Environmental Statements

1. ¹ Some definitions of KPIs have changed over time. A direct comparison over all years is therefore not fully applicable. See the respective topics in this section [the Environmental Statements section] for further information.

Environmental key figures include data from consolidated industrial sites where Umicore has operational control. The following sites are no longer reported compared to 2020: Wickliffe (United States; Cobalt & Specialty Materials) and Guarulhos (Brazil; Jewelry & Industrial Metals, Precious Metals Chemistry, Precious Metals Refining) – all three entities moved from Guarulhos to the existing site in Americana (Brazil; previously Automotive Catalysts only). The following site was added to the environmental reporting scope in 2021: Jiangmen Site 2 (China; Rechargeable Battery Materials). This brings the total number of consolidated industrial sites that report environmental data in 2021 to 54, down from 55 in 2020. The site in Lyngby (Denmark; Automotive Catalysts) moved location and is reported as Hørsholm (Denmark; Automotive Catalysts) as of 2021. At our site in Tulsa (United States; Precious Metals Chemistry), the activities of Automotive Catalysts were discontinued and no longer reported in 2021.

Within the scope of Umicore’s reporting framework, most of the sites report their environmental data at the end of the third quarter together with a forecast for the fourth quarter. In January, the forecasted values are checked by the sites for significant deviations and, if needed, corrected. The 8 sites with the largest environmental impact for 2021 are: Hanau (Germany; Catalysis, Recycling), Olen (Belgium; Energy & Surface Technologies, Corporate R&D), Hoboken (Belgium; Recycling), Jiangmen Site 1, Jiangmen Site 2 ( both China; Energy & Surface Technologies), Cheonan Site 1, Cheonan Site 2/3 (both Korea; Energy & Surface Technologies), and Kokkola (Finland; Energy & Surface Technologies). These sites reported their full year figures in 2021. Only sites running from the 1st of January are included. A sensitivity analysis, undertaken for the 2021 data on energy consumption data, indicates that the potential deviation of the Group environmental performance would be 1% in case of a 20% error in the forecasted data.

Please note that due to improved analytical and reporting methods, some of the data published in the 2020 annual report have been restated in the 2021 report.

In 2021, Umicore revised the definitions to allocate raw materials between primary and secondary materials. This revision was initiated due to the changes in the external environment where an increased focus on resource efficiency and circularity led to new metrics. The definitions below are inspired by these developments and have been refined for the Umicore context. The raw materials in scope for this indicator are the metals purchased to development-based applications. The percentage is expressed in total raw materials weight.

The resource efficiency indicator provides information on the nature - primary or secondary - of the raw materials processed at the operational sites into final (Umicore) product.

The following definitions apply for primary and secondary raw materials¹:

Primary raw material: Material which has never before been subjected to use or processed into any form of end‑use product (or part thereof) other than that required for its manufacture. In the absence of information from the supplier on the nature of the raw materials supplied, these raw materials are considered as primary. The collected data are expressed in terms of total tonnage of incoming material.

Secondary raw material: Material which has been used and/or processed before and can be reused or processed again into any form of end‑use product (or part thereof).

Secondary raw materials consist of two sub-groups²:

Secondary pre-consumer raw material: Material resulting from the industrial processes in the value chain before that material has been processed into a product. Please note that this includes waste materials originating from intermediate manufacturing steps in the value chain using primary raw materials as input. In all cases the material should not be suitable for consumption in the intermediate manufacturing steps from which it originates.

Secondary post-consumer raw material: Material resulting from products ending at least one lifetime. Please note that this includes waste materials originating from intermediate manufacturing steps in the value chain using secondary raw materials (pre- and or post- consumer raw materials) as input. This also includes material recovered from waste generated by industrial facilities in their role as end-users of a finished product. In all cases the material should not be suitable for consumption in the intermediate manufacturing steps from which it originates.

Waste is defined as the total volume of generated waste expressed in tonnes/year. The distinction between hazardous and non-hazardous waste is made based on the local regulation for the region where the reporting entitiy is located.

The waste recycling rate is the ratio of the waste recovered by third parties (including waste recovered as energy through incineration) to the total waste.

Water withdrawal figures for 2021 includes withdrawn produced water (mainly, this is the water/moisture content of incoming raw materials and liquid solutions), groundwater withdrawn for remediation purposes and cooling water withdrawn from and returned to surface water, while the ‘water use’ figures from 2017-2020 do not include these amounts.

4.1 Water, by business group

We focus on the metals that are present in Umicore’s material flow and that are relevant to the environment in terms of impact. A detailed assessment to evaluate and define the relevant metals was carried out in 2010 and implemented in 2011. A procedure is in place to evaluate the effect of changes to Umicore’s material flow at existing sites as well as at plants that are newly established or joining the company, to ensure that the list of metals is up to date and relevant. Since 2011, no changes to the list were needed.

Metal emissions to water (load) are defined as the total amount of metals emitted after treatment to surface water from effluent(s) expressed in kg/year. If sites make use of an external wastewater treatment plant, the efficiency of that treatment is considered if known to the site.

Metal emissions to air (load) are defined as the total amount of metals emitted to air, after emissions abatement where applicable, in solid fraction by all point sources expressed in kg/year. For mercury and arsenic, vapor/fume fractions are counted as well.

For each of the metals emitted to water and air, an impact factor is applied to account for the different toxicity and ecotoxicity levels of the various metals when they are emitted to the environment. The higher the impact factor, the higher the toxicity is to the receiving water body (for water emissions) or to human health (for air emissions).

The impact factors for water emissions are based on scientific data generated (“predicted no effect concentrations” or PNECs) for the REACH regulation for most metals and on Tatsi et.al (2015) for thallium¹. An impact factor of 1 was attributed to the antimony PNEC of 113 μg/l. The impact factors for emissions to air are based on the occupational exposure limits (OELs) (reference: American Conference of Industrial and Governmental Hygienists, 2021) and the binding EU OELs. An impact factor of 1 was attributed to the zinc (oxide) OEL of 2 mg/m³. Subsequently, an impact factor for all relevant metals was calculated based on these references. The metal impact to air and to water is expressed as “impact units/year”.

In 2021, a review of the PNECs and OELs for each metal of concern was carried out to update the scientific basis for the impact factors. This led to a revision of impact factors for several metals. The revised set of impact factors for metals to air and water is applied from 2021 onward, and the values for 2020 have been recalculated to allow a direct comparison with the 2020 metal emissions impact. The impact figures for 2017-2019 were not recalculated with the new set of impact factors and are presented only for the sake of completeness – a comparison with those years is meaningful, but not precise.

Other emissions tracked by Umicore are SOx and NOx emissions, which are reported in tonnes/year. The majority of the data for SOx and NOx are obtained from direct measurements (online analyzers), complemented to a lesser extent by data based on calculations based on site-specific data. Our sites emit further compounds to a certain extent, but these are not considered material, based on a thorough review of trends in the years 2011-2015, when data on VOCs, COD, etc. were also collected. All sites that have joined Umicore since 2015 have been reviewed for potential additional material compounds, and no such addition was deemed required.

At all relevant locations with environmental emissions, Umicore is compliant with the applicable laws and legislation that regulate and control emissions to the environment. Legal obligations drive most of our data collection related to emissions; however, additional compounds may be analyzed at higher frequencies in excess of the strictly legal requirements to improve data reliability, where this is meaningful. Emission of compounds that are not legally required to be monitored and that we have not voluntarily added to our analysis campaigns may occur, but the impact of such untracked emissions is considered negligible.

Umicore has applied the materiality principle to emissions since 2016, meaning that only the sites with a material impact in comparison to the Group total are required to report. An assessment of the emissions of 2015, the last year when all industrial sites were required to report emissions, identified 10 or fewer sites that made up 95% or more of the Group total for each (set of) parameter(s) (assessed in terms of load for SOx and NOx and in terms of impact for metals emissions to water and air). Sites that have joined Umicore since 2015 have been reviewed for their materiality impact and were grouped for each (set of) parameter(s) as material or not, based on a comparison with the sites in these two categories of the 2015 assessment. This renders the previously used ‘95% or more’ assessment rule somewhat less accurate, but it is clear that we are still very close to above 95% of the Group total emissions for any material compound. All non-material sites are requested to assess if there were any significant upward deviations from their 2015/recent emissions baseline, triggering a discussion of whether or not they are to be considered material in the reporting year; this was not applicable in 2021.

It should be noted that during the analysis of the 2021 metal emissions data, a notable uncertainty was encountered with regard to the waste water discharged at one site, having an impact on the total emissions of metals to water. The detailed analysis shows that this might lead to an overestimation of up to 5% in terms of metal emissions to water in terms of impact for the whole Group. However, as this leads to an overestimation of our environmental impact and as it is unclear how large exactly that uncertainty is, 2021 metal emissions were not revised down, not least because this also allows for a better comparison with the 2020 figures, for which a similar level of overestimation at the concerned site can be assumed. In addition, at another site, stormwater events led to unplanned discharge of untreated and partly treated process water in 2021, resulting in metal load and impact to the water environment in addition to the standard operational discharge. The contribution in terms of metal impact from this unplanned discharge represents 23% of the total Group impact in 2021. However, this is likely an underestimation of the total load and impact, as the unplanned discharge is only analysed for about half of the metals that the standard discharge is analysed for.”Measures were taken to increase the capacity of buffer tanks at the site to prevent unplanned water discharge in the future, and to fill the mentioned data gaps with a view to aligning the analyzed parameters for full comparability between these two types of water discharge.

Diffuse metal emissions

The concentration of suspended particulate matter (PM10) in air of relevant metals (lead, arsenic and cadmium) is measured in µg/Nm³ daily at three measurement stations related to our production site in Hoboken. The monthly averages result in an annual moving average concentration, which is then multiplied by the impact factors to air for the respective metals. The data were normalized at the end of 2020, giving the baseline for this Let’s go for Zero target.

In 2021 and 2022 a screening is ongoing across the Group to identify which other sites may be material in contribution to this target. We expect to include the performance of other identified sites in future reporting. Their impact will be added to the baseline by projecting their impact backward to end of 2020 and the baseline will be adapted when new sites are acquired.

6.1 Energy, by business group

Indirect energy consumption: energy from purchased electricity, steam, compressed air and heat

Direct energy consumption: energy from fuel, gas oil, natural gas, LPG, coal, cokes, pet cokes etc.

The definition of renewable energy as given in the Greenhouse Gas Protocol scope 2 Guidance (2015 amendment) has guided us in defining the scope of this indicator. Only the following energy sources are considered in scope for this KPI: wind energy, solar energy, energy from biomass (including bio- and other naturally produced gas), hydropower (including marine hydro) and geothermal energy.

7.1 scope 1 & scope 2 emissions, group data

7.2 scope 1 + 2 emissions, by business group

Umicore reports its absolute CO₂e emissions as per the scope of sites outlined in E1. The absolute CO₂e emission volumes are calculated using the Greenhouse Gas Protocol definition and reporting methodology for scopes 1 and 2 (WBCSD and WRI 2004 and amendment for scope 2 of 2015). scope 2 for Umicore includes not only purchased electricity but also steam, compressed air and heat purchased from third parties (from industrial parks or utility companies). CO₂e includes the greenhouse gases CO₂, CH₄ and N₂O for scope 1 and major process emissions. Other greenhouse gases are not relevant in Umicore’s operations. The scope 2 emissions take only CO₂ into account.

The calculation of scope 2 emissions for each site is done in two ways: once using market-based CO₂ emission factors and once using location-based CO₂ emission factors. The market-based emission factors allow calculating the CO₂ emissions based on the specific contracts that sites have in place with their energy suppliers, considering the relevant energy mix for these contracts (including green energy attributes, where applicable). The location-based CO₂ emission factors facilitate calculating the CO₂ emissions based on grid average emission factors in a country/region where these data are available. The total CO₂ emissions for the Group are then presented as two separate values based on this differentiation, and the metrics are abbreviated as: CO₂e market-based and CO₂e location-based.

The WBCSD Chemical Sector Working Group on GHG Measurement and Reporting established additional guidance to cope with observed anomalies in GHG reporting. Umicore has implemented these guidelines since the 2012 reporting. The sector guidelines are published on the WBCSD website.

GHG emissions intensity is calculated using the total CO₂e market-based emissions divided by the total revenues excluding metals.

7.3 Upstream scope 3 emissions, group data

The estimation of the scope 3 GHG emissions covers all upstream scope 3 categories except for business travel, which has been excluded due to its assumed low impact. Data were collected at business unit level. The emission factors used come from databases (Ecoinvent 3.4, ADEME, DEFRA, EEIOA, etc.) consulted between June and August 2019, when such emission factors were valid. We list here the main encountered limitations and their related assumptions:

• For purchased goods and services: proxies have been selected whenever the emission factors of the related products were not available. The emissions factor for recycled materials is assumed to be equal to 0.

• For capital goods, the most conservative EFs have been selected from the types of investment available in the environmentally extended input/output analysis database.

• Upstream transportation and distribution: as a conservative approach, in the absence of specific data on destination or starting point in the same country, the distance travelled by the goods was estimated to be equivalent to a large distance between two cities in different parts of the country (e.g., Bruges-Arlon for Belgium). When transportation mode was not provided, it was assumed to be by sea whenever the trip was intercontinental or shorter by sea, and otherwise by road (truck) if on the same continent and shorter by road. When transportation was multimodal, only the biggest part of the journey was taken into account (e.g., for goods shipped from Japan to Germany, only the sea journey was considered and not the truck parts of the journey from ports to facilities).

• For waste treatment, emission factors from the French carbon database were used for all countries, as no other appropriate EFs that take into account waste collection were encountered in available databases. The emissions for the waste sent to recycling or recovery were considered in this category. Considering that no EF was available to represent the recycling treatment process without taking into account the avoided emission linked to remanufacturing new products, the EF for treating hazardous waste was used for calculating the emissions linked to recycling treatment. It is a conservative approach, as “hazardous waste treatment” is the highest EF used in this assessment to estimate waste treatment emissions.

• For employee commuting, emissions were extrapolated on the emissions reported for this category in the last scope 3 inventory made by Umicore (reported year: 2007). The amount of GHG emissions per employee in 2007 was multiplied by the number of employees in 2018.

The scope 3 emissions for the covered categories have been reported in our CDP submissions in 2021, accessible here.

Scope 4, avoided emissions

Estimating avoided emissions requires taking assumptions that have an influence on the results. The main assumptions taken in this case are the following:

Cathode materials for electric mobility

• The solution to compare: we compared the emissions of a medium passenger car with a rechargeable battery containing our cathode materials, with the emissions of a medium passenger car containing an internal combustion engine running on diesel or gasoline, considering the European split between diesel and gasoline in the period 2016-2020.

• We considered the NMC (nickel manganese cathode) materials produced between 2016 and 2020 for electric mobility applications, assuming that the entire volume is used for full electric vehicles. We made our calculations under the assumption that the vehicles are charged using the European average grid mix.

• The comparison covers the following steps: mining, production of the cathode materials by Umicore, processing into batteries, use of the batteries in full electric vehicles and recycling of the batteries at end of life. Literature or LCA data from commercial databases have been used for all processes not carried out by Umicore.

• The production of the car and its recycling has not been considered as it has been assumed that it was the same for both the Umicore technology and the solution to compare.

Recycling

• We compared Umicore’s secondary production with the primary production of an equivalent tonnage of each metal considered.

• As much as possible, we applied the industry average climate change impact provided by the metal associations for primary production. Data from commercial LCA databases had to be used for some metals in the absence of such industry average.

• We have considered the recovery of a selected number of metals by Umicore during the period 2016-2020. For these metals, the climate change impact for the scope 1, 2 and 3 upstream has been considered.

The compliance excess rate is the ratio between the total number of excess results and the total number of compliance measurements. An excess result is a monitoring result that violates a limit value defined in a permit, regulation or other relevant regulatory standard.

The total number of measurements is the total number of environmental impact measurements as required by the operational permit, environmental permit, or comparable standard in the region where the reporting entity is operating (this may include higher frequency measurements of permit-related parameters where deemed useful for internal quality reasons). The total number of measurements means the number of measurement events multiplied by the number of parameters per measurement event.

A complaint is a formally registered notification made by an external claimant, authorities excluded, to the entity / site, concerning an EHS-related issue with a perceived negative impact.

Group data