Oxygen measurement
Our oxygen measurement systems included as its centerpiece a dynamic oxygen sensor based on zirconium dioxide and were able to measure failsafe if necessary.
How does the dynamic oxygen sensor work?
The active layer of the dynamic oxygen sensor is composed of porous platinum and the oxygen ion conductor YSZ, that means yttrium (Y) stabilized (S) zirconium dioxide (Z).
Platinum catalyzes the reduction of oxygen to oxygen anions that diffuse through the zirconium dioxide. Thereafter, the platinum catalyzes the oxidation of oxygen anions to oxygen. During the reduction, four electrons are required, four electrons are released during oxidation.
At this so-called voltage probe electricity flows only until the same oxygen concentration prevails on both sides. An electricity probe was integrated to make the measuring process dynamic. This results in a hermetically sealed chamber, into which the current probe alternately pumps or extracts oxygen ions. An oxygen concentration gradient is artificially generated, the diffusion process and thus the current flow at the voltage probe are retained
The measuring cycle starts at an EMF of 40 mV. The pump current sets in and evacuated the chamber until an EMF is reached of 90 mV. After that, the pumping current is reversed, oxygen is transported into the chamber, until a EMF prevails of 40 mV. The time required for this is dependent on the oxygen content of the gas mixture to be measured.
How does the oxygen measurement system detect errors?
The oxygen measuring system outputs two measuring signals via two different channels:
- at channel K1, the measured value is available as an analog signal (0-10 V or 4-20 mA)
- the bidirectional channel K2 as a digital pulse lengths modulated alternating signal (low / high phase: 0/5 V). The length of the low phase is the measure for the oxygen concentration.
The oxygen measuring system works without error if the analogue signal from channel K1 matches the signal of digital channel K2 (maximum deviation 4%). In addition, the low phase of channel K2 must be between min. 0.05 s and max. 1.5 s long, which corresponds to an oxygen concentration of 0.1 and 25 vol%.
If the signal falls from a time window of 0.05 s to 4 s or the switching signal changes to a constant error signal of 5 V, the measured value is outside the measuring range. If the hardware is defective, the output signal remains constant at 0 V.
As the measuring method is dynamic, the proper functioning of the oxygen measuring system can be checked at any time, even during operation, and is ideally cyclical.
For this purpose 24 V are applied to a separate test channel and the sensor current is thus lowered by 20% from the outside. The measuring system gets simulated a lower oxygen concentration than actually exists.
As a result, the measurement signals must drop equally at the channels K1 and K2, which means that the measuring system must correctly calculate the virtual oxygen concentration.
Here is a bandwidth of 4% permissible, that means during the self-test, the measured value must be between 0.76 and 0.84 times of the previous measured value.
Rule of thumb: The measuring signals must equally fall by at least 20%.
This test arrangement makes it possible for the first time to detect faults not only on the hardware of the measuring system but also on the sensor itself, so on the zirconium dioxide chamber!
External monitoring unit of the user
An external device which is connected downstream by the user must take the evaluation of the measuring signals as well as the execution and the monitoring of the cyclic self-test.
The reaction to the error message occurs accordingly to the specifications of the user and is also managed by its external monitoring unit.
Therefore, it has to fit certain requirements:
- The unit must be fail-safe, if the operations listed below are to be executed without errors, the input signals must be read in without errors and the output signals must be out without errors, too.
- Within the permissible tolerance time for the application, the measured values of the channels K1 and K2 must be permanently compared.
- The time of the output signal K2 must be constantly checked for plausibility. Static signals are to be regarded as internal errors.
- At cyclic intervals, a self-test is to be triggered and its effect on the measuring signals to be recorded and evaluated. The time interval between two test cycles must not exceed a certain value.
- An error message must cause the process to be transferred to a safe state.
When the oxygen measuring system is working failsafe?
The oxygen measuring system is working failsafe if:
- the analog and digital output signals match,
- the measurement signal of channel K2 is within a defined time window and is not static, and
- the self-test is cyclic and correct.
This provides for the first time a fail-safe oxygen measuring system, which during operation monitors its entire system itself and also manages with only one oxygen sensor.
Partial pressure is not equal to volume%
The dynamic oxygen sensor determines the oxygen concentration directly in the gas mixture to be measured. It determines the absolute oxygen content, that means the oxygen partial pressure and not the relative (vol.%).
Atmospheric air is composed of the gases nitrogen, oxygen, carbon dioxide, argon and water vapor. The total pressure of the air results from the sum of the partial pressures.
Dalton's law: P(tot)= P(N2) + P(O2) + P(CO2) + P(Ar) + P(H2O)
The oxygen partial pressure thus changes depending on (1) atmospheric air pressure P (tot) and (2) the moisture content of the air P (H2O).
With increasing air pressure and constant humidity of the oxygen increases the partial pressure, the volume% of oxygen remain the same! (The absolute oxygen content increases, but the relative content remains unchanged.)
At a constant air pressure and modified humidity contrast, oxygen partial pressure and volume % of oxygen are changing!
By the way: the more humid the air, the less oxygen is contained in it. This is especially noticeable at high temperatures.
Harmful substances
As the oxygen sensor contains zirconium dioxide and platinum, the following substances can destroy him:
- Heavy metals
- sulfur compounds
- silicone vapors
- fluorine
- NH3 (from1000 ppm)
- Halogenated hydrocarbons (above100 ppm)
- Phosphate ester
- Chlorine
- SF6
- Carbon
- Salts
- long time in a reducing atmosphere
Dust, vibrations, contamination, moisture, oils, greases, boiler cleaners, heavy fuel oil, pyrolysis gases and silicon oxide (when silicones burn in the boiler) shorten the life of the oxygen sensor.
The list does not guarantee completeness.