|
Cross-sensitivity of electrochemical sensors and how to handle it
When the first electrochemical sensors were introduced years ago, cross-sensitivity was a serious handicap for the sensors as well as for the flue gas analysers where they were used. Madur electronics had the same challenging task to solve as everybody else.
There are several ways to handle the problem now.
The first idea was eliminating the influencing agents. That was and still is done by means of a NOx/SOx filter to eliminate their respective interferences on the CO cell. At first, the supplier of the sensors equipped the sensors directly with the respective filters. The problem that occurred when using filters was that they used to be used-up long before the sensors were consumed.
That is a particular problem when the filter is incorporated in the sensors. Since they can not be replaced, the whole sensor is useless when the filter is gone.
The need to eliminate the H2 interferences on CO sensors was initiated by the TÜV in Germany. This is partly done using filters, but there are also sensors where the hydrogen concentration is measured independently of the CO and the compensation carried out electronically. These are then manufactured as 4-electrode sensors with a second sensing electrode. This is the preferred method and madur electronics always uses this approach when there is the risk of high levels of hydrogen.
The other, and much more reliable, method is to compensate electronically for these effects.
When calibrating the instrument, every influence detected is stored and goes into the calculation for that gas. Hence, the signal produced on each sensor by a gas it is not designed to measure, but which it reacts to, is cancelled out in use. This gives instruments equipped in this way an outstanding performance in comparison to most other instruments, although it is often difficult to convince people that the new instrument is correct when they have been used to incorrect readings for years! We first applied this method to the GA-60, the mainstay of our production for many years. The calculation has since been applied to all the multi-gas analysers we produce.
The only real and unavoidable problem with this method of calculating and then compensating for the effects of other gases present is that the cross-sensitivities can, of course, only be calculated for the gases measured directly i.e. no cross-sensitivities can be calculated between gas components the respective instrument is not equipped to measure directly. This means that some form of filtration is still necessary in many cases.
Particularly when testing for HCl, the cross-sensitivity can become a real problem due to the large response caused by SO2. The HCl sensor has a 35% cross-sensitivity to SO2 and a maximum operating concentration of 100 ppm. This means that a concentration of SO2 in excess of about 250 ppm will cause the HCl sensor to go over-range. This is a simple sensor problem that can not be cured by any type of calculation or changes to the software. When the sensor exceeds the maximum range the response becomes unpredictable. The signal can climb to a high value or drop off without warning, thus invalidating the calculations carried out by the instrument. The only real way to avoid this is to know of the problem and to use filters to remove the SO2 when testing for HCl.
At lower concentrations of SO2 and, to a lesser extent, NO2, there is no problem and the results shown by the analyser are trustworthy. Most modern analysers will be equipped with an over-range warning.
Certain sensors are prone to poisoning by high concentrations of the target gas. This does not generally cause permanent damage, but may In some cases, particularly with CO sensors, lead to a very prolonged return to zero after measurement. Here, one solution is the use of a purging pump to isolate the sensor from high concentrations to prevent these effects. With this method, the other sensors are still in operation, only the overloaded CO sensor is out of operation. Our modern analysers, such as the Ga-20plus, GA-21plus and the GA-40(T)plus are all fitted with purge pumps either optionally or as standard to combat this difficulty. The sensor will be purged for one minute and will then "try" the gas again. If the concentration is acceptable, the sensor will resume measurement, otherwise it will repeat the purging. Alternative methods, such as dilution of the sample gas have always proved to be very inaccurate and hence we are not prepared to use them.
|
|
