S-Space College of Engineering/Engineering Practice School (공과대학/대학원) Dept. of Chemical and Biological Engineering (화학생물공학부) Theses (Ph.D. / Sc.D._화학생물공학부)
Advanced Process Design and Operation in a Natural Gas Supply Chain using Superstructure Optimization and Multi-modular Approach : 천연가스 공급망 내 초구조 최적화 및 다중모듈방식을 이용한 공정설계 및 운전
- 공과대학 화학생물공학부
- Issue Date
- 서울대학교 대학원
- 학위논문 (박사)-- 서울대학교 대학원 : 공과대학 화학생물공학부, 2019. 2. 이원보.
- 본 논문은 공정시스템 분야의 최신기술 수요에 상응하는 최적 공정설계 및 운전기술 개발을 주목적으로 한다. 최근 셰일가스 등 변화하는 천연가스 자원으로부터 지속적인 부가가치 창출과 플랜트의 내재적 안전성을 제고할 수 있는 설계 및 운전을 도모하였다는 점에서 실제 산업에의 응용가치가 매우 높다.
첫 번째로 천연가스 가솔린회수 및 액화 통합공정에 질소회수공정을 추가하여, 저품질 천연가스로부터 지속적인 액화천연가스 생산이 가능한 공정을 설계하였다. 열교환망 및 분리공정 최적화를 위해 공정요소들의 엑서지를 최소화할 수 있는 초구조를 설계함으로써 기존의 연구가 찾지 못하였던 새로운 최적 구조 및 운전조건을 결정하였다. 나아가 서로 다른 천연가스 조성에 따라 각기 적용이 가능한 대안공정을 추가 설계·최적화함으로써 변화되는 천연가스 자원에 지속적인 가치창출을 위한 해답을 제시하고 있다.
두 번째로 공정의 예비설계단계에서 내재적 안전성의 개념을 도입하여, 경제성과 안전성의 균형을 유지하기 위한 새로운 다목적최적화 알고리즘을 개발하였다. 잠재적 위험도가 높은 천연가스 액화공정을 대상으로 액화사이클에 따른 초구조를 모사하여 두 가지 목적함수의 가중치에 따른 최적해를 결정함으로써 기존 최적화의 한계를 보완하였다.
마지막으로 플랜트 안전운전을 위해 공정이상에서부터 사고의 발생 및 전파과정을 실시간으로 구현할 수 있는 시뮬레이션 모듈을 개발하였다. 동적공정시뮬레이션 및 사고시뮬레이션의 두 가지 독립된 모듈을 객체연결매입 기법을 이용하여 연동함으로써 사고상황에서 운전원의 임의조치가 모듈에 실시간 반영되도록 하였다. 해당 모듈은 임의의 사고상황에서 제어실 및 현장 운전원의 적절한 대응을 효과적으로 유도할 수 있으며 나아가 플랜트 안전시스템설계에 객관화된 지표를 제시할 수 있었다.
본 논문은 위와 같이 실제 산업의 기술적 수요를 충족시키고 이를 발전시킴으로써 공정시스템 학술분야에 기여하였다.
Recently in the field of process systems engineering in natural gas processing, various researches trying to make changes in the existing framework of process design and operation have been studied with the emerging need of sustainability and safety in the chemical processes. These two considerations of sustainability and safety either result in a totally new solution for a certain decision making or require far different methods or technologies for it.
Especially for a natural gas supply chain broadly from drilling of the gas/oil reservoirs to distributing the product gas to end-users like households or offices, new frameworks of process design and operation critically influence the way of producing desired products and supplying them to the users in the associated industries. Then it determines the structure, operating conditions, and operation procedures of chemical processes which are economically powerful and good in operability. Recently, as the natural gas sources becomes unconventional varying from mid-to-small size reservoirs or shale gases, this change makes the offshore natural gas plants emerge as an alternative and vital site of producing LNG (liquefied natural gas) with strict requirements of safety. It also makes additional processing units like a cryogenic nitrogen recovery be necessary for sustainable production of LNG with leaner feed natural gases.
Among various processes in the overall natural gas supply chain, this thesis dealt with largely three parts including gas pre-treatment, liquefaction, and distribution to the end-users, attempting to design new processes or develop new methods of decision making in the context of the new framework considering sustainability and safety in process systems engineering. In this thesis, I will discuss the process synthesis, intensification, and optimization for sequential units, multi-objective optimization for economic feasibility and inherent safety, and multi-modular approach for interactive simulation of dynamic process and 3D-CFD (computational fluid dynamics) accident models.
First of all, for designing a sustainable process of producing LNG from feed natural gases with high amounts of nitrogen, two cryogenic nitrogen recovery processes integrated with LNG production and NGL (natural gas liquid) recovery were designed and optimized based on the structural analysis of components separation: one for integrated nitrogen recovery unit and the other for standalone one. The difference of each process is the way the nitrogen is removed from the natural gas. The former recovers nitrogen in the integrated heat and mass transfer structure with natural gas liquefaction while the latter separates the nitrogen recovery unit into an independent structure apart from the liquefaction section. These sophisticated nitrogen recovery solutions follow the recent demand of highly efficient electric motors as alternative compressor drivers which require less or no fuel gas, the major sink of nitrogen in the feed gas.
These two processes were compared with each other in terms of specific power (kWH/kg_LNG), which is equivalent to the overall process efficiency, with respect to the nitrogen content in the feed gas from 0mol% to 20mol%. Consequently, as the nitrogen content in the feed gas increases, the specific power of each process also increases while the standalone solution has a priority over the other until around 17mol% of nitrogen and after that point the integrated solution becomes relatively more efficient. It should be noted that all of the optimization results of each configuration were improved with the reduced specific power by up 38.6% compared to those from previous studies which have similar configurations. The way this study aimed could be reasonable guidelines for other chemical process designs as well as nitrogen recovery in natural gas processing.
Secondly, for designing a safer process of natural gas processing, two different systematic approaches were newly proposed in this study: one for risk reduction method based on rigorous QRA (quantitative risk assessment) results through process design modification of an existing plant which already finished up to the detailed design stage, and the other for deciding an optimal process design through multi-objective optimization for minimizing both the TAC (total annual cost) and the risk (fatality frequency) at the preliminary design stage. This latter approach could largely lower the cost required for finalizing the design as it doesnt need to follow the general QRA procedure where the recursive loop is recycled until the risk is reduced to an acceptable level. But before this approach starts to be applied, the suitability of its method should be verified as it has to make some assumptions in assessing the safety level of the process with limited information. Also the computation load would be higher as it needs to simultaneously consider the economic feasibility and inherent safety in designing a process. Despite the differences these two approaches have each other, however, they are essentially in the same context in that they share the same purpose of deciding a process design which is safer and/or even cheaper than the existing processes.
Consequently, for the former approach of which the target process is the GTU (gas treatment unit) of an existing GOSP (gas oil separation plant) for processing associated natural gas, the modified design with different operation conditions reduced the total risk integrals by 27% at the expense of only the additional $50,000 for capital cost. In addition, sensitivity analysis of total risk with respect to probability of success for safety barriers was carried out in order to show the preferences of process design modification, this study proposed, over the improvement of safety systems. Meanwhile, the latter approach of superstructure formulation and multi-objective optimization for designing an optimal heat transfer structure and operating conditions was applied to the natural gas liquefaction processes, deciding that the SMR (single-stage mixed refrigerant process) structure with the TAC of 626.6MM$/yr and fatality frequency of 1.28E-03/yr has the highest priority over all possible solutions.
Finally, with the aim of safely operating a chemical plant, a new operator training module which mainly targets the interactive cooperation of control room operators and field operators was developed through using multi-modular approach with advanced simulations and data processing technologies. This interactive simulation modeling delivers the online simulation results of process operation to the operators and induces them to take proper actions in case of a random accidental situation among pre-identified scenarios in a chemical plant. Developed model integrates the real-time process dynamic simulations with the off-line database of 3D-CFD accident simulation results in a designed interface using OLE (Object Linking and Embedding) technology so that it could convey the online information of the accident to trainees which is not available in existing operator training systems. The model encompasses the whole process of data transfer till the end of the training at which trainees complete an emergency shutdown system in a programmed model.
The developed module was applied to a natural gas pressure regulating station where the high pressure gas is depressurized and distributed to the end-users like households or offices. An overall scenario is simulated in the interactive simulation model, which starts from an abnormal increase of the discharge (2nd) pressure of the main valve due to its malfunction, spreads to an accidental gas release through the crack of a pressure recorder, and ends with gas dispersion and explosion. Then the magnitude of the accident outcomes with respect to the lead time of each trainees emergency response is analyzed. Consequently, the module could improve the effectiveness of operator training system through interactively linking the trainee actions with the model interface so that the associated accident situations would vary with respect to each trainees competence facing an accident.