On the Existence of Steady-State Solutions to the Equations Governing Fluid Flow in Networks

Shriram Srinivasan, Nishant Panda, Kaarthik Sundar

Introduction

Engineers managing massive pipeline networks for natural gas, water, and carbon capture rely on decision-support tools to solve complex nonlinear algebraic equations. A persistent challenge in this field is algorithmic failure: when a simulation fails to converge, it is often unclear whether the issue lies in the numerical stiffness of the problem, a poorly tuned algorithm, or the physical non-existence of a solution. This paper addresses this theoretical gap by providing a rigorous proof for the existence and uniqueness of steady-state solutions for fluids following a "scaled monomial" equation of state.

Theoretical Framework: Potential-Flow and Homotopy

The authors bridge the gap between theory and practice using Topological Degree Theory and Homotopy Invariance. The approach involves two primary steps:

  1. Potential-Flow Formulation: The equations are transformed from pressure space ($p$) into a "potential" space ($\pi$), simplifying the mathematical structure.
  2. Homotopy and Degree Theory: The researchers define a continuous transformation (a homotopy) that begins with a simple, guaranteed linear system and gradually "warps" it into the target nonlinear system. By demonstrating that the topological degree remains non-zero and that solutions remain within a bounded "hypercube" throughout this transformation, they prove that a solution must exist for the final system.

Approximating Non-Ideal Gases

While the existence proof holds strictly for "scaled monomial" equations of state (like ideal gases), many real-world applications involve non-ideal gases (e.g., using the CNGA equation). To handle these, the authors developed a high-fidelity approximation using a least-squares approach to find an optimal monomial function that mimics non-ideal behavior.

Approximation error comparison for different monomial functions $g(\alpha)$ across pressure and temperature ranges.

This "optimal approximant" ensures that the resulting system is mathematically well-behaved while maintaining high physical accuracy, providing a robust starting point for numerical solvers.

Numerical Validation and Results

The theoretical findings were validated using the GasLib-40 benchmark library. The proposed approximate solution for non-ideal gases demonstrated remarkable accuracy and computational utility:

  • High Accuracy: Across 1,000 test instances, the approximation maintained a relative error of approximately $O(10^{-3})$ and a residual of $O(10^{-2})$.
  • Computational Speedup: Using this guaranteed approximate solution as an initial guess for standard Newton-Raphson solvers resulted in convergence in exactly two iterations across all test cases.
Cumulative distribution of relative error and residual for 1,000 instances of the GasLib-40 network, demonstrating the high accuracy of the monomial approximation.

Conclusion and Diagnostic Power

This work effectively "closes the book" on the existence question for ideal gas steady-state flows and provides a robust framework for non-ideal gases. Beyond theoretical interest, these results serve as a powerful diagnostic tool. If a mathematical solution exists but requires physically impossible parameters (e.g., negative pressures), operators can definitively identify the system as infeasible rather than blaming software failure. Future work aims to extend these homotopic methods to transient flows and more complex pumping models.

Summary updated

BibTeX
@article{srinivasan2024existence,
  author = {Shriram Srinivasan and Nishant Panda and Kaarthik Sundar},
  title = {{On the Existence of Steady-State Solutions to the Equations Governing Fluid Flow in Networks}},
  journal = {IEEE Control. Syst. Lett.},
  year = {2024},
  doi = {10.1109/lcsys.2024.3394317},
}
Figure