GHG quantification

General

General GHG reduction quantification rules can be found in the Riverse Standard Rules.

Calculations of GHG emissions for the baseline and project scenarios shall follow the method detailed below, based on .

Electronic device refurbishing projects are only eligible for avoidance Riverse Carbon Credits.

Electronic device refurbishing projects serve two functions: waste treatment from a device’s first life (Device A), and the provisioning of a “new” device in its second life (Device B). Both of these functions are included in the project and baseline scenario. See Figure 1 and Figure 2 for a depiction of project and baseline scenario system boundaries.

The baseline scenario represents the functionally equivalent set of activities that would occur in the absence of the project. Therefore, the baseline scenario is the average e-waste treatment of Device A, and the market mix for production of a new Device B. This market mix includes the fraction of refurbished devices that are already on the market (see Appendix 6).

The distribution, packaging, use, and waste treatment of Device B are not included in the calculations because they are assumed to be the same in both scenarios. Therefore, the downstream system boundary is Device B at the factory gate.

Calculations and data collection are based on annual project operations.

Functional Unit

Electronic device refurbishing projects are multifunctional (see General section above) so the functional unit is twofold:

• production of one electronic device (Device B), plus

• treatment of the corresponding amount of e-waste treated (from Device A) to generate this one device.

Data Sources

The required primary data for GHG reduction calculations from projects are presented in Table 2:

Table 2 Summary of primary data needed from projects and their source. Asterisks (*) indicate which data are required to be updated annually during verification (see Monitoring Plan section).

Parameter	Unit	Source proof
Amount of sold devices (sold in a functioning state) during the reference year, by type and sourcing country (listed in Table 3).*	Units of device, by type	Track records from the refurbishing site
Mass of devices (optional)	grams/device type	Internal document containing this parameter
Distance traveled during collection from the sourcing place/country until the refurbishing site, and mode of transport (road or air freight).*	km	Track records from the refurbishing site
If applicable, secondary transport to send collected devices from the project site to another more specialized refurbishing site.*	Number of devices by type, and distance (km)	Track records from the refurbishing site/invoices
Percent of input used devices, broken down by device type, that undergo: light refurbishment, full refurbishment, are recycled, and are saved for spare parts or sold as non-functional devices.*	Percentage (%)	Track records from the refurbishing site
(Optional) The average buyback price per device category.	Currency (Euro - € or dollar - $)	Invoices

Secondary data taken from the literature are used to define default values for the following elements:

• Device expected lifetime (new and refurbished)

• Device mass (if not provided by the Project Developer)

• Emission factors from device production (when not available in the ecoinvent database, see paragraph below)

These values and their sources are provided in Table 3 in the Assumptions section

The (hereafter referred to as ecoinvent) shall be the main source of emission factors unless otherwise specified. Ecoinvent is preferred because it is traceable, reliable, and well-recognized. The ecoinvent processes selected are detailed in Appendix 1 : Ecoinvent processes

Assumptions

1. Electronic devices are evaluated in categories of device types rather than specific device models to facilitate data collection. It is assumed that devices in the same device type category have similar characteristics (mass, emission factor, lifetime), as defined in Table 3.

2. Some devices are not able to be refurbished to a functioning state, but contain some functional parts. Typically, the device is disassembled to harvest those scrap parts to use as spare parts for other refurbished devices. To maintain a conservative approach, these devices are assumed to go to electronic waste recycling.

3. In the baseline scenario, the distance for e-waste collection of Device A and transport to the waste treatment center is assumed to be 100 km.

4. The distribution of devices in the baseline and project scenarios is assumed to be the same, and is therefore excluded from quantifications. This is a conservative assumption, because and transported long distances. In contrast, the project scenario consists of mostly inter-EU shipping of devices across much shorter distances.

5. Packaging, use, and waste treatment of Device B are assumed to be the same in the baseline and project scenarios, and are therefore excluded from quantifications.

6. Refurbished devices are assumed to have shorter lifetimes than new devices, as presented in Table 3. To account for this difference, it is assumed that the amount of device production avoided in the baseline scenario is proportional to the ratio of new and refurbished device lifetimes. See section Substitution for more details. Lifetimes for Apple and non-Apple devices are assumed to be the same.

7. Detailed project data on the refurbishing process is rarely available. It is a manual process, and most impacts in the life cycle come from production of spare parts. Therefore, impacts of the refurbishing process are assumed to equal the ratio of impacts of new and refurbished devices in the detailed life cycle assessment on electronic device refurbishing, published by The French Agency for Ecological Transition (Agence de la transition écologique, ADEME), referred to hereafter as the . Refurbishing impact ratios for Apple and non-Apple devices are assumed to be the same. See Refurbishing process Full refurbishment impacts section, for more details.

Table 3 Summary of assumed lifetimes, masses, and emission factors of new and refurbished electronic devices.

Device type
Smartphone	49	4	0.2	3	2
iPhone	64	5	0.2	3	2
Laptop	170	19	1.6	5	3
MacBook	161	18	1.7	5	3
PC	190	19	5.4	5	3
iMac	211	22	4.5	5	3
Tablet	87	10	0.5	3	2
iPad	61	7	0.5	3	2
Screen	366	37		7	4

Residual value of input devices is detailed in Refurbishing process, Residual value of input devices section and refers to the allocation of impacts from the production of Device A to the refurbished Device B, based on its residual economic value. This is calculated using the ratio of the buyback price to the price of the newly manufactured device. This ratio is calculated for each device type, and assumed to be the same for all models within that category.

Project Scenario

Figure 1 The system boundary and scope of the project scenario are shown. They are broken down into three life cycle stages, each described in the corresponding sections below. X% allocation is described in Refurbishing process section

The project scenario consists of refurbishing used electronic devices, which serves two functions: 1) waste treatment of the device after its first life (Device A) and 2) refurbishing to produce a “new“ device (Device B). This process is broken down into 3 life cycle stages, and displayed in Figure 1:

• Device A e-waste collection

• Device A e-waste treatment of scrap materials

• Device B refurbishing process

E-waste collection

The mass of e-waste collected equals the total mass of input used devices collected at the refurbishing site annually.

Total mass of devices shall be calculated using the number of devices collected for each device type (provided by the Project Developer), multiplied by the assumed mass of each device type shown in Table 3.

For calculating transport distance, Project Developers shall provide the country and/or city where used electronic devices are transported from, and provide the average distance from the collection source to the refurbishing project site.

It is assumed that transport within Europe is done 100% by truck, and overseas transport is done by long-distance air freight.

Calculations project e-waste collection

This step calculates the GHG emissions from transporting used devices during the collection process ($Total$ $E_{P.collection}$).

$\textbf{(Eq.1)}\ N_{i.\ collected} = \sum({\ N}_{i.sold\ }/\ (1\ - \ Re_{rate.\ i}))$

where,

• $N_{i.\ collected}$ represents the amount of input collected devices of type $i$ collected by the project, in number of devices.

• $N_{i.\ sold}$ represents the number of devices by type $i$ sold in a functioning state, and shall be provided by the Project Developer for each verification.

• ${{Re}_{rate.i}}_{\ }$represents the fraction of input used devices of device type i that are recycled, saved for spare parts, or not successfully refurbished to a functioning state by the project, and shall be provided by the Project Developer for each verification.

$\textbf{(Eq.2)}\ Total\ E_{P.collection} = \sum(N_{i.collected}*W_{i}*D_{C.i}*R_{C.i})*\ EF_{\ transport}$

where,

• ${Total\ E}_{P.collection\ }$ represents the sum of GHG emissions due to the transport of devices collected for refurbishing in the project scenario, in kgCO$_2$eq.

• $N_{i.collected}$ was calculated in Equation 1.

• $W_{i}$ represents the weight in kilograms of device i, according to the presented in Table 3.

• $D_{C.i}$ represents the distance traveled for device collection in km, provided by the Project Developer per sourcing country/city ($C$) and device type ($i$).

• $R_{C.i}$ represents the fraction of the devices collected per sourcing country/city ($C$) and device type ($i$).

• $\ EF_{\ transport}$ represents the emission factor for transport in kgCO$_2$eq/kg.km according to the ecoinvent database and includes truck or air freight. Refer to Appendix 1 : Ecoinvent processes for the ecoinvent processes used.

E-waste treatment of non-refurbished devices

Devices collected by the project that cannot be refurbished undergo e-waste recycling. Refurbishing projects typically have contracts with e-waste recycling companies that collect and recycle such devices.

Project Developers shall provide the fraction of devices that are recycled, and they will be modeled as mechanical e-waste recycling with shredding and separation (see ecoinvent processes in Appendix 1).

Some non-refurbished devices may be kept onsite to harvest spare parts in the future, but due to limited project data on this topic, they are assumed to be recycled.

Devices that are sold by the project in a non-functional state shall be treated in the calculations as recycled devices.

Calculations project e-waste treatment

This step calculates the GHG emissions from transporting and recycling the used electronic devices that are unsuitable for being refurbished (${Total\ E}_{P.waste\ treatment}$).

$\textbf{(Eq.3)}\ E_{recycling} = \sum(N_{i.collected}*W_{i}*Re_{rate.i}*\ EF_{recycling.\ i})$

where,

• $E_{recycling}$ represents the sum of GHG emissions due to the recycling process of devices/scrap not suitable for refurbishing, in kgCO$_2$eq.

• $N_{i.collected}$ and ${{Re}_{rate.i}}_{\ }$ were described in Equation 1.

• $W_{i}$ is described in the section section of the Project scenario.

• $EF_{recycling.\ i}\$represents the emission factor of recycling each device type. Refer to Appendix 1 for the ecoinvent processes used.

$\textbf{(Eq.4)}\ E_{transport} = \sum(N_{i.collected}*W_{i}*Re_{rate.i})*D_{scrap}*\ EF_{\ truck\ transport}$

where,

• $E_{transport}$ represents the sum of GHG emissions due to the transport of devices/scrap not suitable for refurbishing that are sent to recycling, in kgCO$_2$eq

• $N_{i.collected}$ and ${{Re}_{rate.i}}_{\ }$ were described in Equation 1.

• $W_{i}$ is described in the section.

• ${D_{scrap}}_{\ }$represents the distance in km until the waste treatment facility. If not known, this value is considered 100km.

• $EF_{truck\ transport}\$represents the emission factor of truck transport. Refer to Appendix 1 for the ecoinvent processes used.

$\textbf{(Eq.5)}\ Total\ E_{P.waste\ treatment} = \ E_{recycling\ } + E_{transport}$

where,

• $\ Total\ E_{P.waste\ treatment}$represents the sum of GHG emissions in the project scenario e-waste treatment of non-refurbished devices, in kgCO$_2$eq.

Refurbishing process

This life cycle stage is composed of four main processes, each described below:

• light refurbishing impacts

• full refurbishing impacts

• residual value of input devices, and

• secondary transport of devices.

Light refurbishment impacts: The refurbishing process is split into two categories: light and full refurbishment, representing the degree of intervention needed to restore the device to a functioning state. Light refurbishment involves cosmetic and software improvements, and does not require the replacement of parts (e.g. new battery, new screen…). This distinction was chosen because

• Light refurbishment includes inputs of cleaning alcohol, tissues, and cloth, and is modeled after the detailed LCA of electronic device refurbishing from the ADEME study.

Full refurbishment impacts: Full refurbishment includes light refurbishment plus repair and replacement of non-functional pieces. Detailed project data on all replacement pieces and inputs are rarely available, so full refurbishment impacts are modeled following the ADEME study.

• Results from this study are used to obtain the ratio of impacts of a refurbished device to the impacts of the corresponding new device (Appendix 3). This ratio is then applied to the new device production impacts summarized in Table 3 to obtain the desired amount of emissions from refurbishing. The emissions from refurbishing are modeled using the mix of ecoinvent processes used in light refurbishment described above, plus production of commonly replaced parts including screens, batteries, microphones and speakers.

For example, the ADEME study found that production of a new laptop emitted 168 kgCO2eq, and the process for refurbishing a laptop emitted 18 kgCO2eq, so the refurbishing impact ratio of laptops is 11%.

In this model, the emission factors used for new laptop and Macbook production are 170 and 161 kgCO2eq. These are multiplied by the refurbishing impact ratio of 11% to obtain refurbishing impacts of one laptop and Macbook of 19 and 18 kgCO2eq.

Residual value of input devices: In life cycle assessments, when a project uses waste as an input, it typically enters the project system boundary with zero environmental impacts. Refurbishing projects collect and refurbish used devices that are not always at the end of their life, and are not truly waste. They may still be functional and hold residual value from their first life. This is evidenced by the fact that Project Developers sometimes pay for used devices, as opposed to waste collection, where the waste generator has to pay for waste treatment.

• In this case, some environmental . It is assumed that only devices that undergo light refurbishment were in good condition and had residual value, and are allocated a share of GHG emissions from the device’s first life. On the other hand, devices that undergo full refurbishment are assumed to be non-functional waste and are not allocated any environmental impacts from their first life.

• The residual value and corresponding allocated emissions are based on the ratio of the buyback price to the selling price of a new manufactured device. An average ratio shall be used for each device type, and is shown in Table 4. Alternatively, Project Developers may provide a similar project-specific database with their own buyback data.

For example, if Smartphone A has an average buyback price of 100€, and it is sold new for 400€, its residual value is 25% of its original value. Then when it undergoes light refurbishment by the project, it is modeled as an input with 25% of its initial production impacts.

If production of a new Smartphone A emits 80 kgCO2eq, and it has a residual value of 25%, then the calculations shall include it as an input with 20 kgCO2eq.

Table 4 Residual value of device types. See Appendix 7 for more details for smartphone, iPhone, tablet, and iPad. The average value of these device types was applied to the remaining device types due to a lack of device-specific buyback data.

Device	Percent of residual value
Smartphone	11%
iPhone	14%
Tablet	20%
iPad	12%
Laptop	14%
Macbook	14%
PC	14%
iMac	14%
Screen	14%

Secondary transport of devices: After the device is collected by the refurbishing project and sorted, it may be sent to a different refurbishment site, for example to do speciality repairs. Project Developers shall report such secondary transport by providing the distance transported, and the number and type of devices making this transport.

Calculation refurbishing process

This step calculates the GHG emissions from the refurbishing process $(Total\ E_{P.refurbishing\ process})$ broken down into four main processes: 1) light refurbishing impacts, 2) full refurbishing impacts, 3) residual value of input devices, and 4) secondary transport of devices.

$\textbf{(Eq.6)}\ N_{light\ ref.i}\ = \sum(N_{i.collected}*R{ef}_{light.i})$

where,

• $N_{light\ ref.i}$ represents the number of devices of type $i$ undergoing the light refurbishing process and sold in a functional state.

• $N_{i.collected}$ is calculated in the section

• $R{ef}_{light.i}$ represents the fraction of devices of type $i$ undergoing the light refurbishing process and sold in a functional state.

$\begin{aligned}\textbf{(Eq.7)}\ E_{light\ ref} = \sum N_{light\ ref.i}\ *(\ &alcohol*EF_{\ alcohol}\\ +\ &paper\ *EF_{paper}\\ +\ &cloth*EF_{cloth})\end{aligned}$

where,

• $E_{light\ ref}$ represents the sum of GHG emissions due to the light refurbishing of a device type.

• $N_{light\ ref.i}\$is calculated in Equation 6.

• alcohol, paper and cloth represent the amount of cleaning alcohol, paper and cloth needed to clean a device. These amounts were taken per device type from the , pages 45, 77, and 103.

• $EF_{\ alcohol}$ represents the emission factor, in kgCO$_2$eq, for cleaning alcohol composed of 70% ethylene and 30% water. Refer to Appendix 1 for the ecoinvent process used.

• $EF_{paper}$ represents the emission factor, in kgCO$_2$eq, of paper. Refer to Appendix 1 for the ecoinvent process used.

• $EF_{cloth}$ represents the emission factor, in kgCO$_2$eq, of cloth used for cleaning. Refer to Appendix 1 for the ecoinvent process used.

$\textbf{(Eq.8)}\ N_{full\ ref.i}\ = \sum(N_{i.collected}*R{ef}_{full.i})$

where,

• $N_{full\ ref.i}$ represents the number of devices of type $i$ undergoing the full refurbishing process and sold in a functional state.

• $N_{i.collected}$ is described in the section .

• $R{ef}_{full.i}$ represents the fraction of devices of type $i$ undergoing the full refurbishing process and sold in a functional state.

$\textbf{(Eq.9)}\ E_{full\ ref} = \sum N_{full\ ref.i}\ *R_{full\ ref.i}*EF_{full\ ref}$

where,

• $E_{full\ \ ref}$ represents the sum of GHG emissions due to the full refurbishing of a device type.

• $N_{full\ ref.i}\$is calculated in Equation 8.

• $R_{full\ ref.i}$ represents the rate of full refurbishment activities modeled per device type i. This reflects the "amount" of refurbishment used as an input for that device. See Appendix 3 for its calculation and the amounts.

• $EF_{full\ ref}$ represents the emission factor, in kgCO$_2$eq, of one full refurbishment activity. This activity includes a mix of ecoinvent processes, described in Appendix 1 and section Refurbishing process.

$\textbf{(Eq.10)}\ E_{residual} = \sum N_{light\ ref.i}\ * \frac{Av.acquisition_{price.i}}{Av.selling_{price.i}} * EF_{new\ i}$

where,

• $E_{residual}$ represents the sum of residual GHG emissions from the device's first life allocated to the refurbished device, for all devices.

• $N_{light\ ref.i}$ is calculated in Equation 6.

• $Av.acquisition_{price.i}\$represents the average price paid for the collected used devices of type i (also called the buyback price).

• $Av.selling_{price.i}\$represents the average selling price of a new device of type $i$.

• ${{EF}_{new.i}}_{\ }$ represents the emission factor in kgCO$_2$eq/kg due to the production of the new device type $i$. The emission factors of new devices are presented in Table 3.

$\begin{aligned}\textbf{(Eq.11)}\ E_{secondary\ transport} = \sum&(N_{secondary\ transport,i}*W_{i}*D_{secondary\ transport})\\ &*\ EF_{\ truck\ transport}\end{aligned}$

where,

• $E_{secondary\ transport}$ represents the sum of GHG emissions from secondary transport.

• $N_{secondary\ transport,i}$ is the number of devices of device type $i$ that are sent for secondary transport.

• $W_{i}$ and $\ EF_{\ truck\ transport}$ are described in the .

• $D_{secondary\ transport}$ represents the distance traveled for secondary device transport in km per device type $i$.

$\begin{aligned}\textbf{(Eq.12)}\ Total\ E_{P.refurbishing\ process} = \ &E_{light\ ref} + {\ E}_{full\ ref}\\ + &{\ E}_{residual} + E_{secondary\ transport}\end{aligned}$

where,

• $Total\ E_{P.refurbishing\ process}$represents the sum of GHG emissions in the project scenario refurbishing process LCA step, in kgCO$_2$eq.

Baseline scenario

Figure 2 The system boundary and scope of the baseline scenario are shown. They are broken down into three life cycle stages, each described in the corresponding sections below

The baseline scenario consists of two main functions: 1) waste treatment of the device after its first life (Device A) and 2) provisioning of a new device (Device B). The system boundary of the baseline scenario is shown in Figure 2. This is broken down into 3 life cycle stages, which are detailed in the following sections:

• Device A collection

• Device A e-waste treatment

• Manufacturing of Device B

The baseline scenario structure remains valid for the entire crediting period but may be significantly revised earlier if:

• The Project Developer notifies Riverse of a substantial change in project operations or baseline conditions, and/or

• The baseline methodology is revised, affecting the scenario.

The specific values within the baseline scenario will be updated annually, using project data to accurately reflect the equivalent of the project’s annual operations.

The structure of the baseline scenario is the same whether the project consists of ongoing operations or an expansion. In the former, project data from all annual site operations is considered, and the baseline scenario is defined as the functional equivalent of all annual operations. For an expansion project, only project data related to the expansion is considered, because the normal annual operations would be the same in the baseline and project scenario, and can therefore be excluded.

For example, for an expansion project, if the expansion allows for the refurbishment of an extra 5,000 devices annually in addition to the business-as-usual (BAU) 10,000 devices, then the baseline scenario could include the BAU refurbishment of 10,000 devices plus the new manufacturing of 5,000 devices. The same BAU refurbishment of 10,000 devices could also be included in the project scenario.

The processes related to the BAU refurbishment are then excluded from the GHG reduction quantification of both the project and baseline scenarios since they would have no effect on the avoided emissions. The remaining project activity to consider is the refurbishment of the extra 5,000 devices from the expansion project.

E-waste collection

It is assumed that e-waste is transported by truck 100 km to its waste treatment center.

The mass of e-waste collected in the baseline scenario equals the total mass of input used devices collected by the refurbishing project annually.

Total mass of devices shall be calculated using the number of devices collected for each device type (provided by the Project Developer), multiplied by the assumed mass of each device type shown in Table 3.

Project Developers may provide more precise information on the mass of collected devices if it is available.

Calculations baseline e-waste collection

This step calculates the GHG emissions from the baseline e-waste collection life cycle stage (${Total\ E}_{B.collection}$).

$\textbf{(Eq.13)}\ Total\ E_{B.collection} = \sum(N_{i.\ collected}*W_{i}\ )*D\ *\ EF_{truck\ transport}$

where,

• $Total$ $E_{B.collection\ }$ represents the sum of GHG emissions in kgCO$_2$eq due to the transport of devices.

• $N_{i.\ collected}$ is calculated in Equation 1.

• $W_{i}$ is described in Equation 2.

• $D$ represents the distance of the device collection in kilometers, which is assumed to be 100 km.

• $\ EF_{truck\ transport}$ represents the emission factor of truck transport in kgCO$_2$eq/kg.km. Refer to Appendix 1 for the ecoinvent process used.

E-waste treatment

The treatment of e-waste is split between recycling, landfilling and incineration (Figure 2).

The proportion of e-waste recycled is based on national statistics obtained from the Eurostat database for small IT devices, as defined by the WEEE directive. Data for other countries where used devices are frequently sourced are taken from the , and extrapolated where necessary. The dataset and more detailed information are in Appendix 4.

First, the fraction of e-waste that is not separately collected is assumed to be collected with municipal waste and incinerated or landfilled. In 2021, for example, this was an average of for the countries included in Eurostat.

The repartition between landfilling and incineration (with and without energy recovery) was taken from Eurostat, and the total repartition for all EU countries from 2020 was used. This resulted in .

Then, the fraction of e-waste that is separately collected is considered (average of ).

• This can be further broken down into the fraction successfully recycled/reused and the fraction that could not be recycled/reused (21%). Country specific fractions are used and are presented in Appendix 4.

• The separately collected e-waste that could not be recycled/reused is assumed to be incinerated and landfilled, with the same proportions described in the Baseline scenario section E-waste treatment.

Calculation baseline e-waste treatment

This step calculates the GHG emissions from the baseline e-waste treatment life cycle stage ($Total\ E_{B.\ waste\ treatment}$).

$\begin{aligned}\textbf{(Eq.14)}\ E_{B.waste} = \sum\ &(N_{i.\ collected}*W_{i}\ *(1 - {RR_{rate. i}}_{}))\\*&(L_{rate}*\ EF_{landfill} + I_{rate}*{EF}_{incineration})\end{aligned}$

where,

• $E_{B.waste\ }$ represents the sum of GHG emissions due to the e-waste treatment of devices not separately collected.

• $N_{i.\ collected}$ and $\ W_{i}$ are described in the section .

• $W_{i}$ is described in Equation 2.

• ${RR}_{rate}$ represents the project's country waste reuse and recycling rate. These rates are presented in Appendix 1.

• $L_{rate}$ and $I_{rate}$ represent the landfilling and incineration rates, respectively, described in section E-waste treatment.

• $EF_{landfill}$ represents the emission factor of treating e-waste via landfill, in kgCO$_2$eq/kg using ecoinvent database, according to the breakdown of materials on pg. 11 of the :

  • treatment of waste plastic, mixture, sanitary landfill = 50%

  • treatment of waste glass, sanitary landfill = 10%

  • treatment of waste aluminum, sanitary landfill = 40%

• $EF_{incineration}$ represents the emission factor of treating e-waste via incineration, in kgCO$_2$eq/kg using ecoinvent database according to the following split :

  • treatment of waste glass, municipal incineration = 10%

  • treatment of waste plastic, consumer electronics, municipal incineration = 50%

  • treatment of scrap copper, municipal incineration = 20%

  • treatment of scrap aluminum, municipal incineration = 20%

$\textbf{(Eq.15)}\ E_{B.separate\ waste} = \sum(N_{i.\ collected}*W_{.i}*{RR_{rate.i}}_{}*\ EF_{recycling.i}\$)

where,

• $E_{B.separate\ waste}$ represents the sum of GHG emissions due to the e-waste treatment of separately collected devices.

• $N_{i.\ collected}$ ,$\ W_{i\ }$, and ${RR}_{rate}$ are describe above.

• $EF_{\ recycling.i}$ represents the emission factor of recycling device i, in kgCO$_2$eq/kg. Refer to Appendix 1 for the ecoinvent process implemented.

$\textbf{(Eq.16)}\ Total\ E_{B.\ waste\ treatment} = E_{B.waste} + {E_{B.separate\ waste}}$

where,

• $Total\ E_{B.\ waste\ treatment}$represents the sum of GHG emissions in the baseline scenario e-waste treatment life cycle stage, in kgCO$_2$eq.

New device production

The number of new devices to consider in the baseline scenario corresponds to the number of devices successfully refurbished and sold in a functional state in the project scenario. Note that this does not necessarily equal the number of used devices collected, because a fraction of devices can not be successfully refurbished.

To quantify avoided GHG emissions, the baseline scenario must consider the market share of the project technology already in use. Currently, new device purchases come from both new manufacturing and existing refurbishing activities, and this is reflected in the baseline scenario (see Figure 2). The proportions of new and refurbished devices are detailed in Table 5.

For example, in 2022, 87% of smartphones sold in Europe were new, while 13% were refurbished (Table 5). Thus, each smartphone refurbished by the project is assumed to replace the manufacturing of 0.87 new devices.

Table 5 Market share of refurbished devices sold annually in Europe. See Appendix 6 for more details.

Device	Percent Refurbished	Percent new
Smartphone/iPhone	13%	87%
Tablet/iPad	7%	93%
Laptop/Macbook	8%	92%
PC/iMac	8%	92%
Screen	6%	94%

The process of manufacturing a new device is taken from the ecoinvent database, without modifications for the following device types: laptop, PC, tablet, and screen (See Appendix 1).

GHG emissions from manufacturing Apple devices (iPhones, iPads, iMacs, and Macbooks) are taken from the production-stage impacts reported in Apple’s Product Environmental Reports. An average emission factor for recent models of devices was taken, and the emission factors considered are presented in Appendix 2.

The emission factor for smartphones was based on ecoinvent data and adjusted to better represent average smartphones. This was necessary because

1. smartphones are one of the most frequently refurbished devices, so special attention should be paid to them

2. , and

3. it has been noted that . See Appendix 5 for full details.

The difference in lifetime between refurbished and new devices, described in the Substitution, is accounted for in this life cycle stage. The amount of new device production avoided in the baseline scenario is proportional to the ratio of new and refurbished device lifetimes.

Note that this ratio is only applied to the new manufacturing of devices, as shown in Table 5, and not applied to avoided refurbished devices in the baseline scenario.

The impacts of refurbishing devices are described in the Refurbishing process section above.

Calculation new device production

This step calculates the GHG emissions from the baseline device production life cycle stage ($Total\ E_{B.\ new\ device\ production\ \ }$).

$\textbf{(Eq.17)}\ E_{new.device} = \sum(\ N_{i.\ sold*{frac}_{New}*EF_{new.i}})$

where,

• ${E_{new.device}}_{\ }$ represents the sum of GHG emissions in kgCO$_2$eq due to the production of new devices (i.e. excluding the market share of refurbished devices that are already in use).

• $N_{i.\ sold\ }$ was described in Equation 1.

• ${frac}_{New}$ refers to the market share (in percentage) of new devices sold annually per device type i, as presented in Table 1.

• ${{EF}_{new.i}}_{\ }$ represents the emission factor in kgCO$_2$eq/kg due to the production of the new device type i. The emission factors of new devices are presented in Table 3.

$\textbf{(Eq.18)}\ E_{Ref\ B} = \sum(\ N_{i.\ sold}*frac_{Refurb}*R_{full\ ref.i}*EF_{full\ ref})$

where,

• $E_{Ref\ B}$ represents the sum of GHG emissions due to the refurbishing of used devices according to the market shares in the baseline scenario.

• $N_{i.\ sold\ }$ was described in Equation 1.

• ${frac}_{Refurb}$ refers to the market share (in percentage) of refurbished devices sold annually per device type i, as presented in Table 5.

• $R_{full\ ref.i}$ and $EF_{full\ ref}$ are described in Equation 9.

Refurbished devices are assumed to have a shorter lifespan than new devices, as described in the Substitution. This is accounted for in the following adjustment to avoided emissions from new device manufacturing:

$\textbf{(Eq.19)}\ E_{new\ device\ lifetime\ adjusted} = \sum(E_{new\ device\ .i}*\ Y_{refurbished.i}/Y_{new.i})$

where,

• ${Y_{refurbished.i}}_{\ }$ represents the expected lifespan of a refurbished device i in number of years, as presented in Table 3.

• $Y_{new.i}$ represents the expected lifespan of a new device i in number of years, as presented in Table 3.

The total GHG emission for this life cycle stage are calculated according to the following equation:

$\textbf{(Eq.20)}\ Total\ E_{B.\ new\ device\ production} = E_{Ref} + E_{new\ device\ lifetime\ adjusted}$

where,

• $\ Total\ E_{B.\ new\ device\ production\ \ }$represents the sum of GHG emissions in the baseline scenario new device production life cycle stage, in kgCO$_2$eq.

Avoided GHG emissions

Avoided GHG emissions are calculated by subtracting the sum of the project scenario GHG emissions from the sum of the baseline GHG scenario emissions.

Comparative GHG assessment calculation

The total baseline GHG emissions, total project GHG emissions, and the project's avoided emissions are calculated as follows.

$\begin{aligned}\textbf{(Eq.21)}\ Project\ emissions\ = \ &Total\ E_{P.collection} \\+ \ &Total\ E_{P.waste\ treatment} \\+\ &Total\ E_{P.refurbishing\ process}\end{aligned}$

$\begin{aligned}\textbf{(Eq.22)}\ Baseline\ emissions = \ &{Total\ E}_{B.collection} \\+ &Total\ E_{B.\ waste\ treatment} \\+ &{Total\ E}_{B.\ new\ device\ production\ \ }\end{aligned}$

$\textbf{(Eq.\ 23)}\ Avoided\ emissions\ = \ Baseline\ emissions\ - \ Project\ emissions$

Uncertainty assessment

Uncertainty shall be evaluated at both the methodology level and the project level. The project level uncertainty assessment must consider the uncertainty in the methodology, which is inevitably passed down to each project.

The uncertainty assessment below must be complemented by a project-specific uncertainty assessment. The outcome of the assessment shall be used to determine the percent of avoided emissions to eliminate with the discount factor.

The assumptions that are estimated to have high uncertainty (i.e. high variability and high impact) are:

• The amount of devices avoided in the baseline scenario is proportional to the ratio of new and refurbished device lifetimes (see Assumptions, and Substitutionsections).

• The ratio of new and refurbished device GHG emissions from ADEME can be extrapolated to represent the refurbishing process of all similar devices (seeAssumptionssection).

• The residual economic value of used devices represents the GHG emissions that should be allocated from production Device A first life to the refurbished Device B (seeAssumptionssection).

The assumptions that are estimated to have medium uncertainty are:

• Similar devices have similar characteristics (mass, emission factor, lifetime), leading to grouping devices into device type categories rather than assessing specific device models and brands (see Assumptionssection).

• The distribution of Device B in the baseline and project scenarios is assumed to be the same (see Assumptionssection).

The assumptions that are estimated to have low uncertainty (i.e. high variability and high impact) are:

• Non-functioning parts are assumed to be recycled (seeAssumptionssection).

• The distance for e-waste collection of Device A in the baseline scenario is assumed to be 100 km (seeAssumptionssection).

• Packaging, use, and waste treatment of Device B are assumed to be the same in the baseline and project scenarios (seeAssumptionssection).

The baseline scenario selection has low uncertainty and is mostly standardized. It accounts for project-specific information regarding the number, type and fate of devices, and national e-waste management statistics.

The equations used in this methodology consist of basic conversions and have low uncertainty.

Many estimates and secondary data are used in this methodology to enable a reasonable amount of project data collection. These data have varying levels of uncertainty, and are assessed in Table 6.

The uncertainty at the methodology level is estimated to be medium. This translates to an expected discount factor of at least 6% for projects under this methodology.

Table 6 Presentation of all secondary data and estimates used, and an assessment of their uncertainty.

Parameter	Reference in document	Uncertainty assessment
Emissions new device (kgCO2eq)	Table 3	For Apple devices, an average emission factor for new device production was taken from recent models, and the data samples are presented in Appendix 2. There is low uncertainty in the data samples, since the LCAs come from the manufacturer for the specific model. There is medium uncertainty related to the distribution of these values, where the different devices have coefficients of variation (standard deviation/mean) of 6-36%. For smartphones, the emission factor from ecoinvent has been thoroughly researched and modified to better represent average modern smartphones. Still, there is large variability in smartphone design, and there is high uncertainty in this one representative value. Other devices come from ecoinvent processes and similar to smartphones, have variable designs, so there is high uncertainty in using one representative value.
Impact of refurbished (%)	Table 3	This percentage comes from the detailed ADEME refurbishing LCA. That study uses high quality primary data so the values themselves have low uncertainty.
Average mass (kg)	Table 3	The same analysis can be applied from “Emissions new device (kgCO2eq)”. However this parameter has a smaller impact on the avoided GHG emissions calculations, so the uncertainty can be considered lower.
Lifetime new and refurbished (years)	Table 3	These values come from the detailed ADEME refurbishing LCA and are well within the range of expected lifetimes found elsewhere in the literature. Nonetheless, these estimates have a large impact on the results and are expected to have medium uncertainty.
Market share refurbished vs new (percent)	Table 5	This secondary data is highly influential and is estimated to have low uncertainty for smartphones, where precise data were available for many countries. For other device types, there is medium uncertainty.
Residual value of input devices	Table 4, Appendix 7	This measurement comes from device buyback prices and new device prices. For smartphones and tablets, there is low uncertainty here because buyback data came directly from Project Developers, and a large and representative sample of new device prices was taken. For other device types, the lack of buyback price data leads to medium uncertainty.
WEEE statistics	Appendix 4	These values have medium uncertainty because they come from macro-level national datasets.