Training On CCA and SCAL Experiments: 1. Overview

Download as pdf or txt
Download as pdf or txt
You are on page 1of 18

Training on CCA and SCAL Experiments

1. Overview
This document presents an example of the training provided by CYDAREX on special
core analysis (SCAL) experiments.
The training can cover the following topics:
Gas permeability/porosity
Formation factor
Liquid Permeability Experiments
Tracer Test Experiments
Pc/RI Experiments
Two-Phase Flow Experiments in steady state
Two-Phase Flow Experiments in unsteady state
All experiments are done under laboratory conditions, with pressure below 5 bars.

2. Porosity measurement
Porosity is defined as:
Vp
(1)
Vt
Where Vp is the volume of pores and Vt the total volume. Porosity can also been
calculated using the volume of solid Vs since Vt=Vp+Vs
Vt Vs
(2)
Vt

For cylindrical plugs, the total volume is derived from the length L and diameter D:
Vt L D2 / 4 (3)
And the volume of solid using the grain density d and the dry mass Mdry
Training on CCA and SCAL experiments (version July 2015)

Vs Mdry / d (4)
Mass is determined with a balance (accuracy 0.01 g) and dimensions with a caliper
(accuracy 0.01 mm)

3. Gas Permeability Experiments


Gas permeability is measured with a TinyPerm commercialized by Vindum
Engineering.
TinyPerm is a portable handheld air permeameter used for measurement of rock
matrix permeability on outcrops and at the core scale.
The measurement is based on the transient decay of the pressure inside the cylinder
when the valve is open. The apparatus deliver a number converted into
permeability with a chart.

4. Rock typing
For the various samples, permeability is plotted as function of porosity in semi-log
scale. The trends correspond to the different types of rocks, sandstones, carbonates,
double porosity…
10000.00

1000.00

FS
permeability (mD)

GVL
100.00 GV
GVX
GVZ
CLA
10.00
BRAU

1.00
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
porosity (frac.)

Page 2
Training on CCA and SCAL experiments (version July 2015)

5. Formation factor
The plugs are provided 100% saturated with brine, NaCl 35 g/l (density =1.024
g/cc). The dry mass before saturation is known.
1) Pore volume is determined from the difference of mass
Pv (Msaturated Mdry) /
2) Porosity is determined using Pv and geometrical total volume Vt: Vp / Vt
Formation factor is derived from the measurement of electrical resistivity of the
plug. Resistance is given from voltage and intensity (1000 Hz generator, 1Volt
maximum voltage) R V / I and resistivity by: R L / S , where L is length and S
surface area. Intensity is obtained my measuring the tension over a calibrated
resistance (200 ohms).
3.5

2.5
log(Formation Factor)
2

1.5

log ff 1
Archie Slope -2
0.5
redone
0
-2 -1.5 -1 -0.5 0
log(porosity)

Formation factor F is defined as the ratio of the plug resistivity by the brine
resistivity given by a diagram (from Schlumberger). The results are in good
2
agreement with Archie’s law: F

6. Liquid Permeability Experiments


Objective
Measuring the variation of pressure as a function of flow rate to measure the
absolute permeability.

Experimental Design
Core holder: Rock samples are typically 25 mm in diameter and 40 mm in length.

Page 3
Training on CCA and SCAL experiments (version July 2015)

Page 4
Training on CCA and SCAL experiments (version July 2015)

Experimental Setup:

Pin Pout

DP

Experimental Protocol:
Injection of 20 g/L NaCl brine at different flow rate Q. Measurements of Pin; Pout is
atmospheric pressure.
5 steps in Q: 0, 100, 200, 300, 400, 499 cc/h. Measure ∆P at plateau. Measurements
done with increasing and decreasing flow rate for quality control.
Here, sample GVI-4, 25 mm in diameter, 40 mm in length.

Results
Q (cc/h) ∆P (bar) at plateau
0 0.0074
100 0.223
200 0.417
300 0.575
400 0.714
499 0.856
400 0.700
Figure 1: Data acquisition showing
pressure as a function of time. 300 0.533
200 0.361
100 0.185
Table 1: Pressure at plateau.

Interpretation
Interpretation using module Permeability in CYDAR, option Steady-State liquid.
Fill in “Information”, “Sample”, “Fluid”, and “Data Points”.
Then “Calculate Permeability”. Results can be seen in “View” menu.

Page 5
Training on CCA and SCAL experiments (version July 2015)

Figure 2: permeability measurement analyzed in CYDAR.

Final calculation: k = 124 mDarcy.

Troubleshooting
Potential problems and corrections:
Inertial effects if Reynold’s number > 1.
Klinkenberg effects for gas if density is low.
Clay within the sample could make it hard to reach a steady pressure.
Air within the sample.

7. Tracer Test Experiment


Objective
A Tracer Test gives a measure of the homogeneity of the sample. If the sample is
heterogeneous, it should not be used for measurements of relative permeability.

Experiment Design
A sample is loaded with 20 g/L NaCl brine. During injection of a 50g/L NaCl brine,
the electrical conductivity of the solution at output is measured. The changed of
conductivity as a function of time will give information on the pore size distribution.

Page 6
Training on CCA and SCAL experiments (version July 2015)

Results
Raw Data: Measurements done on GP4.

Figure 3: 50 g/L replacing 20 g/L. Figure 4: 20 g/L replacing 50 g/L.

Interpretation

Figure 5: Voltage as a function of time. Figure 6: First derivative using Cydar.

CYDAR curve fitting tool is used to fit the experimental data with a spline function,
and calculate the first derivative. Time can be normalized as the time needed to
inject one pore volume. Voltage (or density) can be normalized between 0 and 1.

Page 7
Training on CCA and SCAL experiments (version July 2015)

Figure 7: normalized voltage vs. time Figure 8: normalized voltage vs. time
and first derivative. and first derivative.

Interpretation:
1
flux derivative

0.5

0
0 1 2 3 4 5
Pore Volume injected

Homogeneous sample (theoretical): spreading << 1 pore volume.

1.5
flux derivative

0.5

0
0 1 2 3 4 5
Pore Volume injected

Symmetrical curve: highly heterogeneous but no preferential paths.

Page 8
Training on CCA and SCAL experiments (version July 2015)

1.5

flux derivative
1

0.5

0
0 1 2 3 4 5
pore volume injected

Dissymmetrical curve: highly heterogeneous with preferential paths.

Troubleshooting
If the sample is heterogeneous, it should not be used for measurements of relative
permeability.

8. Pc/RI Experiments
Objective
To obtain the capillary curve (Pc) curve as a function of water saturation, and water
saturation as a function of Resistivity Index (RI).

Experiment Design
Definitions:
Drainage: experiment where water is pushed with oil. Starts at Sw ~ 1.
Imbibition: experiment where oil is pushed with water.
Capillary Pressure: Pc = Poil – Pwater

Resistivity Index: RI = R(Sw)/R(Sw=1) = Sw-n


Formation Factor: quantify effect of rock on electric conductivity
fR = Rrock / Rbrine = brine/ rock
Archie’s law used for log calibration: fR = -m

Page 9
Training on CCA and SCAL experiments (version July 2015)

With the porosity; m, Archie’s law exponent or cementation exponent. For


a clean formation, m = 2, m < 2 with clays
Experimental cell:

Experimental Set-Up:

Measurements of Resistivity:

Page
10
Training on CCA and SCAL experiments (version July 2015)

Figure 9: measurement with 4 electrodes. Figure 10: measurement with 2 electrodes.

To avoid contact resistance due to oil on the electrodes, a technique with 4


electrodes is preferred.

Actual Set-Up:

The pressure is imposed with the air tank on the oil surrounding the sample. A
porous plate allows the water to exit the sample but not the oil. For each pressure
step, the volume of water produced and the resistivity are measured as a function of
time.

Experimental Simulation using CYDAR


Determining the optimum oil pressure steps using CYDAR:
Two-Phase Flow experiments, Porous plate Experiment,
Sample size, sample characteristics, fluid characteristics,
one porous plate in outlet,
block times, starts with 1 day per step and 5 pressure steps,
Kr is entered as a Corey function,

Page
11
Training on CCA and SCAL experiments (version July 2015)

Pc input is usually obtained from mercury measurements, and is copy/paste.

CYDAR is used to simulate the time to reach equilibrium, and the total volume of
water produced.

The constraints are to have a measurable production of water at each step, and a
maximum pressure below 5 bars. Here, for instance, we see that the first pressure
step doesn’t quite reach equilibrium and should be longer.

Results and Interpretation


Production of water as a function of time is recorded. And for each measurement,
the tension VI, V2, and V4 are measured.
The initial water saturation Sw was 1; measuring the water produced gives Sw as a
function of time, imposed Pc, and RI.

Figure 12: Measured Pc (red) compared to


Figure 11: Imposed pressure (red) and water Mercury Pc (blue).
saturation (blue) as a function of time.

Page
12
Training on CCA and SCAL experiments (version July 2015)

Figure 13: Log RI as a function of Log of water


saturation measured with 2 (blue) and 4 (red)
electrodes.

The experiment presented here shows a much smaller Pc (Figure 12) compared to
what expected. The sample seems to have produced too much water for the applied
pressure.

Troubleshooting
Risk of breaking the porous plate when closing the cell.
Risk of imposing an initial pressure that is too high and emptying the sample
in one step. Need for simulation.
Make sure to lock the cell with screws to avoid loss of pressure over time.

9. Two-Phase Flow Experiments – steady state


Objective
Determining the relative permeabilities Kroil and Krwater injecting two fluids. Once
the absolute permeability is determined (with a Perm experiment), ∆P and Vwater
produced are used to determined Kr.
Evaluating the volume of water produced is easier since the system is at
equilibrium, so the ratio of water to oil in dead volume can be measured.

Page
13
Training on CCA and SCAL experiments (version July 2015)

Experiment Design

(Oil and water filters are now 15 µm).

Experimental Simulation using CYDAR


Numerical simulation in CYDAR is used to determine the oil and water flow rates,
the duration to reach equilibrium, and the corresponding water and oil productions
and pressure.

For sample GVI-3, with an absolute permeability of 168 mD and a porosity of 24%:

Page
14
Training on CCA and SCAL experiments (version July 2015)

Figure 14: block times and QW and WO. Figure 15: pressure for each block time.

Figure 16: average water saturation for each Figure 17: water saturation profile for each block
block time. time.

Note: At the end of measurements, possibility to make bumps (increase QO with


QW = 0) to have information on Pc because the profile saturation is linked to the
capillary pressure.

Results and Interpretation


Water production and pressure measurements are used in history matching to
determine the optimal Kr.

Figure 18: Relative permeabilities after optimization.

Page
15
Training on CCA and SCAL experiments (version July 2015)

Troubleshooting
In SCAL experiments, numerical simulations are required by major companies
(Shell, Total, Chevron …).
More accurate since they can take into account the real physics: capillary
pressure, gas compressibility, heterogeneities, non-constant injection
conditions…
As quality control, comparison between raw data (pressure, effluents,
profiles…) and simulated results.
Difference for the final saturation:
with analytical calculation (Figure 19, symbols) Sw(final) = 0.68
with numerical simulation (Figure 19, solid line) Sw(final) = 0.80
The analytical calculation does not take into account the capillary end
effect and uses the average saturation derived from the effluent balance

Figure 19: Difference between analytical and numerical results.

10. Two-Phase Flow Experiments – unsteady state


Objective
Determining the relative permeabilities Kroil and Krwater injecting one fluid, and
using the shape of the transient ∆P and water production Vwater.
Either one step (not used anymore) or multi-steps.

Experiment Design

Page
16
Training on CCA and SCAL experiments (version July 2015)

Figure 20: actual set-up. A double separator with a


capacitance is used to record the variation of
water level with time.

Results

Interpretation
JBN interpretation
"Welge" method or JBN for analytical interpretation in mono-step or multi-steps.
Assumes Pc = 0. Analyze data after breakthrough.
LQ w LQ o
Pw*
AK
; Po*
AK
Vw Vo Qt ; Vw' Vo' Q

Sw Si w (Vo tVo' ) / VP
Vo' Pw* Vo' Po*
Krw (1 ) ; Kro
Q P t P' Q P t P'
V’=dV/dt.
Example of one step; here breakthrough is at maximum pressure (but not always).

Page
17
Training on CCA and SCAL experiments (version July 2015)

Page
18

You might also like