Rotary Gas Meter K-Factors in Air and Natural Gas

This article examines whether rotary gas meter K-factors obtained in air can be carried over to natural gas service. It draws on Tang's tests of DN50 and DN80 meters at six operating flow points per size. The comparison uses atmospheric air results from Chongqing and Central Sichuan and pressurized natural gas results from Chengdu. Air generally yielded higher K-factors in the reported comparison, but the size and direction of the difference varied with flow and test condition. The results help identify where transfer between conditions needs verification. The study does not isolate the effect of the medium from laboratory and pressure differences.

A rotary gas meter may be tested with air while its intended service medium is natural gas. The K-factor is the recorded pulse count divided by the reference operating volume. In a pulse-based calculation, volume is obtained by dividing the pulse count by the applicable K-factor. A change in that factor therefore matters even when the meter and nominal flow point are the same.

How closely do the reported air and natural gas K-factors agree across operating flow points for the two tested sizes? The practical purpose is to identify where an air-derived factor needs additional verification before use under natural gas conditions. The DN designations identify meter size; the comparison still has to be made at the actual operating flow points. The test cases address the measured K-factor under the listed laboratory conditions. Applying one factor to a different condition is a separate engineering decision that depends on how much error the application can tolerate.

Tang tested DN50 and DN80 rotary meters on air flow standards at Chongqing Gas Field and Central Sichuan Oil and Gas Field and on the Chengdu branch’s natural gas reference system. The rotary meter test plan lists six operating flow points per size, with 120 seconds of acquisition and six readings per point. It also lists four stability test rounds separated by at least one week. The comparison below uses the two Chengdu pressure conditions and the mean results reported for the air laboratories. These design details are documented in Chapters 3 and 5 of the thesis, especially Table 5-2 and §5.3.2.

In the Chongqing Gas Field tests, Tables 4-20 to 4-23 and Figure 4-8 show that the gas pressure was nearly identical in the first two runs. The first run was warmer above 100 m³/h and cooler below 100 m³/h. The K-factor curves for these two runs showed no clear pattern. The third and fourth runs had lower pressures and higher temperatures than the first two, yet their K-factors changed little. In contrast with the turbine meter results, the author found little apparent effect of changes in gas pressure and temperature on the K-factor in this particular set of tests.

The author selected two sets of performance-test data at different pressures from the Chengdu branch and compared them with the mean test data from Chongqing Gas Field and Central Sichuan Oil and Gas Field. Table 5-12 gives the DN80 data and Figure 5-9 compares the curves from the different laboratories.

DN80 K-factor comparison. Redrawn in English from Tang (2018), Table 5-12. Lines connect the reported points. The lower Chengdu pressure is 0.9 MPa in the table and 0.8 MPa in original Figure 5-9.

Figure 5-9 shows a general tendency toward higher K-factors in air than in natural gas. The thesis reports a maximum deviation of 0.96% and a minimum of 0.02% between atmospheric air and natural gas at the condition it labels 0.8 MPa in that passage. For atmospheric air versus natural gas at 1.5 MPa it reports 1.15% and 0.01%, respectively. The next passage compares natural gas at 0.9 MPa and 1.5 MPa, reporting similar trends and a maximum deviation of 0.35%, with the closest points nearly coincident.

For Chongqing Gas Field, Tables 4-14 to 4-17 and Figure 4-6 show very small differences in K-factor. The author nevertheless observed a tendency toward higher K-factors at higher temperatures and judged the temperature effect in this comparison to be larger for the rotary meter than for the turbine meter.

For the Chengdu branch, the thesis cites Tables 4-77 to 4-78 and Figure 4-31. The author interprets the comparison as showing a higher K-factor at 1.5 MPa than at 0.9 MPa, and attributes the trend to pressure. Two Chengdu datasets at different pressures were compared with the mean test results from Chongqing Gas Field and Central Sichuan Oil and Gas Field. Table 5-13 presents the DN50 data and Figure 5-10 shows the comparison curves.

DN50 K-factor comparison. Redrawn in English from Tang (2018), Table 5-13. Lines connect the reported points.

Figure 5-10 shows the general tendency toward higher K-factors in air than in natural gas. For atmospheric air versus natural gas at 0.9 MPa, the thesis reports maximum and minimum deviations of 1.17% and 0.01%. For atmospheric air versus natural gas at 1.5 MPa, it reports 0.78% and 0.01%. The natural gas curves at 0.9 and 1.5 MPa followed similar trends. The largest deviation was 0.39% and the closest points were nearly coincident.

The measurements identify flow points where transferring the air-derived factor deserves closer scrutiny. At 70 m³/h for DN50, both reported air factors exceed both reported natural gas factors. At 40 m³/h, the Chengdu value at 1.5 MPa exceeds the values from both air laboratories. This change in the ordering shows why a single correction inferred from the overall trend would miss important pointwise differences. The comparison changes medium, pressure and reference facility together, so it quantifies differences between test conditions without isolating a causal contribution for each variable. The reported extrema are useful for locating the range of observed differences across the selected comparisons. They are not uncertainty estimates or allowable errors. When reviewing the data for a proposed test, request the flow-point readings and the measurement uncertainty for each condition; an overall average alone can conceal points that need a separate check.

For the tested DN50 and DN80 meters, agreement between air and natural gas K-factors depended on the operating flow point and comparison condition. The tables support checking transfer at the intended flow points and recording the test medium, pressure and reference facility with the reported factor.

A follow-up comparison could test the same meter at matched flow, pressure and temperature while varying the medium on a compatible reference system. Repeating those paired tests across additional meters and reporting the associated uncertainty would help distinguish a repeatable medium effect from bench and specimen variation.

Table 5-12 gives 0.9 MPa for the lower-pressure Chengdu condition. Figure 5-9 and the first comparison paragraph use 0.8 MPa. Both labels are retained where they occur in the source and should be reconciled before using this condition in a specification. The DN50 pressure trend is the author’s interpretation of the selected comparison. The tabulated K-factor at 10 m³/h is 1408.62 m⁻³ at 1.5 MPa and 1409.69 m⁻³ at 0.9 MPa.

Tang Yixin. Comparative Study of Gas Flowmeter Testing with Air and Natural Gas [translated title]. Master’s thesis, Southwest Petroleum University, 2018, §5.3.2, pp. 71–72 (PDF pages 79–80); test plan in Table 5-2, p. 61 (PDF page 69). English charts were independently drawn from Tables 5-12 and 5-13. The source scans of Figures 5-9 and 5-10 are retained in the source verification file.

Editorial cover photo: Sandia mechanical operations photograph, U.S. Department of Energy. U.S. Government Work per source record. Cropped for cover layout with title overlay. The photograph illustrates engineering work.