Integrated Urban Transport and Land-Use Policies in Reducing CO₂ Emissions and Energy Consumption: Case Study of a Medium-Sized City in Thailand

Abstract

In developing cities, transport activities have become one of the primary sources of CO₂ emissions and energy consumption owing to rapid economic growth, urbanization, and motorization. Khon Kaen City, Thailand, was chosen as a representative mid-sized city of a developing country to investigate the potential influences of transit-oriented development (TOD), light rail transit (LRT), and electric vehicle (EV) policy integration scenarios on CO₂ emission and energy consumption reductions in 2016, 2026, and 2046. The TOD did not significantly reduce CO₂ emissions or energy consumption because it was only applied in one area of the city. The LRT development also had a small effect because of the small proportion of modal shifts to LRT. However, EV utilization offered the greatest potential for reducing both CO₂ emissions and energy consumption. In addition, the integrated scenario combining the three policies had a promising effect, diminishing both CO₂ emissions and energy consumption, because it gathered the potential merits and benefits of each individual policy.

1. Introduction

At the beginning of the 21st century, 50% of the world’s population resided in urban areas and this figure is predicted to increase to 60% by 2030 [1]. In particular, the developing world has experienced rapid urbanization. By 2050, approximately 65% of the global population will live in urban areas in developing nations. From 2000 to 2030, the urbanization of Asian and African cities is expected to increase at the highest rate, double the current rate [2]. Urbanization results from the relocation of people from rural areas to cities because of (i) higher salaries, (ii) better health services, and (iii) superior educational opportunities [3]. The growth of urbanization (and therefore, urban passenger and freight transportation) leads to an increase in CO₂ emissions and energy consumption in both developed and developing countries [3,4,5,6,7].

Transport activity in urban areas is a principal generator of global CO₂ emissions and energy consumption, and this trend continues to increase despite vast improvements in advanced vehicles, engines, and fuel technologies. The transport sector accounted for 51.9% of global oil demand in 2022 [8] and 23% of global CO₂ emissions in 2021 [9]. Moreover, the transport sector still relies on oil products for more than 90% of its final energy needs [10]. Many countries have been attempting to promote various international laws and regulations for climate change mitigation, aiming to limit the rise in global temperature to below 2 °C compared to pre-industrialized levels [11]. In 2015, the UNDP announced 17 sustainable development goals (SDGs) associated with 169 targets to promote an equilibrium among economic, societal, and environmental constituents of the sustainable development principle and stimulate the implementation of important actions over the next 15 years [12]. The SDGs directly related to global warming and climate change include SDG 11: “Take urgent action to combat climate change and its impacts”, and Target 11.6: “Reduce the adverse environmental impact of cities” [12]. The SDGs and their associated targets were set up to encourage both developed and developing countries to combat global climate change and energy utilization crises. More recently, the Intergovernmental Panel on Climate Change (IPCC) [13] indicated the need for rapid decarbonization, requiring increased attention to planning more adaptive and resilient cities, with the revised aim of limiting the global temperature rise to no more than 1.5 °C above the 20th century levels by 2050. However, CO₂ emissions and energy consumption reductions in the transport sector may be difficult and complex to achieve [14].

Therefore, it is crucial to propose potential integrated transport and land-use strategies and policies for both developed and developing cities to minimize greenhouse gas (GHG) emissions and energy usage from transport. Several transport strategies and policy instruments have been proposed to reduce private car dependency, decrease transport energy consumption, and consequently minimize GHG emissions. Nakamura and Hayashi [15] proposed three important sustainable transport strategies to mitigate GHG emissions from the transport sector: avoid (A), reducing or eliminating unnecessary travel demand; shift (S), shifting necessary travel to lower GHG-intensive modes, and improve (I), improving the GHG emission reduction performance, together with four individual policy instruments related to each mentioned strategy, including technology (T), regulation (R), information (I), and economics (E). More recently, a fourth strategy, share (Se), shifting from individual to communal car ownership and use, has been proposed in response to new developments in modal choice and vehicle ownership (car sharing, ride sharing, bike sharing, etc.) spurred on by the new capabilities of information technologies [16,17].

The impacts of these strategies have not been as expected. Banister [18] suggested a wide variety of transport and land-use strategies leading to smart growth in cities that can minimize private vehicle trips and their associated travel distances. According to the then-British GHG emission reduction target of 80% by 2050, Hickman et al. [19] evaluated the potential reduction of CO₂ emissions in London using various transport strategies. They concluded that it was difficult to control GHG emissions from the transport sector by applying separate transport policies, such as speed restrictions, alternative fuels, public transport systems, air emission taxes, and pricing regimes. On the other hand, Banister [20] noted that several advanced technologies, such as electric vehicles (EVs) and alternative fuels alone, might never effectively deal with the rapid growth rate of automobile usage and might ultimately increase travel demand. Bubeck et al. [21] noted that neither bus rapid transit (BRT) nor a high-speed rail system could achieve the GHG emission reduction goal. They suggested that the integration of other minor public transport instruments (e.g., feeder systems) into the main public transport system can assist in meeting the GHG emission reduction targets. Furthermore, in-depth research is needed. Philp and Taylor [22] conducted a comprehensive literature review on the low-carbon mobility of New World cities, suggesting the following research topics: (1) urban design and transport–land use integration, (2) low-carbon mobility policies via behavioral changes, (3) the adoption of novel technologies and their applications, and (4) analysis and tools for the decision-making process.

In developing countries, most cities have suffered from the joint phenomena of urban sprawl, a rapid increase in car dependency, a lack of sustainable transport infrastructure, and inefficient services provided by the existing public transport systems. Some of the most pronounced adverse effects are traffic congestion, road accidents, environmental deterioration, inefficient energy use, and social inequity, which are further confounded by global warming and climate change. In Thailand, the transport sector has become the second-largest contributor to energy consumption (38.4%) and GHG emissions (29.2%) from the energy sector [23], because of the rapid expansion of the economy, population, urbanization, and motorization. Consequently, the growth rate trends of both energy use and its associated CO₂ emissions generated from the transport sector have altered significantly [24]. Pita et al. [24] suggested that energy efficiency improvement, public transport implementation, the use of biofuels, and EV utilization are the crucial policy measures to diminish energy use and CO₂ emissions from the road passenger transport sector in Thailand.

Several studies were conducted to assess the impacts of the implementation of both separated and integrated urban land-use and transport strategies and policy measures on road transport GHG emissions. For example, the S6-5 Group [25] applied the leapfrog/back-casting approach to the Khon Kaen town plan (KKTP) area transport network and introduced various potential scenarios for CO₂ emission reduction. They concluded that the integration of the avoid (A) strategy (transit-oriented development (TOD) measures), shift (S) (BRT measures), and improve (I) (use of electric and hybrid vehicles measure) strategies could lead to an approximately 50% reduction in the total CO₂ emissions from transport in 2030 compared to the business-as-usual (BAU) scenario. Chindaprasirt et al. [26] applied the multi-modal travel demand model (MTDM) in association with the bottom-up 2 method to quantify the amount of CO₂ emission reduction derived from implementing the blue line (MRT transit system) extension project in Bangkok. Ratanavaraha and Jomnonkwao [27] applied statistical techniques to estimate the amount of CO₂ emissions from energy use. Their analysis suggests a range from a minimum of 91.68 to a maximum of 225.33 million tons of CO₂ per year from the transport sector in Bangkok. Phdungsilp [28] found that in the transport sector, the modal shift from typical passenger cars to mass transit modes clearly plays an important role in reducing energy demand, CO₂ emissions, and air pollution. Klungboonkrong et al. [29] proposed different (both individual and integrated) transport and land-use scenarios for Khon Kaen University (KKU) in Khon Kaen city, Thailand. They noted that the use of electric motorcycles (EMCs) had the greatest effect on combating CO₂ emissions, followed by land-use control instruments, restrictions on private vehicle utilization, and public transport modifications. Long et al. [30] recently adopted the KKTP area in Khon Kaen, Thailand, as a case study and found that the implementation of the light rail transit (LRT) system and other potential measures (such as park and ride (P&R) and improvement of the feeder system) could, to a certain extent, reduce CO₂ emissions. APERC [31] evaluated combined fuel economy and urban planning policies in Asia-Pacific Economic Cooperation (APEC) economies and assumed that efficient urban planning will lead to a 5–20% lower level of vehicle saturation compared to the business-as-usual (BAU) scenario. The compact city and sustainable urban planning contributed about one-third of the total energy savings in the improved efficiency scenario in 2040 compared to BAU.

The previous studies found that TOD (avoid, A), public transport improvement (shift, S), and the use of EVs (improve, I) have the potential to reduce GHG emissions and energy consumption. However, comprehension of their explicit impacts is lacking, particularly in medium-sized cities in developing countries with high-growth urbanization. The main objectives of the research are to analyze, evaluate, and compare the CO₂ emissions and energy consumption reduction performances for three individual scenarios, TOD, LRT, and EVs, and two integrated scenarios in Khon Kaen city, Thailand.

The remainder of this paper is organized as follows: Section 2 describes the framework employed for the potential reduction of CO₂ emissions and energy consumption levels for a mid-sized city. Section 3 discusses model development. Section 4 presents the results of the analysis and discussion. Finally, Section 5 draws the conclusions.

2. Research Methodology

The research methodology employed in the study used the steps shown in Figure 1. The overall study process was to examine the potential reduction of CO₂ emissions and energy consumption levels in a mid-sized city (Khon Kaen) in Thailand under a range of transport and land-use policy scenarios.

As step 1, illustrated in Figure 1, was previously described in Section 1, the remainder of this section outlines the work undertaken to develop the transport and land-use scenarios for the study area (steps 2–4). The following section (Section 3) considers the model development (step 5) and Section 4 describes the modeling results and the analysis and evaluation of the policy scenarios (steps 6–8).

2.1. The Case Study Area

The Khon Kaen town plan (KKTP) area, covering approximately 228 km², is located in the northeastern region of Thailand (see Figure 2). Khon Kaen is located approximately 445 km from Bangkok. In 2022, approximately 380,557 people resided in the KKTP [32]. The general trends for gross provincial product (GPP), employment, population, and registered vehicles for Khon Kaen are shown in Figure 3. Typically, GPP, population, and registered vehicles increased throughout the period 2006–2022. Employment fluctuated over the period but showed overall net growth. The KKTP was selected as the case study area because it is one of the most strategic regional cities in Thailand. Khon Kaen city has been recognized as a hub for transport and logistics, education, medical services, meetings, and incentive travel, in addition to being a convention and exhibition (MICE) city, a low-carbon city, and a smart city. Khon Kaen has been actively working towards becoming a low-carbon city through various initiatives, such as public transport improvement, the integration of renewable energy sources into its infrastructure, waste management, and urban planning to reduce GHG emissions [33]. The Khon Kaen smart city emerged as part of Thailand’s broader initiative to develop smart cities across the country. This initiative focuses on leveraging information and communication technology (ICT) systems to enhance urban living [34]. Several previous studies have used it as a case study, ranging from a public transport master plan and detailed design study to a low-carbon society research study [25,32,35,36]. Khon Kaen city could be representative of a medium-sized city in a developing country.

2.2. Data Collection

The study area was divided into 182 internal zones and 5 other external zones for primary data collection. In 2016, origin and destination (O–D) surveys (for internal trips) were conducted using a household interview survey (HIS). Approximately 6100 households (4.0% of the total households) were directly interviewed. Additionally, other remaining trips (such as internal–external, external–internal, and external–external trips) in the study area were observed via a roadside interview survey (RIS) at six sites, with a sample size of approximately 5600 interviews. Traffic counts at 44 mid-block locations and their associated traffic compositions were collected. For secondary data collection, most of the data were derived from the SIRDC [32]. Current and future socioeconomic and demographic characteristics (including population, students, and employment), road network physical characteristics (including road hierarchy classes, number of lanes, lane width, and others), current and future highway and transit network data, and land-use characteristics (including land-use types, densities, and others) were obtained from the SIRDC [32].

2.3. Current Traffic and Transport Problems

There was considerable growth in the built-up area of the KKTP between 1990 and 2015. There is clear evidence of sprawling urban development (residential areas, commercial areas, and service facilities) away from the central business district (CBD) during this period [39,40]. This has been a major land-use issue in the KKTP area. Based on the surveyed traffic and transport data [32], the key traffic and transport problems are traffic congestion, road safety, and inefficient and unsuitable public transport systems (including problems such as unsafe services, no service timetables, inappropriate bus stop locations, excessive travel times, and unreliable services). For general road traffic, excessive mean delays and long queue lengths at several intersections during peak hours were observed. Furthermore, the average operating speeds of most roads in the Khon Kaen central business district (CBD) are below 10 km/h during peak hours [32].

2.4. Scenario Development

To deal with the climate change issues, the individual and integrated scenarios developed in this research study utilized the Comparative Study on Urban Transport and the Environment (CUTE) framework [41], including three main strategies: avoid (A) (reduce unnecessary travel), shift (S) (change travel to lower CO₂ emission modes), and improve (I) (improve the CO₂ emission reduction performance of transport technologies). Furthermore, three strategies were associated with four policy instruments, including technology (T), regulation (R), information (I), and economy (E). Three individual scenarios and two integrated scenarios were proposed and compared to the baseline (without project) scenarios, in terms of the CO₂ emissions and energy consumption performance, in the designated years 2026 and 2046. These scenarios are briefly discussed below and are summarized in

Table 1. Summary of the proposed scenarios.

Scenarios	Details
0	Without a project (baseline)
1	Five LRT lines being developed with feeder systems, P&R facilities, ITS signalized intersections, and other related facilities (shift)
2	Scenario 1 + TOD development (shift + avoid)
3	Electric vehicle usage by replacing existing motorcycles with electric motorcycles (EMCs) and changing typical passenger cars to electric passenger cars (EPCs) (improve)
4 = 1 + 3	Integration of scenarios 1 and 3 (shift + improve)
5 = 2 + 3	Integration of scenarios 2 and 3 (shift, avoid, and improve)

.

Scenario 0—without a project, i.e., baseline case: This scenario assumes that the existing road network and transport system characteristics will remain constant during 2016 and 2046, except for the inclusion of some previously committed government projects, as shown in Figure 4. Future socioeconomic data (e.g., population, students, employment, economic situation, etc.) in specific years were extrapolated based on time series trends. The impacts of megaprojects under Thailand’s transport infrastructure development plans (e.g., the high-speed train project, double-track railway project, motorway project, etc.) could then be estimated accordingly. Based on the historical trends in the growth rates of population and employment in the study area and the future potential trends of those growth rates (as a result of Thai national transport infrastructure development projects, such as high-speed train projects) and Khon Kaen’s status as an educational, medical, transport, and logistics hub [32], the growth rates of population and employment in the study area were assumed to be 2.75% in 2016 and 3.25% in 2046. In contrast, based on historical census statistics [32], the number of students was assumed to decline by 1.8 percent per year from 2016 to 2026 and be constant from 2026 to 2046 because of the expected impact of the future potential trends described previously.

Scenario 1—light rail transit (LRT) measure: This scenario falls under the shift (S) strategy, and technology (T) policy instruments. Scenario 1 assumes that all five LRT lines, the associated four P&R facilities, several feeder systems, the intelligent transport system (ITS) signalized intersections, and other facilities will be completely developed and fully implemented by 2026 [32], as shown in Figure 4. In addition, the existing paratransit systems (i.e., “Songthoew” modified pick-up trucks, public vans, and shuttle buses) will be reorganized as the feeder systems for all LRT service lines.

Scenario 2—transit-oriented development (TOD) plus light rail transit (LRT) measure: This scenario fits both the avoid (A) and shift (S) strategies and technology (T) policy instruments. Scenario 2 assumes that all five LRT lines and their associated infrastructure and facilities (Scenario 1), as well as the TOD development, will be in place by 2026 [32], as shown in Figure 4. TOD development can reduce CO₂ emissions and energy consumption by increasing urban densities and mixed land use to minimize unnecessary travel and encourage shorter trips, active transport modes (walking and cycling), and the use of LRT services. Three potentially integrated TOD sites were selected. For the TOD site with a radius of approximately 500 m around the LRT station, the average densities of housing surrounding the station were predicted to be 80 and 135 housing units/hectare [32] in 2026 and 2046, respectively. The scenario assumes that a considerable number of people, students, and jobs concentrated in the TOD were previously situated in the hosting traffic zone and that these aspects will increase over the years. The population, students, and employment were assumed to be completely relocated from areas outside the KKTP area. In addition, half of the total additional trips produced in the TOD development were assumed to be intra-zonal TOD trips (e.g., walking and cycling), while the other half included travel to other traffic zones similar to those living in the hosting traffic zones.

Scenario 3—electric vehicles (EVs) development measure: This scenario falls under the improve (I) and technology (T) policy instruments. In Scenario 3, the new vehicle types, electric motorcycles (EMCs) and electric passenger cars (EPCs) will replace the existing gasoline motorcycles (MCs) and gasoline or diesel passenger cars (PCs) in the KKTP area. The scenario assumes that 20% and 60% of the existing motorcycles will be shifted to EMCs in 2026 and 2046, respectively, whereas 10% and 30% of the existing passenger cars will be replaced with EPCs in 2026 and 2046, respectively. The projected adoption rates of EPCs in 2026 and 2046 seem reasonable when determining the actual sales share of EPCs (i.e., plug-in hybrid and battery electric cars) in the Khon Kaen province in recent years, which were 1.9% and 11.1%, respectively [42]. Such circumstances resulted from Thailand’s ambitious commitments to combatting climate change, as well as the recently reduced upfront costs of EPCs in Thailand. In addition, EVs have been predicted to become the next generation of low-emission vehicles (LEVs) [15].

Scenario 4—scenarios 1 + 3: This scenario involves the integration of scenarios 1 and 3. The combined effects of the two scenarios on CO₂ emissions and energy consumption reduction performance were examined.

Scenario 5—scenarios 1 + 3: This scenario is a combination of scenarios 2 and 3. This study investigated the co-benefits of the integrated strategies (avoid, shift, and improve) in terms of CO₂ emissions and energy consumption reduction outcomes.

Figure 4. The five proposed LRT lines in KKTP (adapted from [25,32,43,44] and designed by Freepik).

Figure 4. The five proposed LRT lines in KKTP (adapted from [25,32,43,44] and designed by Freepik).

3. Model Development

Exploration of the potential impacts of the scenarios required the development of suitable large-scale models. The first of these was a multimodal travel demand model (MTDM) to estimate future usage of the transport system by travel mode. The second model was for the CO₂ emissions and energy consumption of different vehicle types to be used in the study area.

3.1. Development of Multimodal Travel Demand Model (MTDM)

Several studies, such as S6-5 Group [25], Klungboonkrong et al. [29], SIRDC [36], and Tankasem et al. [45] similarly adopted the MTDM to forecast travel demands in the KKTP area. In this study, the MTDM was applied to forecast the travel demands in the KKTP transport network for 2016, 2026, and 2046. CUBE software (Version 5.0) [46] was used as the software development platform for the MTDM. In the trip generation step, the category analysis trip rates per capita per day, classified according to vehicle ownership classes—no vehicle, one motorcycle (MC), more than two MCs, one passenger car (PC), more than two PCs, one MC and one PC, and more than one MC and more than one PC [32]—were used to estimate the total trip production, while the multiple linear regression analysis method was employed to generate the trip attraction equations. In 2016, the total calculated trip generation was 656,500 trips per day. In 2026 and 2046, the total trip generation was predicted to be 860,670 and 1,518,130, respectively. The trip purposes of the trip generation step were divided into four categories: home-based work (HBW), home-based school (HBS), home-based other (HBO), and non-home-based (NHB). Based on the total trips made, 59%, 24%, 10%, and 7% of trips were associated with HBW, HBE, HBO, and NHB, respectively. In the trip distribution step, the well-known gravity model [25,29,45] was applied to predict the trip interchanges between zones [32]. In the modal choice step, the logit model was utilized to split the travel modes into either transit or private modes. The modal splits of the KKTP area in 2016 were as follows: motorcycle, MC (54%); passenger car, PC (32%); public transport, PT (9.0%), and walking and cycling (5.0%) [30,32]. In the trip assignment step, an equilibrium approach was adopted to assign all trips to the KKTP highway and transit networks. In this step, the total travel demand (trips) comprises the internal trips derived from the household interview survey (HIS) and the external trips obtained from the roadside interview survey (RIS). The internal trips are composed of trips made by MC, PC, PT, walking, and cycling, while the external trips comprise PCs, MCs, buses, pick-up trucks, and other trucks.

Finally, the predicted trips based on the MTDM were validated against the observed trip volumes on the six-screen lines for 16 validation sites, as shown in

Table 2. The results of model validation [32].

Screen Lines	Validating Points	Number of Traffic (PCU/hr)		RMS (%)
		Observed	Modeled
SL1 Railway line	4	15,652	14,330	−8.4
SL2 North–south (north)	1	2252	2391	6.2
SL3 North–south (south)	2	4947	4987	0.8
SL4 East–west (west)	2	4949	4887	−1.3
SL5 East–west (west)	2	6067	5248	−13.5
SL6 Cordon line (ring road)	5	13,989	14,639	4.6
Total	16	47,856	46,482	5.5

. The results of such validation clearly illustrate that the modeled trips were reasonably well fitted with the surveyed data, with an average root mean square (RMS) error of 5.5% [32], suggesting that the developed MTDM can predict multimodal travel demand.

3.2. CO₂ Emissions and Energy Consumption Estimation

The estimation of CO₂ emissions and energy consumption under the different scenarios required models to be applicable to the actual vehicle types used in the KKTP. Previous research had produced such models for use in Thailand, as described below.

Currently, three main approaches are used to estimate CO₂ emissions from the urban transport sector, including the top-down method, the bottom-up 1 method, and the bottom-up 2 method. Klungboonkrong et al. [29] and Long et al. [30] provided technical comparisons of these three methods. The three principal methods for calculating CO₂ emissions are briefly summarized as follows.

First, the top-down approach is commonly adopted to estimate the quantity of CO₂ emissions from the total amount of energy consumed by examining the historical statistics of energy consumption. Subsequently, the energy consumed is utilized to calculate the amount of CO₂ emissions by adopting the appropriate CO₂ emission factors for each vehicle type and energy type.

Second, the bottom-up 1 approach is normally applied to quantify the amount of CO₂ emissions from each vehicle. Then, the sum of all of the CO₂ emissions of all vehicles yields the total CO₂ emissions.

Third, the bottom-up 2 approach is typically employed to compute the CO₂ emissions of the traffic volume on each individual road segment of the transport network. Subsequently, the sum of the total CO₂ emissions of each road segment is the total CO₂ emissions of the entire road network.

The bottom-up 2 approach is appropriate for utilizing the MTDM output data, which include traffic volumes and average speeds on all segments of the transport network. Consequently, the bottom-up 2 approach was adopted in this study.

This method predicts CO₂ emissions and energy consumption using link-based traffic volumes, disaggregated by vehicle type and engine type, and their associated average speeds and link distances in the designated years. Based on the bottom-up 2 approach, the total CO₂ emissions and energy consumption can be calculated using Equation (1) [29,36,47]. Additionally, Equation (1) can calculate the amount of energy consumption by replacing the following input variables: (1) TE_k is to be replaced by “fuel economy in liters” and (2) Ef_kij is to be replaced by “specific fuel economy rate in liters per kilometer (L/km)”.

TE_k = ∑_i∑_j∑_k (Ef_kij × D_e × V_ije)

(1)

where:

TE_k = Total daily emissions from energy type k from all vehicle types i and engine type j of all links (e) in a road network (g);

Ef_kij = Emission factor of energy type k for vehicle type i and engine type j (g/km);

D_e = Road distances of link (e) (km);

V_ije = Traffic volume of vehicle type i and engine type j on all links (e) (number of vehicles per day).

The baseline CO₂ emissions and energy consumption were estimated for both the without-project (baseline) cases and the with-project CO₂ emissions and energy consumption scenarios for 2026 and 2046. Assuming leakage emissions (LE_y) are zero, the potential CO₂ emission (and energy consumption) reductions (ER_y) will be the difference between the CO₂ emissions in the without-project (Baseline Emission, BE_y) cases and those in the with-project (project emission, PE_y) cases. The reduction in CO₂ emissions and energy consumption can be estimated using Equation (2) [29].

ER_y = BE_y − PE_y − LE_y

(2)

where:

ER_y = CO₂ emission (or energy consumption) reductions in year “y” (tonne/year);

BE_y = Baseline CO₂ emissions (or energy consumption) in year “y” (tonne/year);

PE_y = Project CO₂ emissions (or energy consumption) in year “y” (tonne/year);

LE_y = Leakage CO₂ emissions (or energy consumption) in year “y” (tonne/year).

For the CO₂ emissions and energy consumption computations, the emission factors and energy consumption rates for each vehicle type and each engine type were obtained from the SIRDC [48], the OTP [49], Klungboonkrong et al. [29], S6-5 [25], and Tankasem et al. [45]. The vehicle types in the KKTP are typically classified into five categories: motorcycles (MCs), passenger cars (PCs), pick-up trucks (PUTs), buses (Bs), and trucks (Ts) [30]. CO₂ emissions and energy consumption at different driving speeds, obtained from the chassis dynamometer tests by the OTP [49], could formulate the relationships between CO₂ emissions and energy consumption with driving speeds for different types of vehicles tested, as shown in

Table 3. The selected emission and fuel economy factors by vehicle type and engine technology, adapted from [48,49].

Vehicle Types	CO2 Emission Factors And Fuel Economy Factors
	CO2	R2	Fuel Economy	R2
Gasoline motorcycle (MC)	Y = 0.0169X2 − 1.3909X + 57.674	0.80	Y = −0.0149X2 + 1.2391X + 37.895	0.68
Gasoline passenger car (PC)	Y = 0.0686X2 − 7.7792X + 340.01	0.92	Y = −0.0028X2 + 0.3642X + 5.4924	0.98
Diesel pick-up truck (PUT)	Y = 0.0695X2 − 8.0605X + 373.14	0.94	Y = −0.0024X2 + 0.3346X + 5.6714	0.98
Medium diesel bus (B)	Y = 0.1955X2 − 21.345X + 836.12	0.90	Y = −0.0003X2 + 0.1277X + 3.0026	0.97
Heavy diesel truck (T)	Y = 0.2889X2 − 33.506X + 1493.7	0.93	Y = −0.0011X2 + 0.1238X + 1.6474	0.87

Remarks: X = average traveling speed (km/h), and Y = CO₂ emission factors (g/km) or energy consumption rates (km/liter).

, with R² values ranging from 0.68 to 0.98. When vehicles travel at lower speeds, CO₂ emissions increase and fuel efficiency decreases for all types of vehicles, as shown by the equations. This trend reverses when vehicles travel at higher speeds. A similar trend could be noticed from the data obtained from the OTP [49]. For instance, a PC model (Toyota Altis Euro II) with a gasoline 91 engine emitted 218.6 g/km of CO₂ and generated fuel consumption of 10.7 km/liter at a speed of 14.6 km/h. On the other hand, it produced 135.1 g/km of CO₂ and achieved 17.2 km/liter fuel consumption at a speed of 73.9 km/h.

In the LRT project (Scenario 1), the operation of the LRT system also required that the electricity needed to power the LRT vehicles be estimated using Equation (3) [30]. Notably, various factors (such as the number of coaches used, service frequency, and distance) contribute to the total amount of electricity consumed by all five LRT systems. The comprehensive details of scheduling the actual operations for each of the five LRT systems are beyond the scope of this study. Based on empirical studies [30,50], the amount of electricity used for servicing all five LRT systems in Khon Kaen city was assumed to be solely dependent on the magnitude of the total estimated LRT ridership to be served. The CO₂ emission factors for electricity generation and distribution were obtained from the Energy Policy and Planning Office (EPPO), as shown in Figure 5. For the electricity used in the public transport (LRT) services (EC_sky), a time factor of 365 days was applied to convert daily transit ridership to annual transit ridership [30,32,49,50]. Assuming that the leakage emission is zero, the net CO₂ emission reductions could be computed by subtracting the difference between the baseline (without project) emission (BE_y) and the project (with project) emissions (PE_y) using the direct project emission (DPE_y), as presented in Equation (3).

DPE_y = EC_sky × EF_grid

(3)

where:

DPE_y = Direct project emission in year “y” (tCO₂/yr);

EC_sky = Amount of electricity used for public transport operation in the year “y” = 0.39 MWh/rider × number of transit ridership/day in year “y” × 365 days (MWh/yr);

EF_grid = The emission factor of grid electricity in the year “y”.

The CO₂ emissions from electric vehicles, both electric passenger cars (EPCs) and electric motorcycles (EMCs), are calculated based on the distance traveled (Equation (4)). The average energy consumption for an electric passenger car (EPC) is about 21 kWh/100 km. For an electric motorcycle (EMC), it is about 4 kWh/100 km. The electricity consumption for electric vehicles (EVs) is assumed to be constant throughout the modeling period. These figures represent the average urban electricity consumption values for medium-sized electric passenger cars and motorcycles available in Thailand. The assumed fuel economy of electric vehicles (EVs) aligns with the assumptions adopted in the Global EV Outlook [51]. The emission factors from electricity consumption are based on national grid emission factors provided by the Energy Policy and Planning Office [52], and future grid emission factors are based on the Power Development Plan (PDP) 2018, Revision 1 [53]. Declining grid emission factors reflect the national policy of increasing the share of renewable energy in power generation to meet Thailand’s nationally determined contribution (NDC) target of reducing GHG emissions by 20–25% by 2030 [54]. However, the future grid emission factors in the PDP are provided up to the year 2037, the end year of the plan. The grid emission factors for the forward years are assumed to be the same as the year 2037. The grid emission factors of Thailand from 2016 to 2046 are shown in Figure 5.

DPE_EV,y = [(PC_EV,y × KT_EV,y × EC_EV,y) + (MC_EV,y × KT _EV,y × EC_EV,y)] × EF_grid,y

(4)

where:

DPE_EV,y = Direct project emissions of electric vehicles in year “y” (tCO₂/year);

PC_EV,y = Number of electric cars in the year “y”;

MC_EV,y = Number of electric motorcycles in the year “y”;

KT _EV,y = Kilometre-travelled of electric vehicles;

EC_EV,y = Electricity consumption for electric vehicles in the year “y” (assumed as a constant value, 21 kWh/100 km for electric cars and 4 kWh/100 km for electric motorcycles);

EF_grid,y = The emission factor of grid electricity in the year “y”.

Figure 5. The grid emission factors from 2016 to 2046 (adapted from [52,53]).

Figure 5. The grid emission factors from 2016 to 2046 (adapted from [52,53]).

4. Results and Discussion

Based on the developed MTDM, the total daily trips generated in 2016 (base year), 2026, and 2046 are 656,500, 860,670, and 1,518,130, respectively. The public transport modal shares in 2016, 2026, and 2046 were 9.0%, 15.5%, and 18.0%, respectively. The percentage growth in the estimated daily trips in 2026 and 2046 compared with those in the base year (2016) was 31.2% and 131.3%, respectively. The total trip distance (vehicle kilometres of travel (VKT) per hour), total travel time (vehicle hours of travel (VHT) per hour), CO₂ emissions, and energy consumption for the base case in 2016 and for the baseline cases in 2026 and 2046 were estimated and are summarized in

Table 4. The total baseline traffic-related indicators, energy consumption, and CO₂ emissions [30].

Indicators	Base Year	Baseline Cases
	2016	2026	2046
Public transport modal shares (%)	9.00	15.50	18.00
Total daily trips (trips/day)	656,500	860,670	1,518,130
Total trip distance, VKT (veh-km/h)	447,544	536,352	790,864
Total travel time, VHT (veh-h/h)	14,447	20,477	82,115
Energy use (ktoe/year)	75.10	94.67	172.94
CO2 (tonne/y)	226,410	289,997	538,456
Population	323,409	424,200	729,805

Remark: Population annual growth rate = 2.75%.

and

Table 5. The baseline traffic-related indicators, energy consumption, and CO₂ emissions per capita.

Indicators	Base Cases	Baseline Cases
	2016	2026	2046
Public transport modal shares (%)	9.00	15.50	18.00
Total daily trips (trips/capita/h)	2.030	2.029	2.080
Total trip distance, VKT (veh-km/capita/h)	1.384	1.264	1.084
Total travel time, VHT (veh-h/capita/h)	0.045	0.048	0.113
Energy use (toe/capita/y)	0.232	0.223	0.237
CO2 (tonne/capita/y)	0.700	0.684	0.738

.

4.1. Analysis and Evaluation of PT Modal Shares, VKT, and VHT Reductions for Each Scenario

The PT modal share performances for each scenario and each year were modeled, and the results are shown in

Table 6. The impacts of each scenario on the magnitudes of PT modal shares, VKT, VHT, CO₂ emissions, and energy consumption per capita in the base year 2016, 2026, and 2046.

Scenarios

Base year 2016

PT modal shares (% modal shares)

VKT (veh-km/capita/h) *

VHT (Veh-h/capita/h) *

CO2 (tonne/capita/y)

Energy use (Toe/capita/y)

Fossil fuels

Electricity

Total

Fossil fuels

Electricity

Total

0

9.00

1.384

0.045

0.700

n/a

0.700

0.232

n/a

0.232

Scenarios

Year 2026

PT modal shares (% modal shares) [Change] (%)

VKT (veh-km/capita/h) * [Reduction] (%)

VHT (veh-h/capita/h) * [Reduction] (%)

CO2 (tonne/capita/y) [Reduction] (%)

Energy use (toe/capita/y) [Reduction] (%)

Fossil fuels

Electricity

Total

Fossil fuels

Electricity

Total

0

15.50

1.264

0.048

0.684

n/a

0.684

0.223

n/a

0.223

1

22.55 [+7.05]

(+45.48%)

1.222 [0.042]

(−3.35%)

0.044 [0.004]

(−8.87%)

0.651 [0.033]

(−4.83%)

0.0205 n/a

n/a

0.671 [0.013]

(−1.83%)

0.213 [0.010]

(−4.63%)

0.0046 n/a

n/a

0.217 [0.006]

(−2.58%)

2

22.84 [+7.34]

(+47.35%)

1.23 [0.032]

(−2.59%)

0.055 [0.002]

(−4.42%)

0.660 [0.023]

(−3.40%)

0.0216 n/a

n/a

0.682 [0.002]

(−0.24%)

0.216 [0.007]

(−3.24%)

0.0048 n/a

n/a

0.221 [0.002]

(−1.09%)

3

15.50 [-]

(-)

1.264 [-]

(-)

0.048 [-]

(-)

0.628 [0.056]

(−8.14%)

0.0243 n/a

n/a

0.652 [0.031]

(−4.58%)

0.204 [0.019]

(−8.42%)

0.0054 n/a

n/a

0.210 [0.013]

(−6.00%)

4 = 1 + 3

22.55 [+7.05]

(+45.48%)

1.222 [0.042]

(−3.35%)

0.044 [0.004]

(−8.87%)

0.598 [0.086]

(−12.58%)

0.0440 n/a

n/a

0.642 [0.042]

(−6.14%)

0.195 [0.028]

(−12.66%)

0.0098 n/a

n/a

0.205 [0.018]

(−8.27%)

5 = 2 + 3

22.84 [+7.34]

(+47.35%)

1.232 [0.032]

(−2.59%)

0.055 [0.002]

(−4.42%)

0.607 [0.077]

(−11.27%)

0.0442 n/a

n/a

0.651 [0.033]

(−4.81%)

0.198 [0.025]

(−11.40%)

0.0099 n/a

n/a

0.208 [0.016]

(−6.99%)

Scenarios

Year 2046

PT Modal shares (% Modal shares) [Change] (%)

VKT (veh-km/capita/h) * [Reduction] (%)

VHT (veh-h/capita/h) * [Reduction] (%)

CO2 (tonne/capita/y) [Reduction] (%)

Energy use (toe/capita/y) [Reduction] (%)

Fossil fuels

Electricity

Total

Fossil fuels

Electricity

Total

0

18.00

1.084

0.113

0.738

n/a

0.738

0.237

n/a

0.237

1

31.11 [+13.11]

(+72.83%)

0.986 [0.098]

(−9.01%)

0.074 [0.039]

(−34.63%)

0.627 [0.111]

(−15.06%)

0.0231 n/a

n/a

0.650 [0.088]

(−11.94%)

0.200 [0.037]

(−15.52%)

0.0063 n/a

n/a

0.206 [0.031]

(−12.88%)

2

31.40 [+13.40]

(+74.44%)

1.037 [0.046]

(−4.27%)

0.095 [0.017]

(−15.26%)

0.638 [0.100]

(−13.51%)

0.0251 n/a

n/a

0.663 [0.075]

(−10.11%)

0.209 [0.028]

(−11.92%)

0.0068 n/a

n/a

0.216 [0.021]

(−9.05%)

3

18.00 [-]

(-)

1.084 [-]

(-)

0.113 [-]

(-)

0.560 [0.178]

(−24.12%)

0.0623 n/a

n/a

0.622 [0.116]

(−15.68%)

0.177 [0.060]

(−25.19%)

0.0139 n/a

n/a

0.191 [0.046]

(−19.34%)

4 = 1 + 3

31.11 [+13.11]

(+72.83%)

0.986 [0.098]

(−9.01%)

0.074 [0.039]

(−34.63%)

0.474 [0.263]

(−35.71%)

0.0798 n/a

n/a

0.554 [0.184]

(−24.90%)

0.149 [0.088]

(−36.94%)

0.0189 n/a

n/a

0.168 [0.069]

(−28.98%)

5 = 2 + 3

31.40 [+13.40]

(+74.44%)

1.037 [0.046]

(−4.27%)

0.095 [0.017]

(−15.26%)

0.482 [0.256]

(−34.67%)

0.0827 n/a

n/a

0.565 [0.173]

(−23.46%)

0.156 [0.081]

(−34.33%)

0.0195 n/a

n/a

0.175 [0.062]

(−26.09%)

Remarks: * In the morning peak hour period; n/a = Not applicable.

. Among the three individual scenarios (1, 2, and 3) proposed in 2026, scenario 2 (transit-oriented development (TOD) plus five light rail transit (LRT) lines measure) performed minimally better than scenario 1 (five light rail transit (LRT) lines measure) in terms of the PT modal share performance. This finding reflects the fact that a single TOD development may not strongly influence PT modal share performance over the entire study area. The PT modal shares of scenario 3 (the utilization of EMC and EPC measures) were zero because the scenario assumes that the traveling patterns and modal split of both EMCs and EPCs remain identical to those of the existing MCs and PCs in the baseline cases. In addition, between the two integrated scenarios 4 and 5 in 2026, the PT modal share performances of both scenario 4 (combining scenarios 1 and 3) and scenario 5 (integrating scenarios 2 and 3) were identical to those of scenarios 1 and 2, respectively. Similar trends for both the individual and integrated scenarios were observed for 2046.

The magnitudes of the VKT and VHT reductions per capita for each scenario in each year were estimated and are shown in Table 6. Among the three individual scenarios (1, 2, and 3) proposed in 2026, scenario 1 performed the best in terms of VKT and VHT reductions per capita, followed by scenario 2. Generally, the integration of LRT development and TOD development leads to shorter trips, fewer unnecessary trips, more usage of active transport modes (such as walking and cycling), and greater public transport shares. Therefore, the magnitudes of VKT and VHT per capita in scenario 2 were expected to be lower than those in scenario 1. However, the degree of VKT and VHT reduction per capita in scenario 2 does not reflect such expectations. This is because, as previously described in scenario 2, the population, students, and employment in the TOD were assumed to be relocated only from the areas outside the KKTP area, and 50% of the total trips produced by such people were intra-zonal trips in the TOD area, while the other 50% of factored trips to destination zones in patterns similar to those living in the original hosting traffic zones. The effects of such circumstances will override the impact of scenario 2 in terms of the VKT and VHT reductions per capita. Hence, the magnitudes of the VKT and VHT reductions per capita in scenario 2 were lower than those in scenario 1. As indicated previously, the travel patterns and modal splits for both EMCs and EPCs were the same as those for MCs and PCs. Therefore, the magnitudes of the VKT and VHT per capita under Scenario 3 were constant, and their VKT and VHT reductions per capita were zero. Among the two integrated scenarios 4 and 5 in 2026, the performance of scenarios 4 and 5 in terms of the VKT and VHT reductions per capita was identical to scenarios 1 and 2, respectively. Similar trends for both the individual and integrated scenarios were observed in 2046. Notably, the growing reductions in both VKT and VHT per capita in scenarios 1, 2, 4, and 5 were affected by the growth in population, employment, and number of students from 2026 to 2046.

4.2. Analysis and Evaluation of CO₂ Emission and Energy Consumption Performance for Each Scenario

Based on the developed MTDM models and the bottom-up 2 approach, the CO₂ emissions and energy consumption reductions per capita for each scenario in each specific year were calculated and are presented in Table 6. Among the three individual scenarios (1, 2, and 3) in 2026, scenario 3 performed the best in terms of CO₂ emission reductions and energy savings per capita, followed by scenarios 1 and 2, respectively. The magnitudes of CO₂ emission and energy use reductions per capita in scenario 3 were much greater than those in scenarios 1 and 2. This indicates that EVs (scenario 3) offer greater benefits and effectiveness than scenarios 1 and 2. Similar reasons, as previously described, can be adopted to explain the inferiority of scenario 2 compared to scenario 1. Among the two integrated scenarios 4 and 5 in 2026, scenario 4 performed better than scenario 5. This is because scenarios 4 and 5 include the combined merits and effectiveness of integrated scenarios 1 and 3, and scenarios 2 and 3, respectively. Similar trends for both the individual and integrated scenarios were observed in 2046.

It should also be noted that in 2026 and 2046, similar patterns of reductions in CO₂ emissions and energy consumption per capita in each scenario could be seen. However, the percentage of energy use reductions per capita for all (individual and integrated) scenarios was generally higher than that of CO₂ emission reductions per capita. This is because the CO₂ emission factors and energy consumption rates for each vehicle and engine type are uniquely different, as shown in Table 3. Consequently, the percentages of CO₂ emissions and energy consumption reduction per capita are distinct.

Electricity consumption is calculated in scenario 1 by introducing five lines of the LRT system and in scenario 3, by replacing internal combustion engine (ICE) passenger cars and motorcycles with electric vehicles. Since the CO₂ emission factors of electricity generation and distribution are applied in this study, CO₂ emissions from electricity use in LRT systems and electric vehicles were calculated. Interestingly, the results show that the proportion of CO₂ emissions from electricity use is higher than that from energy use in all scenarios. This shows that the PDP2018 Revision 1 emission factors are not low enough to decarbonize transport systems.

The integrated scenarios performed better than the individual scenarios because the integrated scenarios usually combined the potential benefits of each scenario in terms of the CO₂ emission and energy consumption reductions per capita. The significantly increased reductions in both CO₂ emissions and energy consumption per capita in the individual scenarios (scenarios 1, 2, and 3) and in the integrated scenarios (scenarios 4 and 5) from 2026 to 2046 arose because of the potential enhanced merits and benefits according to the increases in transport-related activities as a result of the transport and land-use shifts and the alterations in socioeconomic characteristics under those scenarios over the study period (from 2026 to 2046).

5. Conclusions

The MTDM method was applied to predict future travel demands and other traffic-related parameters, including the number of total trips, the VHT per capita per hour, and the VKT per capita per hour in the study area. The KKTP area may be considered representative of a medium-sized city in a developing country. Subsequently, the bottom-up 2 method was adopted to compute the magnitudes of CO₂ emissions and energy consumption under each scenario in each designated year. Three individual scenarios (scenarios 1, 2, and 3) and two integrated scenarios (scenarios 4 and 5) were then analyzed, compared, and assessed regarding the CO₂ emission and energy consumption reductions in 2026 and 2046. CO₂ emissions from the electricity grid are also considered in this study. The key findings of the study include:

• The TOD did not significantly reduce CO₂ emissions and energy consumption because it was applied only in one area of the city. This was also because of the characteristics of the TOD and the residential culture’s housing preferences. In practice, in small- and medium-sized cities, people do not prefer to live in high-rise buildings and high-density urban areas;
• The LRT development had a small effect because of the small proportion of modal shifts to the LRT. The development of mass transit alone makes it difficult to attract private vehicle users in developing countries [55]. A few barriers need to be well developed together, e.g., coverage area, quality of accessibility (walking, cycling, and feeder), interchange, and affordability. Developing such a seamless public transport system in a medium-sized city (low density of population, low demand, and limited resources) in developing countries is very difficult;
• EV usage showed the highest potential for reducing CO₂ emissions and energy consumption. However, it depends on how many vehicles would be switched to EVs and what to do with the old vehicles (them either still being in the city or moving to use in other cities would reduce the benefit of the EV usage policy). Moreover, EV usage cannot tackle other transport problems, particularly traffic congestion and accidents;
• The integrated strategy combining the three policies had certain effects on reducing CO₂ emissions and energy consumption, but most of the effects came from switching to EV usage. It seems that the integration of “pull” policies (preferable to the public) cannot generate a synergistic effect;
• Future studies could consider the improved adoption of electric vehicles (EVs) for all vehicle types, as Thailand aims to achieve carbon neutrality by 2050 and zero greenhouse gas (GHG) emissions by 2065. Furthermore, it is essential to explore alternative CO₂ emission factors in the electricity grid to meet these targets.

Overall, the study suggests that the integration of these policies would reduce CO₂ emissions and energy use more than a single policy. However, the selection of an optimal combination of policies is the key to achieving the target. To achieve the synergy of integration, “pull” policies and also some “push” policies (e.g., parking management and access control) should be considered for application in medium-sized cities, even if the transport problems are not as serious as those in mega-cities. According to the feasibility context, medium-sized cities should also consider the low investment cost of seamless public transport systems, e.g., bus rapid transport (BRT) or busways based on EV buses, to encourage modal shift as much as possible. This would deal with traffic congestion, accidents and other environmental impacts. The remaining motorcycles and cars used should be changed to EVs. Land-use planning and TOD in medium-sized cities in developing countries should be further studied to determine how to enable them to effectively fit local needs and culture. Finally, although the best integrated scenario regarding the CO₂ emission and energy consumption reductions per capita is achievable, actual operation and implementation are complicated. Crucially, successful operation and implementation rely on several key factors, including the strong leadership and commitment of both local and central authorities, appropriate financial resources, public participation and support, planning and design experiences, and professional expertise [29]. In practice, various barriers exist, including legal and institutional barriers (such as land acquisition of the P&R, depot, control center, and TOD development), financial barriers (such as the derivation of financial support and subsidies for transport infrastructure, TOD development, and EV purchase), political and cultural barriers (such as existing local public transport providers protesting against new public transport development), and practical and technological barriers (such as driving distance limitations, excessive charging time, and lack of charging infrastructure for EVs). Moreover, this study has had some limitations and drawbacks. They are as follows: (i) the MTDM (trip-based) model was applied to predict travel demand and other associated parameters — the more appropriate activity-based model (e.g., land use–transport interaction (LUTI) models) should be developed and adopted in future studies; (ii) the impacts of the COVID-19 pandemic were beyond the scope of this research, hence, the COVID-19 influences were not determined in this research, and (iii) only the LRT, TOD, and the EV policy measures were considered. Additional high potential policy measures should be adopted (as a package) to generate synergy and diminish those barriers. Eventually, the sustainable goal of CO₂ emission and energy consumption reduction will be achieved only when the optimal integrated transport and land-use policies and plans are publicly accepted and successfully implemented.