Development of cost reduction mathematical model for natural gas transmission network system
Natural Gas Transmission Systems (NGTSs) have been designed to transmit a huge amount of Natural Gas (NG) at high pressure from the refineries to Natural Gas Distribution Systems (NGDSs) and some of consumers such as power plants and exportations. In these systems, NG is sent to sales point through...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2012
|
| Online Access: | http://psasir.upm.edu.my/id/eprint/34037/1/FK%202012%207R.pdf http://psasir.upm.edu.my/id/eprint/34037/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Natural Gas Transmission Systems (NGTSs) have been designed to transmit a huge amount of Natural Gas (NG) at high pressure from the refineries to Natural Gas Distribution Systems (NGDSs) and some of consumers such as power plants and exportations. In these systems, NG is sent to sales point through pipelines, and Compressor Stations (CSs).
The main problem in the NGTSs was a lack of researches that minimize the total cost of the natural gas transmission network, which is included most transmission pipeline
and compressor stations parameters simultaneously. The next problem was referred to solve NGTSs models. Because developed mathematical models for optimization NGTS problems have been known as a complex problem especially in the real-world cases.
In this research, the objective was to develop a new mathematical model to minimize the costs of the Natural Gas Transmission Supply Chain (NGTSC) system by considering various design factors such as diameter, thickness, pressure, temperature, length, and compressor ratio. To achieve this, other objectives were targeted such as,
to design a multi-echelon supply chain for the NGTSCs, to develop Extended Genetic Algorithms (EGAs), for solving the model, and to validate the mathematical model with a real world case study in natural gas industry.
This research separated Natural Gas Gathering Systems (NGGSs), which was relating to the collected natural gas from the wells and transmit it to the refineries from the
NGTSs. Therefore, a new classification proposed on the natural gas networks from the oil and gas wells to the consumers. In addition, a multi-echelon supply chain was
designed for NGTSs named Natural Gas Transmission Network Supply Chain (NGTSC). It was used for integrating the operational parts of the natural gas industry for transmitting gas through pipelines and better coordination for the flow of gas and information in the system.
A new Mixed Integer Non-Linear Programming (MINLP) mathematical model was developed for the optimization of the NGTSC that named NGTSCM. The model formulated by using gas hydraulic equations such as Weymouth, compressor
horsepower, Reynolds number and gas velocity equations. Some factors, such as the gas pipeline diameters and compressor power, had a greater impact on network costs, and therefore, there was a need to reduce them. To achieve this, the problem was presented as a formula.
Since the optimization problem in NGTSs was non terministic polynomial and this complexity would be increased for the development of the NGTSC in the proposed MINLP model, exact methods had to be replaced with metaheuristic methods for
finding the global optimum solution. In this research, an Extended Genetic Algorithms (EGA) was investigated to solve the proposed mathematical model and to achieve global or near to global optimum solutions in a reasonable time.
To verify the model, a natural gas transmission example from literature was considered, and to present the model validity one real case study was studied. The results obtained from problem were analyzed based on the objective function, and the design parameters of the network. Analysis of results illustrated the priority of the
NGTSCM compared to the other design methods. Through one to one comparison of the costs of the networks, it was clear that, the costs, as calculated using the optimal method, were reduced by 2.91 % in first case, and 0.94 % in second case in comparison with another method. By using this model, the compressors' power or ratios were decreased and the pressure and distance between compressors were
increased. Thus, the total cost of the network was decreased. Therefore, the data clearly exhibit that the proposed method provides a solution that was nearer to an
optimized network. |
|---|