The degradation of amine-based solvents, such as methyldiethanolamine (MDEA) and diethanolamine (DEA), due to the accumulation of heat-stable salts (HSS), poses a significant challenge to the efficiency, safety, and sustainability of acid gas removal systems. Ion exchange using strong-base anion resins has been widely adopted as a practical method for HSS removal; however, the optimization and control of this process remain challenging due to its nonlinear and multivariable nature. In this study, a predictive model based on artificial neural networks was developed to estimate the residual HSS concentration and final solution pH following ion exchange purification. A comprehensive dataset
representing industrially relevant variations in key process parameters-initial HSS concentration, amine strength, flow rate, initial pH, and temperature-was generated and used for model training in MATLAB. The ANN architecture consisted of a two-hidden-layer feedforward network trained using Bayesian regularization, enabling robust learning without overfitting. The model achieved high correlation coefficients of R = 0.9586 for overall prediction, R = 0.9173 for HSS concentration, and R = 0.9898 for pH prediction. Error histograms demonstrated low and symmetrically distributed residuals, confirming the model’s accuracy and generalization capabilities. These results confirm that ANNs can serve as a reliable surrogate model for real-time monitoring and predictive control in solvent purification systems. The proposed methodology contributes to the development of intelligent process optimization in chemical engineering, enhancing operational efficiency and reduced environmental impact.
The degradation of amine-based solvents, such as methyldiethanolamine (MDEA) and diethanolamine (DEA), due to the accumulation of heat-stable salts (HSS), poses a significant challenge to the efficiency, safety, and sustainability of acid gas removal systems. Ion exchange using strong-base anion resins has been widely adopted as a practical method for HSS removal; however, the optimization and control of this process remain challenging due to its nonlinear and multivariable nature. In this study, a predictive model based on artificial neural networks was developed to estimate the residual HSS concentration and final solution pH following ion exchange purification. A comprehensive dataset
representing industrially relevant variations in key process parameters-initial HSS concentration, amine strength, flow rate, initial pH, and temperature-was generated and used for model training in MATLAB. The ANN architecture consisted of a two-hidden-layer feedforward network trained using Bayesian regularization, enabling robust learning without overfitting. The model achieved high correlation coefficients of R = 0.9586 for overall prediction, R = 0.9173 for HSS concentration, and R = 0.9898 for pH prediction. Error histograms demonstrated low and symmetrically distributed residuals, confirming the model’s accuracy and generalization capabilities. These results confirm that ANNs can serve as a reliable surrogate model for real-time monitoring and predictive control in solvent purification systems. The proposed methodology contributes to the development of intelligent process optimization in chemical engineering, enhancing operational efficiency and reduced environmental impact.
Amin asosidagi erituvchilar, xususan, metildietanolamin (MDEA) va dietanolamin (DEA)ning issiqlikka bardoshli tuzlar (HSS) to‘planishi natijasida yemirilishi kislotali gazlarni ajratib olish tizimlarining samaradorligi, xavfsizligi va barqarorligiga jiddiy tahdid soladi. Kuchli asosli anionli smolalar yordamida ion almashinish usuli HSSni yo‘qotish uchun keng qo‘llanadigan amaliy usullardan biridir. Biroq ushbu jarayonning optimallashtirilishi hamda boshqaruvi uning noaniqva ko‘p omilli tabiati sababli murakkab bo‘lib qolmoqda. Ushbu tadqiqotda ion almashinish orqali tozalangandan so‘ng eritmada qolgan HSS miqdori va yakuniy pH darajasini bashorat qilish uchun sun’iy neyron tarmoqlar asosidagi bashorat-
lash modeli ishlab chiqildi. Modelni o‘rgatish uchun MATLAB muhitida sanoatga oid asosiy texnologik parametrlar – dastlabki HSS konsentratsiyasi, amin miqdori, oqim tezligi, boshlang‘ich pH va harorat bo‘yicha o‘zgaruvchanlikni o‘z ichiga oluv-chi keng qamrovli ma’lumotlar to‘plami yaratildi. Sun’iy neyron tarmoq arxitekturasi ikki yashirin qatlamli feedforward (oldinga yo‘naltirilgan) neyron tarmog‘idan tashkil topgan bo‘lib, u Bayesiy regulyarizatsiya usuli asosida o‘qitildi. Bu usul ortiqcha moslashuvga (overfitting) yo‘l qo‘ymasdan kuchli o‘rganishni ta’minladi. Model yuqori korrelyatsiya koeffitsiyentlariga erishdi: umumiy bashoratlashda R = 0.9586, HSS konsentratsiyasi uchun R = 0.9173 va pH bashorati uchun R = 0.9898. Xatoliklar gistogrammalari kam va simmetrik taqsimlangan qoldiq qiymatlarni ko‘rsatdi, bu modelning aniqligi va umumlashtirish qobiliyatini tasdiqlaydi. Natijalar shuni ko‘rsatdiki, sun’iy neyron tarmoq erituvchilarni tozalash tizimlarida real vaqt rejimida monitoring qilish va bashoratli boshqaruv uchun ishonchli model sifatida xizmat qilishi mumkin. Taklif etilgan metodologiya kimyo muhandisligi sohasi-da intellektual jarayonlarni optimallashtirishga hissa qo‘shib, operatsion samaradorlikni oshirish va atrof -muhitga salbiy ta’sirni kamaytirishga yordam beradi.
Разложение аминосодержащих растворителей, таких как метилдиэтаноламин (MDEA) и диэтаноламин (DEA), вследствие накопления термостойких солей (HSS) создаёт серьёзные проблемы для эффективности, безопасности и устойчивости систем удаления кислых газов. Ионообмен с применением сильнoосновных анионных смол
широко используется как практический метод удаления HSS; однако оптимизация и управление этим процессом остаются затруднительными из-за его нелинейного и многопараметрического характера. В рамках данного исследования была разработана прогнозная модель на основе искусственных нейронных сетей для оценки остаточной концентрации HSS и конечного уровня pH раствора после очистки методом ионообмена. Для обучения модели в среде MATLAB был создан обширный набор данных, отражающий промышленно значимые вариации ключевых технологических параметров: начальной концентрации HSS, содержания амина, скорости потока, исходного уровня pH и температуры. Архитектура искусственных нейронных сетей представляла собой прямораспространяющуюся сеть с двумя скрытыми слоями, обучение которой осуществлялось методом байесовской регуляризации, что обеспечило устойчивое обучение и позволило избежать переобучения. Модель продемонстрировала высокие коэффициенты корреляции: R = 0,9586 – для общего прогнозирования, R = 0,9173 – для концентрации HSS, R = 0,9898 – для предсказания pH. Гистограммы ошибок показали низкие и симметрично распределённые остаточные значения, что подтверждает точность модели и её способность к обобщению. Полученные результаты демонстрируют, что искусственные нейронные сети могут использоваться как надёжная заменяющая модель для мониторинга и прогнозного управления процессами очистки растворителей в реальном времени. Предложенная методология способствует развитию интеллектуальной оптимизации процессов в
области химической инженерии, повышая операционную эффективность и снижая негативное воздействие на окружающую среду.
| № | Muallifning F.I.Sh. | Lavozimi | Tashkilot nomi |
|---|---|---|---|
| 1 | Norqulov J.F. | Senior Lecturer | Karshi State Technical University |
| 2 | Muradov R.S. | Doctor of Technical Sciences, Professor | Bukhara State Technical University |
| 3 | Kodirov O.S. | PhD in Technical Sciences, Associate Professor | National University of Uzbekistan |
| № | Havola nomi |
|---|---|
| 1 | Aripdjanov, O. Y., Turobjonov, S. M., Nurullaev, S. P., & Azimova, S. A. (2023). Absorbent composites based on DEA and MDEA and using water-soluble polyelectrolytes. EBSCOhost content item. Retrieved June 3, 2025. https://openurl.ebsco.com/contentitem/ gcd:173505949?sid=ebsco:plink:crawler&id=ebsco:gcd:173505949 |
| 2 | Akmalxon, K., Suvankul, N., & Jamshidbek, X. (2024). Cleaning of methyldiethanolamine and diethanolamine solutions used in natural gas cleaning. Universum: Technical Sciences, 10(5[122]), Article 5(122). Retrieved June 3, 2025. https://cyberleninka.ru/article/n/cleaning-of- methyldiethanolamine-and-diethanolamine-solutions-used-in-natural-gas-cleaning 7universum.com |
| 3 | Turayev, T. B., Igamkulova, N. A., Mengliyev, S. S., & Rakhimov, H. N. (2024). Recovery of spent methyldiethanolamines and reduced environmental impact. E3S Web of Conferences, 491, 04010. https://doi.org/10.1051/e3sconf/202449104010 e3s-conferences.org+1 |
| 4 | Verma, N., & Verma, A. (2009). Amine system problems arising from heat-stable salts and solutions to improve system performance. Fuel Processing Technology, 90(4), 483–489. https://doi. org/10.1016/j.fuproc.2009.02.002 |
| 5 | Grushevenko, E. A., Bazhenov, S. D., Vasilevskii, V. P., Novitskii, E. G., & Volkov, A. V. (2018). Two- step electrodialysis treatment of monoethanolamine to remove heat-stable salts. Russian Journal of Applied Chemistry, 91(4), 602–610. https://doi.org/10.1134/S1070427218040110 |
| 6 | Meng, H., Zhang, S., Li, C., & Li, L. (2008). Removal of heat-stable salts from aqueous solutions of N-methyldiethanolamine using a specially designed three-compartment configuration electrodialyzer. Journal of Membrane Science, 322(2), 436–440. https://doi.org/10.1016/j.memsci.2008.05.072 |
| 7 | Kikhavani, T., Mehdizadeh, H., Van der Bruggen, B., & Bayati, B. (2021). Removal of heat-stable salts from lean amine of a gas refinery via electrodialysis. Chemical Engineering & Technology, 44(2), 318–328. https://doi.org/10.1002/ceat.202000375 |
| 8 | Chen, F., et al. (2020). Removal of heat-stable salts from N-methyldiethanolamine wastewater by anion exchange resin coupled three-compartment electrodialysis. Separation and Purification Technology, 242, 116777. https://doi.org/10.1016/j.seppur.2020.116777 |
| 9 | Ghorbani, A., et al. (2020). Application of NF polymeric membranes for removal of multicomponent heat-stable salts (HSS) ions from methyl diethanolamine (MDEA) solutions. Molecules, 25(21), Article 4911. https://doi.org/10.3390/molecules25214911 |
| 10 | Mansoori, S. A. A., et al. (2022). HSS anions reduction combined with the analytical test of aqueous MDEA in South Pars gas complex. Natural Gas Industry B, 9(3), 318–324. https://doi. org/10.1016/j.ngib.2022.06.004 |
| 11 | Afan, H. A., El-Shafie, A., Yaseen, Z. M., Hameed, M. M., Wan Mohtar, W. H. M., & Hussain, A. (2015). ANN-based sediment prediction model utilizing different input scenarios. Water Resources Management, 29(4), 1231–1245. https://doi.org/10.1007/s11269-014-0870-1 |
| 12 | Afram, A., Janabi-Sharifi, F., Fung, A. S., & Raahemifar, K. (2017). Artificial neural network (ANN)-based model predictive control (MPC) and optimization of HVAC systems: A state-of-the-art review and case study of a residential HVAC system. Energy and Buildings, 141, 96–113. https://doi. org/10.1016/j.enbuild.2017.02.012 |
| 13 | Moon, J. W. (2012). Performance of ANN-based predictive and adaptive thermal-control methods for disturbances in and around residential buildings. Building and Environment, 48, 15–26. https://doi.org/10.1016/j.buildenv.2011.06.005 |
| 14 | Eshbobaev, J., Norkobilov, A., Usmanov, K., Khamidov, B., Kodirov, O., & Avezov, T. (2024, September). Control of wastewater treatment processes using a fuzzy logic approach. In The 3rd International Electronic Conference on Processes (p. 39). MDPI. https://doi.org/10.3390/engproc2024067039 |
| 15 | Pappada, S. M., et al. (2020). An artificial neural network-based predictive model to support optimization of inpatient glycemic control. Diabetes Technology & Therapeutics, 22(5), 383–394. https://doi.org/10.1089/dia.2019.0252 |
| 16 | Meenal, R., & Selvakumar, A. I. (2018). Assessment of SVM, empirical and ANN-based solar radiation prediction models with most influencing input parameters. Renewable Energy, 121, 324–343. https://doi.org/10.1016/j.renene.2017.12.005 |
| 17 | Crystals Editorial Office. (2025). Data-driven ANN-based predictive modeling of mechanical properties of 5Cr-0.5Mo steel: Impact of composition and service temperature. Crystals, 15(3), 213. Retrieved June 3, 2025. https://www.mdpi.com/2073-4352/15/3/213 |
| 18 | Society of Petroleum Engineers. (2019). Development of ANN-based predictive model for miscible CO₂ flooding in sandstone reservoir. In SPE Middle East Oil and Gas Show and Conference (MEOS 2019). Retrieved June 3, 2025. https://onepetro.org/SPEMEOS/proceedings- abstract/19MEOS/3-19MEOS/D032S086R002/218309 |
| 19 | Lee, J., Lim, J., Cho, H., & Kim, J. (2020). Optimization strategies for amine regeneration process with heat-stable salt removal unit. Applied Chemical Engineering, 31(5), 575–580. https://doi. org/10.14478/ace.2020.1073 koreascience.kr+1 |
| 20 | Pal, P., AbuKashabeh, A., Al-Asheh, S., & Banat, F. (2014). Accumulation of heat-stable salts and degraded products during thermal degradation of aqueous methyldiethanolamine (MDEA) using microwave digester and high-pressure reactor. Journal of Natural Gas Science and Engineering, 21, 1043–1047. https://doi.org/10.1016/j.jngse.2014.11.007 |
| 21 | Eshbobaev, J., Khamidov, B., & Fallanza, M. (2025). Application of an adaptive neuro-fuzzy inference system to control the wastewater treatment process. Chemical Technology, Control and Management, 2025(1), 52–60. https://doi.org/10.59048/2181-1105.1654 |