How Continuous Testing Further Reduces Gas USM Uncertainty

With the significant growth in the usage of ultrasonic meters, and industry’s drive for improvements, ongoing testing and development has resulted in enhanced performance of these meters. One such test was presented in 2009 at the CEESI USM Conference.

It involved testing two brands of 12-inch USMs with a series of elbows and tees upstream of the meter, and comparing the meter’s accuracy to a baseline consisting of straight upstream piping. A second test involving three elbows in and out of plane was also included in this test program. In addition to the upstream piping effects, the meter was not only tested in the upright (normal) position, but rotated 90° and rotated 180° (upside down) to see if the results were different.

The meter run assembly consisted of 12-inch Schedule 40 pipe configured as follows: 10D + CPA + 10D + Meter + 5D. The meter was a FLOWSIC600 2plex 4+1 meter as shown in Figure 1. It consists of a conventional 4-path fiscal chordal meter and an additional independent single path meter incorporated into the same body. Each meter has its own independent electronics.

Figure 1: FLOWSIC600 2Plex 4+1 Meter design

For this article only the 4–path meter data is presented. The single path meter is typically used to identify added measurement uncertainty due to field conditions such as contamination. Since the testing was conducted at a calibration facility using clean, dry gas, the single-path data was not analyzed.

To establish a baseline, the meter run assembly was installed with straight piping (no elbows or tees) upstream of the assembly. The test data was collected at three flow rates (pipe velocities), approximately 20, 40 and 60 fps. After the baseline testing, the assembly was installed in a piping configuration consisting of three elbows and one tee upstream (Figure 2).One tee was used as this is typical of many customer installations. It permits internal inspection and cleaning of the meter run without removal from the pipeline. In addition to the meter being tested at the three flow rates, the rotational orientation of the meter was also added as a test variable. The entire meter package was rotated for these tests, not just the meter.

Since the data for the three elbows in and out of plane showed less than 0.05% shift, this presentation focuses on data from the more complex piping that included three elbows and one tee. The results showed a meter shift of +0.12% for the meter in its normal orientation, +0.13% with the meter rotated 90° and +0.22% with the meter package rotated 180° (upside down). See Figure 3.

Figure 3: Original CEESI test results

Figure 4: Meter Error - Before and After for Coefficient
Correction

Although the results were well within the AGA 9 required maximum of ±0.3% measurement uncertainty due to the installation effects, customer interest in the meter orientation effect led to further investigation at the manufacturer’s atmospheric test facility in Dresden, Germany. Although a different 12-inch meter was utilized, the upstream piping configuration and testing results were duplicated at atmospheric pressure.