Real-Driving Emissions of an Aging Biogas-Fueled City Bus

Abstract

Transition to low emission transportation and cleaner cities requires a broad introduction of low- and zero-carbon alternatives to conventional petrol- and diesel-powered vehicles. New-generation gas buses are a cost-effective way to reduce local air pollutants from urban transportation. Moreover, major greenhouse gas (GHG) savings may be achieved using biogas as the power source. The main objective of this research was to investigate CH₄ and other gaseous emissions of a biogas-fueled urban bus equipped with a three-way catalyst (TWC) in real-world conditions. The study focused on emissions from a six-year-old gas-powered city bus, supplementing emission data from aging bus fleets. Impaired CH₄ oxidation and NO_x reduction were observed in the catalyst after its service life of 375,000 km–400,000 km. The main reason for low CH₄ and NO_x conversion over the TWC was concluded to be the partial deactivation of the catalyst. Another critical issue was the fluctuating air-to-fuel ratio. The results show that the efficiency of exhaust after-treatment systems should be closely monitored over time, as they are exposed to various aging processes under transient driving conditions, leading to increased real-world emissions. However, the well-to-wheels (WTW) analysis showed that an 80% GHG emission benefit could be achieved by switching from diesel to biomethane, giving a strong environmental argument for biogas use.

1. Introduction

There is a worldwide consensus that significant reductions in greenhouse gas (GHG) emissions are needed to avoid the worst impacts of climate change, and various laws and regulations have already been implemented to combat and respond to global warming. In July 2021, the European Commission adopted an extensive legislative package, Fit for 55, with the goal of reducing the economy-wide GHG emissions by at least 55% by 2030 compared to 1990 levels [1]. This is a substantial increase from the previous 40% target. Achieving the 55% reduction in GHG emissions over the next decade is crucial for Europe to achieve climate neutrality by 2050. Moreover, Finland has set itself the goal of becoming carbon neutral by 2035 [2]. This is one of the most ambitious targets of any country in the industrialized world.

In 2019, GHG emissions from domestic transportation accounted for 21 percent of Finland’s total greenhouse gas emissions and about 30 percent of the energy sector’s GHG emissions [3]. Road transportation is likely to remain a significant contributor to air pollution in the coming decades, especially in urban areas [4]. Transition to low emission transportation and cleaner cities will undoubtedly require a broad introduction of low- and zero-carbon alternatives to conventional petrol- and diesel-powered vehicles.

New generation gas buses are a cost-effective way to reduce CO₂ and local pollutants from urban transportation. Fueling with gas reduces pollutant emissions, including carbon monoxide (CO), nitrogen oxides (NO_x), and particulate matter (PM), as shown, e.g., by Biernat et al. [5]. Moreover, major GHG savings can be achieved by using biogas as the power source. This is based on the fact that producing biomethane from organic waste material results in fuel that contains only biogenic carbon, and combustion of such fuel releases only biogenic CO₂, which is, unlike CO₂ from fossil fuels, not considered to contribute the climate change [6].

Buses running on biogas are becoming more common in Finland as cities and transportation companies invest in greener alternatives. For example, in the western coastal city of Vaasa, biogas buses have been touring since 2017. Life cycle GHG emissions from biogas vehicles largely depend on the extent of methane (CH₄) leakage throughout the fuel life cycle, and unintended CH₄ emissions from different stages of the fuel chain can narrow their potential climate benefits. Methane is a powerful greenhouse gas with a global warming potential (GWP) 28–34 times that of CO₂ over a 100-year timescale [7]. In addition, due to the strong C–H bonds of methane, it is one of the most difficult hydrocarbons to treat catalytically [8], and insufficient removal rates of exhaust after-treatment systems at low loads and low exhaust temperatures may lead to increased real-world CH₄ emissions [9].

Besides exhaust gas temperature, another critical issue is the effect of rapid changes in exhaust gas composition—typical in real-world driving conditions—on after-treatment devices. This phenomenon is particularly evident when dealing with stoichiometric gas engines using three-way catalytic converters (TWC), requiring a very precise control of air-to-fuel ratio (AFR), as some deviations from the stoichiometric lambda value can interfere with the catalyst efficiency [10]. For example, Rodman Oprešnik et al. [11] reported instantaneous, local rises of THC emissions as a result of occasional inadequate lambda control of a CNG bus during transient regime and, consequently, increased cumulative emissions.

The main objective of this research was to investigate CH₄ and other gaseous emissions plus fuel consumption of a biogas-fueled urban bus in real-world operation. The actual driving emissions were recorded using a portable emissions measurement system (PEMS). The key advantage of on-board measurements is that they can truly demonstrate the emission characteristics of vehicles under various traffic conditions, operating cycles, and ambient conditions, including those that are challenging to replicate in the laboratory, such as varying road gradients [4]. The load on the lines that buses serve and the number of passengers may also affect exhaust emissions under actual traffic conditions [12].

Exhaust emissions under real-world conditions were examined by Lv et al. [13]. The authors showed an underestimation of road emissions of gas- and diesel-powered heavy vehicles; emission factors under real-driving conditions were significantly higher than in previous chassis dynamometer studies, likely caused by frequent accelerations, decelerations, and start-stop operation. In a recent study, Rosero et al. [14] investigated the effects of passenger load, road grade, and congestion level on real-world emissions and fuel consumption of urban Euro VI CNG and Euro V diesel buses. As the road grade and congestion level increased, both buses’ fuel consumption and CO₂ emissions increased by 6–55%. Gallus et al. [15] studied the impact of driving style and road grade on gaseous exhaust emissions of Euro V and Euro VI diesel vehicles. CO₂ and NO_x emissions, measured with PEMS, showed a linear increase with road grade. Chen et al. [16] investigated the impact of speed and acceleration on emissions of heavy-duty (HD) vehicles in Shanghai. They found that congestion conditions with low speed and frequent deceleration and acceleration increased THC and CO emissions. Ozener & Ozkan [17] reported that the acceleration effect on both fuel consumption and emission values was significant. They concluded that the real-driving emission data could be effectively used in developing cleaner engine calibrations and more economical operations.

In addition, gaseous emissions are strongly affected by starting conditions. The cold-start emissions challenge has been highlighted, e.g., in [18,19]. During the first minutes of operation, emissions are high because the after-treatment equipment has not reached the appropriate temperature required to efficiently remove gaseous pollutants. Faria et al. [20] also showed a substantial increase in energy consumption for cold-start, leading to increased CO₂ emissions during the cold-start period. The problem of cold-starts is considered more pronounced at low ambient temperatures, as lower ambient temperature increases the cold-start running duration [20,21].

One crucial topic rarely addressed in real-driving emissions (RDE) studies is the catalyst deactivation and deterioration over time. Indeed, the presence of catalyst poisons and other impurities in the feed, the fluctuating exhaust gas composition and flow rate in the converter, as well as high temperatures and temperature gradients, all increase the possibility of catalyst deactivation [22]. Therefore, to ensure a significant reduction of emission levels throughout the vehicle’s useful life, EU regulation has adopted dedicated “emission durability” periods, i.e., the minimum mileage or time after which the engine is still expected to comply with applicable emission limits. For example, for category M3 buses, the required emission durability period is six years or 300,000 km, whichever comes first [23]. However, the useful life of urban buses is usually much longer; e.g., the Finnish bus fleet’s average age is 12.5 years [24]. Therefore, emission levels after the emission durability period and closer to the service life of the vehicles need to be investigated.

This study focused on emissions from a six years old gas-powered city bus, supplementing emission data from aging bus fleets. PEMS measurements were performed in real-traffic conditions on a regular bus line in Vaasa in collaboration with the University of Vaasa and RISE Research Institutes of Sweden. In addition to methane emissions, gaseous emissions of NO_x, CO, and CO₂ were measured. Both cold-start and warm-engine emissions were recorded. We conducted two measurement campaigns, the first in March 2022 and the second in June 2022. In addition, the total carbon footprint of compressed biogas (CBG) is discussed in terms of its GHG reduction potential, defined as the percentage reduction in life cycle GHG emissions relative to its fossil counterpart natural gas and traditional diesel fuel.

2. Materials and Methods

2.1. Test Vehicle

Exhaust emission tests in real-driving conditions were carried out on a Scania Euro VI bus owned by the City of Vaasa and operated by Wasa Citybus. The CBG-fueled bus was equipped with a spark ignition engine with a displacement of 9.3 dm³ and a power of 206 kW. The vehicle was equipped with exhaust gas recirculation (EGR) and a three-way catalytic converter.

Table 1. Vehicle technical specifications.

Parameter	Value
Model name	Scania Citywide LE
Model year	2016
Gross vehicle weight (kg)	19,100
Curb weight (kg)	12,960
Max passenger number	75
Axle configuration	4 × 2
Gearbox	6-speed automatic transmission
Accumulated mileage (km)	375,000 (Test 1), 400,000 (Test 2)
After-treatment system	TWC
Other systems	EGR
Exhaust emission norm	Euro VI-C

presents the characteristics of the test vehicle and

Table 2. Engine technical specifications.

Parameter	Value
Model	Scania OC09 101
Engine type	Spark ignition engine
Fuel	CNG/CBG
Number of cylinders	5
Compression ratio	12.6:1
Total displacement (L)	9.3
Maximum power (kW@rpm)	206 kW@1900 rpm
Engine peak torque (Nm@rpm)	1350 Nm@1000–1400 rpm

summarizes the engine technical specifications.

2.2. Portable Emissions Measurement System

The real-driving gaseous emissions of CH₄, CO, CO₂, NO, and NO₂ from the tested city bus were measured and recorded using an on-board VARIOplus Industrial device manufactured by MRU Messgeräte für Rauchgase und Umweltschutz GmbH. VARIOplus measures CH₄, CO, and CO₂ concentrations using a non-dispersive infrared (NDIR) sensor, and NO_x concentrations are measured using electrochemical cells.

Table 3. Technical characteristics of VARIOplus Industrial.

Parameter	Measurement Method	Accuracy
CH4	NDIR—Non-dispersive infrared, range 0–10,000 ppm	±2%
CO	NDIR—Non-dispersive infrared, range 0–10%	±0.03% or * ±3% reading
CO2	NDIR—Non-dispersive infrared, range 0–30%	±0.05% or * ±3% reading
NO	electrochemical, range 0–1000 ppm	±5 ppm or * 5% reading
NO2	electrochemical, range 0–200 ppm	±5 ppm or * 5% reading
O2	electrochemical, range 0–10%	±0.2 Vol-% abs.
Sampling	1 Hz

* = whichever is larger.

shows the technical characteristics of the measurement apparatus used in this work.

The engine speed, torque, coolant temperature, air flow, lambda, and the vehicle speed were recorded from the vehicle engine control unit (ECU) via an on-board diagnostics (OBD) system using Scania Diagnosis & Programmer (SDP3) software version 2.50.3 (in Test 1) and version 2.52.1 (in Test 2), copyright Scania CV AB, Scania Suomi Oy, Vaasa, Finland. The vehicle’s position in terms of latitude, longitude, and altitude, and the vehicle speed data were registered using an external global positioning system (GPS). A dedicated weather station was used to register the ambient temperature, pressure, and relative humidity. The real-world emission data obtained with PEMS and the GPS and the weather data were collected and stored with the DEWESoft data acquisition system. All data were recorded with a frequency of 1 Hz. Prior to the data processing, the SDP3 and DEWESoft data were synchronized based on the vehicle speed from the ECU and the GPS.

An external power unit supplied the electrical power to the PEMS system. Figure 1 depicts the system set-up.

2.3. Test Route

Emission tests were performed in real-driving conditions on an urban route in Vaasa, i.e., in normal traffic and with normal driving patterns and typical passenger loads. The selected test route was the same route the bus usually travels daily. The measurements started in the morning at the same time and the same driver from Wasa Citybus was used in both measurement campaigns. Figure 2 shows the driving circuit chosen for the tests. The length of one circuit was 25.5 km, and the same circuit was run three times. The total test duration was approx. 3 h. The route included both urban and rural driving. The speed profile of the driving circuit is presented in Figure 3.

Table 4. Shares of driving speed ranges.

	Speed Range	Time (min)	%	Mean Velocity
Urban driving	0–30 km/h	102	56	12
Urban driving	30–50 km/h	65	36	38
Rural driving	50–75 km/h	16	9	57
Total		182		25

shows the percentages and mean velocities for three different driving speed ranges. The passenger load varied between 5 and 30 percent during the tests.

The first measurement campaign was performed in March 2022, and the second in June 2022. In June, only warm engine measurements were recorded, while in March, both cold-start and hot-start emissions were investigated.

2.4. Fuel

The fuel used in the test was CBG from a commercial filling station. The methane content of the fuel was 97% by volume. The other main components of the fuel were CO₂ (2.2 vol.-%), nitrogen (0.5 vol.-%), and oxygen (0.3 vol.-%), so the energy content of the fuel was solely related to the methane concentration. The calculated lower heating value (LHV) of the gas was 46.4 MJ/kg.

2.5. Calculation Procedure

2.5.1. Calculation of Fuel Mass Flow

The instantaneous fuel flow (ṁ_fuel) in kg/s was calculated based on the recorded instantaneous air flow (ṁ_air) and lambda (λ) values and the stoichiometric air-to-fuel ratio (AFR_stoich), according to Equation (1).

$${\mathrm{ṁ}}_{fuel}=\frac{{\mathrm{ṁ}}_{air}}{AF{R}_{stoich}\times \lambda }$$

(1)

To determine AFR_stoich, the stoichiometric oxygen demand (n_O2,stoich) in moles per kg of fuel was calculated first, based on the chemical composition of the fuel (Equation (2)). In the equation, w_c, w_H2 and w_O2 are the fuel mass fractions of carbon, hydrogen and oxygen in the fuel.

$$n_{O2, stoich=}\frac{w_c}{0.012011}+\frac{1}{2}\times \frac{w_{H2}}{0.002016}-\frac{w_{O2}}{0.031999}$$

(2)

As air contains 20.95% of oxygen, the stoichiometric air demand (n_air,stoich) in moles per kg fuel could be determined by Equation (3):

$$n_{air,stoich}=\frac{n_{O2,stoich}}{0.2095}$$

(3)

Finally, the stoichiometric air demand in kg of air per kg of fuel was calculated by multiplying n_air,stoich by the molar mass of air (M_air), see Equation (4):

$$AF{R}_{stoich}=n_{air,stoich}\times M_{air}$$

(4)

2.5.2. Calculation of Fuel Consumption

The total fuel mass (ΣFC_i) over the test cycle was calculated based on the instantaneous (second-by-second) fuel mass flows according to Equation (5).

$$\Sigma F{C}_i=\left(\frac{1}{2}{\mathrm{ṁ}}_{fuel,0}+{\mathrm{ṁ}}_{fuel,1}+{\mathrm{ṁ}}_{fuel,2}+\dots +{\mathrm{ṁ}}_{fuel,n-1}+\frac{1}{2}{\mathrm{ṁ}}_{fuel,n}\right)$$

(5)

2.5.3. Calculation of Exhaust Mass Flow

The instantaneous exhaust gas mass flow (ṁ_exh.,i) (wet basis) in kg/s was determined based on the recorded air flow and the calculated fuel flow values (Equation (6)):

$${\mathrm{ṁ}}_{exh.,i}={\mathrm{ṁ}}_{air,i}+{\mathrm{ṁ}}_{fuel,i}$$

(6)

2.5.4. Emissions Dry–Wet Correction

The emission concentrations were measured on a dry basis. Dry concentration (c_dry) was converted to a wet basis with the dry–wet conversion factor (K_d–w):

$$c_{wet}=K_{d–w}\times c_{dry}$$

(7)

K_d–w was calculated according to the UN/ECE Regulation 49 [25], Equation (8):

$$K_{d-w}=\left(\frac{1}{1+a\times 0.005\times \left(c_{CO2}+c_{CO}\right)}-k_{w1}\right)\times 1.008$$

(8)

where a is the molar hydrogen to carbon ratio of the fuel, and

$$k_{w1}=\frac{1.608\times H_a}{1000+\left(1.608\times H_a\right)}$$

(9)

where H_a is the intake air humidity in g water per kg dry air.

2.5.5. Calculation of Mass Emissions

Second-by-second mass flow of the pollutant (ṁ_gas) in g/s was calculated using Equation (10):

$${\mathrm{ṁ}}_{gas}=u_{gas}\times c_{gas}\times {\mathrm{ṁ}}_{exh.}$$

(10)

where u_gas is the ratio between the density of pollutant and the density of exhaust gas, and c_gas is the instantaneous concentration of the pollutant in raw exhaust in ppm (wet basis). The instantaneous u values were calculated following the UN/ECE Regulation No 49 [25], according to Equations (11)–(14):

$$u_{gas,i=\frac{{\rho }_{gas}}{\left({\rho }_{exh.,i}\times 1000\right)}}$$

(11)

$${\rho }_{gas}=\frac{M_{gas}}{22.414}$$

(12)

where M_gas is the molar mass of the gas component in g/mol, ρ_gas is the density of the gas component in kg/m³, and ρ_exh.,i the instantaneous density of the exhaust gas in kg/m³, derived from Equation (13):

$${\rho }_{exh.,i}=\frac{1000+H_a+1000\times \left(\frac{{\mathrm{ṁ}}_{fuel,i}}{{\mathrm{ṁ}}_{dry air,i}}\right)}{773.4+1.2434\times H_a+k_{fw}\times 1000\times \left(\frac{{\mathrm{ṁ}}_{fuel,i}}{{\mathrm{ṁ}}_{dry air,i}}\right) }$$

(13)

where k_fw is the fuel specific factor of wet exhaust, obtained from Equation (14):

$$k_{fw}=0.055594\times W_{\alpha }+0.0080021\times W_{\Delta }+0.0070046\times W_{\epsilon }$$

(14)

where W_α is the hydrogen content (wt%) of the fuel, W_∆ the nitrogen content (wt%), and W_ε the oxygen content (wt%) of the fuel.

The mass of gaseous emissions (m_gas) in grams per test cycle was calculated using Equation (15).

$$m_{gas}={{\displaystyle \sum }}_{i=1}^{i=n}{u}_{gas,i}\times c_{gas,i}\times {\dot{m}}_{exh.,i}\times \frac{1}{f}$$

(15)

where f is the data sampling rate in Hz.

The final results are expressed in g/kWh and in g/km, i.e., the total mass of each pollutant over the test cycle was divided by the engine cycle work or by the distance covered in km.

2.5.6. Calculation of Cycle Work

The engine work (W_i) in kWh over the test cycle was calculated based on the instantaneous (second-by-second) engine power values (P_e), according to Equation (16):

$$W_i=\frac{\left(\frac{1}{2}{P}_{e,0}+P_{e,1}+P_{e,2}+\dots +P_{e,n-2}+P_{e,n-1}+\frac{1}{2}{P}_{e,n}\right)}{3600 }$$

(16)

2.5.7. Calculation of Effective Power of the Engine

The instantaneous engine power in kW was calculated by using each pair of recorded engine speed and torque values (Equation (17)):

$$P_e=\frac{2\times \mathsf{\pi }\times N\times \tau }{60\times 1000}$$

(17)

where N is the engine speed in rpm and τ is the engine torque in Nm.

3. Results and Discussion

3.1. Ambient Conditions

Table 5. Ambient conditions during the tests.

Ambient Condition	Test 1	Test 2
	March 2022	June 2022
Temperature (°C)	−5 °C	+18 °C
Pressure (kPa)	102.5	100.5
Humidity (%)	65.5	54.7

summarizes the average ambient conditions during the tests.

3.2. Gaseous Emissions

In the current legislation, the regulatory in-service conformity (ISC) emission test applies the 20% power threshold as a boundary condition for Euro VI-C bus engines. However, Mendoza Villafuerte et al. [26] showed that a large fraction of urban operation is not considered if the current power threshold boundary for post-processing the PEMS data is applied, and up to 80% of the data may be excluded from the emission analysis. They also showed that cold-start emissions, which are currently also excluded from the analysis, could account for a significant proportion of total emissions. To give a more accurate depiction of real-driving emissions, no power threshold boundaries were applied in this study. In addition, in Test 1, both cold-start and hot-start emissions were recorded. In Test 2, unfortunately, only hot-start emissions were successfully recorded.

3.2.1. Hot-Start Emissions

A test was considered a hot-start once the coolant temperature had reached 70 °C for the first time or stabilized within ±2 °C over a period of 5 min, whichever came first [27]. Specific emissions were calculated in both g/kWh and g/km, and the results are presented separately for the total trip and for urban and rural sections of the circuit (Figure 4). Although the tests performed did not fully reflect the ISC tests in the type-approval procedure regarding boundary conditions and route requirements, the Euro VI standard limits (ISC limit) are also presented for comparative purposes.

CO emission values were low and well below the ISC limit of 6 g/kWh in both tests, indicating efficient oxidation of CO in the catalyst. In contrast, relatively high values were observed for CH₄ and NO_x, indicating impaired CH₄ oxidation and NO_x reduction in the catalyst after its service life of 375,000 km (Test 1). After 400,000 km (Test 2), the catalyst efficiency had further deteriorated. Here, it should be noted that according to EU Regulation EC 595/2009 [23], the minimum mileage or time after which the engine is still expected to comply with applicable emission limits for category M3 buses, is 300,000 km or six years, whichever comes first. Hence, the required “emission durability” period had already been exceeded in our case. Nevertheless, the bus has passed the regular technical inspections valid in Finland, including CO₂ and HC measurements.

The primary reason for relatively high CH₄ and NO_x emissions after the TWC was assumed to be the low CH₄ reactivity due to a partial deactivation of the catalyst. In addition to the low CH₄ oxidation rate, low CH₄ reactivity also means that methane-based reducing agents for NO_x reduction do not work, leading to substantial NO_x breakthrough from the catalyst, also concluded by Van den Brink & McDonald [28].

One of the most important reasons for the deactivation of the TWC in automotive applications is chemical deactivation [29], mainly caused by lubricating oil additives and other impurities in the exhaust gases. For example, Winkler et al. [30] observed a significant increase in hydrocarbon emissions during CNG operation over a relatively short TWC lifetime of 35,000 km. Contaminants originating from the lubricating oil, such as calcium, phosphorus, and magnesium, detected on the catalyst’s surface, appeared to affect especially CH₄ oxidation. In addition to lubricating oil, another source of catalyst poisons is the impurities in the fuel. The CBG used in this study contained small traces of commonly encountered catalyst poison sulfur (<2.3 mg/Nm³) and siloxanes (0.7 mg/Nm³). Although the amounts of these compounds were very low, they could have had a deactivating effect on the emissions control system.

Furthermore, the light-off of a TWC in gas-fueled engine exhaust typically occurs at higher temperatures compared to gasoline engines [31]. Indeed, methane is the most difficult hydrocarbon to oxidize due to its high stability [8]. A typical light-off temperature for methane is 400 °C [8], but in a deactivated catalyst, significantly higher temperatures, up to 500–600 °C [32], may be required to break the strong C–H bonds in methane. At low loads (Figure 5), common in a city bus’s driving profile, the exhaust gas temperature was too low to allow the deactivated catalyst to work effectively.

Thus, restoring the catalytic activity of a deactivated TWC is a critical consideration. In some cases, depending on the adsorbed poison, the activity of the poisoned catalyst can be at least partially restored by regeneration [22]. For example, SO₂ can be removed from the catalyst under elevated temperatures and anoxic or very rich conditions, as shown by Auvinen et al. [32]. Careful control of the exhaust gas composition during regeneration could provide significant benefits in terms of CH₄ emissions. However, under real-driving conditions, the rapidly and dramatically varying exhaust gas temperature and composition between oxidizing and reducing environment make the on-board regeneration difficult to control.

Another possible deactivation mechanism for the TWC is thermal degradation. Three-way catalysts are known to lose their activity when exposed to high temperature (>800 °C) oxidizing environments, typically occurring during fuel shut-off phases [33]. Switching off the fuel flow, e.g., during engine braking, is a strategy of the automotive industry to improve fuel economy. Thermal degradation is critical to the catalyst’s performance since these changes are typically irreversible.

In addition to the partial deactivation of the catalyst, another probable reason for the relatively high emissions was the fluctuating lambda value. Indeed, close control of the exhaust gas composition is essential for high emission conversion as the composition of the gas entering the TWC significantly affects its catalytic efficiency [34]. For simultaneous conversion of HC, CO, and NO_x species in the TWC, the engine must be operated within a very narrow AFR window—near stoichiometric conditions—due to a rapid drop in NO_x conversion efficiency on the lean side and a non-complete conversion of hydrocarbons both in lean and rich stoichiometry [10]. For example, Lou et al. [34] detected the highest TWC conversion efficiency when AFR was controlled between 0.995 and 1. The narrow AFR range over which significant conversion of natural gas exhaust emissions is possible presents a challenging control problem. As seen in Figure 6, lambda was outside the optimal range for a significant part of the time in our experiments.

In sum, deterioration of the exhaust after-treatment systems over time should be monitored as they are exposed to different aging processes resulting in elevated real-world emissions. Our results indicate a catalyst replacement need after 375,000 km of service life. In addition, a precise lambda control is absolutely necessary to ensure high conversion rates throughout the vehicle’s lifetime.

3.2.2. Cold-Start Emissions

Cold-start emissions were recorded from the moment the coolant temperature had reached 30 °C for the first time and continued until the coolant temperature was stabilized within ±2°C over 5 min [27]. In Test 1 (at −5 °C), the cold-start period lasted 14.5 min. The combined cold- and hot-start emissions were calculated according to EU Regulation 1718 [35]: the vehicle was driven over a cold-test cycle followed by nine hot-test cycles, identical to the cold one in a way that the work developed by the engine was the same as the one achieved in the cold cycle.

Figure 7 illustrates CH₄ and NO_x emissions during cold-start versus hot-start. During the cold-start, CH₄ emissions were 2.3 times higher and NO_x emissions 1.4 times higher than those during the hot-start. This highlights the temperature sensitivity of catalytic emission control systems, which is also evidenced in Figure 8.

Over the combined cold- and hot-start cycles, CH₄ emissions increased by 30%, NO_x by 13%, and CO by 33% compared to hot-start-only measurements.

The cold-start emissions challenge is more pronounced at low ambient temperatures because it then takes longer for the TWC to reach effective operating temperature, leading to a prolonged period of high emission rates [18,20,21].

3.3. Well-to-Wheels Analysis

In the transport sector, well-to-wheels (WTW) analysis is a commonly used method for assessing the carbon intensity of a fuel. Carbon intensity refers to the amount of greenhouse gases—including CO₂, nitrous oxide, and methane—released during the production and consumption of a transportation fuel, measured in grams of carbon dioxide equivalent per megajoule of energy (g CO₂-eq./MJ).

3.3.1. Fuel Consumption

The total fuel consumption in the hot-start test at −5 °C was 21.9 MJ/km (6.1 kWh/km), corresponding to 0.306 kg/kWh and 47.1 kg/100 km. In June, at +18 °C, the vehicle showed better fuel economy with fuel consumption of 19.8 MJ/km (5.5 kWh/km), corresponding to 0.283 kg/kWh and 42.7 kg/100 km (Figure 9).

3.3.2. Biogas Production Process

The life cycle steps for CBG investigated in this study are feedstock collection and transportation, biogas production, biogas processing to biomethane, biomethane compression, and finally, combustion in an engine. The CBG was produced at Stormossen waste treatment plant near Vaasa. The anaerobic digestion process at Stormossen is divided into two separate process lines. Biogas reactor 1 is fed with wastewater sludge, and the raw material used in biogas reactor 2 is municipal biowaste, supplied within a radius of 40 km [36].

In 2020, raw biogas production at Stormossen was 2.7 million Nm³, of which 52% was upgraded into biomethane, 32% was used for heat and electricity production, and the rest was flared [37]. The methane content of the raw biogas was 62%.

The biogas upgrading is executed by an amine scrubber. The main advantages of chemical absorption with amine solvents are a high methane recovery rate in the upgraded biogas and a low methane slip of <0.1% [38]. In addition, amine solvents are effective at near atmospheric pressure and thus consume a low quantity of electric energy [39]. On the other hand, chemical scrubbing liquids require substantial thermal energy during regeneration, which must be supplied as process heat [39]. After the refining stage, biogas contains 97–98% methane. Finally, the processed biomethane is piped to a gas filling station near the biogas plant. At the refueling station, the gas is pressurized to 300 bar and stored in gas cylinders.

Although the combustion of waste-based biomethane is considered carbon-neutral in Finland’s national GHG inventories (CO₂ emissions from biogas combustion are reported as zero), the use of biomethane may still have climate impact from the above-mentioned earlier stages of the fuel chain. For CBG production, the major contributors of GHG emissions are energy consumption and fugitive losses of methane during digestion and upgrading processes [40]. In addition, some GHG emissions form during the collection of wastes and residues.

3.3.3. GHG Inventory

In this study, the calculation of GHG emissions begins with feedstock collection and transportation. GHG emissions from these steps are based on the following assumptions. Transportation distance 40 km and diesel B7 fuel consumption 20 l/100 km. The lower calorific value of diesel B7 fuel is 43 MJ/kg. The biocomponent of diesel fuel was assumed to be hydrotreated vegetable oil made from waste materials, so the calculated well-to-tank emission factor for diesel B7 was 14.7 g CO₂-eq./MJ fuel, based on the JRC [41] data. Tank-to-wheels CO₂ emission factor for diesel B7 was set at 68.4 g CO₂-eq./MJ fuel [42]. The heat and electricity needs of biogas production and upgrading processes are covered internally by the plant’s own CHP biogas engine and were, therefore, ignored in the GHG inventory. Methane emissions were calculated assuming a methane loss of 1% during anaerobic digestion [43] and 0.1% during the upgrading process [39]. Methane emissions are converted to CO₂-equivalents using a 100-year time horizon global warming potential (GWP) factor of 28 [7]. The energy demand for biomethane compression to 300 bar is 0.25 kWh/m³ (NTP) [44], and the electric energy for compression is taken from the public grid. The CO₂ emission factor for electricity generation in Finland in 2020 was 68.6 g CO₂-eq./kWh [45].

Table 6. CBG well-to-tank GHG emissions.

Parameter	Value	Unit	g CH4/MJbio-CH4	g CO2-Equivalent /MJbio-CH4	Source
Feedstock collection and transportation
Diesel trucks, diesel fuel biocomponent 7%	40	km		1.95	[41,42]
Biogas production and refining
Total biogas production	2,716,000	Nm3			[37]
52% of raw gas for upgrading	1,412,320	Nm3			[37]
Methane content (62%)	875,638	Nm3			[37]
Total biomethane production	31,522,982	MJ
Heat demand * - Anaerobic digestion - Upgrading	0.19 0.110	kWh/Nm3raw gas kWh/kWhbio-CH4			[43]
Electricity demand * - Anaerobic digestion - Upgrading	0.14 0.0136	kWh/Nm3raw gas kWh/kWhbio-CH4			[43]
Methane losses - Anaerobic digestion, 1% - Upgrading, 0.1%	6368 630	kg kg	0.202 0.020	5.66 0.56	[43] [39]
Compression
Electricity demand	0.25	kWh/m3 (NTP)		0.48	[44,45]
CBG well-to-tank GHG emissions				8.65

* Covered internally by the plant’s own CHP biogas engine.

summarizes the main assumptions and input data used in the calculation.

After the anaerobic digestion, the digestate is dewatered and composted to be used as a soil improvement product or as landscaping soil. The digestate treatment is not included in the above table. Any fertilizer or sludge credits are also not considered in GHG calculations.

The GHG benefits associated with transition from fossil-based natural gas or diesel to biomethane were calculated by comparing well-to-wheels CO₂-Equivalent emissions, shown in

Table 7. Well-to-wheels CO₂ Equivalent emissions for CBG, CNG, and diesel B7.

	CBG	CNG	Diesel B7
GHG emissions
Well-to-tank (g/MJfuel)	8.65	13.0	14.7
Tank-to-wheels
CO2 (g/MJfuel)		46.6	68.4
CH4 (g/MJfuel)	0.1708	0.1708
Total GHG (g CO2-eq./MJfuel)	13.4	64.4	83.1
Fuel consumption (MJ/km)	20.8	20.8	16.7
Specific GHG (g CO2-eq./km)	279	1342	1385

. Well-to-tank GHG emission factors for compressed natural gas and diesel fuel were taken from the JRC report [41]. Tank-to-wheel GHG emissions for gas buses are based on the CO₂ and CH₄ emission results recorded in this study, but CO₂ emissions are considered only for fossil CNG. Tank-to-wheels CO₂ emission factor for diesel buses was taken from [42]. The average fuel consumption from Test 1 and 2 in this study was 20.8 MJ/km, and this value is applied to both CBG and CNG bus. It is well known that compression-ignition diesel engines have higher thermal efficiency compared to spark ignition gas engines. Therefore, the fuel consumption of a diesel bus was set at 80% of that of a gas bus, based on the VTT’s (Technical Research Centre of Finland) comprehensive report on city bus emissions measurements [46].

Figure 10 shows the percentage changes in life cycle GHGs for the studied fuels. Shifting from conventional diesel to fossil natural gas does not show meaningful GHG benefits, bearing in mind the higher thermal efficiency of compression-ignition engines compared to spark-ignition gas engines. However, for biomethane, the situation is very different; 80% GHG emission benefit is achieved by switching from diesel to biomethane. With more precise methane emission control, GHG emission savings would advance towards 90%.

This gives a strong environmental argument for biogas use. Increasing biogas use would be a quick and cost-effective way to reduce GHG emissions from urban traffic. Unfortunately, the potential of renewable gas is not acknowledged in the current EU emission standards, which only focus on tank-to-wheels emissions. Changing the measurement method to life cycle-based WTW instead of tailpipe measurement would enable a proper assessment of GHG emissions of future vehicle technology and fuel combinations. However, the results of this study can be utilized in designing strategies for transitioning to sustainable urban transport systems.

4. Conclusions

Transition to low-emission transportation and cleaner cities requires a broad introduction of low- and zero-carbon alternatives to conventional petrol- and diesel-powered vehicles. This paper presents the results of real-driving emission measurements from a Euro VI biogas-powered city bus equipped with a TWC. In addition, the lifetime carbon intensity of CBG was investigated and compared to its fossil counterpart CNG and traditional diesel fuel. The main findings were, first, for the bus:

• The rapid changes in exhaust gas temperature and composition under transient driving conditions seemed to be a critical challenge to an efficient operation of the TWC.
• Unimpressive CH₄ oxidation and NO_x reduction were observed in the catalyst after its service life of 375,000 km–400,000 km. In contrast, CO emissions were low, indicating efficient oxidation of CO in the catalyst.
• The primary reason for deficient CH₄ and NO_x conversion over the TWC was assumed to be the low CH₄ reactivity due to a partial deactivation of the catalyst. At low loads, common in a city bus’s driving profile, the exhaust gas temperature was too low to allow efficient CH₄ oxidation. In addition to the low CH₄ oxidation rate, low CH₄ reactivity also means that methane-based reducing agents for NO_x reduction do not work, leading to substantial NO_x breakthrough from the catalyst.
• In addition, during the cold-start, CH₄ emissions were 2.3 times and NO_x emissions 1.4 times as high as those during the hot-start, highlighting the temperature sensitivity of catalytic emission control systems.
• Based on the above, deterioration of the exhaust after-treatment systems over time should be monitored as they are exposed to different aging processes resulting in elevated real-world emissions.
• Another critical issue was the fluctuating air-to-fuel ratio. Lambda was outside the optimal range for a significant part of the time, likely reducing the TWC efficiency. This highlights the need for precise lambda control to ensure high conversion rates throughout the vehicle’s lifetime.

Additionally,

• The WTW analysis showed an 80% GHG emission benefit by switching from diesel to biomethane, giving a strong environmental argument for biogas use. With more precise methane emission control, GHG emission savings would advance towards 90%.

The presented real-driving emission results are of great importance in supplementing the emission data for aging gas-powered HD vehicles, filling the gap of data on emissions closer to the service life of the vehicles. After all, the average age of bus fleets in Finland, for example, is over 12 years. The results of this study can also be utilized in scheduling catalyst maintenance or replacement activities.

In the future, it would be worthwhile to repeat the weather-related comparison with a completely new bus or with a new catalyst on an old bus.