Select Page

Suppose there is a fault that you need to detect which occurs periodically but most often clears itself within a few minutes without affecting the process. You want an Alarm only if the fault doesn’t clear after 30 minutes and ignore the rest.

In this example we wish to monitor GL_CLIENT_FAIL  which is a Digital Output. This becomes True if any of the GRIDlink software processes stop.   It rarely happens but if so it usually clears in 3 to 4 minutes.


Since tags report on change of state and the Alarm monitors that change, an Alarm will occur immediately if there is no delay configured. You will see all Alarms.  Once you set a delay, the logic requires a second data point to calculate if there has been a change or more importantly no change within the delay period. The tag normally reports on exception so it will not send a data point until it changes state so there will be no second data point to show the fault is still in effect and therefore no alarm.

You must first make sure the GRIDlink is reporting often enough so the Alarm monitor can receive comparative datapoints during the 30 minute parameter. This is done by editing the tag GL_CLIENT_FAIL property: Max Delay. Change the value from 0 to 600,000 which is 10 minutes in milliseconds. This tag will be updated every 10 minutes regardless if it changes state.



Edit the GL_CLIENT_FAIL Alarm property Delay from 0 to 1,800,000 which is 30 minutes in milliseconds.


You are now finished. All faults less than 30 minutes will be ignored and only the important ones will be received.