![]()
|
Data Spikes and Corrupt Data
Occasionally there may be extremely large increases or decreases in the logged data coming from NiagaraAX, which in turn shows up on
shortBrandName reports as large spikes. The root source of this problem could be an incorrectly configured meter or unreadable incoming data (problematic pulse counts). If there are occasional large negative spikes occurring after nearly equal time periods, there may be a problem with the meter's energy capacity. For example, if a delta log is taking consumption values at 15-minute intervals and the meter resets its energy count at a certain limit, say 5000 pulses or kilowatt-hours, it will go back to zero during an interval, resulting in a large negative delta value. To illustrate, assume a consumption meter has a total reading of 4990 kWh with a reset function at 5000kWh. If the meter increases by more than 10 kWh, there will be a large negative value recorded to the 15-minute delta log (delta logs take the difference in value from one 15-minute period to the next). Assume there is 50 kWh used in that 15-minute interval. Instead of a value of 50 in the delta log, there will be a value of around -4950 kWh (10 kWh before the reset, a change of -5000 kWh after the reset, and then 40 is consumed after the reset to yield -4950kWh). Again, this problem arises only if the meter is programmed to reset or clear after a certain threshold has been reached, and is usually meter-dependent.Regardless of the underlying reason, the
shortBrandName reports can be fixed using one of the following methods:
- Edit history data
This method involves actually editing the bad data values in the logs using the History Editor or Database Maintenance views. These views are described in the NiagaraAX User Guide.
- Using Data Cleansing
This method involves leaving the bad data in the logs, but masking them in the
shortBrandName reports so that they do not show up. If you do not want to change any of the records and only want to have theshortBrandName reports not display the bad values, the Data Cleansing function can be used. The only things that change with this feature are the reports themselves, not the actual data values. Data Cleansing is described in “About Data Cleansing,” page 3-64.
|
|
|
|