.png){kind=logo}
.png){kind=logo}
Higher parameter value issue for RO
Background: We were facing issues in the average values in the RO system.
Example- Average Conductivity values, TDS values and False alerts when the RO is out of line.
Description: There was an issue of higher values of conductivity, TDS and negative values of Alerts when the RO was out of line in Coca-Cola Al Ain Plant. This could be due to the reason of Forward Osmosis when the system is out of line.
We have observed trends of increasing Conductivity when the RO is out of line. The below graph from our system shows the increase in Conductivity values when the RO flow rate is 0.
The system was taking the complete average (even when the system was out of line). This was also creating an issue with Alerts as when the RO system was out of line alerts of high TDS, conductivity was getting triggered. At times, the average product conductivity was even higher than the feed conductivity.
Actions taken:
-
The data was downloaded in excel and the average values were checked while considering both the cases.
-
The system was taking simple average of the values that were getting in the system and there were huge differences between the actual average and the average with outliers. Eg- For 8th July- Product Conductivity- 307 (system average), 80 (actual average).
-
We took all the product conductivity & TDS parameters and created KPIs using filters with suitable ranges.
-
Once the ranges were filtered, those KPIs were used in place of actual parameters and the average issues were resolved.
-
The same was done for alerts. Those KPIs were used to define the alerts and Sensor noise filter was checked.
New values of Product Conductivity:
The new average conductivity values are well in the range and could be used in use cases.
