Sensitivity analysis of using diethanolamine instead of methyldiethanolamine solution for GASCO’S Habshan acid gases removal plant

Abstract

Sweeting natural gas processes are mainly focused on removing carbon dioxide (CO₂) and hydrogen sulfide (H₂S). The high-energy requirements and operational limitations make amine absorption process sensitive to any change in conditions. This paper presented a steady-state simulation using Hysys to reasonably predict removal amounts of carbon dioxide and hydrogen sulfide from natural gas with the diethanolamine (DEA) solvent. The product specifications are taken from the real plant (GASCO’S Habshan) which uses the methyldiethanolamine (MDEA) solvent, while this simulation uses DEA under the same operation conditions. First, the simulation validation has been checked with the data of the real plant. The results show accurate prediction for CO₂ slippage and accepted agreement for H₂S content compared with the data of the plant. A parametric analysis has been performed to test all possible parameters that affect the performance of the acid gases removal plant. The effects of operational parameters are examined in terms of carbon dioxide and hydrogen sulfide contents in clean gas and reboiler duty.

Keywords

acid gas / diethanolamine / methyldiethanolamine / carbon dioxide capturing / HYSYS simulation

Introduction

The demand for natural gas increases dramatically around the world as a source of energy. The trade of natural gas increased by 60% between 2006 and 2014 which was almost twice of the amount 10 years ago [¹]. However, natural gas may contain nonhydrocarbon components like carbon dioxide (CO₂) and hydrogen sulfide (H₂S). These components are considered as undesirable contents which can affect the quality of natural gas, cause corrosion problems and greenhouse gas emissions [²]. Therefore, there are many technologies to purify natural gas. One of the most mature methods to capture carbon dioxide and hydrogen sulfide from natural gas is the alkanolamine solution [³]. The advantages of using the alkanolamine absorption technology are cost effectiveness, the ability to handle large amount of acid gases, and the ability to reuse the absorbents again [⁴]. On the other hand, the disadvantages can be illustrated to corrosion problems, degradation, and high energy requirements for rich amine regeneration [⁵]. The choosing of specific solvent of the alkanolamine group depends on the operational conditions and the process specifications. Table 1 is a comparison of different alkanolamines solvent with various operational properties.

Table 1 shows that the alkanolamines group solvent comes with almost the same range of temperature. In addition, the solvent strength mainly depends on the operation conditions of the plant and the product specifications. The real plant uses the MDEA solvent to absorb acid gas while the simulation suggests the application of the DEA solvent to check their ability to reach the product specifications of the real plant. As mentioned, the absorption technology needs high energy which is considered as the disadvantage, and the main consumption equipment is the reboiler of the regeneration column [⁹]. Many studies have been conducted to investigate the reduction of the required energy for the absorption process and the solvent choosing can play a key role in reducing the energy consumption. MDEA is regarded as a good choice for aqueous solutions because of its high absorption rate and low-cost [¹⁰] while DEA can be a competitor option because of its low pressure operation and low heat of CO₂ reaction [¹⁰]. DEA, as secondary amines, is less reactive to sulfur and their reaction products are not corrosive which can make diethanolamine attractive to be chosen to capture acid gases [¹¹]. The object of this study to apply DEA instead of MDEA for real acid gases plant (GASCO’S Habshan) to check the amount of CO₂ and H₂S captured for simulation and plant data. The analysis study has been performed on the factors that have a direct impact on the removal efficiency to test the effects on the process performance.

Process description

The sour feed gas enters the absorption column with a high pressure from the bottom and the lean solvent from the top to configure counter current. The solvent used in this simulation is diethanolamine (DEA) which is the solvent employed to capture acid gases. The design conditions and the plant capacity used in this simulation are based on the data provided by GASCO’S Habshan, where the plant capacity is up to 7.198 × 10³ m³/h. The sour feed gas enters the absorption column which has 25 plates with a temperature of 50°C and a design pressure of 6700 kPa while the lean amine feeds to the absorption column with a temperature of 55°C and the same design pressure. The reaction of DEA with acid gas (CO₂ and H₂S) is exothermic, therefore low temperature and high pressure are favored.

The clean or sweet gas which is free of H₂S and CO₂ after sufficient reaction with DEA, gathers at the top of the absorber with a temperature of 74.47°C, and is then cooled by using a fin fan cooler to reach a specified temperature of 57°C. The rich DEA solution leaving the absorber at the bottom with a temperature of around 59°C and a pressure of 700 kPa passes through the control valve to reduce its pressure to about 240 kPa. Then, the solution is sent to the flash tank to separate any dissolved hydrocarbons. The DEA solution is heated up by the heat exchanger with the lean solution (bottom of regenerator) to 102°C and 225 kPa before finally pumped to the regenerator column. The regeneration process is endothermic; therefore a high temperature and a low pressure are favored. Hence, the gaseous CO₂ and H₂S leave at the top of the column with the product specifications, while the rich DEA solution leaves at the bottom of regenerator which is then cooled and pumped to the absorber column again. Sweeting of natural gas process by using DEA is illustrated by Fig. 1.

Simulation basis

The natural gas purification process is simulated by using the HYSYS V8.8 program which is used commonly in the oil and gas process. HYSYS offers different types of thermodynamic property fluid packages to deal with hydrocarbons systems such as Peng-Robinson, SRK, glycol package, and acid gas. The fluid package for this simulation is chosen to be acid gas which is recommended for hydrocarbons due to its largest binary interaction parameter database. The absorption and desorption columns are considered as the most critical steps for this simulation. Therefore, some assumptions have been made to simplify the complexity of simulation. First, the liquid-vapor leaves each stage into the columns in equilibrium. Secondly, the heat loss of the absorption and desorption process are assumed to be negligible. Other assumptions for the whole process are counter current for all of the heat exchangers used, and the minimum temperature approach is 10°C. The reactions for the process follows the approach taken by Kent and Eisenberg [¹²], as expressed below.

{{\displaystyle \mathrm{H}}}_{\mathrm{2}}\mathrm{O}\leftrightarrow {\mathrm{H}}^{+}+{\mathrm{O}\mathrm{H}}^{-}

{{\displaystyle {{\displaystyle \mathrm{C}\mathrm{O}}}_{\mathrm{2}}+\mathrm{H}}}_{\mathrm{2}}\mathrm{O}\leftrightarrow {{\displaystyle \mathrm{H}\mathrm{C}\mathrm{O}}}_{\mathrm{3}}^{-}+{\mathrm{H}}^{+}

{{\displaystyle \mathrm{H}\mathrm{C}\mathrm{O}}}_{\mathrm{3}}^{-}\to {{\displaystyle \mathrm{C}\mathrm{O}}}_{\mathrm{3}}^{-\mathrm{2}}+{\mathrm{H}}^{+}

{{\displaystyle {\mathrm{D}\mathrm{E}\mathrm{A}\mathrm{C}\mathrm{O}\mathrm{O}}^{-}+\mathrm{H}}}_{\mathrm{2}}\mathrm{O}\to \mathrm{D}\mathrm{E}\mathrm{A}\mathrm{H}+{{\displaystyle \mathrm{H}\mathrm{C}\mathrm{O}}}_{\mathrm{3}}^{-}

\mathrm{D}\mathrm{E}\mathrm{A}\mathrm{H}+{\mathrm{H}}^{+}\leftrightarrow {{\displaystyle \mathrm{D}\mathrm{E}\mathrm{A}\mathrm{H}}}_{\mathrm{2}}^{+}

where CO₂ may react directly with DEA and consist carbonic acid with water as shown in the reactions above. The reaction of DEA with H₂S is exothermic as shown by the reaction below.

\mathrm{D}\mathrm{E}\mathrm{A}+{{\displaystyle \mathrm{H}}}_{\mathrm{2}}\mathrm{S}\to {\mathrm{D}\mathrm{E}\mathrm{A}\mathrm{H}}^{+}{\mathrm{H}\mathrm{S}}^{-}

About 0.02% of vapor enters the environment from the drum flash which can be considered fairly insignificant.

Results and discussion

Simulation validation

The simulation findings have been compared with the data of the plant to check whether the results are reasonable and reach the plant specifications shown in Table 1. Generally, the product specification for sweet gas is that the carbon dioxide percent is less than 3% with a H₂S content of around 20 ppm. The simulation results indicate that the carbon dioxide content in clean gas is 2.23% which is in the accepted range compared with the data of the plant while the H₂S content is 65.8 ppm which is higher than the specification data. The design data show an error comparison to real plant of about 1.4% for CO₂ and 68% for H₂S contents in sweet gas. There are two possible reasons for the deviation in H₂S percent. First, the solvent used by the real plant is methyldiethanolamine but that used in the simulation is diethanolamine for each absorbent specific range of loading capacity. Mokhatab and Poe mentioned that MDEA had the ability to react instantaneously with H₂S by proton transfer [¹³]. Secondly, acid gases solubility plays a key role in the process of sweeting natural gas, where the solubility of H₂S and CO₂ depends on many factors that can result in the best removal efficiency such as temperature, pressure, and amine concentration. The solubility of H₂S is more sensitive to methane partial pressure, where increasing partial pressure of methane results in reducing the H₂S solubility [¹⁴,¹⁵]. Therefore, the difference in methane content between simulation specification and real plant data can be another reason in reducing H₂S removal efficiency.

Table 2 is a comparison of the simulation results and the data of the real plant. To conclude, the model is able to reach a reasonable percent for CO₂ in sweet gas and even the H₂S content is still in the accepted range where the acceptable range of H₂S content in sweet gas varies by country, application of natural gas, and agency [¹³]. For instant, Texas Railroad Commission, a state agency that regulates the oil and gas industry, defines that sour gas can contain more than 100 ppm of H₂S [¹⁶].

Performance analysis

An energy analysis has been performed in order to compare the plant performance with DEA and MDEA solvents. The results show that the reboiler is the energy intensive unit which consumes up to half of the energy of the whole plant. The results show that the energy consumption at the regeneration unit using the MDEA solvent is higher than that using the DEA solvent under the same condition. Mokhatab and Poe mentioned that MDEA and H₂S reacted instantaneously with a high attraction which consumed more energy at the regeneration unit by the reboiler to separate H₂S from the solvent [¹³]. Therefore, the plant using the DEA solvent needs low energy than that using MDEA. DEA has a lower vapor pressure which means lower solvent losses compared to MDEA [¹⁴]. The low energy consumed at the regeneration unit results in reducing the required heat for lean/rich heat exchanger and the cooler as well. The lean amine from the bottom of desorption at 128.9°C which passes through lean/rich heat exchanger reduces the temperature to around 93.3°C. This temperature is still high to feed the absorber; therefore, the stream passes through the cooling unit to reduce the temperature to 54.19°C. The cooling unit is considered as the second energy consumption in the whole plant. The temperature of lean amine using the MDEA solvent is higher than that using the DEA solvent which means that more energy is required at the cooler unit. Therefore, using DEA as a solvent is more efficient in terms of energy consumption, and CO₂ loading capacity for this plant conditions.

Sour gas feed temperature

In this simulation, the sour gas feed temperature is set to 50°C, which can be varied around ∓20% depending on the season and the operation conditions. A simulation study has been conducted to test the effects of varying sour feed temperature on both CO₂ and H₂S contents in clean gas. The results, as plotted in Fig. 2, show that increasing sour feed temperature leads to the increase in carbon dioxide and hydrogen sulfide contents in sweet gas (Fig. 2 (a)). The reason is attributed to the fact that the CO₂ and H₂S solubility decreases as the temperature increases. The optimum feed gas temperature appears in the range between 40°C and 50°C to balance the product specification and the energy requirements. On the other hand, Fig. 2(b) demonstrates that the reboiler duty slightly increases as the inlet temperature of sour gas increases.

Diethanolamine solution temperature

The temperatures of both sour feed gas stream and solvent solution stream can play a key role in controling the reaction in the absorber. The absorber performance can generally be improved by reducing the column temperature; however it is difficult to manipulate the feed gas temperature. Therefore, the lean amine temperature plays a vital role in optimizing the performance of the absorber column. In this regards, Addington and Ness advised that the temperature difference between feed natural gas stream and solvent solution stream be set to 2°C–5°C, where sour feed gas temperature for this design is 50°C and 55°C for lean amine stream [¹⁷]. A simulation study has been conducted to check the effect of diethanolamine solution temperature on both carbon dioxide content and the reboiler duty. Figure 3 depicts the effect of lean amine temperature, which shows that as the diethanolamine solution temperature increases, the carbon dioxide content decreases. However, the carbon dioxide in the sweet gas starts decreasing slightly at a temperature of above 45°C. The possible reason for this is that the solubility of carbon dioxide reaches the limitation with a temperature around 42°C. The reboiler duty decreases as the temperature of lean amine increases, as shown in Fig. 3. The H₂S content is almost constant with all range of lean amine temperatures.

Absorber column pressure

The feed gas should be received by the sweeting unit at the operational pressure, otherwise the plant will be influenced. A simulation study has been conducted to investigate the effect of absorber pressure on carbon dioxide and hydrogen sulfide contents in clean gas stream. The results illustrate that as column pressure decreases, carbon dioxide and hydrogen sulfide contents in clean gas increase, as shown in Fig. 4(a). The reason for this is that the partial pressure of CO₂ in feed stream is reduced as the absorber pressure decreases while the regenerator reboiler duty decreases slightly with increasing column operational pressure, as shown in Fig. 4(b).

Plant capacity

The fluctuation of feed flow is another factor which needs to be paid attention to. Therefore, it is important to design plants which are able to accommodate different rates of feed gas. Warudkar et al. mentioned that the GASCO Habshan Sweeting Plant can handle smoothly down to 30% of design capacity as claimed by the original equipment manufacturer [¹⁸]. A simulation study has been conducted to test the effect of varying feed gas flowrate on the contents of CO₂ and H₂S in sweet gas. The design capacity is 7.198 × 10³ m³/h, where the results show the design can successfully handle 29% of the design capacity. Figure 5 shows that the carbon dioxide and hydrogen sulfide contents in sweet gas stream increases as sour gas feed rate increases.

Lean amine concentration

Solvent concentration plays a vital role in the capturing process and the management of the energy consumption. Figure 6 displays the effect of different DEA concentrations on carbon dioxide content and reboiler duty of regenerator column. The results indicate that at low concentrations of lean amine of around (30 wt.%), the CO₂ concentration is high. As the DEA concentrations increases, a sharp decline in the amount of carbon dioxidein clean gas is noticed with the increase of reboiler duty. The amount of reboiler duty and the carbon dioxide does not vary much between 50 wt.% and 60 wt.% of DEA concentrations where it is almost constant. The design specifications can be achieved with DEA concentrations between 40 wt.% and 45 wt.%. Solvent concentration and its circulation rate can be manipulated to optimize plant performance and reduce energy consumption.

Regenerator feed temperature

The rich amine temperature can be considered as another factor that can be varied to enhance the process performance. Stewart and Arnold reported that the feed temperature to the regenerator column is around 99°C in literature and operational plants [⁸]. The design temperature of regeneration feed stream is 102°C to match the plant conditions, as the reaction in stripper is endothermic. A simulation study has been conducted to check the effects of varying regen feed temperature to stripper column on carbon dioxide content in clean gas and reboiler duty. Figure 7 shows that at low inlet temperature of rich amine, the CO₂ in clean gas is reduced significantly; however the energy needed for the reboiler is high. As regen inlet temperature increases, the carbon dioxide slippage in clean gas increases and reboiler duty decreases almost linearly. H₂S content is almost constant in all range of regen temperatures.

The inlet temperature of regen can affect carbon dioxide vapor in the feed stream. Figure 8 shows that the amount of carbon dioxide vapor in the feed stream increases as the temperature of regen stream increases. The total regen vapor increases sharply with a temperature above 100°C which agrees with literature finding of 99°C.

Columns number of plates

The stages number of the absorber can be considered as another factor that can be manipulated to enhance the performance of the acid gas removal process. Figure 9 displays effect of adding/removing stages for the absorber on carbon dioxide and hydrogen sulfide contents in the clean gas stream. The results suggest that as the number of plates increases, the content of carbon dioxide and hydrogen sulfide in clean gas decreases slightly, where, beyond 25 stages the decrease in CO₂ percent in the clean gas stream becomes too small which is less than 1 ppm. Increasing stages number comes with additional material and energy cost, and the improvements in reducing acid gas beyond 25 stages is small. Therefore, the number of plates of absorber column is set to 25 stages to balance the removal efficiency, cost, and column performance.

Figure 10 exhibits the effects of adding/removing plates of regenerator column on acid gas content and reboiler duty. The results demonstrate that as the number of plates increases, there is a slight decrease in the carbon dioxide slippage in the clean gas and constant when the number of plates increases more than the designed number of stages. Based on Fig. 9, 15 stages can be considered as the number of the plate which is sufficient enough to converge the regenerator column, as after 15 stages the CO₂ content is constant. The reboiler duty increases slightly with the increase of the number of plates. The H₂S contents remain unchanged with respect to adding/removing number of stages to the regenerator column.

Conclusions

The simulation executed using Aspen HYSYS V8.8 program and applying acid gas as a new fluid package, can predict very well acid gas removal on different operational conditions. The design plants are compared with GASCO’S Habshan Plant in terms of carbon dioxide and hydrogen sulfide amounts in clean gas steam and acid gas stream, operational conditions, and other parameters. In regards of the amount of CO₂ captured, the simulation gives perfect prediction with a difference of less than 1.4% to the actual plant while the 65.8 ppm of H₂S content in clean gas shows significant difference with respect to the data of the actual plant, which is, however, still acceptable in industrial range. A sensitive analysis has been perfumed to check various factors that can affect the process performance. The sour gas feed temperature is varied to test its effects on the amount of acid gas in clean gas stream and reboiler duty. The results show that as sour feed temperature increases, the acid gas in clean gas stream and reboiler duty increases. The absorber pressure is another factor that is investigated to reach the conclusion that as the pressure increases, the acid gas percent in clean gas stream and the reboiler duty decreases. The proposed simulation can successfully handle 29% of the design capacity, as the increase of sour feed rate causes the increase of acid gas contents in sweet gas. The increase of lean amine concentration in feed leads to the decrease of the acid gas content in clean gas and the increase of reboiler duty. The temperature of lean amine has a direct effect on the absorption process, where the increase of lean temperature leads to the decrease of both acid gas and reboiler duty. In addition, regeneration feed temperature has been tested. It is found that as regeneration feed temperature increases, the reboiler duty decreases and carbon dioxide content increases in clean gas stream. In the same vein, the total vapor and carbon dioxide vapor in the regeneration feed stream have been checked. The results show that increasing temperature of feed regen causes increasing of both vapor percent in the stream. To conclude, this simulation gives a reasonable prediction of capturing acid gas from natural gas with chosen operational conditions. The future prospects of this plant focuses on design development in terms of reducing the required energy.

References

[1] Kurle Y M, Wang S, Xu Q. Dynamic simulation of LNG loading, BOG generation, and BOG recovery at LNG exporting terminals. Computers & Chemical Engineering, 2017, 97: 47–58

[2] Seo Y, Park S, Kang H, Ahn Y H, Lim D, Kim S J, Lee J, Lee J Y, Ahn T, Seo Y, Lee H. Isostructural and cage-specific replacement occurring in sII hydrate with external CO₂/N₂ gas and its implications for natural gas production and CO₂ storage. Applied Energy, 2016, 178: 579–586

[3] Gupta A K, Ibrahim S, Al Shoaibi A. Advances in sulfur chemistry for treatment of acid gases. Progress in Energy and Combustion Science, 2016, 54: 65–92

[4] Chu F, Yang L, Du X, Yang Y. Mass transfer and energy consumption for CO₂ absorption by ammonia solution in bubble column. Applied Energy, 2017(190): 1068–1080

[5] Li H, Yan J, Campana P E. Feasibility of integrating solar energy into a power plant with amine-based chemical absorption for CO₂ capture. International Journal of Greenhouse Gas Control, 2012, 9: 272–280

[6] Kohl A, Nielsen R. Gas Purification. Boston: Gulf Professional Publishing, 1997

[7] Mofarahi M, Khojasteh Y, Khaledi H, Farahnak A. Design of CO₂ absorption plant for recovery of CO₂ from flue gases of gas turbine. Energy, 2008, 33(8): 1311–1319

[8] Stewart M, Arnold K. Gas Sweetening and Processing Field Manual. Boston: Gulf Professional Publishing, 2011

[9] Rufford T E, Smart S, Watson G C Y, Graham B F, Boxall J, Diniz da Costa J C, May E F. The removal of CO₂ and N₂ from natural gas: a review of conventional and emerging process technologies. Journal of Petroleum Science Engineering, 2012, 94 –95: 123–154

[10] Aaron D, Tsouris C. Separation of CO₂ from flue gas: a review. Separation Science and Technology, 2005, 40(1–3): 321–348

[11] Diab F, Provost E, Laloue N, Alix P, Fürst W. Effect of the incorporation of speciation data in the modeling of CO₂-DEA-H₂O system. Fluid Phase Equilibria, 2013, 353: 22–30

[12] Kent R L, Eisenberg B. Better data for amine treating. Hydrocarbon Processing, 1976, 55(2): 87–90

[13] Mokhatab S, Poe W A. Chapter 7–Natural gas sweetening. In: Handbook of Natural Gas Transmission and Processing. 2nd ed. Boston: Gulf Professional Publishing, 2012, 253–290

[14] Huttenhuis P J G, Agrawal N J, Hogendoorn J A, Versteeg G F. Gas solubility of H₂S and CO₂ in aqueous solutions of N-methyldiethanolamine. Journal of Petroleum Science Engineering, 2007, 55(1–2): 122–134

[15] Blanchon Le Bouhelec-Tribouillois E, Mougin P, Barreau A, Brunella I, Le Roux D, Solimando R. Simultaneous solubilities of CO₂ and H₂S in diethanolamine aqueous solution. Oil & Gas Science and Technology, 2008, 63(3): 363–372

[16] Øi L E. Aspen HYSYS simulation of CO₂ removal by amine absorption from a gas based power plant. SIMS2007 Conference, Gøteborg, 2007,

[17] Addington L, Ness C. An evaluation of general “Rules of Thumb” in amine sweetening unit design and operation. In: 89th Annual Convention of the Gas Processors Association, Austin, Texas, USA, 2010

[18] Warudkar S S, Cox K R, Wong M S, Hirasaki G J. Influence of stripper operating parameters on the performance of amine absorption systems for post-combustion carbon capture: part I. high pressure strippers. International Journal of Greenhouse Gas Control, 2013, 16: 342–350

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature