Analyzer Calibration, Accuracy and Stability Overview

Gas analyzers are manufactured to close tolerances, and calibrated on certified gas blends at the factory.

The published gas accuracy for the analyzer is a representation of the accuracy of the analyzer after the calibration process, when calibrated and checked on the factory gas blends.

Determining the accuracy of the as-delivered gas analyzer at the customer site is complicated by the intrinsic inaccuracy of the gas blend used by the customer for accuracy verification.  As the customer generally checks analyzer gas accuracy by using a 2% relative accuracy certified gas blend, the accuracy of this reference blend must be added to the published accuracy.  Often this leads to apparent inaccuracies which should be avoided when the analyzer is being used to check the viability of the users process.

Improving gas analyzer accuracy – calibration to a local reference gas blend

In practice, this consideration has a relatively simple solution.  Customers generally have a certified reference gas blend replicating their process gas values available to confirm their manufacturing processes, or can readily obtain one from a local gas blender.  If the customer sets the gas analyzer to read this certified reference gas blend to the values on its tag, he can easily remove any blending errors which may exist between the factory cylinders and the customer cylinders.

Once these errors are removed, the analyzer becomes a better tool to validate his manufacturing process and it is far easier to assess the analyzer for accuracy, stability and repeatability.

This White Paper defines a simple, yet structured and generic method which may be used to set-up the analyzer to measure the customer reference gas blend accurately and then assess analyzer accuracy over time and use.

Checking analyzer stability and repeatability

After calibration to the local certified gas blend, the analyzers utility as a test tool is primarily a function of stability and repeatability.  This stability and repeatability of the analyzer can be determined by periodically checking its readings with the same customer certified blend gas used for local calibration.  If this is done in a rigorous manner, it can also be used as a determination of analyzer stability for use.

This White Paper also has a structured method for making this stability and repeatability determination.

Analyzer Calibration and Accuracy Verification Test Procedure

Make sure there is a cylinder of certified calibration gas blend replicating the feedgas mix available with a gas regulator and that it has exit tubing of an inside diameter considerably larger than the sample needle outside diameter.

    1. Turn on the analyzer and operate it in continuous sampling mode (the default startup condition for Bridge Analyzer products).
    2. Let the analyzer warmup according to the manufacturers calibration method instructions (generally 15-30 minutes or so).
    3. Ambient Air Zero Calibrate the analyzer and confirm that it gives stable gas readings on room air. (The Oxygen reading should stabilize at 20.60%, and the other gas readings should be zero.)
    4. Make sure that the calibration gas cylinder is turned on, the regulator is turned completely off (counter-clockwise to zero pressure), and the regulator output gas valve is turned on. In this condition, there should be no gas flow from the calibration gas cylinder regulator, even though the gas cylinder valve is open.
    5. Insert the needle into the open end of the exit tubing, and confirm with the oxygen reading of the gas analyzer that the gas being read is (mostly) room air. (Residual calibration gas may still remain in the exit tubing, but the oxygen reading should gradually move toward ambient 20.6% as the residual gas gets diluted with ambient air.)
    6. Turn the regulator clockwise to slowly start the gas flow from the reference gas blend cylinder.
    7. Observe the oxygen gas reading change in the gas analyzer. It should move from the ambient air level (20.6%) toward the oxygen gas level in the calibration gas.  Gradually adjust the regulator until further opening does not change the gas readings.  This is the point at which the analyzer is now measuring calibration gas instead of room air.
    8. Slowly turn the regulator counter-clockwise to reduce the gas flow from the cylinder until you begin to see signs of air dilution in the analyzer. (Non-Oxygen readings go down – Oxygen readings go toward 20.6% O2)  This is the point at which the analyzer is measuring room air in addition to calibration gas.
    9. Turn the regulator clockwise to adjust the gas flow to be above this transition value to ensure that the analyzer is seeing only the calibration gas blend as in step 7 above. (As a substitute for this method, a rotometer or other flow rate measuring device may be used on the exit gas stream, and the regulator adjusted to produce 500 ml/min gas flow.)
    10. Let the gas readings stabilize to their final value.
    11. Note the readings, and adjust them using the analyzer calibration capability until they match the cylinder certified ‘Tag Values’.
    12. Once the readings match, save the calibrated gas values and go out of calibration mode, but remain in continuous sample mode.
    13. Remove the needle from the calibration gas tube end and let it sample room air.
    14. Allow the readings to stabilize on room air and note their accuracy. (Ambient Air Zero if necessary – but allow the readings to stabilize.)
    15. Go into the test sample mode and run a test on room air. (At the end of the test the analyzer sample pump should be off.)
    16. Put the needle back in the tube from the known gas mix. Note that the gas should still be flowing – but the analyzer should not respond because the sample pump is not on.
    17. Sample the gas three times in sequence and note the readings. (The readings should become more and more stable and closer to the cylinder tag value as you test.  This test sequence allows you to observe the transit time and response time effects of the analyzer.  You may have to adjust the sample time or number of initial samples to achieve the desired accuracy.)
    18. Confirm that the final analyzer test results match the ‘Tag Value’ of the gas mix. If they do not within your test tolerance– you may have to recalibrate the analyzer as above or adjust the test time or number of samples to obtain the desired repeatability.)
    19. Turn off the gas flow from the reference gas cylinder.
    20. This validated / calibrated analyzer may then be used to test product on the line.
    21. Once line testing is done, the analyzer may be used to repeat testing of the customer gas blend to confirm that it was maintaining accuracy during the line testing.

Carrying the local calibration forward

Good practice is to check the analyzer using the customer gas blend to confirm its accuracy before line testing – then perform the line testing required.  Once a history of stable accuracy is established, the pre-test accuracy validation step may be deleted from the testing process.

Analyzer Stability/Repeatability Test Procedures

The intention of these tests are to determine the short and long term stability of the analyzer when used in practice.  In essence, if the analyzer can be made to measure calibration gas correctly, its ability to maintain that accuracy over time is measured to assess its utility in the application.

Short Term Repeatability Test

    1. Power up and validate analyzer accuracy on a certified reference gas mix. (As above – let the analyzer fully warmup to the manufacturer calibration stability level.)
    2. Power-down the analyzer and store it in the normal location for at least an hour. (Let it equilibrate to the normal storage environment.)
    3. Retrieve the analyzer, power it up and go through the initialization routines recommended by the manufacturer.
    4. Test the same certified reference gas mix used in step 1 at 5, 15, 30 minutes from power on. (Do not power down between tests – but service all ambient air calibration per the manufacturer’s instructions.)
    5. Power down the analyzer after the 30 minute test and store again for 30 minutes.
    6. Repeat the above two operations (Test and Rest) for a total of 6 cycles. (This is 6 hours of testing – and should be performed contiguously during the same shift.)

Long Term Repeatability Test

  1. Make sure the analyzer is accurate at the start of the test using the certified reference gas mix as in Step 1 above.
  2. Measure the certified reference gas mix gas using a repeatable routine (Same time of the day, time from power on, time from Zero, etc) once or twice a day for 5 consecutive days.