Teledyne ULTRAFLOW 150, Operation Manual, Page: 78 / 162

ULTRAFLOW 150 GAS FLOW AND TEMPERATURE MONITOR

Transducer Types

Tone Bursts per Transmit

Flow Volume Units
Averaging Properties:

Integration Period Seconds

Integration Periods in Average
Medium Properties:

Minimum Medium Temperature

Maximum Medium Temperature

Minimum Flow Velocity

Maximum Flow Velocity
Calibration Properties:

Span High Volume Set Point

Span Low Volume Set Point

Zero Volume Set Point

Calibration Tolerance

Hour of Auto Calibration

Minute of Auto Calibration

Calibration Interval Hour

Integration Periods in Auto Cycle Span High

Integration Periods in Auto Cycle Span Low

Integration Periods in Auto Cycle Zero
Typically, for maximum accuracy in Continuous Emission Monitoring
Systems (CEMS), the Ultraflow 150 is field calibrated versus a multipoint
pitot tube flow measurement done according to US EPA Reference
Method 40CFR60 Appendix A Methods 1 and 2 (extractive test). While
the ultrasonic time of flight method is among the best methods for flow
measurement in difficult applications, there are situations where its
response is not perfectly linear and does not exactly match the Reference
Method. Flow stratification and cross-flow, i.e., nonzero pitch angles, are
the most common sources of departure from linearity. Also, the
aforementioned Reference Methods 1 and 2 can be significant sources of
error themselves.
The flow velocity channel of the Ultraflow150 supports two approaches to
“linearize” the instrument to compensate for these effects: 1) Look Up
Table (LUT) and Polynomial (POLY). The user must select one of these
two approaches via the “Correlation Curve Source” variable. Both
approaches apply the linearizer curve to the actual flow velocity as derived
directly from the time of flights (“Raw” Actual Velocity) in feet/second
(or meters/second if units are in metric
).
NOTE: It is very important to use actual flow velocity in the indicated
units since these are nonlinear functions. If volume units are used in the
generation process, the linearizer curve will be in a totally different
numerical region and the behavior will not be as anticipated.
8-2