Modeling gas flow directions as state variables: Does it provide more flexibility to power systems?

Gas Flow Models and Computationally Efficient Methods for Energy Network Optimization

“…Furthermore, temporal aggregation techniques can reduce problem size without sacrificing essential details. Future work can explore replacing the linear static model or SOC dynamic model with a linear relaxation for enhanced accuracy or computational efficiency.…”

“…Indeed, fixing the integer variables predetermines flow directions in dynamic flow optimization; however, it does not mandate unidirectional flows. Indeed, the flow directions may change from one time step to another to ensure the economic gains of bidirectional flows. , Notably, other researchers have also adopted the concept of fixing flow directions, but via other simplified optimization forms. ,,,, Furthermore, we extend this concept to fix flow directions in compressors and regulating valves.…”

“…A major challenge in these optimization problems is the nonconvex nature of the equations governing the gas flow dynamics in network pipes. Thus, techniques such as semidefinite relaxation, , second-order cone (SOC) relaxation, ,, and linear approximation − have been deployed to obtain computationally tractable dynamic flow models for operations optimization on daily horizons.…”

Optimal coordinated operation of integrated natural gas and electric power systems: A review of modeling and solution methods

“…Exact Relaxation Linear Conic Dynamic (DY) [4], [15], [16] --Quasi-dynamic (QD) [17] [18]- [20] [21]- [23] Steady-state (ST) [24] [25] [26], [27] generally referred to as linepack [9], [10]. By efficiently utilizing linepack, the gas network can provide flexibility to the power system by enabling rapid changes in the power generation of GFPPs at a low additional gas supply cost.…”

“…Both QD and ST models have been extensively adopted [17]- [27]. They are usually solved either with an outer linear [18]- [20], [25] or conic relaxation [21]- [23], [26], [27]. Both solution choices have been proven to be insufficiently tight under many operating conditions, resulting in physically infeasible solutions [21], [23], [26], [28].…”