Technological aspects of computer control of the secondary condensation complex in ammonia production under uncertainty

Abstract

The object of the study is the technological complex of secondary condensation of large-tonnage ammonia synthesis units of the AM-1360 series, which provides final cooling and separation of condensed production ammonia from circulating gas. The possibility of increasing the energy efficiency of production by modernising the equipment and technological design of the TCC has been established. This is achieved by removing the energy-intensive turbocompressor refrigeration unit (TRU) with electric drive from the circuit and creating an adaptive system of optimal software control. The feasibility of applying a systematic approach to solving such a complex problem, based on mathematical modelling and process identification, has been demonstrated. The operating conditions of the TCSC and the primary condensation unit have been analysed. The results of the research have established uncertainties in the functioning of such components as the condensation column (CC) and low-temperature evaporators (LTE), which are connected to the operating circuit of two absorption refrigeration units (ARU) and TRU. Algorithms have been developed for forming an information array of experimental data and numerical assessment of uncertainties, in particular heat transfer coefficients in the CC and LTE, as well as ammonia concentration at the outlet of the primary condensation unit and CC. The algorithms provide for the separation of transient modes in the operation of the TCSC, verification of stationarity, reproducibility of the process and the hypothesis of normality of empirical distribution, which determines the possibility of using a stochastic approximation method for numerical estimation of uncertainties. Based on the results of processing the experimental data, a discrepancy between the actual and design heat transfer coefficients was established, which is due to an underestimation of the condensation thermal resistance. The heat exchange processes in the CC and LTE as part of the ARU were identified, and equations were obtained for determining the heat transfer coefficients, heat transfer, condensation thermal resistance, and ammonia concentration in the CG at the CC inlet and outlet. Mathematical modelling was used to determine the conditions for the necessary temperature distribution in the TCSC to exclude the TRU from the circuit and reduce the cooling temperature of the CG in the LTE by 5°C compared to the initial version at maximum heat load from the CG at the inlet of the complex. The developed TCSC scheme is characterised by greater energy efficiency due to the use of only heat-utilising refrigeration systems of the ARU and steam ejector units (SEU), which utilise the heat of material flows with both low temperature potential (up to 150°C) and ultra-low (up to 90°C). Using mathematical modelling of LTE, a pattern of extreme dependence of cooling capacity and cooling temperature of the central heating system on the phlegm flow rate has been established. Achieving maximum cooling capacity, and therefore minimum cooling temperature of the central heating system at a certain temperature head, is determined by the critical regime of bubble boiling of the refrigerant. The dependencies of the cooling temperature of the central heating system on the control action of the phlegm flow rate have been determined, which characterise the shift of the extreme under conditions of changing values of the disturbance vector coordinates, and, consequently, the change in the energy efficiency indicators of ammonia production (annual natural gas consumption). Algorithmic support has been developed to solve the problems of identification, obtaining a mathematical model of the LTE evaporator and numerical estimation of the optimal state vector (cooling temperature of the central heating system). The use of the algorithm implemented in the MATLAB package provides a solution to the optimisation problem in real time using a step-type non-gradient method with the application of one-dimensional extremum search methods. The technical structure of a computer-integrated system for optimal software control of the temperature regime of a low-temperature evaporator, adapted to the existing information system of an industrial synthesis unit, has been determined.