PRACTICAL RESERVOIR MONITORING

Material Balance

Reservoir Material Balance

and
Analytical Reservoir Simulation

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Material Balance

F = N . Et + We

in which
• F is the underground fluid withdrawal
• N is the original oil in place (OOIP)
• Et is the total expansion of the reservoir-fluid system
• We is the water influx from the aquifer.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Underground Withdrawal

F = N p [ Bo + ( Rp – Rs ) Bg ]

in which
• Np is the cumulative oil production
• Bo is the oil FVF
• Rp is the production GOR
• Rs is the solution GOR
• Bg is the gas FVF

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Total System Expansion

Et = Eo,g + Egc + Ecw – Epv

in which
• Eo,g is the expansion of the oil and dissolved gas.
• Egc is the expansion of the gas cap.
• Ecw is the expansion of the connate water.
• Epv is the contraction of the pore volume under reduction of pressure.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Expansion of the Oil and Dissolved Gas

Eo,g = N [ ( Bo – Boi) + ( Rsi – Rs ) Bg ]

N is the OOIP
Bo is the current oil FVF
Boi is the initial oil FVF
Rsi is the initial solution GOR
Rs is the current solution GOR
Bg is the gas FVF

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Expansion of the Gas Cap

Egc = m N Boi [ Bg / Bgi – 1 ]

m is the ratio of initial reservoir free gas volume, over the initial reservoir oil
volume.
N is the OOIP.
Boi is the initial oil FVF.
Bg is the gas FVF.
Bgi is the initial gas FVF.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Expansion of the Connate Water

Ecw = ( 1 + m ) N Boi Swc Cw ∆P / ( 1 – Swc )

m is the ratio of initial reservoir free gas volume, over the initial reservoir oil
volume.
N is the OOIP.
Boi is the initial oil FVF.
Swc is the connate water saturation.
Cw is the connate water compressibility.
∆P is the pressure drop since start of production.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Contraction of the Pore Volume

Epv = ( 1 + m ) N Boi Cf ∆P / ( 1 – Swc )

m is the ratio of initial reservoir free gas volume, over the initial reservoir oil
volume.
N is the OOIP.
Boi is the initial oil FVF.
Cf is the total fluid compressibility.
∆P is the pressure drop since start of production.
Swc is the connate water saturation.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Material Balance: History Matching Applications.

History matching is the most effective way to use material balances. Both graphical
and analytical methods are available.

• Determination of the drive mechanism: gas cap, solution gas, aquifer,

combination.
• Modeling of the aquifer: geometry (edge drive radial or linear, bottom drive),
size, and strength.
• Determination of the OOIP.
• Fractional flow matching and optimization of total recovery factors.
• Expected movements of fluid contacts.
• Reservoir compartmentalization.
ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Aquifer Models

• Small Pot. The influx is only a function of the reservoir pressure.

• Schilthuis. The reservoir is assumed to be in steady state. Model produces an overall
influx coefficient with summation of the pressure drops.
• Hurst – Van Everdingen. Uses a superposition of states for a reservoir in radial flow
with an aquifer. Widely used but needs a high density of steps at early times. A recent
variation of this method interpolates pressure steps to approximate a linear pressure
profile. Greatly improves accuracy of matching.
• Vogt – Wang. A variation of the above mostly used in case of scarce data.
• Fetkovitch. Straight analytical computation without superposition. Now mostly
superseded.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Pressure and Production History

• Because a material balance is zero-dimensional, the average reservoir pressure must

be computed at each time step.

• This can be done by planimetering isobaric and pore volume maps. In case these are
not available, volumetric data can be inferred from the production history under a
constant compressibility hypothesis.

• The production history must be entered for the whole reservoir. Alternatively, and
especially in the case of multi-tank systems, wells must be pooled into tanks, or
sometimes entered individually. Both pressure and cumulative production data will be
entered.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
History Matching Methods

The idea is to plot the pressure – production history in a number of ways to solve the
material balance equation
F = N . Et + We
1. Graphical.
• F / Et vs F
• F – We vs Et
• F / Et vs We / Et

• F / Et vs Σ [ ( δP Qtd ) / Et ] (Havlena & Odeh)

• P / Z for gas reservoirs
2. Analytical.
• Multivariate regression methods ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
F / Et vs We / Et : Aquifer and OOIP

• F / Et = N + We / Et
• The plot will yield a unit
slope straight line when
We has been properly
estimated.
• The Y-axis intercept
yields N, the OOIP.
• This method has a
greater sensitivity than the
F – We vs Et method.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Diagnostics: WD vs tD Plot

• The plot of the dimensionless

aquifer influx WD vs dimensionless
time tD allows to determine whether
the aquifer boundary has been felt
(upper plot), or whether there is
insufficient production history to
distinguish between an infinite acting
or a bounded aquifer (lower plot).
• If the aquifer is found to be
bounded, the WD type curve match
yields the aquifer volume.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Gas Wells : Plot of P / Z vs Cumulative Production

• This plot is interpreted by

extrapolating the gas
potential function P / Z down
to abandonment pressure.
• The X-axis intercept yields
the GOIP.
• If the gas pressure is
abnormally high, the P / Z
function must take account
of variations in gas
compressibility.
• The Havlena – Odeh
method also applies to gas
reservoirs.
ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Analytical History Matching

• Once the reservoir

parameters have been
determined by previous
analyses, the history
match is corroborated by
an analytical simulation.
• The effect of the
aquifer can be gauged by
comparing the data
points to the curve
representing the
performance of the
reservoir if no aquifer
had been present.
ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Relative Permeabilities

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Fractional Flow Matching

• Fw matching is used to
extrapolate the production
history to the value Fw = 1.
• For that value, all the
producible reserves have
been produced under the
assumptions governed by
the model.
• Reservoir optimization
strategies attempt to shift
the end water saturation to
the right of the plot,
bringing in additional oil at
high water saturations.
ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Fractional Flow Matching

• The fractional flow of water is defined as

Fw = Qw Bw / ( Qw Bw + Qo Bo ), or

Fw = 1 / [ 1 + ( Kro / µo) / ( Krw / µw ) ]

• The field-wide fractional flow can be matched with field-wide values of the water
saturation in order to yield an estimate of the total hydrocarbon recovery from the
field (corresponding to Sw = 1).

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Fractional Flow Matching

• The fractional flow of water or gas can be history matched with saturation data to
validate the choice of Corey exponents for the relative permeability curves:

Krx = Ex [ ( Sx – Srx) / ( SMx – Srx ) ] ** nx

Sx is the phase saturation,
Srx is the phase residual saturation,
SMx is the maximum saturation of this phase (end point of the Kr curve),
nx is the Corey exponent and Ex is the end point value for this phase.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Movements of Fluid Contacts

• If the distribution of the

pore volume vs depth is
known, material balances can
be used to calculate the
expected level of the fluid
contacts at any point in time.
• This can be used in
predictive mode to predict
water or gas breakthrough in
wells.
• Because of the zero-
dimensional nature of this
approach, measurements
must be generated to validate
the results. ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Reservoir Compartmentalization

• It may not be possible to match the pressure of a reservoir over the entire
production history.
• Pressure maintenance schemes may reveal heterogeneities. In particular, wide
sections of a reservoir may be unswept by injection waters, or by an active aquifer.
• Field data such as wireline pressure measurements, interference tests, production
logs etc... might further contribute to the need to consider that the reservoir may be
composed of several compartments or ‘tanks’.
• Each compartment may function as a separate reservoir, only partially in
communication with other compartments, although sharing some common features,
for example a gas cap or an aquifer, or a common waterflood scheme.
• Compartments may be separated by horizontal permeability barriers, by partially
sealing faults, or by stratigraphic features lacking lateral continuity.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Reservoir Compartmentalization (cont’d)

• Material balance calculations can still be done if the reservoir is

compartmentalized.
• Each of the compartments must have a distinctive pressure-production history.
• Average pressures must be calculated for each of the coimpartments.
• Specific transmissibility parameters must be assigned to each of pre-defined
‘leaks’ between contiguous compartments.
• Even though this methodology allows to work out a solution, it increases the
degrees of freedom of the whole system, hence it enhances the potential for non-
uniqueness of the analytical solution.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Forward Prediction Applications

• Evaluation of development options.

• Compare prediction runs with and without water injection.
• Design drilling schedules to enable and lengthen plateau production.
• Use voidage replacement to estimate water injection requirements.
• Investigate the feasibility of recycling the condensate.
• Completion economics.
• Compare production prediction under different scenarios.
• Production forecasting.
• Predict future production rates.
• Define and use constraints at well and platform level.
• Meet production and reservoir performance targets. ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Forward Prediction Options

• Given a production schedule, calculate the reservoir pressure.

• Given a production schedule, calculate the reservoir pressure and the manifold
pressure.
This option uses individual well lift curves to find the manifold pressure
required to meet the production schedule.
• Given the manifold pressure, calculate the reservoir pressure and the production
rate.
This option also needs well lift curves.
• Gas contract calculations.
• Make reservoir pressure predictions to honour contract schedules: DCQ and
swing.
• Find the DCQ for a particular swing and field configuration.
ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance

Numerical Simulation

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Analytical vs Numerical Reservoir Simulation

1. Analytical reservoir simulation (Material Balance).

• will provide a good, dynamic estimate of the OOIP.
• will identify the reservoir drive, the components of the reservoir energy,
and the reservoir compartmentalization.
• allows to calculate the movement of fluid contacts if the vertical
distribution of the pore volume is known.
• is zero-dimensional and does not require a geological model.
• cannot be expected to deliver unique results under all conditions.
• requires PVT data of high quality.
• is suggested as a necessary, prior step to carrying out a numerical
simulation study.
ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel

 PRACTICAL RESERVOIR MONITORING
Material Balance
Analytical vs Numerical Reservoir Simulation (cont’d)

2. Numerical reservoir simulation.

• Assumes N from the geological model.
• Uses estimates for the water influx We and total fluid compressibility Cf,
and history matches pressure and saturations.
• The quality of history matching feeds back onto the model assumptions.
• Is better adapted than MB to deliver forward predictions.

ΓΧ Gamma Experts
Petroleum Engineering

September 2002 Yves Chauvel