Kyungjae Tak

Optimization of mixed-refrigerant system in LNG liquefaction process

Abstract LNG liquefaction is an energy intensive process. For this reason, various liquefaction processes for saving energy exist. Searching optimal process condition is very difficult because of its high nonlinearity. Pressure level, refrigerant flowrate, and refrigerant composition are key operation variables to reduce energy consumption. These variables play an important role in affecting the overall performance. Compressor is a major energy-consumption unit in LNG plant and it spends much energy depending on operating conditions. In this paper, searching optimal condition is carried out by using simultaneous optimizations. NLP model is applied for the SMR process. The optimization result shows that refrigerant composition is a major key variable and half of energy consumption can be reduced by changing operating conditions and refrigerant composition only.

Multi-scale/multi-physical modeling in head/disk interface of magnetic data storage

The model integration of the head-disk interface (HDI) in the hard disk drive system, which includes the hierarchy of highly interactive layers (magnetic layer, carbon overcoat (COC), lubricant, and air bearing system (ABS)), has recently been focused upon to resolve technical barriers and enhance reliability. Heat-assisted magnetic recording especially demands that the model simultaneously incorporates thermal and mechanical phenomena by considering the enormous combinatorial cases of materials and multi-scale/multi-physical phenomena. In this paper, we explore multi-scale/multi-physical simulation methods for HDI, which will holistically integrate magnetic layers, COC, lubricants, and ABS in non-isothermal conditions.

Kriging models for forecasting crude unit overhead corrosion

Crude unit overhead corrosion is a major issue in the refinery field. However, the corrosion models in the literature are difficult to apply to real refinery processes due to the characteristics of corrosion. We propose a Kriging model, an advanced statistical tool for geostatistics, to forecast the corrosion rate in a real refinery plant. Instead of spatial coordinates, the proposed model employs the non-spatial coordinates of six key corrosion variables: H2S, Cl−, Fe2+, NH3, pH, and flowrate. The Kriging model is compared with two well-known forecasting models, multiple linear regression and an artificial neural network. To overcome the insufficiency of the number of data sets measured in the plant to use the six non-spatial coordinates, the significance probability is applied to reduce the dimensions from six to four. Among all the developed models in this paper, the Kriging model with four corrosion variables showed the best forecasting performance.

Optimization of Petrochemical Process Planning using Naphtha Price Forecasting and Process Modeling

Abstract A naphtha price forecasting model based on the time series method is developed to predict the monthly variation of naphtha price using statistics. We used the model to forecast future naphtha prices by looking at historical time series data from January 2008 to September 2011. After forecasting, we perform the normalization of the observed period and implement a simulation using the price forecasting model. In order to check the accuracy of the model, the predicted naphtha price variations are compared with the actual naphtha variation. If the predicted variation of the normalized naphtha price has the same trend as the actual naphtha variation, our prediction value is called “T.” Otherwise, the predicted value is called “F.” The accuracy of the predicted value is relatively higher than the price forecasting for other products such as crude oil. As a result, our model is useful to industry decision-makers for forecasting the price of naphtha.

Optimization of Pure-Refrigerant Cycle Compressing Ratio on C3-MR Process

Abstract The natural gas liquefaction process is an energy intensive process due to its low temperature. Compressor units are one of primary factors in minimizing energy consumption of natural gas liquefaction process. Among various natural gas liquefaction processes, C3-MR (Propane Pre-cooled Mixed-Refrigerant) process is the most commonly used in LNG market. C3-MR process uses two types of refrigerants such as mixed-refrigerant and pure-refrigerant. In this study, pure-refrigerant cycle is simulated, along with different compressing ratios and different pressure levels. The simulation of case studies shows that energy consumptions mainly depend on both compressing ratios and pressure levels.. This study has achieved the energy consumption savings by 27.7% through different case studies with an emphasis on compressing ratios and pressure levels.

Corrosion effect on inspection and replacement planning for a refinery plant

Abstract This paper presents an optimization model of inspection and replacement planning for a refinery plant under the consideration of corrosion in terms of cost. The management of corrosion is an essential task for processes that operate over several years without a shutdown. This is because corrosion can cause severe failures by thinning the wall thickness and eventually cause pipes or equipment to burst. However, required safety measures, such as the corrosion management, involve costly inspection and replacement. Therefore, a cost-effective safety-action strategy is proposed in this paper. The developed model presents an optimal combination of steel grade, design wall thickness, inspection number, and inspection timing under a given corrosion rate to minimize the cost of design, inspection, replacement, and failure. Three case studies using sensitivity analyses are applied to three major processes in a refinery plant: a crude distillation unit, visbreaker, and hydrocracker.

A multistream heat exchanger model with enthalpy feasibility

Abstract A temperature feasibility constraint is an important part of multistream heat exchanger (MSHE) modeling. However, temperature feasibility of an MSHE model makes a numerical issue when a physical property package is used to obtain highly accurate temperature-enthalpy relationships in equation-oriented modeling environment. To resolve the issue, this study proposes a new MSHE model with enthalpy feasibility using the fact that enthalpy is a monotonically increasing function of temperature. A natural gas liquefaction process, called a single mixed refrigeration process, is optimized using the proposed MSHE model under an equation-oriented modeling environment with a physical property package as a case study.

Analysis of air blast effect for explosives in a large scale detonation

Open Burning/Open Detonation (OB/OD) has been widely used for demilitarization of expired explosives. However, OB/OD effects a variety of hazardous damages to environment. Therefore, using incinerators to treat expired explosives is required instead of OB/OD. To guarantee the safety of these demilitarization methods, the blast wave of the explosives should be previously recognized to evaluate the impact of detonations. Although various materials are used to produce explosives, most researches have focused on trinitrotoluene (TNT). Other representative explosives such as research department explosives (RDX) and high melting explosives (HMX) are seldom studied in the literature. Therefore, our aim was to understand the blast wave of three materials under different geometry throughout simulations. To improve accuracy and reduce computational time, a zoning technique with Euler-Lagrange coupling method was used. Due to limitations and difficulties of detonation experiments, simulations were verified by theoretical models. In case of semi-confined bunker, the simulation results were compared with experimental data, showing a close match. As a result, cylinder type is the safest incinerator among semi-confined bunker, cylinder, and cube incinerators, in terms of the blast wave.

Web-based multi-dimensional education system for the simulated moving bed process

This paper describes the development of an operator-education program for a simulated moving bed (SMB) pilot plant that separates para-xylene isomer from mixed xylenes. The education program was motivated from reliable and safe operations for a newly developed SMB technology, in particular, a priori education on preventing accidents such as explosion, fire and toxic material release, most of which result from the lack of understanding of the processes being operated. This SMB process education program has overall SMB process description, sixteen operating screen and response procedure for accident of leak and explosion. The proposed system was implemented based on client/server architecture in company-wise LAN environment and applied previously operator heuristics. In addition, it features multi-dimensional animation for virtual reality of process flow, using multi-media techniques. Operators are expected to benefit from education that is free of time and space constraints. The education system for the pilot plant will contribute to keeping the full scale SMB operation workforce performing at a high level of proficiency as well as refreshing the operational skills of senior operators.

Current status of optimal design of natural gas liquefaction process

Abstract Natural gas liquefaction process is an energy intensive process due to its cryogenic condition. Therefore, one of the major objectives for the design and the optimization is the minimizing total energy consumption. Another important objective is the minimizing cost of the energy supply system. This research focused on the cost based optimization. First, the energy minimization for the mixed and the pure refrigerant systems was performed by deterministic optimization model. Then, the cost minimization for energy supply to the process was performed by the driver selection model. The driver selection model was built as mixed integer linear programming (MILP). As the result, 15 to 17% of energy saving was checked for the refrigerant system. At that time, optimal driver set which is the minimum cost for the energy supply system was found.