Investigation of process stream systems for targeting energy-capital trade-offs of a heat recovery network

Abstract

The process industries consume a huge amount of heat energy contributing to environmental impact. Energy recovery is a key instrument of energy-saving that can be implemented via a heat exchanger network. Pinch-based approaches presume analysis of energy targets of industrial processes to find the optimal ΔTmin for heat exchanger network design. Classical Pinch Analysis do not account for the stream splitting in the super targeting procedure while parallel branches are usually needed. The splitters and mixers contribute a lot to the capital cost of the heat exchanger network. Current work proposes the update of super targeting procedure accounting stream splitting and mixing in a Pinch problem. The original algorithm of Composite Curves construction is proposed to analyse the distribution of process streams and stream splitting in subsystems above/below the Pinch before the design of the heat exchanger network. It was then used in a super targeting procedure to precise the capital cost and, as a result, the optimal ΔTmin. The process stream distribution and stream splitting are analysed in a whole range of ΔTmin. Identifying all possible starting points for heat exchanger network design. Two new criteria were proposed to estimate the topological complexity of network pre-design and the specific ratio of stream splitting. The case study analyses the ethylene oxide process and calculation of trade-off between capital and energy costs were performed and optimal ΔTmin = 16 °C. The result was compared with two known approaches, which account for the number of heat exchangers without stream splitting. Total annual costs and optimal ΔTmin was also calculated for different energy prices to show a possible deviation of starting point for heat exchanger network design. The range of optimal ΔTmin from 9 to 54 °C resulted in the range of hot utility prices from 42 to 291 $/kWy, and the emission targets will be from 14,380 to 77,919 tCO2/y. The methodology can be used for the pre-design of the heat exchanger network to better precise the optimal ΔTmin, capital cost targets, and check the optimum changing for different energy prices.