PE 4521 Experiment 1: Measurements of Oil Density, API Gravity and Viscosity

1. Objectives. To determine density and viscosity of three liquid hydrocarbons as a function of temperature and to examine if any correlation exists between the liquid density and viscosity To determine viscosity of three inverse emulsions and to examine the effect of changing water/oil volume ratio on the emulsion rheology

2. Discussion. Detailed analysis of a crude oil using its complex chemical composition is very difficult if not impossible. Therefore, crude oils are classified according to their physical properties. Density (specific gravity or API gravity) and viscosity are the most important ones. Measurement of API gravity and viscosity is simple, yet essential in the petroleum industry. The specific gravity of a crude oil is defined as the ratio of the density of the liquid to the density of pure water, both measured at specified conditions of pressure and temperature. (Density of pure water is given in Table 3-28.) Dynamic viscosity of a crude oil can be loosely defined as the internal resistance of the oil to flow. Viscosity appears in the mobility coefficient of Darcys law during the description of flow of phases in porous medium. The latter is the characteristic fluid flow parameter that controls the rate at which fluid can be produced under an applied potential gradient. Density, on the other hand, is important when the flow takes place under the influence of gravity; hence it appears as the coefficient of depth gradient in Darcys equation. An emulsion is a fluid system containing two liquid phases, one of which is dispersed as droplets in the other. The liquid which is broken up into droplets is termed the dispersed phase, whilst the liquid surrounding the droplets is known as the continuous phase or dispersing medium. The two liquids, which must be immiscible or nearly so, are frequently referred to as the internal and external phases, respectively. There are two types of emulsions, direct emulsions and invert emulsions. In direct emulsions, oil droplets are dispersed in water, which were also called oil-in-water emulsion (o/w). In invert emulsions, the opposite is true, since the water droplets are dispersed in oil, which is also called water-in-oil emulsion (w/o). Generally invert emulsions are preferred when a large amount of oil is desired. Emulsions are formed using two phases (such as oil and water) in the presence of an emulsifying agent, i.e., emulsifier. The most common emulsifiers are surfactants. A surfactant, or surface active agent, is a macromolecule with a polar (hydrophilic) head and a long non-polar hydrocarbon (hydrophobic) tail. A good analogy for the surfactant molecules would be lollypops. These molecules locate themselves at the wateroil interface and give an elastic behavior to the dispersed droplets in the emulsion system and, hence, create a thermodynamically stable emulsion. Crude oil emulsions are generally of the water-in-oil type, which are more viscous than either of their constituents. On the other hand, oil-in-water emulsions have lower viscosity than that of the oil phase. Measuring the emulsion viscosity is one of the objectives of this experiment.

Figure 1 Dyed mineral oil mixed with water to produce a pink-colored emulsion.

Figure 2. Appearance of emulsion with its dispersed and continuous phases under the microscope.

Figure 3. A water-in-oil emulsion droplet shown at a microscopic level.

3. Equipment. 25 ml pycnometer, electronic balance, Cannon-Fenske viscometers, API hydrometers, graduated cylinders, electronic balance, stopwatch, latex gloves, lab coat and safety goggle

4. Materials. Various crude oils, water and an emulsifying agent. 5. Procedure. To measure the API gravity and density of oil at room temperature using hydrometer: a) Dip the API hydrometer in the graduated cylinder containing the liquid at room temperature. Spin the hydrometer gently and allow it to come to rest. Note the API gravity reading on the stem of the hydrometer b) Calculate the specific gravity o=141.5/(131.5+oAPI) c) Calculate apparent molecular weight of the crude oil using MW=42.43 o / (1.008-o) d) Calculate the oil density oil = o water (g/cc) of the crude oil at room temperature. Here the water density is read from the attached pure water table To measure the density of oil at room temperature using pycnometer: a) Wash the pycnometer by water, clean it, and let it dried using oven b) Measure the weight of the empty pycnometer using electronic balance c) Fill the pycnometer with one of the oils used in hydrometer experiments above, put the cover, and allow the excessive liquid to come out through the capillary tube in the cover d) Weight the filled pycnometer, using electronic balance e) Record the difference in the weight of the pycnometer between the filled and the empty. f) The density of the liquid is the ratio of the difference in the weight of pycnometer to the volume of pycnometer. g) Repeat the measurement of density for the other oil samples used in hydrometer experiments above To measure oil viscosity: a) See the attached instructions for Cannon-Fenske viscometer. Fill the viscometer as indicated in step 4. It will turn out that when the viscometer is inverted the liquid will fill about half of the bulb in the bigger arm. Apply suction to draw the liquid from the big arm so that it is in the bulb above the upper etched mark. Allow the liquid to fall freely down past the lower etched mark. Note the time required for the liquid meniscus to pass from the upper to the lower mark. You will want to replicate this step at least one more time to find the average flow time b) Repeat at several temperatures c) Calculate the kinematic viscosity from the average time and the viscometer constant. d) Repeat the measurements at three other temperatures. Make sure that the viscometer is kept in the temperature bath for sufficient long time for the liquid to come to the temperature of the bath e) Use density values obtained from step b to find the dynamic viscosity values at three temperatures

f) Clean all equipment and the work area To measure w/o emulsion viscosity: a) Prepare the invert emulsions at varying emulsifying agent concentration. Place a suitable volume of diesel oil into a mixer. Add the required volume of emulsifying agent to diesel oil in the mixer slowly and mix the mixture. Finally, add the suitable volume of distilled water and mix it about 20-30 minutes b) Measure the emulsion viscosity using the Cannon-Fenske viscometer as explained above for the oil c) Repeat the steps using emulsions with two changing water/oil volume ratios and hence two different volumes of continuous phase 6. Report a) In the introduction part of your report discuss the significance and practical applications of density, API gravity and viscosity measurements in oils and oil field emulsions In the discussion part of your report: b) Compare the values of oil density and API gravity estimated using hydrometer and pycnometer. c) Examine and comment on the temperature dependence of viscosity for the various fluids you considered d) Compare the emulsion viscosity with the viscosity of the oil phase. Discuss the impact of water/oil volume ratio on the viscosity e) Comment on controllable measurement error and its impact on calculated results

DENSITY OF PURE WATER

(FROM PERRYS CHEMICAL ENGINEERING HANDBOOK, ON PAGES 375 AND 376)

continue next page

Unit Conversion: To convert kilogram per cubic meter to pounds per cubic foot, multiply by 0.06243 oF=9oC/5+32

PE 4521 Experiment 2 Measurement of Gas Viscosity using Molecular Dynamics Simulation 1. Objective. To predict the viscosity of a mixture of gas with known composition using molecular dynamics (MD) simulation along with Einsteins equation 2. Discussion. Accurate prediction of gas viscosity using instrumentation is difficult and often not reliable. Here we learn to use molecular simulation as the method of predicting gas viscosity at a particular pressure and temperature. Einsteins equation to predict the viscosity (Pa-s) of any fluid is given as:

gas

V 10 30 = 2 k B T t

[ G(t ) G(t = 0) ] 2

eqn (1)

where V is the volume of a cubic computational cell with a 40 length, kB the Boltzmann constant equal to1.3806E-23 J/K, T the temperature (273.0K) and t the time step for the simulation, which is 10sec in this case. The term in the bracket is known as the canonical ensemble average square mean displacement of the molecules with G(t) described as:
G (t ) = 1 V

r m (t )
i i i =1

Here, N is the number of gas molecules in the computational cell, ri is the position of molecule i in the -direction, and mi the momentum of the same molecule in the -direction at time t. 3. Equipment. DL-POLY, parallel computing using LINUX cluster at the OU Supercomputing Center for Education & Research.

Figure 1. A snap shot from the MD simulation showing the location of gas molecules in a slit-shape pore.

4. Procedure. Below, the square mean displacement of the molecules is given at 10 second intervals for a number of gas molecules at 273.0K temperature and 1.028MPa pressure.
172.06 172.08 172.09 172.04 172.1 172.24 172.08 172.04 172.15 172.09 172.04 172.08 172.06 172.12 172.11 172.03 172.04 172.09 172.12 172.02

[G(t)-G(0)]2 (Pa-sec)2

[G(t)-G(0)]2 (Pa-sec)2

Take the average value of these and use it in equation (1) to estimate the gas viscosity 5. Report Report the estimated viscosity in centipoises (cp) 6. Bibliography. 1. Allen, M.P. and Tildesley, D.J. (2007) Computer Simulation of Liquids, Oxford University Press, London 2. Frenkel, D. and Smit B. (2002) Understanding Molecular Simulation From Algorithms to Applications, Academic Press, Computational Science Series, San Diego

PE 4521 Experiment 3: Statistical Analyses of Core Data

1. Objective: The goal of this study is to use data that can be obtained from core analyses and/or well logs to determine statistical measures for distributions of reservoir properties (e.g., porosity, water saturation and permeability) and determine whether the properties can be spatially correlated. 2. Discussion: The volumetric estimation of original oil in-place is conceptually simple; see Craft et al. (1990), and Tiab and Donaldson (1996). The fundamental estimate is the reservoir bulk volume occupied by the liquid and/or vapor hydrocarbon phases. However, there are other parameters in the equation shown below. These include porosity, net to gross and phase saturation. In this exercise you will examine the character of these properties measured on cores believed to be from the same reservoir.

= N ( STB) 7758 Vbulk (1 S wi ) N G

Bo

In practice engineers normally use average quantities for porosity, and water saturation, Swi. Unfortunately, they usually just calculate the averages without subjecting them to statistical analyses. The confidence with which the averages could be used would be much greater if the engineer understood how much statistical confidence could be placed in them. The goal of this lab experience is for you to develop insight into the use of statistics to understand core data. In addition, current trends in reservoir engineering are to use simulators for reservoir analysis and management. For some problems, it is important to examine the effects of reservoir heterogeneity, namely the values of reservoir parameters such as porosity and permeability are developed as functions of position within the reservoir. Consequently, it is convenient to identify if relationships exist between the parameters or if they are independent distributions. It seems appropriate to consider this experiment in two parts. First, examine core reports to identify correlated sections for all wells. Then check those sections, see if the porosity and water saturation frequency distributions appear to be from the same population. You will also check to see if they follow a normal distribution. Many people believe that they do (e.g. Amyx et al.1960). If the two cores are drawn from the same population, the sample frequency diagrams should be quite similar. Similarity would indicate that the two cores are from parts of the same reservoir that were formed by the same geologic processes. If the frequency diagrams are different you might conclude that they might be formed by different processes or at different times. If the frequency diagrams are the same, calculating average porosity and water saturation for the reservoir should be done using the combined frequency distributions, particularly since you will have proven that they are from the same population. If the distributions follow the normal or standard distribution (bell shaped curve), there are specific tests, such as the t-test; you can use to establish confidence limits that the measures of

central tendency and dispersion for the two cores were drawn from the same population. If they do not follow a normal distribution, you will need to use more words and more complicated analysis (thinking) to show how you determined the average values. The second part of the experiment does not directly impact on estimating the quantity of hydrocarbons originally in the reservoir. Rather it has to do with identifying relationships between reservoir parameters that can be used for scaling or assigning parameter (permeability) values to parts of the reservoir in a reservoir simulator. In this section, you will look for relationships between the parameters. It is reasonable to hypothesize that porosity is the fundamental parameter and connate water saturation and permeability could be functions of porosity. It would be reasonable to plot water saturation versus porosity to see if there is an identifiable relationship. (See Chapra and Canale, chapter 17 Regression.) Many people believe that permeability is log normally distributed. That means that the logarithm of permeability is normally distributed. Check your data to see if it confirms this belief. Plot permeability (and log-permeability) versus porosity and saturation to identify any relationship. You should report the results of your regression analysis with measures of quality. 3. Equipment: Core reports for two wells. These core sections are thought to be from the same unit. Use porosity, measured water saturation and maximum K [md] as data for your statistical analysis. 4. Procedure: Use the data, to estimate representative values of porosity and permeability. Also identify relationships that may exist between porosity, water saturation and permeability. This will go more easily in a spreadsheet since they have sort functions, graphical capability and some statistical analysis capability. (Make sure you know what the Excel functions mean and do.) 5. Report: Be sure to include your porosity and permeability histograms. It would be useful to have the histograms for all wells on the same figure (also the combined histogram if the data warrant it). Also include the cross-plots of water saturation and permeability versus porosity. Put the statistical calculations in the appendix. 6. References: Amyx, J.W., D.M. Bass, Jr and R.L. Whiting, Petroleum Reservoir Engineering, Physical Properties, McGraw-Hill, Inc., pp 536560, 1960. Chapra, S.C. and R.P. Canale, Numerical Methods for Engineers, 3rd Edition, McGraw-Hill Publishing Company, Part 5 and Chapter 17, 1998. Craft, B.C., M. Hawkins and R.E. Terry, Applied Petroleum Reservoir Engineering, 2nd Edition, Prentice-Hall, pp. 69-76, 148-150, 1991. Tiab, D. and E.C. Donaldson, Petrophysics, Gulf Publishing Co., pp. 118-122, 1996.
i.y.a. 2012

PE 4521 Experiment 4 Volumetric estimation of OOIP

1. Objective. The goal of this project is to determine the original oil in place using volumetric method. 2. Discussion. The volumetric estimation of original oil in place is conceptually simple; see Craft et al. (1990). The basic estimate is the estimate of the reservoir bulk volume occupied by the liquid and/or vapor hydrocarbon phases. Petroleum reservoirs are irregularly shaped bodies. To properly determine the volume one should integrate the irregular functions between the limits that bound the reservoir and its separation into distinct phase regions.

S
Vb

/ Bo ( x, y, z ) dx dy dz

That means determining the bounding top and bottom surface functions and the horizontal limits of the formation. Since these are not simple functions, integration must be approximate rather than analytic, Chapra and Canale (1998). Historically, geoscientists have determined structure maps from seismic imaging. They have then converted those maps to isopachous maps of the oil and gas zones using oilwater and gas-oil contact data where available. Graphical methods are generally used to contour areas of constant saturated thickness and a numerical method is used to integrate the volume, Vb, contained in the reservoir as mapped. The actual formula used is simple, Craft et al. (1991). N = C Vb (1-Sw) / Bo In oilfield units for liquid hydrocarbon saturated reservoirs, C is 7758 rb/ac-ft, Vb is in acft, and Bo is in rb/stb and N is in stb. Porosity and water saturation is expressed in fraction. Even though the formula is simple there are several sources of errors. An excellent discussion on uncertainty estimation in volumetrics determination is given by Floris and Peersmann (1998). 3. Equipment. Planimeter or another graphical integration device. 4. Procedure. Use the data in Attachment 1, particularly the structural map and the isopach map of the pay zone thickness, excerpted from Andrews, et al. (1995 or 1996), to prepare a net oil sand isopachous map. Integrate the areas within contours using a planimeter and using

an appropriate numerical method to find the volume of the reservoir filled with hydrocarbons. There are three ways to calculate the volume: trapezoidal rule, Simpsons rule and equation for the volume of the frustum of a pyramid. Trapezoidal rule: h is fixed = 10 ft

Volume

Simpson rule:

1 h(a0 2a1 2a2 .......... . 2an 1 an ) tn an 2

Volume

1 h(a0 4a1 2a2 4a3 ....... 2an 2 4an 1 an ) tn an 3

where h = contour interval (ft) ao = area enclosed by zero contour ( at oil water contact) (acres) a1 = area enclosed by first contour an = area enclosed by nth contour tn = average formation thickness above the top contour Simpson rule is the more accurate of the two for irregular curves. However, either will usually give satisfactory results. Simpsons rule is slightly more tedious and has the limitation that an even number of contour intervals (odd number of contours) must be used. The equation for the frustum of a pyramid can be expressed by:

Find the OOIP and compare with the one from the Booch Play where a0 = area enclosed by the lower contour in section (acres) an = area enclosed by upper contour in section (acres)

1 Volume h(a0 an a0 an ) Do the calculation of the bulk volume associated with this reservoir 3

Use the volume described by your map and other data from the table of reservoir engineering data to estimate the original oil in place by all three methods. (For the Booch reservoir, you will notice that a fault is mapped between the Hall #1 discovery well and other wells in the reservoir. The results of the reservoir simulation study suggest that the fault is a seal. Compare your estimate of original oil in place with the value reported in Attachment 1. 5. Report: Be sure to include your isopach map of the oil zone. Include calculations as part of the appendix material. Performing calculations in a spreadsheet would be helpful.

6. Questions: a) One of the figures in Attachment 1 is the type log for the field; it shows the information used to determine the top and bottom of the pay zone interval. How precise do you think the thickness can be measured? What impact can that have on the estimate of original oil and gas in place? b) What are the various sources of errors in volumetric calculations? How will you minimize them? 7. References: Andrews, R.D., Northcutt, R. A., Knapp, R. M. and Yang, X. H. 1995. FluvialDominated Deltaic (FDD) Oil Reservoirs in Oklahoma: The Booch Play, Oklahoma Geological Survey, Special Publication 95-3, 67 pp. Andrews, R.D., Rottman, K., Knapp, R. M., Bhatti, Z. N. and Yang, X. H. 1996. FluvialDominated Deltaic (FDD) Oil Reservoirs in Oklahoma: The Skinner and Prue Plays, Oklahoma Geological Survey, Special Publication 96- 2, 106 pp. Chapra, S.C. and Canale, R. P. Numerical Methods for Engineers, 1998. 3rd Edition, McGraw-Hill Publishing Company, Chapter 21. Craft, B.C., Hawkins, M. and Terry. R. E. 1991, Applied Petroleum Reservoir Engineering, 2nd Edition, Prentice-Hall, pp. 69-76, 148-150. Floris, F. J. T., and Peersmann, M. R. H. E. 1998. Uncertainty estimation in volumetrics for supporting hydrocarbon exploration and production decision-making, Petroleum Geosciences, 4, pp. 33-40.

iya2012

PE 4521 Experiment 5 Hydrocarbon Phase Behavior

1. Objective Study the pressure-volume behavior of hydrocarbon fluids and CO2water mixture to determine experimental values for i) bubble point, ii) gas solubility, iii) formation volume factor, and iv) isothermal compressibility. 2. Discussion Crude oil physical properties such as gas solubility Rso, and formation volume factor Bo are function of pressure, temperature and composition (McCain, 1990). For a given hydrocarbon, the reservoir depletion can be approximated as an isothermal process and as such the above properties can be considered to be a function of pressure alone. In this experiment you will be determining bubble point, Rso, Bo and compressibility of a selected number of hydrocarbon fluids and CO2water mixture at room temperature. Bubble point is the pressure Pb at which the bubbles of free gas first appear. Gas solubility, Rso, is the number of standard cubic feet of gas that will dissolve in one stock-tank barrel of oil at reservoir temperature and pressure, [scf/stb]. Henrys law states that the solution of gas within a liquid is directly proportional to the pressure exerted by the gas above the liquid. However, this is an ideal law that only applies in limited circumstances. Standings correlation, equation 1.26 in Craft et al. 1991, corrects for the real nature of reservoir fluids by using an exponent of 1.204 rather than 1.000. Standings correlation is an approximation and may not be appropriate for specific oils and gas systems. When gas is forced into solution in crude oil by pressure there is an increase in the total liquid volume. This increase in liquid volume is described by the oil formation volume factor, Bo, and is based on stock tank volume. The formation volume factor may be defined as the volume in barrels at reservoir conditions occupied by one stock-tank barrel of oil plus the dissolved gas, [rb/stb]. Isothermal compressibility is a measure of change in volume as the pressure is changed. It can be calculated experimentally using the following equation:

co =

1 V1 V 2 V P1 P2

where V1 and V2 are volumes at pressures P1 and P2. V is the reference volume generally assumed to be average of the two volumes. 3. Equipment PVT cells, hydrocarbons, water and CO2, latex gloves, and safety goggles.

4. Procedure Briefly, the experiment consists of measuring pressure and volume of hydrocarbon and gas mixture in a PVT cell at room temperature. You will be using three PVT cells. Detailed instructions on how to operate the two cells will be provided to you in the laboratory. As you decrease the pump pressure in 500 psig increments down to the bubble point, read the pump displacement corresponding to each pressure stage. Check the cell inspection window during each pressure reduction. The appearance of small bubbles indicates that the saturation (bubble point) pressure has been reached. Expand the system below the bubble point in several increments (3-5) using the remaining pump displacement. Agitate the cell after each expansion to establish equilibrium before reading exact system pressure and volume. Use the data collected to determine the bubble point pressure by plotting system volume versus pressure. Above the bubble point all gas is in solution and you are directly measuring volume changes with pressure. Below the bubble point you are measuring the volume occupied by the liquid (oil plus dissolved gas) and the vapor phase volume occupied by the gas that has come out of solution. This is often referred to as the two phase or total formation volume factor, Bt. The liquid has volume N*Bo and the vapor has volume N*Bg (Rsoi Rso). Use the form of the correlations listed in Craft et al. 1991, along with your data to develop the PVT suite of Bo and Bg, Bt and Rso. You will need to make some assumptions. You can make good use of equations (1.26), p. 33, (1.28), p. 35, (1.29), p 37 and (1.1, 1.31), p. 38 in Craft et al. 1991. Note- you do not use them exactly as they are written but rather use their forms and use values at the bubble point pressure to solve for coefficients so that you preserve the forms while adjusting the correlations for your experimental results. 5. Report a) Graphs (experimental points clearly indicated) for all the liquids and gas combinations: i. Cell pressure versus sample volume. ii. Two phase formation volume factor, Bt and Bo, versus cell pressure. iii. Solution gas, Rso, versus cell pressure. iv. Compressibility of liquids vs. pressure. b) Report bubble point for the studied hydrocarbons. c) Show sample calculation for all the above parameters for one hydrocarbon. 7. References 1. Craft, B.C., Hawkins, M., and Terry, R. E.1991. Applied Petroleum Reservoir Engineering, 2nd Edition, Prentice-Hall, pp.31-44. 2. McCain, Jr., W. 1990. The Properties of Petroleum Fluids, 2nd edition, PennWell Books, Tulsa, OK.
IYA 2012

PE 4521 Experiment 6 Interfacial Tension, Contact Angle & Capillary Pressure Measurements
1. Objectives. To perform high pressure mercury injection experiment on three samples two sandstones and one limestone; from the collected data, to calculate air-brine capillary pressure curve, pore throat size distribution, permeability and J-function. To measure solid-air-brine (with and without surfactant) interfacial tensions (IFTs), contact angles using Du Nuoy tensiometer and contact angle meter. 2. Discussion. Several techniques such as porous plate method, centrifuge method, and high-pressure mercury intrusion are used to obtain capillary pressure (see Amyx et al. 1960; Tiab and Donaldson 1996 etc.). Of all these methods, high pressure mercury injection is the most common one. Capillary pressure, IFT and contact angle data is very important to reservoir engineers as it is used to predict water saturation, calculate free water level, evaluate rock quality (e.g., wettability), calculate and interpret relative permeability curves, predict pore size distribution, develop residual hydrocarbon saturations etc. 3. Equipment. Rock samples, electronic scale, mercury injection high pressure porosimeter, DuNouy tensiometer (to measure interfacial tension) and contact angle measurement apparatus. See the attached supplementary notes for the equipment. Detailed directions on how to use the porosimeter, tensiometer and contact angle meter will be handed out in the laboratory. Caution: In this experiment you will be using mercury. Be very careful; at all times wear gloves. As in the other laboratory sessions, no eating and drinking is allowed in this lab. 4. Data and report. The basic data that you will be using consist of pressure and mercury intrusion volume. From the data do the following: 1) plot intrusion pressure versus Hg saturation, 2) calculate and plot pore throat size histogram, 3) convert and plot Hg-air data to air-brine capillary pressure, 4) estimate R35 and calculate permeability using Windland equation, and weighted geometric mean approach, 5) plot J-function versus wetting phase saturation and 6) discuss which of the three samples will be a better reservoir and why? 5. Bibliography: Amyx, J. W., Bass, D. M., and Whiting R. L.1960. Petroleum Reservoir Engineering: Physical Properties, McGraw Hill Book Co., NY, 610pp. Tiab, J. and Donaldson E. 1996. Petrophysics-Theory and Practice of Measuring Reservoir Rock and Fluid Transport Properties, Gulf Publishing Co. Houston, 706pp. Vavra, C. L., Kaldi, J. G., and Sneider, R. M. 1992. Geological Applications of Capillary Pressure: A review, AAPG Bull. 76, 840-850. Dastidar, R, Sondergeld, C. H. and Rai, C. S. 2007 An Improved Permeability Estimator from Mercury Injection for Tight Clastic Rocks, Petrophysics, 48, 186-190.
iya2012

Pc =

2C cos r

Pc (air brine) = Pc (air Hg )

.2166Pc (K / )1/ 2 J (Sw ) = 2 cos( )
Winlands Equation:

( cos )(air brine) ( cos )(air Hg )

LogR35 = 0.732 + 0.588LogK 0.864Log

LogK = 1.245 + 1.469Log + 1.701Log(R35 )

LogK = 2.51 + 3.06Log + 1.64Log(R wgm )

where Pc: capillary pressure in psi : contact angle (130o for air-Hg, take 0o for air-brine) : interfacial tension in dynes/cm (485 for air-Hg; you will determine in the lab for air-brine using tensiometer) r : capillary radius in microns C : conversion constant (0.145) K : permeability in millidarcy : porosity (%) R35 :pore radius in microns at 35 percentile non wetting phase saturation Rwgm: weighted geometric mean radius

Supplementary Notes for Du Nouy Tensiometer

Du Nouy Tensiometer No 70545

Correction factor for surface and interfacial tension

Supplementary Notes for CAM-PLUS MICRO Contact Angle Meter

AutoPore IV Series
Automated Mercury Porosimeters

Fast and Accurate Porosimety Analyses AutoPore IV Series Automated Mercury Porosimeters
The term porosimetry is often used to include the measurements of pore size, volume, distribution, density, and other porosity-related characteristics of a material. Porosity is especially important in understanding the formation, structure, and potential use of many substances. The porosity of a material affects its physical properties and, subsequently, its behavior in its surrounding environment. The adsorption and permeability, strength, density, and other factors influenced by a substances porosity determine the manner and fashion in which it can be appropriately used. The mercury porosimetry analysis technique is based on the intrusion of mercury into a porous structure under stringently controlled pressures. Besides offering speed, accuracy, and a wide measurement range, mercury porosimetry permits you to calculate numerous sample properties such as pore size distributions, total pore volume, total pore surface area, median pore diameter, and sample densities (bulk and skeletal). The AutoPore IV Series Mercury Porosimeters can determine a broader pore size distribution (0.003 to 360 micrometers) more quickly and accurately than other methods. These instruments are enhanced with features that enable them to more accurately gather the data needed to characterize the porous structure of solid materials. They also offer new data reduction and reporting choices that provide more information about pore geometry and the fluid transport characteristics of the material.

Wide Variety of Benefits

Controlled pressure can increase in increments as fine as 0.05 psia from 0.2 to 50 psia. This allows detailed data to be collected in the macropore region. A quick-scan mode allows a continuous pressure increase approximating equilibrium and providing faster screening. The high repeatability and reproducibility of this method by the AutoPore means that small, but significant, differences between samples will be detected. You can use this technique to screen a sample for conformity to specification because repeatability remains high. High-resolution (sub-microliter) measurement of intrusion/extrusion volumes produces extraordinary precision allowing the development of tighter sample specifications, improved production processes, and high-quality research data. A choice of correction routine for baseline (automatic, differential, or manual) produces greater accuracy by correcting for compressibility and thermal effects caused by high pressure. Choice of pressure ramping methods lets you choose the scanning technique for high-speed or on-demand results, or equilibration techniques for more accurate results with greater detail. The instrument allows the user to program data collection using a minimum number of data points. However, during intervals of unexpectedly large amounts of intrusion, the AutoPore will automatically collect additional data points. Choose between three data plot routines constructed of collected data points, a continuous curve plotted from interpolated data, or a combination of points and curves. A variety of pore volume, pore area, and pore size plots is available as well as the ability to calculate total intrusion volume, total pore (surface) area, median pore diameter, average pore diameter, bulk density, and apparent (skeletal) density.

Four Models
The AutoPore IV Series is available in four models to best match the needs of individual quality assurance and research labs. AutoPore IV 9520: 2 high-pressure (60,000 psia maximum pressure) and 4 low-pressure analysis ports AutoPore IV 9510: 1 high-pressure (60,000 psia maximum pressure) and 2 low-pressure analysis ports AutoPore IV 9505: 2 high-pressure (33,000 psia maximum pressure) and 4 low-pressure analysis ports AutoPore IV 9500: 1 high-pressure (33,000 psia maximum pressure) and 2 low-pressure analysis ports

Typical AutoPore IV Applications

Pharmaceuticals Porosity and surface area play major roles in the purification, processing, blending, tableting, and packaging of pharmaceutical products as well as a drugs useful shelf life, its dissolution rate, and bioavailability. Ceramics Pore area and porosity affect the curing and bonding of greenware and influence strength, texture, appearance, and density of finished goods. Adsorbents Knowledge of pore area, total pore volume, and pore size distribution is important for quality control of industrial adsorbents and in the development of separation processes. Porosity and surface area characteristics determine the selectivity of an adsorbent. Catalyst The active pore area and pore structure of catalysts influence production rates. Limiting the pore size allows only molecules of desired sizes to enter and exit, creating a selective catalyst that will produce primarily the desired product. Paper The porosity of print media coating is important in offset printing where it affects blistering, ink receptivity, and ink holdout. Medical Implants Controlling the porosity of artificial bone allows it to imitate real bone that the body will accept and allow growth of tissue. Electronics By selecting high surface area material with carefully designed pore networks, manufacturers of super-capacitors can minimize the use of costly raw materials while providing more exposed surface area for storage of charge.

AutoPore IV Advantages
Ability to measure pore diameters from 0.003 to 360 m Available with two low- and one high-pressure ports or four low- and two high-pressure ports for increased sample throughput Available in 33,000 psi or 60,000 psi models Quiet, high-pressure generating system Upgradeable without the need for more lab space Enhanced data reduction package; includes tortuosity, permeability, compressibility, pore-throat ratio, fractal dimension, Mayer-Stowe particle size, and more Equilibration by sample-controlled, rate of intrusion Operates in scanning and time- or rate-equilibrated modes Collects extremely high-resolution data; better than 0.1 L for mercury intrusion and extrusion volume Controlled evacuation prevents powder fluidization

Aerospace Surface area and porosity of heat shields and insulating materials affect weight and function. Fuel Cells Fuel cell electrodes require controlled porosity with high surface area to produce adequate power density. Geoscience Porosity is important in groundwater hydrology and petroleum exploration because it relates to the quantity of fluid that a structure can contain as well as how much effort will be required to extract it.

Filtration Pore size, pore volume, pore shape, and pore tortuosity are of interest to filter manufacturers. Often, pore shape has a more direct effect upon filtration than pore size because it strongly correlates with filtration performance and fouling. Construction Materials Diffusion, permeability, and capillary flow play important roles in the degradation processes in concrete, cement, and other construction materials.

Broad Choice of Analysis and Report Parameters

Operating Software
The AutoPore offers various options for obtaining important sample information as quickly as possible and for presenting the data in a format which you can design. Analysis options include choice of analysis variables, equilibration techniques, and pressure points at which data are collected. After operating conditions for the instrument have been chosen, they can be stored as a template and then reapplied to other samples, saving time and reducing the potential for human error. A selection of report options lets you customize many aspects of the data pages. You can select a specific range of data to be used in calculations; arrange columns of tabular data; select cumulative, incremental, or differential plots; scale the X-axis to display in either logarithmic or linear format for pore size; report actual or interpolated data; and select data presentation units such as psia or MPa, diameter or radius, and micrometers or Angstroms.

Data Reduction
The AutoPore IV generates tabular and graphical reports of percentage pore volume vs. diameter, and a summary report of percentage porosity in user-defined size ranges. The user has the ability to average several analyses and to use the resulting average analysis as a reference with which to compare subsequent analyses. A standard, single, user-defined analysis may also be entered and used for subsequent comparisons. SPC reports are available with collected data or user-defined parameters. In addition to the standard data reduction methods, the AutoPore IV Series also provides the following:

Mayer-Stowe Particle Size - Reports equivalent spherical size distributions Pore Tortuosity - Characterizes the efficiency of the diffusion of fluids through a porous material Material Compressibility - Quantifies the collapse or compression of the sample material Pore Number Fraction - Reports the number of pores in different size classes Pore-throat Ratio - Reports the ratio of pore cavities to pore throats at each percent porosity filled value Pore Fractal Dimensions - Quantifies the fractal geometry of a material Permeability - Reports the ability of the sample to transmit fluid

The AutoPore software includes a full report system for producing publicationquality graphics and user-specified reports.

The pore size distribution of two alumina samples are overlaid to provide a comparison of pore structure.

Penetrometer Characteristics
The penetrometer consists of a sample cup bonded to a metal-clad, precision-bore, glass capillary stem. The sample is placed in the sample cup; during analysis, mercury fills the cup and capillary stem. As pressure on the filled penetrometer increases, mercury intrudes into the samples pores, beginning with those pores of largest diameter. The mercury moves from the capillary stem resulting in a capacitance change between the mercury column inside the stem and the metal cladding on the outer surface of the stem. The AutoPore detects very slight changes in capacitance (equivalent to a difference of less than 0.1 microliter of mercury) so extraordinary resolution is achieved.

Micromeritics also offers a large selection of penetrometer bulbs, volumes, stems, and closures designed to fit most sample forms, shapes, porosity, and quantity. The better the match between the sample, its porosity, and the measurement range of the sample cell, the more precise the results.

The computer will not accept keyboard instructions to overpressurize the system. The high-pressure system is mechanically unable to generate unsafe pressures. A circuit stops the generation of pressure in the event of a failure in the computer. The operating specifications of the pressure systems (low = 50 psia, high = 60,000 psia) are well below the actual designed safety limits.

Safety Systems
The AutoPore features several levels of mechanical and electro-mechanical safety devices:

The extensive report system includes pore structure calculations including Cavity-toThroat Size Ratio, Fractal Dimension, Material Compressibility, and Statistical Process Control reporting.

A convenient data summary is automatically generated with each sample reported.

The material compressibility is easily calculated and may be reported both graphically and in tabular form using the AutoPore report system.

Additional pore structure calculations are included in the AutoPore software.

To request a quote or additional product information, visit Micromeritics web site at www.micromeritics.com, contact your local Micromeritics sales representative, or our Customer Service Department at (770) 662-3636.

WATERFLOODING AND ENHANCED OIL RECOVERY PE 4521 Experiments 7, 8 and 9 Fall 2011

1. Objectives A) To become familiar with the application of core (sand pack) evaluation tests to predict recovery factors from waterflooding, surfactant flooding and gas flooding. B) To determine petrophysical and multi-phase flow properties of the sand pack such as porosity, permeability, irreducible water saturation, residual oil saturation and end-point relative permeability for each recovery system. 2. Discussion Primary production seldom depletes an oil reservoir. Common practice has been to water flood partially depleted reservoir after an initial primary production. The series of experiments you are scheduled to perform will help you to create a porous medium using mixtures of sands, to saturate a sand pack to establish oil saturation at connate water saturation and to flood. The latter stage also involves determination of the efficiency values of imbibition during water displacement, surfactant flooding and gas drive. Surface tension at the oil-water interface of a reservoir system has a major influence upon residual oil saturation in the immiscible flooding of a porous medium. The displacement efficiency of a flood system increases as the interfacial tension decreases. Part of this net effect from a change in surface tension can be attributed to a change in wettability that also influences the residual oil saturation (see Craft et al. 1991, Figure 9.19, p. 381). One method of decreasing the interfacial tension between the displacing fluid and crude oil is by adding surfactant to the displacing fluid. When interfacial tension is sufficiently reduced, oil recovery can approach 100 per cent of the oil remaining after water flooding. Performance of a surfactant flood depends on several factors such as 1) pore geometry, 2) interfacial tension, 3) wettability or contact angle, and 4) pressure gradient. A number of factors, including brine composition, have major effect on interfacial tension. Many parameters must be considered in the design of a surfactant flood process. Gas drive is one of the techniques used to recover additional oil when recovery from primary energy becomes uneconomic. A gas flood is applicable in reservoirs with high permeability and high relief. Gravity segregation may occur in a gas drive system. Under some conditions segregation can result in high recovery factors. In other cases, low recovery may occur when gas overrides a zone of high oil saturation. Gas may also be used for pressure maintenance. In this case the recovery process is more complex. As a pressure maintenance agent it can also be useful for reservoirs with low permeabilities (see Craft et al. 1991, section 9.3.2, pp. 353360).

3. Equipment Flow experiment assembly, Lucite sand pack tube, graduated cylinders, electronic balance, pycnometer, stopwatch, latex gloves and safety goggles. You will be working with dyed brine that may stain clothes. In general keep the pressure difference across the sand pack relatively low; a few psi (never more than 10 psi); this will give you enough time to watch what happens and to record data. 4. Materials Mineral oil, NaCl brine, sand, sulfonate surfactant mixture (10% by volume in deionized water) and laboratory compressed air. 5. Procedure Waterflooding a) Build sand pack using 100 mesh quartz sand (grain density = 2.65 gm/cc). Prior to making the sand pack weigh all empty tube and all other necessary parts those are needed to make full sand pack assembly. Start preparing the sand pack. Tap the side of the Lucite tube while filling to ensure a tight pack with minimum voids. After filling, put the tube in the metallic holder. Measure dimensions of the sand pack. Obtain the dry weight of the sand pack. Calculate pore volume and dry porosity of the pack. The porosity should not be more than 27%. If it is more, repeat the packing till you get porosity <27%. This is to ensure that the pack has reasonable permeability for you to do flow through experiments. b) Install sand pack on the waterflood apparatus. Evacuate the sand pack to remove all the air and any moisture, then saturate with brine. Recover at least one pore volume of brine to ensure that the pack is fully brine saturated with no air left inside the tube. Disconnect and weigh. Calculate the saturated porosity. The difference between dry and saturated porosity should not be more than 3 porosity units. c) Establish a flow rate with brine and determine absolute permeability. (Use brine viscosity and density measured earlier in the semester) d) Flood the sand pack with oil until no water is produced. Measure displaced water. Determine irreducible water saturation and initial oil saturation. Determine effective (relative) oil permeability at irreducible water saturation [Craft et al. 1991]. Weigh the sand pack at this point to confirm saturations using a mass balance. e) Waterflood sand pack with 2-3 pore volumes of water. Catch oil displaced in graduated cylinder. (If you do this slowly enough, you may be able to plot cumulative oil production (Np) versus amount of water injected (Wi). This is the fundamental performance plot for waterflooding. You can record Np+Wp (water produced), Np (or Wp), versus time. Determine residual oil saturation to waterflood. Establish steady flow rate of brine and measure effective (relative) permeability to water at residual oil saturation. Weigh core at residual oil saturation to confirm water and oil saturations using a mass balance.

Surfactant Flooding a) Flood the sand pack with oil until no more water is produced. Measure displaced water. Calculate starting oil and water saturation. Then flood the sand pack with surfactant solution. Catch effluent in graduated cylinders. 250 cc graduated cylinders are provided in the lab. Break through of the surfactant solution should occur between 50 and 100 cc. You will be able to see this by color of the effluent. You may take the density of the surfactant 1g/cc and the viscosity 0.96cst. b) During the last 1-2 pore volumes, determine the effective permeability to the surfactant solution. c) Allow the cylinders to set so that the oil and water will separate. Measure the oil and water recovered. It is useful to do this in a plot of Np versus Wi. You can also determine WORs for the separate cylinders and on a cumulative basis. d) Clean all equipment including the sand packs and the work area. Gas Flooding a) For this experiment you will have to prepare a new sand pack, measure its properties, flow brine and oil to bring the sand pack to irreducible water saturation. (Plan to prepare the sand pack while other members of your team are doing the surfactant expt.) b) Flow oil through the sand pack so as to reach original hydrocarbon saturation. Check this by weight measurements. c) Displace oil from the sand pack with compressed air at constant pressure. Measure oil production as a function of time. Record gas injection rate at the same times you measure oil production. You can use this data to determine injected gas. Continue flooding until oil production is zero. 6. Report results: a) Pore volume, porosity and absolute permeability of sand pack. b) Initial oil saturation, connate water saturation and Kro (Swc). c) Residual oil saturation and Krw (Sor). d) Recovery factor at the end of the water flood, surfactant flood and gas flood. See displacement efficiency, Ed[Craft et al. 1991]. e) Mobility ratio for the water flood. f) Determine oil saturation, water saturation after the surfactant flood and Krw (Sor) g) Efficiency of the surfactant flood in terms of the oil in the core at the beginning of the surfactant flood. h) The final oil saturation for the sand pack after surfactant flooding. i) Oil recovery factor for the gas flooding experiment. j) The final oil saturation for the sand pack after surfactant flooding. k) Np versus injected gas. 7. Things to keep in mind when writing the report (You may want to address these points in your report) a) What special safety concerns are there about these experiments? b) What factors might reduce the water flood recovery factor? Why? c) What properties of the reservoir rock or fluids could cause those factors to change? Why? d) Does the water flooding process seem efficient? Why? f) What factors might reduce the surfactant flood recovery factor? Why?

g) Does the surfactant flooding process seem efficient? Why? h) What are special economic concerns to monitor / overcome in surfactant flooding? j) Does the gas flooding process seem efficient? Why? (You might want to compare to your water flooding results.) 8. References and Bibliography Craft, B.C., M. Hawkins and R.E. Terry, 1991. Applied Petroleum Reservoir Engineering, 2nd Edition, Prentice-Hall, pp.148153 and 335-360, and 380-386.

iya2012

Page 1 of 11

Experiment 1 & 2 Measurements of Oil Density, API Gravity and Viscosity Measurement of Gas Viscosity using Molecular Dynamics Simulation

PE 4521 002Reservoir Fluid Mechanics Laboratory Team 002E

Role Manager Researcher Technician Analyst Name Axel Hannenberg Lemmy Oshenye Lucas Gurgel De Carvalho Nor Ashraf Norazman Performance Score 1.0 1.0 1.0 1.0 Signature

Academic Integrity Statement

On my honor, I affirm that I have neither given nor received inappropriate aid in the completion of this exercise. Name: _________________________________ Date: _____________________________

Page 2 of 11 Abstract An emulsion is a dispersion (droplets) of one liquid in another immiscible liquid. The viscosity of emulsion depends on several factors such as viscosities of oil and water, volume fraction of water dispersed, droplet-size distribution, temperature, shear rate, and amount of solids present (Kokal 2006, I538). The oil densities obtained from hydrometer and pycnometer are 0.858-, 0.843-, 0.825-g/cm3 and 0.887-, 0.890-, 0.863-g/cm3 for crude oil 1, crude oil 2, and diesel respectively. The results are relatively similar to each other. The API gravity values acquired from hydrometer and pycnometer are 33-, 35.9-, 39.5API and 33.7-, 38.4-, 40-API for crude oil 1, crude oil 2, and diesel respectively. The API gravity of each samples measured from hydrometer are slightly lower than what are given due to equipment sensitivity to temperature. Temperature has a significant effect on viscosity. At 23C, the dynamic viscosities for crude oil 1, crude oil 2, and diesel are 5.25-, 3.34-, 1.31-cP. At 33C, the dynamic viscosities for crude oil 1, crude oil 2, and diesel are 19.9-, 7.71-, 2.58-cP. At 60C, the dynamic viscosities for crude oil 1, crude oil 2, and diesel are 11.72-, 5.94-, 2.31-cP. The kinematic and dynamic viscosities decrease as temperature increases. At room temperature, the kinematic viscosities of crude oil 1, crude oil 2, diesel, emulsion 1, and emulsion 2 are 23.3-, 9.28-, 3.14-, 118.45-, 143.8-cSt. The kinematic viscosity for the emulsion is higher compared to the crude oils and diesel. Accurate prediction of gas viscosity using instrumentation is difficult. Molecular simulation method is used to predict gas viscosity at a particular pressure and temperature. The average square mean displacement of molecules is 172.1 (Pas)2. The calculated gas viscosity is 0.000146 cP. Both the density and viscosity are inversely related to temperature. The measurement of the viscosity and density of reservoir fluid could be a simple procedure but nonetheless an important aspect in the economics of the well.

Page 3 of 11 Introduction Although determining density and viscosity of petroleum fluids is simple, it is very important. For instance, the heaviest oils may contain a high percentage of contaminant or nonhydrocarbon molecules, resins and asphaltenes. The economic importance of understanding these high contaminant percentages is demonstrated in the additional cost occurred in the processing time. Extra chemical separation means special equipment is required and that contribute to greater cost. The lighter oils are easier to deal with chemically because they require less exhaustive refinement. The viscosity is important because it affects the pressure losses when the fluid is being transported or produced and it affects the cost. Density and viscosity vary with temperature, pressure, and composition. Temperature causes the major variations in these properties. Thus, it is important to study how these properties change with temperature. In the petroleum industry, the density is measured in API gravity. A hydrometer can be used to measure API gravity and density. The fluid composition will dictate whether the density will increase or decrease. When temperature increases, the oil expands and its lightest fractions that were previously dissolved leave the solution. For gas, as temperature increases, it expands and hence the density decreases. When temperature increases, the viscosity can decrease nonlinearly (Fig. 1). On the other hand, the viscosity of the gas increases with the increase of temperature due to the increase of the intermolecular collisions. Some enhanced oil recovery (EOR) methods exploit these temperature/pressure concepts. For example, steam injection and in-situ combustion are used to increase the reservoir temperature in order to lower the oil viscosity so it will be easier to produce. The increase in temperature seen in the EOR methods adds more energy to the reservoir. Fig. 1 shows that the greater the viscosity is, the greater it decreases with the increase of temperature. Many petroleum reservoirs are linked to an aquifer. In many cases, reservoir water is also produced along with the oil. Water is a polar substance and the oil is apolar, meaning they are immiscible in nature. Since oil and water do not mix, there is an interfacial tension between the two liquids. To reduce this interfacial tension, an emulsifier can be added to form an emulsion which is more stable. However, the emulsion has completely different properties.

Fig. 1Viscosity vs. temperature for three different oils (Tarcilio 2011).

Page 4 of 11 Experimental Procedure Experiment 1 Hydrometer, pycnometer and Cannon-Fenske routine viscometer are used for this experiment (Fig. 2). Dip the API hydrometer in the cylinder containing the oils at ambient temperature. Spin the hydrometer and let it stabilize before taking the API gravity reading on the stem of the hydrometer. Calculate the density of oil from the API gravity reading. Use electronic balance to record the pycnometer filled with oil. Calculate the difference between an empty pycnometer and filled pycnometer. Take note of the volume of the pycnometer to calculate the density of the oil. The Cannon-Fenske viscometer is designed to measure viscosity of Newtonian liquids from of 0.5 to 20,000 cSt (Viswanath et al. 2007, 18). The viscometer is kept in a beaker containing water. The water is heated to achieve the desired temperatures. Apply suction to the tube until the oil in the viscometer is above the upper mark on the tube. Start the stopwatch once the oil passes the upper mark and stop it once the oil reaches the lower mark. Multiply the efflux time by the viscometer constant to obtain the kinematic viscosity. Repeat measurements several times for different oils and temperatures. Diesel, emulsifying agent and water are mixed into a mixer. To measure the water/oil emulsion viscosity, use the Cannon-Fenske viscometer after preparing the invert emulsion. Use the stopwatch to record the time for the emulsion to reach the lower mark. Repeat the steps using emulsions with different water/oil volume ratios. The mixture contains 150 mL of diesel, 150 mL of water, and 2 mL of emulsifier. It is assumed that the water/oil ratio is 50/50 from the relative volumes.

Fig. 2From left to right: Pycnometer and electric balance; hydrometer to measure API gravity; and Cannon-Fenske routine viscometer used to measure oil viscosity.

Experiment 2 The viscosity of gas with known composition is determined by using molecular dynamics simulation together with Einsteins equation. The simulation is set at 273K temperature and 1.028 MPa pressure. The square mean displacement of the molecules is provided at 10-second intervals. Accurate gas viscosity is important to understand the composition of the fluid within the reservoir.

Page 5 of 11 Results Experiment 1

TABLE 1OIL DENSITY VALUES OBTAINED FROM TWO METHODS oil (g/cm3) Crude oil 1 Hydrometer Pycnometer 0.858 0.887 Crude oil 2 0.843 0.890 Diesel 0.825 0.863

TABLE 2API GRAVITY VALUES OBTAINED FROM TWO METHODS API Crude oil 1 Hydrometer Pycnometer 33 33.7 Crude oil 2 35.9 38.4 Diesel 39.5 40

0.860 0.855 0.850 0.845 0.840 0.835 0.830 0.825 0.820 0.815 0.810 0.805 20 30 40 50 60 Temperature, C

Oil density, , g/cm3

Crude oil 1 Crude oil 2 Diesel

70

Fig. 3Oil density vs. temperature. Density of oil decreases with increase in temperature.

Page 6 of 11

25

Kinematic viscosity, , cSt

20 15 10 5 0 20 30 40 50 60

Crude oil 1 Crude oil 2 Diesel

70

Temperature, C

Fig. 4Kinematic viscosity vs. temperature. Viscosity of oil decreases with increase in temperature.

25 20 15 10 5 0 20 30 40 50 60 70 Temperature, C

Dynamic viscosity, , cP

Crude oil 1 Crude oil 2 Diesel

Fig. 5Dynamic viscosity vs. temperature. Viscosity of oil decreases as temperature increases.

Page 7 of 11

0.860 0.855 0.850 0.845 0.840 0.835 0.830 0.825 0.820 0.815 0.810 0.805

Oil density, , g/cm3

33C 23C

60C

60C 33C 33C 23C 23C

60C

Crude oil 1 Crude oil 2 Diesel

10

15

20

25

Dynamic viscosity, , cP

Fig. 6Oil density vs. dynamic viscosity. The rate of change for viscosity is dependent upon the oil density as a function of temperature. The higher the oil density, the lower the API gravity, brings greater change in density with temperature. TABLE 3KINEMATIC VISCOSITIES OF EMULSION & OIL PHASES Crude oil 1 (cSt) 23.29 Crude oil 2 9.28 Diesel 3.14 Emulsion 1 118.45 Emulsion 2 143.80

Experiment 2 The average square mean displacement of molecules is 172.1 (Pas)2. The calculated gas viscosity is 0.000146 cP. Discussion of Results Experiment 1 The oil densities obtained from hydrometer and pycnometer methods are relatively similar to each other (Tables 1 and 2). The results of differences may be due to the sensitivity of the equipment. Each of the equipment has its own unique uncertainty. The uncertainty of the hydrometer could be contributed by parallax error. The API gravity of each sample measured from hydrometer is slightly lower than the API gravity measured from pycnometer due to equipment sensitivity to temperature. For the Cannon-Fenske equipment, the constant of each station has to be interpolated or extrapolated by assuming it has a linear relationship with temperature. The kinematic and dynamic viscosities have similar correlations to temperature change. Viscosity decreases as temperature increases. This could be as a result of the intermolecular attraction getting weaker. As temperature increases, the volume of oil increases, the density of oil decreases (Fig. 3). The change in viscosity is largest for crude oil 1 followed by crude oil 2 and diesel. Figs. 4 and 5 show that there is a trend that the smaller the API gravity, the larger the change in viscosity (Fig. 6). The change in temperature can be related to the conditions in the subsurface. As hydrocarbon fluid is being produced, the temperature decreases until it reaches the surface

Page 8 of 11 temperature. Understanding the dynamics of the fluid properties such as viscosity, density, and API gravity can help the production team in designing the well for similar fields or for well intervention purposes. The kinematic viscosity for the emulsion is higher compared to the oil phase. The emulsion kinematic viscosity is five times higher than crude oil 1, 12 times higher than crude oil 2, and 38 times higher than diesel at room temperature of 23C (Table 3). The small quantity of emulsifier contributes to the change of the properties of the immiscible phases. The surface interfacial tension is reduced and then reaches thermodynamic equilibrium. This means that there is lower energy required to increase the surface area of two phases. The emulsion viscosity is substantially greater than the oil phase because emulsion shows non-Newtonian behavior. This behavior is the result of droplet crowding or structural viscosity (Kokal 2006, I-538). The emulsion viscosity shows a similar behavior to the viscosity of the oil phase with changes in temperature. Determination of the fluid viscosity could aid reservoir fluid engineers in running simulations for wells with similar properties to predict future production. Experiment 2 The gas viscosity is calculated using the Einsteins equation. The gas viscosity determination can tell the quality of the gas produced from the reservoir. Conclusion and Recommendation The viscosity and temperature are inversely related in a nonlinear behavior. This is also true for density. Gas viscosity can be determined from the combination of molecular dynamics simulation and Einsteins equation. There is a trend that the smaller the API gravity, the larger the change in viscosity. However, there is no exact correlation about this trend. Determining the viscosity and density of reservoir fluid is an important aspect in the economics of the producing wells.

Page 9 of 11 References Kokal, S.L. 2006. Petroleum Engineering Handbook, Vol. 1, I-533I-538. Richardson, Texas, SPE. Tarcilio Viana Dutra Junior, Petrobras. Viswanath, D.S., Ghosh, T.K., Prasad, D.H.L. et al. 2007. Viscosity of Liquids: Theory, Estimation, Experiment, and Data. Dordrecht, The Netherlands: Springer. Appendices = 0 +
(0 )(1 0 ) 1 0

, ..............................(1)

= 131.5+, ..............................(2)
= 1.008 , ..............................(3)

141.5

42.43

= , ..............................(4) = , ..............................(5) =
1030 [() 2 1

( = 0)]2 , ..............................(6)

() = =1 (), ..............................(7)

Page 10 of 11 Raw Data

TABLE 4SQUARE MEAN DISPLACEMENT OF MOLECULES AT 10 SECONDS INTERVAL AT 273.0K TEMPERATURE AND 1.028 MPA PRESSURE Square mean displacement of molecules (Pas)2 172.06 172.08 172.09 172.04 172.1 172.24 172.08 172.04 172.15 172.09 172.04 172.08 172.06 172.12 172.11 172.03 172.04 172.09 172.12 172.02

Station 6

60C

Crude oil 1 U909

Crude oil 2 T375

Diesel Z753 100 50 0.004141 0.004124

Size Constant @ 40C Constant @ 100C mm2/s2 (cSt/s) mm /s (cSt/s) Time 1 2 API Gravity (API) Station 1 Size Constant @ 40C Constant @ 100C 23C (Ambient temp) mm2/s2 (cSt/s) mm2/s2 (cSt/s) 6:47:04 6:21:68
2 2

100 0.01584 0.01576

0.01609 0.01601

4:21:60 4:06:15 33.7 38.40 Crude oil 2 P242 200 0.09521 0.09474

6:20:19 6:32:00

Crude oil 1 J276 200 0.1107 0.1102 Time 3:30:72 3:29:44 33.7

Diesel U166 100 0.01548 0.01541

1 2 API Gravity (API)

1:36:72 1:37:91 38.40

3:21:49 3:24:14

Page 11 of 11

Station 4

33C

Crude oil 1 M178

Crude oil 2 G505 200 200 0.1136 0.1131

Diesel U920 100 0.01486 0.01478

Size Constant @ 40C Constant @ 100C mm2/s2 (cSt/s) mm2/s2 (cSt/s) Time 1 2 API Gravity (API) 23C (Ambient temp) Volume mL Diesel 2:31:19 2:31:12

0.09092 0.09045

1:02:99 1:03:20 33.7 Water 150 150 38.40

3:09:94 3:09:43 40.00 Emulsifier 2

Emulsion Time 1 2 Equipment Constant Constant @ 40C Constant @ 100C API Experiment (Hydrometer) Crude oil 1 API 33 Crude oil 2 35.9 Diesel 39.5 Water 10.4 mm2/s2 (cst/s) mm /s (cst/s)
2 2

1:38:40 1:59:90

1:38:41 1:59 1:59:50 Equipment size 1.202 1.196 400

Weight Measurement (Each pycnometer is 10 mL) Crude oil 1 Dry weight Measured weight 1 2 3 25.3 25.4 25.4 25.3 25.3 25.3 25 24.9 24.9 16.5 Crude oil 2 16.4 Diesel 16.3

Experiment 3 Statistical Analyses of Core Data

PE 4521 002Reservoir Fluid Mechanics Laboratory Team 002E

Role Manager Researcher Technician Analyst Name Lemmy Oshenye Lucas Gurgel De Carvalho Nor Ashraf Norazman Axel Hannenberg Performance Score 1.0 1.0 1.0 1.0 Signature

Academic Integrity Statement

On my honor, I affirm that I have neither given nor received inappropriate aid in the completion of this exercise. Name: _________________________________ Date: _____________________________

Page 2 of 21 Abstract Statistical analyses are an integral part of modern reservoir analysis; yet engineers frequently use only the average quantities for porosity and connate water saturation to estimate the oil in place volume of the reservoirs. Statistical core analyses allow the team to correlate values of permeability, porosity, and water saturation values among themselves. Histograms and cross plots of permeability and water saturation vs. porosity are plotted to interpret whether the data from both cores indicates similar formation origins or not. Data analysis shows that: 1. For both wells, the average values for permeability, porosity, and water saturation are unique according to their zones. For each zone, there is a different trend for each parameter. 2. XAN 5 has a significant frequency of larger pore spaces with limited connectivity. The water and oil saturation distributions are compared with each other in the results. XAN 5 has a higher frequency of high oil saturation and low water saturation. XAN 12 has a significant frequency of intermediate size pore spaces with a noticeable numbers of small pore spaces. The small pore spaces have high connectivity between the pores, which may indicate the pore throats connect the smaller pores. 3. The cross plots saturation, permeability and porosity seem to indicate a positive relation between saturation and permeability. Permeability is not necessarily to be log-normally distributed. The porosity and water saturation frequency distributions do not appear to be from the same population. The teams recommendation is to conduct a seismic survey to analyze the type of formation for the wells. With the absence of relative permeability curves, it is hard to determine whether the formations are waterwet or oil-wet rocks.

Page 3 of 21 Introduction Oftentimes statistical analyses are done improperly, or even worse, neglected in the petroleum industry. For example, engineers frequently use only the average quantities for porosity and connate water saturation to estimate the oil in place volume of the reservoirs. This simplification of data can be due to the lack of time, the lack of data, economic reasons and so on. Statistical analyses can be very important to give a more precise understanding about the reservoir properties, thus providing companies with more certainty when making critical production decisions. For instance, statistical analysis can provide necessary input data for numerical modeling, or it can be implemented in subsequent analysis such as reservoir analyses and management using simulators. Simulations are the current trend in reservoir engineering and the use of statistical data will only increase as companies wish to exploit more complicated reservoirs. Therefore, it is very important for engineers to develop an insight into the use of statistics in petroleum engineering. It is very common that the reservoir exhibit heterogeneity and anisotropy, especially the unconventional reservoirs. When reservoirs do not exhibit heterogeneity and anisotropy, it becomes very important to examine the effects and the problems that this may cause to producing the reservoir. Some companies offer rock analyses for characterization of reservoirs which includes statistical analyses. Basically, analysis starts with taking core samples of different wells within a predefined area and then measuring their degree of similarity by analyzing the statistical distributions of rock properties. Relationships are then defined between the parameters if they are dependent, and develop functions of the properties with respect to the position within the reservoir. Finally, a reference model can then be used for any well within a certain range. This model will be continuously compared with other samples taken from other wells to quantify the overall confidence of this previous model. Typically, the model may start with only a few wells and have a high degree of confidence limited to the neighborhood of the cored wells. As new samples are obtained, they might update the model to increase the region of confidence (Schlumberger 2011; Trudgen and Hoffmann 1967). Core data of two wells are analyzed. The team looks to verify if the data indicated of the two wells are from the same reservoir by analyzing the frequency distributions. In addition, relationships between the parameters were drawn by applying regression techniques to add additional credence to the inferences made from frequency distributions. Experimental Procedure Use data from the core reports for two wells to estimate representative values of porosity and permeability. Plot the porosity, permeability, water and oil saturations frequency distributions and analyze the figures. Plot the natural logarithm of permeability vs. porosity and water saturation vs. porosity. Use suitable regression techniques in finding the relationship between the parameters.

Page 4 of 21 Results

kair,max Histogram
45 40 35 30 25 20 15 10 5 0

XAN 5 XAN 12

Frequency

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Fig. 1Histogram showing the maximum permeability distribution for XAN 5 and XAN 12 wells. Both of the wells display skewed right distribution.

Histogram
14 12 10 8 6 4 2 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

XAN 5 XAN 12

Frequency

17

18

Fig. 2Histogram showing the porosity distribution for XAN 5 and XAN 12 wells. Both of the wells display slightly skewed left distribution. XAN 12 has a noticeable frequency of small pores.

Page 5 of 21

Soil Histogram
60

XAN 5 XAN 12

50
Frequency
40 30 20 10 0 1 2 3 4 5 6

Fig. 3Histogram showing the oil saturation distribution for XAN 5 and XAN 12 wells. XAN 5 shows skewed left distribution and XAN 12 shows slightly skewed right distribution.

Swater Histogram
35 30 25 20 15 10 5 0

XAN 5 XAN 12

Frequency

10

Fig. 4Histogram showing the water saturation distribution for XAN 5 and XAN 12 wells. XAN 5 shows skewed right distribution and XAN 12 shows a normal distribution.

Page 6 of 21

XAN 5: -kair,max
10,000 1,000

y = 20.561e0.098x R = 0.0567

kair,max, md

100 10 1 5 7 9 11 13 , % 15 17 19

XAN 12: -kair,max

10,000 1,000 100

y = 2.7469e0.1234x R = 0.1173

kair,max, md

10 1 0 0 0 , % 2 7 12 17 22

-3

Fig. 5Exponential regression technique is used for both wells. The coefficient of determination R2 which relates to the goodness of fit of data is not close to 1.0. The statistical model fits better for XAN 12 well compared to XAN 5 well.

Page 7 of 21

XAN 5: -Swater
35 30 25

y = -0.4039x + 19.457 R = 0.0466

Swater, %

20 15 10 5 0

11

13
, %

15

17

19

XAN 12: -Swater

100 90 80 70 60 50 40 30 20 10 0 0 5 10 , % 15

y = 0.9685x + 36.786 R = 0.0584

Swater, %

20

Fig. 6Linear regression technique is used for both wells. The coefficient of determination R2 which relates to the goodness of fit of data is not close to 1.0. The statistical model fits better for XAN 12 well compared to XAN 5 well.

Page 8 of 21

XAN 5: -Soil
60 50 40

y = 0.9778x + 20.982 R = 0.0878

Soil, %

30 20 10 0

11

13
, %

15

17

19

XAN 12: -Soil

50 45 40 35 30 25 20 15 10 5 0

y = -1.0828x + 24.055 R = 0.2341

Soil, %
-3

7
, %

12

17

22

Fig. 7Linear regression technique is used for both wells. The statistical model fits better for XAN 12 well compared to XAN 5 well. The oil saturation shows an upward trend for XAN 5 and a downtrend for XAN 12.

Page 9 of 21

XAN 5: Lorenz Plot

1.0

Normalized kh

0.8
0.6 0.4 0.2 0.0 0.0 0.2 0.4 0.6 0.8 1.0 Normalized h

XAN 12: Lorenz Plot

1.0

Normalized kh

0.8 0.6 0.4 0.2 0.0 0.0 0.2 0.4 0.6 0.8 1.0 Normalized h

Fig. 8Lorenz plots for both wells to determine the heterogeneity of the formations. The Lorenz coefficients are 0.76 and 0.75 for XAN 5 and XAN 12, respectively. Both of the wells might be highly heterogeneous formations.

Discussion of Results The permeability distribution is compared with the porosity distribution for both wells (Figs. 1 through 4). XAN 5 has a significant frequency of larger pore spaces with limited connectivity. The water and oil saturation distributions are compared with each other. XAN 5 has a higher frequency of high oil saturation and low water saturation. The analysis indicates there are more hydrocarbons with low connate water saturation trapped within the unconnected large pore spaces. It is recommended that the engineers fracture the reservoir to connect the pores and then produce the hydrocarbons economically. XAN 12 has a significant frequency of intermediate size pore spaces with a noticeable numbers of small pore spaces. The small pore spaces have high connectivity between the pores, which may indicate the pore throats connect the smaller pores. By comparing the water and oil saturations, XAN 12 has higher frequency of

Page 10 of 21 high water saturation and low oil saturation. The analysis indicates that the well has most of the pore spaces containing water with low oil saturation. It is recommended that the drilling and production team drill and complete the well with caution to avoid fracturing or damaging the reservoir. Fracturing of the reservoir has the potential to cause high water production. Without relative permeability curves, it is difficult to definitively determine whether the formations are water-wet or oil-wet rocks. More data and further analysis are needed if the reservoir engineers are to determine the feasibility of waterflooding. The porosity and water saturation frequency distributions do not appear to be from the same population. The location of both wells is Guatemala. However, the histograms of different properties of both wells show different frequency diagrams. The formation of both wells might be formed by different geologic processes or at different times. The team is recommended to conduct a seismic survey to analyze the type of formation for the wells. The plotted porosity and permeability for both wells show there is a relationship between the two parameters (Figs. 5 through 7). By using exponential regression, the coefficient of determination R2 of XAN 5 is 0.0567, which does not truly indicate an accurate model to determine the relationship. The model predicts the permeability (md) is a function of porosity (%) for XAN 5 by the following equation: ln() = 0.098 + 3.0234, ..............................(1) Similar regression technique with R2 value of 0.1173 is used for XAN 12 and the relationship between permeability and porosity is shown by the following equation: ln() = 0.1234 + 1.0105, ..............................(2) The permeability used in the equations is the maximum permeability at the direction of the maximum permeability. The direction of maximum permeability can be obtained by solving the horizontal and vertical permeabilities using vector analysis. The goodness of fit of the models shows that the permeability is not necessarily log-normally distributed. The porosity and water saturation for both wells are related using a linear regression model. The R2 value is 0.0466 for XAN 5 and the model (in percentages) shows the relationship is as follows: = 0.4039 + 19.457, ..............................(3) The R2 value for XAN 12 is 0.0584 and the porosity and water saturation are related from the following equation: = 0.9685 + 36.786, ..............................(4) For XAN 5, water saturation decreases as the pore size increases. XAN 12 shows an opposite trend, in which the water saturation increases as the pore sizes increases. A simple linear regression technique is used to determine the general trend of oil saturation with porosity. Fig. 7 displays an upward trend for XAN 5 but XAN 12 has a downward trend for oil saturation with increase in porosity. There is more probability of the bigger pore spaces containing more organics for XAN 5 but there is more probability of smaller pore spaces containing more organics for XAN 12.

Page 11 of 21 However, interpretation of the probability should be used carefully since the coefficient of determination R2 values are not close to 1.0. The relationships developed from the models should be used with caution while performing reservoir simulation for both wells. The coefficient of determination R2 is used to indicate the quality of the results when running the simulation from the predicted models. To quantify the reservoir heterogeneity, Lorenz coefficient is determined from Fig. 8. The permeability is arranged from highest to the lowest, and the normalized flow capacity is plotted against the normalized the fraction of total volume. The Lorenz coefficient can range from zero for a homogeneous formation to one for a completely heterogeneous formation. Both of the wells might be highly heterogeneous with = 0.76 for XAN 5 and = 0.75 for XAN 12. The limitation of this method is that the Lorenz coefficient is not uniquely related to the permeability distribution (Craig Jr. 1970, 1239). Information about the heterogeneity of the reservoir can help the engineers to consider the nonuniform nature of the reservoir while estimating recovery during an immiscible displacement. Analysis of permeability, porosity, water and oil saturations interactions are done at respective depths by superimposing two plots containing two different parameters in one figure. The intention of superimposing the plots is to observe similarities and anomalies with each parameter at specific range of depths for both wells. For XAN 5, significant anomalies are observed at depths of 7,620 to 7,630 ft. At these depths, the formation has high permeability with low oil saturation and high water saturation. The porosity varies from 6 to 12%. The small pore spaces with high permeability might indicate pore throats. The high water saturation within the pore throats might indicate that the particular zone is a water-wet formation. Between 7,630 ft and 7,670 ft, the average porosity is approximately 15%, and the oil saturation varies from 19.8 to 46.9% with an average of 36.9%. The average permeability is approximately 102 md and average water saturation is 13%. This might indicate that the formation in the zone has larger pores that contain mainly hydrocarbons with good connectivity. From 7,670 to 7,685 ft, there is a noticeable drop in porosity, oil saturation, and a decline in permeability with an increase in water saturation. This might related to the water-wet formation in this zone. Between 7,685 ft and 7,710 ft, all the parameters show high variations; this might be due to the presence of anhydrites in the formation. For XAN 12, between 7,714 ft and 7,722 ft, the presence of asphalt might contribute to low porosity and high oil saturation. Within this zone, high grain density values are observed. Substantial amounts of clay might be present within this zone; however, water saturation is relatively low compared to oil saturation. The team could not determine what causes the increase in grain density. At 7,772 ft, there is a sudden drop in porosity and water saturation. From 7,796 to 7,819 ft, the porosity drops to an average of 2.33% with average oil saturation close to 19%. At depths from 7,742 to 7,790 ft, the formation has high water saturation low oil saturation. It is best to avoid perforating in this zone because the well will produce water which is not favorable for the operating company.

Page 12 of 21 Conclusion Statistical analysis of core data shows that air permeability measurements as practiced in the industry have a systematic negative bias that must be addressed to improve overall core analysis results (Thomas and Pugh 1989). Reservoir rock properties are a sample of quantitative statistical observations that can be condensed and represented by frequency distributions which may be described by calculated statistical parameters (Trudgen and Hoffmann 1967). Based on analysis, the team concludes that: 1. Correlations involving water saturation, porosity, and permeability are enhanced when their distributions are known and accounted for. 2. Permeability is not necessarily log-normally distributed. 3. The porosity and water saturation frequency distributions do not appear to be from the same population. 4. The average values for permeability, porosity, and water saturation are unique according to position in the formation. For each zone, there is a different trend for each parameter.

Page 13 of 21 References Craig Jr., F.F. 1970. Effect of Reservoir Description on Performance Predictions. Journal of Petroleum Technology: 12391245. SPE-2652-PA. http://dx.doi.org/10.2118/2652-PA. Schlumberger. 2011. Terratek Tight Rock Analysis, http://www.slb.com/~/media/Files/core_pvt_lab/product_sheets/terratek_tight_rock_analysis_ overview_ps.pdf (downloaded 16 September 2012). Thomas, D.C. and Pugh, V. J. 1989. A Statistical Analysis of the Accuracy and Reproducibility of Standard Core Analysis. The Log Analyst 30 (2): 7177. SPWLA. Trudgen, P. and Hoffmann, F. 1967. Statistically Analyzing Core Data. J. Pet Tech 19 (4): 497503. SPE-1574-PA. http://dx.doi.org/10.2118/1574-PA. Appendices Equations = = =
=1 =1

, ..............................(5) , ..............................(6) , ..............................(7)

=1 =1

=1 =1

= =1

, ..............................(8) , ..............................(9)

= =1

= , ..............................(10) = , ..............................(11)

= =1 , ..............................(12)
2 2 = 1 =1( ) , ..............................(13) 1

= (1)3, ..............................(14) = (1)4, ..............................(15)

( )4

( )3

Page 14 of 21 Supplemental Figures and Tables

XAN 5: -Swater
25 20

Swater 35 30 25 20 15 10 5 0 7,720

, %

15 10 5 0 7,620

7,640

7,660

7,680

7,700

Depth, ft Soil 60 50 40

XAN 5: -Soil
25 20

30
10 20 10 7,640 7,660 7,680 7,700 0 7,720

5
0 7,620

Depth, ft Kair,max 10,000 1,000 100 10 5 0 7,620 10 1 7,720

XAN 5: -kair,max
25 20

, %

15

7,640

7,660

7,680

7,700

Depth, ft

kair,max, md

Soil, %

, %

15

Swater, %

Page 15 of 21

XAN 5: -grain
25 20

grain 2.87 2.85 2.83

, %

15

10
5 0 7,620

2.81
2.79 2.77 7,720

7,640

7,660

7,680

7,700

Depth, ft Soil Swater 35 30 25 20 15 10 5 7,640 7,660 7,680 7,700 0 7,720

XAN 5: Soil-Swater
60 50 40

30 20 10 0 7,620

Depth, ft Soil Kair,max 10,000 1,000

XAN 5: Soil-kair,max
60 50 40

30
20 10 0 7,620 7,640 7,660 7,680 7,700

100
10 1 7,720

Depth, ft

kair,max, md

Soil, %

Swater, %

Soil, %

grain, g/cm3

Page 16 of 21

XAN 5: Swater-kair,max
35 30 20 15 10 5 0 7,620 7,640 7,660 7,680 7,700 25

Swater Kair,max

10,000 1,000 100 10 1 7,720

Depth, ft So+w

XAN 5: So+w-kair,max
70 60 40 30 20 10 0 7,620 7,640 7,660 7,680 7,700 50

Kair,max 10,000
1,000 100 10 1 7,720

Depth, ft So+w 70 60 50 40 30 20 10 7,640 7,660 7,680 7,700 0 7,720

XAN 5: -So+w
25 20

10 5 0 7,620

Depth, ft

So+w, %

, %

15

kair,max, md

So+w, %

kair,max, md

Swater, %

Page 17 of 21

XAN 12: -Swater

25 20

Swater 120 100

, %

15 10 5 0 7,670

60 40 20 0 7,720 7,770 Depth, ft Soil 50 40 7,820

XAN 12: -Soil

25 20

10 5 0 7,670

20 10 0 7,720 7,770 Depth, ft Kair,max 10,000 1,000 100 7,820

XAN 12: -kair,max

25 20

, %

15 10 5 0 7,670 10 1 7,720 7,770 Depth, ft 7,820

kair,max, md

Soil, %

, %

15

30

Swater, %

80

Page 18 of 21

XAN 12: -grain

25 20

grain 2.87 2.85 2.83

, %

15

10
5 0 7,670

2.81
2.79 2.77 7,720 7,770 Depth, ft Soil Swater 100 80 60 40 20 0 7,820

XAN 12: Soil-Swater

50 40

Soil, %

30 20 10 0 7,670

7,720

7,770
Depth, ft

7,820

XAN 12: Soil-kair,max

50 40

Soil Kair,max 10,000 1,000 100

Soil, %

30 20 10 0 7,670 10 1 7,720 7,770 Depth, ft 7,820

kair,max, md

Swater, %

grain, g/cm3

Page 19 of 21

XAN 12: Swater-kair,max

100 80

Swater
Kair,max

10,000 1,000 100

60 40 20 0 7,670 10 1 7,720 7,770 Depth, ft So+w Kair,max 10,000 1,000 100 40 20 0 7,670 10 1 7,720 7,770 Depth, ft So+w 100 80 7,820 7,820

XAN 12: So+w-kair,max

100 80

So+w, %

60

XAN 12: -So+w

25 20

10 5 0 7,670

40 20 0 7,720 7,770 Depth, ft 7,820

So+w, %

, %

15

60

kair,max, md

kair,max, md

Swater, %

Page 20 of 21
35 30 25 10,000

Water Saturation Air Permeability

XAN 5: Cross Plot

1,000

Swater, %

20 100 15 10 5 10

0
0 5 10 , % 15 20 25

100 90 80 70

Water Saturation Air Permeability

XAN 12: Cross Plot

10,000.00 1,000.00 100.00 10.00 1.00 0.10 0.01 0.00

50 40 30 20 10 0 0 5 10 , % 15 20 25

kair,max, md

Swater, %

60

kair,max, md

Page 21 of 21
TABLE 1DESCRIPTIVE STATISTICS OF PERMEABILITY, POROSITY, OIL AND WATER SATURATIONS FOR XAN 5 Parameter Mean Standard Error Median Mode Standard Deviation Sample Variance Kurtosis Skewness Range Minimum Maximum Sum Count Confidence Level (95.0%) kair,max 262.0 86.1 75 61.5 774.8 600,322.3 28.9 5.2 5,128.36 8.64 5137 21,223.4 81 171.3 13.3 0.4 13.7 12.8 3.2 9.9 -0.6 -0.3 13.8 5.4 19.2 1,080.8 81 0.7 Soil 34.0 1.2 35.9 39.4 10.4 108.3 1.1 -1.0 50.2 2.5 52.7 2,756.4 81 2.3 Swater 14.1 0.7 13.4 20.1 5.9 34.8 -0.1 0.3 28.6 2.5 31.1 1,139.5 81 1.3

TABLE 2DESCRIPTIVE STATISTICS OF PERMEABILITY, POROSITY, OIL AND WATER SATURATIONS FOR XAN 12 Parameter Mean Standard Error Median Mode Standard Deviation Sample Variance Kurtosis Skewness Range Minimum Maximum Sum Count Confidence Level (95.0%) kair,max 71.2 27.2 12.8 6.68 264.8 70,139.8 42.1 6.1 2,131.995 0.005 2132 6,764.5 95 54.0 11.6 0.5 12.5 11 5.2 26.6 -0.6 -0.5 20.1 1.5 21.6 1,106.3 95 1.1 Soil 11.4 1.2 8.5 1.3 11.5 133.0 0.4 1.1 46.1 0.5 46.6 1,087.3 95 2.3 Swater 48.1 2.1 47.7 29.8 20.7 426.7 -0.8 0.2 80.8 12.1 92.9 4,566.2 95 4.2

Experiment 4 Volumetric Estimation of OOIP

PE 4521 002Reservoir Fluid Mechanics Laboratory Team 002E

Role Manager Researcher Technician Analyst Name Lucas Gurgel De Carvalho Nor Ashraf Norazman Axel Hannenberg Lemmy Oshenye Performance Score 1.0 1.0 1.0 1.0 Signature

Academic Integrity Statement

On my honor, I affirm that I have neither given nor received inappropriate aid in the completion of this exercise. Name: _________________________________ Date: _____________________________

Page 2 of 10 Abstract The calculation of a reservoirs original oil in place (OOIP) is a vital task in determining a reservoirs value and potential. There are many different ways to approach calculating a reservoirs OOIP, and it can be useful to utilize a combination of these. In this experiment, the team uses both deterministic and stochastic approaches to estimate OOIP in the Booch play, discovered in 1961 in Hughes County, Oklahoma. The Booch oil reservoir is a channel sand formed from fluvial deposition. In the deterministic approach, trapezoidal, Simpsons, and pyramidal frustum methods are used to calculate the OOIP. Trapezoidal method yields the highest estimated OOIP compared to Simpsons and pyramidal frustum methods. Pyramidal frustum overestimated the OOIP compared to Simpsons rule but underestimated the OOIP compared to trapezoidal rule. The approaches used are made under the assumption that the Booch reservoir is fairly homogeneous. However, fluvial reservoirs are often very heterogeneous due to their discontinuity and facies changes during the depositional process. Inaccurate modeling and various sources of error may have contributed to the large differences between the data provided and the three methods values, which is approximately 0.7 million STB. Human error can cause random error to occur when using the planimeter. Taking measurements using the planimeter requires high skill and precision. The error of the measurements is amplified when converting the unit from in.2 to ft2 to acre. The magnified error is then carried in the volumetric calculations. Random error can be minimized by repeating the measurements and then averaging the values. Human error aside, numerical approaches are utilized to obtain an accurate OOIP estimate. The uncertainties in geological and geophysical data have an impact in the estimation of reserves. The team needs to obtain the original geophysical and geological data of the play to track the major contributor of the overall uncertainty to get a more accurate estimate of OOIP.

Page 3 of 10 Introduction The determination of the amount of fluid existing in petroleum fields is essential for the correct implementation of an exploratory project. The basic concepts of volumetric estimation of OOIP are simple, yet imprecise. For instance, if the reservoir had a simple square shape, a defined thickness and constant values for all the parameters, the volumetric estimation would be done using Eq. 1. Volumetric calculations are made less accurate because of the complex reservoir shapes, as well as changes in the values of parameters needed such as water saturation, oil saturation and porosity. Unconventional reservoirs are more likely to have complex shapes and the parameters are not likely to always be constant. In order to analysis more sophisticated reservoirs volumetric estimation should be done through integrations of the functions of the reservoir shapes. Since these functions are virtually impossible to determine, estimates are by default only numerical approximations. The volumes of oil or gas of the reservoir may be computed in different ways. One of them is when the geoscientists prepare a geological map from the subsurface data; for example, from seismic imaging and core analysis. These maps are known as isopach maps. Contours of the hydrocarbon rich areas are made from the top to the bottom of the reservoir, which allows for better indications of the corresponding thicknesses of these areas. The water-oil contacts are also shown. The areas are then measured with a planimeter, converted to the real scale of the reservoir, and finally the volume is estimated by numerical integration (Garb 1985). The most common numerical methods are the trapezoidal rule, which is less accurate, the Simpsons rule, which is more accurate but requires an even number of contoured areas, and the frustum of a pyramid. These three methods are used to calculate the volume of OOIP of a reservoir and the values obtained are compared with each other and with the original reservoir data provided. Additionally, volumes of oil or gas can be estimated through material balance equations (MBE). These equations depend on the drive mechanisms acting in the reservoir, and allow for how a reservoir fluids mass balance changes with pressure (Rosa 2011). MBEs are written in terms of the rock properties and the fluid behavior as a function of pressure, rock-fluid properties, and production history. Experimental Procedure Use a planimeter for this experiment. The planimeter is used to trace along the boundary of a given contour line on the net pay isopach map. When using the planimeter, pick a point on the desired interval and carefully trace along the boundary of that interval clockwise. The planimeter finds the areas between the isopach contours (Dean 2007). With a conversion in the map legend, convert the square inches given by the planimeter to acres. Once the acreages of the different levels are known along with the heights, use one of several numerical methods to calculate the volume. Use frustum of a pyramid, Simpsons rule, or the trapezoidal rule to calculate a volume which can be used in the volumetric equation along with values for porosity, water saturation and formation volume factor to get an estimate for the volume of stock-tank barrels (STB) in place.

Page 4 of 10

Fig. 1The planimeter. The view glass is the piece on the far right and is used to follow the contour lines of the isopach map. The wheels across the back on the left side provide stability when tracing along the contour horizon.

Results
TABLE 1ESTIMATED BULK VOLUME AND ORIGINAL OIL IN PLACE Vb Method Trapezoidal Simpsons Pyramidal Frustum Data Provided acre-ft 4,602 4,549 4,551 No data N million STB 2.545106 2.516106 2.518106 3.200106

Millions N, STB

2.550 2.545 2.540 2.535 2.530 2.525 2.520 2.515 2.510 2.505 2.500 Trapezoidal Simpson Pyramidal Frustum

Fig. 2Trapezoidal method presents the highest estimated OOIP. Estimated OOIP values for Simpsons and pyramidal frustum method are relatively similar to each other.

Page 5 of 10

Millions N, STB

3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Trapezoidal Simpson Pyramidal Frustum Data Provided

Fig. 3The estimated OOIP for the three methods are relatively similar with each other compared the data provided in the engineering table. The three methods underestimate the OOIP compared to the data provided.

Discussion of Results Trapezoidal method presents the highest estimated OOIP compared to Simpsons and pyramidal frustum methods. Pyramidal frustum overestimated the OOIP compared to Simpsons rule but underestimated the OOIP compared to trapezoidal rule. However, the difference in the estimated OOIP values from the three methods are insignificant compared to the OOIP provided from the engineering data. The difference between the three methods values to data provided is approximately 0.7 million STB. Craft et al. (1991) stated that whenever the ratio of the areas of any two successive isopach lines is less than 0.5, then pyramidal frustum is applied. However, if it is more than 0.5, then trapezoidal rule is applied. From the above statement, the team recommends the application of pyramidal frustum from the 60 to 50 ft contour interval because the ratio of the areas is smaller than 0.5. The team also recommends applying Simpsons rule on other intervals because the ratios are larger than 0.5 and Simpsons rule is usually more accurate than trapezoidal rule for irregular curves. Based on the representative electric log for the Booch oil reservoir, the thickness can be measured precisely at 10 ft intervals. The height interval impacts the volume calculations and thus affects the estimated OOIP. It is assumed that the area is uniform throughout each of the 10 ft interval. The uncertainty decreases if lower thickness interval is used. There are various sources of errors that contribute to cumulative error in the volumetric calculations. Human error can cause random error to occur when using the planimeter. Taking measurements using the planimeter requires high skill and precision. Since the measurement is recorded in in.2, the error is amplified when converting the unit from in.2 to ft2 to acre. The magnified error is then carried in the volumetric calculations. Random error can be minimized by repeating the measurements and then averaging the values.

Page 6 of 10 Uncertainties in geophysical data have an impact on the geological data, which leads to uncertainties in the estimated OOIP. It is very important to track the source with the largest uncertainty because that would have the largest impact on the overall uncertainty in the volumetric calculations (Floris and Peersmann 1998). Although the estimated OOIP is calculated using the provided porosity, water saturation, and formation volume factor values, the uncertainties in these data are not considered at all in the volumetric calculations. The largest source of uncertainty could be contributed from any of these parameters or the calculated bulk volumes from the numerical methods. At this point, the team needs to obtain the original geophysical and geological data of the play to track the major contributor of the overall uncertainty to get a more accurate estimate of OOIP. Conclusion and Recommendation Trapezoidal method presents the highest estimated OOIP followed by pyramidal frustum and Simpsons methods. However, the difference in the estimated OOIP values from the three methods are insignificant compared to the OOIP provided from the engineering data. The uncertainties in geological and geophysical data have an impact in the estimation of reserves. The team needs to obtain the original geophysical and geological data of the play to track the major contributor of the overall uncertainty to get a more accurate estimate of OOIP.

Page 7 of 10 References Andrews, R.D., Northcutt, R. A., Knapp, R. M. and Yang, X. H. 1995. Fluvial-Dominated Deltaic (FDD) Oil Reservoirs in Oklahoma: The Booch Play, Oklahoma Geological Survey, Special Publication 95-3, 167. Craft, B.C., Hawkins, M. and Terry, R.E. 1991, Applied Petroleum Reservoir Engineering, 2nd Edition, New Jersey: Prentice-Hall, 71. Dean, L. 2007. Volumetric Estimation. Reservoir Engineering for Geologists: 1114. Floris, F.J.T., and Peersmann, M.R.H.E. 1998. Uncertainty estimation in volumetrics for supporting hydrocarbon exploration and production decision-making, Petroleum Geosciences 4 (1): 3340. http://dx.doi.org/10.1144/petgeo.4.1.33. Garb, F.A. 1985. Oil and Gas Reserves Classification, Estimation, and Evaluation. Distinguished Author Series, J Pet Technol 37 (3): 373390. SPE-13946-PA. http://dx.doi.org/10.2118/13946-PA. Rosa, Adalberto Jose. 2011. Engenharia de Reservatorios de Petroleo, Rio de Janeiro, RJ, Brazil. Appendices Equations =
(1 ) ,

..............................(#)

Where = (7758), = , = ,

= , = , = , = (0 + 21 + 22 + + 21 + ) + , ..............................(#)
= (0 + 41 + 22 + 43 + + 22 + 41 + ) + , ..............................(#)
3 1

1 2

Where = , 0 = ( ), 1 = ,

Page 8 of 10 = , = , = , = 3 (0 + + 0 ), ..............................(#) Where 0 = , = , Supplemental Figures and Tables Some of the following tables and figures are taken from the Oklahoma Geological Survey 1995, Special Publication 95-3, Fluvial-Dominated Deltaic (FDD) Oil Reservoirs in Oklahoma: The Booch Play.
1

Fig. 4Isopach map of net Booch oil sand (pay sand), Booch oil reservoir, Greasy Creek field, Hughes County, Oklahoma.

Page 9 of 10

Fig. 5Representative electric log for the Booch oil reservoir, Greasy Creek field, Hughes County, Oklahoma, showing log patterns for spontaneous potential, resistivity, and conductivity measurements. Perforated interval is marked.

TABLE 2GEOLOGICAL/ENGINEERING DATA FOR THE BOOCH OIL RESERVOIR, GREASY CREEK FIELD, HUGHES COUNTY, OKLAHOMA Reservoir size Well spacing (oil) Oil/water contact Gas/oil contact Porosity Permeability Initial water saturation Thickness (net sand in reservoir) Reservoir temperature Oil gravity Initial reservoir pressure Initial formation volume factor Original oil in place (volumetric) Cumulative primary production Recovery efficiency Cumulative gas 140 acres 10 acres -1,555 ft below mean sea level none 15% 35 md 45% 40 ft average 103F 39.5API 940 psia 1.157 3.2 million STB 692,315 BO 22% no data

Page 10 of 10 Raw Data

TABLE 3AREAS RATIO AND AVERAGE DATA Reading 1 in.2 0.356501 0.697501 1.534503 2.681505 3.8905 5.440511 7.1145 8.959 Reading 2 in.2 0.3720007 0.6820013 1.519003 2.681503 3.7975075 5.2390104 6.9905139 8.7265 Average in.2 0.3642507 0.6897513 1.526753 2.681504 3.84400375 5.3397606 7.05250695 8.84275 0.511111 0.454545 0.572254 0.689244 0.715098 0.764707 0.794118 average 0.643011 Areas Ratio

TABLE 4CALCULATED STANDARD DEVIATIONS Vb Method Trapezoidal Simpson Pyramidal Frustum acre-ft 4,601.51 4,548.69 4,551.18 Vb 29.80 N STB 2,545,484.80 2,516,268.36 2,517,646.21 N 16,484.77

TABLE 5ESTIMATED BULK VOLUME AND ORIGINAL OIL IN PLACE Vb Method Trapezoidal Simpsons Pyramidal Frustum Data Provided acre-ft 4,602 30 4,549 30 4,551 30 No data N million STB 2.545106 1.6104 2.516106 1.6104 2.518106 1.6104 3.200106

Experiment 5 Hydrocarbon Phase Behavior

PE 4521 002Reservoir Fluid Mechanics Laboratory Team 002E

Role Manager Researcher Technician Analyst Name Nor Ashraf Norazman Axel Hannenberg Lemmy Oshenye Lucas Gurgel De Carvalho Performance Score 1.0 1.0 1.0 1.0 Signature

Academic Integrity Statement

On my honor, I affirm that I have neither given nor received inappropriate aid in the completion of this exercise. Name: _________________________________ Date: _____________________________

Page 2 of 16 Abstract It is important to determine the pressure-volume relationship of reservoir fluids because it helps to obtain the bubblepoint pressure. Knowing the bubblepoint pressure is significant in understanding the phase behavior of reservoir fluids by calculating integral parameters such as gas solubility, formation volume factor (FVF), and coefficient of isothermal compressibility. The evaluation and proper production of oil and gas reserves depends heavily on the knowledge of reservoir fluid properties. These properties can either be obtained from laboratory experiments or empirical correlations. In this study, experimental measurements and empirical correlations are incorporated to determine values of bubblepoint, gas solubility, FVF, and coefficient of isothermal compressibility. The gas solubility, FVF, and coefficient of isothermal compressibility are plotted as function of pressure at a constant temperature to observe the changes of each parameter with respect to pressure change. Data analysis shows that: 1. The bubblepoint pressures for the 32API oil, 38API oil, and CO2-water mixture are 807.2-, 597.6-, and 874.6-psi respectively. 2. At pressures below the bubblepoint, the coefficient of isothermal compressibility is higher due to the presence of the free gas. Once the gas is in solution, coefficient of isothermal compressibility varies only slightly with increasing pressure because liquids are slightly compressible. 3. The solution GOR above and at the bubblepoint for the 32API oil, 38API oil, and CO2-water mixture are 144.53-, 125.47-, 203.08-scf/STB. Solution GOR is constant at pressures above the bubblepoint because all gas is in solution. At pressures below the bubblepoint, solution GOR decreases because there is now free gas that has evolved out from the solution. 4. At pressures above the bubblepoint, the total FVF and FVF of liquid have the same value. The FVF of liquid decreases at pressures below the bubblepoint but the total FVF increases significantly. It is recommended to perform hydrocarbon analysis on both crude oils to have a better understanding of the compositional changes during the phase transition at pressures below the bubblepoint.

Page 3 of 16 Introduction Phase behavior is the behavior of vapor, liquid, and solids as a function of pressure, temperature, and composition (Whitson and and Brule 2000). As oil and gas are produced from the subsurface, they undergo changes of temperature, pressure and composition. Such changes affect the volumetric and transport behavior of the fluids directly impacting the surface volumes and quality of produced oil and gas (Devegowda 2011). Some physical properties such as gas solubility, formation volume factor, bubblepoint and coefficient of isothermal compressibility exist to help the study of hydrocarbon phase behavior. To determine these properties, experiments are usually performed on fluid samples using PVT cells and data is collected. PVT cells basically consist of a cylinder containing mercury in which the liquid sample to be studied is placed. The change from single-phase to two-phase affects the physical properties. There are two basic types of gas liberation: flash and differential (Rosa 2011). Flash liberation is when the gas that comes out of solution due to reduction in pressure is kept in contact with the liquid. Likewise, gas is said to be differentially liberated when the gas coming out of solution is removed from contact with the liquid. In this experiment, flash liberation is conducted for two different samples of oils and for a mixture of water and CO2. PVT analysis is integral in the calculation of reserves, production forecasts, and the efficiency of enhanced oil recovery methods. It is also useful for surface separator design and to calculate flow in pipe (Whitson and Brule 2000). Experimental Procedure The temperature is set to 60F. Set the pressure at approximately 2,000 psig for all samples. Reduce the pressure in 200 psig increments before reaching the bubblepoint and in 100 psig increments or lower when the pressure is close to the anticipated bubblepoint. For each reading, the pressure is allowed to stabilize for 3 minutes before recording the pressure and volume. Pressure is plotted against volume and the pressure at which the slope changes is the bubblepoint pressure. This experiment is performed using CO2-water mixture, 32API crude oil, and 38API crude oil. The bubblepoint pressure and the corresponding volume are observed and recorded and used as a reference volume (Ahmed 2001).

Page 4 of 16 Results
Pressure-Volume for 32API Oil
2,500

2,000

Pressure, psig

1,500

1,000 y = -284.96x + 43132

500
y = -4.6454x + 1497.2 0 142 144 146 148 150 152 154 156 Volume, mL
Fig. 1The bubblepoint for 32API oil is 807.2 psi at 148.5 mL at 60F temperature.

Pressure-Volume for CO2-Water Mixture

2,500

2,000

Pressure, psig

1,500

1,000

y = -282.85x + 40564

500

y = -3.3361x + 1342.7

0 136 137 138 139 140 141 Volume, mL

Fig. 2The bubblepoint for water with dissolve carbon dioxide is 874.6 psi at 140.3 mL at 60F temperature.

142

143

144

145

146

Page 5 of 16

Pressure-Volume for 38API Oil

2,500 2,000

Pressure, psig

1,500 1,000

y = -386.58x + 54778 y = -90.975x + 13348

500
0 136 137 138

139 Volume, mL

140

141

142

Fig. 3The bubblepoint for 38API oil is 597.6 psi at 140.2 mL at 60F temperature.

TABLE 1 BUBBLEPOINTS FOR THREE FLUIDS Fluid Type Bubblepoint, psig 32API Oil 807.2 CO2-Water Mixture 874.6 38API Oil 597.6

Coefficient of Isothermal Compressibility of 32API Oil 0.0006 0.0005 0.0004

co, psi-1

0.0003

0.0002 0.0001 0
700 900 1,100 P, psig 1,300 1,500 1,700

pb

Fig. 4The coefficient of isothermal compressibility of 32API oil is relatively constant above the bubblepoint and increases significantly below the bubblepoint at constant temperature.

Page 6 of 16

Coefficient of Isothermal Compressibility of CO2-Water Mixture 0.00035

0.0003
0.00025

cw, psi-1

0.0002 0.00015 0.0001 0.00005 0 800 900 1,000 1,100 1,200 1,300 1,400 1,500

pb

P, psig

Fig. 5The coefficient of isothermal compressibility of water is relatively constant above the bubblepoint and increases significantly below the bubblepoint at constant temperature.

Coefficient of Isothermal Compressibility of 38API Oil 1.20E-03 1.00E-03

8.00E-04
co, psi-1

6.00E-04 4.00E-04 2.00E-04 0.00E+00

400 600 800 1,000 1,200 1,400 1,600 1,800

pb

P, psig

Fig. 6The coefficient of isothermal compressibility of 38API oil is relatively constant above the bubblepoint and increases significantly below the bubblepoint at constant temperature.

Page 7 of 16

Solution Gas/Oil Ratio for 32API Oil 145 144 143

Rs, scf/STB

142
141 140 139 138 700 900 1,100 1,300 1,500 1,700 1,900 2,100

pb

P, psig

Fig. 7The solution gas/oil ratio of 32API oil is constant above the bubblepoint and decreases significantly below the bubblepoint at constant temperature.

Solution Gas/Liquid Ratio for CO2-Water Mixture 203.5 203 202.5 202 201.5 201 200.5 200 199.5 199 198.5 198 800 1,000 1,200 1,400 P, psig 1,600 1,800 2,000

Rs, scf/STB

pb

Fig. 8The solution gas/liquid ratio of water is constant above the bubblepoint and decreases significantly below the bubblepoint at constant temperature.

Page 8 of 16

Solution Gas/Oil Ratio for 38API Oil 130 125 120 115 110 105 100 95 90 85 80 400 600 800 1,000 1,200 P, psig 1,400 1,600 1,800 2,000

Rs, scf/STB

pb

Fig. 9The solution gas/oil ratio of 38API oil is constant above the bubblepoint and decreases significantly below the bubblepoint at constant temperature.

Formation Volume Factor for 32API Oil

Formation Volume Factor, res bbl/STB

1.06 1.055

1.05 1.045 1.04 1.035 1.03 1.025 1.02

700 900 1,100 P, psig 1,300 1,500 1,700 Bt Bo

pb

Fig. 10The total formation volume factor of 32API oil increases significantly below the bubblepoint at constant temperature.

Page 9 of 16

Formation Volume Factor for CO2-Water Mixture

Formation Volume Factor, res bbl/STB

1.11 1.105 1.1 1.095 1.09 1.085 800 900 1,000 1,100 1,200 1,300 1,400 1,500 Bt Bw

pb

P, psig

Fig. 11The total formation volume factor of water increases as pressure decreases at constant temperature.

Formation Volume Factor for 38API Oil

Formation Volume Factor, res bbl/STB

1.20000 1.18000 1.16000 1.14000 1.12000 1.10000 1.08000 1.06000 1.04000 1.02000 1.00000 400 600 800 1,000 1,200 1,400 1,600

Bt Bo

1,800

pb

P, psig

Fig. 12The total formation volume factor of 38API oil increases significantly below the bubblepoint at constant temperature.

Page 10 of 16 Discussion of Results At 60F temperature, the bubblepoint is the highest for CO2-water mixture followed by the 32API oil and 38API oil. The 32API oil has higher bubblepoint pressure compared to 38API oil due to higher amount of heavy hydrocarbon components within the oil. It is assumed that methane gas is evolved when the oils reach the bubblepoint. It is also assumed that the pressure, temperature and z-factor at standard conditions are 14.65 psia, 520R and 1.0 respectively. It is assumed that the correlations used for hydrocarbons can be applied to CO2-water mixture. However, the values should be used with caution because CO2-water mixture has different properties than hydrocarbons. It is assumed that the fluids in the PVT cells mimic the behavior of the reservoir fluids. The liquid is said to be undersaturated (singlephase) at pressures above the bubblepoint and saturated (two-phase) when the pressure is at and below the bubblepoint. The evolution of gas from the oil has a significant impact on the FVF. This causes a large decrease in volume of the oil when there is a lot of gas. As pressure decreases, the oil expands slightly at constant temperature. When pressure decreases below the bubblepoint, gas evolve within the cell and the remaining oil has less gas in solution, thus resulting in a lower FVF. The laboratory analysis indicates the FVF of both oils are below 2.0 res bbl/STB, which is consistent with a typical black oil. At pressures above the bubblepoint pressure, the solution GOR is constant due to no gas evolved inside the cell. When pressure is reduced below the bubblepoint pressure, gas evolves in the cell making less gas dissolve in the liquid thus decreasing the solution GOR (McCain 1990). Similar gas solubility behavior is shown for CO2-water mixture. The total FVF and the FVF of oil are identical at pressures above the bubblepoint because no gas evolves from the solution inside the cell. At pressures below the bubblepoint, the total FVF is more than the FVF of oil because of the evolution of gas. The evolved gas is called free gas. The difference between the total FVF and the FVF of liquid at pressures below the bubblepoint represent the volume of gas released in the cell. The coefficient of isothermal compressibility of liquid at pressures above the bubblepoint is virtually constant except at pressures near the bubblepoint. This finding is supported by McCain (1990). Below the bubblepoint pressure, the volume of liquid decreases as pressure is reduced. There is a large shift in the compressibility because of the evolution of gas. It is important to determine the pressure-volume relationship of reservoir fluids because it helps to obtain the bubblepoint pressure. Knowing the bubblepoint pressure is significant in understanding the phase behavior of reservoir fluids by calculating integral parameters such as gas solubility, FVF, and coefficient of isothermal compressibility. The bubblepoint and relative volumes can be used as inputs in tuning equation of state which is useful in reservoir modeling or simulation. The team recommends performing hydrocarbon analysis on both crude oils to have a better understanding of the compositional changes during the phase transition at pressures below the bubblepoint.

Page 11 of 16 Conclusion The bubblepoint of a fluid can be determined by the significant change in slopes on a pressure vs. volume plot. The evolution of gas at bubblepoint has a significant impact on gas solubility, FVF, and coefficient of isothermal compressibility values for liquid. The team recommends performing hydrocarbon analysis on both crude oils to have a better understanding of the compositional changes during the phase transition at pressures below the bubblepoint.

Page 12 of 16 References Ahmed, T. 2001. Reservoir Engineering Handbook, second edition. Houston, Texas: Gulf Professional Publishing/Elsevier. Devegowda, D. 2011. Phase behavior. Lecture notes on phase behavior. The University of Oklahoma, Oklahoma, United States. McCain, W.D. Jr. 1990. The Properties of Petroleum Fluids, second edition, Tulsa, Oklahoma: PennWell. Rosa, Adalberto Jose. 2011. Engenharia de Reservatorios de Petroleo, Rio de Janeiro, RJ, Brazil. Whitson, C.H., and Brule, M.R. 2000. Phase Behavior, Vol. 20, 12. Richardson, Texas: Monograph Series, SPE. Appendices Equations = +131.5, ..............................(1) = , ..............................(2) = , ..............................(3)

141.5

= ( ) , ..............................(4)

1 [( )

) ],

..............................(5)

() = + 1.4, ..............................(6) 18.2 =

0.83

10(0.000910.0125)

()

0.83

+1.4 18.2 10(0.000910.0125)

, ..............................(7)

= (), ..............................(8) () = (

0.5

+ 1.25, ..............................(9)

5 = 0.9759 + 12 105 ()1.2 [ ( = 0.9759 + 12 10

0.5

1.2

+ 1.25] , ..............................(10)

14.65 1520

= 0.0282

3 / 5.615 3, ..............................(11)

= + ( ), ..............................(12)

Sample Calculations 32API Oil

Page 13 of 16 141.5 = 0.865 32 + 131.5

= 0.78 28.01 + 0.21 32 + 0.01 39.94 = 29 = 16.04 = 0.554 29

= 284.96 + 43132 = 4.6454 + 1497.2 Solve for and , = 807.2 = 148.5

0.83

807.2 18.2 + 1.4 0.554 = 144.5 / 10(0.00091600.012532)

5

= 0.9759 + 12 10 Read from chart

0.554 0.5 [144.5 ( ) + 1.25 60] 0.865

1.2

= 1.0413 /

= 0.912 = 0.912 (60 + 460) 14.65 = = 0.0029 / 1 520 (807.2 + 14.7) 5.615

= 1.0413 + 0.0029(144.5 144.5) = 1.0413 / Above bubblepoint pressure = 924.4 1033.2 = 108.8 = 148.2 147.7 = 0.5 = = 148.2 + 147.7 = 148.0 2

1 0.5 1 = 29.7 106 1 148.0 (108.8)

6 (807.2924.4)

= 1.0413 29.710 Below bubblepoint pressure =

= 1.038 /

1 1.039 1.040 138.6 140.7 ) 0.003 ( )] = 542 106 1 [( 1.039 779 789 779 789

Supplemental Tables

Page 14 of 16
TABLE 2DATA TO CALCULATE APPARENT MOLECULAR WEIGHT OF AIR Molecular Composition Mj yj Component weight mole fraction lb/lb mole lb/lb mole Nitrogen Oxygen Argon 0.78 0.21 0.01 28.01 32 39.94 21.8478 6.72 0.3994 Ma lb/lb mole 28.9672

TABLE 3MOLECULAR WEIGHT OF METHANE AND CARBON DIOXIDE Molecular weight Component lb/lb mole CH4 CO2 16.04 44.01

TABLE 4PRESSURE-VOLUME RELATIONS AT 60F 32API Oil P psi 2,002.2 1,623.8 1,474.4 1,212.6 1,033.2 924.4 807.2 800.6 794.2 789.0 779.0 0.9726 0.9802 0.9836 0.9898 0.9947 0.9980 1.0000 1.0098 1.0174 1.0286 1.0399 Relative volume P psi 1,928.4 1,691.2 1,410.0 1,200.6 998.8 865.8 794.6 597.6 590.8 490.8 466.0 0.9758 0.9797 0.9848 0.9887 0.9926 0.9953 0.9978 1.0000 1.0015 1.0064 1.0113 38API Oil Relative volume P psi 1,980.6 1,457.2 1,274.4 1,117.6 1,030.4 955.6 874.6 870.6 869.2 858.2 0.9733 0.9842 0.9888 0.9933 0.9964 0.9992 1.0000 1.0067 1.0137 1.0343 CO2-Water Mixture Relative volume

Page 15 of 16
TABLE 5RELEVANT DATA FOR 32API OIL

API

T F

Pb psi

Vb mL 148.52883

32

60

0.8654

0.5537

807.2

P psi 2,002.2 1,623.8 1,474.4 1,212.6 1,033.2 924.4 807.2 800.6 794.2 789.0 779.0

V mL 144.46614 145.58734 146.0958 147.01853 147.74858 148.22719 148.52883 149.99046 151.11122 152.77672 154.45

dP psi

dV mL

Vave mL

co psi-1

Rs scf/STB 144.53

Bo res bbl/STB 0.0000 1.0240 1.0252 1.0312 1.0348 1.0376 1.0413 1.0408 1.0404 1.0400 1.0393

Bg res bbl/scf

Bt res bbl/STB 0.00000 1.02404 1.02518 1.03116 1.03479 1.03764 1.04126 1.04486 1.04840 1.05138 1.05720

0.82 0.819 0.818 0.819 0.831 0.891 0.912 0.914 0.915 0.921 0.924

0.00106 0.00130 0.00143 0.00174 0.00207 0.00248 0.00290 0.00292 0.00295 0.00299 0.00304

-378.4 -149.4 -261.8 -179.4 -108.8 -117.2 -6.6 -6.4 -5.2 -10.0

1.12120 0.50846 0.92273 0.73005 0.47861 0.30164 1.46163 1.12076 1.66550 1.67328

145.027 145.842 146.557 147.384 147.988 148.378 149.260 150.551 151.944 153.613

2.04E-05 2.33E-05 2.40E-05 2.76E-05 2.97E-05 2.97E-05 5.21E-04 5.26E-04 5.33E-04 5.42E-04

144.53 144.53 144.53 144.53 144.53 144.53 143.14 141.81 140.72 138.65

TABLE 6RELEVANT DATA FOR CO2-WATER MIXTURE

API

T F

Pb psi

Vb mL 140.3197

10

60

1.5193

874.6

P psi 1,980.6 1,457.2 1,274.4 1,117.6 1,030.4 955.6 874.6 870.6 869.2 858.2

V mL 136.57417 138.10641 138.75307 139.38406 139.81319 140.21222 140.31968 141.26158 142.24203 145.1371

dP psi

dV mL

Vave mL

cw psi-1

Rs scf/ST B 203.08

Bw res bbl/STB 0.0000 1.0865 1.0889 1.0923 1.0940 1.0967 1.1001 1.0994 1.0992 1.0975

Bg res bbl/scf

Bt res bbl/STB 0.00000 1.08647 1.08887 1.09234 1.09402 1.09666 1.10005 1.10124 1.10167 1.10520

0.2811 0.21763 0.19589 0.17704 0.1676 0.16024 0.56094 0.56605 0.56605 0.57589

0.00037 0.00039 0.00040 0.00041 0.00042 0.00043 0.00165 0.00167 0.00167 0.00172

-523.4 -182.8 -156.8 -87.2 -74.8 -81.0 -4.0 -1.4 -11.0

1.53224 0.64666 0.63099 0.42913 0.39903 0.10746 0.94190 0.98045 2.89507

137.340 138.430 139.069 139.599 140.013 140.266 140.791 141.752 143.690

2.13E-05 2.56E-05 2.89E-05 3.53E-05 3.81E-05 3.80E-05 2.73E-04 2.73E-04 2.86E-04

203.08 203.08 203.08 203.08 203.08 203.08 202.00 201.62 198.64

Page 16 of 16

TABLE 7RELEVANT DATA FOR 38API OIL

API

T F

Pb psi

Vb mL 140.15325

38

60

0.8348

0.5537

597.6

P psi 1,928.4 1,691.2 1,410.0 1,200.6 998.8 865.8 794.6 597.6 590.8 490.8 466.0

V mL 136.7551 137.30453 138.01804 138.56381 139.11948 139.49596 139.84311 140.15325 140.36311 141.04649 141.73407

dP psi

dV mL

Vave mL

co psi-1

Rs scf/STB 125.47

Bo res bbl/STB 0.0000 1.0168 1.0204 1.0241 1.0276 1.0302 1.0287 1.0358 1.0352 1.0275 1.0256

Bg res bbl/scf

Bt res bbl/STB 0.00000 1.01681 1.02039 1.02408 1.02757 1.03015 1.02868 1.26446 1.27323 1.43047 1.47984

0.819 0.817 0.815 0.814 0.831 0.895 0.915 0.957 0.957 0.958 0.959

0.00110 0.00125 0.00149 0.00175 0.00214 0.00265 0.00295 0.00408 0.00412 0.00494 0.00521

-237.2 -281.2 -209.4 -201.8 -133.0 -71.2 -197.0 -6.8 -100.0 -24.8

0.54943 0.71351 0.54577 0.55567 0.37648 0.34715 0.31014 0.20987 0.68338 0.68758

137.030 137.661 138.291 138.842 139.308 139.670 139.998 140.258 140.705 141.390

1.69E-05 1.84E-05 1.88E-05 1.98E-05 2.03E-05 3.49E-05 1.16E-03 8.88E-04 1.07E-03 1.11E-03

125.47 125.47 125.47 125.47 125.47 125.47 125.47 123.83 100.04 94.28

Experiment 6 Interfacial Tension, Contact Angle & Capillary Pressure Measurements

PE 4521 002Reservoir Fluid Mechanics Laboratory Team 002E

Role Manager Researcher Technician Analyst Name Axel Hannenberg Lemmy Oshenye Lucas Gurgel De Carvalho Nor Ashraf Norazman Performance Score 1.0 1.0 1.0 1.0 Signature

Academic Integrity Statement

On my honor, I affirm that I have neither given nor received inappropriate aid in the completion of this exercise. October 26, 2012 Name: _________________________________ Date: _____________________________

Page 2 of 11 Abstract It is important to have knowledge of the interfacial tension, contact angle & capillary pressure because they help in understanding the nature of the reservoir and designing enhance oil recovery methods such as water flooding. Knowing capillary pressure curves of reservoir rocks helps to determine fluid saturations and connate water saturation. In this study, contact angles and interfacial tensions are measured and data from mercury injection are used to plot intrusion pressure versus Hg saturation, calculate and plot pore throat size histogram, convert and plot Hg-air data to air-brine capillary pressure, estimate R35 and calculate permeability using Winlands equation, and weighted geometric mean approach, and plot J-function versus wetting phase saturation. The three samples, Cl, Red, and Tc are analyzed to determine which will be a better reservoir. Data analysis shows that: 1. Permeability and capillary pressure are controlled by pore radii. Permeability affects the capillary pressure curve. Low displacement pressure indicates good connectivity. Leverett J-function can be used to group similar formations. 2. The air-brine contact angle is higher than air-oil contact angle for both glass and Teflon surfaces. The contact angle of air-fluid is higher on Teflon surface compared to on glass surface. 3. Tc will be a better reservoir because it has the highest permeability and porosity, lowest displacement pressure and is also well sorted. The team recommends performing drainage and imbibition methods on the samples.

Page 3 of 11 Introduction The capillary pressure is the pressure difference between two immiscible fluids in contact with each other, more specifically the pressure difference in the interface of these two fluids. The capillary pressure is inversely proportional to the radius of curvature of this interface and to the radius of the pore. This phenomenon occurs in oil reservoirs because of the presence of at least two immiscible fluids in reservoirs. Molecules located in the interface between these fluids are surrounded by different types of molecules. The molecules located in the fluids have a normal state of equal attraction in all directions but when two immiscible fluids are in contact this equality no longer occurs. The capillarity phenomenon occurs within the porous media of oil reservoirs. A reservoir rock is a porous media which contain an infinite number of pores of different sizes. As the capillary pressure depends on the radius of the pore, we can notice that it will vary for different zones in the same reservoir. For each zone of the reservoir there will be a different capillary pressure curve. For a given capillary pressure the lesser the permeability, the greater the water saturation. The Leverett J-function correlates these different curves to combine them into one curve for the entire reservoir (Rosa, 2011). Another important concept is the wettability. Wettability is the preference of a solid to contact one liquid or gas, known as the wetting phase, rather than another. The wetting phase will tend to spread on the solid surface (Schlumberger). The wettability can be determined by measuring the contact angle. When the contact angle of the fluid is less than 90, the solid has a wetting preference for this fluid. In this experiment, the contact angles of water and oil in different solids are measured using a contact angle meter. Interfacial tension of a brine sample was taken using a DuNouy tensiometer. Finally data from a high pressure mercury injection experiment in conjunction with a penetrometer was used to calculate air-brine capillary pressure curve, pore throat size distribution, permeability, and J-function. Experimental Procedure The interfacial tension of air-brine was measured using a CSC-DuNouy Tensiometer No. 70545 (Fig. 1). First, the stand was filled with liquid. The ring on the tensiometer was calibrated and submerged into the liquid. The adjustment of a combination of knobs gradually increases the tension in the liquid until the ring finally brakes through the liquid. The interfacial tension recorded was the tension required to break ring from contact with the fluid.

Fig. 1CSC-DuNouy Tensiometer No. 70545 used for measuring interfacial tension in the lab. The unit of measurement is dynes/cm.

Page 4 of 11 The contact angles for water/glass, water/Teflon, mineral oil/glass, and mineral oil/Teflon were evaluated using a CAM-PLUS MICRO contact angle meter (Fig. 2). The liquid samples (mineral oil or water) were inserted onto the solid surfaces (glass or Teflon) using a micro syringe and a light was shined through the sample. The radius for each sample (liquid on solid surface) was recorded. The contact angle was observed by taking half of the radius and measuring the angle intersecting the surface of the liquid at the intersection point.

Fig. 2CAM-PLUS MICRO contact angle meter used for measuring contact angles in the lab.

The AutoPore IV Series mercury porosimeter as shown in Fig. 3 was used to compute the absolute porosity of the porous medium. Rock samples were inserted into the porosimeter. The pressure was then increased until the first injection of mercury occurred. Mercury is the ultimate non-wetting phase, when pressure is applied it will go to the largest pores first. It must be forced into pores at a pressure equal to or greater than its capillary pressure. Low pressures are needed for larger pore throats while small pores need high pressure. The procedure was repeated until the pressure reaches 60,000 psi.

Fig. 3AutoPore IV Series Mercury Porosimeter (left). Penetrometers used in the AutoPore IV Series Mercury Porosimeter (right).

Page 5 of 11 Results
100,000 10,000 1,000 100 10 1 100% Cl Red Tc

80%

60%

40%

20%

0%

Hg Saturation

Fig. 4Intrusion pressure vs. Hg saturation plot. Displacement pressure is the highest for Red followed by Tc and Cl.

Frequency Cumulative % 45 40 35 30 25 20 15 10 5 0 0.001

Histogram for Cl, Red, and Tc

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 0.01 0.1 Pore Throat Radius, m 1 10

Fig. 5Histogram for all three samples. According to the bin values used, all three samples show the same histogram.

Frequency

Intrusion Pressure, psia

Page 6 of 11
Cl Red

70,000 60,000 50,000

Tc

Pc(air-Hg), psia

40,000 = 4.04 30,000 20,000 10,000 0 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000

Pc(air-brine), psia
Fig. 6Capillary pressure for air-Hg vs. capillary pressure for air-brine. The reciprocal of the slope represents a constant value to convert Pc(air-Hg) into Pc(air-brine). TABLE 1R35, RWGM, AND PERMEABILITY VALUES FOR THREE SAMPLES R35 m Cl Red Tc 1.53952283 0.56499651 5.87745512 Rwgm m 0.63749607 0.23099749 2.32206718 kR35 md 7.1 1.3 74.3 kRwgm md 7.5 1.4 71.6 kaverage md 7.3 1.4 73.0

10

Cl

Leverett J-function, dimensionless

9
8 7 6 5 4 3 2 1 0 0.0 0.2 0.4 0.6 0.8 1.0 Water Saturation, fraction

Red
Tc

Fig. 7Leverett J-function vs. wetting phase saturation. J-function is used to categorize similar types of formations.

Page 7 of 11
TABLE 2CONTACT ANGLES FOR DIFFERENT COMBINATION OF PHASES Brine/Glass 32 33 33 Brine/Teflon 78 78 78 Oil/Glass 8 8 8 Oil/Teflon 33 33 33

TABLE 3INTERFACIAL TENSION OF AIR-BRINE USING TENSIOMETER Reading Tensiometer dynes/cm 1 2 3 Average 76.7 77.8 77.2 77.2

Discussion of Results The mercury capillary pressure curves for three samples are compared with each other. The entry pressure is the pressure where Hg first enters the pores and is determined by the largest pore aperture size. The displacement pressure is the pressure where Hg becomes connected and is determined by the inflection point from the capillary pressure curve. The mercury injection method is a drainage method. The displacement pressure is lowest for Tc followed by Red and Cl. Low displacement pressure indicates good connectivity. Fig. 11 shows that the critical pore throat aperture is the highest for Cl followed by Tc and Red. The apexes represent the critical pore throat aperture for the samples. The histogram shows that the samples are dominated by pore throat sizes within the range of 0.01 to 9.0 microns. Permeability value affects the capillary pressure behavior. Higher permeability shifts the capillary pressure curve downward while lower permeability shifts the capillary pressure curve upward. Tc has the highest permeability followed by Red and Cl. Table 1 and Fig. 4 show this trend for the samples. Fig. 9 shows that Tc has more Hg saturation per incremental pressure, which supports the discussion about permeability and displacement pressure above. The permeability values are calculated using Winlands equation and the weighted geometric mean approach. The R35 is estimated by using linear interpolation. Hysteresis is the difference between the advancing and receding angles. Hysteresis makes the relationship between capillary pressure and water saturation nonunique (Sondergeld 2011). Capillary pressure cannot be determined from knowing water saturation alone without knowing the past wetting-drying history of the reservoir. All of the samples show no hysteresis. Leverett J-function is used to group similar formations together. Red and Tc samples show similar J-function curves and Cl has a different J-function curve. This means that Red and Tc are from sandstone formation and Cl is from a limestone formation. Fig. 6 shows the relationship between air-Hg and air-brine capillary pressures. The capillary pressure of air-Hg can be converted into the capillary pressure of air-brine by multiplying it with a constant. The

Page 8 of 11 constant is a function of contact angles and surface tensions of the phases. In this experiment, the contact angles and surface tensions are constant. The contact angle of air-Hg, air-Hg surface tension, and air-brine contact angle are given as 130, 485 dynes/cm, and 0 respectively. Contact angle is always measured through the denser fluid phase. The air-brine contact angle is higher than air-oil contact angle for both glass and Teflon surfaces. Since the fluids are dropped onto the solid surface, the contact angle measured is the advancing angle. The contact angle of air-fluid is higher on Teflon surface compared to on glass surface because the coefficient of friction of Teflon is higher than glass. Hence, both brine and oil molecules have a tendency to slip on a glass surface. Wettability is related to the contact angle. At the subsurface, oil-water contact depends on rock quality and is not uniform. Permeability and capillary pressure are controlled by pore radii (Sondergeld 2011). In the industry, the capillary radius in the lab and the field is assumed to be the same. Thus, the capillary pressure can be converted to equivalent height or height above free water level at reservoir conditions. The normalized capillary pressure curve (plot of Leverett J-function vs. water saturation) can also be used to determine the transition zone in the reservoir. This is done by conducting drainage and imbibition methods on the samples and plotting J-function vs. water saturation. Thus, the team recommends performing these methods on the samples. The contact angle is one of the factors affecting the transition zone other than the interfacial tension and density difference of multiple phases in the reservoir. The interfacial and surface tension forces in porous medium, wettability and capillary pressure affect microscopic displacement efficiency, which is important in water flooding (Akkutlu 2012). Both Tc and Red are well sorted formations because the plateaus of the capillary pressure curves are nearly horizontal. Compared to the other two samples based on the shape of the capillary pressure curve, Cl is poorly sorted. This might indicate that that Cl formation is a shaly limestone formation. Experimental results suggest Cl is the worst candidate for a reservoir. From the analysis, the team concludes that Tc will be a better reservoir because it has the highest permeability and porosity, lowest displacement pressure and is also well sorted. Conclusion and Recommendation Permeability and capillary pressure are controlled by pore radii. Permeability affects the capillary pressure curve. Low displacement pressure indicates good connectivity. Leverett J-function can be used to group similar formations. The air-brine contact angle is higher than air-oil contact angle for both glass and Teflon surfaces. The contact angle of air-fluid is higher on Teflon surface compared to on glass surface. The team concludes that Tc will be a better reservoir because it has the highest permeability and porosity, lowest displacement pressure and is also well sorted. The team recommends performing drainage and imbibition methods on the samples.

Page 9 of 11 References Akkutlu, Y.I. 2012. Displacement of oil and gas. Lecture notes on displacement of oil and gas. The University of Oklahoma, Oklahoma, United States. Rosa, Adalberto Jose. 2011. Engenharia de Reservatorios de Petroleo, Rio de Janeiro, RJ, Brazil. Schlumberger Oilfield Glossary, http://www.glossary.oilfield.slb.com/Display.cfm?Term=wettability (accessed 24 October 2012). Sondergeld, C.H. 2011. Capillary pressure. Lecture notes on capillary pressure. The University of Oklahoma, Oklahoma, United States. Appendices Equations =
2 cos ,

..............................(1)
( cos )() ( cos )()

() = () ( ) =
0.2166 , cos

, ..............................(2)

..............................(3)

log 35 = 0.732 + 0.588 log 0.864 log , ..............................(4) log = 1.245 + 1.469 log + 1.701 log 35, ..............................(5) log = 2.51 + 3.06 log + 1.64 log , ..............................(6) = = 0 +
ln( ) [ =1 ] =1

, ..............................(7) , ..............................(8)

(0 )(1 0 ) 1 0

Page 10 of 11 Supplemental Figures

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 100 10 1 0.1 0.01 0.001 Pore Radius, m Cl Red Tc

Fig. 8Cumulative Hg saturation vs. pore radius on a semilog scale.

0.028 0.026 0.024 0.022 0.020 0.018 0.016 0.014 0.012 0.010 0.008 0.006 0.004 0.002 0.000 0 0.2 0.4 SHg 0.6 0.8

Hg Saturation/Pressure

Fig. 9Hg saturation/pressure vs. Hg saturation. The plot shows that Tc has more Hg saturation per incremental pressure.

Cumulative Hg Saturation
Cl Red Tc 1

Page 11 of 11

100 90 80 70 60 50 40

Cl Red Tc

30
20 10 0 100 10 1 0.1 0.01

0.001

Pore Radius, m

Fig. 10Incremental pore volume vs. pore radius on a semilog scale.

0.006
0.005 0.004 0.003

Cl

Tc

0.002
0.001 0 100 10 1 0.1 0.01

0.001

Pore Throat Diameter, m

Fig. 11Incremental pore volume vs. pore throat diameter on a semilog scale.

Incremental Pore Volume, mL/g

Red

Incremental Pore Volume, %

Experiments 7, 8 and 9 Waterflooding and Enhanced Oil Recovery

PE 4521 002Reservoir Fluid Mechanics Laboratory Team 002E

Role Manager Researcher Technician Analyst Name Lemmy Oshenye Lucas Gurgel De Carvalho Nor Ashraf Norazman Axel Hannenberg Performance Score 1.0 1.0 1.0 1.0 Signature

Academic Integrity Statement

On my honor, I affirm that I have neither given nor received inappropriate aid in the completion of this exercise. December 07, 2012 Name: _________________________________ Date: _____________________________

Page 2 of 12 Abstract As production from oil reservoirs matures, improving the recovery factor will play a decisive role in offsetting the decline in production. Many methods exist to improve the recovery factor. In this experiment, waterflooding, surfactant flooding, and gas flooding were conducted. Significant findings include: 1. The mobility ratios of waterflooding, surfactant flooding, and gas flooding are 1.79, 0.63, and 2.5 respectively. Oil recovery decreases as mobility ratio increases. 2. The microscopic displacement efficiency of waterflooding, surfactant flooding, and gas flooding are 57.7, 49.2, and 33.9 respectively. High microscopic displacement efficiency results in low residual oil saturation after flooding. 3. Surfactant flooding has the highest recovery factor followed by waterflooding and gas flooding.

Page 3 of 12 Introduction Recovering oil from a petroleum reservoir can be achieved by primary recovery, secondary recovery and tertiary recovery. Primary oil recovery normally refers to the production of hydrocarbons using the natural driving mechanisms in the reservoir. Secondary oil recovery describes the additional recovery that results from water injection or immiscible gas injection. Waterflooding is the most common method of secondary recovery. Tertiary (enhanced) oil recovery refers to the additional recovery that is beyond what could be recovered by primary and secondary recovery methods (Ahmed 2001). Tertiary recovery includes chemical, thermal and miscible processes. In this experiment, waterflooding, surfactant flooding and gas flooding are conducted. Waterflooding Waterflooding is the use of water injection to increase the production from oil reservoirs. This is accomplished by the injection of water to increase the reservoir pressure to its initial pressure and maintain it near that pressure (Warner Jr. 2007). The water displaces oil from pore spaces but the microscopic displacement efficiency is affected by the interfacial tension in porous medium, rock wettability, water-/oil-relative permeability, and capillary pressure, which in turn affect waterflood recovery factor. Waterflood recovery factor is also influenced by intrinsic factors such as mobility ratio, reservoir heterogeneity, pore geometry, and initial water-/oil-saturation distribution. The grain shape, size, and sorting determine the pore-geometry heterogeneity. The wettability affects the capillary pressure and relative permeability. The viscosity and relative permeability of the fluids contribute to the mobility ratio. The initial water-oil-saturation distribution is important because it controls the efficiency of the waterflood in portions of the reservoir and also relates directly to the residual oil saturation that can be achieved at the end of a waterflood (Warner Jr. 2007). Higher oil saturation at the beginning of flood operations increases the oil mobility and hence gives higher recovery factor (Ahmed 2001). Connatewater saturation and residual oil saturation after waterflood are the most important numbers in waterflooding because they are used to determine the displacement efficiency. Surfactant Flooding Surfactant flooding is a method where surfactant is injected into the reservoir either for wettability alteration or to reduce the water-/oil-interfacial tension (Zitha et al. 2011). When a surfactant is added to the sand pack that consists of oil and brine, the surfactant molecules adsorb at the interface, displacing some of the water and oil molecules. Accumulation of surfactant at the interfacial zone disrupts the fluid structure and decreases interfacial tension (Akkutlu 2012). Surfactant flooding recovery factor is affected by the same factors that affect waterflooding recovery factor. The front-end cost of surfactant is very high. To overcome this, the surfactant should be injected in a small volume for mobility control. Retention of surfactants, which involves adsorption, precipitation, and phase trapping, is one of the main factors for the unfavorable economics in chemical flooding (Austad & Milter 2000). A proper amount of electrolyte is essential to maintain low interfacial tension with the oil globules trapped in the sand pack. In the industry, the surfactant flooding process requires a pre-flush to condition reservoir, followed by surfactant solution for releasing oil, followed by polymer solution for mobility control, and water to drive the chemicals and oil bank towards the production well.

Gas Flooding There are two different types of gas flooding: miscible gas injection and immiscible gas injection. In the miscible gas injection, the injected gas mixes with oil, makes the oil lighter and reduces the oil viscosity and surface tension of oil and rock. Consequently it becomes easier to produce the oil. In the immiscible

Page 4 of 12 gas injection the injected gas does not mix with oil. It only provides formation energy via incremental pressure. Typically it yields only about half the recovery of miscible floods. Each technique has advantages and disadvantages and is applicable to different types of reservoirs and conditions. CO2 miscible gas flooding is a major source of oil production in the U.S., particularly in West Texas, Wyoming and Mississippi where natural sources of CO2 can be obtained at reasonable cost. CO2 applications are increasingly popular due to the increasing desire to sequestrate manufactured CO2. Besides the miscible CO2 injection, there is also the enriched gas injection and the high pressure dry gas injection (Rosa 2011). Experimental Procedure Initial Preparation Before conducting the experiment, special safety requirements should be met. The use of safety goggles and gloves is required. A sand pack is built using a pre-weighed 300 ml Lucite tube and various sized (coarse, medium, and fine) sand grains. The relationship of coarse to medium to fine grains in the tube was approximately 1:1:2. In an attempt to obtain 27% or less porosity, the sand grains are packed tightly by pounding and continuously shaking the Lucite tube as it is being filled with the sand mixture. For additional packing, the container is connected to a vacuum to pack the sand as tightly as possible. The container is weighed again to calculate the pore volume and dry porosity. Next, the sand pack is completely saturated with brine at a constant flow rate and then weighed. The wet porosity and absolute permeability were computed.

Fig. 1Weighing tubing assembly.

Waterflooding The sand pack is flooded with mineral oil at a constant flow rate to determine the irreducible water saturation and relative permeability of the fluids. A waterflood is conducted by injecting water at a constant flow rate to displace the mineral oil. The volume of mineral oil displaced is collected to determine the residual oil saturation and the effective permeability. The sand pack is weighed again after waterflooding. Surfactant Flooding The sand pack is flooded with mineral oil until no water (or only fine oil bubbles) is produced to prepare it for surfactant flooding. The displaced water is measured to determine the irreducible water saturation, Swir, initial oil saturation, and relative permeabilities. The sand pack is weighed again. Surfactant is injected into the sand pack at a constant rate to displace the mineral oil. The effective permeability to the surfactant solution was determined during the last 1-2 pore volumes. The cylinders are allowed to set so that the oil and water will separate so that the oil and water recovered can be measured. All equipment and work area is cleaned as a safety precaution.

Page 5 of 12

Figure 2Injection of surfactant and volume of mineral oil and surfactant.

Gas Flooding A second sand pack is prepared with the same process as the initial preparation. The sand pack is flooded with oil so as to reach original hydrocarbon saturation. This is checked by weight measurements. Oil is then displaced from the sand pack with compressed air at constant pressure. Oil production is measured as a function of time and gas injection is recorded at the same times. Flooding is continued until oil production is zero.

Page 6 of 12 Results
40 35

Np vs. Wi for Waterflooding

Np, cm3

30 25 20 15 0 20 40 60 Wi, cm3 80 100 120 140

Fig. 3The cumulative oil production vs. amount of water injected for waterflooding.

180 160 140

Np + Wi vs. t for Waterflooding

Np + Wi, cm3

120 100 80 60 40 20 0 20 40 60 80 t, s
Fig. 4The cumulative fluids produced vs. time.

100

120

140

160

180

Page 7 of 12

45 40 35

Np vs. Wi for Surfactant Flooding

Np, cm3

30 25 20 15 0 20 40 60 80 Wi, 100 cm3 120 140 160 180

Fig. 5The cumulative oil production vs. amount of surfactant injected for surfactant flooding.

45

40
35

Np vs. Wi for Waterflooding and Surfactant Flooding

Waterflooding
Surfactant flooding

Np, cm3

30 25 20 15 0 20 40 60 80 Wi, 100 cm3 120 140

160

180

Fig. 6Comparison of the performance plots for waterflooding and surfactant flooding.

Page 8 of 12

30 25

Np vs. Gi for Gas Flooding

Np, cm3

20 15 10 5 0 50 100 150 200 Gi , cm3 250 300 350 400 450

Fig. 7The cumulative oil production vs. amount of gas injected for gas flooding.

70

Recovery Factor Displacement Efficiency Mobility Ratio

3.0 2.5

Recovery Factor, % Displacement Efficiency, %

60 50 40

1.5 30 20 10 0 Surfactant Flooding Waterflooding Gas Flooding 1.0

0.5
0.0

Fig. 8Comparison of RF, Ed, and M. Surfactant flooding has the highest RF and Ed, followed by waterflooding and gas flooding. The figure also shows a reverse trend for mobility ratio.

Mobility Ratio

2.0

Page 9 of 12

TABLE 1COMPARISON OF SAND PACKS 1 AND 2 Sand pack Pore volume cm3 1 2 73.09 75.29 dry % 26.91 27.34 saturated % 25.24 26.00 kabs darcy 3.84 3.74 Soi % 97.14 92.97 Swi % 2.86 7.03 kro fraction 0.24 0.23

TABLE 2COMPARISON OF THREE ENHANCED RECOVERY METHODS RF % Surfactant flooding Waterflooding Gas flooding 57.7 52.1 38.6 M fraction 0.63 1.79 2.50 Ed % 57.7 49.2 33.9

TABLE 3WATERFLOODING AT RESIDUAL OIL kw at Sor darcy 3.4330 krw at Sor fraction 0.8935 Sor fraction 0.4938 Sw fraction 0.5062

TABLE 4SURFACTANT FLOODING AT RESIDUAL OIL ksurfactant at

Sor

krsurfactant at
Sor

Sor fraction 0.4104

Sw fraction 0.0286

Ssurfactant fraction 0.5609

darcy 3.4482

fraction 0.8975

The efficiency of the surfactant flood in terms of the oil in the sand pack at the beginning of the surfactant flood is 0.2033.

Page 10 of 12 Discussion of Results The mobility ratio is lowest for surfactant flooding, and increasing from waterflooding to gas flooding. A low mobility ratio is favorable for flooding operations because it will be easier for the injected fluid to displace the recoverable oil. The recorded relative permeability after surfactant flooding is higher than that of waterflooding. The microscopic displacement efficiency is highest for surfactant flooding followed by waterflooding and gas flooding. High relative permeability and high displacement efficiency are favorable for flooding operations because they result in low residual oil saturation after flooding. The surfactant flooding method is the most efficient with the highest recovery factor, microscopic displacement efficiency, and lowest mobility ratio. The waterflooding method also seems efficient and can be a viable alternative if surfactant flooding is unavailable or too costly. The gas flooding process is least efficient with the lowest recovery factor because it has poor sweep efficiency. Also, the mobility ratio is highest for gas flooding and there is tendency for viscous fingering to occur. The oil recovery after breakthrough is lower than other processes. However, it must be acknowledged that gas flooding is more efficient in volatile oil reservoirs. Conclusion It can be concluded that the mobility ratio, relative permeability after flooding, microscopic displacement efficiency, and the initial water-/oil-saturation distributions for flooding operations have significant effect on the recovery factor. Based on the analysis of the three methods, without the consideration of cost of operation and type of reservoir, surfactant flooding is most favorable for oil recovery.

Page 11 of 12 References Ahmed, T. 2001. Reservoir Engineering Handbook, second edition. Houston, Texas: Gulf Professional Publishing/Elsevier. Austad, T. and Milter, J. 2000. Surfactants: Fundamentals and Applications in the Petroleum Industry. United Kingdom: Cambridge University Press. Evolution Petroleum Corporation. 2012. http://www.evolutionpetroleum.com/solutions_flooding.html. Rosa, Adalberto Jose. 2011. Engenharia de Reservatorios de Petroleo, Rio de Janeiro, RJ, Brazil. Warner Jr., H.R. 2007. Petroleum Engineering Handbook, Vol. 5, V-1037V-1096. Richardson, Texas, SPE. Zitha, P., Felder, R., Zornes, D., et al. 2011. Increasing Hydrocarbon Recovery Factors. www.spe.org/industry/docs/recoveryfactors.pdf. Appendices Equations = 1 2 , ..............................(1) =

= ( ) = 1 2 , ..............................(2) , ..............................(3)

2 2

= , ..............................(4) = , ..............................(5) = =

, ..............................(6) ..............................(7)

( )

, ( )

, ..............................(8)

1 = 1 + 2 + 3 , ..............................(9) = =
,

..............................(10) , ..............................(11)

Supplemental Tables
Waterflooding p p Vtotal Voil Vbrine t q Sw So kw krw Np+Wi

Page 12 of 12
psig atm 1.3816 5.59 1.3836 5.62 1.3816 5.59 1.3625 5.31 1.3625 5.31 1.3625 5.31 1.3597 5.27 At residual oil RF Vbrine fraction cm3 0.5211 40 cm3 20 40 60 80 104 117 125 t s 21 cm3 20 28 33 34 36 37 37 q cm3/s 1.9048 cm3 0 12 27 46 68 80 88 kw at Sor darcy 3.4330 s 22.1 54.4 80.6 102.64 121.23 139.94 158.23 krw at Sor fraction 0.8935 cm3/s 0.9050 0.6192 0.7634 0.9074 1.2910 0.6948 0.4374 Sor fraction 0.4938 fraction 0.2736 0.3831 0.4515 0.4652 0.4925 0.5062 0.5062 Sw fraction 0.5062 fraction 0.7264 0.6169 0.5485 0.5348 0.5075 0.4938 0.4938 Weight g 1,311.6 darcy 1.6053 1.0967 1.3541 1.6322 2.3222 1.2498 0.7883 Relative error fraction 0.9894 fraction 0.4178 0.2854 0.3524 0.4248 0.6044 0.3253 0.2052 M fraction 1.7853 cm3 40 68 93 114 140 154 162 Ed fraction 0.4917

Surfactant flooding p psig p atm t s Voil cm

3

Vsurfactant
injected

Vsurfactant cm
3

q cm /s 1.0983 0.5765 1.3018 1.4579 0.9840 1.2252 0.8201 y 0.6978 0.77393 0.5473 Sor fraction 0.4104
3

Sw fraction 0.0286 0.0286 0.0286 0.0286 0.0286 0.0286 0.0286 x 18.2100 1 54.19 Sw fraction 0.0286

Ssurfactant fraction 0.2736 0.4241 0.4925 0.5199 0.5473 0.5609 0.5609

So fraction 0.6978 0.5473 0.4789 0.4515 0.4241 0.4104 0.4104

ko darcy 4.1119 2.1659 4.9348 5.6024 3.8862 4.8616 3.2560

kro fraction 1.0702 0.5637 1.2844 1.4582 1.0115 1.2654 0.8474

ksurfactant darcy 1.7053 0.8983 2.0466 2.3235 1.6117 2.0162 1.3503

krsurfactant fraction 0.4438 0.2338 0.5327 0.6047 0.4195 0.5248 0.3515

WOR fractio n 0.0 0.0 0.5 1.9 2.2 2.6 3.0

cm

1.3775 5.53 18.21 20 1.3727 5.46 37.29 31 1.3604 5.28 54.19 36 1.3420 5.01 94.66 38 1.3058 4.48 110.92 40 1.2997 4.39 128.06 41 1.2990 4.38 146.35 41 At the beginning of the surfactant flood Ed fraction 0.2033 At residual oil RF fraction 0.5775 Vbrine cm3 40 t s 19.1 q cm3/s 2.0942

20 31 53 112 128 149 164

0 0 17 74 88 108 123 0 1

ksurfactant
at Sor

krsurfactant
at Sor

Ssurfactant fraction 0.5609

Weight g 1,306.9

darcy 3.4482

fraction 0.8975

Relative error fraction 0.9894

M fraction 0.6276

Ed fraction 0.5775

Gas flooding p t psig s 5.08 30 60 90 120 150 180 210 At residual oil RF M fraction fraction 0.3857 2.4962

Vgas injected cm3 50 128 184 256 291 357 381 Ed fraction 0.3393

Voil produced cm3 10 15 21 24 26 27 27

qg cm3/s 2.1

So fraction 0.8571 0.7857 0.7000 0.6571 0.6286 0.6143 0.6143

Sw fraction 0.0703 0.0703 0.0703 0.0703 0.0703 0.0703 0.0703

Sg fraction 0.0725 0.1440 0.2297 0.2725 0.3011 0.3154 0.3154

kg darcy 0.8841

krg fraction 0.2363