Technical note

Core Flooding Experiments: Setup, Equipment, and Common Mistakes

How a core flood system is assembled, the eight-step startup sequence, steady-state versus unsteady-state methods, and the five mistakes that quietly ruin SCAL data.

Coretest Systems, Inc. · Core analysis reference

A core flooding experiment injects fluid through a reservoir rock sample at controlled pressure and temperature to measure how fluids flow through pore space. It is the primary laboratory method for generating relative permeability data, screening enhanced oil recovery formulations, and quantifying injectivity under reservoir conditions. When the setup is correct, coreflooding produces data that goes directly into reservoir simulation. When it is not, it produces numbers that look plausible but bear no relationship to reservoir behavior.

What a core flooding experiment is

Coreflooding is the workhorse of special core analysis (SCAL). A cylindrical plug of reservoir rock, typically 1.5 inches in diameter and 2 to 12 inches in length, is mounted in a pressurized holder. Fluids are injected at controlled flow rates or pressure differentials while the system maintains net confining stress conditions representative of the reservoir.

The primary applications are:

• Saturation-dependent relative permeability (kr) curves used directly in reservoir simulation models
• Wettability characterization through Amott-Harvey or USBM methods
• EOR chemical screening for surfactant, polymer, or low-salinity water formulations
• Injectivity assessment for CO2 storage, water disposal, or WAG projects
• Formation damage evaluation before and after drilling fluid or scale inhibitor exposure

The data quality is entirely dependent on how closely the laboratory conditions replicate the reservoir: effective stress, temperature, fluid saturation history, and flow geometry all need to match the target formation.

Core flood system components

The core holders are the central component and dictate many other specifications. Precision pumps determine your minimum measurable flow rate and the stability of the injection profile. For viscosity-sensitive EOR experiments, flow rate stability at low rates (below 0.01 mL/min) is critical.

Step-by-step core flood setup

• Step 1. Sample preparation. Clean the plug with appropriate solvents (toluene then methanol for most sandstones). Dry at 60–80°C until constant weight. Measure dry weight, dimensions, and porosity by gas expansion. Measure initial gas permeability at multiple net stresses.
• Step 2. Initial saturation. Evacuate the sample to below 0.1 mbar. Introduce formation brine under vacuum, then apply back pressure to dissolve any residual gas. Confirm full saturation by comparing pore volume from saturation weight gain against porosity measurement. A discrepancy greater than 2–3% indicates incomplete saturation.
• Step 3. Mount in core holder. Install end caps with appropriate filter discs matched to formation grain size. Insert into sleeve. Confirm the sleeve is free of damage before assembly.
• Step 4. Connect flow lines. Connect inlet, outlet, confining pressure, and differential pressure lines. Purge all lines of air before connection. Confirm all fittings are made up to torque specification.
• Step 5. Apply confining pressure. Increase confining pressure to the target net stress in increments of 500–1,000 psi, pausing at each step to confirm the sleeve is seating correctly.
• Step 6. Set pore pressure. Set the back pressure regulator to the target pore pressure. Pressure up the pore space from the inlet at low flow rate. Confirm the BPR is holding setpoint before proceeding.
• Step 7. Set temperature. Ramp the heating system to target temperature at no more than 2°C per minute. Monitor pore pressure throughout and confirm the BPR is managing thermal expansion. Wait for full thermal equilibration, indicated by stable temperature and pressure readings over a 30-minute window, before injection.
• Step 8. Begin injection. Start injection at the planned flow rate. Record differential pressure, pore pressure, confining pressure, and temperature at minimum one-minute intervals. Collect effluent in fraction tubes. The experiment begins when all parameters are stable at setpoint.

Steady-state vs. unsteady-state flooding

Steady-state flooding co-injects oil and water at fixed fractional flow ratios, stepping through a series of ratios until the full kr curve is defined. At each ratio, flow continues until saturation and differential pressure stabilize, then kr values are calculated directly from Darcy's law. Steady-state is the reference method: it generates accurate saturation values and is not affected by capillary end effects in the same way as unsteady-state. The drawback is time. A full steady-state kr curve on a low-permeability plug can take several weeks.

Unsteady-state flooding injects a single displacing fluid at constant rate or pressure and measures the production history. Johnson-Bossler-Naumann (JBN) analysis or numerical history matching is used to derive kr curves from the production data. Faster and uses less fluid, but the JBN method becomes unreliable at low production rates or when viscosity ratios are unfavorable. Numerical history matching is more robust but requires careful treatment of capillary pressure.

The same steady versus unsteady trade-off applies to single-phase gas permeability; we cover it in gas permeability measurement methods. For EOR screening, unsteady-state displacement efficiency measurements are often sufficient to rank candidate formulations. For reservoir simulation input, steady-state kr data is preferred where schedule allows.

Common mistakes and how to avoid them

Fluid bypassing around the core (poor sleeve seal). If the Viton sleeve fails to seal against the core plug face, injected fluid routes around the sample rather than through it. The symptom is differential pressure that is too low for the expected rock permeability, combined with rapid fluid breakthrough with no saturation change. Prevention requires proper endcap design with the endcap shoulder seating flush against the sleeve, inspection of the sleeve for cuts or swelling from previous fluid exposure, and confirmation that confining pressure exceeds pore pressure by the specified net stress before injection begins.

Capillary end effects distorting saturation profiles. At the outlet face of the core plug, capillary continuity is broken. Wetting-phase fluid accumulates near the outlet, causing measured average saturation to be higher than the true in-situ saturation. End effects are most severe at low flow rates, high interfacial tension, and in short core plugs. Solutions: use the longest core plugs available (12 inches preferred over 2 inches for kr measurements); increase injection rate to reduce the capillary number; or apply numerical correction methods.

Incomplete initial saturation. Residual air or gas in the pore space compresses during pressurization and then gradually dissolves, creating a moving gas bank that contaminates the saturation history. The symptom is irreproducible permeability measurements and anomalous early effluent volumes. The correct protocol is vacuum evacuation followed by saturation under back pressure at a minimum of 500 psi above the bubble point of the brine.

Starting injection before thermal equilibration. Injecting at reservoir flow rate before the system has equilibrated thermally means the early data is collected at a temperature gradient, not at reservoir temperature. Fluid viscosity, interfacial tension, and relative permeability are all temperature-dependent. The equilibration criterion requires stable temperature readings at both inlet and outlet thermocouples, plus stable pore pressure at the BPR setpoint, over a minimum 30-minute hold period.

Dead volume errors in saturation calculations. The volume of fluid held in tubing, valves, end caps, and fittings between the pump and the core inlet must be subtracted from injected and produced volumes when calculating saturations. Failing to account for system dead volume causes saturation values to be offset, sometimes by more than 5% pore volume in compact systems with short core plugs. Calibrate the system dead volume by flowing tracer through the system with a non-porous steel blank in the core holder before every new test series.

Getting the fundamentals right separates SCAL data that reservoir engineers trust from data that ends up flagged as an outlier in the database: material selection, system calibration, saturation protocols, and thermal management. Pumps that drift at low flow rates, endcaps with poorly toleranced sleeve seats, and back pressure regulators that chatter at setpoint will produce data that no analysis method can fully correct for.

Related equipment: Coretest special core analysis systems, precision HPHT pumps, and back pressure regulators are the building blocks of a core flood laboratory.