Scribd
Upload
6 views3 pages

Transient Reservoir Simulation Methods

The document discusses transient problems in numerical reservoir simulation, focusing on time-dependent changes in pressure and saturation. It outlines three time discretization techniques: Explicit, Implicit, and Crank-Nicolson, along with their pros and cons. Additionally, it covers stability analysis, emphasizing the importance of ensuring numerical methods do not amplify errors.

Uploaded by

Emre Köçkar
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
6 views3 pages

Transient Reservoir Simulation Methods

The document discusses transient problems in numerical reservoir simulation, focusing on time-dependent changes in pressure and saturation. It outlines three time discretization techniques: Explicit, Implicit, and Crank-Nicolson, along with their pros and cons. Additionally, it covers stability analysis, emphasizing the importance of ensuring numerical methods do not amplify errors.

Uploaded by

Emre Köçkar
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

PetE 507 - Numerical Reservoir Simulation

8. Unit and After: Deep Conceptual Review

I. Transient (Time-Dependent) Problems

▶ A. Concept

• These are problems where the state of the system evolves over time.
• In reservoir simulation, this typically involves pressure and/or saturation changing in the
domain.

▶ B. Governing Equation (1D Diffusive Flow)

∂p ∂ ∂p
ϕ = (λ )
∂t ∂x ∂x

Where: - p : pressure - ϕ : porosity - λ : mobility (λ = k


μ )- t : time - x : spatial coordinate

This is a parabolic PDE (diffusion-type).

▶ C. Time Discretization Techniques

We aim to calculate pn+1


i : pressure at grid point i and time n + 1 .

1. Explicit Method (Forward Euler)

λΔt
pn+1 = pni + (pn − 2pni + pni−1 )
i
ϕΔx2 i+1
- Pros: Simple to code, fast per step - Cons: Conditionally stable (requires small Δt )

2. Implicit Method (Backward Euler)

pn+1 − pni λ
i
= (pn+1 − 2pn+1 + pn+1
i−1 )
Δt ϕΔx2 i+1 i

- Pros: Unconditionally stable, can use large Δt - Cons: Need to solve a system of equations at each
time step

3. Centered (Crank-Nicolson)

pn+1 − pni λ
i
= [(pni+1 − 2pni + pni−1 ) + (pn+1 n+1
i+1 − 2pi + pn+1
i−1 )]
Δt 2ϕΔx2
- Pros: Second-order accuracy in time - Cons: Still requires solving a linear system

1
▶ D. Physical Insight

• Think of pressure "propagating" through the reservoir.


• Explicit: quick, but potentially unstable — like taking big steps in darkness.
• Implicit: careful and stable — like slowly feeling your way forward.
• Crank-Nicolson: a balance — predicting and correcting at once.

▶ E. Sample Numerical Output (t = 0.1s, 5 grid points)

Method p2 p3 p4

Explicit 10 1 0

Implicit 8.47 1.36 0.11

Crank-Nicolson 4.56 1.23 0.06

II. Stability Analysis

▶ A. Why?

We must ensure numerical methods don’t amplify errors. A numerically unstable scheme leads to non-
physical and divergent results.

▶ B. Von Neumann Stability Analysis (for linear PDEs)

Assume a solution of form:

ϵni = An eikxi

Where: - ϵ : numerical error - A : amplification factor - k : wave number

If ∣A∣ ≤ 1 , the scheme is stable.

For Explicit Scheme:

A = 1 − 4r sin2 ( ),
kΔx λΔt
with r =
2 ϕΔx2
1
- Stability condition: r ≤ 2

For Implicit Scheme:

A ≤ 1 always → Unconditionally stable

▶ C. Physical Interpretation

• Explicit: like stepping ahead without knowing what's next — can stumble unless careful.
• Implicit: like recalculating based on what's to come — always plays it safe.

2
▶ Final Tips

• In exam, if asked for "time discretization methods" → mention all three with stability and matrix
insight.
• For stability → always discuss amplification factor and concept of unconditional stability.

You might also like