Breakdown of primary specific energy consumption (SEC) for ethylene... | Download Table

Industry is responsible for 40\% of global greenhouse gas (GHG) emissions, which drive anthropogenic climate change. It is also an important consumer of energy and materials, accounting for 40\% of final energy consumption. While other sectors such as power generation, heat and transport are expected to reduce emissions in the near future, industry is often harder to decarbonise. Efficiency in the use of both materials and energy can be a powerful tool for addressing a wide range of industrial emissions mechanisms, but these approaches are often pursued separately. Exergy analysis is a powerful assessment tool that can combine material and energy efficiency approaches into one, while also accounting for resource quality as well as quantity. This thesis investigates how exergy analysis can be applied in industrial sites to achieve real resource and emissions savings. Production of ammonia and syngas through the steam methane reforming (SMR) process are chosen as the case study, given the high production volume, resource consumption and emissions arising from this industry. This thesis develops a comprehensive exergy analysis methodology and applies it to a detailed ammonia plant simulation at various level from the site to the equipment-level. Specific modelling choices are made, to allow for disaggregation of exergy inefficiencies according to the mechanisms they are caused by. The approach considers multiple aspects such as exergy calculations, flow visualisation, efficiency definitions and other metrics. In addition, network theory is used to assess the ammonia site using network analysis metrics, taking into account the complex nature of the site. The results show that SMR is the main source of inefficiency with an exergy efficiency of 68\% and exergy destruction of 165 MW, followed by the ammonia synthesis and water gas shift plants. The SMR plant is then investigated in detail. The two main combustors and two heat exchangers are the highest contributors to exergy destruction. Overall, combustion and heat exchange are the main exergy destruction mechanisms with 57 and 39 MW respectively, while reactions are responsible for 11 MW. Total exergy efficiency definitions are found to consistently overestimate efficiency compared to transit and fuel-product definitions. Network analysis is shown to detect communities of tightly connected plants in the ammonia site. Network metrics rank areas of the site with eigenvector centrality found to be the most relevant metric, but additional tailoring is required to make them applicable to an industrial setting. Industrial decision-makers at the level of the site can have immediate impact on reducing emissions from industry. However, typical exergy analysis studies have been addressed primarily to process designers, based on simulated, static data. So the second step involved applying this methodology to real, dynamic data from an ammonia site. Two years of data from 311 at minute-level frequency are collected. This thesis develops boundary definition and data reconciliation methods for missing and un-metered data, consistent with preserving the dynamic nature of the plant data. The analysis finds average conventional and transit exergy efficiencies for the plant (71\%, 15\%) and its constituent processes: primary reformer (86\%, 40\%), secondary reformer (96\%, 71\%), high-temperature shift (HTS) (99.7\%, 77\%), combustor (56\%, 55\%) and heat exchange section (85\%, 82\%). Overall exergy loss and destruction is 80 MW; the primary reformer and combustor contribute 35 MW and 33 MW respectively. Efficiencies often fluctuate much lower than the maximum value attained and are consistently lower than in the simulation, presenting significant opportunity. The data-driven approach is then extended to evaluate long-standing arguments in favour of exergy such as its ability to evaluate real resource loss and to reduce emissions. Previous studies have focused on analytical, model-based approaches to compare emissions to exergy performance, while parametric studies on the influence of operating parameters on exergy efficiency have also relied on simulations. Different intensity, efficiency and performance metrics are evaluated for the plant and statistically analysed using correlation coefficients, univariate and multivariate first-order and second-order regression models. Exergy efficiency is found to correlate well with a number of performance indicators such as energy intensity. It is the best correlated performance metric to both carbon intensity and cost per tonne of product, quantitatively confirming a long-held qualitative assumption in literature. Exergy efficiency is better correlated with reduced emissions from combustion than with reduced process emissions, indicating its limitations in dealing with that mechanism of industrial emissions. The overall contribution of this thesis is to develop and apply an exergy analysis methodology to a real plant for the first time. It identifies important aspects of the analysis and outlines methods to overcome data collection and cleaning challenges in real sites. Finally, it assesses the ability of exergy efficiency to lead to real resource and emissions savings in industrial sites. Industry needs to decarbonise rapidly by 2050, if the worst effects of climate change are to be avoided. Exergy analysis proves a useful tool for assessing improvement opportunities in industrial sites and can aid towards that goal.