Towards improved and multi-scale liquefied natural gas supply chains: Thermodynamic analysis
Graphical abstract
Introduction
Driven by the continuous population growth, economic development, and other factors, the global energy demand is projected to grow by about 20% by 2040, with fossil fuels being the primary energy source (ExxonMobil, 2019). Now, given the growing global warming concerns and the difficulties associated with decarbonizing the energy sector, it is deemed necessary to foster more deployment of low carbon and sulfur energy sources such as natural gas (NG). In addition to its relatively clean-burning characteristics, NG adaptability and low price have increased its contribution to the world energy supply (Al-musleh et al., 2015b). While excessive use of NG over a prolonged period can adversely impact the climate, NG's high energy content per mole carbon makes such a fossil fuel attractive to phase in renewable carbon-free energy sources (Mason, 2019). NG can also complement renewable energy by addressing peak demands and renewables' intermittent nature. Methods of supplying NG to end-users include pipelines and liquefaction. While pipelines are simpler and more economical for short distances, liquefied NG (LNG) is more suitable for transporting terawatt-levels of energy over long distances. According to recent forecasts, LNG trade will account for approximately 20% of the globally traded NG in 2040, with an annual growth rate of 3.6% (ExxonMobil, 2019). This may indicate the need for expanding current LNG supply chain capacities, which could give rise to several concerns.
As shown in Fig. 1, a baseload LNG supply chain consists of four successive systems; each section is made of sophisticated and integrated subsystems that are highly interdependent. While the chain supplies end-users with the most environmentally friendly fossil fuel, crude NG transformation across the chain's subsystems requires a substantial amount of fuel, which gives rise to considerable amounts of harmful emissions (carbon dioxide, NOx, SOx, and hydrocarbons). For example, our previous analysis showed that carbon dioxide emissions from a baseload chain are higher than those emitted from an advanced oil refinery processing crude oil with a similar rate of energy input (Katebah et al., 2020). Excessive fuel usage in the chain's subsystems also limits its energy delivery per NG feed rate. Moreover, LNG supply chains discharge considerable amounts of liquid and solid wastes. Minimizing such wastes, whether by recycling or other techniques, requires enormous amounts of energy. These issues, along with economic considerations and heightened environmental/emissions regulations, make design and operational optimization crucial tasks.
Optimization of LNG processes is widely discussed in the open literature. A comprehensive literature review regarding the use of optimization techniques for LNG process design and operation was conducted by Austbø et al., 2014 (Austbø et al., 2014). Significant efforts have been dedicated to energy efficiency enhancement using deterministic and stochastic optimization algorithms (Xin-She, 2010). In the majority of the published work, the optimization of the LNG process is performed either using a local search optimization algorithm (such as sequential quadratic programming, SQP) for rigorous process models or advanced global methods for simplified process models. For instance, Aspelund et al. (2010), developed an optimization-simulation tool, where both stochastic and local search methods are combined, for Aspen Hysys® simulated processes. In this work, the method was applied to minimize the energy requirement of the Prico liquefaction cycle. Alabdukkarem et al.(2011) used a genetic algorithm (GA) from Matlab® optimization toolbox to optimize an LNG plant modeled in Aspen Hysys®. The connection of the Matlab® optimization toolbox with Hysys® has saved the efforts and time to build the mathematical model in Matlab® (or other coding platforms) (Alabdulkarem et al., 2011). Wang et al. (2005) proposed a synthesis framework for screening a low-temperature heat-integrated separation system. The authors illustrated their approach by examining several separation options for an LNG plant and carried out the optimization using generic algorithms (Wang and Smith, 2005). In another study, a process software (LNG-Pro) was developed for the optimal (minimum operating and capital cost) synthesis of refrigeration cycles (Kamalinejad et al., 2014). Furthermore, optimization techniques based on thermodynamic consideration and process knowledge have been widely used to improve the energy efficiency of LNG plants (see for example the work of Wang et al., al.,2011)
As shown above, considerable efforts have been dedicated to developing optimization methods suitable for energy systems. Many of these require high-fidelity mathematical models (Austbø et al., 2014; Schulz et al., 2005). However, rigorous modeling and optimization of highly integrated systems, such as a baseload LNG chain, is a non-straightforward and complicated task. The complexity of the current technologies making up the chain and the high number of variables available for optimization can make the model's convergence very challenging. This is compounded due to the issues pertaining to global optima search. Considering these shortcomings, one may be tempted to adopt less predictive techniques to simplify the modeling and optimization. These include empirical models, linear programming, or surrogate models. Alternatively, and more favorably, attention could be focused on optimizing and analyzing critical subsystems that can result in substantial impacts on the overall performance of the system. Concentrating on central units makes the utilization of rigorous models more efficient as convergence and global optimum issues, being proportional to the model size, are alleviated. Thermodynamic-based approaches, such as energy, exergy, and pinch analysis, could be used as effective pinpointing tools
Energy, exergy, and pinch analysis have been widely applied in the literature to assess and improve various components of the LNG supply chains. For example, Pellegrini et.al. conducted energy and exergy analysis to evaluate and compare the performance of different NG purification (also known as sweetening) and liquefaction technologies (Pellegrini et al., 2019). Gas sweetening exergy destruction was also investigated by Banat et.al (Banat et al., 2014). The sulfur recovery unit (SRU), which is needed to process the removed acid gases in the sweetening section, was evaluated by Samimi et al. (Samimi et al., 2014). On the other hand, Ghorbani et al. employed exergy analysis to evaluate an NG liquids (NGLs) recovery process and its integration with the NG liquefaction system (Ghorbani et al., 2018). Replacing the mechanical pre-cooling stage of the refrigeration system with an absorption process was also investigated. Aiming to identify opportunities to enhance process efficiencies, Lee et al. performed detailed exergy analysis for cryogenic energy storage systems combined with an LNG regasification process (Lee et al., 2017a, Lee et al., 2017b). Morosuk et al. used exergy to evaluate the single mixed refrigerant (MR) liquefaction process, including its compressor gas turbine (GT) driver and highlighted improvement opportunities (Morosuk et al., 2015). Kanoğlu performed an exergy analysis on the multistage cascade refrigeration cycle (pure refrigerants) to identify exergy destruction and exergetic efficiency (Kanoğlu, 2002). Tirandazi et al. assessed the effect of a multistage refrigeration cycle in an NGL recovery plant to determine the components causing major exergy destruction (Tirandazi et al., 2011). Palizdar et al. also conducted exergy analysis on three mini-scale nitrogen expander refrigeration cycles for NG liquefaction (Palizdar et al., 2017). Tesch et al. applied the same analysis to evaluate an LNG regasification process integrated with an air separation system (Tesch et al., 2016). The following conventional liquefaction processes were assessed by Vatani et al. using exergy and energy analysis: 1) Linde single MR, 2) Air Products and Chemicals (APCI) single MR, 3) Linde propane MR (C3MR), 4) APCI dual MR, 5) Linde Mixed fluid cascade. A simple GT cycle, conventional compressors drivers in LNG plants, was analyzed by Morosuk and Tsatsaronis, also using exergy (Morosuk and Tsatsaronis, 2009). Khan et al. performed energy and exergy analysis to enhance the efficiency of a three-stage propane pre-cooling cycle, where the effect of changing the evaporators’ operating conditions was analyzed. In their work, exergy losses and efficiencies were reported for several cases (Khan et al., 2016). Zargarzadeh et al. developed a tool, referred to as Olexan, using Visual Basic Application (VBA) to enable dynamic and online exergy analysis in an interactive manner at various levels of the equipment and plant (Zargarzadeh et al., 2017).
Different pinch and exergy tools were also developed for assessments and optimization. For example, Al-Sobhi et al.,utilized pinch techniques to evaluate and optimize an LNG process. In his study, different useful cogeneration options were proposed and evaluated. However, the study did not include the SRU and rigorous liquefaction cycles. The author found that through heat integration, heating, and cooling utilities could be reduced by 15 and 29%, respectively (Al-Sobhi, 2007). Aspelund et al. analyzed the design of sub-ambient processes using the ExPAnd tool, a new tool based on pinch and exergy principles. The tool can also be utilized to reduce NG liquefaction compression power. According to the authors, using the ExPAnd tool resulted in a 36% exergy efficiency improvement for a novel LNG process (Aspelund and Gundersen, 2007). Malham et al. also introduced a hybrid pinch-exergy tool to couple both analyses and overcome their individual limitations. The proposed methodology was applied to a basic LNG process (C3MR) to assess exergy losses and determine needed changes in the process conditions and structure (Malham et al., 2018). On the other hand, the hybrid approach proposed by Gourmelon et al., referred to as PiXAR, can be used for the analysis and retrofitting of existing industrial processes (Gourmelon et al., 2016). Last but not least, Wechsung et al. and many others developed an optimization formulation model that combines pinch analysis, exergy analysis, and mathematical programming. Their model was used for the synthesis of heat exchanger networks and can be useful for NG liquefaction cycles (Wechsung et al., 2011).
However, such analyses were targeted for specific units within the LNG chain, with a lot of emphasis on the liquefaction process. To the best of our knowledge, there is no rigorous and comprehensive study assessing a fully integrated baseload LNG chain using exergy and pinch techniques. A holistic and thorough analysis is essential to avoid missing non-obvious opportunities. For example, the sulfur recovery unit (SRU), an integral component of every baseload chain, is highly exothermic, and neglecting it would mean ignoring a vital heat source for the chain's operation. The fuel balance in the chain is another critical component that is usually not rigorously considered in the literature. Proper fuel rate and Wobbe index balance would require full chain consideration, as most of the fuel is generated at the tail of the LNG process train and importing terminal (see Fig. 1). We also believe that the utility subsystems (especially for the LNG train) are usually ignored or not adequately assessed. Such vital sections should be rigorously evaluated in a non-standalone aspect. Additionally, their performance on the different equipment they serve should be accurately represented and analyzed to extract valuable engineering insights pertaining to the processes’ inefficiencies.
Section snippets
Problem statement
The problem we are addressing in this paper pertains to “the where and the magnitude of losses” across the LNG value chain. Exploiting these to generate enhancement options is also an integral part of our approach in tackling the above problem. By doing this, we are not only addressing the above-mentioned literature gaps, but we are also reporting rigorous exergy and pinch analysis results for an actual chain that we have previously modeled and simulated. (Katebah et al., 2020). Some of the
LNG chain description
Fig. 1 illustrates the LNG supply chain considered in this work with detailed flowsheets shown in Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6. The chain has an LNG capacity of 3.35 million tones per annum (MTA) at the end-user storage tank during the prevalent import terminal holding mode. This is equivalent to a crude NG feed rate of 522 million standard cubic feet per day (MMSCFD at 0 °C and 1 atm). Referring to Fig. 1, the chain consists of the following interdependent segments: LNG process
Approach
As pointed above, thermodynamic-based approaches, such as energy, exergy, and pinch analysis, could be used as effective pinpointing and process options generation tools. In a typical energy analysis (also known as first law analysis), subsystems’ efficiencies are identified based on feeds, fuel, and products’ lower heating values (LHVs) (Katebah et al., 2020). On the other hand, employing exergy analysis allows the efficiencies to be determined while considering the streams’ maximum potential
Overall
The chain under consideration was modeled and simulated to process near 522 MMSCFD of natural gas feed (7332 MW LHV). Our modeling/simulation results were obtained using a methodical approach and actual data supplied by an operator in Qatar. In essence, our approach capitalized on the concept of paring manipulated variables with the process constraints. In the cases where the number of variables are more than the constraints, we would vary them, using a parametric analysis approach, to optimize
Conclusions
In this work, rigorous exergy and pinch analysis were carried out to analyze an actual LNG baseload chain processing ~522 MMSCFD of NG feed. Results were used to identify and recommend optimization opportunities within the chain for further investigations. Our exergy study showed that the main contributor to the power loss (from the NG feed) is the utility system (mainly GTs) serving the LNG process train, followed by the sulfur recovery, liquefaction, gas sweetening, and regasification
CRediT authorship contribution statement
Zineb Bouabidi: Methodology, Software, Validation, Formal analysis, Data curtion, Investigation, Writing – original draft, Writing – review & editing. Mary A. Katebah: Visualization. Mohamed M. Hussein: . Abdur Rahman Shazed: . Easa I. Al-musleh: Conceptualization, Resources, Supervision, Project administration, Funding acquisition.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgment
This paper was made possible by NPRP grant No. NPRP8-964-2-408 from the Qatar National Research Fund (a member of Qatar Foundation). The statements made herein are solely the responsibility of the authors.
References (56)
- et al.
Efficient electrochemical refrigeration power plant using natural gas with ~100% CO2 capture
J. Power Sources
(2015) - et al.
Efficient electrochemical refrigeration power plant using natural gas with~ 100% CO 2 capture
J. Power Sources
(2015) - et al.
Optimization of propane pre-cooled mixed refrigerant LNG plant
Appl. Therm. Eng.
(2011) - et al.
A new process synthesis methodology utilizing pressure exergy in subambient processes
Comput. Aided Chem. Eng.
(2007) - et al.
A liquefied energy chain for transport and utilization of natural gas for power production with CO2 capture and storage–Part 2: the offshore and the onshore processes
Appl. Energy
(2009) - et al.
An optimization-simulation model for a simple LNG process
Comput. Chem. Eng.
(2010) - et al.
Annotated bibliography—Use of optimization in LNG process design and operation
Comput. Chem. Eng.
(2014) - et al.
Energy and exergical dissection of a natural gas sweetening plant using methyldiethanol amine (MDEA) solution
J. Nat. Gas Sci. Eng.
(2014) - et al.
A novel energy efficient LNG/NGL recovery process using absorption and mixed refrigerant refrigeration cycles–Economic and exergy analyses
Appl. Therm. Eng.
(2018) - et al.
PiXAR: pinch and eXergy for the Analysis and Retrofit design of industrial processes
(2016)