Gas Recombination (Composition Restoration)

PVT Analysis of Reservoir Fluids › 14. Gas Recombination (Reservoir Composition Restoration)

Reviewed by Coretest reservoir-fluids engineers · Updated 21 June 2026

Gas recombination is not a "measurement" in the same sense as a CCE or CVD experiment: it is the sample-preparation and quality-control step that creates the live reservoir fluid on which every subsequent gas-condensate or wet-gas PVT study depends. For a gas reservoir sampled at separator conditions, recombination is the bridge between what was physically collected at surface and what actually exists in the reservoir at \(p_R\) and \(T_R\). Done wrong, it silently biases dew point, liquid dropout, \(Z\)-factor, and CVD recovery for the entire study. Done right, it delivers a single, well-characterized live fluid that all downstream tests can share.

Purpose↑ Contents

The recombination test answers a deceptively simple question: what is the true, in-situ composition of the reservoir gas, and how do we physically reconstitute it in the laboratory from the streams that were sampled at surface?

Surface samples are never the reservoir fluid. A separator splits the produced stream into a gas phase and a liquid (condensate) phase, and only their proportional recombination reproduces the original single-phase reservoir gas. Even single-phase bottomhole samples can require validation against surface streams. The practical questions it serves:

• Establish the master live fluid (correct molar composition \(z_i\), density, and gas/condensate ratio) used to initialize the dew-point test, Constant Volume Depletion (CVD), Constant Composition Expansion (CCE), p-V/Z gas-depletion, and any compositional or EOS-tuning work.
• Confirm that the recombined fluid is single phase at reservoir conditions (above the dew point) so that the sample is representative rather than a depleted or partially condensed fluid.
• Provide the validated GOR / GCR and the reconstructed composition that feed material-balance, deliverability, and surface-process design.

In short: it restores the reservoir composition, so that everything measured afterward describes the reservoir and not the separator.

Physical principle↑ Contents

The governing idea is conservation of moles across the surface separation. The produced reservoir stream flashed to separator conditions (\(p_{sep}\), \(T_{sep}\)) partitions into a gas phase and a condensate phase in vapor-liquid equilibrium. Recombination reverses that flash by remixing the two phases in exactly the ratio at which they were produced.

The overall (reservoir) mole fraction of component \(i\) is the mole-weighted average of the separator gas and separator liquid compositions:

$$z_i = \frac{n_g\,y_i + n_L\,x_i}{n_g + n_L}$$

where: - \(z_i\) = mole fraction of component \(i\) in the reconstituted reservoir fluid (dimensionless), - \(y_i\) = mole fraction of component \(i\) in the separator gas, - \(x_i\) = mole fraction of component \(i\) in the separator liquid (condensate), - \(n_g\) = total moles of separator gas, - \(n_L\) = total moles of separator liquid.

Equivalently, defining the recombination ratio on a molar basis, \(F = n_g / n_L\) (moles of separator gas per mole of separator liquid):

$$z_i = \frac{F\,y_i + x_i}{F + 1}$$

This is the molar-balance form of the source equation \(z_i = (n_{i,\text{initial}} + n_{i,\text{added}}) / \sum n_{total}\): the "initial" moles are the dominant separator-gas stream and the "added" moles are the condensate (or, in a build-up recombination, the make-up components or a standard reference mixture introduced to hit the target composition). The closure condition is the source's \(z_{i(\text{final})} \approx z_{i(\text{reservoir})}\).

Each phase composition itself obeys vapor-liquid equilibrium, characterized by the equilibrium ratio (K-value):

$$K_i = \frac{y_i}{x_i}$$

where \(K_i\) is the equilibrium ratio of component \(i\) at the separator \(p\) and \(T\). \(K_i > 1\) means the component prefers the gas phase (methane, nitrogen); \(K_i < 1\) means it prefers the liquid (C7+, heavy ends). The \(K_i\) set is exactly why surface gas and surface liquid have different compositions and why both must be recombined to recover \(z_i\).

The field measurement that fixes the ratio \(F\) is the gas/condensate ratio, the gas-phase analogue of GOR. The producing ratio is reported volumetrically, for example:

$$GCR = \frac{V_{g,sc}}{V_{cond,sc}} \quad [\text{scf/STB or sm}^3/\text{sm}^3]$$

To convert the measured volumetric ratio to the molar ratio \(F\) used in the balance, the standard-condition gas volume is divided by molar volume and the condensate volume is converted through its density and apparent molecular weight:

$$n_g = \frac{V_{g,sc}}{V_{M,sc}} \quad \text{where} \quad V_{M,sc} = \frac{R\,T_{sc}}{p_{sc}}$$

(\(= 23.69\ \text{sm}^3/\text{mol}\) at \(15\ \text{°C}\), \(101.325\ \text{kPa}\); \(379.5\ \text{scf/lb-mol}\) at \(60\ \text{°F}\), \(14.696\ \text{psia}\))

$$n_L = \frac{V_{cond}\,\rho_{cond}}{M_{cond}} \quad \text{where } \rho_{cond} \text{ is condensate density and } M_{cond} \text{ its apparent molecular weight.}$$

So:

$$F = \frac{n_g}{n_L} = \frac{V_{g,sc} / V_{M,sc}}{V_{cond}\,\rho_{cond} / M_{cond}}$$

Symbols: \(V_{g,sc}\) = standard-condition separator-gas volume; \(V_{cond}\) = condensate (stock-tank or separator) liquid volume; \(\rho_{cond}\) = condensate density (\(\text{kg/m}^3\) or \(\text{g/cm}^3\)); \(M_{cond}\) = condensate apparent molecular weight (\(\text{g/mol}\) or \(\text{kg/kmol}\)); \(R\) = universal gas constant; \(T_{sc}\), \(p_{sc}\) = standard temperature and pressure; \(V_{M,sc}\) = molar volume at standard conditions.

A consistency check uses real-gas behavior, since the separator gas is not ideal:

$$Z = \frac{pV}{nRT}$$

with \(Z\) the compressibility factor of the separator gas at (\(p_{sep}\), \(T_{sep}\)); this is needed to convert separator metered gas volumes to true moles when laboratory gas is dispensed at separator or cell conditions rather than standard conditions. Symbols: \(p\) = pressure, \(V\) = gas volume, \(n\) = moles, \(T\) = absolute temperature.

The single, non-negotiable validation criterion is thermodynamic: the recombined fluid must be a single dense-phase gas at reservoir conditions, that is, its dew-point pressure \(p_d\) at \(T_R\) must satisfy \(p_d < p_R\) (or, for a saturated gas-condensate, \(p_d \approx p_R\) within tolerance). If the recombination ratio is wrong, the reconstructed dew point shifts and the whole study is biased.

Apparatus and conditions↑ Contents

• Sample sources: a pair of matched separator samples, one separator gas (captured in a piston/floating-piston gas cylinder or evacuated bottle) and one separator liquid (condensate, in a piston cylinder), both with documented separator pressure and temperature at the time of sampling and the simultaneously metered GCR. Alternatively a single-phase bottomhole sample for validation, or pure make-up components / a certified standard mixture for a build-up reconstruction.
• Recombination/PVT cell: a high-pressure, high-temperature visual or blind cell. A mercury-free floating-piston PVT cell (variable volume, hydraulically driven piston) is the modern standard. Older mercury-displacement cells achieve the same volumetric control but are being phased out for HSE reasons. A windowed (sapphire) cell is strongly preferred so the operator can confirm single phase and later observe the dew point.
• Charging system: precision high-pressure positive-displacement (syringe-type) pumps with volume readout to ~\(0.01\ \text{cm}^3\), a calibrated balance for gravimetric charging of condensate, and accurate pressure (transducer, \(\pm 0.1\%\ \text{FS}\)) and temperature (RTD, \(\pm 0.1\ \text{°C}\)) instrumentation.
• Agitation: a magnetic mixer, rocking mechanism, or circulation loop to drive the charged streams to a single equilibrated phase.
• Conditions: charging is done at controlled separator or cell conditions; the cell is then brought to reservoir temperature \(T_R\) and pressurized above the expected dew point, typically up to or above reservoir pressure \(p_R\), to ensure a single phase before any test begins. What is held fixed: the target composition \(z_i\) (the design target) and reservoir temperature for validation. What is varied: cell pressure (to compress to single phase and later to find the dew point) and the relative amounts of the two streams charged.

Procedure↑ Contents

1. Compositional analysis of the surface streams: analyze the separator gas (GC to C7+ or extended C36+) and the separator liquid (GC plus distillation/density for the C7+ fraction) to obtain \(y_i\) and \(x_i\) with their molecular weights and densities. Record separator \(p_{sep}\), \(T_{sep}\), and the field-metered GCR.

2. Establish reservoir targets: from field data and the analyses, fix the target reservoir composition \(z_i\) (or the design GCR), the reservoir pressure \(p_R\), and reservoir temperature \(T_R\) that the recombined fluid must reproduce.

3. Compute the recombination ratio: convert the volumetric GCR to a molar ratio \(F = n_g / n_L\) using \(V_{M,sc}\), condensate density and molecular weight, and the separator gas \(Z\)-factor for any non-standard-condition conversions. This sets the moles of each stream to charge.

4. Condition and validate the source samples: heat each separator sample to single phase, restore it above its own saturation pressure, and verify the cylinder pressure/volume against the sampling record. Confirm the gas and liquid are each homogeneous before transfer.

5. Charge the heavy stream first: meter the calculated amount of separator condensate into the evacuated, clean PVT cell, recording exact mass (gravimetric) or volume at known density. Charging the liquid first reduces flashing losses.

6. Charge the gas stream: add the calculated moles of separator gas to the cell, tracking pumped volume, pressure, and temperature so the true moles delivered are known (apply \(Z\) at the charging condition). Keep the cell pressure rising so the mixture stays toward single phase.

7. Pressurize above the dew point: drive the floating piston to compress the combined fluid to a pressure clearly above the expected \(p_d\) (commonly to or above \(p_R\)) at reservoir temperature \(T_R\).

8. Equilibrate and homogenize: hold at \(T_R\) and mix (rock/stir/circulate) for an extended period, hours to overnight, until pressure and any optical readings stabilize, indicating a single equilibrated phase. Through the window, confirm no second phase (no liquid film, no haze).

9. Verify single phase and composition: sample a small aliquot of the recombined fluid and analyze by GC; confirm the measured \(z_i\) matches the target within tolerance (closure \(z_{i,\text{final}} \approx z_{i,\text{reservoir}}\)) and that the C7+ fraction is reproduced.

10. Adjust if off-target (build-up case): if \(z_i\) deviates, add the deficient component(s) or a metered amount of one stream / standard mixture, re-equilibrate, and re-sample until \(z_i\) and GCR converge to the reservoir values.

11. Validate the dew point: perform a quick constant-composition expansion to confirm \(p_d < p_R\) (or \(p_d \approx p_R\) for a saturated reservoir). This is the acceptance gate.

12. Release to downstream tests: with composition, density, GCR, and dew point validated, designate the cell contents as the master live fluid for dew-point, CVD, CCE, and p-V/Z studies, or transfer to a storage piston cylinder under single-phase conditions.

Measured parameters↑ Contents

• Separator gas composition \(y_i\) and separator liquid composition \(x_i\) (mole fractions through C7+ or extended).
• Separator conditions \(p_{sep}\), \(T_{sep}\) at sampling, and the field GCR (\(V_{g,sc} / V_{cond,sc}\)).
• Condensate density \(\rho_{cond}\) and apparent molecular weight \(M_{cond}\); separator-gas molecular weight and specific gravity.
• Separator-gas compressibility factor \(Z\) at charging conditions.
• Charged quantities: moles/mass/volume of gas and of condensate actually delivered to the cell, with the pump volume, pressure, and temperature at each charge.
• Cell pressure and temperature during compression and equilibration; equilibration time to stability.
• Post-recombination composition \(z_i\) from a GC aliquot of the live fluid.
• Validation dew-point pressure \(p_d\) at \(T_R\) from the confirmatory CCE.

Calculated / derived parameters↑ Contents

• Molar recombination ratio: $$F = \frac{n_g}{n_L} = \frac{V_{g,sc} / V_{M,sc}}{V_{cond}\,\rho_{cond} / M_{cond}}$$

• Reconstituted reservoir composition (per component): $$z_i = \frac{F\,y_i + x_i}{F + 1} \quad \left(\text{equivalently } z_i = \frac{n_g\,y_i + n_L\,x_i}{n_g + n_L}\right)$$

• Closure / acceptance condition: $$z_{i,\text{final}} \approx z_{i,\text{reservoir}} \quad \text{and} \quad GCR_{\text{final}} \approx GCR_{\text{reservoir}}$$

• Standard molar volume used in the conversion: $$V_{M,sc} = \frac{R\,T_{sc}}{p_{sc}}$$

• Live-fluid (reservoir) gas density: $$\rho_R = \frac{p_R\,M_{app}}{Z_R\,R\,T_R} \quad \text{where } M_{app} = \sum z_i\,M_i \text{ is the apparent molecular weight of the recombined fluid, } Z_R \text{ is its compressibility at reservoir conditions, and } M_i \text{ is the molecular weight of component } i.$$

• Apparent molecular weight of the live fluid: $$M_{app} = \sum (z_i\,M_i)$$

• Moles of each surface stream from metered volumes: $$n_g = \frac{V_{g,sc}}{V_{M,sc}} \;;\quad n_L = \frac{V_{cond}\,\rho_{cond}}{M_{cond}}$$

• Acceptance offset (QC): \(\Delta z_i = z_{i,\text{measured}} - z_{i,\text{target}}\), required within the lab tolerance band before release.

Reporting and typical curves↑ Contents

The recombination report is mostly tables plus a few diagnostic plots:

• Recombination composition table: \(y_i\) (separator gas), \(x_i\) (separator liquid), the computed \(K_i = y_i/x_i\), the recombination ratio \(F\), and the resulting \(z_i\), side by side with the target reservoir composition and \(\Delta z_i\). This is the central deliverable and is read by checking that \(z_i\) tracks the target and that the C7+ mole fraction and molecular weight are honored.
• Live-fluid properties table: apparent molecular weight \(M_{app}\), reservoir-condition density \(\rho_R\), GCR, and validated dew point \(p_d\) at \(T_R\).
• Material-balance / charge log: moles and mass of each stream charged, pump volumes, and the closure check \(\sum z_i = 1\).
• Validation p-V (CCE) curve at \(T_R\): cell volume versus pressure, read to locate the dew point as the break in slope (or the first observed liquid drop in the window). The acceptance reading is \(p_d\) relative to \(p_R\): \(p_d < p_R\) confirms a single-phase (undersaturated) gas; \(p_d \approx p_R\) indicates a saturated gas-condensate.
• Optional sensitivity plot: dew-point pressure versus recombination GCR, used to show how strongly the reconstructed saturation pressure depends on the chosen ratio and to justify the selected \(F\).

A correct report reads as a closed loop: surface analyses in, mole balance applied, recombined composition out, and a dew point that places the fluid on the right side of reservoir pressure.

Practical notes and pitfalls↑ Contents

• The GCR is the master lever. The single largest error source is an inaccurate field gas/condensate ratio, because \(F\) scales directly with it and the reconstructed dew point is highly sensitive to \(F\). Use simultaneously and accurately metered separator gas and liquid rates; reconcile field meters before trusting the ratio.
• Representative, matched samples: the gas and liquid must come from the same separator at the same, recorded \(p_{sep}\) and \(T_{sep}\). A gas sample and a liquid sample taken at different conditions cannot be validly recombined. Verify each source is restored to single phase in its cylinder before charging.
• Protect the C7+ fraction: the heavy ends control the dew point. Avoid flashing or "weathering" the condensate during transfer (charge liquid first, keep it pressurized and warm), and use an extended GC plus distillation so the C7+ molecular weight and density are right. A few percent error in C7+ moves \(p_d\) substantially.
• Equilibration time: under-mixing leaves the cell heterogeneous and yields a non-representative GC aliquot. Mix and hold until pressure and optical readings are truly stable (often overnight); confirm visually that no second phase persists.
• Validate, do not assume: always run the confirmatory dew-point/CCE check. If \(p_d\) comes out wrong relative to \(p_R\), the recombination ratio or a stream composition is off; do not proceed to CVD on an unvalidated fluid.
• Mercury-free operation: prefer floating-piston cells and pumps over mercury displacement for HSE compliance and to avoid mercury contamination of the sample; mercury can also amalgamate and bias volumes.
• Dead volume and lines: account for the dead volume of valves, transfer lines, and the cell head. Unpurged dead volume retains residual fluid or air and corrupts both the charged moles and the GC; evacuate and, where appropriate, flush lines with the stream being charged.
• Air/contaminant exclusion and \(Z\) corrections: evacuate the cell before charging to exclude air (which shows up as spurious \(N_2\)/\(O_2\) in the GC), and apply the gas compressibility factor \(Z\) when converting volumes to moles at non-standard charging conditions. Skipping the \(Z\) correction systematically miscounts the gas moles and shifts \(z_i\).
• Build-up recombination caveat: when reconstructing by adding pure components or a standard mixture rather than the real condensate, the synthetic C7+ characterization may not match the true reservoir heavy ends; flag this clearly, because EOS tuning and dew point inherit that approximation.