Benefits of the swinging door deadband

One benefit of using a swinging door deadband is that it reduces the disk space required to store data. However, because the storage system already provides a good compression ratio, the amount of disk space that is saved by applying this type of deadband for slow-changing tags (changing less than twice in a 15-minute interval) is negligible. For example, a tag that changes 12 times per hour will use 2K bytes of disk space in a 24-hour period. Even if only every fifth point is stored, the savings is only 1.5K bytes per day.

Another benefit of the swinging door deadband is that it captures the data value before the rate change, which is something that a value deadband does not do. If you trend data, the peaks and valleys of the trend curve are more defined to provide a more accurate picture of what is happening in your plant.

Generally, using a swinging door (rate) deadband provides better representation of the value change curve with the same or less number of values stored than regular value or time deadbands for delta storage.

The following graphics compare the trend curves of the same raw data, but with different deadbands applied.

The following graph shows the trend of the actual raw data values:

The following graph shows the trend of the data values with a value deadband applied. Notice how only the first data value that deviates by the deadband from the previous value will be stored, and not any of the values between the starting value and the first deviating value.

The following graph shows the data values that will be stored for both a value deadband and a swinging door deadband. Notice how the swinging door deadband captures data before the deadband change, allowing for a more complete view of the data.

A swinging door deadband is most useful for tags that have a steady increase and decrease in slope, such as a tank level or tank temperature that rises and falls. A swinging door deadband may not be appropriate for "noisy" signals, in which the value of the tag constantly fluctuates around a certain point for long periods of time. Also, the reduction in storage requirements offered by the swinging door deadband may not have much of an impact if you have an application with a small tag count (for example, 500 tags). In this case, it may not be necessary to use a deadband at all.

A swinging door deadband is applicable for analog tags that receive data from the following sources:

• Real-time data values from I/O Servers, MDAS, or HCAL

• Store-and-forward data from a remote IDAS

• Late data from an I/O Server topic that was configured for late data

• A "fast load" CSV import

• Real-time inserts of data using a Transact-SQL statement

A swinging door deadband is not applicable for manual inserts of data through a CSV import of a Transact-SQL statement.

To best visualize the tag that uses swinging door storage, plot a trend using the Historian Client Trend application and set the plot type from to "line" (rather than "step-line").