Diaphragm Gas Meters: What Changes When the Reference Volume Is Condition-Corrected
Abstract
A diaphragm meter and a laboratory reference can indicate different volumes even when the same quantity of gas passes through both. Wang's 2021 technical article examines this problem through temperature and pressure correction during air-based verification. This analysis retains the article's measurement model, distinguishes the indicated reference volume from its value at the meter conditions, and checks the scale of a one-kelvin temperature difference. It also identifies ambiguous printed notation and inconsistent acceptance intervals that prevent the article from serving as a stand-alone compliance procedure. The practical outcome is a clearer comparison basis: record both measurement locations, correct their gas states, and assess the resulting indication error against the applicable documented procedure.
Background and research question
Gas volume depends on the conditions at the measurement location. In a diaphragm-meter test, the reference device and the meter inlet can have different temperatures and pressures. Comparing their displayed volumes directly therefore combines two effects: the meter's indication error and a difference in the physical volume occupied by the gas. The question is how the reference reading should enter the calculation when those conditions differ.
Wang discusses bell-prover and reference-meter methods, with particular attention to laboratory temperature management [1]. This is a review of that published method and its worked example. It does not report a new meter test or establish the performance of a current commercial product. For buyers commissioning a test bench, the relevant deliverable is a traceable comparison at defined conditions.
Verification methods and measurement model
The article first describes environmental control and temperature equilibration, then gives an indication-error equation. Its printed Eq. (1) has ambiguous bracket placement. With the intended difference expressed explicitly, the calculation is:
A = [(Vm − Vref) / Vref] × 100% (1, notation clarified)
Vm is the meter's indicated volume. Vref is the reference volume expressed at the meter-inlet conditions. Both volumes must use the same unit, such as litres. A percentage result refers to their relative difference; it is not a volume in litres.
The article's condition correction is retained as:
Vref = Vs × (Psa × Tma) / (Pma × Tsa) (2)
Vs is the reference-device reading. Psa and Pma denote the gas pressures at the reference device and meter inlet; Tsa and Tma denote the corresponding thermodynamic temperatures. Temperatures enter in kelvin. The pressure ratio requires absolute pressures, so any gauge readings need an appropriate atmospheric-pressure conversion before use.
This equation transfers the reference reading to the meter location. It is distinct from converting a customer's consumption to a contractual base condition. Keeping those operations separate helps prevent a correct temperature calculation from being applied to the wrong volume.
Temperature effects in the published example
The source discusses a 200 L bell-prover installation and illustrates a 100 L reading near 20°C. It associates a temperature difference of approximately one kelvin with a volume effect of roughly 0.33 L, or around 0.33–0.34% of the example volume. These are the author's example conditions, not a measured error distribution for all diaphragm meters.
A transparent independent check shows the scale. Taking equal absolute pressures, Tsa = 293.15 K, Tma = 294.15 K and Vs = 100 L gives Vref = 100.3411 L. If the meter also indicates 100 L, Eq. (1) gives approximately −0.340%. Reversing which location is warmer reverses the direction of the correction. This check is an editorial calculation from Eq. (2), not an additional experiment.
Quantity | Illustrative value | Role |
Reference reading Vs | 100 L | Uncorrected reference volume |
Reference temperature Tsa | 293.15 K | 20°C |
Meter-inlet temperature Tma | 294.15 K | 21°C |
Absolute-pressure ratio Psa/Pma | 1 | Assumed equal for this example |
Corrected Vref | 100.3411 L | Independently calculated |
A when Vm = 100 L | about −0.340% | Independently calculated |
The example explains why room temperature alone is an incomplete record. Gas at the reference and at the meter inlet may not follow the room sensor immediately. A small percentage effect can matter when the available margin to an acceptance boundary is similarly small.
Proposed solutions and their limits
The source next considers simplifying routine calculations when the laboratory is sufficiently stable. Its underlying idea is to avoid unnecessary work while retaining a reliable decision near the acceptance boundary. That idea requires evidence that any omitted state correction is small enough for the particular test and uncertainty budget.
The printed intervals in its solutions section do not form a consistent partition. One line labels 0.12% < |A| ≤ 0.18% as unacceptable, another labels |A| ≤ 0.12% as acceptable, and a subsequent line refers to a range between those cases. Those statements also need reconciliation with the other error figures discussed in the paper. They are documented here as a source inconsistency, and are not adopted as meter pass/fail rules.
A laboratory can instead make its decision basis reviewable by retaining the measured conditions, raw volumes, applied correction and resulting error together. Where a simplified calculation is proposed, compare it with the corrected calculation over the actual working range. The permitted simplification and acceptance decision must come from the applicable controlled procedure.
Conclusions
Condition correction changes the reference volume used in the denominator of the indication-error calculation. In the published example, a one-kelvin difference near room temperature can produce an effect of about one third of one percent. Thermal control and a recorded comparison basis are therefore substantive parts of the test.
Limitations and outlook
The article does not provide a full uncertainty budget, repeated raw test series or a validated acceptance algorithm. Future laboratory work should compare corrected and simplified results using traceable sensors and documented equilibration conditions. For procurement, request that evidence together with the actual test range and reporting format.
Notes
The original article contains no numbered body figures or data tables. The table above is an editorial calculation and is labelled accordingly. The intended parentheses in Eq. (1) are clarified; the source's inconsistent numerical intervals remain disclosed. Historical procedure references in [1] have not been treated as proof of current regulatory compliance.
References and Image Sources
[1] Wang Guojian. Temperature effects on diaphragm gas meter verification and analysis of solutions [original Chinese title: 温度对膜式燃气表检定的影响及解决方法分析]. Light Industry Standards and Quality, 2021(2):75–76. DOI printed in the source:
https://doi.org/10.19541/j.cnki.issn1004-4108.2021.02.016 . Full text consulted: both pages; Eqs. (1)–(2). The public DOI resolver title and author were matched to the archived original on 19 November 2026.