6.4 Reliability and Maintenancebility of ZrO2 sensor
Continuous on-board measurements so far have proven ZrO2-type NOx sensor's high reliability except for a few trouble cases occurred on-board. In this section, countermeasures for such troubles will be reported.
6.4.1 Sensor Trouble and Countermeasure in IKOMASAN
.1 Summary
Unexpected readings have been observed in weekly report of data starting at around December, 2001. Once it seemed to be recovered, however, it ended up as over the reading range in January, 2002. It can be inferred that such unexpected readings were due to the failure of ZrO2 in the sensor caused by exposure to water.
.2 Outbreak of Trouble
The following were reported from IKOMASAN on January 30, 2002.
Unexpected readings were observed on January 29, 2002, after ship left Kharg Island for service:
NOx reading: 5000 ppm
Corrected NOx reading: 5000ppm
O2 reading: 38.02 vol %
(the NOx and corrected NOx are far above upper limit)
.3 Field Investigation
At Kawasaki Harbor, on February 17,2002
- Receiver display Error "E-09; low heater electric current"
- Probe Normal (without problem such as clog)
- Power supply line of sensor heater No circuit (should be 3 A in normal condition)
- Sensor Normal operation is confirmed after being replaced.
Figure 6.4.1.1 shows damaged sensor that has been taken out. It can be inferred that deposit of mist or the kind caused the tip of the sensor to turn white. This portion is the stainless protection cover, and the deposit has not formed on the surface of ZrO2 itself. Figure 6.4.1.2 is a sensor holder, which is normally covered by fiber filter. It can be inferred that water has been entered through the crack.
Figure 6.4.1.1 Detached Damaged Sensor
Figure 6.4.1.2 Sensor Holder
.4 Investigation of Sensor
Tortoise-pattern crack was confirmed on the removed sensor element. Figure 6.4.1.3 shows the sensor tip.
.5 Presumption of Causes
From the following reasons, it can be inferred that the sensor's exposure to water has caused the problem.
Figure 6.4.1.3 Photo of the ZrO2 sensor tip
- Tortoise-pattern crack was confirmed on the sensor element. This type of crack on ceramics is due to exposure to water.
- It is highly possible that water has entered through the probe and dampened the ZrO2 element of the sensor in the process of cleaning the stack using water. Cleaning was done in Kharg Island immediately before the failure of the sensor.
.6 Countermeasure
Note below when cleaning the stack using water.
- Be very careful to prevent water enter from the sample gas inlet or ejector air outlet at the tip of probe.
- Turn the power off and let the sensor cool off prior to cleaning, to prevent the risk of damaging the sensor by water. Do not turn the power on until everything dries out completely.
In IKOMASAN, the tip of probe is covered when cleaning the economizer to avoid water entry. This has effectively prevented the same problem.
6.4.2 Countermeasure to the Reading Problem in ANTARES
ANTARES equips an engine which belongs to the largest category in sizes and flow rate. Increase of flow rate of ejector air for acquisition of adequate amount of exhaust sample gas caused ejector temperature drop, and lead to the accumulation of dust and particles in the exhaust gas at the ejector surrounding. Measurement was interrupted due to this condition. Heating the ejector showed some improvements, but was not a definite solution. At the end, stable measurement was achieved by following countermeasures, making ANTARES' measurement condition equivalent to that of other two ships.
|