In ETAP (Electrical Transient Analyzer Program), the Newton-
Raphson method is the primary and most widely used calculation method for load
flow studies. While ETAP provides several options to balance speed and accuracy,
the Newton-Raphson method is preferred for its fast convergence and high precision
in large, complex power systems.
Available Calculation Methods in ETAP
ETAP allows users to select from several numerical techniques in the Load Flow
Study Case Editor based on the specific requirements of the electrical network:
Newton-Raphson (N-R): The industry standard for complex systems. It uses
iterative calculations with a Jacobian matrix to solve non-linear power
equations, offering quadratic convergence.
Adaptive Newton-Raphson (ANR): A variation of the standard N-R method
that automatically adjusts parameters to improve convergence in difficult-to-
solve networks.
Fast Decoupled (FD): A faster but slightly less accurate method compared to
N-R. It is often used for very large systems where computational speed is
more critical than extreme precision, as it decouples real and reactive power
equations.
Accelerated Gauss-Seidel: Though ETAP typically uses a few Gauss-Seidel
iterations at the start of an N-R calculation to establish stable initial values, the
full method is generally only used for smaller, simpler networks because it
converges more slowly than Newton-Raphson.
Summary of Method Characteristics
Method Best Use Case Key Strength Weakness
Newton- Large, complex, or meshed Fast convergence and High computer memory
Raphson systems high accuracy requirements
Fast Very large networks or real- 5x faster than Newton- Less accurate due to
Decoupled time management Raphson assumptions
Gauss-Seidel Small, radial networks Simple and low memory Very slow to converge on
usage large systems
Pro Tip: For most industrial and utility-scale projects, you should stick with
the Newton-Raphson method in ETAP to ensure the most reliable steady-state
results for voltage magnitude and phase angle at each bus.
Newton-Raphson or adaptive Newton-Raphson, which method is efficient
In ETAP, the choice between
Newton-Raphson (N-R) and Adaptive Newton-Raphson (ANR) depend on how
"well-behaved" your electrical system is. While both are highly efficient, they serve
slightly different purposes.
1. Newton-Raphson (The Standard)
The standard Newton-Raphson method is generally considered the most
efficient for systems that are properly modelled and have a clear solution.
Efficiency: It features "quadratic convergence," meaning the error decreases
very rapidly with each iteration.
Best For: Robust, well-designed networks where voltage profiles are
relatively stable.
Why use it: It is the fastest way to get a high-precision result if the system is
not near its physical limits.
2. Adaptive Newton-Raphson (The "Troubleshooter")
The Adaptive Newton-Raphson method is more "efficient" at reaching a solution
in difficult cases where the standard method might fail or diverge (blow up).
Efficiency: It is technically slower per iteration because it uses a "Step-Size
Control" (acceleration factor). It automatically adjusts the calculation steps to
prevent the math from overshooting the target.
Best For:
o Systems with extremely high R/X ratios (like low-voltage distribution
lines).
o Networks with very low voltages or heavy loading.
o Systems that show a "Divergence" error when using the standard N-R
method.
Why use it: It saves time by finding a solution for "stiff" or problematic
networks that would otherwise require you to manually troubleshoot
convergence errors.
Comparison Summary
Feature Newton-Raphson Adaptive Newton-Raphson
Speed Faster (fewer iterations) Slower (more cautious steps)
Stability Moderate High
Complexity Standard High (Adjusts to system conditions)
Default Choice Use this first Use if Standard N-R fails to converge
The Verdict
For 90% of studies, the Newton-Raphson method is the most efficient choice.
However, if ETAP returns an error saying "Exceeded Maximum Iterations" or the
system is failing to converge, switching to Adaptive Newton-Raphson is the most
efficient way to get a result without redesigning your entire model.
Would you like me to show you how to change the Iteration
Limit or Precision settings in the Study Case editor to help a stubborn model
converge?