A new life cycle impact assessment methodology for assessing the impact of abiotic resource use on future resource accessibility

Based on the same depletion concept, which demonstrates the reduction of stocks, many methods were proposed based on different theoretical foundations for assessing resource depletion in LCA, besides ADP method.

Schneider et al. (2011) derive the anthropogenic stock extended abiotic depletion potential (AADP) based on stocks in the economy and in the environment, since it is widely agreed that depletion does not occur if the resource is still available after extraction into the economic stock. The definition of the problem was changed to a decrease in resource availability in the environment and the economy for AADP method (Van Oers and Guinée 2016).

Lately, van Oers et al. (2020a) updated the reserve data and production rate data used to calculate ADP that were recorded and published in 2002 by USGS and revisited the calculation procedure for ADP to address the time dependence of the production rate data. As results, a time series of ADPs on the basis of moving average of production and cumulative production can be calculated. If the aim is to reflect trends, the most appropriate choice is to use a 5-year average production ADP, as shown in Eq. (1) and Eq. (2), where \(m\) indicates the number of years used in the calculation.

If the aim is to reflect how problematic is the cumulative use of primary natural resources by society, then a cumulative ADP over a period of time for which production data are available is the most appropriate, according to Eq. (3) and Eq. (4).

Notably, a new concept of abiotic resource dissipation has been proposed and widely considered in recent years. With continuous discussion and research on the problem of abiotic resource use, more authors argue that extracting elements from the environment to the technosphere does not necessarily mean they are lost to future generations if they can be reused, recycled, or recovered. Elements cannot be transformed other than through nuclear fission or decay (Frischknecht 2016; Vadenbo et al. 2014). As a result, the concept of dissipation of resources emerged, which is defined as material flows to sinks or stocks and becomes inaccessible to future users due to various constraints that prevent the potential functions of resources from being utilized (Zimmermann 2017).

As the view that resource dissipations cause resource inaccessibility became accepted, a new method, environmental dissipation potential (EDP), was proposed by van Oers et al. (2020b) to assess the impact of resource dissipation to the environment. Different from ADP, EDP defines the problem as a decrease in the accessibility of abiotic natural resources. Based on the assumption that extracted resources will eventually dissipate completely into the environment over very long-term scales, the characterization model equation for EDPs for the very long-term time horizon is shown in Eq. (5), and Eq. (6) shows the equation of environmental dissipation (ED).where M_t,i is the global amount of primary resource i consumed by all products in year t, equaling the world annual extraction of resource i (kg/year); E_t,VTL,i is the cumulative global emissions of resource i over very long-term time horizon (kg); and e_i is the amount of resource i emitted by a product system (kg/FU).

Further research was conducted on assessing the impact assessment of resource dissipation. Helbig et al. (2020) contended that dissipative losses of metals exist in any process of the global metal cycle, from primary production to waste management, with different receiving media (i.e., tailings and sludges, the environment, landfills, or other material streams), and are diluted so as to make a recovery technically or economically infeasible. They developed dynamic and static material flow analysis (MFA) models for different metals by analyzing each metal’s global cycle and uniform yield parameters.

However, the dissipated resources do not mean to be permanently inaccessible since the dissipated elements are considered potentially accessible in the future, with the development of technology and the availability of economic conditions. Dewulf et al. (2021) mentioned that the destination of dissipation determines the inaccessibility duration of dissipated resources. Based on literature studies and interviews with experts in various fields, they make estimates for the inaccessibility duration of inaccessible stock due to four compromising behaviors (Table 1). These four compromising behaviors are the main causes of long-term dissipation (Dewulf et al. 2021).

Due to the current lack of a common view of resource use and a global consensus on potential issues related to resource use, a project called “Sustainable Management of Primary Raw Materials through a better approach in Life Cycle Sustainability Assessment” (SUPRIM) was initiated. As demonstrated by the SUPRIM project report, stakeholders have some consensus regarding the “role of resources,” i.e., that abiotic resources are valued by humans because of their use in the environment, primary and secondary production, and the “goal and scope,” i.e., the accessibility of resources is more consistent with the goals humans face than the availability, with global geographical scope (Schulze et al. 2020). Hence, this new characterization model defines the role of abiotic resources as their use in the environment, in primary and secondary production, on a global geographical scope, with the goal of resource accessibility.

Regarding the choice of resource type, whether it is limited to elements only (e.g., antimony) or aggregates (e.g., sand) or includes both elements and aggregates, the currently operational resource types are limited to elements. Identification of aggregate equivalents based on composition is considered non-feasible due to the complex geological characteristics of the deposit and the material in use (van Oers et al. 2020a).

Four flows are considered the main cause of resource dissipation in the new method: flows to tailings, flows to landfills, dissipative flows to the environment, and flows enter into other materials. Tailings, as waste from primary production, still contain some resources due to technical limits. According to some reports, mining actions on tailings are meaningful. For example, the contribution of tailings to copper production is estimated at 2% of global production (Graedel et al. 2004). Landfills have been popularly considered a resource reserve. However, some studies have shown that the net present value of recovering resources from landfills is negative (Winterstetter et al. 2016), which does not make economic significance. However, based on the possibility of future technological improvements, such as more efficient energy technologies or reduced aftercare costs, landfill mining is still considered commercially viable and justified as a resource reserve. Non-functional recovery, in which flows enter into other materials, occurs in the end-of-life recycling of resources incorporated as impurity elements in the associated bulk material streams. It is difficult to recycle such an impurity element because its concentration is far too small. Dissipative flows to the environment refer to emissions that occur at any stage. The destination of such a dissipative flow is open and cannot be determined, as it is impossible to track where the dissipated resource is entering. Generally, these two dissipative flows are challenging to recover due to their extremely low concentration.

The choice of temporal scope for the new method in this paper depends on the inaccessibility duration of the dissipated flows considered. As a result, the temporal scope of the new method was chosen to be the mid-to-long-term, that is, the time beyond 65 years (> 65 years), which is the same as the inaccessibility duration of dispersed stock in the environment, landfills, tailings, and dispersed stock in the technosphere (Dewulf et al. 2021).

Based on the above definitions of “role of resources” and “goal and scope,” the method defines the problem caused by current resource use for future generations as a decrease in the accessibility of primary and/or secondary elements in a global scope over the mid-to-long-term (> 25 years) due to the dissipation of the resource.

In cases where the extracted resources are dissipated, additional resources need to be extracted from the environment or anthropogenic stock to meet the demand. Thus, resource dissipation is the underlying reason for the depletion of total stock (i.e., the sum of natural reserves and anthropogenic stock). In this context, dissipated resources include flows into tailings, the environment, other material flows, and landfills. Dissipated resources cannot be used further and are challenging to recycle and return to use, so future generations will not be able to take advantage of the functions they may have in the technosphere.

In the context of hibernating stocks, which are obsoleted and are not recovered as well. However, the point of when hibernating stocks are recovered is uncertain, since hibernating materials are only recovered when they obstruct new activity (Daigo et al. 2015). Hibernating resources may all be at the critical point between being recovered and staying where they are. In addition, the inaccessibility duration of hibernating resources is considered to be very short, ranging from 0 to 5 years (Dewulf et al. 2021). As a result, hibernating stocks are not considered the leading cause of the problem of reduced resource accessibility.

The system model for this method is shown in Fig. 1, which is consistent with the definition of the role, goal, and scope of resources identified in the previous section. The model establishes a boundary of resource stocks in the environment and technosphere and flows of resources between the stock in the environment and the technosphere. Accessible resources consist of resource reserves in the environment and anthropogenic stocks in the technosphere. Extraction is the flow from the environment to the technosphere, while emission is the flow from the technosphere to the environment. In addition to emissions, there are three other types of dissipative flows within the technosphere (tailings, landfills, and other material flows) that end up in inaccessible stocks. Within the technosphere, there is hibernating resource flows, which are recovered in the short term.

Accessible resources consist of resource reserves in the environment and anthropogenic stocks in the technosphere. The reserves in the environment enter the technosphere through the extraction by humans. This process does not result in a loss of resource accessibility because the resources are only transferred locationally in the process. The extraction is only the pathway for humans to obtain natural resources from the environment, so the rate of resource extraction does not influence the accessibility of resources, which is why the ADP method has been questioned.

The extracted resource must then undergo a smelting and refining process, during which a large number of tailings are created. Due to the low concentration of functional elements included in tailings, many actions have been reprocessing the tailings over the last 100 years. Then, the refined resources are fed into product manufacturing and subsequently used. At the end of life, the products flow into landfills. In each stage of a product’s entire life cycle, dissipation of resources to the environment and other material flows can occur. Resources flowing into tailings, landfills, and other material streams travel from the technospheric accessible stocks into the technospheric inaccessible stocks. Resources dissipated to the environment flow from the technospheric accessible stocks into the inaccessible environmental stocks. Within the technospheric accessible stocks, there are also resource flows marked as hibernating resources, which are not considered dissipative streams and circulated within the technosphere due to their relatively short inaccessible duration and the uncertainty of functional loss. Since they represent potential recycling but do not necessarily enter the recycling process, the arrows denoting the hibernating resource flows are indicated by dashed lines.

The characterization factors (CFs) of an element will reflect the relative severity of the inaccessibility of that element and the reference element (Eq. (7)). Consequently, inaccessibility is quantified for each element, although it is a relative measure.

The equation for inaccessibility in the new method takes the ratio form of the original ADP, i.e., the ratio of dissipation to total reserve, and the reserve is squared, as shown in Eq. (8). Although the squared total reserve in the original ADP equation has been criticized as a less objective choice, the square is the lowest power that satisfies the requirement of invariance in the change from kilograms to cubic meters.where \({D}_{i}\) is the dissipation rate of resource i in kg/year and \({R}_{i}\) is the total reserve of resource i in kg.

Due to the different degrees of inaccessibility caused and cause-effect mechanisms, the impact of different dissipative flows varies. Therefore, in this assessment method, four different impact categories can be distinguished to assess the impact of abiotic resource use:

As the problem of aggregating different impact categories in this method is very similar to the aggregation of different energy resources in CED, the weighting approach proposed by Frischknecht et al. (2015) is adopted here. The above four impact categories result in different resource inaccessibility based on the previous elaboration. As the inaccessibility duration is used to measure the degree of resource inaccessibility, the weighting factors for the different impact categories take their different inaccessibility durations as criteria to influence the importance of each category in calculating the final impact score for resource use. The weighting factors can be calculated with Eq. (9).where \({w}_{\mathrm{df}}\) is the weighting factors of different dissipative flows, \(I\) is the inaccessibility duration of the different dissipative flow entering tailings in year, \(\mathrm{df}\) is the dissipative flows which is calculated.

The weighting factors (w_df) for the impact categories are derived by Eq. (9). The best estimates obtained in Dewulf et al. (2021) study were adopted to calculate the inaccessibility duration of the dissipative flows required in the calculation. The best estimates of the inaccessibility duration for different impact categories adopted in the calculations and the derived results of the weighting factors (w_df) are given in Table 3.

Based on the input data in Table 2, the AEDP for the eight metals can be calculated. The calculation process for AEDP follows the description in Sect. 2. The intermediate parameters in the calculation are listed in Table 4, as well as the final results of AEDP. The year of assessment here is determined as the termination year of the primary production data, i.e., 2015. Due to the limitations of primary production data, for metals with very long dissipation durations (iron, aluminum), unavailable data can only be counted as zero in the calculation.

For the space constraints, the calculations of primary production, the net extraction rates, and each year’s contribution to the global annual dissipation rate for these eight metals can be seen in Appendix.

Both the AEDP for the eight elements and their ADP₂₀₁₅ derived from the conventional ADP characterization model with updated data (van Oers et al. 2020a, b) are shown in Table 5 for further analysis. Due to data limitations, the AEDP for antimony (Sb) is temporarily unavailable, and therefore, silver (Ag) is used as the reference substance for the AEDP. And the ADP of each element are normalized based on the ADP of Ag, that is, dividing the ADP of each element by the ADP of Ag. Since the result of the ADP is the relative depletion potential of resource i and the reference substance Sb, the ratio of ADP_i to the ADP_Ag indicates the relative depletion potential of resource i and Ag. The normalized ADPs are expressed in kg Ag/kg.

The controversy regarding the total reserve’s power choice in the ADP has not been resolved. The conclusion proposed by Guinée and Heijungs (1995) is that the power of reserves must be one greater than the de-accumulation. Therefore, only powers higher than 2 are reasonable. The AEDPs using the squared, cubed, and quadratic total reserves are listed in Table 7.

The AEDPs with different powers of total reserve from Table 7 are plotted in the bar chart, Fig. 6 for a more visual comparison. By comparing the different AEDPs in Fig. 6, it can be observed that the power of the total reserves does not significantly affect the ranking and relative size of the AEDPs of elements. The relationship between AEDPs with different powers of total reserve is shown in Eq. (23). The ratio between the AEDPs with cubed total reserve and the AEDPs of squared total reserve of resource i is determined by the ratio of the total reserve of the reference substance to that of resource i.

Natural reserve is dominant in the total reserve, while anthropogenic stock has a negligible impact on the total reserves. Therefore, the higher the power of the total reserve, the more the AEDP is influenced by resource’s natural reserve. Evidence of this phenomenon can also be found in Fig. 6.

As the choice of natural resource reserves is still a controversial issue. This study uses “ultimately extractable reserve,” “resources,” and “economic reserve” for sensitivity analysis. The data for the “economic reserve” of the eight elements in this study are collected from the Mineral Commodity Summaries released by the USGS (2016). The data for “resource” do not have a uniform source and are obtained from Deutsches Kupferinstitut (2011), Frondel et al. (2006a, b), and Hill and Sehnke (2006). The data for the ultimately extractable reserves are taken from the data for option 1 published by Schneider et al. (2015), which was calculated based on the assumption of UNEP (2011) that 0.01% of the total amount in the crust at a depth of 3 km would eventually be available. Ultimately extractable reserve data for silver (Ag) are not directly available. However, they can be derived from the composition of the upper continental crust published by Rudnick and Gao (2003), following the calculations mentioned by Schneider et al. (2015).

Since the natural reserves have significant influence on the CFs, a variation of ± 10% was considered in sensitivity analysis, so that uncertainties of reserve data are included. The economic reserve, resource, and ultimately extractable reserve data for the eight elements concerned in this study and CFs calculated from the different reserve data with ± 10% variations are listed in Table 8. For comparison purposes, the different reserve data in Table 8 are compared in the bar chart in Fig. 7. The CFs in Table 8 with different reserve data are compared in the bar chart in Fig. 8. The AEDPs based on the “ultimately extractable reserve” are almost proportional to the baseline AEDP. In addition, there is no influence on the AEDP ranking of the elements. The reason for this is that the calculation of the “ultimately extractable reserve” is similar to that of the “ultimate reserve,” both of which are related to the concentration of elements in the upper crust.

Nevertheless, there are significant differences between the AEDP based on “economic reserves” and “resources” and the baseline AEDP. These two reserves are influenced by the resource price of the metal and the cost of extraction. Thus, they are much lower than the amount of “ultimately extractable reserves.” Nickel (Ni) and aluminum (Al) are notably affected by economic factors, with a significant difference between their “economic reserves” or “resources” and their “ultimate reserves.” Possible reasons for this are that aluminum (Al) is more energy-intensive in production than iron, and nickel (Ni) cannot be economically extracted at low mineral concentration. In some cases, the extractable reserves of a resource may be abundant, but lower market prices compromise its economically extractable reserves. Moreover, demand often influences market prices, making the issue more complicated and harder to assess. Regarding the inclusion of economic factors, as mentioned by van Oers et al. (2002), there is no optimal choice of data for the characterization factor, and the choice of data depends only on the definition of the abiotic depletion problem. However, as mentioned in the previous section, in this paper, the goal is resource accessibility. The role of resources is their use in the environment, primary and secondary production, rather than their economic value. Hence, economic factors should be excluded.

While the method achieves the stated objectives, it also has a variety of limitations and shortcomings. In summary, the limitations of the method are mainly related to the data, the theory, and the results.

In the data aspect, the main limitations of the method are the small range of final operable elements and the influence on the accuracy of the results due to missing data. Not only data on residual tailings content and data used to calculate resource dissipation rates are missing but also recorded years for world primary production. It does not influence the resources with high dissipation rates. However, for those with low dissipation rates, such as iron (Fe) and aluminum (Al), it reduces the accuracy of the final results. As resources with low dissipation rates usually require more years of production data, the missing data are represented by zeros.

In the theoretical aspect, the limitation of this method is the assumptions made in the characterization model. In the characterization model, the annual dissipation of an arbitrary resource is assumed to be constant to overcome the arbitrariness of the dissipation curve. However, this is not the case—the dissipation curves for resources record dissipation occurring in several continuous processes.

In the aspect of results, there are changes in the characterization factors obtained by the new method compared to the original ADP method, but they are not significant. As the new method retains the emphasis on reserves in the characterization model, and as the reserve data is usually of considerable magnitude, the new characterization factor of the resource is dominated by the reserves.