Estimating the Dominant Life Phase Concerning the Effects of Battery Degradation on CO₂ Emissions by Repetitive Cycle Applications: Case Study of an Industrial Battery System Installed in an Electric Bus

Abstract

Many studies have evaluated CO₂ emission from batteries. However, the impact of Li-ion battery (LiB) degradation on the CO₂ emissions from the material through operation phases has not been sufficiently examined. This study aims to clarify the dominant CO₂ emission phase and the impact of the degradation of general industrial LiBs from repetitive cycle applications. We developed a model common to general LiB composition and calculated CO₂ emissions by the LCA method using the IDEA database. Our model simplifies the degradation process, including capacity decrease and internal resistance increase. We used it in a sensitivity analysis of the carbon intensity of electricity charged to a LiB. The loss mechanism was determined by experimental data for an electric bus with an industrial LiB. The results illustrate that the carbon intensity of electricity affects CO₂ emissions dominance, the operation phase for mix (65.9%), and the material phase for renewables (77.3%), and that battery degradation over six years increases the total amount of CO₂ emissions by 10.8% for mix and 3.0% for renewables equivalent. Although there are limitations regarding the assumed conditions, the present results will contribute to building a method for monitoring emissions and to standardizing degradation calculations.

1. Introduction

In the context of the widespread use of renewable energy, batteries are playing an important role in balancing the demand and supply of energy. In view of the rapid introduction of large numbers of batteries throughout society, demand for battery-related environmental treatment has become pressing. From the perspective of climate impact mitigation, many regulations concerning carbon footprints (CFPs) are being introduced in Europe and other regions [1]. To contribute to substantial carbon reduction, it is necessary to apportion responsibility for the total emissions related to the whole life phase of a battery regardless of the kind of battery used.

Toward this end, individual industrial fields related to battery use are promoting various measures to reduce CO₂ emissions according to the product characteristics themselves. In the automotive industry, under the progression of the electrification of vehicles, the development of battery-powered and hydrogen-powered vehicles has become a representative trend [2,3]. Additionally, various studies from the viewpoint of eco-driving are being conducted with the aims of reducing fuel consumption [4], optimizing remaining battery power [5], examining the engine energy balance [6], and designing a method for machine-learning-based driving environment classification [7]. For battery-powered vehicles, coordination with EPA/NEDC/WLTP measurement standards is under way to improve the accuracy of CO₂ reduction measurements.

For industrial batteries, IEC 63369 (Methodology for the Carbon Footprint calculation applicable to Lithium-ion batteries) is progressing toward the standardization of product environmental footprint category rules for industrial batteries. In these industries, individual users have not yet begun to disclose such data.

IEC 63369 requires the evaluation of CO₂ emissions caused by production. For batteries used repetitively throughout their life, CO₂ emissions during the operation phase should be a matter of concern. In addition, as charging and discharging are repeated, the inside of the battery becomes degraded. This state of degradation affects the charging and discharging loss of the battery and promotes increased CO₂ emissions. Additionally, the fact that information about the internal degradation of batteries is hardly disclosed to the public makes the calculation difficult.

To address these problems, we constructed a model of an electric bus (E-bus) equipped with a general industrial lithium-ion battery (LiB) based on public information and estimated its CO₂ emission from the extraction and processing of raw materials to operation and considering degradation. To construct a degradation model, we referred to the results of a demonstration experiment of an E-bus equipped with an industrial LiB for which actual data, such as internal resistance variation, had already been obtained. Under this condition, we focused on the losses caused by inevitable battery degradation, with capacity decrease and internal resistance during battery operation. In the operation phase, certain factors may exert considerable influence on the total amount of CO₂ emissions. This study also took account of the sensitivity of these factors.

Even though we here developed a general virtual model, reflecting such characteristic changes due to degradation will be sufficiently useful to evaluate CO₂ emissions. At the same time, this research can provide useful information for standardization studies such as IEC 63369 and will have the positive effect of promoting the acquisition of data on industrial batteries in the operation phase, allowing for more accurate calculations.

This study aims to clarify the impact of the battery degradation of general LiBs in E-buses caused by repetitive cycling on life cycle CO₂ emissions. We develop a general industrial LiB model based on publicly available information to examine the detailed emissions that are closely related to the actual battery conditions, the power conversion loss scenario, and the electricity mix in the operation phase. Economic aspects are excluded from the scope of this paper, which focuses on batteries rather than on vehicles, since we view the economic value obtained by battery installation in terms of the whole vehicles, such as E-buses and locomotives equipped with batteries.

2. Literature Review

The present work seeks to clarify the impact of battery degradation of general LiBs by repetitive cycling on life cycle CO₂ emissions. Within this scope, many studies have reported on CO₂ emissions of lithium-ion batteries (LiBs), determining either the cradle-to-gate impact or the impact at the end-of-life (EoL) phase [8,9,10,11,12].

In the operation phase of a battery system, certain factors may considerably influence the total amount of CO₂ emissions. These are classifiable into internal and external factors [13]. Internal factors include power conversion losses in the battery system and the weight of the battery itself [14]. The electricity losses due to diminishing batter efficiency are considered a significant factor in CO₂ emissions in the operation phase [15]. Battery degradation is another major factor that can influence power conversion performance and result in increased CO₂ emissions. The impact of the degradation-induced CO₂ increase in the operation phase was analyzed based on an operation scenario considering the battery degradation condition [16,17]. However, a review of the literature suggests the impact of degradation is unclear because of a lack of information that could be used to compare efficiency between the initial state and the degraded state. In addition, the absence of a detailed calculation formula makes validity evaluation difficult.

Various verifications are undertaken by the manufacturers of electric automobiles and buses [18,19]. The battery degradation effects on GHG emissions from EV operation were investigated in each state of the United States. [20,21].

The mix ratio of the electricity charged to the battery is one of the external factors. Many studies mention the relationship between the carbon intensity of the electricity charged to the battery in the production phase [22] or the operation phase [23]. So far, it has been pointed out that the unavailability of data from the operation is likely to prevent improvement of the quality of assessment. Therefore, there is a pressing need to conduct a study based on a scenario with sufficient technical rigor reflecting both internal and external battery factors.

From the above, many studies on CO₂ emissions in individual life cycle phases originated from batteries. However, no study has evaluated emissions related to the complete life cycle of a battery, from the extraction and processing of raw materials to operation, while factoring in both degradation and the carbon intensity of the electricity used to charge the battery.

3. Methods

3.1. Conceptual Assumption

The conceptual assumption of this paper is that the degradation of batteries affects CO₂ emissions over the total battery life cycle, excluding the disposal stage. Based on this assumption, the aims of the present work are as follows.

• To estimate the potential CO₂ emissions of individual life cycle phases from a battery installed in an E-bus under the conditions of a practical, technically proven scenario.
• To assess how the operation phase conditions, e.g., battery degradation (an increase in the internal resistance and a reduction in capacity) and electricity mix, affect the total CO₂ emissions per function unit.

3.2. Tool Description and Inventory

LiB performance in an electric bus is assessed by applying the Life-Cycle Assessment (LCA) methodology according to ISO [24,25], which is widely used to assess the environmental performance of battery systems and other products and systems.

The secondary data of carbon emission factors are based on the IDEA ver2.2 database. IDEA is one of the most established and valid databases in Japan. IDEA (Inventory Database for Environmental Analysis), the standard equipment inventory database for life cycle assessment, has been jointly developed since FY2008 by the National Institute of Advanced Industrial Science and Technology (AIST) and the Japan Environmental Management Association for Industry (JEMAI) [26]. The data of IDEA are from Japan, and the statistical information used to formulate the data ver2.2 was published in 2018. For LiNiO₂ and LiPF₆, CO₂ emissions corresponding to the precision of the stoichiometric process were estimated. From these estimations, the emissions of LiNiO₂ and LiPF₆ are calculated individually.

LiNiO₂ is manufactured by means of the sintering reaction of Li₂CO₃ and NiO following Equation (1). The carbon emission intensities of Li₂CO₃, NiO, and O₂ in the IDEA ver2.2 database are applied. Assuming a stoichiometric reaction, the CO₂ emission intensity of LiNiO₂ is calculated as 18.35 kg CO₂/kg. Because the energy consumption data lack clarity, only ΔfH0 is considered. The carbon emissions of ΔfH0 of LiNiO₂ (69.6 kJ/mol) are calculated by converting to LNG.

$$\frac{1}{2}{\mathrm{Li}}_2{\mathrm{CO}}_3+\mathrm{NiO}+\frac{1}{4}{\mathrm{O}}_2\Rightarrow {\mathrm{LiNiO}}_2+\frac{1}{2}{\mathrm{CO}}_2$$

(1)

LiPF₆ is manufactured by means of the reaction of LiF and PF₅ following Equation (2).

$$\mathrm{LiF}+{\mathrm{PF}}_5\Rightarrow {\mathrm{LiPF}}_6$$

(2)

In view of the lack of carbon emission intensity data in the IDEA ver2.2 database for LiF and PF₅, as well as for related submaterials, additional calculations are performed. LiF is manufactured by means of the reaction of Li₂CO₃ and HF following Equation (3). PF₅ is manufactured by means of the reaction of PCl₅ and HF following Equation (4). PCl₅ is manufactured by means of the reaction of P₄ and Cl₂ following Equations (5) and (6). Assuming a stoichiometric reaction, from this condition, the CO₂ emission intensity of LiPF₆ is calculated as 6.52 kg CO₂/kg.

$$\frac{1}{2}{\mathrm{Li}}_2{\mathrm{CO}}_3+\mathrm{HF}\Rightarrow \mathrm{LiF}+\frac{1}{2}{\mathrm{H}}_2{\mathrm{CO}}_3$$

(3)

$${\mathrm{PCl}}_5+5\mathrm{HF}\Rightarrow {\mathrm{PF}}_5+5\mathrm{HCl}$$

(4)

$${\mathrm{PCl}}_3+{\mathrm{Cl}}_2\Rightarrow {\mathrm{PCl}}_5$$

(5)

$$\frac{1}{4}{\mathrm{P}}_4+\frac{3}{2}{\mathrm{Cl}}_2\Rightarrow {\mathrm{PCl}}_3$$

(6)

The CO₂ emission intensity of the average Japanese grid electricity mix of 2015 IDEA (0.573 kg-CO₂/kWh) was used for the electricity “mix”, and 10 MW class solar-based electricity [27] (0.059 kg-CO_2/kWh) was used for “renewables”.

The transportation scenario was assumed to be 700 km by a truck with a carrying capacity of 10 tons in Japan (0.105 kg-CO_2/ton km).

3.3. Analysis Scheme

3.3.1. Boundary

The system boundary of the present work and the individual targeted components are illustrated in Figure 1. The model includes the material (raw material extraction and processing), production (component processing and assembling), transportation, and operation phases. The EoL phase was excluded because of the unavailability of reliable information for modeling.

3.3.2. Assessed Battery Model and Components

The model in this study configures an industrial LiB installed in a WEB-3 E-bus [28]. To assess the LiB installed in the bus, a generic data model common to both the composition and components of the LiB is designed to ensure the generality of the product structure and the reality of the operational scenario. The composition of the cell and module is based on the WEB-3 battery unit, and the details on the LiB components are based on the inventory list obtained from a previous study of a standard LiB inventory dataset [29]. Individual material quantity values of the battery were broken down in proportion to the 45 kWh unit of WEB-3. The configurations of the cell and the module are listed in

Table 1. System specification.

Cell	Rated power capacity (Ah)	40
	Nominal voltage (V)	3.8
	Rated power capacity (Wh)	152
	Volumetric energy density (Wh/L)	143
	Weight energy density (Wh/kg)	72
	Dimensions (mm)	W170 × D47 × H133
	Weight (kg)	2.1
Module	Cell configuration	1 in Parallel × 7 in Series
	Rated power capacity (Wh)	1064
	Weight energy density (Wh/kg)	63
	Weight (kg)	17
Unit	Module configuration	3 in Parallel × 14 in Series (for E-bus)
	Rated power capacity (kWh)	44.68
	Rated current (A)	120
	Service life (year)	6
	Weight (kg)	714

[28,30]. The unit component and individual weights are shown in

Table 2. Unit components of generic data model.

Component	Material	Weight (kg)
Cathode (Positive terminal)	LiNiO2	205.256
	Carbon black	4.729
	Binder (PVDF)	4.729
	NMP	18.681
	Al foil	38.308
	PP	0.071
Anode (Negative terminal)	Graphite	87.967
	Binder (PVDF)	1.892
	NMP	7.804
	Cu foil	72.360
	NiO	2.838
Electrolyte	PC, DEC	47.294
	LiPF6	5.912
Separator	Cellulose	30.978
Gasket	PP	2.365
Insulating plate for plus terminal	PP	0.946
Intermediate mating body	PBT	1.419
Strip	Al	2.578
Outer film	Polyester film	5.770
Safety valve	Al	5.770
Sub-disc	Al	0.733
Cell casing	Al	39.727
Cap	Fe	11.587
Insulation for cap	PP	0.899
Center core	Fe	9.695
Fixing tape	PP	2.128
Insulation for baseplate	PP	1.419
Upper insulation sheet	PP	0.236
Insulation for upper cover	PP	1.182
PTC element	PE	2.128
Parts for module	Cu	14.862
	PP	55.731
	Silicone phenolic resin	26.008
Total		714.000

.

For production and assembly, only electricity was assumed to be used. The amount of electricity was calculated based on the energy consumed in the production of the 45 kWh battery system. The system specifications are shown in

Table 3. Electricity used for production.

	Component/Process	Electricity (kWh)
Cell	Cathode	1046.57
	Anode	1476.15
	Electrolyte	167.04
	Assembly	267.27
Module	Assembly	291.61

. Energy input into the LiNiO₂ process was examined to identify the share of electricity in the production phase.

3.3.3. Functional Units

To quantify the performance of the storage system, two types of functional units are used. One is Wh of delivered electricity over the total service life (from material to operation) of the batteries. The other is per Wh of battery capacity (Wh_bc, cradle to gate). The batteries are assumed to be in operation for six years.

3.3.4. Operation Phase Scenario

The operation phase is analyzed, reflecting the battery losses both with and without factoring in degradation and the electricity mix.

In this paper, we define the power conversion loss as the energy loss due to decreasing system efficiency as a result of charging and discharging during charge and discharge. as is shown in Figure 2. As illustrated at the top of Figure 2, when the battery degrades, internal resistance increases and capacity decreases. When capacity decreases, the charge/discharge cycle number increases, resulting in an internal resistance increase. Since the charge/discharge loss of the battery is calculated from the internal resistance, an internal resistance increase results in an increase in the charge/discharge loss. Battery system loss is divided into three types based on a previous summary of battery efficiency [31]. The values for the degradation of batteries installed in E-buses are obtained from reference values [28]. According to Table 2 of reference [28], the data acquisition period was from 20 December 2011 to 31 January 2014, and it represents the general characteristics of Japan through the three lots of clearly distinguished four-season data. The initial value of the internal resistance was calculated from Figure 6 in reference [30], and the decrease in battery capacity and the increase in internal resistance due to degradation were obtained from Figures 9 and 11 in Reference [28]. We quoted the idea of loss when considering a storage battery as a system from [31].

Figure 2 illustrates the losses of a general battery system on the left and the losses applied to this evaluation for an EV on the right in an image simplified from [21]. These are composed of losses of PCS, auxiliary machines, and the storage device, and the losses are assumed to be equally organized. For the PCS and auxiliary machine losses for an EV shown on the right, the percentage of the battery system on the left was applied to the rated capacity of the E-bus.

• Loss in the storage device (Loss_Sd)
• Loss in the system auxiliary machines (Loss_Aux)
• Loss in the power condition system (Loss_PCS)

Loss_Sd is calculated from rated current and internal resistance. Equation (7) describes the percentage of the Loss_Sd in terms of rated power capacity.

$$\mathrm{Loss}\_\mathrm{Sd}={\mathrm{I}}^2\ast (\mathrm{Rsys}+\mathrm{Rci})/\mathrm{CP}$$

(7)

I: Rated current

Rci: Cable impedance

CP: Rated power capacity

I = 120 (A), CP = 44.68 (kWh) (ref. Table 2)

The value of internal resistance (Rsys) is calculated from internal resistance (Rcell) in accordance with the module configuration using the following Equation (8).

$$\mathrm{Rsys}=\mathrm{Rcell}\ast \frac{Num\_Cell\_series}{Num\_Cell\_parallel}\ast \frac{Num\_Module\_series}{Num\_Module\_parallel}$$

(8)

Rsys: Internal resistance of battery system

Rcell: Internal resistance of battery cell

Num_Cell_series: Number of cells in series in module

Num_Cell_parallel: Number of cells in parallel in module

Num_Module_series: Number of modules in series in unit

Num_Module_parallel: Number of modules in parallel in unit

Rcell is estimated to be 0.00165 (ohm) from the specification of discharge I-V [30] at SOC 50%. Rsys is calculated to be 0.054 (ohm) ($\mathrm{Rsys}=0.00165\left(\mathrm{ohm}\right)\ast \frac{7}{1}\ast \frac{14}{3}=0.054\left(\mathrm{ohm}\right)$). Rci is calculated to be 0.006 (ohm) from the length of the E-bus and the maximum conductor resistance at 20 °C (80 mm² of cab tire cable) [32]. In this study, this calculation applies equally to both a general battery system and an evaluation of an E-bus. Loss_Sd is doubled to account for charge and discharge. From the above setting, Loss_Sd is calculated to be 3.8% ($\mathrm{Loss}\_\mathrm{Sd}={120}^2 \left(\mathrm{A}\right)\ast \left(0.054\left(\mathrm{ohm}\right)+0.006\left(\mathrm{ohm}\right)\right)$/44.68 (kWh) ∗ 2 = 3.8%).

Loss_Aux and Loss_PCS are determined by referring to the values in a previous study [31]. Loss_Aux is set as 4.0%, which is inferred from the double (charge and discharge) ratio of a 2 kW auxiliary loss per 100 kW system. Loss_PCS is set as 7.9% based on the 7.9 kW loss in the 100 kW power system.

From the assumption above, the battery loss without degradation is calculated to be 15.7% as the sum of the values of Loss_Sd, Loss Aux, and Loss PCS.

For the battery degradation scenario, we estimated that degradation causes an increase in internal resistance and a reduction in capacity. Related values were set based on the actual measurement [28], 30% for the range of the state of charge, 148% for the increase in the rate of internal resistance, and 76% for the battery capacity retention rate for 6 years of service. Total yearly discharged electricity (D_Energy) is calculated to be 13.0 MWh based on the estimation of the steady discharge amount (${\mathrm{D}}_{\mathrm{Energy}}=44.68\left(\mathrm{kWh}\right)\ast 30\%\ast 365\mathrm{day}\ast \frac{2.65\mathrm{cycle}}{1000}=13.0\left(\mathrm{MWh}\right)$).

Internal resistance increases 12% per year in response to the degradation scenario, and it increases the value of Rsys. In addition, battery power capacity decreases 6% per year. The voltage change caused by the volume decrease is compensated for by the electric current, and therefore the amount of Loss_Sd increases. Loss_Sd (one way) in the second year is calculated to be 2.4% (Loss_Sd(one way)=[(120 ∗ 1.06)]² (A) ∗ (0.054(ohm) ∗ 1.12 + 0.006 (ohm))/44.68 (kWh) = 2.4%), under the condition that Loss_Aux, and Loss_PCS, and annual total discharged electricity are steady.

Table 4. Degradation calculation.

	Loss	Loss_Sd (One Way)	Charge/Discharge Cycle	Loss (MWh/Year)	Discharge Energy (MWh/Year)	Charged Energy (MWh/Year)
1st year	15.7%	1.9%	2.65	2.0	13.0	15.0
2nd year	16.6%	2.4%	2.81	2.2	13.0	15.2
3rd year	17.7%	2.9%	3.00	2.3	13.0	15.4
4th year	18.9%	3.5%	3.21	2.4	13.0	15.7
5th year	20.2%	4.1%	3.46	2.6	13.0	16.0
6th year	21.7%	4.9%	3.74	2.8	13.0	16.4
Sum of 6 years				14.4	77.8	93.8
				Loss/Discharge	18.5%

summarizes the degradation calculation during the 6-year service lifetime. The ratio of electricity loss per unit of discharged electricity is calculated to be 18.5% in 6 years. This is a 17.4% increase compared to the value of 15.7% (=Loss/Discharge energy) with no consideration of degradation. The carbon emissions from the battery during operation are calculated by means of the emissions equivalent to the power conversion losses. The carbon emission factors of the average Japanese grid electricity mix and the predominance of renewable energy are the same in the operation phase as in the production phase.

4. Results

4.1. Material to Production Phase

The results of CO₂ emissions in the material to production phases are illustrated in

Table 5. CO₂ emissions of material to production phases.

Components	Wt% of Total Battery System	CO2 Emission
		Mix			Renewable
		%	kg-CO2	kg-CO2/wh_bc	%	kg-CO2	kg-CO2/wh_bc
Cathord	38.1%	52.9%	4480.6	99.6	65.9%	4480.6	99.6
Anode	24.2%	9.9%	838.6	18.6	12.3%	838.6	18.6
Electrolyte	7.5%	1.6%	134.1	3.0	2.0%	134.1	3.0
Separator	4.3%	0.3%	26.2	0.6	0.4%	26.2	0.6
Cell casing	5.6%	7.8%	663.7	14.8	9.8%	663.7	14.8
Other component for cell	6.8%	3.0%	256.1	5.7	3.8%	256.1	5.7
Parts for module	13.5%	2.5%	209.3	4.7	3.1%	209.3	4.7
Energy for cell manufacture	-	20.0%	1693.9	37.7	2.6%	174.5	3.9
Energy for module manufacture	-	2.0%	167.0	3.7	0.3%	17.2	0.4
ToTal	100%	100%	8470	188.3	100%	6609	151.2

* kg of battery unit.

. CO₂ emissions per Wh of battery capacity are calculated to be 188.3 g-CO₂/Wh_bc for mix and 151.2 g-CO₂/Wh_bc for renewables.

The CO₂ emissions per battery unit storage capacity are illustrated in Figure 3. CO₂ emissions per FU are 191 g-CO₂/Wh_bc for mix and 153 g-CO₂/Wh_bc for renewables. These results indicate that the carbon intensity of electricity production affects cradle-to-gate CO₂ emissions.

4.2. Total CO₂ Emissions

The total lifetime CO₂ emissions are illustrated in Figure 4. CO₂ emissions are 22.5 t-CO₂ for mix, and 8.3 t-CO₂ for renewables without degradation, and 25.0 t-CO₂ for mix, and 8.5 t-CO₂ for renewables with degradation. These results reveal that the dominant CO₂ emission phases are the material phase for renewables and the operation phase for mix. The ratio of CO₂ emission per total life cycle CO₂ of the dominant phase are 65.9% for mix and 77.3% for renewables with degradation, and 62.2% for mix and 79.6% for renewables without degradation. These results indicate that determination of the dominant CO₂ emission phase depends on the carbon intensity of electricity.

The CO₂ emissions per total lifetime of delivered electricity are illustrated in Figure 5. CO₂ emissions per FU are 0.145 g-CO₂/Wh for mix, and 0.053 g-CO₂/Wh for renewables without degradation, and 0.161 g-CO₂ for mix, and 0.055 g-CO₂ for renewables with degradation. These results indicate that battery degradation during the six years contributes to the total amount of CO₂ emissions, the increases are 10.8 % for mix and 3.0% for renewables equivalent to the charge and discharge power. It was clarified that the impact of degradation on CO₂ emissions, which is 3.0% of total life cycle CO₂ even in the case of renewable energy, or 17.4% of operation phase CO₂, cannot be ignored. These results are similar to those in a previous study [21].

Regarding the CO₂ emissions of individual phases over the total life cycle, the present study revealed that the carbon intensity of electricity affects the life phase that is dominant in terms of CO₂ emissions. Additionally, these emissions are associated with power conversion losses considering the impact of degradation.

5. Discussion

5.1. Overviews

To assess CO₂ emitted from the LiB life cycle, we developed an inventory analysis model of common to general LiB composition. CO₂ emissions per battery weight and Wh_bc have a wide range determined by the stoichiometric precision from the material to production phases. This is because of the differences in the composition of anode and cathode and in the solvent compositions of batteries, as well as the system boundary for evaluation. In the literature review of greenhouse gas (GHG) emissions per kWh of battery capacity relating to the production phase [10], GHG emissions per kWh battery capacity can range from 53 to more than 300 kg CO₂eq/kWh_bc for the different technologies and depending on the age of the studies.

This study focuses on the loss mechanism of battery degradation caused by decreasing battery capacity and increasing internal resistance as well as the carbon intensity of electricity charged to a LiB. The results reconfirmed that degradation during a LiB’s six years of service contributes to the total amount of CO₂ emissions by 10.8% for mix and 3.0% for renewables equivalent to the charged and discharged power.

In summary, we evaluated the life cycle CO₂ emissions of a LiB installed in an E-bus from battery raw material extraction and manufacturing through to operation and revealed that the dominant phase for CO₂ emissions changes depends on the carbon intensity of electricity. A comparison between the present study and previous studies using different degradation models [21] showed that similar values of CO₂ emissions were obtained. We also confirmed the validity of our simple model. The results highlight the importance of CO₂ emissions derived from battery degradation. This contributor to emissions should not be ignored even in the emerging era of renewable energy.

5.2. Limitations

5.2.1. Limitations to CO₂ Calculations

The limitations in our CO₂ calculations are listed below.

• Limitation of primary data: It was difficult to acquire upstream information in the supply chain (data on electrode material manufacturing, etc.)
• Limitations due to the LCA unit database associated with the use of secondary data: There are two limitations to the IDEA data: it was formulated in 2018 so it is not the latest one, it was only used for evaluation in Japan.
• Only two types of carbon intensity of electricity are used: the average power source in Japan and typical PV values. Regional differences are not taken into account.
• We relied on hypothetical rather than actual electricity power configurations supplied to the battery.
• There is little information on the disposal/recycling stage. Although our evaluation this time did not include this phase, the impact of recycling should be taken into consideration in order to fully elucidate CO₂ emissions for the whole life cycle of a battery, along with the improvement of data accumulation specific to that end stage.

5.2.2. Limitation Regarding Battery Condition Management

The limitation of this study regarding battery condition management is as follows:

• There is only a small amount of available data on internal resistance and its change as the battery degrades.

There are few published data that can simultaneously refer to changes in battery capacity and internal resistance. In the present study, we designed the model by combining several published cases of different batteries. We believe that the accuracy of this research will improve when relevant data become available.

5.3. Further Steps

5.3.1. Further Steps in CO₂ Calculations

The following are listed as further steps for CO₂ calculation.

• Practical use of the primary database that can be used to trace a supply chain (utilization of data linkage frameworks such as Digital Product Passport).
• To ensure reduction in CO₂ emissions, a methodology must include monitoring during operation in accordance with the principles of MRV (measuring, reporting, and verifying) methodology for implementing energy efficiency measures [33]. In addition, the present results can provide useful information for the standardization of the calculation method for CO₂ emissions over the entire life cycle of IEC 63369, which is being studied for industrial batteries.
• Utilization of the Global LCA Data Access network (GLAD), that is the largest directory of LCA datasets to achieve wide usage of LCA through better accessibility and interoperability of LCA data [34].
• Evaluation of the disposal/recycling stage.

5.3.2. Further Steps for Battery Condition Management

The following is a further step for battery condition management.

• Rule formulation for monitoring KPIs (Key Performance Indicators) to identify battery conditions such as internal resistance and capacity reduction concerning battery degradation.

This study clarified that the carbon intensity of the grid electricity for charging and discharging affects total CO₂ emissions. It is necessary to build a system to collect data on such changes in the power source composition along with the KPIs of battery degradation.

LiBs are utilized for diverse purposes, not just for mobility but also in the stationary power sector for both normal and emergency situations. Consequently, there are many types of scenarios to consider, ranging from short-term frequency disturbance to relatively long-term hourly or daily balance. To properly evaluate battery performance depending on several usage scenarios and conditions, a methodology should be established that can be tailored to individual requirements.

Comparison with multiple kinds of actual data and analysis of the impact of differences in the operation phase would be required for statistical analysis, but the future accumulation of battery degradation data is awaited. Regarding the creation of systems that will make it possible to obtain data in the future, it is assumed that disclosure will start from batteries for industrial systems using the data linkage framework. It is also possible to develop a battery degradation calculation model by matching simulation results from our proposed simplified model with actual information based on measurement standards. The results of this research will surely contribute to the construction of such future systems.

6. Conclusions

In the present work, an LCA study was performed to ascertain the life cycle CO₂ emissions of a LiB installed in an E-bus, focusing on phases from material to practical operational scenarios that reflect power conversion losses and degradation conditions as internal factors, while considering the carbon intensity of electricity charged to the LiB as an external factor.

Regarding the CO₂ emissions of individual phases over the total life cycle, our results revealed that the carbon intensity of electricity affects the life phase that is dominant in terms of CO₂ emissions. Additionally, these emissions are associated with power conversion losses considering the impact of degradation.

The results also illustrate that that the carbon intensity of electricity affects which life phase will dominate CO₂ emissions: the operation phase is dominant for mix (65.9%) while the material phase is dominant for renewables (77.3%); and degradation during the battery’s six years of service increases the total amount of CO₂ emissions by 10.8% for mix and 3.0% for renewables equivalent to the charge and discharge power. These results highlight the fact that a real and precise battery condition scenario during operation contributes to the accuracy of CFP calculations and the fact that improving the resilience of battery performance will be indispensable to reducing total CO₂ emissions in the coming era, in which renewable energy is expected to predominate.