week 2
1.05
1.05
1.03
1.03
1.01
1.01
UWL
LWL
UCL
0.99
0.99
0.97
0.97
0.95
0.95
1
2
3
4
Sample
5
6
7
LCL
mg/L iron
Week 2 – Three
or more points
in one direction
indicates a
possible bias in
analytical
results.
Investigate!
Example Control Charts
• Control Analysis Results – Week 3
Sample
mg/L Iron
Thurs
0.986
Mon
1.012
Fri
0.994
Tues
1.000
Sat
0.968
Wed
1.015
Sun
0.997
Example Control Charts
mg/L Iron
Iron Control Chart - Week 3
1.05
1.05
1.03
1.03
1.01
1.01
UWL
LWL
UCL
0.99
0.99
0.97
0.97
0.95
0.95
1
2
3
4
Sample
5
6
7
LCL
mg/L iron
Week 3 – Data
has a high
degree of scatter
to the LCL.
Investigate!
Quality Assurance
Quality Control
• Certification of Analyst
Competence
• Recovery of Known Additions
• Analysis of Standards
• Analysis of Reagent Blanks
• Calibration with Standards
• Analysis of Duplicates
• Maintenance of Control Charts
Quality Assessment
• Performance Evaluation Samples
• Performance Audits
Performance Evaluation Samples
• Standards provided by an outside agency
– ‘Blind’ Samples
Performance Audits
• Inspection to document sampling handling from
receipt to final reporting of results
– To detect any variations from SOPs
– Checklists developed for each analysis type
•
•
•
•
Sample entered in log book?
Meter calibrated?
Standard Analyzed?
Etc., etc…..
LABORATORY MANAGEMENT
and QUALITY ASSURANCE
References
• Standards Methods
• “Handbook for Analytical Quality Control in Water and
Wastewater Laboratories”
– EPA 1979
• Hach Water Analysis Handbook
• “An Introduction to Standards and Quality Control for the
Laboratory”
– Barbara Martin, Hach Company