Systems modelling design for waste resource recovery

Considerable amounts of carbon-containing and nutrient-rich waste resources in solid, liquid and gas phases are generated every year and could be converted to value-added bio-products. By far the most widespread recovery technology in the wastewater treatment and bio- solid sector is anaerobic digestion (AD). By 2016, there were 540 operational AD facilities in the UK with a total capacity of 708 MW (equivalent to powering 850,000 households); over 90 AD plants injected bio-methane into the natural gas grid, accounting for 2.2% UK gas consumption [2]. However, concerns have been raised regarding the energy-focused incentive strategies which could hinder the penetration of other environmentally favourable resource recovery options [1] e.g. phosphorus recovery, the AD biogas reformed to hydrogen for microbial protein production. Moreover, AD digestate could be further processed via thermochemical routes (e.g. hydrothermal liquefaction) to recover value-added products. A shift to technology integration to proactively manage waste and co-recover value-added products is underway. However, the composition variability and traceability of diverse waste streams from various sectors hinder the waste valorisation from current passive response to future proactive management. Another challenge lies in the disconnection between two-level design i.e. conversion process synthesis (single site) and enterprise-wide optimisation (multi-site). Process synthesis determines the flowsheet and equipments for given waste and product stream; whereas multi-site network design involves wider temporal and spatial scales such as the short-term operational decisions and long/mid-term planning. In particular, with the evolving Fourth Industrial Revolution (Industry 4.0) [3], many of the data barriers are being overcome through interconnection of smart machinery, which support collection and analyses of real-time data across multi-sites and supply chain. Despite of the advances in technology and process design and integration, sustainability assessment and mathematical optimisation, research gaps emerge in the modelling approach, that can tackle the challenges above and provide multi-level decision-support.

This PhD programme will build on some preliminary work and focus on modelling approach research, aiming to bridge two-level decision spaces (process synthesis and multi-site network design), tackle the bottleneck (e.g. waste variability) and solve the scheduling and planning problems under uncertainty (e.g. uncertainty in waste supply). This can be achieved by incorporation analytical results and operational data (both from lab and industry) within a modelling framework, that integrates waste-recovery process design and integration, life cycle approach, optimisation approach (e.g. rolling horizon algorithms) and machine learning techniques. Capability of the methodology will be tested in one or two applications in UK/China AD-based bioconversion design for wastewater and bio-solid waste recovery where 'plug-and-play' design solutions for the existing centralised and decentralised AD facilities will be explored. The trade-offs between conflicting economic and environmental criteria will be accounted for in optimisation and smart decision-making under coordinated waste-to-resource systems (waste supply, machines at multi-site are interconnected via internet) will be explored.

1. BIS, Building a high value bioeconomy: opportunities from waste, 2015.
2. Anaerobic Digestion & Bioresources Association, ADBA Market & Policy Reports 2016 & 2017. 2016.
3. Recommendations for implementing the strategic initiative INDUSTRIE 4.0. 2013, National Academy of Science and Engineering.