Optimal Design of Natural Gas Gathering Systems
259
parameters that affect the investment and operating costs, this topic deserves special
attention.
Nowadays, phased development is widely applied in the gas field to enhance production efficiency. It gives priority to develop the blocks with better reserves and then
gradually explore other blocks. In the mid to late period of gas field development, the
pressure of the gas wells may drop and result in a decrease of production [2]. Accordingly, new blocks would be connected to the existing gas gathering system (GGS) for
stable production and new CSs and pipelines are needed to adjust the original network
system to minimize energy consumption.
1.2 Literature Review
Optimal pipeline networks design is a complex task involving pipeline layout, equipment
selection, hydraulics and reliability which makes it difficult to be solved. Most of the
previous studies, therefore, considers the simpler problem of component design and in
particular optimal pipe sizing problems [3–7]. However, the separate optimization of
pipeline topological structure and design parameters would lead to the local optimal
solution due to the discrete choices crossed with physical non-linear constraints.
To solve this problem, studies focus on the combination of those two problems that
have become increasingly popular. Wang et al. [8] developed a model of subsea wells
partition for the layout of pipelines and cluster manifolds. Then, the optimization of
the layout scenarios of cluster manifolds with pipeline end manifold is studied in Ref
[9]. Zhang et al. [10] established a MILP model with considering terrain and obstacle
conditions. The optimal parameters were obtained integrally by solving this model with
GUROBI solver. Zhou et al. [11, 12] decomposed the optimization problem into several
subproblems and applied a hierarchical-optimization strategy to determine the layout of
star-tree and tree-tree pipeline networks. Hong et al. [13] applied ant colony optimization algorithm for route optimization and a piecewise method was employed to linearize
the nonlinear hydraulic equations. Wang et al. [14] developed an mixed-integer linear
programming (MILP) model for designing GPNs to determine the optimal locations of
the central processing facility and manifolds, detailed topological structure, diameter
and route of each pipeline, pipeline flow and node pressure. Liu et al. [15] proposed a
combined optimization strategy for large-scale oil and GGS optimization. MPSO algorithm was applied to solve the problems. In the mid to late development stages of the
gas field, reconstruction of GPNs is needed for new blocks connection and satisfy the
development of low pressures and low production rates. Wang et al. [16] proposed a
MILP model for network reconstruction optimization design. In this problem, the variables to determine were which transferring stations should be abandoned and which
reserve stations should be linked to the testing stations. In the following studies, terrains
and obstacles were taken into consideration for multi-period natural gas transmission
network construction [17] and site optimization of gas gathering stations [18]. He et al.
[2] established a model aiming at optimization of improved pipeline network layout
considering the newly developed blocks. Genetic simulated annealing hybrid algorithm
was applied to determine the design parameters.
In previous researches, various approaches were proposed for the pipeline design
optimization problem. However, few articles have focused on the renewal and reform of
259
parameters that affect the investment and operating costs, this topic deserves special
attention.
Nowadays, phased development is widely applied in the gas field to enhance production efficiency. It gives priority to develop the blocks with better reserves and then
gradually explore other blocks. In the mid to late period of gas field development, the
pressure of the gas wells may drop and result in a decrease of production [2]. Accordingly, new blocks would be connected to the existing gas gathering system (GGS) for
stable production and new CSs and pipelines are needed to adjust the original network
system to minimize energy consumption.
1.2 Literature Review
Optimal pipeline networks design is a complex task involving pipeline layout, equipment
selection, hydraulics and reliability which makes it difficult to be solved. Most of the
previous studies, therefore, considers the simpler problem of component design and in
particular optimal pipe sizing problems [3–7]. However, the separate optimization of
pipeline topological structure and design parameters would lead to the local optimal
solution due to the discrete choices crossed with physical non-linear constraints.
To solve this problem, studies focus on the combination of those two problems that
have become increasingly popular. Wang et al. [8] developed a model of subsea wells
partition for the layout of pipelines and cluster manifolds. Then, the optimization of
the layout scenarios of cluster manifolds with pipeline end manifold is studied in Ref
[9]. Zhang et al. [10] established a MILP model with considering terrain and obstacle
conditions. The optimal parameters were obtained integrally by solving this model with
GUROBI solver. Zhou et al. [11, 12] decomposed the optimization problem into several
subproblems and applied a hierarchical-optimization strategy to determine the layout of
star-tree and tree-tree pipeline networks. Hong et al. [13] applied ant colony optimization algorithm for route optimization and a piecewise method was employed to linearize
the nonlinear hydraulic equations. Wang et al. [14] developed an mixed-integer linear
programming (MILP) model for designing GPNs to determine the optimal locations of
the central processing facility and manifolds, detailed topological structure, diameter
and route of each pipeline, pipeline flow and node pressure. Liu et al. [15] proposed a
combined optimization strategy for large-scale oil and GGS optimization. MPSO algorithm was applied to solve the problems. In the mid to late development stages of the
gas field, reconstruction of GPNs is needed for new blocks connection and satisfy the
development of low pressures and low production rates. Wang et al. [16] proposed a
MILP model for network reconstruction optimization design. In this problem, the variables to determine were which transferring stations should be abandoned and which
reserve stations should be linked to the testing stations. In the following studies, terrains
and obstacles were taken into consideration for multi-period natural gas transmission
network construction [17] and site optimization of gas gathering stations [18]. He et al.
[2] established a model aiming at optimization of improved pipeline network layout
considering the newly developed blocks. Genetic simulated annealing hybrid algorithm
was applied to determine the design parameters.
In previous researches, various approaches were proposed for the pipeline design
optimization problem. However, few articles have focused on the renewal and reform of