Mapping supply dynamics in renewable feedstock enabled industries: A systems theory perspective on ‘green’ pharmaceuticals

Considering that the focus of this research is mapping industrial systems enabled by renewable feedstocks, it is necessary to define the terms ‘industrial system map’ and ‘enabling [technology]’ in this context. As ‘renewable feedstock’ we identify any biological or non-biological material that, under certain processing stages, can be converted to another form of intermediate and/or end-product that can be used as value-added input to manufacturing operations.

Firstly, industrial system maps are tools which are used to capture key elements for a range of industrial supply networks and explore the underlining relationships that define networks’ performance and progressive evolution. More specifically, in this study we adopt the view provided by Srai and Gregory (2008) who describe supply network configuration maps as a useful mechanism for capturing and visually representing supply network configurations for further allowing a consistent and coherent analysis. Notably, supply chain maps need to capture the following elements: (i) network shape and structure, (ii) ownership, (iii) levels of vertical and horizontal integration, (iv) relationships and inter-dependencies between network partners, (v) unit operations, (vi) product offerings, (vii) infrastructure, and (viii) network dynamics.

Secondly, considering the inherent dynamism encapsulated to the aforementioned industrial systems’ mapping elements, along with our aim to understand network (re)configuration dynamics stemming from the utilisation of renewable feedstocks in manufacturing, we adopt the view of Zhang et al. (2012) who define enabling technologies in the bio-based economy as: “… technologies for efficient utilization of inexpensive and renewable resources for the production of target compounds.”. Enabling technologies can have quite massive spillover effects via driving technological change and influencing business models (Teece 2018).

The present research work shares the systems view of the Supply Chain and Operations Management domain, with research focus being on discovering cause-and-effect relationships within an objective reality (Gold 2014). Complementarily, the ability to capture and analyse the development of industrial supply chains over time generally supports the apprehension of operating model amendments to leverage subsequent growth opportunities and support future evolution pathways in operations management (Srai 2017). In this regard, we leverage the System Dynamics theory proposed by Forrester (1961) to map complex systems while further capturing the causalities that define the dynamic behaviour of the industrial networks of interest. In supply chains, System Dynamics allows the capturing of the relationships among the involved network stakeholders and product and process entities, from an end-to-end standpoint, to understand and assess systems’ behaviour over time (Saavedra et al. 2018). System Dynamics also allows for the development of representative simulation modelling tools that facilitate the quantitative analysis of complex systems’ behaviour over the course of time. The feedback control characteristics of System Dynamics along with the associated causal loop diagrams and stock and flow maps render the approach as an appropriate methodology for decision-making at the strategic level (Tsolakis and Anthopoulos 2015), applicable to a wide range of managerial, socioeconomic and organisational cases (Sterman 2000). Typically, a causal loop diagram uses directed arrows to represent the relations between dependent (i.e. cause) and independent (i.e. effect) variables in a system, while a polarity is also assigned at each arrow. A positive (+) polarity denotes that the effect changes towards the same direction as the cause (reinforcing feedback, R). On the other hand, a negative (−) polarity denotes that the effect changes towards the opposite direction of the cause (balancing feedback, B).

To enable a more detailed analysis, the approach elaborated in this study embraces the mapping framework proposed by Srai (2017) and Srai and Christodoulou (2014). In this regard, our mapping activity includes five stages of stakeholders and product and process related entities, capturing: (i) institutional players, (ii) sector specialists, (iii) products and intermediates, (iv) production operations, and (v) firms within the supply chain. An industrial systems’ mapping methodology should be able to inform about the role of institutional, regulatory and technological advancements in the manufacturing sectors through (Harrington and Srai 2012):

Then, business stakeholders could compare current states and possible future scenarios of already deployed industrial manufacturing systems to further outline interventions and (re)configuration policies to support competitive transformations in industries.

In addition, considering upcoming advances that could shape industrial manufacturing systems in a circular supply chain context, enabled by the utilisation of renewable feedstocks, the mapping process presented in this study encapsulates four essential theme areas (Tsolakis et al. 2016), including: (i) renewable chemical feedstocks, (ii), production processes and technologies utilised to manufacture value-added intermediates and/or end-products, (iii) market and institutional arrangements, and (iv) value and commercial viability constituents. The aforementioned four theme areas are confirmed as essential for exploring emerging (re)configuration opportunities for supply chains enabled by renewable feedstocks (Srai et al. 2018). In particular, Srai et al. (2018) provide a comprehensive decision-making process and a concise framework for exploring commercially viable supply chain configurations arising from renewable feedstocks, towards delivering value-added intermediates and/or end-products in a circular economy context. Empirical evidence from primary industries, such as chemicals and pharmaceuticals, demonstrate the necessity to tackle a series of key-decisions in these four theme areas for the systematic analysis, both qualitative and quantitative, and design of renewable feedstock-based supply chains (Tsolakis et al. 2016; Kawaguchi et al. 2016). Whereas the primary analysis of traditional networks revolves around a specific product or a firm, supply chains that use renewable feedstocks have to be primarily analysed from a compound-level (Böhmer et al. 2012; Srai et al. 2018), e.g. feedstock availability, quality specifications, physicochemical properties. Following that, synthesis pathways and technology options determine upscaling production opportunities for the resulting intermediates and/or end-products (Xu et al. 2012), along with any possible by-products. At a third stage, investigation of established or new markets for the derived goods, either as substitutes or novel offerings, can dictate any product supply chain (re)configuration options (Black et al. 2016). Finally, the integration of the previous elements leads to the evaluation of the elaborated renewable feedstocks’ economic viability (Paulo et al. 2013).

To sum up, our proposed mapping methodology uses the System Dynamics approach to map five categories of stakeholders and product and process related entities (i.e. institutional players, sector specialists, products and intermediates, production operations, firms) at each of the four identified theme areas (i.e. renewable feedstock, technology, market, value and viability). Thus far, the understanding and analytical capability on industrial and supply network evolution, especially for the domain of renewable feedstocks, is rather limited (MacCarthy et al. 2016). In this regard, we suggest that the mapping of key stakeholders and product/processes entities at each of the ‘feedstock-technology-market-value and viability’ theme areas allows for the consistent analysis and improved understanding of enablers in renewable feedstock-based supply chains, in a repeatable manner.

The development of the analysis framework and the integrated mapping approach for the study of renewable feedstock-based supply chains is theoretically grounded on a fourfold nexus of theme areas, as described in subsection 2.2. The proposed mapping methodology, further detailed in Srai et al. (2018), is underpinned by literature evidence and fourteen interviews with industry experts and academics alike who have a long-term involvement in the field of renewable chemical materials. The informants’ specialisation refers to: (i) renewable chemical feedstocks – three experts, (ii) chemical engineering – four experts, (iii) chemistry – three experts, (iv) biology and biochemistry – one expert, (v) pharmaceuticals – one expert, and (vi) systems engineering – two experts. Following that, a multi-stakeholder workshop was organised to collectively refine the literature and empirical outputs and validate the proposed mapping approach.

The mapping framework and the integrated mapping approach for the investigation of the specified pharmaceutical case are based on our collaboration with leading European chemical supply chain businesses and academic stakeholders in the context of the Engineering and Physical Sciences Research Council (EPSRC) funded project “Terpene-based Manufacturing for Sustainable Chemical Feedstocks” (EPSRC Reference: EP/K014889/1). The project consortium recognises System Dynamics as a valuable methodology for: (i) mapping pharmaceutical industrial systems, (ii) comprehending the systems’ evolving dynamics at a strategic setting, and (iii) advancing network (re)configuration theories in the pharmaceutical supply chains’ field.

Generally, pharmaceuticals are manufactured through complex chemical processes that utilise petrochemical-based feedstocks as raw materials. However, a number of recent research initiatives emphasises the transition from fossil-based feedstocks to renewable alternatives. Therefore, prior to the API formulation, a fundamental step is the identification of renewable chemical feedstocks that conform to specific technical, quality and economic feasibility standards that facilitate their further exploitation for manufacturing pharmaceuticals at an industrial scale. Indicatively, the API for ‘green’ pharmaceuticals could be synthesised from terpenoid feedstocks available in the flora or extracted from compounds present in industrial waste (Tsolakis et al. 2016). Following the feedstock focus strand, specialist suppliers procure the appropriate raw materials required by primary manufacturers. Typically, such key suppliers have a limited supply capacity, are geographically scattered across the globe and are vulnerable to risks hence impacting the resilience of the entire supply chain. To that end, research initiatives and business stakeholders demonstrate a vivid interest on the exploration of a range of diversified chemical feedstocks for producing similar APIs. In this sense, the use of renewable feedstocks can lead to resilient supply networks in terms of upcoming increased market needs, while they can assist in developing a risk mitigation capacity to effectively adapt to the on-going pharmaceutical landscape metabolisms.

For the exemplar case of paracetamol, the industrial formulation of the API is typically based on petrochemical feedstocks like phenol or p-nitrochlorobenzene sourced from specialist chemical industries/laboratories located mainly in China or India (Settanni et al. 2016). For the ‘green’ paracetamol case investigated in our participating project we find that the API could be manufactured from terpenoid feedstocks, either limonene or β-pinene. The identification of suppliers of limonene – found in significant concentrations in citrus waste – or β-pinene – extracted in substantial volumes from crude sulphate turpentine found in waste from kraft paper and pulp industries – is essential for: (i) ensuring adequate supply of renewable chemical feedstock volumes to the API manufacturing firms, and (ii) assuring a reasonable procurement pricing strategy.

The complexity and non-linear behaviour in the ‘Renewable Chemical Feedstock’ theme area is captured through the identification of feedback loops. Indicatively, in the balancing loop B1, an increase in the number of ‘Chemical Feedstock Quality Standards’ imposes constraints that limit the ‘Chemical Feedstock Suppliers’ Number’ which in turn decreases the available ‘Chemical Feedstock Installed Capacity’. Decreased feedstock installed capacity entails limited ‘Chemical Feedstock Supply’, resulting in an increase in ‘Chemical Feedstock Price’ and thus limiting the renewable ‘Chemical Feedstock Demand’. In response to the resulting lower demand, ‘Chemical Feedstock Quality Standards’ in generic applications are lessened to motivate utilisation of renewable feedstocks in industrial applications. The causal loop diagram that captures the mental model of the feedstock theme area which managers could conceive is depicted in Fig. 3.

A significant aspect of mapping approaches in the technology theme area of analysis relates to the demonstration of the way that alternative manufacturing processes result in different value-added intermediates and/or end-products (Srai 2017). Particularly, in pharmaceutical industrial systems technology should be considered across the following echelons, namely:

The aforementioned three supply chain echelons of operations are valid for the industrial production of the majority of medications, following the research and development and clinical trial strands. Therefore, in the technology theme area any technological innovations or advancements in chemical synthesis pathways are expected to have an impact on the unit operations level, further determining the performance of the entire pharmaceutical supply network.

Based on the specific case of ‘green’ paracetamol that we discuss, the selection of renewable chemical feedstocks as opposed to petrochemical alternatives for manufacturing the desired API has major repercussions in terms of synthesis routes and required infrastructure. Currently, paracetamol manufacturing is based on iron reduction as opposed to the ‘green’ chemistry route referring to the direct amination of hydroquinone (Joncour et al. 2014). Within the context of the participating project, the elaboration of either limonene or β-pinene as renewable feedstocks to provide the targeted API is associated with different production inputs and equipment, cost elements and resulting by-products. In addition, a prominent trend at the technology theme area is the shift towards continuous as opposed to batch manufacturing of APIs.

Following the primary manufacturing stage, pharmaceutical dosage forms are produced at the secondary manufacturing echelon (Meneghetti et al. 2016). Dosage forms may include tablets, capsules, suppositories, inhalers, powders or any other form. Nowadays, digital manufacturing technologies and especially 3-D printing are already applied for the manufacturing of pharmaceuticals’ dosage forms aiming to increase inventory control efficiency, ensure market responsiveness and provide personalised medications according to patients’ health profile and related needs (Srai et al. 2014).

Thereafter, medication packages are produced and distributed to the market. Advancements in the field refer to ‘smart’ packages and the exploitation of radio-frequency identification tags to monitor storage conditions and enhance the visibility across the pharmaceutical supply chain.

Prior to the diffusion of the aforementioned advancements in the technology theme area, technoeconomic feasibility is required to be ensured by the responsible research and development institutions following the support of funding initiatives. Furthermore, institutional bodies responsible for ensuring and promoting public health have to approve and certify those innovations to allow for their adoption by the actors at each supply chain echelon. Then, manufacturers will be able to plan their capacity and schedule the production volumes of the selected API, the range of dosage forms and the number of packages necessary to cover the projected market demand.

Similarly to the ‘Renewable Chemical Feedstock’ theme area, the complexity and non-linear behaviour in the ‘Technology’ theme area is captured through corresponding feedback loops. Indicatively, in the reinforcing loop R2, an increase in the ‘Packaging Manufacturing Rate’ implies increased ‘Pharmaceuticals’ Number of Packages’, which in turn results in elevated ‘Packages’ Inventory’. The conceptual system under study is illustrated in Fig. 4 via the relevant causal loop diagram.

Demand for medications is commonly expressed in terms of packages of dosage forms and can be considered stochastic in nature, often estimated by historical data over actual dispensed quantities. Particularly, the demand function should be formulated based on statistics gathered by national healthcare systems to forecast the anticipated quantities to be prescribed by either pharmacy contractors, dispensing doctors and appliance contractors, or dosage units supplied under personal administration. Notably, marketing professionals and academics alike should consider that branded and generic pharmaceuticals follow different life cycle patterns. Therefore, they have to explore different avenues for finding appropriate forecasting models for pharmaceutical data to then allow companies in the sector to prepare relevant marketing strategies and implement recommended inventory control policies (Nikolopoulos et al. 2016). Following the circular economy notion, for the case of generic pharmaceuticals like paracetamol, consumers could be environmental sensitive towards the sustainable nature of the offered medications (Tsolakis and Srai 2017), hence highlighting future market opportunities.

National healthcare systems comprise the major direct market for pharmaceutical firms while these further approve the dosage forms to be prescribed by medical specialists and dispensed by generic retailers to a specific market. Thereafter, wholesalers, retailers and logistics services’ providers should apply appropriate inventory control and management policies to avoid product stock-outs and price inflations.

Analogously to the previously examined theme areas, the complex nature and the non-linear system functionality in the ‘Market’ theme area is appreciated through the respective feedback loops. For example, in the balancing loop B4, lessen ‘National Healthcare System Regulations’ that allow greater financial support to the national healthcare sector would result in enhanced ‘Doctors’ and Practitioners’ Prescriptions’. The increased number of prescriptions augments the ‘Demand for Packaging Units of Pharmaceutical Dosage Forms’, which in turn lowers the ‘Retailers’ Pharmaceuticals’ Inventory’. A lower inventory typically increases, either in number or size, the ‘Orders for Packaged Pharmaceuticals’, thus further motivating ‘Research & Development Initiatives for Pharmaceutical Alternatives’. Research and development in ‘green’ pharmaceuticals are necessitated as renewable chemical feedstocks can expand the installed capacity for the synthesis of either existing or novel pharmaceuticals’ API. Figure 5 visualises the causal loop diagram that captures the connections and interrelations among the system’s variables.

Following a supply chain perspective, devising robust and commercially viable renewable feedstock-based networks requires consideration of key issues (Gold and Seuring 2011), like: limited availability of key compounds, constrained production capacity of existing platform technologies, dispersed geographical allocation of network echelons, and applied risk mitigation strategies to secure continuity of business operations. More specifically, in the pharmaceutical sector, value is primarily harnessed through delivering medication safety and efficacy while economic viability is ensured through the respective monetary flows from downstream to upstream supply chain operations. Following that, personalisation of pharmaceuticals based on individual patients’ healthcare profile should be positioned at the core of every medical treatment solution (Stegemann 2015).

To foster supply chain value and commercial viability, a pharmaceutical product has to be acclaimed by both medical and consumers’ associations, while for ‘green’ medications environmental groups can have a significant role in supporting the provision of environmentally sustainable offerings. Following that, doctors and practitioners could be inclined towards prescribing ‘green’ pharmaceuticals that both have: (i) similar or even improved pharmacological properties to conventional medications, and (ii) reduced environmental impact from an end-to-end supply network perspective. The elaboration of third-party logistics services providers to outsource any closed-loop network operations could support the efforts towards better management of pharmaceutical waste streams and possibly circulation of production resources.

Alike the previous three theme areas, the complex interrelations and the dynamic functionality of the ‘Value and Viability’ theme area is captured through the relevant feedback loops. Indicatively, in the balancing loop B1, increased ‘Doctors’ and Practitioners’ Prescriptions’ result in augmented demand and consumption of ‘Pharmaceutical Dosage Units’, which in turn increases the discarded ‘Pharmaceutical Waste Volumes’. In closed-loop supply networks of specific nature, like in the pharmaceutical industry, an increased amount of waste necessitates the establishment of more contractual agreements with ‘Third-Party Logistics Services Providers’ to address the enhanced ‘Pharmaceutical Waste Recovery Volumes’. However, the increased closed-loop operations of ‘green’ pharmaceuticals that contain chemical feedstocks, which could be recovered for the synthesis of similar APIs, mediate the ‘Environmental Groups’ Complaints’ thus motivating ‘Doctors’ and Practitioners’ Prescriptions’ on renewable feedstock-based pharmaceuticals. Figure 6 illustrates the corresponding interrelations among the involved actors and system’s entities.

The presented mapping approach systematically captures the dynamics among institutional stakeholders, industrial actors, material/product transformations and processing technologies, and firms, following a supply chain perspective. The analysis also extends to four interconnected and mutually interacting theme areas, namely “renewable chemical feedstock – technology – market – value and viability”, which are fundamental for exploring supply chain (re)configuration opportunities arising from renewable feedstocks. The applied approach enables the effective cross-theme area analysis towards comprehending the critical industrial systems’ mechanisms along with the recommended data gathering tools and sources. The System Dynamics view of the proposed mapping approach enables the understanding of the connections, level of engagement and causalities among the different actors across industrial value chains. We argue from theory and a synthesis of empirical findings that decision-makers should explore alternative supply chain (re)configuration options considering that the four theme areas of analysis are interconnected and mutually interacting, as depicted in Fig. 7. Direct linkages among stakeholders and product/process entities, among different themes areas, are case dependent and should be considered. The effectiveness of the proposed mapping approach is discussed for the case of pharmaceutical industrial systems focusing on the exemplar case of ‘green’ paracetamol. More specifically, we indicate generic emergence patterns from a supply chain perspective that stakeholders and actors should reflect to ensure business competitiveness within the today’s circular economy era.

The review of selected industrial systems’ mapping techniques along with the provided approach indicate that mapping approaches should reflect the granularity levels of the intended outputs (Srai 2017). For pharmaceutical industrial systems, a cross-case analysis could foster the identification of sector specific emerging trends and challenges to better inform local industry and policy-makers to outline realistic interventions and value chain reforms. In addition, operations research scholars are recommended to maintain and compare systems’ maps over several time horizons to further enable the analysis of industry-specific patterns of progressive evolution that are particularly evident in the pharmaceutical sector.

The mapping methodology developed in this research provides a basis for conceptualising industry transformations and facilitates industry evolution through capturing the dynamic relations among enablers in industrial manufacturing sectors. Our proposed framework differs from the original business model canvas approach proposed by Osterwalder and Pigneur (2010), principally in terms of the considered feedback effects and the inclusion of causal loops and inherent polarity. In addition, the business model canvas does not take into consideration any potential innovations (Joyce and Paquin 2016); our proposed approach specifically elaborates on the ‘Technology’ theme area. Another key differentiating element of our approach is the consideration of major external factors to a supply system (e.g. policies), a building block that is neglected by Osterwalder and Pigneur (2010).

Through the analysis of the cross-dynamics captured via the provided mapping process, we can identify three main external drivers that foster transformations and supply chain (re)configuration opportunities and impact the respective systems’ long-term commercial viability. The identified drivers span catholically across the four theme areas of analysis and should be evaluated on a multi-stage multi-layer perspective to better inform decision-makers and analysts about sustainable and commercially viable options.

Firstly, regulations have a detrimental role in supporting the diffusion of renewable feedstocks to industrial applications in geographically diversified markets. For example, the demand and production of biofuels from biomass and agricultural residues was increased in the United States of America mainly due to the legislation enacted by the Energy Independence and Security Act of 2007 (Castillo-Villar et al. 2017). In the pharmaceutical industry, governmental regulations should contemporarily ensure the triplet “quality – safety – efficacy” of renewable feedstock-based medications (Desai et al. 2018) and alleviate sectoral barriers to growth (Festel 2011). In particular, owing to the often volatile physicochemical properties of renewable feedstocks, regulatory requirements concern about potential changes in the properties of excipients and biopharmaceuticals’ API during the scale-up and down-stream production processes that could hinder manufacturing consistency and reproducibility (Parr et al. 2016). During the last decades, Desai et al. (2018) report that pharmaceutical industry experts and regulators from the United States, Europe, and Japan have jointly worked towards streamlining, harmonising, and accelerating the development and approval processes of new bio-based pharmaceutical products. Institutional arrangements and regulations could benchmark market needs and drive the use of renewable feedstocks for specific offerings; then technology advancements could be supported to promote value creation in a synergistic manner. Indicatively, for the case of ‘green’ paracetamol, renewable chemical feedstocks like terpenes posse a promising alternative to petrochemicals that can be found either as a by-product during the pulp digestion in kraft paper mills (Roberge et al. 2001; Yumrutaş et al. 2008) or in citrus waste generated in significant volumes by the citrus (i.e. orange and orange juice) industry (Negro et al. 2016).

Secondly, bio-based fine chemicals’ manufacturing requires that synthesis routes for the value-added offerings are developed, tested and optimised at a laboratory and unit operations scale (Jiménez-González et al. 2011), respectively. Thereafter, the future of bio-based supply networks relies on either the technical capability to integrate novel synthesis routes to existing infrastructure and processing facilities (Lapkin et al. 2017), or the economic feasibility of investments in new biorefineries (Gunukula et al. 2018). Especially, commercialisation of new biopharmaceutical molecules requires consideration of key technoeconomic parameters (Mupondwa et al. 2015), including: medications’ biological and pharmacological potency, capital investment and operating costs, market uncertainty analysis, production scalability and profitability. In terms of primary manufacturing, the selection of the renewable feedstock for manufacturing the API has major repercussions in terms of the utilised equipment and chemical synthesis routes, always complying with the intermediate and/or end-product specifications required by the market. Despite existing knowledge about pharmaceuticals’ chemical synthesis pathways, detailed data for manufacturing ‘green’ medications at an industrial scale are sparse. For example, Settanni et al. (2016) report that classic routes to synthesise paracetamol are principally based on a reduction-acetylation system (Mitchell and Waring 2002). On the other hand, a promising ‘green’ chemistry route for the synthesis of paracetamol is documented to be based on the amination of hydroquinone (Joncour et al. 2014).

Thirdly, uncertainty conditions, particularly on the market demand side of ‘green’ offerings in a circular economy landscape, create domino effects to the manufacturing processes thus shaping dynamic negotiations and interactions among supply chain actors (Primo 2010). A demonstrable-confirmed efficacy of biopharmaceuticals and their acceptance by the healthcare community will contribute to the growing market demand for such ‘green’ medications (Kesik-Brodacka 2018). Demand for ‘green’ pharmaceuticals could be projected through available historical data from prescriptions for both patented and generic corresponding medications (Makridakis and Hibon 2000). Conversely, conceptualising the impact of the pharmaceuticals’ ‘green’ image on the respective demand patterns could unveil further market opportunities; however, capturing the ‘green’ image dynamic effects is rather ambiguous to understand and requires the elaboration of specific modelling methods. For the case of ‘green’ paracetamol, the market diffusion dynamics are closely related to inventory management requiring counterfactual analysis (Tsolakis and Srai 2017).

To sum up, three external drivers are fuelling dynamic transformations in supply networks enabled by renewable feedstocks, namely: (i) regulatory conformance with market requirements, (ii) system level technoeconomic feasibility assessment of specified renewable feedstocks, and (iii) target market volume demand. To this effect, the competitiveness and commercial viability of the respective industrial network systems could be ensured by mapping and comprehending the sources of these drivers.