(690e) Modeling and Optimization of Mixed Hydrogen-Natural Gas Flow in Pipeline Network

The decarbonization of the energy grid is a necessary step to reduce global greenhouse emission levels and limit the effects of climate change. One aspect of this decarbonization process includes the use of low or zero emission hydrogen fuel. At present, pure hydrogen cannot be transported using the current gas pipeline network and needs to be mixed with natural gas for transmission. Moreover, the optimal operation of gas pipeline network is a non-trivial problem which requires the modeling of pressure drop and flow equations inside the pipes and nodes of the network.

In this talk, we present a model for the mixing and transport of hydrogen-natural gas (H₂-NG) along a pipeline network. The mixed gas network model captures the mixing, pressure drop and flow equations inside each node and pipe of the network. The nonlinear and non-smooth equations of mixing are reformulated using linear polyhedral relaxations and integer variables or complementarity constraints. We discuss and prove the uniqueness of solution for the model for special cases using graph theory. We show multiple reformulations and compare the performance when solved using standard optimization solvers.

Subsequently, we discuss the dynamic model which is derived by lumping the full-scale Partial Differential Algebraic Equation (PDAE) model into a Differential Algebraic Equation (DAE) model. We present both dynamic simulation and optimization model for the H₂-NG mixed gas flow problem and solve literature test cases to show the effectiveness and need for optimization.