How Relative Density Is Used in Gas Flow Calculations
Most discussions of gas relative density stop at the definition. This page is about the working side: what the number actually does once it is entered into a flow computer. In AGA-style gas measurement the relative density is one of the highest-leverage inputs on a meter run - it helps establish the density of the flowing gas, it drives the compressibility computed by the gross characterization methods, and through both it scales every volume the flow computer reports. The worked example below shows, with plain algebra and no invented numbers, how far a gravity error travels, and the final section catalogs the data-entry mistakes that put such errors into real flow computers.
Relative density in flow calcs in one line: In gas measurement the relative density is a working input to the flow calculation, not just a descriptive property. Orifice flow computation uses it to establish the density of the flowing gas, and the gross characterization methods use it as a primary summary of what the gas is. Because orifice flow varies with the square root of density, a relative error in the entered gravity propagates as roughly half that relative error into the computed flow, which is why a wrong or stale gravity entry in a flow computer quietly biases every volume it reports.
Where the Value Enters the AGA Calculation Chain
An orifice meter does not measure flow; it measures a differential pressure, and the flow computer turns that differential into a flow rate using, among other things, the density of the flowing gas. That density is not measured directly on most runs. It is constructed from the static pressure, the temperature, the compressibility, and the relative density, which is the term that summarizes what the gas is made of. In the AGA 3 computation the relative density is therefore not background information: it sits inside the density term that every computed flow rate depends on.
The same number enters a second time through compressibility. The AGA 8 gross characterization methods exist so a flow computer can estimate compressibility from a few summary inputs, chiefly the relative density and the inert content, instead of a full component analysis. So one entered gravity value shapes both the density used in the flow equation and the compressibility factor that corrects gas behavior to base conditions. A single wrong entry contaminates the calculation through two doors at once, which is why gravity deserves more respect than a set-and-forget configuration field.
None of this works unless the number is on the basis the calculation expects. The definitional side of that question - what the ideal molar-mass ratio and the real density ratio each mean and why they diverge on heavy or CO2-rich gas - is covered in the companion explainer on relative density vs specific gravity of gas. This page takes those definitions as given and follows the number into the arithmetic.
A Worked Example in Symbols
Write the relative density as Gr, so the molar mass of the gas is M = Gr times M_air, where M_air is the molar mass of air. At pressure P and temperature T, with compressibility Z and gas constant R, the density of the gas is rho = P M / (Z R T), which is rho = Gr times (P M_air / (Z R T)). Hold the conditions fixed and the conclusion is immediate: the density the flow computer constructs is directly proportional to the gravity it was given. Enter a gravity that is high by some fraction and the computed density is high by the same fraction.
Now push that through the orifice equation, where the flow rate Q is proportional to the square root of the product of density and differential pressure. Suppose the entered gravity is wrong by a small relative error e, so the flow computer uses Gr times (1 + e) instead of Gr. The computed density becomes rho times (1 + e), and the computed flow becomes Q times the square root of (1 + e), which for small e is approximately Q times (1 + e/2). The algebra says a relative error in gravity surfaces as roughly half that relative error in flow, in the same direction, on every calculation interval.
The reason this matters operationally is that the error is a bias, not noise. It does not average out over the day; it accumulates in every totalized volume, feeds straight into imbalance and allocation figures, and never announces itself, because the flow computer keeps producing plausible numbers with nothing in fault. A gravity error survives until a reconciliation, an audit, or a check meter exposes it, and by then it has been multiplying quietly for as long as the wrong value has been in the box.
Common Data-Entry Mistakes in Flow Computers
The first family of mistakes is basis confusion. A gravity arrives from a chromatograph on one basis and is entered into a gross-method configuration that expects the other, or a value measured by a gravitometer as a real density ratio is treated as an ideal molar-mass ratio. On lean gas the discrepancy hides inside normal scatter; on heavy or inert-rich gas it becomes a genuine volume bias. A related error is applying the gravity from the wrong stream to a meter, which is easy to do on a site where several runs share an analyzer.
The second family is staleness. Many flow computers carry a fixed gravity entered at commissioning from a design-basis gas analysis, and the value is simply never revisited while the actual gas changes with well decline, new production mixes, or seasonal blending. A fallback value keyed in during an analyzer outage and never cleared behaves the same way: the live feed returns, but the override stays in force. Nothing fails, no alarm sounds, and the reported volumes drift away from the truth at half the rate the gravity is wrong.
The defenses are provenance and visibility. The configuration should make explicit where the gravity comes from, whether a live analyzer feed or a fixed entry, and any period running on a fallback should be annunciated rather than silent. A cloud SCADA platform such as Merobix supports this by trending the gravity the flow computer is actually using alongside the analyzer's reported value, so a mismatch or a forgotten override is visible as a diverging pair of lines, and by keeping an audit trail of configuration edits so a step change in volume can be traced to the gravity entry that caused it.
Frequently Asked Questions
How much does a relative density error change the computed gas flow?
For an orifice run, the flow rate is proportional to the square root of the gas density, and the constructed density is directly proportional to the entered gravity. Working the algebra through, a relative error e in the gravity produces approximately e/2 of relative error in the computed flow, always in the same direction. Because it is a bias rather than random noise, it accumulates in every totalized volume until the wrong entry is found and corrected.
Should a flow computer use a live gravity from the analyzer or a fixed value?
Use the live value where a healthy analyzer feed exists, with a documented fallback for outages, because the actual gas composition changes with well decline, production mixes, and blending while a fixed entry does not. If a fixed value must be used, it needs a review discipline: a recorded source analysis, a revisit schedule, and an annunciation whenever a fallback or override is the value actually in service, so a temporary entry cannot quietly become permanent.
Why does the relative density basis matter when entering the value?
The gross characterization methods and the flow equations are defined around a particular basis, and they use whatever number they are given as if it were on that basis. Feeding an ideal molar-mass ratio where a real density ratio is expected, or vice versa, biases the compressibility and the volume. On lean gas the two bases nearly coincide and the mistake hides; on heavy or inert-rich gas they diverge and the mistake becomes a measurable, persistent volume error.
Sources and verification
This page references the standards, specifications, and official documentation published by the organizations below. Editions, product capabilities, and documentation change over time - confirm current requirements and specifications directly with the source.
- AGA Measurement Standards (Report No. 3 / No. 8) - American Gas Association
Merobix is not affiliated with, endorsed by, or sponsored by these organizations; their names are used only to identify the standards and products discussed.
Automation services
Need help turning this into a working system?
Merobix integrates SCADA, programs Allen-Bradley and Siemens PLCs, and designs and fabricates industrial control panels.
Meeting requests are reviewed before confirmation.