Modeling of Chemical Kinetics for Methanol-to-Olefin (MTO) Process

이민경 포항공과대학교 일반대학원 2020 국내박사

In this research, the chemical reaction behavior of the methanol-to-olefin (MTO) process was formulated by kinetic models to predict the yield of product and selectivity. Two types of kinetic models with different complexity depending on the application were presented for the MTO process under the SAPO-34 catalyst. In Chapter 2, a mechanistic kinetic model that takes into account the elementary steps including ions of MTO reaction was developed. Referring to the preceding studies, an overall reaction mechanism was established that reflected the autocatalysis nature of the MTO reaction and the interplay between the hydrocarbon pools. For kinetic modeling based on the complex chemical reaction mechanism, the approximate approach based on transition state theory, Evans-Polanyi relation, and thermodynamic constraints are applied to reduce the kinetic parameters to be estimated. The kinetic parameters are determined using experimental data obtained from literature by the genetic algorithm. This model provides satisfactory information on product distribution in various operating conditions based on highly theoretical approaches. In Chapter 3, a lumped kinetic model was developed to predict the seven lumps in which the main products of MTO are grouped. The deactivation kinetics of the MTO reaction is studied based on the proposed 7-lump kinetic model. The model is based on assumptions that methanol conversion is a first-order reaction and the active catalyst reduction is proportional to the conversion. The kinetic parameters were determined using experimental data measured in a fixed bed reactor by the genetic algorithm. By defining the deactivation of SAPO-34 as the loss of the active catalyst, the deactivation constant is the only intrinsic parameter required to describe the effect of catalyst deactivation on the conversion and product yields with time on stream. This theoretical approach has been demonstrated to be effective in modeling the complex deactivation kinetics of MTO. Both proposed models have been theorized in many parts to reduce the computational loads compared to the previous studies, but they also guarantee reasonable results. The validity of the two models has been verified by experimental data and statistical methods. These models are expected to be used in mechanism research, catalytic design, reactor design, and operating condition optimization. 본 연구에서는 메탄올-올레핀 전환 공정 (MTO)의 화학 반응 거동을 수학적으로 모델링하고 생산물 수율과 선택도를 예측하기 위한 반응 역학 모델 (kinetic model)을 수립하였다. SAPO-34 촉매 하의 MTO 공정에 대하여 적용 목적에 따라 다른 복잡도를 가지는 2가지 유형의 반응 역학 모델을 제시하였다. 제2장에서는 MTO 반응의 이온 단위 거동까지 고려한 메커니즘 기반 역학 모델 (mechanistic kinetic model)을 개발하였다. 선행 연구들을 참고하여 MTO 반응의 자촉매 (autocatalysis) 특성과 탄화수소 중심 (hydrocarbon pool) 간의 상호작용을 반영하는 전반적인 반응 메커니즘 (reaction mechanism)을 구축하였다. 이 복잡한 화학 반응 메커니즘을 기반한 반응 속도 모델링을 위해 전이상태이론에 기반한 근사적 접근법 (approximate approach), 에반스-폴라니 관계식 (Evans-Polanyi relation), 그리고 열역학적 조건들이 적용되면서 반응 매개변수 추정량을 크게 감소시켰다. 반응 매개변수들은 문헌에 공개된 실험 데이터를 바탕으로 유전 알고리즘 (genetic algorithm)으로 추정하였다. 본 모델은 강건한 이론적 접근을 기반으로 다양한 조업 조건에서의 각 성분들의 거동에 대한 신뢰성 있는 정보를 제공한다. 또한 반응 내 탄화수소 중심의 영향을 정량화할 수 있는 모델이라는 점에서 독창성을 가진다. 제3장에서는 총 7개의 주요 성분의 거동만 예측하는 집중 역학 모델 (lumped kinetic model)을 개발하였다. 촉매 비활성화에 대한 MTO 반응 역학은 개발된 집중 역학 모델을 기반으로 연구되었다. 본 모델은 모든 반응 단계는 1차 반응이며, 유효 활성 촉매는 원료 MDOH (methanol, dimethyl ether) 전환율에 비례한다는 가정을 전제로 한다. 반응 매개변수는 고정층 반응기에서 수득한 실험 데이터를 이용하여 유전 알고리즘으로 추정하였다. SAPO-34 촉매의 비활성화를 유효 활성 촉매의 감소로 정의함으로써, 비활성화 상수는 촉매 비활성화가 각 반응 성분들에 미치는 영향을 설명하기 위한 유일한 고유 매개변수이다. 이러한 이론적 접근은 MTO의 복잡한 비활성화 동역학 (deactivation kinetics)을 모델링에 효과적임이 확인되었다. 제안된 두 모델은 기존 연구들에 비해 모델의 많은 부분을 이론화하여 계산 부하를 줄이면서 합리적인 결과를 도출했다는 것에 공통점이 있다. 두 모델의 결과는 실험 데이터와 통계적 방법으로 그 정합성이 검증되었으며, 메커니즘 연구, 촉매 설계, 반응기 설계, 조업 조건 최적화에 활용될 것으로 기대된다.

사과 중 농약의 잔류특성에 따른 Kinetic Model 적용

김지환 경북대학교 대학원 2013 국내석사

사과에 대한 살충제 novaluron, lufenuron, carbaryl, teflubenzuron, flubendiamide와 살균제 diniconazole, metcoanzole, kresoxim-methyl의 생물학적 반감기와 잔류양상을 조사하였다. Carbaryl을 제외한 7가지 농약의 경우 0일차 기준량과 3배량 처리구 모두 잔류허용기준을 넘지 않았다. 재배 기간 중 사과의 잔류농도는 약제 살포 후 14일 경과 시 novaluron 기준량 및 배량에서 각각 70.6% 및 68.5%의 농약이 분해되었다. Lufenuron은 기준량 및 배량에서 각각 88.9% 및 87.5%의 농약이 분해되었다. Carbaryl은 기준량 및 배량에서 각각 81.0% 및 60.6%의 농약이 분해되었다. Teflubenzuron은 기준량 및 배량에서 각각 39.1% 및 29.5%의 농약이 분해되었다. Flubendiamide는 기준량 및 배량에서 각각 53.8% 및 53.4%의 농약이 분해되었다. Diniconazole은 기준량 및 배량에서 각각 88.9% 및 87.5%의 농약이 분해되었다. Metconazole은 기준량 및 배량에서 각각 60.0% 및 56.0%의 농약이 분해되었다. Kresoxim- methyl은 기준량 및 배량에서 각각 74.5% 및 42.1%의 농약 분해율을 보였다. 본 실험에서는 carbaryl 기준량의 경우 first order kinetic model을 이용하였을 때 보다 second order kinetic model을 이용할 경우 3일 가량 짧아지는 경향을 보이고 3회 처리의 경우 2.4일 가량이 짧아지는 경향을 보였다. 그리고 teflubenzuron 기준량의 경우 2일 가량이 길어지는 것을 볼 수 있었으며, flubendiamide 3회 처리의 경우 1일 가량이 짧아지는 경향을 볼 수 있었다. 또한 metconazole 기준량의 생물학적 반감기에서는 2일 가량 짧아지는 것을 볼 수 있었고 3회 처리의 경우 1일 가량이 짧아지는 경향을 보인다. 본 실험 결과를 이용하여 생산단계 잔류허용기준량을 설정 하였으며, 생산단계 잔류허용기준량 설정에 있어서 기준량을 기준으로 하여 설정하였으며 일자별 잔류량이 검출한계 아래로 나오는 경우는 3회 처리량을 기준으로 설정하기에 diniconazole은 기준량이 아닌 3회 처리량을 기준으로 생산단계 잔류허용량을 설정하였다. Dissipation patterns of insecticides (novaluron, lufenuron, carbaryl, teflubenzuron, flubendiamide) and fungicides (diniconazole, metconazole, kresoxim-methyl) during the cultivation of apple were established by utilizing the dissipation curve and the biological half-lives of the pesticides. The tested pesticides except for carbaryl did not go beyond the acceptable limit of maximum residue limits (MRLs) for them in apple at a recommended dose and the level of three times of the recommended dose. Novaluron at 14 days after pesticide spraying was degraded 70.6 and 68.5% at a recommended dose and the level of three times of the recommended dose, respectively. Lufenuron was degraded 88.9 and 87.5% at a recommended dose and the level of three times of the recommended dose, respectively. Carbaryl was degraded 81.0 and 60.6% at a recommended dose and the level of three times of the recommended dose, respectively. Teflubenzuron was degraded 39.1 and 29.5% at a recommended dose and the level of three times of the recommended dose, respectively. Flubendiamide was degraded 53.8 and 53.4% at a recommended dose and the level of three times of the recommended dose, respectively. Diniconazole was degraded 88.9 and 87.5% at a recommended dose and the level of three times of the recommended dose, respectively. Metconazole was degraded 60.0 and 56.0% at a recommended dose and the level of three times of the recommended dose, respectively. Kresoxim-methyl was degraded 74.5 and 42.1% at a recommended dose and the level of three times of the recommended dose, respectively. The pattern of carbaryl dissipation at the recommended dose showed 3 days shorter in the second order kinetic model when compared to the first order kinetic model. The pattern of carbaryl dissipation at the level of three times of the recommended dose showed 2.4 days shorter in second order kinetic model when compared first order kinetic model. The pattern of teflubenzuron dissipation at the recommended dose showed 2 days shorter in second order kinetic model when compared to first order kinetic model. The pattern of flubendiamide dissipation at the level of three times of the recommended dose showed 1 day shorter in second order kinetic model when compared to first order kinetic model. The biological half-life of metconazole at the recommended dose showed 2 days shorter in second order kinetic model when compared to first order kinetic model. The biological half-life of metconazole at the level of three times of the recommended dose showed 1 day shorter in second order kinetic model when compared to first order kinetic model. Taken together, results were used to establish pre-harvest residue limits for the tested pesticides and the pre-harvest residue limit for diniconazole was created with the amount of three treatments because the daily residual amount of diniconazole was found under the limit of the detection.

Determination of integral stereoselectivity on medium-chain triacylglycerol hydrolysis with interface-based kinetic model

박재현 서울대학교 대학원 2023 국내석사

본 논문에서는 기존 입체선택성 분석법의 한계를 극복하기 위해 도입된 통합적 입체선택성을 저열량 지방과 빠른 에너지원으로의 가치를 지닌 중쇄지방에 적용하기 위해 대표적인 중쇄지방인 트라이카프릴로일글리세롤(TCG)에 대한 라이페이스의 통합적 입체선택성을 구명할 수 있는 분석법과 동역학 모델을 구축하였다. CHIRALPAK AY-3 칼럼을 이용한 HPLC-UV/ELSD 시스템을 통해TCG와 그의 가수분해 산물인 모든 종류의 다이카프릴로일글리세롤(DCG)과 모노카프릴로일글리세롤(MCG) 이성질체를 직접적인 방식으로 동시에 분리하였다. 15 분의 분석시간내에 모든 분석물이 2.4 이상의 높은 분리능으로 분리되었다. 구축된 분석법으로 돼지 췌장, Chromobacterium viscosum, Pseudomonas fluorescens 유래 라이페이스 및 Candida antarctica 유래 라이페이스 A를 이용하여 역미셀계에서의 TCG 가수분해 반응을 분석하였다. 분석 결과 4가지 라이페이스 모두 이전에 보고된 것과 동일한 선택성을 보여 구축된 방법이 TCG 가수분해를 분석하기에 정확하며 효율적임이 입증되었다. 분석결과를 통해 라이페이스의 통합적 입체선택성을 정량적으로 나타내기 위해 새로운 계면 기반 동역학 모델을 구축하였다. 계면 기반 동역학 모델은 이전 연구에서 구축된 동역학 모델을 기반으로 반응물과 생성물 및 효소의 계면에서의 작용을 추가적으로 고려하여 구축되었다. 계면 기반 동역학 모델을 통해 산출된 예상값은 TCG, DCG, MCG 및 CA의 반응곡선을 이전 모델(R2>0.94)에 비해 더 우수하게 예측하였다(R2>0.98). 또한 산출된 동역학 상수를 비교하였을 때 TCG에서 DCG로의 속도를 나타내는 동역학 상수의 대소관계는 4가지 모델 라이페이스의 선택성과 일치하였다. 추가적으로 DCG와 MCG에 대한 동역학 상수를 통해 TCG부터 글리세롤까지 각 라이페이스들의 주 반응경로를 확인할 수 있었다. 결론적으로 중쇄지방에 대한 통합적 입체선택성을 구명하기 위해 구축된 분석법과 동역학 모델은 라이페이스를 통한 반응을 분석하고 연구하는데 도움을 주어 저열량 지질과 같은 영양학적 가치가 높은 중쇄지방 기반 재구성 지질 개발 및 생산에 큰 보탬이 될 것으로 기대된다. Lipase is an enzyme that hydrolyzes triacylglycerol and is used in various fields such as food, medicine, and cosmetics. An important characteristic of lipase-catalyzed hydrolysis of triacylglycerol (TAG) is that lipase exhibits a different preference for the sn-position of the glycerol backbone. The stereoselectivity of lipases on TAG has been evaluated by measuring the ratio of diacylglycerols (DAGs) produced at the early phase of the reaction. However, the conventional methods have a limitation in that it does not consider the stepwise hydrolysis reaction of acylglycerols, i.e., the conversion of diacylglycerol to monoacylglycerol (MAG) and glycerol, and acyl migration between acylglycerol isomers. To overcome these limitations, the concept of integral stereoselectivity was suggested, which considers the entire process of TAG hydrolysis reaction with trioleoylglycerol (TOG) as a model substrate. In recent years, the utilization of medium-chain triacylglycerol (MCT) has gained much attention in food industry because it can be used as low-calorie lipids. Therefore, in this study, an analytical method to determine the integral stereoselectivity of lipases on MCT was established using tricapryloylglycerol (TCG), a representative MCT as a substrate. The method to simultaneously analyze TCG and its hydrolysis products was constructed through the HPLC-UV/ELSD system. During a short analysis time of 15 min, a total of 8 analytes including all capryloylglycerols were clearly separated with a separation factor of 2.4 or higher. Subsequently, the TCG hydrolysis by lipase was carried out through four model lipases: lipase from porcine pancreas (PPL), Pseudomonas fluorescens (PFL), Chromobacterium viscosum (CVL), and lipase A from Candida antarctica (CALA) representing the different stereoselectivity (sn-1,3, sn-1, sn-3, and sn-2, respectively). A novel interface-based kinetic model was constructed to determine integral stereoselectivity by considering the interfacial characteristics of lipases and components in reverse micelle based on the previous kinetic model. The interface-based kinetic model has improved the accuracy of data fitting (R2>0.98) compared to the previous kinetic model (R2>0.94). From kinetic parameters calculated from the kinetic modeling, the main reaction pathways of TCG hydrolysis were from TCG to 1,2-sn or 2,3-sn-DCG and then 2-sn-MCG for PPL, PFL, and CVL, while for CALA that was from TCG to 1,3-sn-DCG to 1-sn and 3-sn-MCG. In conclusion, this study developed the analytical method and the interface-based kinetic model that enable the determination of integral stereoselectivity on MCT. Through these advancements, the present study provides stereochemical information of lipase on a wider substrate as well as long-chain TAG, which would accordingly contribute valuable insights into the utilization of MCT and lipases to produce a wide range of structured lipids in the food industry.

Biomass Conversion and Catalytic Hydrodeoxygenation of Bio-oil Model Compound

VO THE KY Kyung Hee University 2018 국내박사

Biomass-derived biofuel has attracted considerable attention as an alternative and renewable fuel owing to petroleum price instability and global warming. Conversion of biomass such as lignocellulosic biomass or microalgae by thermal decomposition considered as feasible approaches for converting biomass into energy as well as valuable chemicals. Recently, pyrolysis and hydrothermal liquefaction (HTL) have received growing attention on conversion of biomass compositions to produce biofuel. Pyrolysis is the thermochemical process, in which dry biomass will be decomposed with the absence of oxygen. While HTL converts wet biomass into gases, biocrude, aqueous –phase product and biochar through thermal and hydrolytic decompositions. Lignocellulosic biomass has been studied extensively in terms of their pyrolysis characteristics as well as bio-oil yield and bio-oil compositions. Most of these biomass were used as wild – type biomass. Recently, transgenic biomass could be considered a promising biomass for production of bio-oil and valuable chemicals when its gene structure was modified from that of the wild-type to change the biochemical components (lignin, hemicellulose, cellulose). Hybrid poplar trees, which are valuable biomass feedstocks because they can grow very fast and are good candidates for genetic improvement with regard to bioenergy feedstock production. A comparative study on pyrolysis characteristics and kinetic of the wild –type and genetically engineered hybrid poplar trees were conducted systematically to understand their thermal decomposition behaviors that were necessary before using these feedstocks. The obtained results indicated that transgenic biomass had lower activation energy and produced higher bio-oil yield compared to that of wild type under the same pyrolysis conditions. In addition to this, chemical compositions of biocrude produced from genetic modified hybrid poplars had higher carbohydrate-derivatives but lower lignin-derivatives compared to those obtained from wild –type biomass. Heterotrophic microalgae Aurantiochytrium sp. is a promising feedstock for biofuel production due to its fast growth rates and high lipid content (50 wt. % of dry biomass). Another kid of microalgae strain, Tetraselmis sp. was cultivated successfully by artificial seawater semi-permeable membrane photobioreactor promising to provide an economic and sustainable biofuel production from microalgae. Pyrolysis characteristics and kinetics of microalgae by means of thermogravimetric analysis (TGA) and pyrolysis on a micro-tubing reactor. The thermal decomposition behaviors of biochemical compositions (carbohydrates, proteins, lipids) of microalgae were investigated and compared. Free-model methods such as Kissinger-Akahira-Sunose (KAS) and Flynn-Wall-Ozawa (FWO) were applied to estimate activation energy for pyrolysis of algal biomass. These methods have widely used for the determination of activation energy since they can be used without knowledge of the reaction model. A lumped kinetic model was applied for the expertiment data. The obtained results indicated that the predominant pyrolysis reaction pathway of Aurantiochytrium sp. was from biomass to bio-oil rather than from biomass to gas, indicating that the feasibility of converting this macroalgae biomass into bio-oil by fast pyrolysis. Hydrothermal liquefaction of microalgae (Aurantiochytrium sp. and Tetraselmis sp.) was conducted at different temperature (250 ~ 400 oC) and time (10~ 90 min). Under these conditions, the biochemical compositions in microalgae cells were decomposed to produce biocrude, gas, aqueous- phase product and biochar. Biocrude with chemical compositions depend on the biomass feedstock as well as experimental conditions. A reaction network that can generally describe the hydrothermal liquefaction of each carbohydrates, proteins and lipids in the biomass cell. Besides, there exists interconversion between product phases as bio-oil and aqueous –phase, heavy –oil and light -oil were also included in the reaction network. The results showed that microalgae were rapidly decomposed for first few minutes of reaction time. With longer reaction time, the interconversion between products phases were predominant reactions. Based on this reaction network, quantitative kinetic model for HTL of microalgae, which can be useful for design, control and optimization of HTL processes, was proposed. The estimated reaction rates and activation energy suggested the dominant reaction pathways as well as the distribution of the biochemical compositions to the bio-oil phase. Kinetic parameters were used to explore the parameter space in order to predict product yields as a function of reaction time and temperature. Bio-oil obtained from pyrolysis and hydrothermal liquefaction of microalgae cannot be used directly since it is high viscosity, high acidity, and low heating value due to the presence of significant quantities of many oxygen-contain compounds such as acids, aldehydes, ketones and phenolic compounds. Therefore, upgrading the quality of biomass –derived biocrude have attracted much attention for recent years. Hydrodeoxygenation (HDO) reaction is one of the most potentially valuable processing routes to selectively cleave C – O and C – C bonds in oxygen-containing substances. In this work, a novel method combining sol-gel and spray pyrolysis was applied to synthesize Mo/Al2O3 –TiO2 catalysts for upgrading of hexadecanoic acid (palmitic acid) that was found to be a major component (ca. 50%) in the biocrude obtained from pyrolysis and HTL of microalgae Aurantiochytrium sp. During spray pyrolysis process, the spherical composite particles were formed directly from the droplets containing a well-dispersed mixture of molybdenum salt, boehmite sol and titania sol through one-step pyrolysis. The results obtained from catalytic activity studies on hydrodeoxygenation of palmitic acid showed that the Mo/Al2O3-TiO2 catalysts exhibited excellent catalytic performance as high HDO conversion (100%) and high saturated hydrocarbon selectivity (93.18%). These results were much better than those of catalyst derived from conventional impregnation method. Effects of TiO2 concentration used to modify γ-Al2O3 on the catalytic activity was systematically investigated. Reusability experiment results showed that there were a slight decrease in metal/metal oxides concentration ratio of reduced catalyst after four time uses.  

팜유의 촉매수소화반응에 관한 반응속도론적 연구

김진홍 공주대학교 테크노전략대학원 2015 국내석사

수소화 반응 촉매를 개발하여 국내의 촉매 시장의 다양화시키며 촉매의 해석방법을 연구함으로써, 국내 수소화 촉매분야의 노하우를 축적하고 세계 촉매 시장에 있어 국가 경쟁력에 크게 기여 하고, 수소화 촉매 제조의 국산화를 이루기 위해 개발한 촉매를 이용하여 반응속도론적 연구를 통한 촉매의 반응 특성을 연구하였다. 오일의 수소화공정에서 운전 변수인 수소 공급 압력, 반응온도를 조절하여 오일내의 각 지방산 성분변화를 얻고, 이를 통한 Kinetic Data를 Pow-law Kinetic Model, Langmuir-Hinshelwood Kinetic Model을 이용하여 비교를 통해, 본 연구에서 개발된 촉매의 수소화 반응에서의 모델식을 선정하였다. 또, 수소화 반응에 있어 공정변수에 따라 요오드값의 변화를 측정하고 이 또한 Kinetic Model을 이용하여 해석하였다. 반응속도 모델 선정은 Power-Law kinetic model 식이 반응실험 결과를 잘 표현했다. Langmuir-Hinshelwood kinetic model 식에서 Linoleic acid의 수소화반응속도에서 반응물과 생성물 그리고 수소가 촉매 활성점에 흡착하는 경우, 반응물과 생성물이 촉매 활성점에 흡착하는 경우, 반응물과 수소가 촉매 활성점에 흡착하는 경우모두를 만족 하였고, Oleic acid의 수소화 반응속도의 경우 반응물과 수소가 촉매 활성점에 흡착하는 경우만을 만족하였다. 수소화 반응속도는 수소의 공급 압력 보다 반응온도의 영향이 더 크다. 요오드가 변화 모델은 수소 공급 압력, 수소화반응온도에 대해 잘 표현했다.

On the Neural Network approach to the Kinetic Model for the Semiconductor Thin Film Deposition

선봉석 포항공과대학교 일반대학원 2025 국내석사

This Master’s thesis addresses the simulation of the kinetic model, which is used to describe the semiconductor thin film deposition. In this work, we consider a ki- netic model, with absorbing, specular reflection, and inflow boundaries. We adopt a thermal Atomic Layer Deposition (ALD) method that does not consider chemical re- actions. Our focus is on the precursor flow during various processes of thermal ALD. Using deep learning algorithms, we derive a Deep Neural Network (DNN) solution for the kinetic model. Through this approach, we observe the behavior of particles and investigate the associated macroscopic physical quantities.

Development of kinetic models for soybean (Glycine Max) and soybean powders in the grinding and storage : 대두와 대두 분말의 분쇄, 저장 공정에서의 kinetic model의 개발

Lee, Youn Ju 강원대학교 대학원 2014 국내석사

Grinding kinetics of soybeans at different moisture content (6%, 8% and 12%) were investigated. The hardness of soybean particles increased and showed brittle characters as the moisture content decreased. Three theoretical models for grinding, such as the Rittinger, Kick, and Bond model, were applied to characterize the grinding process of soybeans. The lower moisture content showed less grinding constants including Bond’s work index. Sigmoid model was successfully applied to describe the changes in particle size of soybeans at different moisture contents during grinding (R2 > 0.96). The TBARS (thiobarbituric acid reactive substances) at different size of soybean flours were measured at 25 °C and 50 °C for 20 days and the yield of oil composition at different size of soybean flours were measured. As the particle sizes decreased, the TBARS values increased during storage, while the oil yield from soybean flours increased. Effects of particle size and heating time during TBA test on the TBARS of soybean (Glycine Max) powder were studied. Effects of processing variables involved in the pulverization of soybean, such as the temperature of soybean powder, the oxygen level in the vessel, and the pulverization time, were investigated. The temperature of the soybean powder and the oxygen level had no significant influence on the TBARS (p < 0.05). The pulverization time and the heating time during TBA test significantly affected the TBARS. Change of TBARS during heating was well described by the fractional conversion first order kinetics model. A diffusion model was introduced to quantify the effect of particle size on TBARS. The major finding of this study was that the TBA test to estimate the level of the lipid oxidation directly from powders should consider the heating time and the mean particle sizes of the sample. 대두의 수분함량에 따른 분쇄 kinetic 은 수분함량을 6, 8, 12%로 조절 하여 특성을 비교하였다. 대두의 수분함량이 감소할수록 경도가 증가하였고 대두 낟알이 더 깨지기 쉬운 특성을 나타낼 수 있음을 확인하였다. 3가지 분쇄 이론 (Rittinger, Kick, Bond’s model)의 모델은 대두의 분쇄 공정의 특징을 설명하는데 적용되었다. 수분함량이 적을수록 각 모델의 상수들은 감소하는 결과를 나타내었다. 이와 같은 결과로 대두의 수분함량이 증가할수록 분쇄 시 더 많은 분쇄 에너지를 요구함을 확인하였다. Sigmoid 모델로 해석한 분쇄 kinetic modeling은 대두의 수분함량에 따른 분쇄물의 양을 예측하는데 적절하게 사용되었다. 수분함량이 적을수록 동일한 분쇄 시간 동안 분쇄하였을 때 더 많은 분쇄물을 얻을 수 있음을 확인하였다. 입자 크기가 산패나 추출 공정에 주는 영향을 확인하기 위하여 다른 3가지 대두 분말을 25 ℃, 50 ℃ 에서 20일간 저장하면서 산패도 변화를 TBARS로 측정하였고 유지를 추출하여 수율을 확인하였다. 입자크기가 감소할수록 저장기간중의 TBARS는 증가하였고 반면에 대두 분말로부터 추출한 유지의 수율은 증가하였다. 대두 분말에 대한 Thiobarbituric acid (TBA) test 에서 입자크기와 가온 시간의 영향에 대한 연구를 수행하였다. 분쇄하는 동안 분쇄기 내부의 온도 변화와 산소 농도의 변화, 분쇄 시간에 대한 효과 또한 확인하였다. 분쇄기 내부의 온도와 산소농도의 변화는 대두 분말의 산패도 지표인 TBARS에 유의미한 영향을 나타내지 않았다. 분쇄 시간과 TBA test를 수행하는 동안의 가온 시간은 TBARS에 유의미한 영향을 주었다. 가온 시간에 따른 TBARS의 변화는 fractional conversion 1차 kinetic 모델로 적절하게 설명되었다. 확산 모델은 대두 분말의 입자크기가 TBARS에 대해 나타내는 효과를 양적으로 나타내는데 적절하였다. TBA test를 수행할 때 입자크기와 가온 시간을 고려한 직접적인 산패도 측정방법으로 적절할 것으로 생각된다. Kinetic modeling은 화학반응뿐만 아니라 분쇄와 같은 단위 공정에서도 대두의 특성에 따른 영향을 설명하기에 적절한 툴로서 유용하게 사용될 수 있을 것이다.

활성슬러지 공정에 관한 Kinetic model의 비교

강상훈 漢陽大學校 大學院 1983 국내석사

Stability Prediction and Control of Anaerobic Digestion Process Based on Artificial Intelligence

JIA RU 국립한국해양대학교 대학원 2025 국내박사

This study investigated the process stability of anaerobic digestion and enhanced intelligent control through applications of machine learning. One stability indicator characterized the dynamic balance of anaerobic biochemical reactions by introducing Recovery Potential (RP) and Deterioration Potential (DP). Another stability indicator diagnosed the state of anaerobic digestion through a comprehensive indicator derived using Principal Component Analysis (PCA). A deep learning model combining a Convolutional Neural Network (CNN) and Bidirectional Long Short-Term Memory (BdLSTM) utilized real-time sensor data to provide insights into system state and performance, ensuring economical and stable digester operation. Finally, intelligent control of anaerobic digestion was implemented using a Deep Q-Network (DQN) reinforcement learning model, integrating stability indicators with the CNN-BdLSTM network. RP and DP were formulated to shed light on the kinetic balance between anaerobic biochemical reactions. RP is gauged by the ratio of the methanogenesis rate (MR) to the acidogenesis rate (AR), while the DP is the sum of the accumulation rate (AcR) and dilution rate (DR) of total VFAs, normalized using the AR. In an anaerobic digester for a mixture of pulverized food waste and liquified sewage sludge, an RP above 1.0 signifies a restorative state in the kinetic balance of anaerobic biochemical reactions across various operational phases, including startup and steady state, and shifts in organic loading rate. Conversely, a DP value of 0.0 or higher denotes a deterioration in the kinetic balance. The instability index (ISI), calculated as the DP to RP ratio, serves as an indicator of an anaerobic digestion state. When the standard deviation of ISI surpasses 0.2, it signifies instability in biochemical reactions; however, an average ISI below 0.05 indicates a stable digestion process. The study underscores the efficacy of RP, DP, and ISI as robust indicators for assessing the stability of anaerobic digestion based on the kinetics of biochemical reactions. A comprehensive indicator based on PCA has been proposed for diagnosing the state of anaerobic digestion. Various state and performance variables were monitored under different operational modes, including start-up, interruption and resumption of substrate supply, and impulse organic loading rates. While these individual variables are useful for estimating the state of anaerobic digestion, they must be interpreted by experts. Coupled indicators combine these variables with the effect of offering more detailed insights, but they are limited in their universal applicability. Time-series eigenvalues reflected the anaerobic digestion process occurring in response to operational changes: Stable states were identified by eigenvalue peaks below 1.0, and they had an average below 0.2. Slightly perturbed states were identified by a consistent decrease in eigenvalue peaks from a value of below 4.0 or by observing isolated peaks below 3.0. Disturbed states were identified by repeated eigenvalue peaks over 3.0, and they had an average above 0.6. The long-term persistence of these peaks signals an increasing kinetic imbalance, which could lead to process failure. Ultimately, this study demonstrates that time-series eigenvalue analysis is an effective comprehensive indicator for identifying kinetic imbalances in anaerobic digestion. The immediate response to the state disturbances of anaerobic digestion is essential to prevent anaerobic digestion failure. However, frequent monitoring of the state and performance of anaerobic digestion is challenging. Thus, deep learning models were investigated to predict the state and performance variables from online sensor data. The online sensor data, including pH, electric conductivity, and oxidation-reduction potential, were used as the input features to build deep learning models. The state and performance data measured offline were used as the labels. The model performance was compared for several deep learning models of CNN, LSTM, dense layer, and their combinations. The combined model of CNN and BdLSTM was robust and well-generalized in predicting the state and performance variables (R2=0.978, root mean square error=0.031). The combined model is an excellent soft sensor for monitoring the state and performance of anaerobic digestion from electrochemical sensors. Reinforcement learning (RL) based on a deep Q-network (DQN) was studied to enable intelligent control of anaerobic digestion processes. Anaerobic digesters operated under statistically designed organic loading rate (OLR) conditions provided sensor data on process states and performance. Variable importance analysis identified key RL components—pH, EC, and ORP as states; OLR (flow rate and COD) as actions; and total reward combining stability and methane production. A deep learning-based environment model was trained to simulate process dynamics, predicting the next states and total reward based on the current states and actions. The architecture of the DQN with ε-greedy and prioritized experience replay was optimized by interacting with the environment model. Offline training effectively pre-trained model parameters, enhancing initial learning performance. The pre-trained DDQN was activated above a total reward threshold, stabilizing process instability and improving methane production under variable OLR conditions. The dueling DDQN (TDQN) showed slower pre-training but rapidly adapted to variability, stabilizing the process and significantly improving methane production. Both pre-trained DDQN and TDQN provide intelligent control frameworks for optimizing anaerobic digestion performance under variable OLR conditions.

The determination of pesticide residues in Yuza (Citrus Junos) and Yuza tea and its application to kinetic model for pesticide degradation

남영성 동국대학교 2011 국내석사

유자 ( Citrus Junos Sieb. ex Tanaka)는 분류학상으로 운향과이며, 감귤에 속하는 과실로서 쓴맛 성분인 리모노이드는 항암효과가 기대되며, 신맛을 주는 구연산, 유기산, 칼슘 및 칼륨 등 미네랄도 풍부하여 피로회복, 감기예방 및 피부미용에 효과가 있다. 국내 주요 산지로는 전라남도 고흥, 완도와 경상남도 거제, 남해 등이다. 한국작물보호협회(Korea Crop Protection Association, Pesticide Handbook 2009)는 유자에 사용되는 8종의 농약을 제시하고 있다. 따라서 본 연구에서는 첫째, 식품공전(Korea Food Code) (Korean Food & Drug Administration) 을 바탕으로 최적의 잔 류 농약 분석 실험 방법을 확립하여 2009~2010년에 생산된 고 흥산 유자 및 유자차에 대해 잔류농약분석 모니터링을 실시하였 다. 분석방법 검증을 위하여 외부표준물질(External Standard) 사용하여 표준곡선을 작성하였으며, 회수율, 일내정밀성(Intraday) 및 일간정밀성(Inter-day)를 수행하였다. 둘째, 잔류농약분석에 가장 적합한 Kinetic Model인 First order Kinetic 을 바탕으로 농약이 살포된 유자로 만든 유자차의 반감 기를 계산하였다. 그리고 MRLs와 비교함으로써 생산단계 잔류 허용기준을 설정하였다. 이는 수확 후 또는 유통 과정 중 부적합 품목을 사전에 차단할 수 있는 기준으로 사용될 것이다. In this study, the rapid analytical method of 8- pesticides residues in Yuza was developed in our laboratory. 8 different pesticides (chlorpyrifos, prothiofos, phosalone, deltamethrin, acequinocyl, spirodiclofen, benomyl and thiophanate methyl) were analyzed in 235 samples (75 Yuza tea, 100 Yuza cultivated by ordinary culture and 60 Yuza cultivated by environmentally-friendly culture) taken from HANSUNG FOOD CO., LTD (Goheung-gun, Jeollanam-do, Korea) during 2009-2010. chlorpyrifos, prothiofos, phosalone and deltamethrin in 8-pesticides were measured LODs(limits of detection), LOQs(limits of quantifications) and recovery, using gas chromatography (GC) with nitrogen-phosphorus detection (NPD) and gas chromatography (GC) with mass spectrometric detection (MSD). GC/NPD column was selected DB-5. GC/MS column was selected DB-5MS. the sensitivity of the MSD further improve and analysis time was shortened. Pesticide analysis by GC was performed Simultaneous analytical method. The rest (acequinocyl, spirodiclofen, benomyl and thiophanate methyl) was analyzed by high performance liquid chromatography (HPLC/PDA). Sample preparation was optimized for high-resolution of each 4-pesticide using HPLC analysis. Analytical separation was individually carried out using a gradient elution with acetonitrile or methanol on C18 column. Also, It could improve the accuracy and save time of the test by confirmed experiment for detecting materials. As expected, the value for Yuza cultivated by ordinary culture seemed to be rather high than Yuza tea and Yuza cultivated by environmentally-friendly culture in 2009 and 2010. The optimized analytical method was utilized to study the decline of the pesticides residues. The kinetics was performed to evaluate the decline of 9-pesticide in Yuza tea. Ten days after spraying, the degradation rates of chlorpyrifos and acequnocyl exceeded 90%, and the acequnocyl and chlorpyrifos being 100% and 90% respectively. The half-lives of the nine pesticides in Yuza were 3.2, 20.4, 25.7, 11.7, 1.18, 2.92 and 5.29 days, respectively. If the Yuza was sprayed out at recommend dosage of 9-pesticides, all pesticides maximum residue limits (MPLs) were not exceeded in Yuza tea made with the sprayed Yuza.