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

Shriram Srinivasan, Nishant Panda, Kaarthik Sundar

TL;DR

This paper resolves a long-standing theoretical gap in pipeline network modeling by proving the existence of steady-state solutions for fluids with monomial equations of state (like water or ideal gases) when formulated using potentials instead of pressures. For complex non-ideal gases, the authors develop a guaranteed-solvable proxy system that drastically accelerates algorithm convergence and enables robust infeasibility detection in network optimization.

What Problem Does This Paper Address?

Engineers and operators rely heavily on decision-support software to manage fluid flow through massive pipeline networks—whether distributing natural gas, managing water networks, or transporting carbon dioxide for carbon capture. These tools solve systems of nonlinear algebraic equations representing steady-state conditions.

While prior work proved that these equations have a unique solution if one exists, nobody had rigorously proven that a solution is mathematically guaranteed to exist. This lack of an existence proof created a major headache for algorithm developers: if a solver failed to converge on a large grid, it was impossible to tell whether the algorithm itself had failed due to numerical stiffness, or if the physical problem simply had no valid mathematical solution.

Methods

The authors tackle this problem using topological degree theory and homotopy invariance. Instead of trying to analyze the complex nonlinear system directly, they start with a simple, structured linear system where a unique solution is easily guaranteed. They then construct a continuous mathematical transformation (a homotopy) that gradually deforms this simple linear system into the target nonlinear network flow system. By proving that the "topological degree" remains non-zero throughout this deformation and that solutions cannot escape a bounded region (a hypercube), they guarantee that the final nonlinear system must possess a solution.

Crucially, this proof requires shifting the mathematical formulation from pressures and flows to potentials and flows. For non-ideal gases governed by the complex CNGA (California Natural Gas Association) equation of state, the monomial property breaks down. To circumvent this, the authors use a least-squares optimization to find an optimal polynomial approximation ($\gamma_{opt}$) that mimics the monomial property, creating a proxy system that is mathematically guaranteed to have a unique solution.

Datasets

The authors validated their approximation and solution method using GasLib-40, a standard, publicly available benchmark library containing realistic data (pipe diameters, friction factors, and nodal injections) for a 40-node pipeline network.

Key Insights and Ideas

  • Variables Matter: Transforming the network equations from the pressure-flow domain into a potential-flow domain is the key that unlocks the proof of existence for scaled monomial equations of state.
  • Guaranteed Initial Guesses: For non-ideal gases where a strict existence proof remains an open question, the proxy system provides a mathematically guaranteed solution that serves as an exceptionally high-quality initial guess for standard numerical solvers.
  • Definitive Infeasibility Detection: Knowing that a unique potential solution always exists allows operators to definitively identify when an Optimal Gas Flow (OGF) optimization problem is physically impossible to satisfy, rather than just numerically stubborn.

Key Results

  • Theoretical Guarantee: The paper proves that a unique generalized potential-solution always exists for ideal gases, water, and CO2 networks under standard, realistic network assumptions.
  • High-Fidelity Proxy: For non-ideal gases, the optimized proxy function ($\gamma_{opt} \approx 0.9\alpha^2 + 0.1\alpha^3$) reduces approximation error drastically compared to standard ideal-gas assumptions.
Relative error distribution for different choices of $g(\alpha)$
  • Rapid Convergence: Tested across 1,000 randomized instances of the GasLib-40 network, using the proxy system's solution as an initial guess allowed a standard Newton-Raphson solver to converge to the true non-ideal gas solution in exactly 2 iterations every single time, yielding a residual error of $O(10^{-2})$ and a relative error of $O(10^{-3})$.

Limitations and Future Work

  • Non-Ideal Gas Existence Still Open: While the proxy system is guaranteed to have a solution and approximates the true system beautifully, a formal mathematical proof of existence for the exact, unapproximated CNGA equation of state remains an open challenge. This is because bounding the solution space during the homotopy transformation becomes highly complex without the monomial property.
  • Future Directions: The authors suggest that future work should focus on developing constructive homotopic continuation methods to trace the solution trajectory directly from the simplified system to the true non-ideal gas system.

Summary updated

Abstract

The steady-state solution of fluid flow in pipeline infrastructure networks driven by junction/node potentials is a crucial ingredient in various decision-support tools for system design and operation. While the nonlinear system is known to have a unique solution (when one exists), the absence of a definite result on the existence of solutions hobbles the development of computational algorithms, for it is not possible to distinguish between algorithm failure and non-existence of a solution. In this letter, we show that for any fluid whose equation of state is a scaled monomial, a unique solution exists for such nonlinear systems if the term solution is interpreted in terms of potentials and flows rather than pressures and flows. However, for gases following the CNGA equation of state, while the question of existence remains open, we construct an alternative system that always has a unique solution and show that the solution to this system is a good approximant of the true solution. The existence result for flow of natural gas in networks also applies to other fluid flow networks such as water distribution networks or networks that transport carbon dioxide in carbon capture and sequestration. Most importantly, our result enables correct diagnosis of algorithmic failure, problem stiffness, and non-convergence in computational algorithms.
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