Professional Documents
Culture Documents
CONTENTS
1 GLOSSARY OF COMMON TERMS AND CONCEPTS IN RESERVOIR SIMULATION AND FLOW THROUGH POROUS MEDIA 1.1 Some General Definitions 1.2 Reservoir Fluid Properties 1.3 Single Phase Rock Properties 1.4 Multi-Phase Rock/Fluid Properties 1.5 Wettability and Fluid Displacement Processes 1.6 Oil Recovery Methods, Waterflood Patterns and Sweep Efficiency 1.7 Terms Used in Numerical Reservoir Simulation 1.8 Numerical Solution of the Flow Equations in Reservoir Simulation 1.9 Pseudo-Isation and Upscaling 1.10 Numerical Simulation of Flow in Fractured Systems 1.11 Miscellaneous - Vertical Equilibrium, Miscible Displacement and Dispersion
Glossary of Terms
1 GLOSSARY OF COMMON TERMS AND CONCEPTS IN RESERVOIR SIMULATION AND FLOW THROUGH POROUS MEDIA
This glossary is intended for use by the reader as a quick reference to terms used commonly in reservoir engineering in general and in reservoir simulation in particular. The student is not expected to work through this from begining to end in a systematic manner. However, the students should make sure that he or she is quite familiar with all the technical terms that appear in the main text of this unit. It is hoped that this is of particular use for distance learning students who may have studied the reservoir engineering distance leasrning unit some time ago but hopefully it will also be of use to our residential students.
Oilfield Units volumes in oilfield units are barrels (bbl or B); 1 bbl = 5.615 ft3 or 0.159 m3. A stock tank barrel (STB) is the same volume defined at some surface standard conditions (in the stock tank) which are usually 60oF and 14.7 psi. A reservoir barrel (RB) is the same volume defined at reservoir conditions which can range from ~ 90oF and 1500 psi for shallow reservoirs to > 350oF and 15,000 psi for very deep (high temperature - high pressure, HTHP) reservoirs. Note that when 1RB of oil is produced it gives a volume generally less than 1B at the surface since it loses its gas. (See formation volume factor.) Oil Types: Dry gas; Wet gas; Gas Condensate; Volatile oil; Black oil; Heavy (viscous) oil; Tar - see Tables 1 and 2 below.
4000
= 127 F
3500
3000
C C1
Critical Point
B1
2500
B2 D 40% 20%
2000
1500
1000
u Liq
m olu dV
10% 5% 0%
l bb 80% Bu
t o in eP
B3
200
250
Path
De w
Po i
of Pr oducti on
Cricondentherm = 250 F
Tc
Reservoir Temperature, T F
nt
350
Reservoir Fluid
Surface Appearance
GOR Range
API Gravity C1
C2
C3
C4
C5
C6+
Almost no liquids >100 Mscf/bbl some clear or straw-coloured liquid 3-100 Mscf/bbl (900-18000 m3/m3)
60o -70o
Condensate
50o-70o
87
4.4
2.3
1.7
0.8
3.8
Volatile or Brown liquid high shrinkage various yellow, red, oil or green hues
40o-50o
64
7.5
4.7
4.1
3.0
16.7
30o-40o
49
2.8
1.9
1.6
1.2
43.5
Heavy oil
10o-25o
20
3.0
2.0
2.0
12.0
71
Tar
Black substance
< 10
90+
Component C1 C2 C3 C4 C5 C6 C7+ Mol. Wt. C7+ GOR, SCF/bbl Tank gravity, 0 API Liquid color
Black Oil 48.83 2.75 1.93 1.60 1.15 1.59 42.15 100.00 225 625 34.3 Greenish Black
Volatile Oil 64.36 7.52 4.74 4.12 2.97 1.38 14.91 100.00 181.00 2000 50.1 Medium Orange
Gas-Condensate 87.07 4.39 2.29 1.74 0.83 0.60 3.80 100.00 112 18,200 60.8 Light Straw
Dry Gas 95.85 2.67 0.34 0.52 0.08 0.12 0.42 100.00 157 105,000 54.7 Water White
Gas 86.67 7.77 2.95 1.73 0.88 .... .... 100.00 .... Inf. ....
Table 2 Mole Composition and Other Properties of Typical Single-Phase Reservoir Fluids
Phase: A chemically homogeneous region of fluid which is separated from another phase by an interface e.g. oleic (oil) phase, aqueous phase (mainly water), gas phase, solid phase (rock). There is no particular symbol but frequently subscripted o, w, g; phases are immiscible.
Glossary of Terms
Inter Facial Tension (IFT): The IFT between two phases is a measure of energy required to create a certain area of the interface. Indeed, the IFT is given in dimensions which are energy per unit area. The symbol for IFT is and units are dyne/cm in c.g.s. units and N/m (newtons per m) in S.I. units. For example, if both gas and oil are present in a reservoir then the gas/oil IFT may be in the range, go ~ 0.1-10 mN/m; likewise. The oil/water value may be in the range, 0w ~ 15 - 40 mN/m. Note that numerically 1mN/m = 1dyne/cm. Component: A single chemical species that may be present in a phase; e.g. in the aqueous phase there are many components - water (H2O), sodium chloride (NaCl), dissolved oxygen (O2) etc.; in the oil phase there can be hundreds or even thousands of components - hydrocarbons based on C1, C2, C3, etc. Some of these oil components are shown in Table 2. Viscosity: The viscosity of a fluid is a measure of the (frictional) energy dissipated when it is in motion resisting an applied shearing force; dimensions [force/area.time] and units are Pa.s (SI) or poise (metric). The most common unit in oilfield applications is centiPoise (cP or cp). Typical example are:- water viscosity at standard conditions, w ~ 1 cP; typical light North Sea oils have o ~ 0.3 - 0.6 cP at reservoir conditions (T ~ 200oF ; P ~ 4000 - 6000 psi); at reservoir conditions, medium viscosity oils have o ~ 1 - 6 cP; moderately viscous oils have o ~ 6 - 50 cP; very viscous oils may have o ~ 50 - 1000s cP and tars may have o ~ up to 10000 cP. Formation Volume Factor: The factor describing the ratio of volume of a phase (e.g. oil, water) in the formation (i.e. reservoir at high temperature and pressure) to that at the surface; symbols Bw, Bo etc. For oil, a typical range for Bo is ~1.1 - 1.3 since, at reservoir conditions, it often contains large amounts of dissolved gas which is released at surface as the pressure drops and the oil shrinks; oilfield units [reservoir barrels/stock tank barrel (RB/STB)]. API Gravity (API): Definition = Gas Solubility Factors (or Solution Gas/Oil Ratios): These factors describe the volume of gas (usually in standard cubic feet, SCF) per volume of oil (usually stock tank barrel, STB); symbol, Rso and Rsw; units SCF/STB. Compressibility: The compressibility (c) of a fluid (oil, gas, water) or rock formation can be defined in terms of the volume (V) change or density () change with pressure as follows:
c =
1 V 1 = V P P
Note that this quantity is normally expressed in units of psi-1. Typical ranges of compressibilities are presented below (from Craft & Hawkins (Terry revision), 1991):
Compressibilities (units of 10-6 psi-1) Formation rock Water Undersaturated Oil Gas at 1000psi Gas at 5000psi 3 2 5 900 50 10 4 100 1300 200
Compressibilities are used in reservoir engineering for Material Balance Calculations. Material Balance Equations: Material Balance applied to a reservoir is simply a volumetric balance. It may be expressed as an equation which relates: The quantities of oil, gas and water produced. The reservoir (average) pressure. The quantity of water influx (e.g. from the aquifer). The initial oil and gas content of the reservoir. Essentially the material balance equations described how the energy of expansion and influx drive production in the reservoir. If there is a sufficiently low (or zero) fluid influx, the reservoir pressure will decline. One form of the Material Balance Equation is given below where each term on the left-hand side described a mechanism of fluid production (from Craft & Hawkins (Terry revision), 1991):
N .( Bt Bti ) +
= N p . Bt + ( Rp Rsoi ). Bg + Bw .Wp
Where the terms have the following meaning: N = initial reservoir oil, STB; Np = cumulative produced oil, STB Boi = initial oil formation volume factor, bbl/STB Bo = oil formation volume factor, bbl/STB Bgi = initial gas formation volume factor, bbl/STB Bg = gas formation volume factor, bbl/STB Bw = water formation volume factor, bbl/STB Rsoi = initial solution gas-oil ratio, SCF/STB Rp = cumulative produced gas-oil ratio, SCF/STB Rso = solution gas-oil ratio, SCF/STB We = water influx into the reservoir, bbl Wp = cumulative produced water, bbl cw = water isothermal compressibility, psi-1 cf = formation isothermal compressibility, psi p = change in average reservoir pressure, psi
6
.p + We
Glossary of Terms
Swi = initial water saturation m = (Initial hydrocarbon vol. of gas cap)/(Initial hydrocarbon vol. of oil) In practice the material balance equation is often applied in the linear form of Havlena and Odeh (J. Pet. Tech., pp896-900, Aug. 1963; ibid, pp815-822, July 1964); see discussion in Craft & Hawkins (Terry revision, 1991). In the above formulation of the Material Balance Equation, the various terms have the following interpretation. Left-Hand Side of the Material Balance Equation The following terms account for the expansion of any oil and/or gas zones that may be present in the reservoir:
N .(Bt Bttii ) +
The following term accounts for the change in void space volume which is the expansion of the formation and the connate water:
The next term is the amount of water influx that has occurred into the reservoir: We Right-Hand Side of the Material Balance Equation The first term of the RHS represents the production of oil and gas:
.p
= N p . Bt + ( Rp Rsoi ). Bg
The second term of the RHS represents the production of water: Bw.Wp
Porosity: the fraction of a rock that is pore space; common symbol, Porosity varies from 0.25 for a fairly permeable rock down to 0.1 for a very low permeability rock; there may be an approximate correlation between k and . Pores & pore throats: The tiny connected passages that exist in permeable rocks; typically of size 1m to 200 m; they are easily visible in s.e.m. (scanning electron microscopy). Pores may be lined by diagenetic minerals e.g. clays. The narrower constrictions between pore bodies are referred to as Pore Throats. See Figure 2:
Institute of Petroleum Engineering, Heriot-Watt University
illite
Figure 2
Permeability: The fluid (or gas) conducting capacity of a rock is known as the permeability; symbol k ; units Darcy (D) or milliDarcy (mD); dimensions -> [L]2. Permeability is found experimentally using Darcy's Law (see below). Permeability can be anisotropic and show tensor properties (see below) - denoted k . Probably the most important quantity from the point of view of the reservoir engineer since its distribution dictates connectivity and fluid flow in a reservoir. Timmerman (p. 83, Vol. 1, Practical Reservoir Engineering, 1982) presents the rule: Classification poor to fair moderate good very good excellent Permeability Range (mD) 1 - 15 15 - 50 50 - 250 250 - 1000 > 1000
k/ Correlations: It has been found in many systems that there is a relationship between permeability, k, and porosity, . This is not always the case and much scatter can be seen in a k/ crossplot. Broadly, higher permeability rocks have a higher porosity and some of the relationships reported in the literature are shown below. Some examples of k/ correlations which have appeared in the literature are shown in Figure 3:
Glossary of Terms
0.01 0.01
6 6
8 8
10 10
18 18
20 20
22 22
10,000 10,000
100 100
10 10
1 10 0
10 10
30 30
Darcys Law: Originally a law for single phase flow that relates the total volumetric flow rate (Q) of a fluid through a porous medium to the pressure gradient (P/x) and the properties of the fluid ( = viscosity) and the porous medium (k = permeability; A = cross-sectional area): Note that Darcy's law can be used to define permeability using the quantities defined in Figure 5 as follows:
k.A P Q = x
Note that the in the equation in Figure 5 is a factor for units conversion (see Chapter 2). Darcy Velocity: This is the velocity, u, calculated as, u = Q/A; this may be expressed as,
u=
k P Q = A x
v=
Q u = A.
P Q L k.A P . L Q
Q = .
Permeability Anisotropy: Since permeability can be directional, then it is possible for kx ky kz in a given system. This is often seen in practice when comparing the horizontal permeability, (kh), with the vertical permeability, (kv) - usually as the ratio, (kv/kh). It is often found (kv/kh) < 1, i.e. there is more resistance to vertical flow than horizontal flow. The value of (kv/kh) must be evaluated with respect to the scale (i.e. the size) of the sample or system. The value of this quantity will be different in a core plug or in a large simulator grid block in which the core plug was a small part. The origin of the anisotropy may be quite different in each case. At the small (core plug) scale, anisotropy may come from fabric anisotropy at the grain level or from lamination at the slightly larger scale (laminaset scale). At the larger scale (grid block), the anisotropy may arise from larger scale heterogeneity, even though locally the component rock facies are completely isotropic. This is illustrated in Figure 6.
10
Glossary of Terms
Core Plug
Small Scale
kh kv Hi k lamina Lo k lamina
Fabric Anistropy
Rock Grains
Lamination
Heterogeneity Anistropy
Low Perm Lenses or Shales (kv/kh) = 1 Low Perm Sand (kv/kh) = 1
Saturation: The saturation of a phase (oil, water, gas) is the fraction of the pore space that it occupies (not of the total rock + pore space volume); symbols Sw, So and Sg ; saturation is a fraction, where Sw + So + Sg = 1. Multi-phase flow functions such as relative permeability and capillary pressure (see below) depend strongly on the local fluid saturations. Residual Saturation: The residual saturation of a phase is the amount of that phase (fraction pore space) that is trapped or is irreducible; e.g. after many pore volumes of water displace oil from a rock, we reach residual oil saturation, Sor; the corresponding connate (irreducible) water level is Swc (or Swi); the related trapped gas saturation is Srg; at the residual or trapped phase saturation the corresponding relative permeability (see below) of that phase is zero. Strictly, we should refer to the phases in terms of wetting and non-wetting phases - the residual phase of non-wetting phase is trapped in the pores by capillary forces. Typically, in a moderately water wet sandstone, Sor ~ 0.2 - 0.35. The amount of trapped or residual phase depends on the permeability and wettability of the rock and a large amount of industry data is available on this quantity: For example, the relationship for k vs. Swc (or Swi) is shown for a range of reservoir formations, Figure 7.
11
10,000 5,000 1
1,000
500
12 100 50 13 10 11 7 9 8
10 5
11A 5 4 6
1.0 0 10 20 30 40 50 60 70 80 90 100
% Connate water
1 2 3 4 5 6 7 8 = = = = = = = = Hawkins Magnolia Washington Elk Basin Tangely Creole Synthetic Alundum Lake St John 9 = Louisiana Gulf Coast Miocene Age-Well A Ditto-Wells Band C North Belridge California North Belridge California Core Analysis Data Dominguez Second Zone Ohio Sandstone
10 = 11 = 11A = 12 = 13 =
Figure 7 Correlation between (air) permeability and the connate water (Swc) for a range of reservoir rocks
Relative Permeability: A quantity (fraction) that describes the amount of impairment to flow of one phase on another. It is defined in the two phase Darcy law (see notes); depends on the Saturation of the phase; e.g. in two phase flow -> krw and kro depend on Sw (since Sw + So = 1). A schematic of the Two Phase Darcy Law showing the definition of Relative Permeability is presented in Figure 8. At steady-state flow conditions, the oil and water flow rates in and out, Qo and Qw, are the same:
12
Glossary of Terms
At steady-state flow conditions, the oil and water flow rates in and out, Qo and Qw are the same Po Qw Qo Pw Qw Qo
Qw =
k.k ro .A . Po Qo = o L
Rel. Perm.
0 0
kro
krw
Sw
Several further examples of relative permeabilities and capillary pressure are given later in this glossary. Note that the Units for the two-phase Darcy Law are given in Figure 2, Chapter 2. Capillary Pressure: The difference in pressure between two (immiscible) phases; defined as the non wetting phase pressure minus the wetting phase pressure; depends on the saturation - for two phases capillary pressure, Pc(Sw) = Po- Pw (for a water wet porous medium). The following figures show schematic figures for Capillary Pressure (Pc(Sw)) and Relative Permeability (krw(Sw) and kro(Sw)) for a water wet system:
13
Capillary Pressure
Swc Sor
Relative Permeability
Swc Sor
Pc
krel
kro krw
Sw
Sw
Figure 9 Schematics of capillary pressure and relative permeability for a water-wet system
Mobility and Mobility Ratio: the mobility of a phase (e.g. w or o) is defined as the effective permeability of that phase (e.g. kw = k . krw ; ko = k . kro) divided by the viscosity of that phase;
k.k rw k.k ro w = ; o = w o
M=
o k ro . w = w k rw . o
Fractional Flow: The Fractional Flow of a phase is the volumetric flow rate of the phase under a given pressure gradient, in the presence of another phase. The symbols for water and oil fractional flow are fw and fo and they depend on the phase saturation, Sw:
fw =
Qw Q ; fo = o ; ;where Q T = Q o + Q w QT QT
The fractional flows play a central part in Buckley-Leverett (B-L) theory of linear displacement which starts from the conservation equation:
Sw f = w ; t x
So f = o t x
Buckley-Leverett Theory: This mathematical theory of viscous dominated water oil displacement is based on the fact that the velocity, vSw, of a fixed saturation value Sw is given by:
f vSw = v. w Sw
14
Glossary of Terms
where v is the fluid velocity, v = Q/(A) and (dfw/dSw) is the slope of the fractional flow curve. The relationship between the fractional flow and Buckley Leverett theory is illustrated in Figure 10.
Fractional Flow
Fractional Flow of Water, Welge Tangent
"Buckley-Leverett" Saturation Profile
Sor Sw
Sor Length, (x/L) Swc Swf Flood Front Height
fw
Swf Swc
Figure 10 Relationship between the fractional flow function and the Buckley-Leverett front height
Water Saturation, Sw
Wettability: This is a measure of the preference of the rock surface to be wetted by a particular phase - aqueous or oleic - or some mixed or intermediate combination. The Wettability of a porous medium determines the form of the relative permeabilities and capillary pressure curves; a very complex subject which is still the subject of much research. We refer to: Water wet, Oil wet and Intermediate wet systems in the following definitions. Water-Wet: Water-wet formation: Where water is the preferential wetting phase. Water occupies the smaller pores and forms a film over all of the rock surface - even in the pores containing oil. A Waterflood in such a system would be an imbibition process (see below). Water would spontaneously imbibe (see below) into a waterwet core containing mobile oil at Sor, hence displacing the oil. Oil-Wet: Oil-wet formation: Where oil is the preferential wetting phase. Oil occupies the smaller pores and forms a film over all of the rock surface - even in the pores containing water. A Waterflood in such a system would be a drainage process (see below). Oil would spontaneously imbibe into an oil-wet core containing mobile water at Swr, displacing the water. Intermediate-Wet: An Intermediate wet formation is where some degree of water wetness and oil wetness is shown by the same rock. Some different types of intermediately wet system have been identified known as Mixed wet and Fractionally wet. Both water and oil may spontaneously imbibe (see below) into such a system to some degree and indeed this forms the basis for certain Wettability Tests known as the Amott Test and the USBM Test (USBM => United States Bureau of Mines). Drainage: A Drainage displacement process is when the non-wetting phase is increasing. For example, in a water wet porous medium, drainage would be oil displacing water. The drainage and imbibition capillary pressure curves and relative permeabilities are different since these petrophysical functions depend on the saturation history. A simple schematic of a drainage process is shown in Figure 11.
Institute of Petroleum Engineering, Heriot-Watt University
15
Qo Qo Oil Injection
Water Wet Core at 100% Water Oil Injection Water Wet Core Qo at 100% Water Figure 11 Drainage
Imbibition: An Imbibition displacement process is when the wetting phase is increasing. For example, in a waterWater wet porous medium, drainage would be water Oil Injection Wet Core displacing oil as shown in Figure 12. at The drainage and imbibition capillary pressure 100% Water curves and relative permeabilities are different since these petrophysical functions depend on the saturation Q history.
w
Water Injection Q
w
Water Wet Core at sor Water Injection Water Wet Core at sor Qw
Figure 12 Imbibtion
Water Wet Core Spontaneous Imbibition: This process occurs when a wetting phase invades a porous at sor Water Injection
medium in the absence of any external driving force. The wetting fluid is sucked in under the influence of the surface forces. For example, if we have a water wet core at irreducible water saturation, Swr, then water may spontaneously imbibe and displace the oil as shown in Figure 13.
Oil
The observed behaviour in a system of Intermediate Wettability is shown in Figure 14 Oil where we seeat that Core s both phases can spontaneously imbibe under certain conditions.
wc
Oil
Water imbibes into core displacing Water imbibes oil-water wet or into core displacing intermediate wet oil-water wet or system intermediate wet Water imbibes into system core displacing oil-water wet or intermediate wet system
Figure 13 Spontaneous Imbibition Figure 14 Intermediate wettability. Both water and oil may spontaneously imbibe into the core displacing the other phase. Shows both water wet and oil wet character
Core at swc
Oil
Core at sor
Water
Primary and Secondary Recovery Processes: Primary and Secondary processes refer to the stage in the fluid displacement when one phase displaces another. For example, in a water wet porous medium (--> means displaces) :
Drainage and Imbibition Capillary Pressure Drainage (d) and Imbibition (i) Relative Permeabilities
Swc Sor
16
Glossary of Terms
Primary drainage: is oil --> water from a core at 100% water saturation to Swr. Secondary imbibition: is water --> oil from a core at Swr and mobile oil to Sor. Examples: Figures 15 and 16 show schematics of typical Drainage and Imbibition capillary pressure (Pc) and relative permeability (krw and kro) curves for a water wet system. Primary Drainage (oil --> water from core at 100% water) and Secondary Imbibition (water --> oil from core at Swr) processes are illustrated:
Drainage and Imbibition Capillary Pressure Drainage (d) and Imbibition (i) Relative Permeabilities
Swc Sor d kro i
Pc
krel
Drainage
d
krw
Imbibition
Sw
Sw
100 80 60 40 20 0 Imbibition
Relative Permeability, %
Drainage
Examples: Further examples of experimental capillary pressures and relative permeabilities in cores are shown for various processes (Drainage and Imbibition) and wettability conditions (Water wet and Intermediate wet) in figure 17,18,19 and 20 on the following pages.
17
OilOil
100 100
1.0 1.0
il il O O
0.1 0.1
0.01 0.01
Wa W arter te
0.001 0.001
0.0001 0.0001 0 0
100 100
100
18
1.0
il
Glossary of Terms
1.0
il
0.1
0.01
0.001
0.0001
Wate r
100
Examples: Experimental Capillary Pressures in cores for various processes (Drainage and Imbibition) and wettability conditions (Water wet, Oil Wet and Intermediate Wet). Figure 21.
19
48 40 Capillary Pressure - Cm of Hg 32 24 16
Water Wet
Venango core VL-2 k = 28.2 md Capillary Pressure - Cm of Hg
48 40 32 24 16 8
Oil Wet
2 8 0 0
1 2
20 40 60 80 Water Saturation - %
100
0 0
20
Capillary Pressure Characteristics, Strongly Water-Wet Rock. Curve 1 - Drainage Curve 2 - Imbibition
Oil-Water Capillary Pressure Characteristics, Ten-Sleep Sandstone, Oil-Wet Rock (After Ref. 29). Curve 1 - Drainage Curve 2 - Imbibition
32 24 16
Intermediate Wet
1 8 0 -8 -16 -24 2
20 40 60 80 Water Saturation - %
100
20 40 60 80 Water Saturation - %
100
Oil-Water Capillary Pressure Characteristics, Intermediate Wettability. Curve 1 - Drainage Curve 2 - Spontaneous Imbibition Curve 3 Forced Imbibition
Figure 21
20
Glossary of Terms
Examples: Relative Permeabilitites Examples of experimental relative permeabilities in cores for Water Wet and Oil Wet systems. Figure 22.
100 Relative Permeability, Fraction Relative Permeability, Fraction 80 60
Oil
100 80 60 40 20 0
Oil
40 20 0
Water
100
Wa te
100
1.0
il O
1.0
O
il
0.1
0.1
0.01
Water
0.01
0.001
0.001
0.0001
Wate r
100
Figure 22
21
A simple table summarising the typical characteristics of water-wet and oil-wet relative permeabilities is given below.
WATER WET Swc Sw where krw = kro (Points A on Figure 23) krw at Sro (Points B on Figure 23) mostly > 20% Sw > 50% krw < 0.3 OIL WET < 15% (Often < 10%) Sw < 50% krw > 0.5 (approaching 1)
Oil-Wet Reservoir
60 40 20
60 40 20 A Swi 0
Wa
te
r
Swi
r Wate
20
100
20
100
At Point A: kro = krw ; Sw > 50% krw at Sor / kro at Swi < 30%
At Point A: kro = krw ; Sw < 50% krw at Sor / kro at Swi > 50%
Figure 23 Influence of wettability on relative permeability (after Fertl, OGJ, 22 May 1978)
22
Glossary of Terms
Tertiary Recovery or Enhanced Oil Recovery (EOR) or Improved Oil Recovery (IOR) - this refers to a range of methods which are designed to recover additional oil that would not be recovered by either Primary or Secondary Recovery. Such methods include: Thermal Methods Gas Injection Chemical Methods Microbial Methods - steam, in-situ combustion,.. - N2, CO2, hydrocarbon gas injection (usually after a waterflood) - surfactant, polymer, alkali,.. - using bugs to recover oil!
Waterflood Pattern: On-land Waterflooding is often carried out with the producers and injectors in a particular pattern. This is known as pattern flooding and examples of such patterns are: Five Spot, Nine Spot, Line Drive etc. as shown schematically in Figure 24.
Injection Well Production Well Pattern Boundary
Regular Four-Spot
Skewed Four-Spot
Normal Nine-Spot
Inverted Nine-Spot
Five-Spot
Seven-Spot
Inverted Seven-Spot
Areal Sweep Efficiency: The Areal Sweep Efficiency refers to the fraction of areal reservoir that is swept at a given pore volume throughput of displacing fluid as shown schematically in Figure 25. For example, the Areal Sweep Efficiency at Breakthrough for various processes (Waterflooding, Gas Displacement and Miscible flooding) is shown as a function of mobility ratio in Figure 26:
Institute of Petroleum Engineering, Heriot-Watt University
23
90 80 70 60 50 0.1
WaterOil GasOil Miscible
10.0
50 0.1
WaterOil
Figure 26 Areal sweep efficiency at breakthrough in a five spot pattern for various displacement processes
10.0
Gas Oil : The Vertical Sweep Efficiency refers to the fraction of Vertical Sweep Effi ciency Miscible vertical section (or cross-section) of reservoir that is swept at a given pore volume throughput of displacing fluid. This is function of the heterogeneity of the system (e.g. stratification), the fluid displacement process (e.g. waterflooding, gas injection) and the balance of forces (e.g. importance of gravity). It is shown schematically in Figure 27.
24
Glossary of Terms
Poor Vertical Sweep (By gravity over-ride or the presence of a high-k streak in this case)
Mass Conservation: This is a general principle which is used in many areas of computational fluid dynamics. It says that: (mass flow rate into a block) - (mass flow rate out) = (the rate of mass accumulation in that block) A reservoir simulation model (for 1, 2 or 3 phases) is basically: A mass conservation equation combined with Darcys law. Black Oil Model: Different types of formulation of the transport equations for multiphase/multicomponent flow are used to simulate the various recovery processes; by far the most common is the Black Oil Model which can simulate primary depletion and most secondary recovery processes. A black oil simulation model is one of the most common approaches to modelling immiscible two and three phase (o, w, g) flow processes in porous media; it treats the phases rather like components; it does not model full compositional effects; instead, it allows the gas to dissolve in the other two phases (described by Rso and Rsw); however, no oil is allowed to enter the gas phase. Grid Structure: This refers to the geometry of the grid being used in the numerical simulation of the system. This grid may be Cartesian, radial or distorted and may be 1D, 2D or 3D (see notes). Spatial Discretisation: This is the process of dividing the grid in space into divisions of x, y and z. In reservoir simulation, we always chop up the reservoir into blocks as shown in the gridded examples below and then we model the block block flows.
25
Temporal Discretisation: This is the process of dividing up the time steps into divisions of t. 2D Areal Grid: This is a 2D grid structure as shown in Figure 28 which is imposed looking down onto the reservoir. For a Cartesian system, it would divide up the x and y directions in the reservoir into increments of x and y.
y
W2
W1
W3
2D Cross-Sectional Model: This is a 2D grid structure which is imposed on a vertical slice down through the reservoir. For a Cartesian system, it would divide up the x and z directions in the reservoir into increments of x and z. Cross-sectional calculations are carried out to asses the effects of vertical stratification in the system and to generate pseudo-function for upscaling. (Figure 29).
Figure 29
3D Cartesian Grid The 3D Cartesian Grid is the most commonly used grid when constructing a relatively simple model of a reservoir or a setion of a reservoir. This is shown in Figure 30.
26
Glossary of Terms
Producer
Water Injector
z (Variable)
Transmissibility: The transmissibility between two grid blocks is a measure of how easily fluids flow between them. The mathematical expression for two phase flow between grid blocks i and (i+1) is (for water):
(i+1/2) Boundary
Block i
Sw i krw i (kA)i Qw
Block i+1
Sw i+1 krw i+1 (kA) i+1
Figure 31
k (P Pi ) Q w = ( kA kA )i +1/ 2 rw . i +1 x w Bw i +1/ 2
where the inter-grid block quantities are averages at the interfaces (where i+1/2 denotes this block to block interface. The single phase Transmissibility, Ti+1/2 , is given by:
Ti +1/ 2 =
( kA)i +1/ 2
x
and the full Water Transmissibility, Tw,i+1/2, between the two grid blocks is given by:
Tw, i +1/ 2 =
( kA)i +1/ 2 k rw
x
Qw = Tw, i +1/ 2 ( Pi +1 Pi )
The water transmissibility is clearly made up of two parts each of which is an average between the blocks. The single phase part is (k.A)av and the two phase part is [krw/(w.Bw)]av (k.A)av - a Harmonic Average between blocks is taken for the single phase part of the transmissibility (see Chapter 4; Section 3.2). [krw/(w.Bw)]av - this term is more complicated. For the average relative permeability term, [krw]av an Upstream Weighting is used; For [(w.Bw)]av the Arithmetic Average between blocks is taken. (See Chapter 4; Section 3.3). Numerical Dispersion: The spreading of a flood front in a displacement process such as waterflooding, which is due to numerical effects, is known as Numerical Dispersion. It is due to both the spatial (x) and time (t) discretisation or truncation error that arises from the gridding. This spreading of flood fronts tends to lead to early breakthrough and other errors in recovery. How bad the error is depends on several factors including the actual fluid recovery process being simulated e.g. waterflooding, water-alternating-gas (WAG) etc. (See Chapter 4; Section 2.2). Grid Orientation: The Grid Orientation problem arises when we have fluid flow both oriented with the principal grid direction and diagonally across this grid as shown schematically in Figure 32. Numerical results are different for each of the fluid paths through the grid structure. This problem arises mainly due to the use of 5-point difference schemes (in 2D) in the Spatial Discretisation. It may be alleviated by using more sophisticated numerical schemes such as 9-point schemes (in 2D).
I = Injector P = Producer
Figure 32 Flow arrows show the fluid paths in oriented grid and diagonal flow leading to grid orientation errors
Local Grid Refinement: Local Grid Refinement is when the simulation grid is made fine in a region of the reservoir where (LGR) quantities (such as pressure or saturations) are changing rapidly. The idea is to increase the accuracy of the simulation in the region where it matters, rather than everywhere in the reservoir. E.g. LGR ==> close to wells or in the flood pilot area but coarser grid in the aquifer.
28
Glossary of Terms
Hybrid Grid LGR: Hybrid Grids are mixed geometry combinations of grids which are used to improve the modelling of flows in different regions. The most common use of hybrid grids are Cartesian/Radial combinations where the radial grid is used near a well. Hybrid Grid LGR can be used in a similar way to other LGR scheme. Examples: A simple example of LGR and Hybrid Grid structure is shown in figure 33.
Injector Producer
Hybrid Grid
Distorted Grids: A Distorted Grid is a grid structure that is bent to more closely follow the flow lines or the system geometry in a particular case. Corner Point Geometry: In some simulators (e.g. Eclipse), the option exists to enter the geometry of the vertices of the grid blocks. This allows the user to define complex geometries which better match the system shape. This option is known as Corner Point Geometry and it requires that the block block transmissibilities are modified accordingly. The idea of Corner Point Geometry is illustrated schematically in figure 34:
29
Fault L1 L2 L3 L4
L1 L2 L3 L4
Distorted Grid
30
Glossary of Terms
1 Extented Refinement
31
History Matching: History Matching in numerical simulation is the process of adjusting the simulator input in such a way as to achieve a better fit to the actual reservoir performance. Ideally, the changes in the simulation model should most closely reflect change in the knowledge of the field geology e.g. the permeability of a high perm streak, the presence of sealing faults etc. The observables which are commonly matched are the field and individual well cumulative productions, watercuts and pressures. Examples: Examples of history matching of pressure and water-oil-ratio (WOR) in two reservoirs are Figure 36. Note that in the WOR match the first pass was very inaccurate but that eventually a suitable match was found. A vitally important point is that a good history match must be obtained for the right reason. It may be possible to get a satisfactory match for the wrong reason i.e. by adjusting a variable that is not the primary cause of the mis-match (indeed, this is very often the case). However, such a model will eventually have very poor predictive properties.
1.0
(a) 3600 3400 3200 Final Match 3000 Pressure (psi) 2800 2600 2400 2200 2000 0 1 2 3 4 5 6 7 Year (b) 3600 11 3400 3200 3000 Pressure (psi) 2800 2600 2400 2200 2000 0 1 2 3 4 5 6 7 Year (c) 3600 3400 3200 3000 Pressure (psi) 2800 2600 2400 2200 2000 0 1 2 3 4 5 6 7 Year 8 9 10 11 12 Calculated Field Data 8 9 10 11 12 Calculated Field Data 8 9 10 11 12 Calculated Field Data
0.8
0.6 Watercut
0.2
Final Pressure Matches of Typical Khursaniyah Field Wells: (a) Reservoir AB (b) Reservoir C (c) Reservoir D (MP)
Figure 36 A field of a history match of watercut and well pressures; redraw from Mattax and Dalton (1990)
Finite Differences: When the derivative in a differential equation is approximated as a difference equation as follows:
32
Glossary of Terms
S (Si Si 1 ) x i x
then this is referred to as a Finite Difference approximation. In this example, which is illustrated below, (S/x)i is the derivative of Saturation (S) with respect to x at grid point i ; Si and Si-1 are the discrete values of S at grid points i and i-1, respectively; x is the size of the spatial grid.
Si-1 Si S(x,t) Si+1 x x i-1 x .... i x i+1
Slope = s x
Figure 37
Linear Equations: When finite difference methods are applied to the differential equations of reservoir simulation, a set of linear equations results. These have the form:
A. x = b
where A is a matrix of coefficients, x is the vector of unknowns and b is the (known) right hand side. Expanded up, this set of linear equations has the form: a11 x1 a21 x1 a31 x1 a41 x1 + + + + a12 x2 a22 x2 a32 x2 a42 x2 + + + + a13 x3 .... a23 x3 .... a33 x3 .... a43 x3 .... + a1n xn = + a2n xn = + a3n xn = + a4n xn = b1 b2 b3 b4
+ ann xn = bn
Direct Solution Of Linear Equations: A Direct Solution method is when the linear equations are solved by an algorithm which has a fixed number of operations (given N, the number of linear equations [unknowns]). If the equations have a solution, then, in principle, a direct method will give the exact answer, x(true), to the machine accuracy. E.g. Gaussian Elimination Iterative Solution Of Linear Equations: An Iterative Solution method is when the linear equations are solved by an algorithm which has a variable number of operations. A first estimate of the solution vector x(0) is made and this is successively refined to converge to the true solution. In a convergent iterative method, then x(v) x(true) as v . It is because of this iterative process that a variable number of steps may
33
be required depending on how accurate the answer must be. Normally, the iterative method would be carried out until: | x(v) - x(true)| < Tol. where Tol. is some small pre-specified tolerance. E.g. Line Successive Over-relaxation (LSOR) Grid Ordering Schemes: When the simulation grid blocks are ordered in various ways, the structure of the non-zeros in the sparse matrix, A, is different. Advantage can be taken of the precise structure when solving these equations. E.g. Schemes known as D2 and D4 ordering.
Upscaling: The process of reproducing the results of a calculation which is carried out on a fine grid on a coarser grid is known as Upscaling. The basic idea of upscaling is shown schematically in Figure 38. The input properties at the coarser scale must take into account the flow effects of the smaller scale structure. These coarser scale properties then become pseudo-properties.
34
Glossary of Terms
Low Sw Oil
Upscaling or Pseudo-Isation
Time
Pseudo-Property: This refers to the value of a property or function (e.g. permeability, relative permeability..) which is an average or effective value at a certain scale usually the grid block scale. For example, we might put the value kx = 150 mD in a simulator grid block which is 200 ft x 200 ft x 30 ft. Clearly, this incorporates a large amount of geological substructure and permeability may vary very significantly in different parts of this block. Pseudo-Relative Permeability: This is probably the most important pseudo-property that is used in reservoir simulation. It refers to the effective relative permeability in the simulation model at the grid block scale and is shown schematically in Figures 39 and 40. It must incorporate the effects of all the smaller scale geological heterogeneity, the balance of forces (viscous/capillary/gravity) and certain numerical effects (numerical dispersion). Methods for calculating the Pseudo-rel perm include: Jacks et al, Kyte and Berry, Stone etc. Newer methods are based on tensor pseudorelative permeabilities (Pickup and Sorbie, 1994).
Institute of Petroleum Engineering, Heriot-Watt University
35
Geopseudos: When the fluid flow upscaling is performed in the correct context of the sedimentary structure up from the lamina, laminaset, bedform.. scales, then the approach is known as the Geopseudo Methodology. This has been developed in work at Heriot-Watt which has extended more conventional approaches by putting in the geology. Figure 40 shows a simple example of a pseudo relative permeability showing holdup of fluid.
krel
Low Sw Oil
Sw
Upscaling or Pseudo-Isation
Pseudo-Relative Permeabilities
krel Sw
Coarse Grid Layered Model
Time
36
Glossary of Terms
Water Flow Sw
No Water Flow
Figure 40 A simple example of a pseudo relative permeability
Sw Swc
Fractured System: In this context, we imply systems (such as in many carbonate reservoirs) where small scale conductive fractures occur but most of the oil is in the rock matrix. In certain non-porous fractured rock reservoirs (e.g. fractured volcanics), it is possible to have all the oil in the fractures but these are much less common. Typically, in porous fractured systems: fracture porosity, f = 0.1 - 1% of bulk volume (i.e. as a fraction f = 0.001 - 0.01). Features of Fracture Geometry: The main geometric features of fractures which are thought to affect fluid flow are the: fracture orientation, width, conductivity, connectivity and spacing (or fracture density). The interface between the fracture and the matrix will also play a very important role in multiphase flow and fluid displacement processes. Stylolites: Stylolites are frequently found in limestones. They are laterally extensive features formed by grain-to-grain sutured contact caused by the phenomenon of pressure solution. These features may significantly reduce vertical permeability thus causing systems containing them to have very low (kv/kh) ratios (at certain scales). Vugs: Vugs are dissolution holes in a carbonate rock caused by diagenetic reactions. Dual Porosity Models: These are the most widely used simulation models for modelling flow in fractured systems. They have separate conservation/flow equations for the matrix and the fractures and matrix fracture flow is represented by Transfer Functions. They are most frequently used to model multiphase flow in fractured carbonates. Variants of this model allow for; (i) flow only in fractures and (ii) flow in both fractures and matrix. Discrete Models of Fracture Systems: A more recent approach to flow in fractured systems tries to represent the fractures explicitly as oriented planes with various shapes in 3D. Single phase (tracer) flow models of this type are used to model radioInstitute of Petroleum Engineering, Heriot-Watt University
37
nuclide transport in fractured media. However, multi-phase models of this type are not commercially available at present. Transfer Function: The function which describes the oil flow rate between the matrix and the fractures is known as the Transfer Function. Approximate analytical equations for this function have been suggested by Birk, Boxerman and Ahronovsky. Sudation: When oil is recovered from the matrix blocks in a fracture by a combination of gravity and capillary forces, the recovery mechanism is sometimes referred to as Sudation.
Vertical Equilibrium: The concept of Vertical Equilibrium (VE) is quite widely used in reservoir engineering. It takes several forms, two of which are listed below (and illustrated schematically in Figure 41: A: the pressure gradients in a particular direction (x, say), P/x, are all equal locally in a long system. See Figure 41a. B: there is virtually instant crossflow vertically - nearly infinite - compared with the horizontal flow. See Figure 41b. The VE assumption is often made in order to simplify the mathematical analysis of certain fluid flow problems in reservoir engineering. Vertical Equilibrium (VE) is known to apply in long thin systems (where x >> z, in Figure 41b). More accurately, the VE limit is approached as the scaling group, RL ; where:
x k R L = . z z k x
1/ 2
and kz and kx are the vertical and horizontal permeabilities, respectively. In practice, if RL is > 10, then VE is a very good assumption.
38
Glossary of Terms
P x
P x
P x
Miscible Displacement: Whereas oil and water are immiscible fluids (i.e. they do not mix and are separated by an interface), some fluids are fully Miscible (i.e. they mix freely in all proportions). When a gas (or other fluid) is injected into an oil reservoir and the fluids are miscible, this is referred to as a Miscible Displacement. When two fluids (e.g. gas and oil) are fully miscible (go = 0), the local pressure and the pressure gradients are the same (there is no capillary pressure since there is no interface). The mixing between the solvent and the oil can occur locally by Dispersion and by Fingering (see Viscous Fingering below). The displacement is described by a generalised Convection-Dispersion Equation where the mixing viscosity, (c) is a function of the concentration of the solvent, c (or oil). Often, the solvent viscosity is below that of the oil (i.e. s < o) which tend to cause an instability to develop in the displacement known as Viscous Fingering. The Miscible Flow Equations: These comprise of a Pressure Equation and a Transport Equation. The pressure equation is derived by inserting Darcys Law (with a viscosity dependent on solvent concentration) into the continuity equation. The transport equation is a generalised convection-dispersion (or convection-diffusion) equation. Continuity equation:
u=
k .P ( c)
where u is the Darcy velocity, c is the miscible solvent concentration and k is the
Institute of Petroleum Engineering, Heriot-Watt University
39
c = .(D.c) v.c t
where D is the dispersion tensor and v is the pore velocity ( v =
u ).
Dispersion and Dispersivity: Hydrodynamic dispersion in a porous medium at the small (core) scale is a frontal spreading or mixing which is due to various flow paths which the fluid can flow along at the pore scale. This mixing is a diffusive process since the growth of the mixing zone, Lf , tends to grow in proportion to t . In a tracer core flood experiment, the Dispersion Coefficient, D, may be measured by fitting the effluent profile to an analytical solution of the Convection-Dispersion Equation (see figure 42). Units of D (cm2/s - at lab scale). Dispersive mixing behaviour can also be seen from the mixing effect of heterogeneities at larger scales in a porous medium. D, has been found to depend linearly on velocity through the relationship: D = . v; where is the Dispersivity and has dimensions of length. Dispersion is actually a tensor in rather the same way that permeability ( k ) in its general form is a tensor.
40
Glossary of Terms
C(x,t1)
Lf
Ce 1
1 time (pv)
In situ concentration profile at time, t1; C(x,t1) Described by Convection Dispersion Equation: C = D C - v C t x2 x C = Dimensionless Concentration (C/Co) D = Dispersion Coefficient v = Pore Velocity (Q/A)
2
Viscous Fingering: When a high mobility (lower viscosity) fluid displaces a lower mobility (higher viscosity) fluid, a type of instability may develop known as Viscous Fingering. For such a displacement, the Mobility Ratio (see above) is high and the process may be observed in either miscible or immiscible displacements, although it occurs more readily in miscible systems. Results from an experiment are shown in Figure 43 where the dark fluid is high viscosity and the light fluid is low viscosity. Clearly, viscous fingering leads to a poorer sweep efficiency in such floods.
41