This paper presents and clarifies the academic and industrial terminology used in the area of product recovery. It is concluded that there exist many different concepts and definitions in academia and industry, several of which are unclearly defined. Given this, a new way to define product recovery is presented through the use of a model. This model is based on actual industrial product recovery cases, existing academic product recovery concepts and definitions and product design theory. The presentation contains a holistic model that can be used for describing and analyzing different product recovery scenarios. In addition, several industry cases are presented as a verification of the model.

Quantitative modelling of reverse logistics networks and product recovery have been the focus of many research activities in the past few decades. Interest to these models are mostly due to the complexity of reverse logistics networks that necessitates further analysis with the help of mathematical models. In comparison to the traditional forward logistics networks, reverse logistics networks have to deal with the quality of returns issues as well as a high degree of uncertainty in return flow. Additionally, a variety of recovery routes, such as reuse, repair, remanufacturing and recycling, exist. The decision making for utilising these routes requires the quality of returns and uncertainty of return flow to be considered. In this research, integrated forward and reverse logistics networks with repair, remanufacturing and disposal routes are considered. Returns are assumed to be classified based on their quality in ordinal quality levels and quality thresholds are used to split the returned products into repairable, remanufacturable and disposable returns. Fuzzy numbers are used to model the uncertainty in demand and return quantities of different quality levels. Setup costs, non-stationary demand and return quantities, and different lead times have been considered. To facilitate decision making in such networks, a two phase optimisation model is proposed. Given quality thresholds as parameters, the decision variables including the quantities of products being sent to repair, disassembly and disposal, components to be procured and products to be repaired, disassembled or produced for each time period within the time horizon are determined using a fuzzy optimisation model. A sensitivity analysis of the fuzzy optimisation model is carried out on the network parameters including quantity of returned products, unit repair an disassembly costs and procurement, production, disassembly and repair setup costs. A fuzzy controller is proposed to determine quality thresholds based on some ratios of the reverse logistics network parameters including repair to new unit cost, disassembly to new unit cost, repair to disassembly setup, disassembly to procurement setup and return to demand ratios. Fuzzy controller’s sensitivity is also examined in relation to parameters such as average repair and disassembly costs, repair, disassembly, production and procurement setup costs and return to demand ratio. Finally, a genetic fuzzy method is developed to tune the fuzzy controller and improve its rule base. The rule base obtained and the results of sensitivity analyses are utilised to gain better managerial insights into these reverse logistics networks.

Increasing legislative and societal pressures are forcing manufacturers to become environmentally-conscious and take responsibility for the fate of their goods after they have been used by consumers. As a result, some manufacturers operate hybrid systems which produce new goods and recover used goods. Product recovery describes the process by which used products are returned to their manufacturers or sent to a specialised facility for recovery, before being sold on the original or a secondary market. The quality of the returned goods is a significant issue in product recovery systems as it can affect both the type of recovery and costs associated with it. Quality in product recovery systems has not been adequately studied, with many authors either ignoring the possibility of receiving lower quality returns, or assuming they are disposed of rather than recovered. However, such assumptions ignore the possibility that the firm might be able to salvage value from lower quality returns by using them for parts or materials. This thesis presents four models that investigate the importance of considering the quality of returns in the management of inventory in a product recovery system, by examining the cost-effectiveness of recovering both high quality and low quality returns. The first model is a deterministic lot-sizing model of a product recovery system. It was found that performing both high and low quality recovery reduced the sensitivity of the optimal cost to operational restrictions on the choice of decision variables. The second model is a discrete-time, periodic-review model formulated as a Markov decision process (MDP) and introduces uncertainty in demand, returns, and the quality of the returns. It was found that performing both types of recovery can lead to cost savings and better customer service for firms through an increased fill rate. The third model addresses those industries where produced and recovered goods cannot be sold on the same market due to customers’ perceptions and environmental legalisation. Using an MDP formulation, the model examines a product recovery system in which produced and recovered goods are sold on separate markets. The profitability of offering two-way substitution between these markets was investigated. It was found that offering substitution can allow firms to increase both their profits and fill rates. The fourth model examines the issue of separate markets and substitution in the continuous time domain using a semi-Markov decision process. The continuous nature of the model allows more detailed examination of the substitution decision. It was found that offering substitution can allow firms to increase their profit and in some cases also increase their fill rate. In some cases, production is performed less frequently when downward substitution can be offered, and recovery is performed less often when upward substitution can be offered. The findings of this thesis could be used to help a firm that is currently recovering high quality returns assess the cost-effectiveness of also recovering lower quality returns. Recovering low-quality items, rather than disposing of them, may allow a firm to increase the amount it recycles. The findings highlight the importance of considering the quality of returns when managing a product recovery system as they show that economic gains can be achieved by reusing rather than refusing low quality returns.

Due to legal, economic and socio-environmental factors, reverse logistics practices and extended producer responsibility have developed into a necessity in many countries. The end results and expectations may differ, but the motivation remains the same. Two significant components in a reverse logistics system -product recovery options and return channels - are the focus of this thesis. The two main issues examined are allocation of the returned products to recovery options, and selection of the collection methods for product returns. The initial segment of this thesis involves the formulation of a linear programming model to determine the optimal allocation of returned products differing in quality to specific recovery options. This model paves the way for a study on the effects of flexibility on product recovery allocation. A computational example utilising experimental data was presented to demonstrate the viability of the proposed model. The results revealed that in comparison to a fixed match between product qualities and recovery options, the product recovery operation appeared to be more profitable with a flexible allocation. The second segment of this thesis addresses the methods employed for the initial collection of returned products. A mixed integer nonlinear programming model was developed to facilitate the selection of optimal collection methods for these products. This integrated model takes three different initial collection methods into consideration. The model is used to solve an illustrative example optimally. However, as the complexity of the issue renders this process ineffective in the face of larger problems, the Lagrangian relaxation method was proposed to generate feasible solutions within reasonable computational times. This method was put to the test and the results were found to be encouraging.

Biosurfactants are naturally occurring surface active compounds with unique properties such as biodegradability, low toxicity and tolerance to extreme conditions. These unique properties promote their use as alternatives to traditional petrochemical and oleochemical surfactants, as they satisfy requirements for environmentally friendly manufacturing processes. However, the cost of biosurfactants is still significantly higher than chemical surfactants which hinders their large-scale commercialisation. This work presents an investigation into the production of surfactin, a lipopeptide biosurfactant, exploiting its foamability characteristics for the design and implementation of a recirculating continuous foam fractionation column operated in parallel with a bioreactor. Surfactin is a powerful amphiphilic compound produced by Bacillus subtilis. It is a plant-elicitor with antimicrobial properties offering a huge potential in the food and agricultural industries. However, surfactin has extreme foamability even at low concentrations. This foamability can lead to production problems such as large volumes of uncontrolled overflowing foam with significant product and biomass losses. Here, it is demonstrated that this overflow can be controlled, or eliminated, by integrating a foam fractionation system to the bioreactor in a recirculating loop. A dual production and separation process was engineered and enabled reaching high surfactin productivity in a controlled manner. After elucidating the surface properties of surfactin-rich broth, a foam fractionation column was designed for bench-scale production. Operation of the recirculating column in parallel with the bioreactor enabled air flow to be independently controlled for each unit. Surfactin solutions of various concentrations were tested to relate foamability to concentration over a range of bubble sizes. The sintered glass pore size was revealed to be the main factor influencing the enrichment, with a positive correlation with increasing pore size. Characterisation of the fermentation production rate enabled fractionation column air flow rate to be controlled to ensure sufficient foam surface area for product adsorption. The airflow rate was identified as the main factor impacting on the surfactin recovery rate. This characterisation enabled broth feed flow rate to be controlled to balance the removal rate with the production rate. Two processes were created coupling the newly designed fractionation column with the bioreactor operated either in aerated or non-aerated conditions. Under aerated settings, controlled surfactin production was successfully achieved at a productivity of 0.0019 g L-1 h-1 whilst simultaneously recovering 91% of the product at a maximum enrichment of 79 and 116 through the column and overflow routes, respectively. Under non-aerated settings, overflowing foam was fully avoided and 90% of the product was recovered solely through the fractionation column at an enrichment ratio of 40 under non-optimised settings. Additionally, up to 14% (g/g) increase in surfactin production was observed with the coupling of the fractionation column demonstrating a further benefit as a bioprocess intensifying device for surfactin production. This work provides a benchmark for a robust system for surfactin production, substantially improving the productivity at bench scale, potentially leading the way to more productive and less costly industrial processes for surface active compounds in a wide range of industrials fields.

The aim of remanufacturing is to retrieve a product’s inherent value when the product no longer fulfils the user’s desired needs. By taking advantage of this inherent value through different product recovery alternatives, there is a potential for both economically and environmental advantageous recovery of products. Remanufacturing is a complex business due to the high degree of uncertainty in the production process, mainly caused by two factors: the quantity and the quality of returned products. These factors have implications both on the external processes, e.g. coordinating input of returned products with the demand for remanufactured products, as well as the internal processes that coordinates the operations within the factory walls. This additional complexity needs to be considered when organising the remanufacturing system. The objective of this dissertation is to explore how remanufacturing companies can become more competitive through analysing and managing material flows and remanufacturing processes. The first issue discussed in this dissertation is the drivers that make companies interested in remanufacturing products in the first place. The conclusion is that the general drivers are profit, company policy and the environmental drivers. In a general sense, the profit motivation is the most prevalent business driver, but still there are situations where this motivation is secondary to policy and environmental drivers. Secondly, the need to balance the supply of returned products with the demand for remanufactured products shows that the possible remanufacturing volumes for a product are dependent on the shape of the supply and demand distributions. By using a product life cycle perspective, the supply and demand situations can be foreseen and support is given on possible strategies in these different supply and demand situations. Thirdly, how used products are gathered from customers is categorised by seven different customer relationship types. These types all have different effects on the remanufacturing system, and the characteristics of these relationships are disused in detail. When considering the remanufacturing process within the factory walls, a generic remanufacturing process was developed that divides the remanufacturing process into five different phases; pre-disassembly, disassembly, reprocessing, reassembly and the post-assembly phase. These different phases are separated by three different key decision points in the process that also have a major impact on the material planning of the process. For the remanufacturing material planning and production planning, the possibility to apply lean principles can be difficult. One foundation for implementing lean principles in new production is the existence of standardised processes that are stable and predictable. In the remanufacturing system, the possibilities to realise a predictable process is limited by the “normal” variations in quantity and the quality of the returned cores. Even though lean principles can be problematic to implement in the remanufacturing environment, this dissertation proposes a number of solutions that can be used to make the remanufacturing process leaner.

In this study, we consider the disassembly problem of end-of-life (EOL) products for recovering valuable parts or assemblies. All parts obtained by disassembly processes of an EOL product may not be profitable due to their high recovery costs. Our problem is to select the parts to be released and determine the associated disassembly tasks so as to maximize the total profit. We first tackle the simple part selection problem, and then introduce a time constraint for the tasks to be performed for selected parts and search for incomplete time constrained sequences. We formulate our first problem as a Mixed Integer Problem and show that the constraint set of this formulation is totally unimodular. We also provide the dual formulation of our problem and its interpretation. For time-constrained part selection problem we propose a branch-and-bound algorithm. We first develop some reduction mechanism to reduce the size of the problem. Our solution procedure is capable of solving problems with up to 94 parts and tasks.

abstract: Emergent environmental issues, ever-shrinking petroleum reserves, and rising fossil fuel costs continue to spur interest in the development of sustainable biofuels from renewable feed-stocks. Meanwhile, however, the development and viability of biofuel fermentations remain limited by numerous factors such as feedback inhibition and inefficient and generally energy intensive product recovery processes. To circumvent both feedback inhibition and recovery issues, researchers have turned their attention to incorporating energy efficient separation techniques such as adsorption in in situ product recovery (ISPR) approaches. This thesis focused on the characterization of two novel adsorbents for the recovery of alcohol biofuels from model aqueous solutions. First, a hydrophobic silica aerogel was evaluated as a biofuel adsorbent through characterization of equilibrium behavior for conventional second generation biofuels (e.g., ethanol and n-butanol). Longer chain and accordingly more hydrophobic alcohols (i.e., n-butanol and 2-pentanol) were more effectively adsorbed than shorter chain alcohols (i.e., ethanol and i-propanol), suggesting a mechanism of hydrophobic adsorption. Still, the adsorbed alcohol capacity at biologically relevant conditions were low relative to other `model' biofuel adsorbents as a result of poor interfacial contact between the aqueous and sorbent. However, sorbent wettability and adsorption is greatly enhanced at high concentrations of alcohol in the aqueous. Consequently, the sorbent exhibits Type IV adsorption isotherms for all biofuels studied, which results from significant multilayer adsorption at elevated alcohol concentrations in the aqueous. Additionally, sorbent wettability significantly affects the dynamic binding efficiency within a packed adsorption column. Second, mesoporous carbons were evaluated as biofuel adsorbents through characterization of equilibrium and kinetic behavior. Variations in synthetic conditions enabled tuning of specific surface area and pore morphology of adsorbents. The adsorbed alcohol capacity increased with elevated specific surface area of the adsorbents. While their adsorption capacity is comparable to polymeric adsorbents of similar surface area, pore morphology and structure of mesoporous carbons greatly influenced adsorption rates. Multiple cycles of adsorbent regeneration rendered no impact on adsorption equilibrium or kinetics. The high chemical and thermal stability of mesoporous carbons provide potential significant advantages over other commonly examined biofuel adsorbents. Correspondingly, mesoporous carbons should be further studied for biofuel ISPR applications. / Dissertation/Thesis / M.S. Chemical Engineering 2011

Reverse Logistics (RL) is complex to be managed due to the uncertainty involved e.g. the quality range of products, timing of product returns, and volume of returns. A robust RL design can contribute to increase the effectiveness of RL operations. Therefore, an RL design framework needs to be formalised. Circular economy (CE) focuses on supporting the separation of treatments between technical and biological materials in maximising the design for reuse to return to the biosphere and retain value through innovations across fields. The aim of the research is to develop a new framework to design RL operations based on CE values that can increase the effectiveness and efficiency of RL operations. This research has been conducted through the qualitative research involving cases in product recovery (PR) options that are analysed in-depth. Literature review and interview are the main methods of this research. A test was conducted by interviewing expert and respondents to obtain the expert view and test the research result which is a developed framework to design RL based on CE values. The testing engaged five criteria (usability, feasibility, consistency, effectiveness, and utility). The formal RL design framework, 15 CE values, framework to design RL based on CE values specifically PR options (repair, refurbishment, remanufacturing, and cannibalisation) are main research findings. The research contributes academically to the development of a formal RL design framework and to the identification, reformulation, redefinition, and implementation of CE values. The research can be used as a basis for an effective design of RL that takes into account the economic, environmental, and social impacts. The research can be used as a guideline or an appraisal tool in designing/modifying RL based on CE values that can support the implementation of a single RL operation and also RL based on CE.

abstract: The aims of this project are to demonstrate the design and implementation of separations modalities for 1) in situ product recovery and 2) upstream pretreatment of toxic feedstocks. Many value-added bioproducts such as alcohols (ethanol and butanol) developed for the transportation sector are known to be integral to a sustainable future. Likewise, bioproduced aromatic building blocks for sustainable manufacturing such as phenol will be equally important. The production of these compounds is often limited by product toxicity at 2- 20 g/L, whereas it may desirable to produce 20-200 g/L for economically feasible scale up. While low-cost feedstocks are desirable for economical production, they contain highly cytotoxic value-added byproducts such as furfural. It is therefore desirable to design facile detoxification methods for lignocellulose-derived feedstocks to isolate and recover furfural preceding ethanol fermentation by Escherichia coli. Correspondingly it is desirable to design efficient facile in situ recovery modalities for bioalcohols and phenolic bioproducts. Accordingly, in-situ removal modalities were designed for simultaneous acetone, butanol, and ethanol recovery. Additionally, a furfural removal modality from lignocellulosic hydrolysates was designed for upstream pretreatment. Solid-liquid adsorption was found to serve well each of the recovery modalities characterized here. More hydrophobic compounds such as butanol and furfural are readily recovered from aqueous solutions via adsorption. The primary operational drawback to adsorption is adsorbent recovery and subsequent desorption of the product. Novel magnetically separable mesoporous carbon powders (MMCPs) were characterized and found to be rapidly separable from solutions at 91% recovery by mass. Thermal desorption of value added products was found efficient for recovery of butanol and furfural. Fufural was desorbed from the MMCPs up to 57% by mass with repeated adsorption/thermal desorption cycles. Butanol was recovered from MMCPs up to an average 93% by mass via thermal desorption. As another valuable renewable fermentation product, phenol was also collected via in-situ adsorption onto Dowex Optipore L-493 resin. Phenol recovery from the resins was efficiently accomplished with tert-butyl methyl ether up to 77% after 3 washes. / Dissertation/Thesis / Doctoral Dissertation Chemical Engineering 2017