CONDUCTIVITY MEASUREMENT AND ITS CALIBRATION Leif Jensen, Insatech - - PowerPoint PPT Presentation

CONDUCTIVITY MEASUREMENT AND ITS CALIBRATION

Leif Jensen, Insatech A/S

Agenda

februar 2017 2

• Conductivity – what is it ? how/why it is measured
• Conductivity – why is it important ? performance and

regulatory requirements

• Conductivity – the importance of temperature and its

compensation

• Conductivity – how to calibrate
• Quality assurance

What is Conductivity in Liquids?

februar 2017 3

• Conductive liquids are

called electrolytes

• Electrolytes have positively

and negatively charged particles called ions.

• Positive charged Ions are

Cations and Negatively charged ions are Anions

Conductivity – What is it?

februar 2017 4

• Conductivity is a measurement of how well electrical

signals pass through a liquid

Conductivity – what is it?

februar 2017 5

• Conductivity is a measurement of how well

electrical signals pass through a liquid

• Used to measure
• Chemical concentration
• Control of water purification plant –

Water for injection and purified water systems

• Concentration of dissolved solids
• Interface detection – is it clean or dirty
• Heat exchanger breakthrough detection
• Defined as reciprocal of resistance,
• Unit of measurement - Siemens –

nS,uS,mS nano,micro,milli

The SI unit of conductivity is S/m and, unless otherwise qualified, it refers to 25 °C (standard temperature)

Cell Constant

februar 2017 6

• To make a measurement of current flow and

the unit siemens useful, it is necessary to define what is being measured.

• The unit Specific Conductivity defined as

conductivity of a liquid column with a length

• f 1cm and a cross section of 1cm

S/cm = S * Length cm / Area cm2

• All sensors are supplied with a cell constant,

a factoring number derived from the relationship between the size of the contacts and their distance apart. CC = Length cm / Area cm2

• The SI unit of conductivity is S/m and,

unless otherwise qualified, it refers to 25 °C (standard temperature)

Conductivity – How is it Measured?

februar 2017 7

• It is measured with a transmitter and

a sensor that has a defined cell Constant! Or an Installation Factor

• The sensor and transmitter will have built in

temperature compensation – incredibly important due to the temperature effect on Conductivity measurement

Conductivity – How is it Measured?

februar 2017 8

• Two methods of measurement
• 1 - Contacting- specific closed cell design & 4 electrode OPEN cells.
• 2 - Inductive – Toroidal.

In a Pharmaceutical manufacturing environment both methods are used

Typically polished sanitary sensors - pure waters and CIP

• Closed cell design for water purity /steam condensate – WFI systems –UP systems
• Range is usually 0.056 uS to 100 uS
• New Open sensor design – wide range from 100uS to 200mS

Toroidal/open cell for concentration - Typically CIP only

• Usually CIP/SIP skids and dilution applications – washers
• Range is typically 1000uS (1mS) to 200mS

MEASURING PRONCIPLE vs. CONDUCTIVITY RANGE

februar 2017 9

How each Measuring Principle is applied currently

Conductivity – Why is it important? WFI UPW

februar 2017 10

For purified Water and Water for Injection (WFI), the USP defines the following conductivity requirements:

• The USP <645> calibration and performance requirements
• Meter reports uncompensated conductivity or uncompensated resistivity.
• The display resolution is 0.1 μS/cm or better.
• The meter reads accurately to ±0.1 μS/cm when a 0.1% precision resistor replaces the sensor

(to calibrate/verify the meter).

• The sensor cell constant is calibrated/verified to ±2%
• Temperature accurate to 2°C (effective USP 28)
• Appropriate dynamic range to meet the above requirements. There are no specific requirements for the

dynamic range of the conductivity system, but the conductivity system operate in pharmaceutical waters typically 0.2 to 4 μS/cm and in the fluid that the sensor is calibrated in (ASTM D1125 solution D, 146.9 µS/cm)

Conductivity – Why is it important? CIP

februar 2017 11

For CIP Clean In Place - it is imperative that the correct cleaning is carried out using the correct conductivity measurement

• Correct measurement leads to correct concentrations at high temperature - 1% Caustic at 80

degC for example

• The incorrect Temperature Compensation will typically cause 20%-30% more Chemical use
• For higher Concentration Aggressive washes – 4% Caustic – 3% Nitric , you will damage your

plant – Diaphragms seal, Valves etc by constantly being at high temperature and larger concentrations than you think .

• To ensure the Final rinse has been successful

Sensor Geometry – specific or contact sensors

februar 2017 12

In practice sensors do not use flat plate construction

• Rod within tube
• Rings
• Facing electrodes
• Polished sanitary type 316L

Open Sensors

februar 2017 13

• Measuring range from 0.1 to

1,000,000µS/cm

• Suitable for pure water and

chemical concentration

• Easy clean construction
• Polished surface finish
• However – very, very challenging

to calibrate!

4 pole cell Field

februar 2017 14

INDUCTIVE – PRINCIPLE OF OPERATION

What is the transmitter doing? DC Voltage Excitation

februar 2017 15

• If a dc excitation voltage is used then the electrically charged ions will gather at the
• ppositely charged plate forming a non conductive layer.
• This is called polarisation.
• Electrochemical reactions can also take place (eg electro plating is carried out in this

way).

Ac Voltage Excitation

februar 2017 16

• When an ac voltage is used the ions move alternately towards first one plate and

then the other.

• Therefore min. polarization occurs (assuming the correct frequency is used).
• All manufacturers uses square signals with changing polarity in complex ways.

Temperature Effect

februar 2017 17

• Temperature has a very pronounced effect on the conductivity of a solution.

The magnitude of this effect is variable with:- Solution type Solution concentration Temperature Temperature change

• Temperature effect is typically non linear – this makes complex matrices or

calibration correction quite challenging.

• Temperature effect can be as large as 7-10% per Degree C – especially in

purer solutions!

• The SI unit of conductivity is S/m and, unless otherwise qualified, it refers to

25 °C (standard temperature)

Cause of temperature dependence

februar 2017 18

• Speed of ionic travel
• Dissociation. Variable up to 10%/°C

Water = H2O non conductive Dissociated H+ OH- conductive

• At higher concentrations is the effect of “speed of ionic travel” is much

greater than the dissociation effect, but in pure water dissociation is the dominant effect

• NOTE: Generally a 10°C temperature increase will improve
• cleaning efficiency by 50% (above 30°C) – this is why CIP is usually at

80 degC ni =

Zi · e 6 · π · h · ri · E

Review of importance of temperature

februar 2017 19

• If temperature compensation is turned on,

temperature sensors should be calibrated within the conductivity sensor

• If temperature compensation is used , choose the

correct algorithm for compensation – it could save you 30% on your chemical costs , and save your plant

• If Temperature compensation is Turned OFF , you

must report the temperature with the conductivity value

• Do an audit of all instruments before commissioning

Look at the varying temperatures in a cycle

februar 2017 20

Temperature Compensation UPS

februar 2017 21

NaOH matrix

februar 2017 22

If you only have 1 transmitter doing two jobs which curve?

Sample

• Salt solution (sodium chloride)
• 5% NaOH (sodium hydroxide)
• Dilute ammonia solution
• 10% HC1 (hydrochloric acid)
• 5% sulfuric acid
• 98% sulfuric acid
• Sugar syrup
• 10% KC1 (potassium chloride)

% per °C from 25C

2.12% 1.72% 1.88% 1.32% 0.96% 2.84% 5.64% 1.88%

februar 2017 23

Incorrect Compensation effects – NAOH 1%

Incorrect Compensation effects – NAOH 1% With NaC1 comp v NaOH or base comp

februar 2017 24

Incorrect Compensation effects – Phosphoric acid1%

Incorrect Compensation effects – Phosphoric1% With NaC1 comp v Matrix or Acid comp

februar 2017 25

Conductivity Calibration

februar 2017 26

TWO METHODS ARE TYPICALLY USED IN CONDUCTIVITY CALIBRATION

• Comparison to another instrument (calibration reference

system-Insacal) usually accredited

• Use a known value standard solution (standard solution)

usually traceable

MANGLER OVERSKRIFT

februar 2017 27

• TRACEABLE CALIBRATION

The term ”measurement traceability” is used to refer to an unbroken chain of comparisons relation an instrument’s measurements to a known standard. Calibration to a traceable standard can be used to determine an instrument’s bias, precision, and accuracy.

• ACCREDITED CALIBRATION

What is Accreditation? – Accreditation is a voluntary, third party-reciewed

• process. As part og accredtitation, a laboratory’s quality management

system is thoroughly evalutated on a regular basis to ensure continued technical competence and compliance with ISO/IEC 17025. Laboratory accreditation can only be granted by an accreditation body, or AB. Although there are a number of accreditation bodies in the US, customers should choose calibration and testing laboratories accredited by Abs having an MRA with ILAC. In Singapore this is the Singapore Accreditation Council (Singlas)

COMPARISON VS. STANDARD

februar 2017 28

Comparison Calibration (Insacal)  In-line – while the sensors is in-situ  At-line – simulate installation  Laboratory – controlled environment Standard Solutions  Stabile conditions required -especially below 100uS Uncontrolled environment Variations in temperature

Comparison Calibration

februar 2017 29

Advantages

• Multi-point Cell Constant calibration in a agitated homogenous mix
• Multi-point temperature calibration possible
• Accurate and 4:1 test accuracy achievable
• It doesn’t matter what value or purity of the solution as it effects both Master and UUT in

same way Disadvantage

• Cell needs to be taken out of the process – unless a single pint verification method is

used

Calibration against Another SUPERIOR calibrated MASTER METER- INSACAL

More examples

februar 2017 30

More examples

februar 2017 31

Conductivity Calibration

februar 2017 32

• Using a Standard Solution compared to a Master

Calibration Reference Standard

Let us look at some of the issues

• Ease of use
• Uncertainty
• Traceability
• Process
• Cost

Ease of use

februar 2017 33

STANDARD SOLUTIONS

• Standard solutions are sensitive to temperature

changes –typically 2% per degC – this means a temperature change of 1 degC can add 2% to the reading – so temperature changes account for almost all the allowable uncertainty of USP

• The temperature of the standard solution must

be measured, and the temperature influence must be defined in order to compensate for this influence.

• Some solutions are only defined at a fixed

temperature, in these cases the standard solution has to be placed in a temperature controlled bath at the fixed temperature.

• Some low conductivity solutions are also

influenced by ambient CO2 and/or humidity, these problems are normally solved by limiting the time the solutions are exposed to ambient

• air. (Time has to be defined)
• INSACALTM Master

Reference Standard

• Hook up the rig l and place sensor and UUT in

an aggitated solution of nominal value ( 1.3 -5- 10 uS ) and then wait for the temperature to stabilise and read the conductivity and temperature.

Ease of use

februar 2017 34

STANDARD SOLUTIONS

• Before any calibration is performed the sensor

under test has to be thoroughly cleaned as any contamination of the standard solution will influence the conductivity, even clean/pure water will dilute the standard solution.

• The temperature compensation in the

transmitter under test has to be switched off or adjusted to the coefficient of the standard

• solution. (If known)
• If the sensor under test is of the open or

inductive type, an highly accurate mock up of the installation has to be used during the calibration as these types are influenced by the physical installation.

• INSACALTM Master

Reference Standard

• Hook up the rig l and place sensor and UUT in

an aggitated solution of nominal value ( 1.3 -5- 10 uS ) and then wait for the temperature to stabilise and read the conductivity and temperature.

Ease of use

februar 2017 35

STANDARD SOLUTIONS

• Wait for the temperature to equalise after the

sensor under test is placed in the solution, then read temperature and conductivity.

• The conductivity of the standard is then

determined by interpolation or by controlling the temperature of the bath to a value where the conductivity of the standard solution is known.

• INSACALTM Master

Reference Standard

• Hook up the rig l and place sensor and UUT in

an aggitated solution of nominal value ( 1.3 -5- 10 uS ) and then wait for the temperature to stabilise and read the conductivity and temperature.

Uncertainty

februar 2017 36

STANDARD SOLUTIONS

• Solutions at low conductivity are at best 1%.

When purchased and usually exhibit errors close to 15 to 20%!!! See Gingerella report and

• thers – demonstrate – open a new bottle!
• High conductivity solutions have accuracies

from 0,15% and up so not so challenging at the CIP type applications

• Commercially affordable high conductivity

solutions with uncertainties of 0.5 to 1% are

• affordable. But the solution is only a part of the

equation

• Factors are :
• Accuracy of temperature probe
• Temperature homogeneity of the solution.
• Contamination
• Interpolation uncertainty
• Resolution
• Skill
• All these factors to be taken into account!
• The end result can easily be an uncertainty of

several percent.

• INSACALTM Master

Reference Standard

• In the range of 100µS/cm to 10 mS/cm it is

possible to determine the cell constant with less than 0,4% uncertainty

• The uncertainty of the transmitter under test

can be enhanced by calibrating the transmitter with resistors prior to the loop calibration.

• The most important factor is that the reference

cell and the cell under calibration have the same temperature during the calibration. If not a temperature compensation has to be performed adding some uncertainties.

• With a test accuracy ratio of 4:1 from the

master to the UUT ( looking for 2%) then you can have a high degree of CERTAINTY that a good calibration has been achieved

Traceability

februar 2017 37

• INSACALTM Master

Reference Standard

• The reference standard is calibrated in an

accredited Lab

• First the indicator is electrically calibrated in all

Sub-ranges, this can be done by an accredited institute.

• Secondly indicator and cell are calibrated with a

standard primary solution.

• This rig is an accredited calibrator with a high

level of accuracy of <0.5% and gives a good 4:1 test accuracy directly to the UUT STANDARD SOLUTIONS

• There are different levels of traceability
• Traceable to a standard method

e.g. ASTM 56 or OIML

• Traceable to an accredited institute e.g.

Hamiltons solutions

• Made by an institute who is accredited by

an organisation which is under the umbrella of B.I.P.M. eg. NIST, DFM (Danish accredited institute when it comes to conductivity solutions)

• This still leaves no direct line to an

accredited calibration unless you pay 1000s of euro per litre for accredited solutions - and these have a very short shelf life with only the best lab conditions allowable to achieve the result

Process

februar 2017 38

STANDARD SOLUTIONS

• Cell has to be dismounted for calibration resulting

in:

• Production interruption.
• Possible pollution in clean applications

e.g.

• WFI. Requiring new sterilisation etc.
• INSACALTM Master

Reference Standard

• Cell can remain in the process
• Calibration can be performed during production

if required – it may only be a 1 point but it is a great insurance

• If flow outlet of the reference fitting is drained

away, then there can be no pollution of the process.

• The calibration can be performed in a closed

loop.

Cost

februar 2017 39

STANDARD SOLUTIONS

• Depending on conductivity, traceability and

uncertainty, the solutions can cost from 50€ up to 500 € per bottle

• Over and above the cost of the standard

solution, a considerable cost of labour has to be added, as well as loss of production and in some cases cleaning/sterilising cost of the installation.

• They have a shelf life
• They have no Value – they are not reliable –

you are not getting what you are paying for

• What is the cost of having an error on a WFI or

Purified water system???

• INSACALTM Master

Reference Standard

• Initial cost is considerably higher compared to a

standard solution; but calibrations are performed faster and require less operator skill. They do not interfere with production and decrease contamination risks. Paperwork is also easier and faster.

• The rig offers very low cost of ownership and is

much more efficient (time) than use of solutions.

Conclusion arguments for INSACAL

februar 2017 40

Industrial roughness Known operator interface Reliable reference equipment Easy calibration on-site Fulfill requirements of pharmaceutical industry Comes with Accredited Calibration incl. documentation

Examples of customers Rigs for INSACAL

februar 2017 41

Quality assurance

februar 2017 42

Quality assurance in the pharma industry is made up by pharmacopeias and metrology! But – Pharmacopeia and metrology do not “goes hand in hand” Then add guidelines and practices like 1:4 rule( Z540.3) and other local practices.

VS.

Quality assurance

Trademarks of the Ph-Eur & USP for WFI Celle constant: must be known within±2%

• According to Ph-Eur -determined at max.

1500µS/cm. No recommendation in USP.

• Neither of them has a position on

uncertainties, and how to deal with them. System calibration – cell & indicator

• Ph-Eur states: accuracy within ±3% ± 0,1µS/cm

februar 2017 43

Quality assurance

februar 2017 44

Conductivity indicator:

• Calibrated by usage of 0,1% resistors.
• Accuracy limit acc. to Ph-Eur & USP is 0,1µS/cm
• Min. display resolution: 0,1µS/cm

Temperature:

• No temperature compensation.
• Accuracy ± 2°C

Quality assurance

februar 2017 45

How abort non WFI applikations? Ph-Eur has an general limit of 5%. But how about temperature compensations for applications like:

• Steam?
• CIP?
• Washing machines?
• Reverse osmosis water?
• ect.

Quality assurance

februar 2017 46

So how do “we” cope? Temperatur compensation:

• Some still uses temeratur compensations, even for WFI.
• Some has switch off temperature compensations in more
• r less all cases.

Limits:

• Most use: ±2% ± 0,1µS/cm for pure water as a system.

Quality assurance

How about the 1:4 ratio?

februar 2017 47

februar 2017 48

Quality assurance

How about the 1:4 ratio? So 1:4 is possible if calibrated correct!

MANGLER OVERSKRIFT

februar 2017 49

The most widely debated issue regarding cell constant calibration is "if I calibrate at 147 uS/cm or 0.0550 uS/cm, how do I know if my sensor is calibrated at 1 uS/cm?". Because of the impact of CO2 on water, there are no standard solutions that can be prepared accurately and precisely in the 1 uS/cm range. Other than trusting the conductivity manufacturer, how do you assure yourself? One valid method that meets all of the technical challenges is to determine the cell constant in pure water and in ASTM solutions. The cell constant is a "geometrical factor" that normalizes the water resistance for the sensor design. The cell constant is a constant over the linear dynamic range of the system. Therefore, if the cell constant is identical at 0.055 u S/cm and 147 uS/cm, then it is the same value at 1 uS/cm. To use a pH analogy, if you calibrate a pH sensor at 4 pH and 7 pH, it may or may not be accurate at 10 pH. However, if you calibrate a pH sensor at 4 pH and 10 pH, it is generally accurate at 7 pH. Although the linearity of the conductivity circuitry is more complex than the circuitry for pH, the same concepts apply.12 A track record of measurement rangability and cell constant consistency has been established for one measuring system. Table 4 shows a series of conductivity sensors that have been repeatedly calibrated in 5 different "solutions" over the last three years : ultra pure water at 3 different temperatures and ASTM D1125-95 solutions "C" and "D". The cell constants were determined by calculating the pure water resistivity based on the temperature which is known to ±0.01°C. The equation relating the temperature (T, °C) and the resistivity (r, MW-cm) is shown below11. r = e (a0 + a1 T1 + a2 T2 + a3 T3 + a4 T4 + a5 T5) where a0 = 4.45656 a1 = -7.33064´10-2 a2 = 5.02097´10-4 a3 = -2.56203´10-6 a4 = 6.43445´10-9 a5 = 1.40405´10-12 The conductivity of the pure water was adjusted from ~0.02 uS/cm to ~0.1 uS/cm (pure water at 15°C to 40°C, respectively). The ASTM solution conductivities were 147.9 and 1409.8 mS/cm, 1.0 mS/cm greater than the values in the Table 3 above, to account for the conductivity of the water. The "constancy" of the cell constants is displayed in a statistical format in Table 5. The last two columns show the relative standard deviation of the cell constant measurement for all 5 "solutions" (3 pure water and 2 ASTM solutions) and for all solutions except solution C. Note that the relative error is less than 0.5% from 0.02 to 150 mS/cm, across nearly 4 orders of

• magnitude. The relative error is less than 0.75% from 0.02 to 1400 mS/cm, across nearly 5 orders of magnitude.

INSACAL CUSTOMERS

februar 2017 50