Retrofit design of an industrial natural gas liquids recovery process based on the pinch technology concept

1. Introduction

Since 2000, the humans' need for energy has increased by more than 40% however, the number of natural energy resources is decreasing (Dong, Xu, Li, Quan, & Wen, 2018). So the importance of energy-saving in the last two decades has increased due to the increase in the cost of fuels(Fard, Pourfayaz, Kasaeian, & Mehrpooya, 2017; Jahromi & Beheshti, 2017). Especially for high-energy consumption such as oil, gas, and NGL refineries. On the other hand, governmental and environmental regulations are forcing the oil and gas industries to reduce the effects of CO₂ emissions by fossil fuels. According to the BP Statistical Review of World Energy, the US total proved reserves of natural gas were equal to 4.5 trillion cubic meters in 1997, which has almost doubled in 2016(Dudley, 2018). Also, Iran, with the fourth rank among the natural gas consumers has a promising future regarding the production of natural gas and NGL recovery industries(Dudley, 2018).

Pinch analysis as a well-established and successful technique is used to improve the energy efficiency of the industries such as oil and gas refineries (Fard et al., 2017). It is an optimizing methodology based on the thermodynamic principles that reduce the external utility load and maximize the heat recovery(Mehdizadeh-Fard, Pourfayaz, Mehrpooya, & Kasaeian, 2018). Linnhoff introduced this method in 1979(Linnhoff & Hindmarsh, 1983). Gai et al.,(Gai, Varbanov, Walmsley, & Klemeš, 2020) considered the Pinch analysis for different types of heat pumps with various processes. They presented a new plan by recovering and upgrading the waste heat of the process after reducing the utility consumption. Gaikwad et al.,(Gaikwad & Ghosh, 2020) used pinch technology for sizing the energy storage system and the fuel cell for fuel cell-based electric vehicles. Lebedev et al., (Lebedev & Yushkova, 2020)presented a mathematical model for optimization by exergy pinch analysis of a heat exchanger system in a Russian Refinery plant. Zhou et al., (Zhou et al., 2020) studied a novel pinch-based mathematical model developed for process integration and optimization of the Kalina Cycle.

The thermo-economic technique has been used for evaluating chemical and oil and gas processes. Jafari et al (Jafari, Ghasemzadeh, Yusefi Amiri, & Basile, 2019) conducted an economic evaluation of membrane technologies in comparison absorption process for CO₂ capturing from specified flue gas by Aspen Process Economic Analyzer v10. Rahimi et al, (Rahimi & Alibabaee) considered Iranian co-production of power and desalinated water. The economic analysis shows that their scenario with inlet air cooling has the highest total annual profit. Mousavi et al(Mousavi, Lari, Salehi, & Torabi Azad), considered the economic indicators in expansion turbines for the gas field of National Iranian Oil Company. Their new scenario economic consideration resulted in an internal rate of return (IRR) of 74.53% and a payback period (PBP) of 1.3 years. Asani et al(Asani, Mukherjee, & El-Halwagi, 2020) considered the economic scenarios in NGL plant using pinch and sensitivity analyses.

Pinch analysis cannot reduce the utility beyond a specific amount due to the limitations on ΔTmin and technical and economic aspects. Also, this analysis only uses the existing process streams and is not able to identify and develop high potential streams reducing by the external utility consumption. Therefore, self-refrigeration methods can be integrated with pinch analysis to reduce the limitations of this method. Opportunities for saving the energy consumption would increase by this method, and utility consumption would decrease. However, process configuration and product composition are essential parameters that should not be changed by the self-refrigeration method. Numerous researches have been done on natural gas refineries to improve energy efficiency. Tianbiao et al.,(He & Ju, 2014) considered pinch analysis on a mixed refrigerant cycle (MRC) integrated with the NGL recovery process for the small-scale LNG plant. Fard et al.,(Fard et al., 2017) represented the pinch analysis and the trade-off between capital cost and energy (super targeting) method in a natural gas refinery and improved the old network by 73.54%. Farahani et al.,(Farahani, Aghili, & Aghaei, 2015) redesigned the existing HEN of the Iranian NGL recovery plant in the Sirri Island. Their results showed that this analysis could save about 7% of the steam consumed by the plant.

Energy consumption in each country is among the critical indicators of economic development, but energy efficiency is a factor that is more important than the energy consumption influencing the country's economic growth(Mehrpooya, Ghorbani, & Hosseini, 2018). In this regard, many studies have been performed on industrial plants including NGL plants. Tahmasebi et al.,(Tahmasebi et al., 2015) considered the effect of different compositions of the feed on the energy and exergy of the NGL recovery plant. Ghorbani et al considered the combined pinch and exergy analysis for the refrigeration cycle in the NGL plant(Ghorbani, Salehi, Ghaemmaleki, Amidpour, Hamedi, et al., 2012). In a more detailed study of NGL plants, Ghorbani et al. considered the effect of energy quality on the economy(Ghorbani, Mehrpooya, Hamedi, & Amidpour, 2017). Also due to economic studies, Ghorbani et al considered the impact of electricity and NGL prices on the payback period of return for mixed fluid cascade LNG/NGL plant (Ghorbani, Shirmohammadi, & Mehrpooya, 2018). In other more detailed reviews of energy quality, Ghorbani et al studied an integrated cryogenic natural gas process with the aid of advanced exergy and exergoeconomic analyses(Ghorbani, Roshani, Mehrpooya, Shirmohammadi, & Razmjoo, 2020).

Several self

Refrigeration NGL recovery processes have been proposed with some limitations on operational flexibility and overall recovery performance that have had different options including JT valve expansion (Joule- Thompson), propane as an external cryogenic refrigerant, turbo-expansion, and side streaming from the column. Campbell (J. M. Campbell, 1979) introduced a procedure in the early 1970s called ISS, which uses turbo-expander design rather than the Joule–Thomson valves for cooling. The ISS scheme was one of the major advances in the gas processing industries. (Getu et al., 2015) The main problem of this scheme is freezing CO₂ during the process. In 1981, Campbell and Wilkinson(R. E. Campbell & Wilkinson, 1981) solved many problems of the ISS scheme by presenting the Gas Subcooled Process (GSP). They used a vapor stream from the top of the demethanizer column and refluxed it. The level of ethane recovery was improved by this scheme. Campbell et al., introduced the CRR (cold residue recycle) process scheme in 1989 (R. E. Campbell, Wilkinson, & Hudson, 1989).This process scheme was introduced to improve the efficiency of ethane recovery. In this way, the feed stream enters and refluxes different trays of the demethanizer column. Yao et al., and Lee et al., introduced the IPSI-1(enhanced NGL recovery process)and IPSI-2 (internal refrigeration for enhanced NGL recovery process)processes in 1999 (Yao, Chen, & Elliot, 1999) and 2007 (Lee, Zhang, Yao, Chen, & Elliot, 2007), respectively. The GSP, CRR, and ISS processes mainly focus on reflux streams of the top of the demethanizer column. However, two processes introduced by the IPSI company, the IPSI-1 and IPSI-2 focus on the bottom of the demethanizer column(Getu, Khan, Long, & Lee, 2012).

Inspired by the processes mentioned above, several new NGL recovery plants have been developed focusing on the improvement of energy efficiency and saving of the process. Mehrpooya et al. proposed the open-closed method in the NGL plant (Mehrpooya, Vatani, & Mousavian, 2010). High levels of hydrocarbon liquid recovery, high performance of the multi-stream heat exchangers, and low required compression power are the three main features of the proposed configuration. Moreover, Geta et al., (Getu et al., 2015) simulated and evaluated the economic performance of different types of NGL recovery processes (ISS, GSP, CRR, RSV, IPSI-1, and IPSI-2).

In this paper, the NGL plant No. 800 from National Iranian South Oil Company (NISOC) with  a production capacity of 120,000 NGL barrels per day located in Ahvaz, Koreit Industrial Zone is chosen as a real-life case study.

In this study, as a novelty, the self-refrigeration method is integrated with pinch analysis to reduce the limitations of the external utility saving. The grass root plant is revamped with side streaming from the demethanizer column, providing an additional refrigeration stream for inlet feed instead of external utility, and then is recycled back to the column. which in turn reduces more energy consumption by using new potential streams. The economic evaluation showed that the pinch and self- refrigeration integrated scheme is feasible and profitable.

2. Methodology

2.1 Process description

Fig. 1 illustrates the process flow diagram of the current NGL operating condition.

Fig 1. Process flow diagram of NGL plant.

In this plant, the feed stream enters the triple shell and tube, heat exchangers. The feed stream, heat exchanges with propane refrigeration cycle by the heat exchangers (E101 and E102), and with refluxed feed stream in heat exchanger E100. And finally is cooled up to -23 °C. Liquid hydrocarbons flow to the demethanizer column after these steps and passing through the separator (V-100) and the throttling valve (VLV-100). After the extraction, outlet gas from the top of the column is sent to the pressure boost units with the exchanged gas in the heat exchanger (E-100). Furthermore, the NGL is discharged from the bottom of the demethanizer column, enters into the pumping station (P-100), and after heat exchanging with the propane refrigeration cycle of the plant in the last heat exchanger (E-103), the final product is sent to the petrochemical companies at a temperature of 48 °C and a pressure of 63 psi for other uses.

In this plant, the external refrigeration cycle is used for the process and cooling of the final product of the NGL 800 plant. This cycle is completely separated from the production process of the NGL 800 plant. Propane is used as a refrigerant for this cycle, with its streams distinguished with the letter "P" in Fig. 1. This cycle consists of four main parts: a) the compressor system, b) the condenser, c) the economizer towers, and d) the heat exchanging system.

Propane passes through three compressors: low pressure (K-101), medium pressure (K-102), and high pressure (K-103), and propane pressure increases up to 23.84 bars in three stages. After that, pressurized propane enters the condenser (E-105) and is cooled down to 65.55 °C. This gas also enters the economizer towers in three steps (V102-V103-V104). The propane gas is returned to the compression system after separation and liquid propane continues for heat exchanging in the refrigeration cycle. This liquid heat exchanges with the inlet feed streams in the heat exchangers (E-101 and E-102) and outlet product stream in the heat exchanger (E-103). Also, the required heat for the reboiler of the demethanizer column is provided by the cooler (E-104).

2.2. Data Gathering

Firstly, it should be noted that the essential step in the pinch analysis is extracting proper data and understanding the process well (Keshavarzian, Verda, Colombo, & Razmjoo, 2015). After identifying all the primary process data of the plant including the temperature changes, the specific heat capacities, and the mass flow rates of the selected process, "Composite Curves(CC)" and "Grand Composite Curve (GCC)" can be drawn and they will determine the amount of energy consumed and the needed utilities in the process(Ghannadzadeh & Sadeqzadeh, 2017; Jahromi & Beheshti, 2017). Once the detailed data extraction was carried out, modeling and simulation of the NGL plant were done in the Aspen Hysys v7.2 simulator.

The data gathering was done in accordance with the Iranian Petroleum Standards (IPS-E- PR-170) ("ENGINEERING STANDARD FOR PROCESS FLOW DIAGRAM- IPS-E-PR-170," 1996). This standard specifies the minimum general and specific requirements for the contents of the process flow diagram which shall be used throughout the process simulations. This standard illustrates the minimum required information for simulating the equipment and streams in Aspen Hysys.

With obtaining this minimum operating information, simulation of the NGL plant was performed. Peng-Robinson Equation of State (EOS) was selected for determining the thermodynamic properties of the NGL plant. This state equation has been used in previous simulations of the NGL plants (AlNouss, Ibrahim, & Al-Sobhi, 2018; Ghorbani, Hamedi, Amidpour, & Shirmohammadi, 2017; Ghorbani, Salehi, Ghaemmaleki, Amidpour, & Hamedi, 2012; Long & Lee, 2013; Tahmasebi et al., 2015). The demethanizer column has ten trays with a 22 bar pressure and a variable temperature ranging from -24.6 to 37.9 °C. The refrigeration cycle for the shaft work of the three compressors consumes energy of 4,410 kW. Also, the required heat for the reboiler of the demethanizer column is equal to 4,260 kW provided by the heater E-104, and the refrigeration cycle condenser consumes energy of 9,950 kW for cooling the cycle by cooler E-105. Moreover, the cooling propane flow rate in the refrigeration cycle is equal to 33.64 kg/s.

Table 1 summarizes the thermodynamic process data obtained from the Aspen Hysys simulator. Also, Table 2 shows the energy consumption of the equipment. Table 3 shows a comparison regarding the simulated and operating conditions of the main feed and product streams of the NGL plant. The error rate in this simulation indicates that the simulation  complies with the operating conditions.

The following assumptions were used to simulate the NGL plant:

• Peng Robinson equation of state is applied, to predict governing equations and calculate the balances in the simulation are considered.
• The NGL simulation is performed in a steady state.
• Pipelines are assumed to be insulated in the HEN and do not have heat loss.

Table.1: Simulator output process data for current operating condition

Table 2: Energy consumption table of the plant equipment

Table 3: Comparison of simulation and operational conditions.

3. Pinch Analysis

The main goal of pinch analysis is minimizing the need for hot and cold utility by maximizing the internal heat recovery in the HEN. Graphical representations of pinch analysis, as well as Composite and Grand Composite Curves (GCC), are among the key tools used to determine the overall required to heat and cool in the process. (Raei, 2011)

The pinch point is referred to as the minimum vertical difference between hot and cold composite curves(Ghorbani, Ebrahimi, Rooholamini, & Ziabasharhagh, 2020). The exact location of the pinch point is shown in the grand composite curve, which is more evident than the composite curves (Kemp, 2005). Firstly, for adjusting the temperatures of the CC with those of the GCC, the separate cold composite temperature should be increased by ½ ΔTmin and the separate hot composite temperature should be decreased by ½ ΔTmin according to the Equations (1) and (2)(Stampfli, Atkins, Olsen, Walmsley, & Wellig, 2019). As a result of this temperature shift, the CCs touch one another at the pinch point. Figs. 2 and 3 show the CCs and GCCs in the current state of the plant.

Aspen Hysys simulation and its process data in according to table 1, is exported to Aspen Energy Analyzer and then, the CC and GCC are obtained. Fig. 2 shows the composite curves. The minimum cold and hot required utility are equal to 10,190 and 410 kW, respectively for ΔTmin of 10.645 °C. The results obtained from the composite curves indicated that the NGL plant requires a large amount of external utility.

The GCC diagram for this process is shown in Fig. 3. In this diagram, the y-axis is the shifted temperature according to the Eqs. 1 and 2. And the x-axis represents the amount of energy consumption. Above the pinch point, is the required hot utility and below the pinch, is the cold utility. Based on this GCC diagram, the pinch temperature is 40.79 C and the cold and hot utilities for the NGL plant are 410 and 10,190kW respectively. As can be seen from these graphical representations, the cold utility consumption (10,190 kW) is higher than the heating utility (410 kW).

Fig 2. Composite Curve (CC) chart for cold and hot streams in the current operating condition.

Fig 3. GCC (Grand Composite Curve) chart for cold and hot streams in the current operating condition.

3.1. Pinch Rules

In applying the pinch method design for HEN, it is essential that the correct stream matches with the pinch rules in order to achieve the minimum energy targeting. The main scope of pinch analysis is decomposing the HEN into above and below pinch points after identifying the pinch point, and considering pinch rules in two separated networks.

There are three main rules in the pinch analysis. These rules must comply to achieve the minimum energy target for the HEN.

1- Heat should not be transferred across the pinch points.

2-No external cooling above the pinch.

3-No external heating below the pinch.

Violating these three rules will result in increased energy requirements(March 1998).

There are also two sub-rules. For stream matching, the outgoing CP of the streams must be higher than incoming streams, and outgoing streams must be higher than incoming ones. (Liew, Alwi, Klemešb, Varbanov, & Manan, 2014; Polley, 1995)

Thus, the results of inequality in CP rule and stream number for two regions of above and below the pinch are as follows(Liew et al., 2014; Polley, 1995):

If these two rules are not satisfied with the streams, then stream splitting is required as shown in Fig. 4(Gundersen, 2013). Also, if needed, stream splitting is done for dividing an existing stream between two heat exchangers and using them more efficiently. Fig. 4 shows a brief description of the pinch rules procedure in the HEN.

Fig 4. A systematic approach for pinch rules(Ebrahimi, Ghorbani, & Ziabasharhagh, 2020).

Two approaches were developed based on these tools for the heat exchanger network(Ghorbani, Ebrahimi, Skandarzadeh, & Ziabasharhagh, 2020):

1) retrofit method.  2) grass-root design method.

The first method focuses on an existing design and modifying it to reduce the heat exchange across the pinch point. While the second method presents a new plant design involving the existing design constraints(Li & Chang, 2017).  The first method is more difficult because, in the retrofitting method, existing HEN, piping, equipment size, plant size, etc. limits the opportunities for efficient energy-saving(Nordman & Berntsson, 2009).

Grass-root design is based on specifying the ΔTmin range and considering the utility consumption rate and its capital cost. ΔTmin should be properly determined in order to achieve a cost-effective and optimized utility consumption, which is an essential parameter in designing a HEN. Utility consumption reduces for a smaller value of ΔTmin, while a small value of ΔTmin causes the excessive heat transfer area.

HEN optimization involves a relation between area and energy saving. A more detailed relation is given in Figure 13, which shows, the energy trade-off with the heat exchanging area of HEN of the NGL plant.

Table 4 shows the typical value for various types of processes according to the industrial-scale size of the chemical plants based on Linnhoff march's application experience. This can provide practical targeting for optimum ΔTmin.

Table 4.Typical ΔTmin values for several types of processes. These values are based on Linnhoff March's application experience(March 1998).

In the case of NGL plants using the refrigeration utilities with low-temperature ranges, lower ΔTmin values (3-5 °C) are used to minimize the expensive demand of the refrigeration cycle consequently leading to lower shaft work in the refrigeration cycle. Thus, in our case, the experimental ΔTmin value is considered as 3 °C for the NGL plant.

3.2. Retrofit Analysis

The current state of HEN is represented in Grid Diagram in Fig. 5. This diagram simply shows the HENs. It also shows thermal utilities and process heat recovery duties, target and supply temperatures (in °C) of streams, and pinch hot and cold temperatures. Moreover, it considers the main rules of pinch analysis.

As shown in Fig. 5, the cold and hot pinch temperatures of the current HEN are equal to 35.46 and 46.11 °C, respectively. There is no cross pinch stream. Also, there is no hot utility below the pinch and no cold utility above the pinch. All of the main pinch rules are met in the current state, and there is no need for any correction in retrofit Pinch Analysis.

Fig 5. The Grid Diagram of the current state of HEN.

3.3. Grass-Root Design

As shown in the current state of operating condition, the ΔTmin is 10.645 °C. Considering the ΔTmin of similar plants (about 3-5 °C), this high value of ΔTmin reflects the plant's significant potential for reducing the energy consumption by the grass-root method.

Targeting both capital cost and utility consumption is necessary and it is done over a range of ΔTmin values. As shown in Figs. 6 and 7, the range-targeting feature of Aspen Energy Analyzer gives a range target plot regarding hot and cold utility consumption and capital costs as a function of ΔTmin ranges. Figs.6 and 7 show the changes in HEN for the ΔTmin range from 1 to 12. For the smaller value of ΔTmin, heat recovery increases in the HEN, and utility consumption decreases but in smaller ΔTmin, more heat is transferred in the heat exchangers, resulting in a larger heat transfer area. Therefore, the trade-off between these two parameters is essential.

The range target plot can easily evaluate the optimal value of temperature difference for the case study. As shown in Figs. 6 and 7; there is no significant change in the total cost of HEN in the ΔTmin range from 4 to 12 °C. As a result, the intersection point of the utility targets and the total cost is considered as the optimal energy consumption point, which is equal to 2.3 °C. Moreover, the experimental aspect (based on the existing chemical industrial-scale plant) should take into account for choosing proper optimum ΔTmin. Therefore, ΔTmin changes from 2.3 °C to the nearest experimental value, which is equal to 3°C. These typical ΔTmin values have been used in much  scientific research .(Al-Mutairi, 2014; Brunner, Slawitsch, Giannakopoulou, & Schnitzer, 2008; Mondal, Uddin, Paul, Deb, & Azad, 2014; Owat & Skolpap, 2019; SinghaΨ & Crosbieb, 2011; Skolpap & Owat).

The process of the NGL recovery plant will change by applying the grass-root scheme and changing the ΔTmin to 3 °C (Fig.8). Moreover, due to this change, the consumption rate of hot and cold utilities changes to 354.44 and 10,156.40 kW, respectively. As shown in Fig. 8, one of the three compressors was removed due to the reduction in utilities, which reduced the required shaft work of refrigeration cycle compressors by 61.63%.

According to the grid diagram in Fig. 9, the arrangement of the heat exchangers changes, and their number increases by 6 in the HEN by changing the ΔTmin from 10.645 to 3 °C. Figs. 9 and 10 show the CCs and GCCs. Table 5 also shows the amount of energy saving. This optimization of energy consumption is only based on the current streams.

Fig 6. Range target plot of variation in the capital costs and hot utility consumption in different ΔTmin values showing optimum ΔTmin=2.2 °C (intersection point).

Fig 7. Range target plot of variation in the capital costs and cold utility consumption in different ΔTmin values, showing optimum ΔTmin=2.3 °C (intersection point).

Fig 8. Process flow diagram of the optimized NGL plant

Fig 9. Grid diagram of the optimized NGL plant

Fig 10. Composite curve of the optimized plant

Fig 11. Grand composite curve of the optimized plant

3.4. Self-Refrigeration and Pinch Methods

The increase in the price of energy and economic problems has caused the cryogenic cycle of the NGL plant to be more efficient and consume less amount of energy. (Mehrpooya, Vatani, & Moosavian, 2011) On the other hand, pinch analysis, as an energy-reducing technique has some limitations. It cannot reduce the utility beyond a specific amount due to the technical and economic constraints and limitations on ΔTmin. Although, it is a very valuable method considering the opportunities for more reduction of energy consumption. The self-refrigeration and pinch method introduced in this paper, as an improved method over the conventional process identifies and creates high potential streams that can further reduce the utility consumption. Also, the application of this method is accompanied by less maintenance and operating costs. (Harzaneh & Abbasgholi, 2011) One of the features of this method is that the products of the NGL plant should be unchanged similar to those of the current situation. For minimizing the process changes, the HEN should be optimized by pinch analysis and then, the opportunities for self-refrigeration of the process should be challenged. Demethanizer column of the NGL plant with ten trays is high potential equipment, which can generate the streams with low temperature and a sufficient flow rate. As shown in the process flow diagram in Fig. 12, and for achieving lower utility consumption with side streaming, a portion of cold hydrocarbon liquid (stream No. 31) from the top trays of demethanizer column is delivered to the heat exchanger E-103 and the heat exchanges with feed streams instead of required utility and then, is flashed back to the demethanizer column (stream No. 33). Therefore, the large amount of utility consumption in the refrigeration cycle can be effectively reduced by this technique without any limitation in product fields. Furthermore, it finally reduces the required shaft work. In this case, one of the other compressors will be removed from the optimized state resulting in the reduction of shaft work of compressors by 83.26% and 51.78% in total energy consumption compared to the current state. This is while the cooling propane flow rate reduces by 56.47 %, with a flow rate of 14.64 kg/s in the refrigeration cycle. Due to the plant's life (40 years), reducing the depreciated equipment, especially the refrigeration equipment, such as the economizer column is very important. Also, the depreciation of the refrigeration equipment, especially in the economizer column causes a leak and a decrease in the propane flow rate. Due to the high cost of propane, it is also very desirable to reduce the depreciated equipment. After revamping the design of the refrigeration cycle in the plant, one of the economizer columns and heat exchanger E-103 are removed from the new design to reduce the heat demand of the refrigeration cycle and propane flow rate. However, pinch rules should be considered again due to the addition of a new stream to HEN. Finally, after pinch analysis, the plant will process the NGL recovery with four heat exchangers (Fig.12).

Table 5 shows a comparison regarding the amount of reduced energy consumption in the current state, optimized state, and the state optimized with self-refrigerated stream analyses. Table 6 shows a comparison regarding the properties of the product optimized by grass-root and self-refrigeration method with the current state. As shown, the characteristics of the output product of the column did not change significantly.

Fig 12. Process flow diagram of self-refrigerated and optimized NGL plant

Table 5: Comparative table of energy consumption in different states

Table 6: Comparative table of output NGL product

As shown in Fig. 10 and Table 7, the potential of the demethanizer column is used for the replacement of the utility streams. After applying this issue and without increasing the number of heat exchangers, the shaft work of the compressors reduced by 83% while the refrigerant flow rate and the number of compressors decreased. Pinch and self-refrigeration scheme caused further heat integration in the system and decreased the required utility for the process, which in turn reduces the required heat load from the refrigeration cycle such that, the introduction of the new propane flow rate covers the network's need for the process. Also, the required energy for the condenser and the reboiler decreased by 34 and 63%, respectively due to the reduction in propane flow rate.

Compared with this study, a review of similar considerations shows the following result. Ghorbani et al applied a new scheme in a similar NGL plant. The shaft work of the compressor was reduced to 170 kW. The chosen ΔTmin was 4 °C(Ghorbani, Salehi, Ghaemmaleki, Amidpour, Hamedi, et al., 2012). In this study, the compressor shaft work of the compressor was reduced by 2717 and 3671 kW for grass root and integrated pinch and self-refrigeration schemes respectively by changing the ΔTmin to 3 °C. Tianbiao et al(He & Ju, 2014), reduced the refrigerant flow rate by 11.68% in pinch analysis of a similar NGL plant. In this study, the refrigerant flow rate was reduced by 65% in the integrated pinch and self-refrigeration method. Asani et al. evaluated the economic parameter of the pinch scheme for a NGL plant. Their scheme payback period was calculated by 536 days after startup (Asani et al., 2020). The economic evaluation of this study results in earlier payback (378 days after startup).

Figure 13 shows the area-energy plot for the current and optimized NGL plant modes. The area-energy plot is used to show a trade-off between the area added in the modified scheme design and the resulting energy savings. The X-axis in this figure represents the amount of energy consumption in two schemes. As the heat duty is reduced, the slope of the modified scheme curve increases. However, this reduction in energy consumption is due to the change of ΔTmin which has increased the heat exchange area. Naturally, this increment in the heat exchange area increases the investment costs. Which once again declares the importance of the optimal ΔTmin definition in energy consumption.

Fig 13. Area-Energy plot of NGL plant HEN.

4. Economic Evaluation

In the early stages of retrofit and grass-root design projects, reliable cost estimation is necessary for a proper decision(Shirazi, 1396). Economic estimation as a key element for preliminary and detailed design projects enables the designer to consider the economic options and make the right decision before performing any action(Ghorbani, Javadi, Zendehboudi, & Amidpour, 2020). The costs of investment and maintenance can be optimized by applying these cost-based decisions and changing some parameters in the design. (Caputo, Pelagagge, & Salini, 2016)

Every process of economic evaluation is done with respect to the basic economic indicators. These basic economic indicators are Total Capital Investment (TCI), Operating Costs (OC), and Payback Period (PBP). TCI is a function of some direct and indirect parameters associated with the NGL plant and includes the costs of purchasing the equipment, design, installation costs, etc. Table 7 represents the TCI parameters. This assessment is called as the Bare Module Cost Factor classified by Turton(Turton, Bailie, Whiting, & Shaeiwitz, 2008). In this paper, the allocated costs for utility plants and related facilities, cost of royalties, cost of land, and working capital are assumed to be zero.

Table 7: breakdown of total capital investment parameters(Kazemi & Samadi, 2016).

Where the T_CB, is the initial cost of the equipment. C_site is the site preparation costs. C_serv includes all service costs required until equipment start-up. These are fixed costs of equipment, which are called total direct permanent investment costs. C_royal, C_cont, C_land, and C_startup are the wage responsible for developing the process, contractor’s fee, the non-depreciation part of the costs, and the pre-commissioning and commissioning costs respectively. Total permanent investment is the sum of the pre-mentioned costs. And the C_WC is the finance for the first few months of operation before revenues from the process started. Finally, the total capital investment is the sum of working capital and the total permanent investment.

The basic cost of a heat exchanger depends on its heat exchanging area and is corrected by the pressure rating of the heat exchanger and construction material.(Kazemi & Samadi, 2016)

Where, F_{P, HX}, and F_{M, HX} are correction factors for operating pressure and material, respectively. Equation (10) calculates the initial cost of the heat exchanger and A is the heat-exchanging area. Equation (11) determines the correction factor for heat exchanger pressure, and P is the pressure for heat exchanger.

B₁, B₂, K₁, K₂, K₃, C₁, C₂, and C₃ in Equations (9) - (11) are cost evaluation coefficients (Table 8). These coefficients have been determined previously in the study by Turton (Turton et al., 2008). Moreover, according to Equation (9), CEPCI (Chemical Engineering Plant Cost Index) values are equal to 668 and 397 for 2020 and 2009, respectively. (Ghorbani, Mehrpooya, & Shokri, 2020; Mignard, 2014; Turton et al., 2008)

Table 8. Cost evaluation coefficients for bare module cost of the shell-and-tube heat exchanger (Karimi & Mansouri, 2018; Kazemi & Samadi, 2016; Turton et al., 2008).

Unlike total capital investment, operating costs are subjected to the operation time of the NGL plant. Considering the maintenance time, annual operation time is as follows:(Kazemi & Samadi, 2016)

The operating costs play a major role in the scheme's feasibility. Which include all expenses that are proportional to the total operating capacity of the NGL plant. Moreover, the operating costs are involved at two places of refrigeration cycle including the thermal costs of heat exchangers and electricity cost of compressors, reboiler, and condenser. Annual total net operating cost (OC) for NGL plant is determined by the following Equation: (Bhran, Hassanean, & Helal, 2016; Mehrpooya, Taromi, & Ghorbani, 2019; Zhang et al., 2018) Data presented in Table 9 are used to calculate the operating cost of NGL plant.

Table 9: Parameters considered in operating cost of NGL plant

Table 10: Energy consumption table of the plant refrigeration cycle for operating costs.

Where the energy content of fuel or electricity is defined as the fuel’s complete combustion under well-defined conditions.

The operating cost may change in the new NGL plant scheme in comparison with the NGL plant in the current state, which can be determined by Equation (18). Based on TCI and total OC, PBP can be estimated by Equation (19) (Osman, Mutalib, & Shigidi, 2016).

The total savings in the purchased equipment costs and operating costs saving of the NGL plant are equal to 3,891,560 (USD) and 3,757,451 (USD/year), respectively for the newly designed plant by the pinch and self-refrigeration scheme. PBP for pinch and self-refrigeration scheme is 378 days after startup.

Conclusion

In this study, the results of applying three practical approaches of pinch analysis including retrofit, grass-root, and self-refrigerated methods were presented for a large-scale NGL recovery plant. As a novelty, the self-refrigeration method is integrated with pinch analysis to reduce the limitations of the external utility saving.

After identifying all the primary process data of the plant, the obtained ΔTmin was 10.645 °C. The results indicated that the NGL plant requires a large amount of external utility. By applying, the grass-root method and changing the ΔTmin to 3 °C, the hot and cold required utility will reduce to 10,156.4 and 354.44 kW, respectively. Which reduced the required shaft work of refrigeration cycle compressors by 61.63 %. The pinch and self-refrigeration scheme resulted in reducing compressors' shaft work by 86% and saving 43% of the needed energy for the reboiler and condenser. The economic evaluation showed that the pinch and self- refrigeration integrated scheme is feasible and profitable. Providing other possible self-refrigeration methods can remove the refrigeration cycle completely.

Acknowledgments

The author is grateful to NISOC, for permission to publish this work.