3.1 Life Cycle Assessment to assess the sustainability of drones
Life Cycle Assessment (LCA) is an analytical tool to measure and compare the environmental impacts of a product, process, system, or service. LCAs are used for decision-making for designers, manufacturers, researchers, and policymakers. The outputs can identify impact hotspots and have the potential to make robust comparisons between different scenarios [11]. LCA studies examine emissions along life cycles or phases: from the extraction of raw materials, transportation, materials processing, manufacturing, distribution, product use, and disposal or recycling at the end of life (EoL) [12] (see Fig. 1). LCA studies separate emissions along life cycles or phases: extraction of raw materials from the earth, transportation, materials processing, manufacturing, distribution, product use, and disposal or recycling at the end of life [12]. LCA studies can be carried out on the entire life cycle such as a cradle-to-grave or a cradle-to-cradle study. Otherwise, LCA studies examine part of a life cycle such as a cradle-to-gate, a gate-to-gate, or a gate-to-grave.
An LCA consists of four iterative phases beginning with the goal definition and scope, followed by Life Cycle Inventory Analysis (LCI), Impact assessment, and interpretation [13]. To define the goal and scope of a product (such as a drone) all the phases in Fig. 1 should be considered. For a drone delivery service study, understanding the operational model of the delivery business and a systems perspective is necessary. Drone delivery services are operated by businesses that are complex and dynamic. A systems approach can capture important interactions and feedback [14]. Furthermore, the functional unit (e.g. a drone or a fleet of drones) employed in any LCA must align with the goal and scope for suitable comparisons [13].
A framework to conduct an LCA is detailed in ISO 14040 series and consists of four phases as in Fig. 2 (with an iterative approach) beginning at the goal definition and scope, followed by Life Cycle Inventory analysis (LCI), Impact assessment and interpretation [13]
To define the goal and scope of a product (such as a drone) all the phases in Fig. 1 should be considered [13]. To define the goal and scope of a product system, such as a drone delivery service, it is necessary to understand the operational model of the delivery business. A systems thinking perspective is important, as drone delivery services are operated by a business. Businesses are complex and dynamic, and systems thinking can capture important interactions and feedback [14]. The boundaries set at the goal and scope phase draw the limits in the process flow diagram.
The functional unit (FU) is a key element of an LCA as it defines how the study is measured, is aligned with the goal and scope, and allows for appropriate comparisons between systems [13]. For example, an LCA study may compare different scenarios using the FU as one drone unit, or a fleet of drones may be more appropriate. A drone delivery system could be 1 delivery or 100 customer deliveries, and this decision may depend on the operational context. An LCA study compares and considers different scenarios which must be modelled on this FU.
The most critical phase of a robust LCA study is the Life Cycle Inventory (LCI) analysis [15]. This consists of four parts which are: creating a flow diagram, developing a data collection plan, data collection (allocation), and evaluation and reporting. The main approach for compiling an LCI is a process analysis that uses data specific to the product under study. However, due mainly to cost and time constraints, it is impossible to exhaustively assess the entire supply chain of any given product. Data is collected by retrieving primary evidence from companies and/or using established databases (such as Eco Invent) that are continuously updated to reflect evolving realities to increase accuracy [15]. A lack of established data on recycling processes has made conducting comparative LCAs difficult when determining favourable scenarios [16]. There is a lack of quantitative, transparent models for handling aircraft and drones at the end of life [17]. The composite materials often end up in landfills, which has led to batteries being a major focus of drone LCAs to date [18]. However, carbon fibre production represents an important aspect of LCA impact categories (such as human toxicity, freshwater aquatic ecotoxicity, marine aquatic ecotoxicity, and terrestrial ecotoxicity) [19], hence more research is needed in this area.
The handling of end-of-life aircraft and UAVs is a relatively undeveloped research topic and little knowledge about the end-of-life process has been reported, as well as a lack of quantitative, transparent models about handling aircraft and drones at the end of their lives [17]. Composites, particularly the most common thermosets, have very little value at the end of life so often end up in landfill, which has led to batteries being a major focus of drone LCAs [18]. However, carbon fibre production represents an important aspect of LCA impact categories (such as human toxicity, freshwater aquatic ecotoxicity, marine aquatic ecotoxicity, and terrestrial ecotoxicity) [19] so more research is needed in this area. Many studies lack transparency in the data, and Bishop et al [20] recommend that additives should be included in plastic studies when there is no clear evidence that their contribution is < 1% to impact categories.
Many LCAs are conducted on something that is already in existence, and results are often used to support decision-makers. Retrospective LCAs on existing products or services are the most common. However, this may not influence a design change, even though the design is the most important aspect to establish the environmental impact [21]. For emerging technologies, there are different challenges [22]. While drones may seem to be commonplace (for amateur uses), drones for delivery are an emerging technology, as industrial-scale drone delivery services are still in development. There are many challenges with conducting LCAs on emerging disruptive technology. Instincts may tell us a drone is ‘better’ than a car or van, but these new technologies need to prove this environmental advantage with scientific evidence. An ex-ante LCA (as opposed to an ex-post) is recommended by some [22]. Modelling future scenarios, such as full-scale operation, maximum efficiency, and market penetration, can in turn inform design improvement [23]. Cucurachi et al [22] define ex-ante LCAs as those that
The literature on ex-ante LCAs classifies LCA assessments as prospective, consequential, dynamic, anticipatory, and mixed [22]. A dynamic LCA will allow the model to include the temporal aspects so that the LCI is more appropriate. A consequential approach models the significant consequences that may happen as a result of the new technology - such as the removal of another technology e.g. in this instance a fossil-fuelled van or car to replace a drone. There are two main methods to model data: attributional and consequential LCA models. Attributional LCAs are more common in a product system such as a drone delivery service (and all of the studies in this paper are attributional). Consequential LCAs function to include consequences across the different scenarios being modelled [20]. For example, there may be “good” consequences (modelled as a negative value) when a road vehicle is no longer manufactured, used, and scrapped because a drone is now replacing it. However, there are negative consequences to be considered with the increased use and demand of batteries at this time [24].
3.2 Drones, composite materials, and circularity
Government policies and action plans for a circular economy (CE) are driving change in the dominant model of the take-make-waste linear economy. CE offers a framework to improve the sustainability of drones by designing out waste and pollution in all phases, making more durable products to extend the use phase. A CE model can focus attention on designing products for repair, reuse, and remanufacturing as a priority before the consideration of recycling.
Aircraft contain large volumes of fibre-reinforced composites and are mostly landfilled when they are retired [25]. The impact can depend on manufacturing processes, additives, the carbon intensity of the grid, recycling (and displacing virgin production), and EoL treatment methods [16]. While emissions from landfills are not of concern due to the inert nature of the waste, the Council Directive 1999/31/EC on the “landfill of waste” addresses a reduction of material disposed of by landfill and has a direct impact on composite manufacturers [26]. Incineration produces larger amounts of GHG emissions mainly from the combustion process as the carbon content of carbon fibre is released to the environment as CO2 [16].
Composite materials for aircraft are highly challenging in the recovery and recycling phases [26]. Pyrolysis recovery of carbon or glass fibres is positive, as it consumes only 5–10 % o the energy required to produce real carbon or glass fibre [27]. However, while composites recycling can reduce impacts, it is often not economically viable [16] outside of non-structural light-weighting applications that do not demand high levels of purity and feedstock quality [16, 28]. A lack of inventory data on recycling processes in life cycle inventory (LCI) databases and in the markets for recycled materials, comparative LCA studies on CFRP recycling are not well established, making it difficult to compare the environmental performance of recycling technologies and determine which one offers the most benefits [16]. The impacts of composites differ greatly due to the different manufacturing processes, additives, carbon intensity of the grid, recyclability, and the EoL treatment methods. The data variation can be so extreme that the carbon footprint of recycled composites can sometimes exceed that of the original production [16].
