Generally, there are two basic types of averages; a BLOCK AVERAGE and a ROLLING AVERAGE.

BLOCK AVERAGE means the average of all valid emission concentrations, measured over a specified period of time from the beginning of the averaging period to the beginning of the next period.

ROLLING AVERAGE means the average of all valid emission concentrations, measured over a specified period of time or measured for a specified number of Quality Assured Valid (QAV) data points required to represent the period.  Therefore, ROLLING with respect to an average means the calculation of an average by dropping the earliest data value and adding the latest data value for the period over which an average is calculated. A first-in first-out (FIFO) procedure is applied to the data; hence a rolling average value can be computed at each data point interval time.

In CEMView, we consider the Rolling Average to have two basic sub-types.

We refer to the first sub-type as simply a basic Rolling Average.  The specified averaging period is a fixed window of time which could contain quality assured valid operating data points and non-valid data points.

The second sub-type is a QAV Rolling Average.  The specified averaging period may not correspond to a fixed window of time.  Rather, the window of time is variable because the rolling average array is made up of only Quality Assured Valid (QAV) data points.  As a result, the QAV data points in the averaging array that constitute the averaging period, could be non-consecutive in time because non-valid data points are excluded from the array.

“Valid” means quality assured data based on the regulatory compliance rules and procedures for operating the CEM system.

For example, the Canadian federal regulations EPS 1/PG/7 states that:

“valid hour” means an hour during which the generating unit burned fuel and the associated continuous emission monitoring system produced a minimum of 30 minutes of valid data. In the case of a time-shared continuous emission monitoring system, the minimum requirement is two data points per valid hour.

In CEMView, there is no inherent limit to the number of regulatory averages that can be defined in a system.

An average type can be defined for each pollutant or variable being measured.  In CEMView, what is defined is the average period, the base period for the data points contained in the average, the minimum number of valid data points required to compute a valid average, and finally, an exception condition to accommodate a daily calibration where there could be a different minimum number of valid data points required to compute a valid average.  It is also possible in CEMView to define a completely unique average type which could be programmatically configured using the CEMView embedded Microsoft script technologies, VBScript and JScript.

For example:

For the Canadian regulations mentioned above, the “valid hour” for a dedicated analyzer CEM system is defined with the averaging period of one (1) hour.  The data points that make up the average are 1-minute data points; hence up to 60 valid data points could be included in the “‘valid hour”, but if there are less than 30 valid data points, the block average for the hour is not considered to be a ‘valid hour’.For a time-shared system, the data points that make up the average are 15-minute data points; hence up to 4 valid data points could be included in the “valid hour”, but if there are less than 2 valid data points, the block average for the hour is not considered to be a “valid hour”.