compositional fluid analysis, Oil & Gas Glossary

What Is Compositional Fluid Analysis?

Compositional fluid analysis (also called PVT analysis or fluid characterization) is the laboratory measurement and equation-of-state (EOS) modeling of a reservoir fluid sample to determine its molecular composition — from methane (C1) through heptane-plus (C7+) fractions, along with CO2, nitrogen, and hydrogen sulfide — and its phase behavior and volumetric properties at reservoir conditions and throughout depletion, providing the thermodynamic foundation for reservoir simulation, surface facility design, and enhanced oil recovery screening. Accurate compositional data allows engineers to predict how a reservoir fluid will behave from the moment it leaves the rock through separator trains, pipelines, and processing plants at surface.

Key Takeaways

Reservoir fluid samples are collected either downhole using wireline formation testers (MDT/RFT) or as recombined separator samples at surface, with downhole samples preferred for compositional accuracy.
The core laboratory tests are flash liberation, differential liberation, constant composition expansion, constant volume depletion (for gas condensates), and swelling tests (for EOR screening).
The Peng-Robinson (PR) and Soave-Redlich-Kwong (SRK) equations of state are the industry-standard thermodynamic models used to match and extrapolate PVT laboratory data.
C7+ characterization — splitting the heavy fraction into pseudo-components — is the greatest source of EOS tuning uncertainty and most significantly affects dew point and retrograde condensate predictions.
Compositional simulation data feeds reservoir models for gas cycling, CO2 injection, miscible flooding, and lean-gas injection EOR designs, where phase behavior changes drive recovery efficiency.

Fluid Sampling, Laboratory Tests, and EOS Modeling

The reliability of any compositional fluid analysis depends first on sample quality. Wireline formation tester tools such as Schlumberger's Modular Formation Dynamics Tester (MDT) or Halliburton's Reservoir Description Tool (RDT) collect fluid directly from the reservoir at in-situ pressure and temperature, avoiding the phase separation that occurs when fluid travels up the wellbore. These samples are transferred to pressurized transfer cylinders and shipped to a PVT laboratory where they are recombined to reservoir conditions before testing. When downhole sampling is not feasible, surface recombination samples are collected: separator gas and separator liquid are taken simultaneously, their compositions measured separately, and the streams mathematically recombined at the producing gas-oil ratio. The recombination method introduces error proportional to uncertainty in the measured GOR and is particularly problematic for near-dew-point gas condensates that are already losing heavy components in the separator.

In the PVT laboratory, a standard suite of tests characterizes the fluid across its production life. Constant composition expansion (CCE) measures the bubble point pressure (for oils) or dew point pressure (for gas condensates) and the compressibility of the fluid above and below saturation pressure. Differential liberation (DL) simulates reservoir depletion in an oil system: at each pressure step below bubble point, the liberated gas is removed from contact with the oil at reservoir temperature, mimicking the behavior of solution gas coming out of a reservoir oil as pressure drops. Constant volume depletion (CVD) simulates gas condensate reservoir depletion: gas is produced at constant volume to track retrograde condensate dropout — liquid that condenses in the reservoir pores and may be lost permanently to production. Separator tests at the planned facility conditions complete the dataset, providing formation volume factors and shrinkage factors for production engineering calculations.

Equation of state modeling transforms the raw laboratory measurements into a predictive thermodynamic tool. The laboratory data — saturation pressures, gas-oil ratios, densities, and viscosities at multiple pressure-temperature steps — are used to tune the EOS by adjusting binary interaction parameters and C7+ pseudo-component properties until the model reproduces the observed data within acceptable tolerances (typically within 2 to 5 percent on saturation pressure and within 5 percent on volumetric properties). Once tuned, the EOS can be used to predict fluid behavior at any pressure-temperature condition not directly measured in the laboratory: at different injection-gas compositions during EOR, at lower abandonment pressures during late-life depletion, or at the extremely high pressures and temperatures encountered in ultra-deepwater fields. This predictive capability makes the EOS the central thermodynamic engine in compositional reservoir simulators used for field development planning.

Fast Facts: Compositional Fluid Analysis

Sample types: Downhole MDT/RFT samples (preferred) and surface-recombined separator samples
Primary lab tests: CCE, differential liberation (oils), CVD (gas condensates), separator tests, swelling tests
Industry EOS models: Peng-Robinson (PR78) and Soave-Redlich-Kwong (SRK) are the two dominant choices
C7+ split: Heavy fraction split into 3–10 pseudo-components via gamma or exponential distribution
GOR range covered: From heavy oil (<200 scf/bbl) to lean gas condensate (>100,000 scf/bbl)
Bubble point accuracy target: EOS tuned to within 2–5% of measured bubble point pressure
EOR applications: CO2 minimum miscibility pressure, lean-gas enrichment ratio, gas cycling efficiency
Key output for facilities: Formation volume factor (Bo/Bg), shrinkage factor, stock-tank density, viscosity curves

EOR Screening Tip:

Before commissioning a full compositional simulation study for a CO2 or miscible gas injection project, run slim-tube displacement tests on the actual reservoir fluid to measure the minimum miscibility pressure (MMP) experimentally. The EOS-predicted MMP is sensitive to C7+ characterization and binary interaction parameters that are difficult to tune without a measured MMP as an anchor point. A slim-tube test costs a fraction of a full compositional study and provides the single most critical number for determining whether a miscible EOR project is technically feasible at the reservoir's current pressure.

Compositional fluid analysis is also referred to as:

PVT analysis — pressure-volume-temperature analysis; the most common field and lab shorthand, referring to the fundamental physical relationships being characterized
Fluid characterization — used in reservoir engineering and simulation contexts to describe the complete process of sampling, lab testing, and EOS tuning
Reservoir fluid study — the formal term used in laboratory reports; typically includes all relevant tests and an EOS model delivered as the final product
Phase behavior analysis — emphasizes the focus on phase envelope, bubble point, dew point, and retrograde behavior rather than just volumetric properties

Frequently Asked Questions About Compositional Fluid Analysis

Why is C7+ characterization the most difficult part of EOS tuning?

The C7+ fraction — all hydrocarbons with seven or more carbon atoms — is analytically reported as a single composite lump with an average molecular weight and specific gravity, because gas chromatography cannot fully resolve individual heavy components above C6 in typical reservoir fluid samples. Yet this fraction disproportionately controls the fluid's saturation pressure, viscosity, and phase envelope shape. Different methods of splitting C7+ into pseudo-components (gamma distribution, exponential distribution, or Whitson's characterization method) can predict dew points that differ by 500 psi or bubble points that differ by 200 psi using the same underlying EOS. Engineers compensate by using extended compositional analysis (TBP distillation or high-temperature GC) on at least one sample per reservoir to better constrain the heavy-end composition, then tuning the pseudo-component critical properties and binary interaction parameters to match all measured PVT data simultaneously.

What is the difference between black-oil simulation and compositional simulation, and when is compositional modeling required?

Black-oil simulation uses simplified, pressure-dependent PVT tables (Bo, Rs, Bg) derived from differential liberation tests, assuming the fluid composition does not change significantly during depletion. This approach is adequate for solution-gas drive oil reservoirs and dry gas reservoirs where only pressure depletion is involved. Compositional simulation tracks the mole fraction of each component or pseudo-component in every grid block at every time step, recalculating phase equilibrium with the EOS at each iteration. Compositional modeling is required whenever composition changes affect recovery: gas condensate reservoirs below dew point (retrograde condensate dropout changes remaining vapor composition), CO2 or enriched-gas injection (injected composition mixes with reservoir fluid), gas cycling projects, miscible flooding, and WAG (water-alternating-gas) processes. The computational cost of compositional simulation is typically 5 to 20 times higher than black-oil simulation for the same grid, which is why black-oil tables are used whenever compositional effects are minor.

How does reservoir fluid sampling depth affect the quality of compositional analysis?

Reservoir fluids frequently show compositional gradation with depth, particularly in large accumulations with tall hydrocarbon columns. Heavier components concentrate at depth due to gravity segregation, while lighter components increase toward the crest of the structure — a phenomenon described by the Schulte or Montel-Gouel compositional grading model. A single fluid sample collected near the crest may underestimate the heavy-component content of oil near the oil-water contact, leading to underestimation of viscosity and bubble point at depth. Best practice for large accumulations is to collect multiple fluid samples at intervals of 100 to 300 meters throughout the hydrocarbon column, verify compositional grading with a thermodynamic model, and use depth-varying PVT tables in the reservoir simulator. Single-point sampling at the producing interval is acceptable for thin reservoirs (less than 30 meters of hydrocarbon column) where grading effects are negligible.

Why Compositional Fluid Analysis Matters in Oil and Gas

Every decision in oil and gas field development — from setting the minimum tubing size that prevents liquid loading in a gas well to designing a multi-billion-dollar LNG liquefaction train — rests on an accurate understanding of how the produced fluid behaves across a range of pressures and temperatures. Compositional fluid analysis provides that foundation. An EOS model tuned to quality PVT data allows reservoir engineers to confidently predict recovery factors under different depletion strategies, facilities engineers to size separators and compression equipment for the full range of wellhead conditions, and EOR teams to screen injection scenarios and calculate economic returns before committing capital. Conversely, a poorly sampled fluid or a carelessly tuned EOS propagates error through every downstream calculation, resulting in oversized or undersized facilities, missed recovery targets, and costly mid-project redesigns. In an era when unconventional plays, deepwater fields, and complex EOR projects demand increasingly precise fluid characterization, high-quality compositional analysis is one of the highest-return investments a project team can make in the early appraisal phase.

Compositional Fluid Analysis: PVT Testing and EOS Modeling of Reservoir Fluids