The phase equilibria of the KBr–SrBr₂–H₂O and NaBr–SrBr₂–H₂O ternary systems at 308 K were measured using the isothermal dissolution equilibrium method. The compositions of the liquid and solid phases were expressed as the mass fraction, and the phase diagrams constructed from the experimental data show that both ternary systems are simple co-saturated types with only one invariant point, two univariant curves, and two crystallization regions. Neither of the two ternary systems forms a double-salt or solid solution. The solid-phase crystallization regions of the KBr–SrBr₂–H₂O ternary system correspond to KBr and SrBr₂·6H₂O, and the invariant point of the NaBr–SrBr₂–H₂O ternary system is saturated with NaBr·2H₂O and SrBr₂·6H₂O. From the phase diagrams, the salting-out effect of SrBr₂ on KBr is clear, which indicates that the solubility of KBr is lower and it is easier to separate KBr from solution in the ternary KBr–SrBr₂–H₂O system at 308 K.

To streamline the salt manufacturing process, a novel method is required to recover and upgrade Ca and Mg from the concentrated seawater discharged by salt manufacturers in Japan. The synthesis of carbonate by reactive crystallization between the dissolved Ca²⁺ and Mg²⁺ in concentrated seawater and CO₂, which is based on the relative solubility of the salts, is an effective separation/recovery method. Dolomite (CaMg(CO₃)₂), which is a double salt, is used in various industrial fields. To improve the functionality of CaMg(CO₃)₂, it is necessary to achieve a high Mg/Ca ratio. Generally, high concentrations of Ca²⁺, Mg²⁺, and CO₃²⁻ are necessary for the synthesis of CaMg(CO₃)₂ with a high Mg/Ca ratio. Therefore, the fine bubble formation technique, which generates regions of high ion concentrations near the minute gas–liquid interfaces of the bubbles, was applied in the reactive crystallization of CaMg(CO₃)₂. In this study, we evaluated several concentrated seawater samples to determine the most suitable one for the synthesis of CaMg(CO₃)₂ with a high Mg/Ca ratio during the reactive crystallization with CO₂ fine bubble injection. Three concentrated seawater samples: ED brine, concentrated brine, and removed-K brine, were used. These samples have different concentration products of Ca²⁺ and Mg²⁺ ([Ca²⁺][Mg²⁺]), as well as various concentrations of coexisting ions, such as Na⁺, K⁺, Cl⁻, Br⁻, and SO₄²⁻. When the removed-K brine, which had a high [Ca²⁺][Mg²⁺], was subjected to a high CO₂ flow rate, CaMg(CO₃)₂ with a higher Mg/Ca ratio was produced in a shorter crystallization time. Moreover, the crystallization phenomena of CaMg(CO₃)₂ in the concentrated seawater samples were compared with those of the CaCl₂/MgCl₂ mixed solutions, and it was found that the effects of the [Ca²⁺][Mg²⁺] on the production of CaMg(CO₃)₂ with a high Mg/Ca ratio were more significant than those of the coexisting ions.

Steam drum water level is recognized as a key factor associated with industrial boiler operations. Therefore, it is important to predict the changes in the steam drum water level to ensure safe operation of the boilers. Due to the complexity of the boiler processes, multiple correlated process variables are found to be responsible for the variations in the steam drum water level. In this study, the key variables that significantly influence the changes in the steam drum water level were successfully extracted by combining the approaches of kernel principal component analysis and maximal information coefficient. Subsequently, a long-term prediction method for the steam drum water level based on gated recurrent unit network was created. To verify the effectiveness, this method was applied to actual historical operating data for steam cylinders. In comparison to long short-term memory neural networks, traditional recurrent neural networks, convolutional neural networks, and deep neural networks, the proposed approach was proved to be more effective.

Continuous manufacturing including process integration is considered to improve the process economy of protein drug production. An operation method known as flow-through chromatography (FTC) is expected to be an efficient method for separating a target protein as a pseudo continuous method. In FTC, a target protein is eluted from the chromatography column without adsorption whereas contaminants are strongly bound. In this paper, we developed a simple and fast method for determining the sample feed volume for FTC processes by ion-exchange chromatography. The method uses the distribution coefficient as a function of mobile phase salt concentration, which can be determined by linear gradient elution experiments. The proposed method was verified by a model separation system (removal of dimer from BSA monomer by anion exchange chromatography).

Continuous or integrated downstream processing is expected to increase the productivity of protein drug production. However, it is not easy to design and operate continuous downstream processing as many batch chromatography unit operations are involved. An operation method known as flow-through chromatography (FTC) is an efficient method for separating a target protein. In FTC, a target protein is recovered from the chromatography column without adsorption whereas contaminants are tightly bound. Although the operation is simple, choosing the right mobile phase, column dimension and operating conditions is not easy as the separation is quite sensitive to mobile phase salt concentration and pH as well as operating conditions. In this study, we developed an optimization method for FTC. The model system was removal of protein dimer from the monomer by anion exchange chromatography. It was shown that by choosing the right mobile phase salt concentration and the flow-velocity (residence time) similar high productivities can be obtained at pH 6.0, 7.0 and 8.0. It was also found that FTC processes are quite sensitive to the mobile phase salt concentration.

The safe distance measurement of stacking in hazardous chemicals warehouses is an important factor to ensure the safety of hazardous chemicals storage. Binocular vision distance measurement is a very promising technology. Aiming at the key problems in that feature points matching in distance measurement is easily affected by illumination and feature similarity, and based on the degree of density, the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) clustering algorithm is proposed to realize Scale-invariant feature transform (SIFT) matching algorithm of variable thresholds in different regions of the image, and then adaptive neighborhood and bidirectional matching algorithms are used to further eliminate the mismatch points. The experimental results show that the matching rate of binocular distance measurement in hazardous chemicals warehouse is 68% and the accuracy of feature points is 98%.