The Impact of Vehicle Engine Characteristics on Vehicle Exhaust Emissions for Transport Modes in Lagos City

ABSTRACT

The study delves into how vehicle engine characteristics impact the release of air pollutants from various vehicle fleets in Lagos, Nigeria. It involved the direct measurement of emissions from the exhaust pipes of 88 vehicles using gas analyzers. The vehicle fleets encompassed motorcycles, tricycles, private cars, minibuses, large buses, and trucks. A statistical analysis was conducted on carbon monoxide (CO) and nitrogen oxide (NOx) emissions to develop a model equation based on vehicle type, engine type, vehicle age, and purchase status. Results indicate that personal cars and minibuses predominantly emit CO from gasoline engines, whereas large buses and trucks significantly contribute to NOx emissions from diesel engines. Further scrutiny revealed that 66% of the vehicles examined were over 10 years old, resulting in a 65% increase in emission levels. Approximately 60% of gasoline and 75% of diesel vehicles exceeded the permissible emission limits, leading to air quality deterioration and heightened health risks. The study underscores the risks associated with ageing vehicles and different engine types, emphasizing the imperative for a gradual transition to low-carbon or electric vehicles in developing African cities to combat air pollution and mitigate health hazards.

KEYWORDS:

• Vehicle fleets
• exhaust emissions
• engine type
• air pollutants

1. Introduction

With an increase in the number of personal cars per capita, which serves as the principal mode of transportation in many developing nations (Pojani & Stead, 2015), urban road transportation has seen increased activity. This trend can be linked to rising household wealth, greater mobility, and households shifting from non-motorized to motorized modes of transportation (Mahadevia & Advani, 2016. Other factors contributing to the increase in per capita vehicle ownership include the availability of quality road infrastructure, the supply and efficiency of public transportation, policy decisions affecting vehicle ownership (such as taxes and insurance), the cost of automobiles, petrol prices, and the expenses associated with alternative modes of transportation (Rong et al., 2022). Rapid urbanization and population growth in West African cities generate significant demand for transport, resulting in increased traffic emissions and atmospheric pollution, despite a high percentage of poorly maintained engines, the importation of used cars, and low fuel quality (Abera et al., 2021). Traffic emissions have emerged as a key contributor to environmental air pollution, deteriorating ambient air quality, particularly since Africa’s vehicle population has grown rapidly (Ayetor et al., 2021).

According to Adama’s (2018) projections, fine particulate matter (PM_2.5) was estimated to have contributed to 2.9 million premature deaths worldwide in 2017 as a result of heart attacks, strokes, lung cancer, pulmonary diseases, acute respiratory infections, and type 2 diabetes. The problem is especially serious in Nigeria, where ambient PM_2.5 pollution has been linked to the greatest number of premature deaths in West Africa, particularly in Lagos, the nation’s commercial hub and one of the fastest-growing megacities in the world. Over 11,200 deaths in Lagos in 2018 were caused by air pollution, which also had a negative impact on other health issues (Croitoru et al., 2020). Although public health and air quality are becoming increasingly concerned with traffic air pollutants, there is no accurate way to measure how vehicle fleets affect pollution levels. To fully quantify the consequences of air pollution on public health, especially for those who live close to roads, it is crucial to conduct health impact evaluations and perception studies (Ajayi, Adams, Dumedah, Adebanji, et al., 2023). In comparison to modern cars with more efficient engines and cutting-edge pollution control technology, a large portion of the cars on Lagos’ roads are outfitted with antiquated emission control systems, which may result in increased emissions of PM and black carbon (Fiebig et al., 2014; Kholod & Evans, 2016). It is estimated that the bulk of Lagos’ current car fleet is over 15 years old and does not have the newest pollution control equipment, while the precise age and technological makeup of the fleet is unknown (LAMATA, 2016). About 400,000 automobiles are imported into Nigeria each year, of which 22% are new and 78% are used, according to the National Bureau of Statistics (LSBS, 2015). The majority of cars in West African cities are secondhand imports from the United States and Europe, and they frequently lack equipment for reducing emissions and noise. These older, poorly maintained cars add to Lagos’ traffic-related air pollution.

Poor gasoline quality is a major contributor to vehicle emissions, especially in Sub-Saharan Africa (Hirota & Kashima, 2020). According to Ayetor et al. (2021), the amount of sulphur in petroleum products used by Nigerian automobiles exceeded the permissible limits for contemporary emission systems by 204 times. By reducing sulphate particles directly, reducing secondary particle formation from sulphur dioxide (SO₂), and enabling installed emission control and catalyst systems for other precursor emissions to function properly, the reduced sulphur content in fuel when combined with advanced emission control technologies can reduce vehicular emissions. Nigeria is one of the major producers of crude oil in West Africa. Still, it imports the majority of its refined petroleum products from other nations, mostly the US and Europe. Many of these nations continue to export high-sulfur (dirty) fuel, although doing so is illegal domestically due to concerns about air pollution. As of right now, imported petrol and diesel are permitted to contain up to 1,000 ppm and 3,000 ppm of sulphur, respectively (Gope et al., 2018). These are far beyond the 10 ppm limit for both products set by the European Union (Gitiaux et al., 2009).

To calculate emissions based on emission characteristics – such as average speed, fuel type, road conditions, vehicle features, traffic demand, driver behaviour, and environmental factors like temperature, humidity, and wind speed – emission models are crucial (Lyu et al., 2021). The various elements that influence vehicle emissions are shown in Figure 1. Vehicle operating factors fall into two groups, according to Lyu et al (Lyu et al., 2021): fuel economy and speed-related problems. Fuel-related vehicle attributes, such as petrol type, fuel consumption, distance travelled, and engine design, are considered while the vehicle is operating (Golbasi & Kina, 2022). According to theory, an engine’s size affects how many pollutants it produces; bigger engines use more fuel and produce more pollutants than smaller engines (Katrašnik, 2007; Pilusa et al., 2012). Fuel efficiency is also influenced by the kind of engine. Vehicle fuel economy and pollution levels will be impacted by variable engine operating resistance and low-speed vehicle inertia (Yu et al., 2022). Emissions are also influenced by the fuel type used (Zhang et al., 2020). This indication illustrates the unique characteristics of driving energy consumption and provides the foundation for the emission. It has to do with things like running distances and transit options. A model was created by Ajayi, Adams, Dumedah, Adebanji, et al. (2023) to evaluate how traffic flow characteristics, such as flow, speed, and vehicle composition, affect the amount of emissions in a heterogeneous traffic situation. They concluded that there is an exponential association between pollution concentration and traffic mobility parameters.

Figure 1. Factors contributing to traffic-related air pollution (source; author).

The measurement of vehicle emissions and their effects on Lagos, Nigeria’s ambient air quality have been the subject of extensive research (Adeleke et al., 2011; Ajayi et al., 2023; Obanya et al., 2018; Odekanle et al., 2017; Ojolo et al., 2007; Olajire et al., 2011), but these studies have not concentrated on the characteristics of vehicle engines and how they affect emission levels. The amount of pollution that cars emit and how varied car fleets and engine types affect emissions, air quality, and public health in pollution zones remain unknown since there is a dearth of information on inventory, machinery, and measurement technology. Thus, by identifying and modelling variables that influence pollutant concentration, this study aims to evaluate the influence of vehicle engine parameters on emission levels from tested automobiles. The following research questions are important for the characterization of vehicle characteristics’ impact on emissions for Lagos;

1. What are the characteristics of vehicle engines in Lagos, Nigeria?

2. How do these vehicle engine characteristics, such as age, fuel type, maintenance, and vehicle engine capacity affect pollutant concentration?

3. What is the drive cycle impact on the pollutant concentration for the different vehicle engine characteristics?

4. What are the emission standards and air quality for the vehicle engine characteristics?

2. Materials and methods

2.1. Study site

Owing to some distinctive characteristics, Lagos was chosen for this investigation. Lagos, one of Nigeria’s largest cities and most populous, has serious problems with air pollution. According to Hoornweg and Pope (2017) and Okimiji et al. (2021), it is also one of the megacities with the quickest rate of growth, and by 2100, it is expected to outnumber all others in Africa in terms of population. A significant portion of the populace has been exposed to elevated air pollution levels as a result of swift industrialization and urbanization, which has detrimental impacts on both environmental circumstances and human well-being (Baldacci et al., 2015). Excessive car fleets, regular traffic jams, inadequate fuel requirements, irresponsible driving, and Sub-Saharan African weather all contribute to high car exhaust pollution levels. Particularly in Lagos, Nigeria’s economic hub and one of the megacities with the fastest rate of growth in the world (Maduekwe, Akpan, & Isihak, 2020), this is a crucial problem. Motor cars on Lagos’ roadways have almost doubled in the last ten years, to a daily total of 5 million (Olvera, Plat, & Pochet, 2020). Around 227 cars per km of road are seen daily in Lagos, which is substantially more than the average of 11 cars per kilometre across the country (Croitoru, Chang, & Akpokodje, 2020). Lagos is known for having extremely heavy traffic, with most commuters spending at least three hours a day stuck in it. Tricycles, minibuses, big buses, motorbikes, and private automobiles are examples of common forms of mobility. On the routes around the city, there are also a lot of diesel-powered commercial buses in Lagos (Fakinle et al., 2020).

2.2. Direct emission measurement for test vehicles

It is necessary to explicitly capture the characteristics of the individual vehicle, such as age and mileage, maintenance condition, vehicle fleet, speed, weight and size, and engine power on the emission levels. Vehicle characteristics data were retrieved from the computerized vehicle inspection, Berger Lagos. The Lagos state government commissions the agency for the implementation of the compliance on Road Traffic Law 2021; it undertakes the duties of ensuring compliance with all stipulated/required vehicle policies such as roadworthiness, vehicle license, hackney permit, testing and training of applicants for driver’s license/rider card. The agency randomly selected 88 vehicles for emission testing to ensure a representative sample of the registered vehicle population in Lagos. The initial testing of this set of vehicles took place in 2021, marking the early stages of the agency’s operations. While the sample size of 88 vehicles may be considered small, it provides a reasonable representation of the broader vehicle population in Lagos. Unfortunately, data on subsequent vehicles were unavailable during the study for additional assessment. The emission results and vehicle information obtained from the initial testing were used for further evaluation in the study. The vehicle information obtained includes; vehicular types, age, maintenance, registration, engine types, plate number, registration, colour, engine type and purchase condition. Also, estimated results of emission tests for carbon monoxide CO and nitrogen oxide NOx in g/km were collected. This information was collected for the vehicle’s fleets such as motorcycles, tricycles, personal/private cars, minibuses, large buses and trucks. Emission testing involved inserting the sampler probe into the exhaust pipe of the vehicles and measuring emissions using a gas analyzer during acceleration. The procedure included starting the vehicle and allowing it to run at idle speed (750rpm −1000rpm) for 15 seconds, followed by acceleration for another 15 seconds, maintaining an idle speed (1500rpm −2500rpm). While the engine idled, the sampler probe was inserted into the exhaust pipe as deeply as possible but not less than 300 mm. Readings for CO and NOx pollutants were taken 20 seconds after using the gas analyzer. To assess emissions at various speeds, the vehicles were driven on a test field for 15 minutes at different speeds (0, 5, 15, 45, and 85 km/hr). Gas emissions of CO and NOx were measured through the gas analyzer and recorded.

2.3. Statistical analysis

To effectively understand how vehicle characteristics, affect the emission levels of CO and NOx, the study employed a multiple linear regression (MLR) approach to quantify the joint impact of vehicle engine characteristics on emission levels of CO and NOx for the 88 tested vehicles. Equation (1)(1) $\left(\mathrm{CO}\right)={\mathrm{b}}_0+{\mathrm{b}}_1{\mathrm{V}}_{\mathrm{t}}+{\mathrm{b}}_2{\mathrm{E}}_{\mathrm{t}}+{\mathrm{b}}_3{\mathrm{A}}_{\mathrm{g}}+{\mathrm{b}}_4\mathrm{P}$(1) and (2(2) $\left(\mathrm{N}{\mathrm{O}}_{\mathrm{x}}\right)={\mathrm{b}}_0+{\mathrm{b}}_1{\mathrm{V}}_{\mathrm{t}}+{\mathrm{b}}_2{\mathrm{E}}_{\mathrm{t}}+{\mathrm{b}}_3{\mathrm{A}}_{\mathrm{g}}+{\mathrm{b}}_4{\mathrm{P}}_{\mathrm{s}}$(2) ) represents the Multiple Linear Regression Model generated from test vehicles for CO and NOx.

Null Hypothesis:

H₀ = Model adequately fits the data

Alternative Hypothesis:

H_A= Model does not adequately fit the data

Regression Equation

(1) $\left(\mathrm{CO}\right)={\mathrm{b}}_0+{\mathrm{b}}_1{\mathrm{V}}_{\mathrm{t}}+{\mathrm{b}}_2{\mathrm{E}}_{\mathrm{t}}+{\mathrm{b}}_3{\mathrm{A}}_{\mathrm{g}}+{\mathrm{b}}_4\mathrm{P}$(1)

(2) $\left(\mathrm{N}{\mathrm{O}}_{\mathrm{x}}\right)={\mathrm{b}}_0+{\mathrm{b}}_1{\mathrm{V}}_{\mathrm{t}}+{\mathrm{b}}_2{\mathrm{E}}_{\mathrm{t}}+{\mathrm{b}}_3{\mathrm{A}}_{\mathrm{g}}+{\mathrm{b}}_4{\mathrm{P}}_{\mathrm{s}}$(2)

Where V_t = vehicle types; E_t = engine types; A_g = age; P_s = purchase condition; b₀ = intercept; b_1- b₅ = coefficients.

Spearman’s correlation was employed to measure the relative strength of the linear relationship between two continuous variables. The National Environmental Standards Regulations and Enforcement Agency, NESREA have outlined emission standards for engine types called the National Environmental (control of vehicular emissions from petrol and diesel engines) regulation in 2015. The vehicle types are grouped according to their Gross Vehicle Weight (GVW). The GVW is the weight of the empty vehicle plus the weight of the maximum payload that the vehicle was designed to carry. The CO and NOx values for each vehicle class were determined and compared with the emission standards.

3. Result and discussions

3.1. Vehicle fleet characteristics

Table 1 provides a summary of the vehicle fleet, indicating the percentage by motorization type and age. The findings indicate that 65 percent of the tested vehicles had petrol engines, while 35 percent were equipped with diesel engines. The breakdown of petrol engine vehicles includes motorcycles (16%), tricycles (11%), personal cars (24%), minibuses (11%), large buses (2%), and trucks (2%). On the other hand, a larger proportion of diesel-engine vehicles comprises minibuses (12%), large buses (9%), and trucks (12%). Notably, about 50% of commercial large buses for transport in Lagos are powered by diesel engines (Croitoru et al., 2020; Maduekwe et al., 2020; Odekanle et al., 2017). Furthermore, the considerably lower cost of petrol compared to diesel may contribute to its prevalent use in transit vehicles like buses, minibuses, and intercommunal sedan taxis. This cost difference is particularly relevant in West African nations characterized by high poverty levels and a lack of widespread awareness of environmental concerns, leading to a higher consumption of petrol (Doumbia et al., 2018). Additionally, as illustrated in Figure 2, the age of vehicles emerges as a crucial characteristic. The results reveal that 66 percent of the total vehicles were older than 10 years, with about 27 percent falling in the 5 to 10 years age bracket, and 7 percent being between 0 and 5 years old. This analysis aligns with previous studies indicating that the vehicle fleets of sub-Saharan countries predominantly consist of outdated imports from Western nations, significantly contributing to traffic-related pollution (Jiang et al., 2017; Tang et al., 2020). Furthermore, Figure 2 highlights that minibuses, followed by trucks and personal cars, have a high proportion of vehicles older than 10 years. In West Africa, some minibuses can be as old as 20 years (Verster & Fourie, 2018). These ageing vehicles often undergo inadequate maintenance, leading to increased emissions and elevated air pollution levels.

Figure 2. Age of the test vehicles.

Table 1. Vehicle fleet and percentage by motorization, type and age.

	Engine type (%)		Age (%)			Vehicle purchase status (%)
Vehicle fleet	Petrol	Diesel	0–5 years	5–10 years	older than 10 years	used- type 1	used- type 2	new
Motorcycle	16	0	1	3	12	9	4	3
Tricycle	11	0	0	4	7	7	3	1
personal car	24	1	3	9	13	9	13	3
Minibus	11	12	2	4	17	6	15	2
large bus	2	9	0	5	6	6	3	2
Truck	2	12	0	1	13	8	2	4
Total	65	35	7	27	66	43	38	19

Figure 3 illustrates the status of vehicles at the time of purchase, distinguishing between used and new vehicles. The used vehicles are further categorized into two types: Type 1 includes vehicles imported into the country, and Type 2 encompasses vehicles acquired from a third party, commonly referred to as ‘second-hand’ vehicles. The findings indicate that 43 percent of the vehicles are imported into the country as used, while 38 percent are classified as second-hand vehicles. The smallest proportion, at 19 percent, represents vehicles purchased as new. This underscores a significant influx of imported vehicles in Lagos (Aminu & Asikhia, 2021; Ibitayo, 2012; Maduekwe et al., 2020). However, income levels play a role in vehicle ownership, contributing to increased reliance on older and imported vehicles (Sun et al., 2021). The results reveal a higher percentage of vehicles aged more than 10 years among motorcycles, tricycles, private cars, minibuses, and trucks. Additionally, a greater proportion of second-hand vehicles is observed among personal cars and minibuses, while used Type 1 vehicles (imported) are more prevalent among motorcycles, tricycles, large buses, and trucks.

Figure 3. Vehicle status at purchase.

3.2. Effect of vehicle age on pollutants concentration

The current age of vehicles plays a crucial role in determining pollutant concentration. Figures 3 to 6 illustrate the relationship between vehicle age and pollutant concentrations of CO and NOx for both engine types. The proportion of vehicles in the age groups of 5–10 years and above 10 years significantly contributes to increased pollutant concentrations in both engine types. Notably, in the case of petrol engines (Figures 4 and 5), a considerable rise in CO and NOx concentrations is observed due to a higher proportion of minibuses and trucks aged more than 10 years. Conversely, in diesel engines (Figures 6 and 7), the pollutant concentrations are predominantly influenced by the older large buses and trucks.

Figure 4. Effects of vehicle age on CO-emissions in petrol engine.

Figure 5. Effects of vehicle age on NOx-emissions in petrol engine.

Figure 6. Effects of vehicle age on CO-emissions in diesel engine.

Figure 7. Effects of vehicle age on NOx-emissions in diesel engine.

3.3. Direct vehicle fleet emission levels

The CO and NOx emissions for the vehicle’s fleets were observed for the different categories of the age of the vehicles and engine type under the static condition. This elucidates the effects of the age of the vehicle and engine type on the emission rate of the vehicles. Table 2 shows the pollutant’s average concentration from the vehicles for both engine types. The percentage of the engine type for each class of vehicles has a direct effect on the pollutant’s concentration of CO and NOx as shown in. Figure 8 depicts the contribution of the vehicle fleet to the emission of CO and NOx gases. The chart revealed the CO is high for personal cars, minibuses, large buses and trucks while large buses and trucks contributed to the high emission of NOx for a petrol engine. The CO and NOx for diesel engines show large buses and trucks as major contributors to their high emissions. The results indicate that the sources of carbon monoxide (CO), and nitrogen oxide (NOx), are mainly passenger cars and trucks with petrol and diesel engine respectively (Shahbazi et al., 2016). claim that although minibuses, buses, and trucks make up only 2.4 percent of Tehran’s fleet, they are responsible for more than 41, 64, and 85 percent of the city’s NOx, SOx, and PM emissions, respectively. The implication of this is that the minibuses, large buses and trucks are contributing major to the emission of CO and NOx in both engine types due to their weight and size, emission rate, maintenance and age of these vehicles. We have a larger percentage of personal cars, minibuses, large buses and trucks in the more than 10 years category. The contribution of CO and NOx from a diesel engine is lowest from a motorcycle and a tricycle (Biona et al., 2007; Giang & Oanh, 2014; Ibeto & Ugwu, 2019).

Figure 8. (a) Contribution of the vehicle fleet to petrol engine; (b) contribution of the vehicle fleet to diesel engine.

Table 2. Pollutants average concentration from the different engine vehicles.

	Percentage (%)		Petrol		Diesel
Vehicle fleet	Petrol	Diesel	CO (g/km)	NOx (g/km)	CO (g/km)	NOx (g/km)
Motorcycle	14	0	2.35	0.15	0	0
Tricycle	10	0	2.53	0.25	0	0
Personal car	21	1	4.21	0.54	0.4	0.4
Minibus	10	10	4.22	0.32	1.88	1.8
Large bus	2	8	4.05	3.25	4.21	6.64
Truck	0	12	4.2	1.2	4.72	7.15

A thorough analysis of each vehicle’s contribution to CO and NOx emissions for both engine types is shown in Figure 9. With the exception of large buses and trucks, where emissions are comparatively greater, NOx emissions from petrol engines are generally lower throughout all vehicle fleets. Conversely, CO emissions are more noticeable in all kinds of cars. Diesel engines show a different trend, with over 70% of emissions coming from NOx emissions. In comparison to petrol or petrol engines, diesel engines run at higher temperatures and pressures, which promotes the production of NOx gas. The intensity and duration of the hottest part of the flame during combustion affect the amount of NOx (Akal et al., 2020; Cho & He, 2007; Thiyagarajan et al., 2022). Due to the heat generated during compression, diesel fuel spontaneously ignites in the engine because the amount of air in the cylinder of a diesel engine is roughly twice that of a petrol engine (Solanki et al., 2020). Remarkably, the findings show that CO emissions from diesel engines are lower than NOx emissions. Diesel engines are said to be lean combustion engines since they have low CO and HC contents (Reşitoğlu et al., 2015). In terms of particular vehicle classes, trucks, large buses, and personal cars all contribute significantly to CO emissions from petrol engines (19%, 20%, 19%, and 19%, respectively). Tricycles and motorcycles contribute the least, at 12% and 11%, respectively. Large buses stand out among petrol-powered vehicles as significant contributors to NOx emissions, accounting for 33% of the total. It is important to note that big buses in Lagos that are meant for public mass transit travel in specific lanes and do not share space with other cars. With percentages of 45% and 38%, respectively, trucks and large buses in diesel engines contribute significantly to CO and NOx emissions. One of the main causes of the increased NOx emissions in the tested trucks is the high percentage of diesel engines. This emphasizes how large vehicles and buses are the main producers of NOx emissions. The combined CO and NOx emissions for the two engine types are shown in Figure 10.

Figure 9. (A) emission of CO and NOx for petrol engine; (b) emission of CO and NOx for diesel engine.

Figure 10. Total emission of CO and NOx gases for the test vehicles.

3.4. Speed profile for drive cycle of vehicles

One of the study objectives is to assess the influence of drive cycles on the pollutant concentration of different vehicle engine types. The speed of vehicles in motion significantly impacts the emission levels of CO and NOx for various engine types. The findings indicate increased emissions across all vehicles, ranging from 85 km/hr to 5 km/hr. Notably, CO emissions peak in the range of 4.61–3.25 g/km for average speeds from 5 km/hr to 85 km/hr, particularly observed in minibuses. For NOx gases in petrol engines, the highest emissions occur in the range of 3.8–3.05 g/km. In diesel engines, CO emissions range from 5.0–4.18 g/km, while NOx emissions range from 7.61–6.67 g/km. Figure 11 represents the speed variation in the drive cycle and its impact on CO and NOx emissions for both engine types. The concentration of NOx is notably high in petrol and diesel vehicles due to the emissions from large buses and trucks, respectively.

Figure 11. (A) total emission of CO for petrol engine; (b) total emission of NOx for petrol engine; (c) total emission of CO for the diesel engine; (d) total emission of NOx for diesel engine.

Vehicle characteristics significantly contribute to the fluctuation in emission levels across various speed cycles. Notably, there is an increase in emissions for minibuses and trucks as the speed decreases. However, the results do not provide insights into the impact of speed dynamics or drivers’ behaviour on emission levels. Assessing emissions at different speeds is crucial for understanding emission rates under diverse traffic and road conditions. The findings highlight that owing to severe traffic congestion, vehicles operating at low speeds with frequent accelerations, decelerations, and idle periods generate higher pollutant levels. This information can aid policymakers in developing a framework for future health assessments.

3.5. Statistical model analysis

To effectively model the influence of vehicle engine characteristics on CO and NOx emission levels, the Multiple Linear Regression (MLR) approach is employed. This approach quantifies the combined impact of independent factors such as vehicle type, vehicle age, engine type, and purchase condition on the tested vehicles. The correlation between vehicle types, engine types, vehicle age, purchase status, and the concentration of CO and NOx is presented in Table 3. Spearman’s correlation results reveal a strong positive association among vehicle age, engine type, vehicle type, CO, and NOx emissions. The statistically significant p-value of less than 0.05 underscores the strength of these correlations.

Table 3. Spearman’s correlation measures the variables.

			Carbon mono oxide	Nitrogen oxide
Spearman’s rho	Vehicle type	Correlation Coefficient	.764**	.794**
		Sig. (2-tailed)	.000	.000
		N	88	88
	Engine type	Correlation Coefficient	.071	.763**
		Sig. (2-tailed)	.773	.000
		N	88	88
	Purchase status	Correlation Coefficient	−.095	−.610
		Sig. (2-tailed)	.377	.928
		N	88	88
	Age of the vehicle	Correlation Coefficient	.798**	.702**
		Sig. (2-tailed)	.005	.000
		N	88	88
	Carbon mono oxide	Correlation Coefficient	1.000	.676**
		Sig. (2-tailed)	.	.000
		N	88	88
	Nitrogen oxide	Correlation Coefficient	.676**	1.000
		Sig. (2-tailed)	.000	.
		N	88	88

Equation (1)(1) $\left(\mathrm{CO}\right)={\mathrm{b}}_0+{\mathrm{b}}_1{\mathrm{V}}_{\mathrm{t}}+{\mathrm{b}}_2{\mathrm{E}}_{\mathrm{t}}+{\mathrm{b}}_3{\mathrm{A}}_{\mathrm{g}}+{\mathrm{b}}_4\mathrm{P}$(1) and (2(2) $\left(\mathrm{N}{\mathrm{O}}_{\mathrm{x}}\right)={\mathrm{b}}_0+{\mathrm{b}}_1{\mathrm{V}}_{\mathrm{t}}+{\mathrm{b}}_2{\mathrm{E}}_{\mathrm{t}}+{\mathrm{b}}_3{\mathrm{A}}_{\mathrm{g}}+{\mathrm{b}}_4{\mathrm{P}}_{\mathrm{s}}$(2) ) represents the Multiple Linear Regression Model generated from test vehicles for CO and NOx.

Null Hypothesis:

H₀ = Model adequately fits the data

Alternative Hypothesis:

H_A = Model does not adequately fit the data

Regression Equation

(3) $\mathit{CO}:\mathit{\hspace{0.17em}}1.948+.807V_t+\mathit{\hspace{0.17em}}2.142E_t+.668A_g$(3)

(4) $\mathit{NOx}:-3.67+.775V_t+\mathit{\hspace{0.17em}}2.803E_t+.300A_g$(4)

Where V_t = vehicle types; E_t = engine types; A_g = age; P_s = purchase status; b₀ = intercept; b_1- b₅ = coefficients.

The equation’s model coefficients are shown in Tables 4 and 5. The null hypothesis is rejected based on the model findings, which show that the p-values for age, engine type, and vehicle type are statistically significant. This suggests that the kind of vehicle, the kind of engine, and the age of the vehicle all affect how much pollution it emits. All motor vehicles release carbon dioxide (CO), but light-duty, gasoline-powered vehicles are the main sources of CO emissions. According to Adeyanju and Manohar (2017), CO is a useful indicator of the movement and dispersion of primary combustion emissions from traffic sources that are inert and do not pose a health risk since it does not react in the near-road environment. The growing significance of NO₂ emissions has also been attributed to the use of diesel emission control technologies, such as Diesel Oxidation Catalysts (DOC). According to Carslaw et al. (2019), these technologies purposefully convert NO to NO₂ and use NO₂ to speed up the oxidation of CO, hydrocarbons, and particulate matter. Notably, a sizable percentage of Lagos’s commercial buses have diesel engines and are frequently seen travelling the city’s routes (Croitoru et al., 2020; Fakinle et al., 2020).

Table 4. Coefficients for the model equation – NOx

Model		Unstandardized Coefficients		Standardized Coefficients	t	Sig.
		B	Std. Error	Beta
1	(Constant)	−3.679	1.013		−3.631	.000
	Vehicle type	.775	.164	.429	4.735	.000
	Engine type	2.803	.561	.466	4.999	.000
	Age of the vehicle	.300	.314	.652	5.958	.000
	Purchase status	.012	.256	.003	.046	.963
a. Dependent Variable: Nitrogen oxide

Table 5. Coefficients for the model equation – CO.

Model		Unstandardized Coefficients		Standardized Coefficients	t	Sig.
		B	Std. Error	Beta
1	(Constant)	1.948	.553		3.522	.001
	Vehicle type	.807	.089	.955	9.041	.000
	Engine type	2.142	.306	−.760	−7.000	.000
	Age of the vehicle	.668	.171	.310	3.903	.000
	Purchase status	−.045	.140	−.025	−.325	.746
a. Dependent Variable: Carbon mono oxide

The model implies that a unit increase in vehicle type will affect the emission of CO and NOx if other variables are constants.

3.6. Model summary

Table 6 summarizes the result of this model. The best values of adjusted R² (coefficient of determination) for CO and NOx are 0.72 and 0.75 respectively. This shows that the goodness of fit of the line obtained from the analysis is accepted since the R² values are above 0.5 and the p-value is less than the alpha value (0.05), we reject the null hypothesis and hence the model provides the best fit for the data. The model was assessed by the analysis of variance ANOVA to determine if there is a significant difference between the groups being studied by using variance as shown in Tables 7 and 8. The model developed for the traffic emission for each pollutant showed a 95 percent level of confidence, therefore a significant relationship exists between the independent variables and the predictor based on the value of p which is less than 0.05 i.e p < 0.05, hence it is statistically significant.

Table 6. Model summary

Model	R	R Square	Adjusted R Square	Std. Error of the Estimate
1	.935^a	.874	.718	.93261
2	.914^a	.835	.746	1.70897
a. Predictors: (Constant), Purchase status, Vehicle type, Age of the vehicle, Engine type

Table 7. Model assessment by ANOVA- CO.

ANOVA^b
Model		Sum of Squares	Df	Mean Square	F	Sig.
1	Regression	84.671	4	21.168	24.337	.000^a
	Residual	72.190	83	.870
	Total	156.860	87
a. Predictors: (Constant), Purchase status, Vehicle type, Age of the vehicle, Engine type
b. Dependent Variable: Carbon monoxide

Table 8. Model assessment by ANOVA- NOx.

ANOVA^b
Model		Sum of Squares	Df	Mean Square	F	Sig.
1	Regression	474.386	4	118.597	40.607	.000^a
	Residual	242.408	83	2.921
	Total	716.794	87
a. Predictors: (Constant), Purchase status, Vehicle type, Age of the vehicle, Engine type
b. Dependent Variable: Nitrogen oxide

3.7. Emission standard for petrol and diesel engines

Table 9 displays the average emissions for CO and NOx alongside their respective emission standards. Vehicle types were categorized based on their Gross Vehicle Weight (GVW) to establish classes according to weight. GVW represents the combined weight of the empty vehicle and the maximum payload it was designed to carry (Krebs & Ehmke, 2021). The findings indicate that motorcycles, personal cars, minibuses, large buses, and trucks powered by petrol engines exceeded the standards for both CO and NOx emissions. Specifically, motorcycles emitted 2.35 g/km for CO and 0.17 g/km for NOx; personal cars emitted 4 g/km for CO and 21 g/km for NOx; minibuses emitted 6.22 g/km for CO and 0.32 g/km for NOx; large buses emitted 4.05 g/km for CO and 3.25 g/km for NOx; and trucks emitted 4.2 g/km for CO and 1.2 g/km for NOx. For diesel engines, the emissions from minibus and trucks exceeded the standards, with minibus emitting 1.88 g/km for CO and 1.8 g/km for NOx, and trucks emitting 4.21 g/km for CO and 6.64 g/km for NOx. These exceedances have implications for the degradation of ambient air quality and pose health burdens for road users and residents near roadways. Furthermore, additional results revealed that approximately 60% of petrol vehicles failed the emission test and exceeded the standards, while 75% of diesel vehicles surpassed the permissible limits. Diesel-powered vehicles generate ten times more particles than gasoline vehicles per kilometer driven (Kontses et al., 2020; Lyu et al., 2021). Interestingly, despite these findings, the vehicle inspection agencies in Lagos, responsible for issuing roadworthiness certificates, do not currently consider failed emission test results as a major criterion for the issuance of roadworthiness certificates for registered vehicles under test.

Table 9. Average emission for CO and NOx and their emission standards.

		Petrol g/km				Diesel g/km
Reference mass (kg)	Vehicles	CO	NER, 2015 standards	NOx	NER, 2015 standards	CO	NER, 2015 standards	NOx	NER, 2015 standards
Class I < 1305	motorcycle	2.35	2.3	0.17	0.15	–	–	–	–
	tricycle	2.53	2.3	0.25	0.15	–	–	–	–
Class II < 1305–1760	personal car	4.21	4.17	0.54	0.18	0.4	0.8	0.4	0.65
	minibus	6.22	5.22	0.32	0.21	1.88	0.98	1.8	0.78
Class III: > 1760	large bus	4.05	4.0	3.25	1.1	4.21	4.0	6.64	7.0
	Truck	4.2	4.0	1.2	1.1	4.72	4.0	7.15	7.0

4. Conclusions

The study analyzed vehicle attributes gathered from a testing facility, investigating how factors such as vehicle type, engine type, age, and purchase status affect the emission levels of CO and NOx from a sample of 88 vehicles. Findings revealed that a significant portion of personal cars, motorcycles, and tricycles relied on petrol engines, while diesel engines were predominant in large buses and trucks. Moreover, a majority (66%) of the examined vehicles were over ten years old, indicating a reliance on older, second-hand vehicles. Petrol-powered minibuses and large buses emerged as notable contributors to CO and NOx emissions, whereas trucks and large buses with diesel engines emitted substantial amounts of these pollutants. Furthermore, emission levels varied across different speed cycles, with trucks emitting higher levels of CO and NOx in diesel engines, and large minibuses emitting more in petrol engines at reduced speeds. Emission levels increased across all vehicles during test drives from 85 km/hr to 5 km/hr. Despite the significant relationship found between CO and NOx emissions and vehicle type, engine type, and age (p < 0.05), the study’s limited sample size restricts its representation of Lagos’ vehicle population, underscoring the necessity for larger emission datasets for improved characterization. The predictive model, while effective for prediction purposes, solely focuses on the impact of vehicle engine characteristics, overlooking variables such as vehicle speed dynamics and driver behaviour.

Future investigations could delve into the influence of vehicle speed dynamics and driver activity on emission levels. Additionally, exploring roadway conditions such as terrain, lane configuration, width, gradient, and weather variables like temperature, humidity, and wind direction could provide further insights into and enhance the prediction of vehicle emissions.

5. Policy implications

The study’s findings emphasize the significance of creating suitable testing facilities and procuring relevant equipment in Nigeria before vehicles may be registered. Legislation must include emission testing as a significant prerequisite for issuing roadworthiness certificates. According to the study, large buses, minibuses, and trucks have a significant impact on pollution levels. As a result, there is a need to minimize the number of trucks on the road and replace ageing minibuses and large buses with modern public transit vehicles. The report emphasizes the importance of enacting laws that limit the importation and usage of automobiles that are more than five years old. This strategy tries to reduce pollutants and carbon emissions while also contributing to general environmental improvement. Furthermore, to achieve long-term pollution reductions, the quality of gasoline products used by vehicles on the road must be improved. This could include improving fuel regulations and encouraging the use of cleaner, more sustainable alternatives. Overall, these recommendations emphasize the need for comprehensive measures such as emission testing, vehicle replacement, import limitations, and fuel refinement in addressing vehicle emissions and mitigating their impact on air quality and the environment in developing African cities.

Authorship

We attest that all authors contributed significantly to the creation of this manuscript. We confirm that the manuscript has been read and approved by all named authors and the order of authors listed in the manuscript has been approved by all named authors. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).