Air vs. Natural Gas Calibration: What DN50 and DN80 Turbine Meter K-Factors Show

A master’s thesis compared the calibration K-factors of DN50 and DN80 turbine gas meters at six operating flow points per size. The study used two atmospheric-air facilities and two natural-gas pressure settings at a third facility. The published tables reveal that the lowest flow point deserves separate attention. The DN80 air values are higher than the gas values at 20 m³/h, while the DN50 air values are lower at 10 m³/h. Recalculation from the printed tables also gives larger low-end differences than the maximum air-to-gas deviations quoted in the narrative. This article reproduces the tables, identifies the calculation basis and explains how to interpret the comparison.

A turbine gas meter with pulse output needs a K-factor to convert pulses into operating volume. A laboratory may calibrate a meter in air while the intended application uses pressurized natural gas. A calibration report therefore needs to show how the factor behaves across the flow range, particularly at the lowest point intended for use. A visually close fit at higher flows can conceal a difference at one end of the range, and a single average cannot show where that difference occurs.

How closely do air and natural-gas K-factors agree at the six tested flow points for each meter size, and does the direction of the difference stay the same near the low end? The comparison also checks whether the narrative’s reported deviation ranges can describe every printed table entry. The scope is limited to the specimens and facilities reported by Tang (2018).

Chongqing Gas Field and Central Sichuan Oil and Gas Field tested in air at local atmospheric pressure. The result tables label both air conditions as 0.1 MPa. The Chengdu branch tested in natural gas at 0.9 and 1.5 MPa. The thesis's turbine test plan lists six readings per flow point, with a 60 s acquisition period. It also specifies four stability rounds separated by at least one week. Tables 5-10 and 5-11 compare selected Chengdu results at both pressures with the mean results from each air facility. The air facilities and Chengdu used different reference installations, so the between-site comparison changes the facility as well as the gas and pressure.

Kᵢⱼ is the factor for run j at point i, Nᵢⱼ is the counted pulse total, Vᵢⱼ is the reference operating volume, and n is the number of repeat readings. Kᵢ is the mean factor at point i. The tables express K in m⁻³. For the editorial check below, the signed difference is (K_air − K_gas) / K_gas × 100%.

The Chongqing runs showed nearly overlapping curves when air temperature and pressure were similar. The thesis observed higher K-factors in warmer runs at similar air pressure. In Table 5-10, the four series lie relatively close together over most of the tested range. At 20 m³/h, both air results rise above both natural-gas results. For example, Central Sichuan reported 2323.94 m⁻³ and Chengdu reported 2285.04 m⁻³ at 0.9 MPa. At 50 m³/h, the two Chengdu pressure values reverse their ordering relative to 20 m³/h. That pointwise reversal limits the use of a single pressure trend to explain the table.

DN80 turbine meter calibration K-factors. Redrawn in English from Tang (2018), Table 5-10; compare original Figure 5-7. Straight lines connect reported points.

In the first two Chongqing runs, air pressures were similar and the temperature bands were 16.46–16.56 °C and 16.24–16.47 °C, respectively. The thesis reports a higher K-factor curve in the first run. In Chengdu, the 1.5 MPa natural-gas run also had higher temperature than the 0.9 MPa run, so the observations do not separate the two influences. At 20 m³/h both air values exceed both natural-gas values. At 10 m³/h both air values are lower. Chongqing reported 7704.33 m⁻³ against Chengdu’s 7869.61 m⁻³ at 1.5 MPa. That reversal occurs within the tested range of a single nominal meter size.

DN50 turbine meter calibration K-factors. Redrawn in English from Tang (2018), Table 5-11; compare original Figure 5-8. Straight lines connect reported points.

At the minimum measured flow, a named air-to-gas pair gives about +1.70% for DN80 and −2.10% for DN50 using the natural-gas reading as the denominator. Even when the two air facilities are averaged first, those same comparisons give +1.63% and −1.89%, respectively. These calculations use the published table values rounded to two decimals. They show the sensitivity of the interpretation to flow point and comparison basis. A laboratory reviewing an air calibration for a natural-gas application should compare corresponding operating flows, state the reference used for any percentage difference and keep the low-flow results visible in the report. The two test sizes exhibit opposite low-end directions here, so pooling them would obscure a feature that matters for verification. A separate check of the Chengdu data illustrates why a pressure comparison should also be made point by point. At 70 m³/h for DN80, the factor at 1.5 MPa was 2286.49 m⁻³ and the factor at 0.9 MPa was 2279.21 m⁻³, about 0.32% apart relative to the lower-pressure reading. For DN50 at 40 m³/h, the corresponding figures were 7829.59 and 7777.02 m⁻³, about 0.68% apart on the same basis. In both cases the percentage describes a difference between measured factors. It is not an estimate of uncertainty or a tolerance for service.

The results support a flow-point-specific review of calibration transfer for the tested DN50 and DN80 turbine meters. Most of the curves are comparatively close, but the lowest measured flow yields clear differences in opposite directions for the two sizes. Any decision to transfer a K-factor should use the relevant flow, medium, pressure and facility information from the actual calibration. These results describe the reported samples and test conditions.

A follow-up comparison could repeat measurements near each meter’s minimum service flow and record pressure, temperature and uncertainty alongside the pulse data. Matching pressure and temperature while changing only the test gas, where the reference system permits, would help separate a medium effect from the effects of the facility and operating condition. Testing additional meters would show whether the observed low-end directions recur beyond these specimens.

The thesis narrative gives maximum air-to-gas deviations of 0.19% and 0.16% for DN80, and 0.49% and 0.43% for DN50, in the 0.9 and 1.5 MPa comparisons respectively. It gives 0.32% and 0.67% for the two natural-gas pressures for DN80 and DN50 respectively. The selected section does not specify an exclusion or denominator that reconciles its air-to-gas maxima with the lowest-flow entries. This article retains the author's figures as reported and presents the independent table calculations separately.

Tang Yixin (2018). Comparative Study of Gas Flowmeter Testing with Air and Natural Gas [translated title]. Engineering master’s thesis, Southwest Petroleum University. §5.3.1, pp. 69–70 (PDF pages 77–78), Tables 5-10 and 5-11 and original Figures 5-7 and 5-8. Test plan: Table 5-1, p. 60. Equations: §3.3, p. 18. The English charts use the tabulated values and are independently redrawn.

Cover: Betty Carrell, U.S. Department of Energy archival photograph. Public-domain U.S. government work per the file record. Editorial crop and title overlay. Photo record