Dynamics of Heterogeneous Gas Mixtures in Pipeline Networks

20 concepts7 papers2 proposals

Traditional energy infrastructure is built on the assumption of homogeneity. When natural gas flows through a continental pipeline network, operators treat the fluid as a single, uniform substance with consistent physical properties like density, viscosity, and calorific value.

However, the transition to low-carbon energy systems upends this assumption. Blending "green" hydrogen into existing natural gas networks is a promising strategy for decarbonization, but it introduces a fundamental physical challenge: the gas flowing through the network becomes a heterogeneous mixture whose composition varies continuously across both space and time. Because hydrogen and natural gas (primarily methane) possess vastly different physical properties, treating the mixture as homogeneous is no longer sufficient.

A growing body of research addresses this challenge by developing mathematical frameworks to model, optimize, and control the spatio-temporal composition variability of multi-substance gas flows.

Why Heterogeneity Breaks Traditional Models

To understand why heterogeneous mixtures require a paradigm shift, consider the physical differences between hydrogen and methane. Hydrogen has a much lower density and a significantly higher wave speed (speed of sound in the medium) than natural gas. As highlighted by Fazeny et al., existing optimal transport frameworks for gas networks typically assume a single substance, completely failing to account for the complex interactions and physical variations that occur when different gases are blended.

When hydrogen is injected at distinct nodes in a network, it does not instantly homogenize. Instead, pockets of varying concentrations move through the pipelines. These concentration gradients alter the underlying fluid dynamics in highly non-linear ways:

  1. Pressure Drop Alterations: The pressure gradient along a pipe depends heavily on the fluid's density and friction. Sodwatana et al. demonstrated that standard steady-state pressure drop equations must be modified to incorporate varying mass concentrations to accurately predict how different hydrogen-to-natural-gas ratios influence pressure gradients across a network.
  2. Extreme Sensitivity: Even minor variations in concentration can cause disproportionately large operational shifts. For instance, Baker et al. observed that a fixed 10% hydrogen injection could cause a substantial 2 MPa pressure increase over just 8 hours, drastically altering compressor workloads.
  3. Emergence of Chaos: Perhaps most strikingly, the introduction of spatial and temporal concentration variations can fundamentally alter the qualitative behavior of the system. Baker et al. discovered that while homogeneous gas flows behave predictably and monotonically, varying hydrogen concentrations can trigger transitions into chaotic regimes. These states are highly sensitive to initial conditions, making long-term operational forecasting exceptionally difficult without exact mixture modeling.

Tracking Composition: The Nodal Mixing Approach

To model a network where composition varies at every point, researchers must track how gases mix at junctions where different pipelines converge.

Inflow A (Pure Methane) --------\
                                 +---> Junction (Mixed Composition) ---> Outflow to Network
Inflow B (Hydrogen Blend) ------/

A key mechanism for solving this is the nodal mixing approach. As formulated by Kazi et al., this method tracks gas composition by defining the concentration inside any given pipeline based on the properties of its immediate upstream junction. By applying conservation of mass and species at each node, the framework determines the exact mixture ratio entering downstream pipes.

To make this computationally tractable for large-scale optimization, the authors developed non-smooth reformulations that can efficiently simulate these heterogeneous mixtures across complex topologies without succumbing to the combinatorial explosion typically associated with tracking multi-component flows.

Managing Transient and Dynamic Flows

In a real-world grid, gas injections and withdrawals are rarely static; solar-powered hydrogen production fluctuates throughout the day, and consumer demand spikes unpredictably. This creates transient flows where concentration waves travel through the network over time.

Managing these fluctuations requires dynamic optimization frameworks. Kazi et al. transformed the complex partial differential equations governing transient fluid dynamics into solvable optimization problems, allowing operators to maximize economic value while keeping pressures within safe physical limits.

Without these composition-aware transient models, control systems are blind to the delayed impacts of upstream injections. As Baker et al. argued, explicit modeling of time-varying mixing dynamics is essential for safe compressor operations and reliable gas delivery, as traditional simplified models fail to capture the lagging pressure waves caused by moving concentration fronts.

The Economics of Localized Physics

The physical reality of spatial composition variability directly impacts the economics of energy markets. Because hydrogen has a lower volumetric energy density than natural gas, a cubic meter of gas at a node with high hydrogen concentration contains less energy than a cubic meter of pure natural gas. Furthermore, the carbon intensity of the gas varies depending on the local blend ratio.

To address this, Sodwatana et al. extended physics-informed modeling into market design through Locational Marginal Pricing (LMP). Traditional energy markets assume a homogeneous commodity, but by analyzing the dual solutions of a mathematical program that embeds the network's physical mixing equations, the authors derived a localized "decarbonization premium." This framework allows market administrators to equitably distribute carbon offset incentives and price the delivered gas based on its exact, locally tracked physical and environmental characteristics rather than a network-wide average.

Key Foundational Work

Transitions from Monotonicity to Chaos in Gas Mixture Dynamics in Pipeline Networks (2023) — contributed 7 concept elements to this direction.
Optimal Control of Transient Flows in Pipeline Networks with Heterogeneous Mixtures of Hydrogen and Natural Gas (2023) — contributed 6 concept elements to this direction.
Optimization of Hydrogen Blending in Natural Gas Networks for Carbon Emissions Reduction (2022) — contributed 2 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_spatial_composition_variability_fa9bf8d3,
  title = {spatial composition variability},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/spatial-composition-variability/rdr_fa9bf8d39d204f26bb1036235d4ab593},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_spatial_composition_variability_fa9bf8d3_v1,
  title = {spatial composition variability},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/spatial-composition-variability/rdr_fa9bf8d39d204f26bb1036235d4ab593/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake