Separation of gaseous mixtures produced from any process plant can be a major issue for the industry. In comparison to other separation techniques available, hydrate formation can be an appropriate choice in terms of fuel efficiency and degree of separation. In this work, a three-stage process was designed to capture CO₂ from a hypothetical flue gas mixture (comprising 15 mol% CO₂+ 85 mol% N₂). For the hydrate based CO₂ capture process, a detailed analysis was carried out covering the material and energy balance, energy integration and economic analysis. The results show that, at steady state, 91% of the CO₂ in the flue gas can be recovered, resulting into two streams: Stream-A (94 mol% CO₂ + 6 mol% N₂) and Stream-B (1.5 mol% CO₂ + 98.5 mol% N₂). Per 100 mol of the flue gas feed, 3.9 MW of energy is required to achieve such degree of separation. This amount was reduced to 2.4 MW after energy integration was applied to the process. Overall, the process can help in reducing CO₂ emission via flue gas because of the huge difference in the carbon foot print of its energy requirement (500 g CO₂ per kWh) and the amount of CO₂ captured by the process (2190 g CO₂ per kWh). GRAPHICAL ABSTRACT

Accurate vapor-liquid equilibria (VLE) calculations of carbon dioxide and hydrogen sulfide mixtures are critical to gas processing and the affordable, safe design of flow assurance technologies. Inaccurate VLE predictions can lead to inaccurate gas hydrate phase equilibria predictions and ensuing safety and economic risks. This research paper explores the potential incorporation of Deep Neural Networks (DNNs) to support conventional expert systems within the context of predicting VLE. It facilitates more flexible and data-driven approaches that are required due to the growing intricacy and dynamic character of chemical processes. Moreover, various cubic and non-cubic equation of state (EoS) models (such as SRK, PR, CPA, SAFT, and PC-SAFT) were also examined to compare predicted VLE for various mixtures of CO2 and H2S. Prior to the comparison of DNN-predicted VLE with EOS models, binary interaction parameters were optimized for all EOS with the available experimental phase equilibria measurements. Model accuracies were compared and analyzed for various binary systems containing CO2/H2S + other associative and non-associative components. The absolute average deviation in vapor and liquid phase composition/bubble pressure was calculated and compared for all five-state EOS with DNN predictions. The DNN and equation of states with BIP gave a reliable illustration of the phase behavior of CO2/H2S-containing systems compared to others as indicated by the lower AADP values. By contrasting the applied DNN model with conventional techniques, we explore the promising channel for future research directions and industry applications, as well as an opportunity for innovation and field advancement for modern expert systems.

In this study, efficient techniques are utilized to design the multivariate Hotelling control chart with sensitizing rules for detecting small-to-moderate variations. The control limit of the proposed chart is derived relative to probability of a single point and number of process characteristics. To calculate probability of a single point for sustained in-control average run length, a generalized single polynomial equation is derived. For evaluation, performance measures are considered based on the average, the median, and the percentile run length. These measures are calculated using Monte Carlo simulation and numerical integration. The results indicate that the proposed control chart has consistent behavior when a process is in control. The in-control average run length is obtained equal to prefixed level which remains valid for all choices of sensitizing rules. This implies that the proposed control chart can resolve the issue of existing control chart in terms of sustained behavior. The effectiveness of sensitizing rules is dependent on process characteristics and variations of mean vector. A comparative analysis of different choices of sensitizing rules is conducted to locate optimal choices of process characteristics. Real-life example, dowel-pin manufacturing, shows that proposed control chart with sensitizing rules is efficient for diagnosing small variations.