1- Introduction

2- Literature review

3- Methods and data

4- Results and discussion

5- Conclusions and implications

References

Abstract

With the rapid development of rail transit, a vast amount of electric power is consumed each year. However, the generation of electricity, and especially coal power generation, is an important source of greenhouse gases. Therefore, it is very important to analyze the potential of carbon emission reduction of urban rail transit based on specific electricity consumption structures. In this study, 18 cities in China were taken as the research object, and backward analysis was used to analyze the proportion limit of coal power consumption for rail transit in each city under three scenarios from 2015 to 2017. By comparing the limit value with the actual coal power consumed by urban rail transit, we analyzed the potential carbon emission reduction of rail transit relative to other traffic modes. The study results supported several conclusions. First, the traffic demand is in direct proportion to the carbon emission reduction potential in rail transit. Second, for cities with high coal power consumption, the development of ground bus transit is more conducive to achieving carbon reduction targets compared with rail transit. Finally, promoting the development of rail transit technology and lowering the energy consumption per capita unit travel distance are the fundamental ways to increase the emission reduction potential of rail transit. Therefore, as a modern transportation tool, rail transit is not an absolute emission reduction advantage for all cities. A city needs to analyze its own resource structure and travel demand together, set up suitable traffic modes to realize green growth and sustainable development.

Introduction

With the increasingly prominent traffic problems globally, the priority development strategy of public transport, especially rail transit, has been adopted by many countries. Rail transit has been vigorously promoted because of its characteristically large transport volume and low pollution potential. According to the China Urban Rail Transit Association, a total of 34 cities in mainland China had opened urban rail transit at the end of 2017, totaling 165 lines and total annual traffic volume of 18.5 billion passengers. The total length of operation lines reached 5033 km, 3884 km (77.2%) of which involved subways. Rail transit is an important part of an urban passenger transport system, and plays a great role in relieving urban traffic congestion. However, urban rail transit operations rely mainly on power resources; with the increase in operating rail length, energy consumption is increasing (Sun et al., 2018). According to the statistics of the China Urban Rail Transit Association, the energy consumption of rail transit in China reached 12.226 billion kWh in 2017. China predominantly generates electricity by burning coal; therefore, the large amount of electricity consumed by rail transit results in large amounts of greenhouse gases from the electricity generation process (Dong et al., 2018). Because of the development of urban rail transit and the huge increase of carbon emissions, the economic cost and the aggravation of environmental pollution have restricted green growth and sustainable development of cities in China and elsewhere (Song et al., 2018). Reducing the carbon emissions of rail transit systems has become a problem of widespread concern in recent years. Therefore, whether rail transit has more emission reduction advantages than other traffic modes is a question worthy of further examination. The carbon emissions arising from the consumption of electric energy in rail transit mainly come from coal power generation, and there are great differences in the power supply structure in various regions of China (Chen et al., 2018). The north of China is rich in mineral resources and relies heavily on coal power. By comparison, water resources in the south of China are abundant, and hydropower occupies an important position in the region’s power structure. The cleanliness (in terms of environmental emissions) of power in different regions is not the same (Pei et al., 2015). Therefore, the carbon emissions from the generation of electric power that is consumed by rail transit are also very different throughout China.

The assessment of transit service quality is a matter of academic interest since over thirty years already, and several research studies highly covered the scientific literature since the beginning of the new millennium. The growing interest for this topic also from transit enterprises and authorities is due to the awareness that the knowledge of the transit service performances can be used for monitoring the services and for identifying potential weak points, with the aim to increase passenger’s satisfaction and finally attract new transit users and retaining the usual ones. In fact, it is well known that in recent years users’ travel behaviour has become more complex, and people prefer private car for moving, because it is perceived as the most comfortable, flexible, and fast mode compared with the other ones. The urban sprawl and low density areas rising also contributed to increase the attractiveness of car as the most convenient transportation mode from all the points-of-view, with undesirable consequences in terms of environmental pollution, energy consumption and quality of life (Nocera and Cavallaro, 2016; Ayyildiz et al., 2017). In order to reduce these problems, travel demand has to be moved from private car to more sustainable mode as public transit systems. Consequently, transport companies have to operate interventions aimed to reduce travel time and increase service quality over all the aspects, since improvements in service quality can have a significant positive impact on the willingness of potential users to choose public over private transport (de Oña et al., 2014; Eboli et al., 2016). Transit service quality can be measured by different approaches. The most common approach is based on transit users’ opinions about the used services, collected by means of surveys. When service quality is measured from the customer’s perspective, transit quality depends on the passengers’ perceptions about each attribute characterising the service (de Oña et al., 2013). In recent years, the most popular methods for capturing transit users’ perceptions were: (1) traditional Customer Satisfaction Surveys (CSS), where users express their opinions by rating the various service characteristics (Weinstein, 2000; Cavana et al., 2007; Joewono and Kubota, 2007; Pakdil and Aydın, 2007; Dell’Olio et al., 2011; Eboli and Mazzulla, 2010, 2012b; Jen et al., 2011; Habib et al., 2011); and (2) Stated Preference surveys, where the importance given by the users to the service attributes indirectly derives from exercises based on the stated preferences according to which users express their opinions by choosing (or rating of ranking) alternative hypothetical services (Cirillo et al., 2011; Eboli and Mazzulla, 2008a, 2008b, 2010; Hensher and Prioni, 2002; Hensher et al., 2003; Dell’Olio et al., 2011). An alternative approach entails using different variables of the transit system demand and operation to calculate the ‘efficiency’ indicators (e.g. Badami and Haider, 2007; Lao and Liu, 2009; Eboli and Mazzulla, 2012a). There are also some studies integrating the two mentioned approaches (e.g. Sheth et al., 2007; Abreha, 2007; Nathanail, 2008; Eboli and Mazzulla, 2011).

Since China’s first high-speed railway passenger transport line, Beijing-Tianjin Intercity High-Speed Railway thoroughly opened in August 2008, Wuhan-Guangzhou (December 2009), Zhengzhou-Xi ‘an (February 2010), Shanghai-Nanning (July 2010), ShanghaiHangzhou (October 2010), Beijing-Shanghai (June 2011) and other high-speed railways were successively put into use, leading to a dramatic increase of the mileage of China’s high-speed railway. By the end of 2014, total operating mileage of high-speed railways in China had accumulated to 16,727 km, which was far more than that of any other country. By the end of 2015, the operating mileage of high-speed railways in mainland China reaches 19,915 km. The railway passenger volume in China also increased from 1085.79 million in 1986 to 2530 million in 2015, with an increase of 133.01%. Functionally speaking, the aviation transport and high-speed railway transport are all relating to transport passengers form one place to another place, to serve the passengers’ needs well. They are substitutable to a certain extent. With the characteristics of comfort, rapidness, convenience and punctuality rate, the high-speed railway has put a powerful influence on the aviation industry (Óscar et al., 2015; Marie and Frédéric, 2015; Daniel and Xavier, 2016; Mu et al., 2015; Chen and Haynes, 2015). Since Beijing-Shanghai High-speed Railway opened in 2010, the airline passenger volume from Beijing to Shanghai has dropped by 9.9%, from 7,473,355 in 2010 to 6,730,218 in 2013. Since Beijing-Guangzhou High-speed Railway opened in 2012, the airline passenger volume from Beijing to Wuhan has dropped by 23.8%, from 1,747,549 in 2013 to 1,330,864 in 2014. The airline passenger volume from Harbin to Shenyang has dropped by 70.9%, from 329,822 in 2010 to 95,856 in 2012. Therefore, the analysis on the shock effect of China’s high-speed railway on aviation transport in China has become such an immediate task for us. As for the study of high-speed railway and aviation transport, at present, the scholars at home and abroad studied the influence of high-speed railway to aviation transport, from the angle of demands of the high-speed railway (Maria, 2014), the influence of the high-speed railway to regional economy (Hyojin and Selima, 2015; Roger, 2015; Shailesh and Sharada, 2014; Guineng and João, 2014; Wang et al., 2012), the service cost of the aviation industry (Frédéric et al., 2014; Yang and Zhang, 2012), the shock to the traveling time brought by the high-speed railway (Kamga, 2015; Mu et al., 2015; Zhao et al., 2015; Daniel and Xavier, 2016), and the competition with high-speed railway from the different air routes (Jiang and Zhang, 2014; Óscar et al., 2015; Marie and Frédéric, 2015; Daniel and Xavier, 2016; Wu et al., 2014; Chen and Haynes, 2015). In which the shock from adjusting the railway network is the most obvious. By using the data from 1995 to 2009, Clewlow et al. (2014) analyzed the influences of the high-speed railway and budget airline to the demand of air travel at three different levels: air routes, traveling within the country and traveling in the selected European area. By taking GDP, population, population density, aviation fuel price, change of the time consumed by the railway, highspeed railway and the routes of budget airline as the variables, he used ordinary regression and panel regression to analyze the influences of the high-speed railway and budget airline to the demand of air travel.

On-the-track competition is a new phenomenon on the railway market, where operators compete on the same line for passengers. Similar reforms on the bus market have in some situations unleashed fierce, unsustainable competition and eventual return to monopoly; concerns have thus been raised over how prices, frequencies, travelvolumes and overall social welfare will develop. In particular, the existence of stable equilibrium with more than one player is questioned. The regulator plays a crucial role in this respect. Because of capacity constraints, it is by necessity involved in decisions regarding departure times and frequencies of service. Tweaks to regulatory proceedings alter the optimal strategies of market participants, and in turn social welfare and other outcomes. Assuming that the regulator strives to achieve a stable market equilibrium and high social welfare through the means of on-rail competition, what policies should be adopted? We suggest a simulation model with realistic parameters where operators compete on frequency and price. The model is built to facilitate comparison between different regulatory settings, including a duopoly market where the regulator allocates departure times; a profit-maximising monopoly; a large number of competitors; and others. It also includes bench-mark scenarios, such as welfare maximisation with a no subsidies constraint. The model allows for individual prices to be set for each departure, and it takes account of each departure’s relative position in time. In this way it builds explicitly on the hypothesis that competition occurs not only between operators but between departures that are close in time. This makes it possible to study how prices vary over the day, depending on the relative intensity of competition at certain times. Possible operator strategies to lessen the pressure of price competition are explored.

مقاله سال 2016

 مجله امگا – Omega

Development of an efficient plan for preventive maintenance is crucial in many fields [28,30,19]. The plethora of activities for which preventive maintenance must be scheduled includes railway transportation. To ensure a good daily service in terms of punctuality and safety, railway infrastructure managers have to plan and perform maintenance operations for whole railway networks. Most of these operations involve preventive maintenance at a tactical level within a medium time horizon, usually of 1 year. The aim of preventive maintenance activities is to control track failure probability to guarantee a stable and safe service and minimize maintenance costs. ACEM-Rail [26] is a European project to improve the optimization and automation of railway infrastructure maintenance. One of the objectives of the ACEM-Rail project is to develop new algorithms for efficient planning of railway infrastructure maintenance tasks based on stochastic data drawn from predictions. Several studies have addressed railway maintenance planning. Cheung et al. [9] studied the railway track possession assignment problem (RTPAP) using constraint satisfaction. The problem involves assigning railway tracks to scheduled maintenance tasks according to the satisfaction of a set of constraints. The aim of the RTPAP is to produce a plan that maximizes the assignment of jobs with the highest priority. To effectively achieve this goal, an engineering work track possession assignment system based on the CHIP constraint programming language [27] substituted manpower used to find a manual solution for the RTPAP. The performance of the system was ten times more efficient than the manual method, and its solutions are free of human errors. Budai et al. [6,7] proposed a preventive maintenance scheduling problem in which routine activities and projects were scheduled within a given time horizon by minimizing possession costs. The authors provided an integer programming model for the problem and proved that it is NP-hard. Moreover, they provided three fast but simple heuristics. To reduce the gap yielded by these heuristics, Budai et al. [8] tackled the same problem using genetic and memetic algorithms. van Zante-de Fokkert et al. [29] studied a problem in which the whole network was divided into basic working zones named single-track grids. The authors proposed a method yielding a maintenance schedule in two phases. Moreover, they provided a mixed integer programming model with a lexicographic objective function that first minimizes the number of nights (i.e., the length of the plan) and then the sum of the maximum workloads scheduled.

Although these methods efficiently address railway maintenance planning at a tactical level, they deal with a known set of maintenance tasks to be scheduled within a given time horizon. However, within the predictive setting of the ACEM-Rail project, future track conditions are not known in advance and the development of the deterioration process can be only predicted. Therefore, maintenance tasks are affected by uncertainty resulting from unknown deterioration processes. The resulting problem involving creation and solution of an efficient adaptive maintenance plan to minimize overall costs when maintenance tasks are not known a priori but subject to uncertainty is the Stochastic Tactical Railway Maintenance Problem (STRMP). We call the deterministic STRMP subproblem that the infrastructure manager has to solve in each month of the rolling horizon Deterministic Tactical Railway Maintenance Problem (DTRMP). Very little has been proposed so far in the literature to address the STRMP. Heinicke et al. [16] outlined a preliminary methodology in which capacity and risk constraints are treated as soft constraints. Nevertheless, their results were limited to three instances. The aim of our study is to complete the work started by Heinicke et al. [16] and provide a formal definition of the problem, improved heuristics to address the problem, and extensive computational results. We extend the preliminary work of Heinicke et al. [16] with a number of contributions. We provide a formal definition of the problem setting for both the STRMP and the DTRMP. Moreover we formulate an integer linear programming model for the DTRMP and treat capacity and risk constraints as hard constraints. The STRMP and the DTRMP show a number of analogies with the bin packing problem (BPP) [18], in particular with one of the latest variants, namely the stochastic generalized BPP (SGBPP) [22], and its deterministic variant, the generalized BPP (GBPP) [2]. Using a linear model and exploiting the analogies with these BPPs, we present three efficient heuristics that can address the DTRMP. The first heuristic, named ADAPTED FIRST FIT DECREASING (AFFD), is an adaptation of the FIRST FIT DECREASING (FFD) heuristic [18], which is widely used for BPPs. The second heuristic is a GREEDY RANDOMIZED ADAPTIVE SEARCH PROCEDURE (GRASP), which uses the AFFD heuristic in each iteration of the algorithm. The last heuristic is a GENETIC ALGORITHM (GA) that uses the GRASP in the initialization phase and the AFFD heuristic in each iteration of the algorithm. We also present extensive computational results for 40 instances, each consisting of 36 plans, for a 3-year overall rolling horizon.

The remainder of the paper is organized as follows. In Section 2 we define the STRMP. Section 3 describes the DTRMP and the three heuristics developed to address it. In Section 4 we present computational results. Section 5 concludes. 2. The Stochastic Tactical Railway Maintenance Problem The STRMP involves scheduling of predictive and preventive maintenance activities within a given time horizon (planning horizon) over a long-term rolling horizon. The goal is to bring the degradation process under control by performing appropriate maintenance activities at minimum cost. A set of maintenance activities, called warnings, is assigned to resolve a set of glitches on the track. Each warning can be assigned to a temporal portion of the planning horizon, called a time slot. In general, one glitch can be resolved by more than one warning. The final assignment of warnings to time slots is called a plan. Each warning is characterized by a cost, an amount of resources for resolving the warning, and a risk. Depending on the maintenance carried out, warnings can be at different degradation levels. The degradation process for warnings is modeled using a set of degradation levels and a Markov chain that defines the transition between the degradation levels. Therefore, the cost, the resource requirement, and the warning risk depend on the degradation level for a given warning. Because of the different probabilities for the degradation levels in the time slots, the above values also depend on the time slot in which a warning is allocated. Each time slot is characterized by the resource amount available to resolve assigned warnings