Calibration and use of Balances in a Quality System
Calibration and Daily use of
Balances
• Choosing the right balance
• Uncertainty of measurement
• Setting appropriate pass fail criteria
• Summary and Open Question Session
Good Laboratory Practice (GLP)
Requires:
• Qualified and competent
personnel
• Defined appropriate equipment
and conditions
• Testing facility designed to
minimise problems
• Record management/Archive
Good Laboratory Practice (GLP)
Requires:
• Suitable Test and Reference Standards and
Internal Audit System
• Documented Methods or SOP’s
• Regular Calibration of Measuring
Equipment According to Accepted Methods
For GLP Balances have printouts to
identify:
• Model & Manufacturer
• Serial number
• Software version
number
• All adjustment routines
For GLP Premium Balances have:
• Internal Adjustment
Weights & isoCAL
• Real Time Clock
• Alpha Numeric Data
Input
• Bar Code Scanner
Connectivity for sample
ID’s
Typical GLP/GMP Compliant
Printout
10.04.2005 11.45.23
Sartorius AG
Model ME235S
Ser. no. 60419914
Ver. no. 01-30-01
ID QC Lab 22
SQmin 45mg
-----------------------------------------------
Internal Calibration
Start: isoCAL/temp
Diff. + 0.00007g
Internal adjustment completed
Diff. + 0.00000g
-----------------------------------------------
10.04.2005 11.46.23
Name:
Calibration and Daily use of
Balances
• Terminology - How to understand
specifications
• Weight Units
• Basic Principle of an Electronic Balance
• Care in the use of balances
• Hazards to avoid
Understanding The Technical data:
Readabilit
y
• The smallest difference that
can be measured.
• Resolution or
Discrimination
Weighing Capacity
• The maximum weight
that can be applied
before overloading
occurs
Reproducibility of a
Balance
± 0.01g
x= average weight
s= standard deviation
68.3% of the weights measured lie within the range x s
95.5% of the weights measured lie within the range x 2s
99.7% of the weights measured lie within the range x 3s
Reproducibility of a
Balance
± 0.03g
x= average weight
s= standard deviation
68.3% of the weights measured lie within the range x s
95.5% of the weights measured lie within the range x 2s
99.7% of the weights measured lie within the range x 3s
Linearity
Deviation from a Theoretical Strait
Line
Weight
readout
1 kg
(500+)
g
0.01g
500 g 1 kg Weight loaded
Sensitivity Drift – with 4 place
balance
• ± 1x10-6 per degree Celsius or:
• ± 1 part per million per degrees Celsius
200.0000g/1,000,000
= 0.0002g
Sensitivity Drift –with 5 place
balances
• ± 1x10-6 per degree Celsius or:
• ± 1 part per million per degrees
Celsius
200.00000g/1,000,00
0
= 0.00020g
Sensitivity Drift
• Normal use of 5 decimal place balance is to
weigh small amounts into containers
10.00000g/1,000,000
= 0.00001g
Response Time
• How long it takes the
balance to become
stable
Terminology
Was this you?
• Accuracy
• Readability
• Sensitivity
• linearity Error
• Repeatability
• Uncertainty
Accuracy
• Is the extent to which the
readout approaches the true
value of the object.
• Can not be specified
• Can be determined
Errors - out of level weighing
• Balance table 1m x 1m
• Raise one side 5mm
• Before 200.00000g
• After 199.99850g
• Difference 0.00150g
Errors - changes in regional gravity
• Ground Floor 200.00000g
• First Floor 199.99974g
• Difference 0.00026g
Good Weighing Practice
• Keep Balance Clean
• Site Balance on Firm Surface
• Level the balance
• Free from Drafts
• Not in Direct Sunlight
• Not near a Radiator
Good Weighing Practice
Hazards to Accuracy
• Electro Static Influences from
• Plastic Containers
• Oven Dried Glassware
• Some Samples
• Magnetic/ Electromagnetic Effects
• Stirring Bar!
• Steel Containers
Calibration and Daily use of
Balances
• Calibration weights
• Design
• Class of weights
• How to Adjust a Balance
• Internal v External Weights
• Daily Calibration Checks
Weight Units
Unit Unit Relation to
Symbol Base Unit
Nanogram ng ÷ µg 1000
Microgram µg ÷ mg 1000
Milligram mg ÷ g 1000
Gram g ÷ 1000
Kilogram kg Base Unit
Ton t x 1000
Calibration Weights and OIML
• Organisation International
de Metrology Legale
• Specifies
Design, Material, Surface
Quality, Tolerances,
Markings, & Manufacture
• RI 111
Classes of Calibration Weights
• E1 Solid Stainless Steel
No markings. 1kg +/- 0.5mg
• E2 Solid stainless Steel
No markings 1kg +/- 1.5mg
• F1 Stainless Steel with
adjustment cavity 1kg +/-
5.0mg
Handling and care of weights
• Weight Calibration Laboratories store
weights under a Glass Dome
• E2 and F1 store in Boxes supplied
• Suitable lining material
• Free from fibres
Handling and care of weights
• Use lifting devices
• Plastic forks
• Tipped forceps – keep clean!
• Cotton or chamois gloves
Handling and care of weights
• Remove dust with camel hair brush
• Never touch weights
• Fingerprints are acidic and greasy
• Clean weights with cotton or chamois
• Alcohol or steam cleaning removes
absorption layer
Calibration of Laboratory Balances
Calibration test with 10g weight
Hardware Validation or IQ/OQ
• Equipment Qualification
EQ
• What is it?
• Why?
• Description!!
Installation Qualification or IQ
• Completeness Check
• Record Serial #
• Software Version
• Location
• Accessories
• Appropriate Safety
Standards
Operational Qualification or
IQ
• Testing against
Specification
• Subject to Environmental
conditions
• Testing Functionality of
Balance
• Essentially a Full
Calibration
Performance Qualification PQ
• Performed by the Operator
• Daily or Weekly
• User Defined
• One or Two Calibration
Weights
• Outlined in the SOP
USP Measurement to 0.1% max Uncertainty
• United States Pharmacopoeia
- section 41
• Implemented by the FDA
Food and Drugs
Administration
USP Measurement to 0.1% max Uncertainty
• Pharmaceutical Laboratories
• All measurements to be
made with intended
accuracy
• 0.1% maximum uncertainty
What is Uncertainty?
• Sum of systematic and random errors
• Systematic errors:- from the balance
• Random Errors:- from operators, and the
environment,
Random Errors
• Operators
• Off Centre Loading
• Occasional Poor conditions
• Occasional Drafts
• Occasional Vibration
Systematic Errors
• Inherent performance flaws
• Errors in Calibration
• Temperature change
What is Uncertainty?
• Measured mass 20.0000g
• Determined Uncertainty 0.0003g
• Result is likely to lie between
19.9997 and 20.0003g
Measurement to 0.1% max Uncertainty
USP
• According to USP - section 41
• Maximum of 0.1% uncertainty in any
weighing
Measurement to 0.1% max
Uncertainty
• For USP Uncertainty is accepted as equal to
3 times the determined SD -in the place of
use
Measurements to 0.1% max Uncertainty
• To determine if balance is acceptable for
300mg
• Standard deviation = 0.1mg
• Uncertainty = 0.1 x 3
• Uncertainty = 0.3mg
Measurements to 0.1% max
Uncertainty
• So we would like to know if
• 0.3 mg ≤ 0.1% of 300mg
Measurements to 0.1% max Uncertainty
USP statement:
SD x 3 /sample weight ≤ 0.001
0.3/300 = 0.001
Measurements to 0.1% max Uncertainty
• Even easier
• Minimum weight is calculated:
• Measured SD x 3 x 1000
• 0.1mg x 3 x 1000
• 300mg
Typical Theoretical minimum
weights
Readability Typical Minimum
weight
• 1mg balance 3000mg
• 0.1mg balance 300mg
• 0.01mg balance 45mg
• 1µg balance 3mg
Determination of Uncertainty with 10g mass
• Reproducibility ≤ 0.07mg
• Temp drift 5C
• 10g x 1 x 10-6 x 5C = ≤ 0.05mg
• Max linearity = ≤ ± 0.15mg
• E2 calibration error = ≤ ± 0.015mg
• (200g max 0.3mg)
Determination of Uncertainty with 10g
mass
According to the Law of
Propagating errors
Sartorius
Sartorius
mg
change corrected
U = 0.21mg
Determination of Uncertainty with10g
sample
• We can now say:
• 10.0000g ± 0.21mg uncertainty with 95%
confidence
Finally: Setting pass fail limits
• You can determine what is possible
from your balance
• You have a measured value and
uncertainty for your calibration weight
• Set limits in accordance with tolerable
error in your assays
Verified Balances according to
90/384/EEC
