Quantifying the volumetric shrinkage of oil and expansion of gas as fluids travel from reservoir conditions to surface standard conditions, the essential conversion linking reservoir to surface volumes in every production calculation.
Every barrel of oil measured in a stock tank was once a larger volume in the reservoir. Every thousand cubic feet of sales gas expanded from a tiny volume at depth. The Formation Volume Factor (FVF) is the conversion factor that links these two worlds. Without it, material balance equations, production forecasts, reserves estimates, and well deliverability calculations are dimensionally incoherent.
FVF appears in virtually every equation in this programme. The Darcy radial flow equations for oil productivity index use Bo, the real gas pseudo-pressure uses Bg, and the total FVF Bt governs solution-gas drive behaviour. A 10% error in Bo propagates directly into a 10% error in calculated production rate. Understanding what drives FVF, dissolved gas, temperature, pressure, composition, enables the engineer to apply correlations intelligently and recognise when laboratory data should override them.
Lecture 1.2A: Formation Volume Factors — Physical Meaning and Engineering Use
23:51 · Standing, Vasquez-Beggs, and real gas deviation
Builds physical intuition for why reservoir oil shrinks and gas expands. Derives B_o and B_g from first principles. Demonstrates Standing's correlation and the real gas law. Uses GK-22 Gashaka condensate as the running case study.
▶
Lecture 1.2B: Reading PVT Reports — Extracting and Applying FVF Data
Walks through an actual North Sea PVT report. Shows how to extract B_o, B_g, R_s tables from differential liberation data, apply the flash correction, and use the corrected data in nodal analysis and material balance.
LEARNING OBJECTIVES
After completing this topic you will be able to:
1. Define oil, gas, and water formation volume factors (Bo, Bg, Bw) precisely and state their standard units and reference conditions. 2. Explain physically why Bo > 1 for live oil and why it peaks at bubble point then declines. 3. Calculate Bo using the Standing and Vasquez-Beggs correlations from GOR, API gravity, gas gravity, and temperature. 4. Calculate Bg from the real gas law using the Z-factor and reservoir conditions. 5. Determine the total FVF (Bt) and explain its role in solution-gas drive material balance. 6. Apply the shrinkage factor and explain how Bo enters the oil productivity index formula. 7. Distinguish between differential liberation and flash/separator test data and apply the flash correction to convert between them. 8. Estimate future Bo as reservoir pressure declines using Standing's correlations and a PVT table.
PREREQUISITES
Required: Topic 1.1 (PVT Behaviour, bubble point, phase envelopes, solution GOR definition). Basic algebra and unit conversion.
Helpful: Familiarity with the ideal gas law (pV = nRT).
Sub-problem SP-2 of the GK-22 PBL asks you to convert separator gas and condensate production rates into reservoir volumes to assess drainage efficiency. You need Bg to convert Mscf/d of produced gas to reservoir barrels, Bo (or Bgc for condensate) to convert STB/d of condensate, and Bt to estimate the total expansion driving production. The correlations and worked examples in this topic provide the tools for that calculation.
Section 1
The FVF Concept: What, Why, and How
Why reservoir volumes differ from surface volumes, what standard conditions mean, and how the FVF family (Bo, Bg, Bw, Bt) is defined.
The Core Problem: Two Different Worlds
A production engineer lives in two simultaneous worlds. World 1 is the reservoir: high pressure (thousands of psia), high temperature (150–350°F), where oil is swollen with dissolved gas and gas is compressed to a tiny fraction of its surface volume. World 2 is the surface: standard conditions of 14.696 psia and 60°F, where gas has expanded enormously and oil has shrunk as dissolved gas escaped. Every equation that connects these worlds requires a conversion factor, the Formation Volume Factor.
The FVF family answers one question in three forms: "How many reservoir barrels does one stock-tank barrel of oil (or one Mscf of separator gas, or one STB of water) occupy at reservoir conditions?"
Figure 1.2.1: Oil shrinks and gas expands between reservoir and surface conditions. FVF converts between the two measurement worlds.
Standard Conditions and Reference State
All surface volumes are reported at standard conditions: 14.696 psia (1 atm) and 60°F. These are the reference conditions for STB (stock-tank barrel) and scf (standard cubic foot). Some countries use different standards (e.g., 15°C and 101.325 kPa in metric jurisdictions), but the US oilfield standard dominates petroleum engineering literature. When comparing FVF values from different sources, always confirm the reference conditions.
The FVF is dimensionless in concept but carries compound units in practice: Bo is in res bbl/STB, Bg is in res bbl/scf or equivalently res ft³/scf, and Bw is in res bbl/STB. The total FVF Bt is in res bbl/STB of original oil.
The Full FVF Family
Bo: Oil Formation Volume Factor
Volume at reservoir conditions (p, T) occupied by 1 STB of stock-tank oil PLUS its dissolved gas. Always > 1.0 for live oil. Peaks at bubble point. Typical range: 1.05–2.5 res bbl/STB. Bo = 1.0 means zero dissolved gas (dead oil).
Bg: Gas Formation Volume Factor
Volume at reservoir conditions occupied by 1 scf of gas at standard conditions. Always << 1.0: gas is much smaller at depth. Typical range: 0.003–0.020 res bbl/scf. Decreases with pressure. Bg drives gas reserves calculation and lift design.
Bw: Water Formation Volume Factor
Volume at reservoir conditions occupied by 1 STB of stock-tank brine. Very close to 1.0: water is nearly incompressible and holds little dissolved gas. Typical range: 1.01–1.07 res bbl/STB. Often approximated as 1.0 in early-stage calculations.
Bt: Total (Two-Phase) FVF
Total reservoir volume occupied by 1 STB of original oil plus all gas that has evolved from it at pressure p. Equals Bo + (Rsi − Rs) × Bg. Used in material balance for solution-gas drive reservoirs. Rises steeply below bubble point.
KEY CONCEPTShrinkage Factor, the reciprocal of Bo
The shrinkage factor Sf = 1/Bo expresses the fraction of reservoir oil that survives to become stock-tank oil. A Bo of 1.35 means 1 res bbl produces only 1/1.35 = 0.741 STB: 26% of the reservoir volume is lost (as evolved gas) at surface. In field production reports you will sometimes see "shrinkage" listed rather than FVF: always check which convention is used.
Similarly, the surface/reservoir ratio for gas is 1/Bg expressed in scf/res bbl, a typical value of 1/0.005 = 200 scf per reservoir barrel.
Why FVF Appears in the Productivity Index
Consider the pseudo-steady-state oil productivity index (from Darcy's law, Topic 1.3 preview):
J* = kh / [141.2 B_o µ_o (ln(r_e/r_w) − 3/4 + S)]
Bo appears in the denominator because the Darcy equation models flow of reservoir fluid, while production rates are measured at surface. A higher Bo means reservoir fluid is more expanded, the same surface STB/d requires moving more reservoir barrels through the formation, which appears as reduced productivity. This is why a gas condensate (Bo equivalent large) has apparently lower productivity than its permeability alone would suggest.
Section 2
Oil Formation Volume Factor (Bo)
Behaviour above and below bubble point, physical drivers, Standing and Vasquez-Beggs correlations, and the undersaturated correction.
Behaviour of Bo with Pressure
Bo is not constant, it changes with pressure along the entire production path. Its behaviour has two distinct regimes separated by the bubble-point pressure pb:
Figure 1.2.2: B_o peaks at bubble-point pressure p_b. Below p_b (saturated region), gas evolves from solution causing rapid shrinkage. Above p_b (undersaturated), liquid compression causes a slight, slow decrease as pressure increases.
Above Bubble Point (Undersaturated)
All gas remains dissolved. Bo decreases slightly as pressure increases (liquid compression). The correction uses oil isothermal compressibility co:
B_o = B_ob × exp[c_o(p_b − p)]
Since co is small (~10⁻⁵ psi⁻¹), Bo changes slowly. Often approximated as constant above pb in quick calculations.
Below Bubble Point (Saturated)
Gas escapes from solution as p drops. Dissolved GOR R_s decreases. Bo falls rapidly because: (1) the oil volume shrinks as gas leaves, (2) remaining oil has lower gas content so lower volumetric expansion at depth. The rate of decrease is governed by Rs(p) correlation.
Physical Drivers of Bo Magnitude
Four factors govern how large Bo is at bubble-point conditions. Understanding these helps engineers anticipate whether a fluid will have a Bo of 1.10 or 2.20 before any PVT data are available:
Driver
Effect on Bo
Physical reason
Example
Solution GOR (Rs)
Dominant ↑
More dissolved gas = more expansion at depth
Rs 50→1,500: Bo rises ~1.08→2.0
Reservoir Temperature
Moderate ↑
Higher T means more thermal expansion of liquid
+50°F adds ~0.02–0.04 to Bo
API Gravity (lighter oil)
Moderate ↑
Lighter oils dissolve more gas and expand more
20°API → 40°API: Bob +0.05–0.15
Gas Gravity (lighter gas)
Slight ↑
Lighter gas (higher H content) occupies more volume
Minor effect vs. GOR and T
Standing's Correlation (1947)
Marshall Standing's correlation, derived from 105 California crude oil samples, remains the most widely used quick-estimate method for Bo at and below bubble point. It uses a single correlating function F:
F = R_s (γ_g / γ_o)^0.5 + 1.25 T
B_o = 0.972 + 0.000147 × F^1.175
Where Rs is solution GOR (scf/STB), γg is gas specific gravity (air=1), γo is oil specific gravity (=141.5/(131.5+°API)), and T is temperature (°F). Valid for 1.02 ≤ Bo ≤ 2.05, Rs from 0 to 1,950 scf/STB.
Vasquez-Beggs Correlation (1980)
Vasquez and Beggs developed a more accurate correlation from 6,000 data points spanning a wider range of oil types, split at 30°API:
B_o = 1 + C₁R_s + (T−60)(γ_API/γ_gc)[C₂ + C₃R_s]
Where γgc is gas gravity corrected to 100 psig separator pressure. Constants: for °API ≤ 30: C₁ = 4.677×10⁻⁴, C₂ = 1.751×10⁻⁵, C₃ = −1.811×10⁻⁸. For °API > 30: C₁ = 4.670×10⁻⁴, C₂ = 1.100×10⁻⁵, C₃ = 1.337×10⁻⁹. The gas gravity correction for separator pressure ps (psig) and temperature Ts (°F): γgc = γg[1 + 5.912×10⁻⁵ × γ_API × Ts × log(ps/114.7)].
Vasquez-Beggs is generally more accurate than Standing for oils outside California, especially lighter crude systems.
WORKED EXAMPLEBo Calculation: Standing vs. Vasquez-Beggs
Typical values: 5–25 × 10⁻⁶ psi⁻¹. The compressibility correction is generally small (<2% per 1,000 psia above bubble point) but becomes significant in high-pressure undersaturated reservoirs (e.g., pre-salt Brazil, deepwater Gulf of America at 15,000+ psia).
Example: Bob = 1.35 at p_b = 3,000 psia. At p = 6,000 psia with co = 12×10⁻⁶ psi⁻¹: Bo = 1.35 × exp[12×10⁻⁶ × (3,000 − 6,000)] = 1.35 × 0.964 = 1.302 res bbl/STB.
PVT laboratory reports present Bo from differential liberation (DL), a process where gas is removed at each pressure step, simulating reservoir depletion. But surface separators operate as a flash process, the composition stays constant until it flashes into gas and liquid in a single vessel.
These two processes give different Bo values because the differential process removes different gas compositions at each step, changing the residual oil composition. The flash correction (separator test) yields a conversion factor R_STO (residual oil shrinkage) that adjusts DL data to match actual separator conditions:
B_o(corrected) = B_o(DL) × (V_o at reservoir T,p) / (V_residual at 60°F after flash)
In practice, the PVT report provides both DL tables and a separator test table. The separator test gives a "shrinkage factor" at specified separator conditions (pressure, temperature). Using uncorrected DL data overestimates Bo by 2–8% for black oils, a systematic optimism in production forecasts.
Section 3
Gas FVF (Bg) and the Z-Factor
Deriving Bg from the real gas law, the compressibility factor Z, Standing-Katz charts, and how B_g governs gas reserves and well deliverability.
Gas Behaves Very Differently from Oil
Unlike oil, gas is highly compressible, a small pressure change causes a large volume change. Gas at 5,000 psia occupies roughly 1/340 of its volume at standard conditions (14.696 psia). The ratio of these volumes is B_g, and it is governed almost entirely by the real gas law. There are no equivalent of the Standing or Vasquez-Beggs semi-empirical correlations for gas, the thermodynamic equation is sufficiently accurate when the Z-factor is correctly determined.
Deriving Bg from the Real Gas Law
For an ideal gas, pV = nRT. For a real gas, the compressibility factor Z accounts for molecular interactions: pV = nZRT. Applying this at reservoir conditions and at standard conditions for the same number of moles n:
p_sc × V_sc = n Z_sc R T_sc (at standard, Z_sc ≈ 1)
p × V_res = n Z R T (at reservoir)
Dividing and rearranging for the volume ratio Vres/Vsc:
Where T is in °R (= °F + 460) and p is in psia. The constants come straight from the standard conditions: because Bg is a volume ratio for the same n moles, the gas constant R cancels, leaving psc/Tsc = 14.696/519.67 = 0.02829 res ft³/scf per unit of (ZT/p). Dividing by 5.615 ft³/bbl converts to barrels: 0.005035 = 14.696/(519.67 × 5.615) res bbl/scf per unit of (ZT/p).
Gas engineers frequently work in Mscf (thousands of standard cubic feet) rather than scf:
B_g = 5.035 ZT / p [res bbl/Mscf]
B_g = 0.02829 ZT / p [res ft³/scf]
Figure 1.2.3: B_g decreases hyperbolically with pressure. The real gas Z-factor causes departure from ideal behaviour, especially above 2,000–3,000 psia. At 5,000 psia and 250°F, B_g ≈ 0.0035 res bbl/scf, 1 Mscf at surface was only 3.5 res bbl at depth.
The Gas Compressibility Factor Z
The Z-factor (also called the gas deviation factor or supercompressibility factor) quantifies the departure of a real gas from ideal gas behaviour. It is defined by Z = pV/(nRT), unity for an ideal gas, and varying between ~0.70 and ~1.30 for reservoir gases depending on pseudo-reduced conditions.
Z is determined from the pseudo-reduced pressure ppr = p/ppc and pseudo-reduced temperature Tpr = T/Tpc, where ppc and Tpc are the pseudo-critical pressure and temperature of the gas mixture. For natural gas, pseudo-critical properties are calculated from either composition or gas gravity using the Kay mixing rule:
p_pc = Σ(y_i × p_ci) | T_pc = Σ(y_i × T_ci)
When composition is unknown, Standing's correlations from gas gravity γg provide pseudo-critical properties. Standing published two distinct curve pairs — one fitted to wet-gas/condensate well fluids, one to dry gases — and choosing the right pair matters at high reduced pressure:
Which to use when: use the condensate/wet-gas pair whenever the wellstream carries recoverable liquids — GK-22, with Rv = 112 STB/MMscf, is firmly in this class, so every GK-22 calculation in this module uses the condensate curves. Use the dry-gas pair for lean, sales-gas-only systems (as in Worked Example 2 in Section 5). Both fits cover roughly 0.55 < γg < 1.5; for gases with substantial heavy ends Sutton's (1985) correlation is the common alternative. For sour gases containing H₂S and CO₂, the Wichert-Aziz correction adjusts the pseudo-critical properties to account for these non-hydrocarbon components' effect on Z.
The Standing-Katz chart is the graphical standard for Z-factor determination. For computer calculations, the Hall-Yarborough (1974) method provides an analytical equation. Given t = 1/Tpr:
Where y is the reduced density (solved iteratively, Newton-Raphson), and A = 0.06125·t·exp(−1.2(1−t)²): note A itself is positive; the minus sign belongs to the A·ppr term in F(y), and Z would otherwise come out negative. Then Z = A ppr / y. This converges in 3–5 iterations and gives Z accurate to ±0.5% across nearly all natural gas conditions.
The Papay correlation is a simpler explicit formula: Z = 1 − (3.52 ppr)/(10^(0.9813 Tpr)) + (0.274 ppr²)/(10^(0.8157 Tpr)). Less accurate but useful for quick estimates at moderate reduced pressures (ppr ≲ 7); beyond that, prefer Hall-Yarborough or the chart.
WORKED EXAMPLEB_g Calculation: GK-22 Gashaka Gas
GK-22 reservoir: p = 7,350 psia, T = 285°F (745°R), γ_g = 0.65
Step 1: Pseudo-critical properties (Standing, condensate/wet-gas curves —
the correct set for GK-22's rich condensate):
T_pc = 187 + 330(0.65) − 71.5(0.65)² = 187 + 214.5 − 30.2 = 371.3°R
p_pc = 706 − 51.7(0.65) − 11.1(0.65)² = 706 − 33.6 − 4.7 = 667.7 psia
Step 2: Pseudo-reduced conditions:
T_pr = 745 / 371.3 = 2.006
p_pr = 7,350 / 667.7 = 11.01
Step 3: Z-factor (from Standing-Katz chart or correlation):
At T_pr = 2.0, p_pr = 11.0 → Z ≈ 1.40 (supercritical region)
(High pressure causes Z > 1: repulsive molecular forces dominate)
Step 4: B_g (res bbl/scf):
B_g = 0.005035 × Z × T / p
= 0.005035 × 1.40 × 745 / 7,350
= 0.005035 × 1,043 / 7,350
= 5.252 / 7,350
= 0.000714 res bbl/scf
= 0.714 res bbl/Mscf
Interpretation: 1 Mscf of sales gas came from only 0.714 res bbl
at depth. Reservoir gas is very compressed, 1 res bbl contains
1/0.714 = 1,401 scf of gas at surface conditions.
Gas Expansion Factor (Eg)
The reciprocal of Bg, expressed as scf/res bbl or Mscf/res bbl, is the gas expansion factor Eg. It tells you how many scf can be produced from each reservoir barrel of gas pore space:
E_g = 1/B_g = p / (0.02829 ZT) [scf/res ft³]
Eg is the basis of volumetric gas reserve estimation: G = Vφ(1−Sw) Eg = Vφ(1−Sw)/Bgi. As reservoir pressure drops from pi to p, the produced gas volume per reservoir pore volume increases proportionally to (Eg − Egi), the classic gas expansion drive mechanism.
COMMON ERROR: Units of B_g
Bg is reported in different unit systems in different references and software packages. Before substituting into any formula, confirm whether Bg is in res bbl/scf, res bbl/Mscf, or res ft³/scf. The total FVF equation Bt = Bo + (Rsi − Rs) × Bg requires Bg in res bbl/scf (not Mscf) when Rs is in scf/STB. Mixing units here generates a 1,000× error in Bt, a surprisingly common mistake in material balance worksheets.
Section 4
Water FVF (Bw) and Total FVF (Bt)
Water swelling, solution gas in brine, and how total FVF captures the combined expansion of oil and evolved gas below bubble point.
Water Formation Volume Factor (B_w)
Water is the least compressible and least thermally expansive of the three reservoir fluid phases. Its FVF is always close to 1.0, but the deviations from unity are important for accurate material balance in water-drive and waterflood reservoirs.
Bw is influenced by three competing effects: thermal expansion (increases Bw above 1.0), dissolved gas (methane solubility in brine increases Bw slightly), and salinity (higher TDS reduces Bw slightly). In practice these effects nearly cancel and Bw falls in the narrow range of 1.00–1.07 res bbl/STB for most reservoir conditions.
B_w ≈ 1 + ΔV_wT − ΔV_wp
Where ΔVwT is the thermal expansion correction and ΔVwp is the pressure compression correction. The McCain correlations (1990) provide polynomial fits:
ΔV_wT = −1.0001×10⁻² + 1.33391×10⁻⁴T + 5.50654×10⁻⁷T² (T in °F)
Natural gas dissolves in formation water, though far less than in oil. The dissolved gas ratio Rsw (scf/STB water) adds a small component to B_w. The Culberson-McKetta correlation gives gas solubility in pure water; salinity corrections (using the Dodson-Standing chart) reduce R_sw for brines.
At typical North Sea conditions (200°F, 3,000 psia, 25,000 ppm TDS): Rsw ≈ 15–25 scf/STB. This is negligible compared to oil R_s values of 300–1,500, but it must be accounted for in high-pressure, high-temperature gas reservoirs where water influx is significant: the produced water carries dissolved gas that appears as "extra" gas production not attributed to the gas reservoir itself.
The Bw including dissolved gas: Bw(total) ≈ Bw + Rsw × Bg, which at 20 scf/STB and Bg = 0.003 res bbl/scf adds only 0.06 res bbl/STB, generally negligible.
Condition
T (°F)
p (psia)
TDS (ppm)
B_w (res bbl/STB)
Shallow reservoir
120
1,500
50,000
1.020
Moderate depth
180
3,000
30,000
1.035
Deep reservoir
250
5,000
20,000
1.055
Very deep
300
8,000
10,000
1.065
Deep, high salinity
200
4,000
200,000
1.015
Total Formation Volume Factor (Bt)
The total FVF Bt, sometimes called the two-phase FVF, measures the total reservoir volume occupied by 1 STB of original oil at initial conditions, plus ALL the gas that has evolved from it at the current reservoir pressure p. It is the key volumetric parameter in solution-gas drive material balance.
B_t = B_o + (R_si − R_s) × B_g
Where Rsi is the initial solution GOR (scf/STB), Rs is the current solution GOR at pressure p (scf/STB), and Bg is in res bbl/scf. The term (Rsi − Rs) represents the gas that has evolved from solution, this gas now occupies additional reservoir volume, expressed through Bg.
Bt is always ≥ Bo, with equality at and above bubble point (no free gas, so Rsi = Rs). Below bubble point, Bt rises steeply as gas evolves and expands. This steep rise is what drives production in a solution-gas drive reservoir, the expanding gas pushes oil toward the wellbore.
Figure 1.2.4: B_t diverges from B_o below bubble point as evolved gas expands. The gap (R_si−R_s)×B_g is the volume occupied by free gas. The rapid rise of B_t at low pressures reflects the large expansion of gas at near-atmospheric conditions.
In a volumetric solution-gas drive reservoir (no water influx, no cap gas injection), the driving energy comes entirely from the expansion of reservoir fluids. The material balance in its simplest form is:
N × B_t = N × B_ti + ΔN_p × B_o
Rearranging for cumulative oil production N_p:
N_p = N × (B_t − B_ti) / B_o
The numerator (Bt − Bti) is the total expansion per STB of original oil. When Bt rises from 1.35 (initial) to 2.80 (at low pressure), the fluid has expanded by 1.45 res bbl/STB original oil, this expansion is what produces the oil. Recovery factor in a solution-gas drive reservoir is typically only 15–25% precisely because much of this expansion goes into producing gas rather than oil, and gas relative permeability rises rapidly as gas saturation builds.
Section 5
Correlations, Selection, and Worked Examples
Comparing correlation accuracy, selecting the right method for the fluid and data available, and three full worked examples at increasing complexity.
Selecting the Right Correlation
Multiple correlations exist for Bo and the supporting parameters (pb, Rs, co). Selecting the appropriate one requires knowing the geographic/geological origin of the fluid, the data available, and the required accuracy. No single correlation is universally superior.
Correlation
Data Source
GOR Range
API Range
Accuracy
Best For
Standing (1947)
105 California oils
0–1,950 scf/STB
16–63°
±106 psia (p_b); ±5% B_o
Quick estimates; California-type crude
Vasquez-Beggs (1980)
6,000+ global data
20–2,070 scf/STB
16–58°
±4.7% (p_b); ±4.6% (B_o)
Most oil systems; preferred for non-CA oils
Lasater (1958)
158 data points
3–2,905 scf/STB
17.9–51°
±3.8% (p_b)
Bubble-point of light/volatile oils
Al-Marhoun (1988)
160 Middle East oils
26–1,602 scf/STB
19–44°
±4% (p_b); ±1.7% (B_o)
Middle East reservoirs; low-GOR
Petrosky-Farshad (1993)
81 Gulf of America oils
217–1,406 scf/STB
25–46°
±3.6% (p_b)
GoA deepwater; moderate GOR
EOS (PR/SRK)
Tuned to PVT data
All
All
±0.5–2% when tuned
Simulation; complex fluids; volatile oil
PRINCIPLE: Laboratory Data Always Wins
All the correlations above are statistical fits to historical datasets. They represent average behaviour, not your specific reservoir fluid. A PVT analysis from a representative fluid sample, properly conducted (CCE + DL + separator flash), will always be more accurate than any correlation. Use correlations for: (1) preliminary screening before PVT data are available; (2) checking whether laboratory data are reasonable; (3) interpolation/extrapolation outside the pressure range of laboratory data. Never replace laboratory PVT data with correlations once the data exist.
IN THE FIELD: PVT Sampling Quality Sets Your IPR Uncertainty
Every FVF and Z value in this topic traces back to a physical fluid sample, and sampling errors propagate directly into deliverability: a 10% error in Bo (or in Z for gas) is a 10% error in every calculated rate, and a mis-measured saturation pressure shifts the predicted onset of gas evolution or condensate banking by months to years. The classic failure modes are: (1) bottomhole samples pulled while flowing below pb or pd — the fluid is already two-phase at the sampler and unrepresentative; (2) separator recombination at the wrong producing GOR: a well still cleaning up or loading liquids gives a false GOR and therefore a false recombined fluid; (3) condensate systems sampled after near-wellbore dropout has begun: the wellstream is leaner than the reservoir fluid, biasing Rv and the dew point low. For GK-22 the dew point (6,200 psia) came from a bottomhole sample taken with pwf = 6,850 psia, comfortably above dew point, so it is carried as high-confidence in the PBL, whereas the provisional skin from a single build-up is not. When you build the SP-5 sensitivity study, let sampling pedigree decide which PVT inputs get the wide uncertainty bands.
Worked Example 1: Full Bo Table from PVT Data
A North Sea black oil reservoir has been PVT-sampled. The differential liberation data are given below. Construct Bo and Bg as functions of pressure, and calculate Bt at each pressure step (initial Rsi = 680 scf/STB).
p (psia)
Rs (scf/STB)
Bo (res bbl/STB) DL
Z
Bg (res bbl/scf)
Bt (res bbl/STB)
4,200 (pb)
680
1.480
0.862
0.00157
1.480
3,500
598
1.440
0.848
0.00184
1.440 + (680−598)×0.00184 = 1.591
3,000
535
1.411
0.840
0.00213
1.411 + (680−535)×0.00213 = 1.720
2,500
470
1.379
0.832
0.00253
1.379 + (680−470)×0.00253 = 1.910
2,000
400
1.344
0.826
0.00314
1.344 + (680−400)×0.00314 = 2.223
1,500
325
1.304
0.823
0.00418
1.304 + (680−325)×0.00418 = 2.788
1,000
245
1.260
0.824
0.00627
1.260 + (680−245)×0.00627 = 3.988
500
155
1.205
0.832
0.01264
1.205 + (680−155)×0.01264 = 7.841
200
80
1.150
0.857
0.03257
1.150 + (680−80)×0.03257 = 20.69
OBSERVATIONThe Dramatic Rise of Bt at Low Pressure
Note how Bt rises from 1.480 at bubble point to 20.69 at 200 psia, a 14-fold increase. B_o only falls from 1.480 to 1.150. The huge Bt rise is driven entirely by Bg (which rises from 0.00157 to 0.03257 as gas expands 20-fold at low pressure). This explains why solution-gas drive reservoirs accelerate production initially (expanding gas is very energetic) but then decline rapidly (gas saturation builds, krg rises, GOR escalates, and reservoir pressure collapses).
Worked Example 2: Bg and Gas Reserves
A dry gas reservoir: pore volume Vp = 1.2 × 10⁸ res ft³, φ = 0.18, Sw = 0.25, p_i = 4,500 psia, T = 210°F, γg = 0.72. Calculate (a) initial gas in place (IGIP) and (b) gas recoverable when pressure drops to 1,500 psia (abandonment), assuming no water influx.
Given: V_res = 1.2×10⁸ ft³, φ = 0.18, S_w = 0.25, T = 210°F = 670°R
Step 1: Pseudo-critical properties (Standing, DRY-GAS curves — this is a
lean dry gas, unlike GK-22 which uses the condensate curves):
T_pc = 168 + 325(0.72) − 12.5(0.72)² = 168 + 234.0 − 6.5 = 395.5°R
p_pc = 677 + 15.0(0.72) − 37.5(0.72)² = 677 + 10.8 − 19.4 = 668.4 psia
Step 2: Z at initial conditions (p_i = 4,500 psia):
T_pr = 670/395.5 = 1.694
p_pr = 4500/668.4 = 6.73
Hall-Yarborough (or Standing-Katz chart): Z_i ≈ 0.949
Step 3: B_gi (res ft³/scf):
B_gi = 0.02829 × 0.949 × 670 / 4,500 = 17.99 / 4,500
= 0.003997 res ft³/scf
Step 4: Initial gas in place (IGIP):
Gas pore volume = V_p × φ × (1−S_w) = 1.2×10⁸ × 0.18 × 0.75
= 1.62×10⁷ res ft³
IGIP = Gas PV / B_gi = 1.62×10⁷ / 0.003997 = 4.053×10⁹ scf = 4,053 MMscf
Step 5: Z at abandonment (p_ab = 1,500 psia):
p_pr = 1500/668.4 = 2.24, T_pr = 1.694 → Z_ab ≈ 0.882
B_gab = 0.02829 × 0.882 × 670 / 1,500 = 16.72 / 1,500 = 0.01115 res ft³/scf
Step 6: Recoverable gas (volumetric depletion):
G_R = IGIP × (1 − B_gi/B_gab) = 4,053 × (1 − 0.003997/0.01115)
= 4,053 × (1 − 0.3585) = 4,053 × 0.6415 = 2,600 MMscf
Recovery Factor = 64.1%: typical for a well-connected, volumetric gas reservoir.
(Note that Z_i and Z_ab move together: recomputing both with Hall-Yarborough
changes IGIP by −7% but leaves the recovery factor almost unchanged, because
RF depends on the ratio Z_i/Z_ab, not on either value alone.)
Worked Example 3: Productivity Index with Bo
Demonstrate how Bo directly affects the oil productivity index J*. A reservoir has k = 150 md, h = 80 ft, re = 2,980 ft, rw = 0.35 ft, S = 0, µ_o = 0.8 cp. Compare J* at initial reservoir pressure (Bo = 1.30) versus after 3 years of production (Bo = 1.18, µo = 1.1 cp, p has dropped below bubble point).
Pseudo-steady-state PI formula:
J* = kh / [141.2 × B_o × µ_o × (ln(r_e/r_w) − 0.75 + S)]
ln(2,980/0.35) − 0.75 = ln(8,514) − 0.75 = 9.050 − 0.75 = 8.300
Initial conditions (B_o = 1.30, µ_o = 0.8 cp):
J*_i = (150 × 80) / (141.2 × 1.30 × 0.8 × 8.300)
= 12,000 / (1,216) = 9.87 STB/d/psi
After 3 years (B_o = 1.18, µ_o = 1.1 cp):
J*_f = (150 × 80) / (141.2 × 1.18 × 1.1 × 8.300)
= 12,000 / (1,521) = 7.89 STB/d/psi
Decline = (9.87 − 7.89) / 9.87 = 20% reduction in PI
Note: B_o alone falls 9% (from 1.30 to 1.18) — but the
viscosity increase (0.8 → 1.1 cp, +38%) has a larger impact.
Both effects compound in the denominator of J*.
Interactive Simulation
FVF Interactive Calculator
Compute Bo, Bg, Bt, and shrinkage factor in real time. Explore how each parameter shifts the FVF and visualise the full pressure profile.
Try these scenarios:
• GOR=100, T=150°F, 22°API → heavy black oil (B_o~1.08)
• GOR=800, T=220°F, 38°API → typical North Sea
• GOR=1,800, T=280°F, 48°API → volatile oil (B_o~1.8+)
• Raise pressure above bubble point → see compressibility effect
• Drop pressure below bubble point → watch B_t diverge from B_o
Gas FVF Calculator: Real Gas Law + Z-factor INTERACTIVE
Loading...
Pressure Profile: B_o, B_g×1000, and B_t INTERACTIVE
Adjust sliders above and the profile updates automatically.
Section 7
FVF in Material Balance & Production Engineering
How Bo, Bg, and Bt enter the material balance equation, reserves estimation, and the Productivity Index, connecting FVF to every calculation in this programme.
▶
Lecture 1.2C: FVF in the Material Balance — From Havlena-Odeh to Production Decline
18:00 · MBE derivation, F vs. E_o plots, reservoir typing
Derives the full Havlena-Odeh material balance. Shows how B_o and B_t appear in the underground withdrawal term F and the expansion terms E_o and E_g. Demonstrates reservoir type diagnosis (solution-gas drive vs. water drive) from straight-line MBE plots. Uses GK-22 condensate as the gas MBE analogue.
The Underground Withdrawal (F)
In the Havlena-Odeh formulation of the material balance equation, all produced fluids are converted to reservoir volumes using FVFs and summed into a single "underground withdrawal" term F:
F = N_p[B_o + (R_p − R_s)B_g] + W_p B_w
Where N_p is cumulative oil production (STB), R_p is cumulative producing GOR (scf/STB), R_s is current solution GOR (scf/STB), W_p is cumulative water production (STB). This equation converts every produced barrel (oil, dissolved gas, free gas, and water) back into the reservoir volume it originally occupied. F must equal the sum of all expansion and influx terms.
Expansion Terms and Reservoir Drive Diagnosis
The right-hand side of the MBE decomposes into expansion terms. Each term uses FVF:
The full MBE becomes: F = N(E_o + m·E_g + E_f,w) + W_e·B_w. Havlena and Odeh showed that plotting F/(E_o + m·E_g) vs. W_e·B_w/(E_o + m·E_g) yields a straight line with slope 1 and intercept N, enabling simultaneous determination of OOIP N and aquifer influx W_e.
Solution-Gas Drive (No Influx)
F = N × E_o. Plot F vs. E_o → straight line through origin, slope = N (OOIP in STB). Curvature indicates additional energy source (water influx, gas cap). E_o is dominated by the (R_si−R_s)B_g term at low pressures, dissolved gas expansion is the drive mechanism.
Water Drive
F = N×E_o + W_e×B_w. Plotting F/E_o vs. W_e/E_o must be straight with slope 1, but W_e is unknown without an aquifer model. Incorrect MBE drive diagnosis (mistaking water drive for depletion) leads to over-optimistic recovery predictions using FVF data alone.
Gas Reservoir Material Balance: p/Z Plot
For a volumetric gas reservoir (no water influx, no cap gas), the MBE simplifies dramatically when B_g = ZT/(p × constant) is substituted:
G_p = G[1 − (p/Z)/(p_i/Z_i)]
This is the famous p/Z plot: cumulative gas production G_p vs. p/Z plots as a straight line. Extrapolating to p/Z = 0 gives IGIP = G. The slope and linearity diagnose reservoir drive mechanism, a curved p/Z line (concave up) indicates water influx, not simple depletion.
Note that this entire result follows from B_g = 0.02829ZT/p, the FVF is embedded in the derivation. The p/Z plot is, in a real sense, a plot of FVF behaviour through depletion.
FIELD EXAMPLEp/Z Plot: Gashaka GK-22 Gas Condensate
GK-22 is a gas condensate (not a simple dry gas), so strictly a modified p/Z must be used with a condensate correction to the Z-factor. However, the method illustrates FVF in practice:
Initial conditions: p_i = 7,350 psia, T = 285°F, Z_i = 1.40 (Topic 1.2 Sec 3)
p_i/Z_i = 7,350/1.40 = 5,250 psia
After 2 years (G_p = 35 Bscf produced):
p = 6,800 psia (from wellhead buildup surveys)
Z at 6,800 psia, 285°F: T_pr = 745/371.3 = 2.006, p_pr = 10.18 → Z ≈ 1.35
p/Z = 6,800/1.35 = 5,037
Slope of p/Z line:
Δ(p/Z) / ΔG_p = (5,250 − 5,037) / (0 − 35) = −6.09 psia/Bscf
Extrapolating to p/Z = 0:
G (IGIP) = 5,250 / 6.09 = 862 Bscf (862×10⁹ scf)
Note: Real condensate needs two-phase Z correction (Walsh method or EOS).
Retrograde dropout below dew-point requires CVD data for accurate IGIP.
FVF and the Productivity Index: Closing the Loop
Every flow equation in Module 02 (Single-Phase Inflow Performance) will contain B_o or B_g in the denominator. The Darcy pseudo-steady-state equation for oil:
gives q in STB/d (surface barrels) from a reservoir-condition Darcy flow. The B_o converts from reservoir barrels per day (what Darcy computes) to surface barrels per day (what the production engineer measures). Similarly for gas:
The FVF therefore appears in every inflow equation, every nodal analysis calculation, every material balance, and every reserves estimate. It is the universal conversion between the two worlds of petroleum engineering. Mastering it here means every subsequent topic in the programme has a solid quantitative foundation.
PROGRAMME CONNECTION
FVF appears explicitly in:
Module 01 Topic 1.3: Relative permeability curves shift the effective kro in the PI denominator, compounding with Bo changes. Module 02: All single-phase IPR equations for oil and gas contain Bo or Bg. Module 04: Multiphase IPR (Vogel, Fetkovich) uses Bo inside the q_max term. Module 07: Nodal analysis VLP matching requires Bg for gas-lift calculations. Course 04: Scale deposition rates depend on actual brine Bw volumes. Course 10: Acid volume design uses Bo to convert from surface to in-situ placement volumes.
Assessment
Knowledge Check
Ten questions covering FVF definitions, correlations, physical behaviour, and engineering applications. Immediate feedback provided.
1. An oil reservoir has B_o = 1.42 res bbl/STB at bubble-point pressure. A well produces 800 STB/d at surface. What volume of reservoir fluid is being withdrawn per day?
B_o = reservoir volume / surface volume, so reservoir volume = surface volume × B_o = 800 × 1.42 = 1,136 res bbl/d. This is the actual volume being withdrawn from the reservoir per day — 42% more than the stock-tank volume because dissolved gas swells the oil at reservoir conditions. This is why material balance calculations always work in reservoir volumes before converting to surface volumes.
2. What happens to B_o as reservoir pressure drops from 4,500 psia to 2,000 psia, when the bubble-point pressure is 3,800 psia?
Above p_b = 3,800 psia: all gas remains dissolved, liquid is slightly compressed as p rises, so B_o decreases slightly from 4,500 to 3,800 psia (using the c_o correction). Below p_b: gas evolves from solution, the remaining oil has less dissolved gas and therefore less volumetric expansion at depth — B_o falls rapidly. At 2,000 psia the oil is significantly "dead-er" than at bubble point. B_t (total FVF including free gas) rises sharply below p_b.
3. Calculate B_g at p = 2,800 psia, T = 180°F (640°R), Z = 0.838. Which answer is correct?
B_g = 0.005035 × Z × T / p = 0.005035 × 0.838 × 640 / 2,800. Step through it: 0.005035 × 0.838 = 0.004219; × 640 = 2.700; ÷ 2,800 = 0.000964 res bbl/scf (equivalently 0.964 res bbl/Mscf). Option D is correct. The distractors each reflect a single common error: A keeps T in °F (must use °R = 180 + 460 = 640), B assumes the gas is ideal (Z = 1.0) when Z = 0.838 here, and C uses a wrong Z. Always apply B_g = 0.005035 Z T / p with T in °R and p in psia.
4. The total FVF is defined as B_t = B_o + (R_si − R_s) × B_g. At a pressure where R_s = 200 scf/STB (initial R_si = 600 scf/STB), B_o = 1.22 res bbl/STB, and B_g = 0.0025 res bbl/scf, what is B_t?
B_t = B_o + (R_si − R_s) × B_g = 1.22 + (600 − 200) × 0.0025 = 1.22 + 400 × 0.0025 = 1.22 + 1.00 = 2.22 res bbl/STB. The 400 scf/STB of evolved free gas now occupies 1.00 additional res bbl per STB of original oil, doubling the effective reservoir volume compared to B_o alone. This rapid rise of B_t below bubble point is the core driver of solution-gas drive, but also why GOR escalates and reservoir pressure collapses quickly.
5. An engineer uses differential liberation B_o data directly (without flash correction) to calculate well productivity. What systematic error does this introduce?
Differential liberation produces a slightly larger B_o (typically by 2–8% for black oils) than the flash-corrected value: the DL process preferentially strips lighter gas components at each step, leaving a heavier, smaller residual oil, so DL B_o = reservoir volume / DL residual volume comes out LARGER than the true B_o referenced to flash stock-tank oil. B_o appears in the DENOMINATOR of J* = kh/[141.2 B_o µ_o (ln(r_e/r_w) − ¾ + S)], so the inflated DL B_o produces a smaller calculated J*, a systematically PESSIMISTIC productivity estimate (option B). Option C states the same B_o direction but inverts the consequence: it is the classic trap of forgetting where B_o sits in the equation. (Note the direction differs for reserves: an inflated B_o makes volumetric STOIIP look smaller too.)
6. A gas reservoir has initial pressure p_i = 5,000 psia with Z_i = 0.91, and abandonment pressure p_ab = 500 psia with Z_ab = 0.95. Assuming volumetric depletion (no water influx), the recovery factor is approximately:
From p/Z analysis: RF = 1 − (p_ab/Z_ab)/(p_i/Z_i) = 1 − (500/0.95)/(5,000/0.91) = 1 − (526.3)/(5,494.5) = 1 − 0.0958 = 0.904 ≈ 90% (option C), reflecting typical high recovery from a well-connected volumetric gas reservoir. Compare this to ~15–25% for solution-gas drive oil — gas reservoirs recover far more in primary depletion because gas expands so dramatically (B_g rises 10× from initial to abandonment pressure).
7. Why does B_w remain very close to 1.0 for most reservoir brines, while B_o can range from 1.05 to 2.5?
The key drivers of large B_o are: (1) high dissolved gas content (R_s up to 2,000 scf/STB for volatile oils) which enormously swells oil at reservoir conditions, and (2) thermal expansion of liquid hydrocarbons (which are more thermally expansive than water). Water dissolves <30 scf/STB of gas at most reservoir conditions, orders of magnitude less than oil. Water's bulk modulus (~300,000 psi vs. ~20,000 psi for oil) makes it nearly incompressible. Therefore both the "dissolved gas swelling" and "liquid compressibility" effects that make B_o large are absent for water, keeping B_w in the 1.01–1.07 range.
8. In the pseudo-steady-state productivity index formula J* = kh / [141.2 B_o µ_o (ln(r_e/r_w) − 3/4 + S)], B_o appears in the denominator. A well initially producing at 4,500 psia (B_o = 1.25) later produces below bubble point at 2,500 psia where B_o = 1.18 but µ_o has increased from 0.9 to 1.4 cp. What is the ratio of new J* to original J*?
J* ∝ 1/(B_o × µ_o). Ratio = (B_o×µ_o)_initial / (B_o×µ_o)_new = (1.25×0.9) / (1.18×1.4) = 1.125 / 1.652 = 0.68. Option D is correct. The productivity drops ~32% — primarily driven by the viscosity increase (µ_o rose 56%) rather than the B_o change (only 6% decrease). This illustrates that below bubble point, the PI changes for two compounding reasons: B_o falls (good, less reservoir volume per STB) but µ_o rises (bad, heavier oil flows slower) and the viscosity effect usually dominates. Additionally, k_ro falls as gas saturation builds, not captured in this simplified calculation but covered in Topic 1.3.
9. Which of the following statements about the Standing correlation for B_o is FALSE?
Option C is FALSE and is therefore the correct answer. Standing's correlation was derived exclusively from California crude oil systems: it is NOT universally superior to other correlations for oils from other basins. The Vasquez-Beggs correlation (1980), derived from over 6,000 global data points, is generally more accurate for non-California oils. Middle East oils are better described by Al-Marhoun (1988). Gulf of America oils by Petrosky-Farshad (1993). Options A, B, and D are all TRUE statements about Standing's correlation.
10. A gas well produces 5 MMscf/d. The reservoir B_g = 0.004 res bbl/scf. What reservoir withdrawal rate does this represent in res bbl/d?
Reservoir withdrawal = surface gas × B_g. Note units: B_g = 0.004 res bbl/scf. Surface production = 5 MMscf/d = 5,000,000 scf/d. Reservoir withdrawal = 5,000,000 × 0.004 = 20,000 res bbl/d. This is a useful sanity check: a 5 MMscf/d gas well withdraws only 20,000 res bbl/d because gas is highly compressed at reservoir pressure. Compare to an oil well producing 5,000 STB/d with B_o = 1.35: that withdraws 6,750 res bbl/d — far less than the gas well in reservoir volume terms, even though the oil production rate seems much lower in surface units.
NEXT STEPS
Topic 1.2 complete! You can now convert confidently between reservoir and surface volumes for oil, gas, and water systems. Proceed to:
→ PBL Sub-Problem SP-2 (GK-22): Calculate B_g for GK-22 at initial conditions and at separator conditions. Use Bg and the condensate yield to estimate total reservoir voidage rate, and verify with the p/Z approach. Use Worked Example 2 as your template.