Material Balance and Reservoir Analysis

Reservoir engineers need to answer two fundamental questions: How much oil is in the ground? and What is pushing it out?

The General Material Balance Equation (GMBE) answers both. It is the single most important equation in reservoir engineering --- a conservation-of-mass accounting that relates everything produced from the reservoir to the energy sources driving that production. When applied correctly, it can estimate original oil in place (OOIP), identify the dominant drive mechanism, predict future performance, and diagnose whether a reservoir is behaving as expected.

The GMBE is also one of the most misapplied equations in the industry. It has many terms, each representing a different physical mechanism. Using it carelessly --- ignoring terms, misidentifying the drive, or working with poor pressure data --- leads to OOIP estimates that can be off by factors of two or more. Every term in this equation has a physical meaning, and this chapter ensures you understand all of them.

When a well produces oil, something must replace the volume that was removed. Otherwise, the reservoir pressure would drop instantly to zero. The energy source that maintains pressure and pushes oil toward the wellbore is called the drive mechanism.

Understanding the drive mechanism is critical because it determines recovery efficiency, pressure behavior, and the appropriate production strategy. The main mechanisms are:

All reservoir fluids are slightly compressible. When pressure drops, oil, water, and rock expand slightly. This expansion pushes oil toward the wellbore. The effect is small --- compressibilities are on the order of $10^{-6}$ psi$^{-1}$ --- so fluid expansion alone typically recovers only 1--5% of the original oil. This drive is most significant in undersaturated reservoirs (reservoirs where the pressure is above the bubble point, meaning all gas is dissolved in the oil).

When reservoir pressure drops below the bubble point, dissolved gas comes out of solution and forms free gas bubbles within the pore space. As the gas expands, it pushes oil toward the well. Recovery factors are typically 5--30%.

The tradeoff: as gas liberates, the oil's viscosity increases and its volume decreases. The gas-oil ratio (GOR) at the surface rises dramatically. Solution gas drive is efficient early but becomes progressively less effective as the gas cap grows and gas preferentially flows to the wellbore (gas is more mobile than oil).

If the reservoir has a gas cap --- a zone of free gas sitting above the oil column --- the expanding gas cap pushes the oil-water contact downward and sweeps oil toward production wells. Recovery factors range from 20--40% depending on the size of the gas cap relative to the oil zone.

Operators must manage gas cap drive carefully. Producing too aggressively from the top of the structure can cause gas coning, where gas breaks through to the wellbore prematurely, reducing oil recovery.

An aquifer connected to the reservoir provides pressure support as water moves in to replace the produced fluids. Water drive is the most efficient natural mechanism, with recovery factors of 30--60% or higher.

The strength of the water drive depends on the aquifer's size, permeability, and compressibility. A strong aquifer maintains reservoir pressure nearly constant throughout the field's life. A weak aquifer provides partial support, and pressure declines gradually. Many Niger Delta reservoirs (including fields in OML 58) benefit from strong aquifer support, making water influx modelling a critical part of reserves estimation in that basin.

In some reservoirs, particularly unconsolidated sands, the reduction in pore pressure causes the rock matrix itself to compact. This squeezes fluid out of the pore space. The Ekofisk field in the North Sea is the most famous example, where seafloor subsidence of several meters accompanied significant compaction drive.

The GMBE is a volumetric balance: the total volume of fluids withdrawn from the reservoir must equal the total expansion of fluids and rock remaining in the reservoir, plus any water that has entered from an aquifer.

where the withdrawal and expansion terms are defined as:

Each term represents a physical mechanism:

Left side --- total withdrawal:

• $N_p[B_o + (R_p - R_s)B_g]$ --- total reservoir volume of oil and its associated gas that has been produced
• $W_p B_w$ --- reservoir volume of water produced

Right side --- sources of energy:

• $N(B_o - B_{oi})$ --- expansion of oil (change in oil formation volume factor)
• $N(R_{si} - R_s)B_g$ --- expansion of gas liberated from oil as pressure drops below bubble point
• $N B_{oi}\frac{(c_w S_{wi} + c_f)\Delta P}{1 - S_{wi}}$ --- expansion of connate water and reduction in pore volume due to rock compressibility
• $\frac{mNB_{oi}}{B_{gi}}(B_g - B_{gi})$ --- expansion of the gas cap, where $m$ is the ratio of gas cap to oil zone volume
• $W_e$ --- net water influx from an aquifer

Every symbol has a specific physical meaning. Every term represents a real phenomenon occurring in the reservoir.

Solving the GMBE directly for $N$ is difficult because it contains multiple unknowns ($N$, $m$, $W_e$). Havlena and Odeh (1963) rearranged the equation into a straight-line form that allows graphical (and computational) determination of these unknowns.

Using the grouped terms ($F$, $E_o$, $E_g$, $E_{fw}$) defined above, the GMBE is already in a compact form:

For different reservoir types, this reduces to specific straight-line forms.

For an undersaturated reservoir with no gas cap ($m = 0$) and no aquifer ($W_e = 0$), the GMBE simplifies to:

A plot of $F$ vs. $(E_o + E_{fw})$ should yield a straight line through the origin with slope $N$.

A straight line through the origin confirms two things: the OOIP estimate is reliable, and the assumed drive mechanism (depletion only, no water influx, no gas cap) is consistent with the data. If the points curve upward, it suggests water influx is present. If they curve downward, the reservoir may have a gas cap that was not accounted for.

The shape of the Havlena-Odeh plot reveals which drive mechanisms are active. This is one of the most powerful diagnostic tools in reservoir engineering.

When the diagnostic indicates water influx, you need a model to estimate $W_e$ at each time step. Three models are commonly used, each with different assumptions about the aquifer.

The simplest model. Assumes water influx is proportional to the pressure drop:

where $k$ is a constant determined by regression.

Treats the aquifer as a finite tank with its own compressibility and initial volume:

where $W_{ei}$ is the initial encroachable water volume in the aquifer, $c_{aq}$ is the total aquifer compressibility, and $\bar{P}_{aq}$ is the current average aquifer pressure.

The most rigorous model. Solves the radial diffusivity equation for the aquifer and uses dimensionless time and pressure functions. This model is appropriate for large or infinite aquifers where transient effects are significant.

The ultimate test of a material balance model is whether it can reproduce the observed pressure-production history. History matching involves adjusting uncertain parameters (OOIP, aquifer properties, gas cap size) until the model's predicted pressure matches the measured pressure at every time step.

• Drive mechanisms determine how much oil can be recovered and how the reservoir pressure behaves. Water drive is the most efficient (30–60% recovery); solution gas drive is the least (5–30%).
• The General Material Balance Equation is a conservation-of-mass balance that accounts for every source of energy in the reservoir. Each term represents a specific physical mechanism.
• Havlena-Odeh rearranges the GMBE into straight-line form. The slope gives OOIP ($N$), and deviations from linearity diagnose the drive mechanism.
• F/Et trending upward indicates water influx. Trending downward suggests an unaccounted gas cap or PVT errors. Stable confirms depletion drive.
• Water influx models range from simple (Schilthuis steady-state) to rigorous (van Everdingen-Hurst unsteady-state). The choice depends on aquifer size and data quality.
• History matching validates the material balance model by checking whether predicted pressures match observed pressures. If they do not, the assumed OOIP or drive mechanism needs revision.
• Material balance is only as good as its input data. Average reservoir pressure, accurate PVT properties, and reliable production records are prerequisites.