No Trends At Startup

Once you've been in production for about one billing period, there should be enough system-generated trend data (depending on the estimation algorithm and how the trend areas/classes are configured) to generate statistically accurate consumption estimates. However, before this time, you will have little or no trend data and therefore the system will not be able to accurately estimate consumption (and therefore high / low validation will not operate).

If you can live with this, then you don't have to set up trends because the system will automatically build them as meter reads are recorded in the system. However, if you opt for this approach, you must turn off high / low validation on your uploaded register reads during the first few months of production. If you fail to do this, every register read will fail high / low validation and reads that fail high / low checks are not trended. To turn off high / low validation populate the uploaded register read's REVIEW_​HILO_​SW with a value of N.

If you can't live with this, then you'll have to populate the trend data manually. To do this, create trend records for every combination of trend area, trend class, unit of measure and time of use code. For each trend, enter the following:

Thread Nbr Enter 1.

Read Date Set this to the day preceding your production date.

Total Service Quantity For peak units of measure (e.g., kW, kVar), this will be the average peak amount of consumption associated with registers belonging to this trend. For non-peak units of measure (e.g., CCF, kWh), this will be the average daily consumption associated with registers belonging to this trend.

Number of Units Enter 1.

Number of Reads Enter 1.

Note: You do not have to change the trend classes to indicate only one read is needed. This is because, if the system cannot find enough trends to satisfy the number of reads required by the trend class, it will use as many as exist. For example, consider a trend class that needs 10,000 reads to form a statistically significant sample for estimation, but at estimation time there is only one trend and this trend contains the results from only one read. In this situation, the system will use the consumption information on the single trend to calculate estimated consumption.
Note: Make sure that your manually entered trend information is accurate and well tested. The quality of the trend information impacts your estimation results (depending on your estimation algorithm) and this in turn impacts the bills you send your customers.