Characterization of Raw Materials Based on Supply Risk Indicators for Europe

electronic supplementary material

social lca in progress

Characterization of raw materials based on supply risk indicators for Europe

Lucia Mancini1 • Lorenzo Benini1 • Serenella Sala1

1European Commission, Joint Research Centre, Institute for Environment and Sustainability, Sustainability Assessment Unit, Via Enrico Fermi 2749; T.P. 270; 21027 Ispra (VA), Italy

Received: 14 July 2015 / Accepted: 19 May 2016

© Springer-Verlag Berlin Heidelberg 2016

Responsible editor: Alessandra Zamagni, Ph.D

 Serenella Sala

Content:

1. Using supply risk factors for deriving resource security CFs: methodological issues
2. Deriving environmental ranking for resources
3. Selection of LC inventories used for testing the different indicators for impact assessment
4. Results of the testing for 50 products on representative resource

1. Using supply risk factors for deriving resource security CFs: methodological issues

The supply risk factors as developed in the EU study can be interpreted in two different ways: as intensive or extensive variables. In case the former interpretation is adopted, the factors represent the supply risk associated to a unit of mass of raw material. If the latter interpretation is adopted, the factors would represent the global supply risk associated to the entire market of that specific raw material. In that case, in order to make these factors usable at the product scale, the intensity of supply risk per unit of mass can be derived by dividing the SRWGI factors by the total volume of material marketed in a given year. Similarly it could be done by dividing for the reserve of the given material, although this choice is more criticisable as the SRWGI factors are calculated on the basis of market shares by country.

Drawbacks of the first interpretation:

If the first interpretation holds, the total impact at world scale would be ideally calculated by the following: SRWGI factors times the production volumes. This would result in a non-sense, as the equation would lead to the following: the higher the production, the higher the global supply risk associated to the market of that material, which is a contradiction of terms.

Moreover, by looking at the results of the cases we calculated, in case of the straight application of SR as characterisation factors in product LCA, the effect of the factors becomesnegligible as for all the groups of products assessed the result is completely dominated bytheinventory (see fig...), astheorder of magnitude of the masses varies verymuch amongresources.The most evident drawback of this choice is the factthat the impact assessmentmodel and indicator fails to highlight products which are composed of resources for which their supply is at risk. A way to compensate this limitation is the introduction of an exponent tothe equation, making hence the SR factors numerically more distant one to each other andthen adding explanatory power to the set of factors. However, the choice of the exponent istotally arbitrary, not being evidencebasedbut driven by the expected performance of themethod.

Whereas thesecond interpretation does not suffer from these limitations, as the difference inthe order of magnitude in inventory is compensated at the level of the characterizationfactors, which take into account global production (or reserve) figures.

Impact pathway and cause-effect chain

The supply risk characterization factors are intended to be proxy indicators for a midpoint impact on the economic resource availability impact category as proposed by Schneider et al. (2014). Unlike environmental impacts and mechanistic processes, which can be modelled through a specific and measurable cause effect chain, supply risk is proposed to represent the potential risk associated with the consumption of specific material resources, in order to perform a hotspot analysis of the supply chain. The scope is indeed to warn users on the use of critical resources, rather than to measure an impact. As for criticality assessment, the nature of this analysis is not predictive and has a strategic nature. The impact assessment at endpoint level, however, could be envisaged as reduced resource availability or increased production costs (for the supply side) and increased prices for consumers.

General limitations

Factors have to be updated frequently as the markets are dynamics, newmines can beexploited and governance of countries might change.

In their current form, factors are meant as global but they do express the fears of resource-constrained economies (e.g. EU) a way better than those of resource rich economies. Thus it mightmake sense for the EU overall, although with some limitations for those resources with high shares ofthe extraction occurring within the EU.

In conclusion, a set of economyspecificfactors should be calculated for matching specificpolicy needs, as supply risk is specificof groups of countries.

1. Deriving environmental ranking for resources

An estimation of overall environmental impact has been carried out by performing the following steps:

1)Normalization of the results by using the factors provided by Sala et al. 2015

2)Calculation of the logarithm base 10 for each of the impact categories in order to quantify the order of magnitude of the impact;

3)As some results were negative (for arguments lower than 1 the logarithms provide negative values), the minimum value was calculated for each impact category, increased by 10% and then added to the results so to obtain positive values different from 0;

4)The average of the values so calculated was made across the impact categories. These results represent the average magnitude of the impact across the impact categories.

The main limitation of this approach is related to the fact that using a composite indicator implies that shifting of burdens among different impact categories would not be visible. However, the environmental score is used in the context of the discussion with an illustrative purpose, in order to highlight that the different perspectives on resources lead to different priorities.

Sala S, Benini L, Mancini L, Pant R (2015) Integrated assessment of environmental impact of Europe in 2010: data sources and extrapolation strategies for calculating normalization factors. International Journal of Life Cycle Assessment, 20(11), 1568-1585.

1. Selection of LC inventories used for the impact assessment and clustering

Table 1 shows the full numbered list of LC inventories selected from Ecoinvent; table 2 shows how these inventories have been clustered, based on similarities in the material composition, and the inventories chosen as representative, and for performing the impact assessment.

Table S1 Numbered list of the selected LC inventories

Product number / Ecoinvent name
1 / 1 kg Ferrochromium, high-carbon, 68% Cr {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
2 / 1 kg Magnesium-alloy, AZ91 {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
3 / 1 kg Pig iron {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
4 / 1 kg Steel, low-alloyed {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
5 / 1 kg Aluminium, cast alloy {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
6 / 1 kg Cobalt {GLO}| production | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
7 / 1 kg Sand {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
8 / 1 kg Fluorspar, 97% purity {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
9 / 1 kg Rare earth concentrate, 70% REO, from bastnasite {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
10 / 1 kg Gypsum, mineral {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
11 / 1 p Computer, laptop {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
12 / 1 p Display, liquid crystal, 17 inches {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
13 / 1 p Keyboard {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
14 / 1 p Printer, laser, colour {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
15 / 1 m Cable, connector for computer, without plugs {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
16 / 1 kg Printed wiring board, through-hole mounted, unspecified, Pb containing {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
17 / 1 kg Battery, Li-ion, rechargeable, prismatic {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
18 / 1 p Hard disk drive, for desktop computer {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
19 / 1 m2 Photovoltaic panel, multi-Si wafer {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
20 / 1 m2 Photovoltaic panel, single-Si wafer {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
21 / 1 tkm Transport, freight, aircraft {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
22 / 1 personkm Transport, passenger, aircraft {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
23 / 1 tkm Transport, freight, lorry 16-32 metric ton, EURO4 {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
24 / 1 km Transport, passenger car, medium size, petrol, EURO 4 {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
25 / 1 personkm Transport, passenger, motor scooter {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
26 / 1 personkm Transport, passenger train {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
27 / 1 personkm Transport, passenger train {DE}| high-speed | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
28 / 1 tkm Transport, freight, sea, transoceanic tanker {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
29 / 1 tkm Transport, pipeline, offshore, long distance, natural gas {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
30 / 1 tkm Transport, pipeline, onshore, petroleum {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
31 / 1 kg Cement, Portland {Europe without Switzerland}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
32 / 1 kg Brick {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
33 / 1 kg Concrete block {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
34 / 1 kg Concrete roof tile {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
35 / 1 m2 Door, inner, glass-wood {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
36 / 1 kg Foam glass {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
37 / 1 kg Rock wool {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
38 / 1 m2 Window frame, wood-metal, U=1.6 W/m2K {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
39 / 1 kg Acrylic varnish, without water, in 87.5% solution state {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
40 / 1 p Air filter, central unit, 600 m3/h {GLO}| market for | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
41 / 1 kWh Electricity, high voltage {GLO}| treatment of bagasse, from sugarcane, in heat and power co-generation unit, 6400kW thermal | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
42 / 1 kWh Heat, central or small-scale, other than natural gas {CH}| heat and power co-generation, biogas, gas engine | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
43 / 1 kWh Heat, central or small-scale, natural gas {Europe without Switzerland}| heat and power co-generation, natural gas, 50kW electrical, lean burn | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
44 / 1 kWh Electricity, high voltage {CH}| heat and power co-generation, wood chips, 6400kW thermal, with extensive emission control | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
45 / 1 kWh Electricity, high voltage {CH}| electricity production, hydro, run-of-river | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
46 / 1 kWh Electricity, high voltage {BE}| electricity production, hard coal | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
47 / 1 kWh Electricity, high voltage {CH}| electricity production, nuclear, pressure water reactor | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
48 / 1 kWh Electricity, high voltage {CH}| electricity production, natural gas, 10MW | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
49 / 1 kWh Electricity, high voltage {CH}| electricity production, wind, 1-3MW turbine, onshore | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)
50 / 1 kWh Electricity, low voltage {CH}| electricity production, photovoltaic, 3kWp slanted-roof installation, multi-Si, laminated, integrated | Alloc Def, S (of project Ecoinvent 3 - allocation, default - system)

Table S2 Clustering of the LC inventories based on dominating resources

Cluster / Description / Inventories included in the group (number as in table 1) / Representative inventory
1 / Iron dominated products / 3, 4, 28, 29, 30,40,43 / Steel, low-alloyed {GLO}| market for | Alloc Def, S
Transport, pipeline, onshore, petroleum {GLO}| market for | Alloc Def, S
2 / Clay dominated products / 31, 32, 33, 34, 37 / Concrete block {GLO}| market for | Alloc Def, S
Rock wool {GLO}| market for | Alloc Def, S
3 / Iron and clay dominated products / 7, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 45 / Computer, laptop {GLO}| market for | Alloc Def, S
Electricity, high voltage {CH}| electricity production, hydro, run-of-river | Alloc Def, S
4 / Aluminium and iron dominated products / 5, 38 / Aluminium, cast alloy {GLO}| market for | Alloc Def, S
Window frame, wood-metal, U=1.6 W/m2K {GLO}| market for | Alloc Def, S
5 / Inventories dominated by iron, clay and another material / 35, 36, 39, 50 / Door, inner, glass-wood {GLO}| market for | Alloc Def, S
Electricity, low voltage {CH}| electricity production, photovoltaic, 3kWp slanted-roof installation, multi-Si, laminated, integrated | Alloc Def, S
6 / Others (inventories dominated by a single material, other than iron, clay and aluminium) / 1, 2, 6, 8, 9, 10 / Ferrochromium, high-carbon, 68% Cr {GLO}| market for | Alloc Def, S
Cobalt {GLO}| production | Alloc Def, S
Rare earth concentrate, 70% REO, from bastnasite {GLO}| market for | Alloc Def, S
Gypsum, mineral {GLO}| market for | Alloc Def, S

1. Results of the testing of the 50 products and process for resources

In order to understand the effect of the choice of the algorithm to be used for translating SRWGI factors into characterization factors for use in LCA, a comparative exercise was done for several raw materials (iron, clay, gallium, platinum, chromium and magnesium). For each of the raw material a comparison between the contribution in mass to the total inventory (x axes) and the contribution to the overall criticality scores calculated (y axes) was performed for each of the 50 processes and products retrieved from Ecoinvent v2…..

The first two charts show the behaviour of raw materials characterized by low criticality but dominant in mass. By applying the methodology SRWGI it clearly emerges that the share in mass dominates the impacts, making the role of characterization almost negligible. The opposite holds when applying SRWGI^6 or SRWGI/P, as the share in mass over the inventory is very often negligible. Discrepancies from this trend are to be interpreted as deviations due to the specific characteristics of the process or product analysed.

Chart 3 and 4 present the typical behaviour of critical raw materials: they are used in small quantities and their share over the total mass is almost negligible, very often below the cut-off values generally used in LCA. IN this case by adopting SRWGI/P it is possible to see that regardless of the little quantities the contribution to the criticality score can be extremely high. Instead, the results obtained by applying SRWGI and SRWGI^6 are dominated by the extremely low mass of gallium and platinum in inventory.

In the case of magnesium, SRWGI^6 shows higher values than SRWGI/P. this is due to the fact that production volume of this raw material is not small as in case of e.g. gallium and platinum.

In the case of chromium it is difficult to identify a clear pattern as both mass-wide and criticality-wise it is not one of the major contributors. The results are likely to be influenced by the performance of other raw materials included in the inventory rather than chromium itself.