2024 年 12 巻 1 号 p. 135-155
Due to budget constraints in the local cities of Japan, there is an urgent need to prioritize the maintenance or removal of aging transportation infrastructure, particularly bridges. Given the declining tax revenues, our study proposes an approach to systematically address this challenge. Using GIS network analysis, we evaluated the impact of bridge removal on the daily travel time of residents, considering the intricate road network and population distribution of the selected study area, Island A. Various public facilities on the island were identified as potential destinations. Our methodology assessed the changes in travel time for residents if a bridge was removed. Four distinct bridge removal scenarios were proposed: two prioritize minimal disruption in resident travel time, and the other two focus on the life-cycle costs (LCC) of the bridges. This analytical process aims to achieve a balance between fiscal responsibility and ensuring resident convenience. By proposing efficient bridge removal plans based on the consequences of each scenario, we hope to guide local governments in making informed decisions. Through this Japanese case study, our methodology offers valuable insights into formulating strategies for other developed areas facing similar infrastructure challenges.
In recent years, Japan has experienced a nationwide population decline, particularly in rural towns, making it difficult to maintain urban functions and residents’ comfortable lives (Ministry of Internal Affairs and Communications, 2017, 2018). Additionally, many urban infrastructures developed during rapid economic growth are now outdated and require updating. Transportation infrastructures, such as roads and bridges, are typical examples, having long provided smooth travel for people in the city. The usefulness of each road or infrastructure should be assessed to decide whether they should be retained or removed (Mittal and Biswas, 2019; Wagale, Singh, et al., 2021).
Bridges, in particular, play an important role in handling urban traffic and connecting two areas separated by a river or sea. If not properly maintained, they risk collapsing and causing significant losses in transportation or impacts to people’s daily lives. Due to depopulation, tax revenues have also declined, making it difficult to maintain aging bridges. In the near future, the number of obsolete bridges and the required costs will increase dramatically. However, in small communities, it will become more challenging to maintain aging bridges due to lower tax revenue. Therefore, it is important to determine the priority of maintenance or demolition for each bridge.
In a Japanese survey, as much as 32.9% of citizens indicated that social infrastructure should be downsized (Ministry of Land, Infrastructure, Transport and Tousrism, 2020). However, because infrastructure is closely related to people’s lives, residents sometimes disagree with its removal (Enomoto, Takao, et al., 2009; Ninomiya, 1992). Therefore, a new way of thinking is needed that simultaneously considers cost savings and the impact on residents’ daily lives. A reasonable and legitimate plan would achieve consensus among residents.
Other developed countries, such as the United States and the United Kingdom, face similar problems (Mitsubishi UFJ research and consulting, 2018; Newsweek Japan, 2019), and in the near future, many cities in developed countries will also face similar problems due to a lack of investment owing to population decline (Markusen, 2003). Therefore, it is useful to share a Japanese case and use it as a reference for considering appropriate countermeasures.
In this study, we performed a network analysis of GIS to evaluate the impact of bridge removal on the daily travel time of local residents and determined the importance of each bridge. Additionally, based on the concept of Life Cycle Cost (LCC), the amount of future financial cost reduction owing to bridge removal was calculated. Based on the results, we proposed efficient bridge removal plans as a case study in Japan. The results of this study provide valuable insights into infrastructure management planning, especially in developed countries.
Literature reviewOur study aimed to propose a bridge removal plan considering both the influence on the daily travel time of local residents and future financial cost reduction. Such policy decisions are already being made in Japan. Moreover, an approach to improving the quality of decisions is to enhance quantitative methods using spatial analysis (Ministry of Land, Infrastructure, Transport and Tousrism, 2023). Therefore, in this section, we mainly focus on the studies that have narrowed down the list of candidate bridges to be reduced and discuss the characteristics and problems of those studies in turn. In this process, we will clarify where our studies stand among the previous studies by comparing and gradually moving closer to the perspective of our study.
First, there are a considerable number of studies contributing to the appropriate maintenance of bridges. One of the most important approaches is to assess the connectivity of road networks, assuming damage or collapse (Serdar, Koç, et al., 2022). Several studies have topologically assessed the emergency road network connectivity and identified vulnerable points (Kozawa, Nakayama, et al., 2021; Zhang, Wang, et al., 2017). Not only emergency but ordinary road networks, including the connectivity of bridges, are being examined (Akbarzadeh, Memarmontazerin, et al., 2019; Bhatia, Sela, et al., 2020; Karamlou and Bocchini, 2016; Liu and Frangopol, 2005b; Testa, Furtado, et al., 2015; Zhang and Wang, 2016).
Alternatively, though topological modelling is an important method to consider the case of bridge collapse or demolition, it would be better to quantitatively assess the impact on local residents. Travel time is a representative indicator for assessing the impacts. Several studies have examined the reliability of the road network concerning disaster damage by using the travel time of residents (Bocchini and Frangopol, 2011; Chen and Miller-Hooks, 2012; Harada, Kurauchi, et al., 2014; Li, Jin, et al., 2019). However, the main purpose of these studies was to evaluate the reliability of the road network as a whole rather than to identify specific bridges that occupy an important position in the network.
To solve this problem, studies have been conducted to identify the specific bridges that should be strengthened via seismic retrofitting or other means to prepare for disasters (Kilanitis and Sextos, 2019; Nagae, Fujihara, et al., 2007; Sato, Sinozaki, et al., 1995). These studies were able to identify the specific bridges that should be preserved by seismic retrofitting. These studies consider arbitrary OD pairs. However, from the perspective of infrastructure management in daily life, the important bridges should be identified considering the OD pairs that are frequently used by residents. Such a study will enable us to make a more appropriate bridge removal plan that is better suited to the actual conditions in the region.
However, in light of population decline and austerity measures, it is necessary to consider which bridges should be maintained or removed based on cost consciousness. The concept of smart shrinkage has been proposed as one of the effective ways to strategically shrink the regional population and infrastructure. Sabo (1998) confirms that shrinking cities have a significant impact on reducing the administrative costs of urban facilities. Furthermore, the application of the LCC concept to individual bridges contributes significantly to cost reduction (Matsuzaki, 2021; Yanev, 1998). Studies have applied LCC assessment to bridges and prioritized the maintenance of bridges (Frangopol and Liu, 2007; Liu and Frangopol, 2005a; Sato, Yoshida, et al., 2005; Yanweerasak, Pansuk, et al., 2018). These studies may lead to effective policies to reduce the infrastructure management costs. However, they are based on the assumption that existing infrastructure will continue to exist rather than the assumption that the number of bridges should be reduced. They also rarely consider the impact on residents’ daily lives, such as travel time.
With this in mind, Arakawa, Sugimoto, et al. (2019) evaluated the impact of removing bridges whose LCC is higher on administrative services such as firefighting from the perspective of changing the time required to reach each district. Sugiura, Kurauchi, et al. (2015); Sugiura, Machi, et al. (2015) discussed the impact of bridge removal on residents’ activities and proposed a road decommissioning strategy by considering an optimization problem that minimizes the LCC of maintenance while maintaining the residents’ travel time. However, these studies were not applied to a real case.
Based on the discussion above, it can be concluded that there is a lack of studies that address bridge removal strategies accounting for both the impact on residents’ lives (travel time) and the financial cost reduction. Nevertheless, a study that considers both these viewpoints simultaneously would open up new ways of thinking and help create a more appropriate infrastructure management plan. It will also contribute toward smooth and amicable consensus building.
City S on Island A in Prefecture H, Japan, was chosen as a municipality facing depopulation and grappling with limitations in maintaining infrastructure. Situated in the central part of the island, the road network of this city operates largely independently of other regions beyond the island. It is crucial to achieve highly accurate results using the methods outlined below.
We primarily employ ArcGIS Pro ver. 2.5 network analysis to calculate the impact of bridge removal on local residents, considering changes in the required time to reach a destination. Additionally, we assess a bridge alignment reduction plan aimed at alleviating the financial burden on the municipality. Previously, the demand for roads or facilities was described from the perspective of the transportation behaviour data of people (e.g., Zhang, Chai, et al., 2016). Therefore, we focus on the change in travel time (OD cost) on the road network between multiple origins and various destinations (i.e., different public facilities) when a bridge is removed. Figure 1 shows the schematic diagrams. The figure demonstrates that the OD cost increased by 15 minutes when bridge #3 was removed.
Figure 2 depicts the distributions of the origins, destinations, bridges, and road network dataset (ND) for the entire Island A. Each grid cell contains a population value based on the 2015 Japan National Census, with the size of each cell being approximately 500 m². The centre of gravity for each grid cell was extracted as the origin point, resulting in 1130 origin points, excluding those without population. During network analysis, each point is connected to the nearest link in the ND.
Various public facilities on Island A were chosen as target points. The selection criteria for these facilities include 1) their availability as point data from the National Land Digital Information Download Service by the Ministry of Land, Infrastructure, Transport and Tourism, and 2) their importance for residents’ access or the provision of services to residents from the facilities. Appendix 1 contains the list of selected destinations.
Unlike the origins, the destinations are not limited to the city of S, as residents of the city of S can also use facilities outside their community. In addition, each destination is selected based on distance, regardless of the type of facility. In the analysis, individuals identify the nearest 20 facilities to the origin point, except in cases where there are one or more breaks in the road network between the origin and destination.
The ND was created from the road centreline data of the Geospatial Information Authority of Japan and contains 522,331 nodes and 535,584 edges. Each edge also contains vehicle movement speed data. We configured that a vehicle can move uniformly, and the average speed is 34.3 km/h. To simplify the analysis, the time taken for right and left turns, pauses at intersections is not considered. For simplicity, we set only vehicles as the target of the study to consider whether each bridge is to be removed or not. Although Island A is connected to the mainland road network by two land routes, the road environment in this study is intentionally independent of the mainland.
Regarding bridges, there are 656 bridges in the city of S managed by local governments and listed in the Annual Report on Road Maintenance (2014-2018). According to a report of the Ministry of Land, Infrastructure, Transport and Tourism, 58 bridges are judged to be in the "III Early Action Stage," and their location is plotted on the ND. Fifty of these bridges are selected as candidates for removal through a series of analyses, and eight bridges are excluded from the candidate list because they are located on routes to agricultural and forest roads.
Methodology of Analyses Analysis 1: Evaluation of the impact of the removal of each bridge’s removal on the life of the inhabitantsIn this section, the performance of the 50 target bridges was quantitatively evaluated and ranked based on their importance. For this, the travel times of the 50 bridges were measured when the bridges were rendered impassable, and the subsequent change in travel time was calculated to evaluate their respective importance.
The change in travel time (OD costs) on the road network between origins and the multiple destinations was evaluated, assuming that one of the bridges is removed. We set case 0 as the maintained status quo and defined case i as the situation in which bridge #i was removed (i = 1,2, …, 50).
The flow of Analysis 1 is shown in Figure 3. First, in case 0, the sum of travel time from grid cell k to the nearest 20 public facilities was calculated (note: if there are facilities of the same distance and order, they are additionally included for the calculation) and defined as the "OD cost of cell k (ODCk)." Then, the population-weighted average of all ODCk was calculated, and the expected value was defined as E0.
Next, we consider the situation of case i, where bridge #i is no longer working. In this modified road network, we calculated the population-weighted average of all ODCk, and the expected value was defined as Ei. The difference between Ei and E0 (hereafter,
To safeguard a shrinking municipality’s infrastructure management budget, we evaluated appropriate bridge removal scenarios to achieve the goal of cost reduction. We considered not only the cost reduction but also the impact on resident travel time. In this section, assuming that the municipality has to implement cost cuts by removing multiple bridges, we quantitatively evaluated its impact on the residents’ travel time (i.e., calculated in Analysis 1).
Two bridge selection methods were compared: 1) descending order of importance as determined in Analysis 1, and 2) ascending order of LCCs. In addition, considering the estimated budget constraints of the municipality in the years 2030 and 2050, we examined four scenarios (a combination of two time periods and two bridge removal methods). For these scenarios, the change in travel time was calculated, and then, each bridge removal plan was evaluated. The flow of Analysis 2 is shown in Figure 4.
First, we set the cost reduction targets needed to maintain the current level of the financial burden per resident in S city from a long-term perspective (target year 2030) and a very long-term perspective (target year 2050). Next, we planned two types of downsizing policies. One is to remove the bridges that have a lower impact on the travel time of residents, based on the results of analysis 1. In this strategy, the bridges are removed in the order of least impact until the cost reduction goal is achieved. The other strategy is to remove the bridges in the order of the highest LCC. For both removal strategies, the change in the expected weighted average travel time to the nearest 20 public facilities from case 0 (defined as ΔE hereafter) is calculated after deciding which bridges are to be removed. Therefore, we use a total of 4 scenarios; 1) removal of the lowest impact bridge at long-term cost reduction (Scenario 1-1), 2) removal of the lowest impact bridge with very long-term cost reduction (Scenario 1-2), 3) removal of the highest cost bridge with long-term cost reduction (Scenario 2-1), and 4) removal of the highest cost bridge with very long-term cost reduction (Scenario 2-2).
First, we proceed as follows in estimating the future maintenance budget and setting cost reduction targets. We assumed that the FY2020 bridge maintenance and management budget is 150 million yen and calculated the financial burden per resident of S city. Assuming that the financial burden of each resident is maintained at the current level, the maintenance budget for each year is estimated based on the population of each future year.
Figure 5 shows the estimated deficit in the budget for each future year. We are considering offsetting this deficit by reducing costs through bridge removal. In this way, the amount of cost reduction by removing the bridges is estimated at 10,603,000 yen in 2030 and 30,526,000 yen or more in 2050.
Next, the method for estimating the LCC for each bridge is applied as follows. The targets for estimation are bridges that are candidates in Analysis 2. Note that we excluded several bridges from the candidates if the bridges are considered unsuitable for candidacy based on the results of Analysis 1.
In this study, the LCC of a bridge is defined as the sum of replacement cost, repair cost, and inspection cost. (Note: the sum of the repair cost and inspection cost is the maintenance cost). Based on the results of an interview with a member of the Construction Division of S city, we made the assumption that the repair cost would be 100 million yen and the inspection cost would be 50 million yen. Therefore, as for the maintenance cost of each bridge, we first divided the maintenance cost of 150 million yen by the sum of the area of all 656 bridges (Ministry of Land, Infrastructure, Transport and Tourism, 2020) managed by S city. This would yield the unit maintenance cost per area. Then, we calculated the maintenance cost for each bridge by multiplying this value by the area of each bridge. Thus, the value of the estimated annual maintenance cost is 3,410 yen/m2.
For the renewal cost, we divided the average renewal cost for each bridge structure and divided it by the number of years it has been in service. The renewal cost of each bridge was calculated by multiplying this value with the area of each bridge. Thus, the estimated value of the annual renewal cost is 13,910 yen/m2.
Table 1 shows the results of the impact of the removal of each bridge. The number of paths between the origin and destinations remains almost constant at 22,602 because the number of endpoint facilities is the same. In contrast, the change of
For the 25 bridges whose cells were coloured, we discuss the relationship between the results and the characteristics of their geographic location in the next section. They are the top four bridges (orange-coloured cells) and the bottom three bridges (blue-coloured cells) with changes in
| Removed Bridge No. |
Number of passes |
Total travel cost (min/person) |
Ei : Expected value of population-weighted average of travel time (min) |
Amount of change in expected value (sec) |
Rank |
|---|---|---|---|---|---|
| - | 22602 | 2036899.825 | 46.380 | 0.000 | - |
| 1 | 22602 | 2037351.536 | 46.390 | 0.617 | 8 |
| 2 | 22602 | 2037325.203 | 46.389 | 0.581 | 9 |
| 3 | 22602 | 2038381.322 | 46.413 | 2.024 | 5 |
| 4 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 5 | 22602 | 2037114.686 | 46.385 | 0.294 | 13 |
| 6 | 22602 | 2044261.470 | 46.547 | 10.057 | 1 |
| 7 | 22602 | 2037223.466 | 46.387 | 0.442 | 11 |
| 8 | 22602 | 2036918.418 | 46.380 | 0.025 | 27 |
| 9 | 22602 | 2037119.822 | 46.385 | 0.301 | 12 |
| 10 | 22602 | 2036905.266 | 46.380 | 0.007 | 31 |
| 11 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 12 | 22602 | 2036978.237 | 46.381 | 0.107 | 18 |
| 13 | 22602 | 2038575.185 | 46.418 | 2.289 | 3 |
| 14 | 22602 | 2038575.185 | 46.418 | 2.289 | 3 |
| 15 | 22602 | 2037036.554 | 46.383 | 0.187 | 16 |
| 16 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 17 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 18 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 19 | 22602 | 2036925.562 | 46.380 | 0.035 | 26 |
| 20 | 22602 | 2036910.998 | 46.380 | 0.015 | 29 |
| 21 | 22602 | 2036959.265 | 46.381 | 0.081 | 21 |
| 22 | 22602 | 2037324.614 | 46.389 | 0.580 | 10 |
| 23 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 24 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 25 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 26 | 22582 | 2036007.262 | 46.359 | -1.219 | 50 |
| 27 | 22602 | 2036900.847 | 46.380 | 0.001 | 32 |
| 28 | 22602 | 2039659.454 | 46.442 | 3.770 | 2 |
| 29 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 30 | 22602 | 2036927.628 | 46.380 | 0.038 | 25 |
| 31 | 22602 | 2037531.599 | 46.394 | 0.863 | 6 |
| 32 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 33 | 22602 | 2036935.513 | 46.380 | 0.049 | 23 |
| 34 | 22602 | 2036983.885 | 46.382 | 0.115 | 17 |
| 35 | 22602 | 2036907.042 | 46.380 | 0.010 | 30 |
| 36 | 22602 | 2036965.749 | 46.381 | 0.090 | 20 |
| 37 | 22602 | 2036977.396 | 46. 381 | 0.106 | 19 |
| 38 | 22602 | 2036952.686 | 46. 381 | 0.072 | 22 |
| 39 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 40 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 41 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 42 | 22602 | 2036916.580 | 46.380 | 0.023 | 28 |
| 43 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 44 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 45 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 46 | 22602 | 2037357.847 | 46.390 | 0.626 | 7 |
| 47 | 22602 | 2037097.108 | 46.384 | 0.270 | 14 |
| 48 | 22602 | 2037097.108 | 46.384 | 0.270 | 14 |
| 49 | 22602 | 2036899.825 | 46.380 | 0.000 | 33 |
| 50 | 22602 | 2036932.174 | 46.380 | 0.044 | 24 |
The first group comprises the bridges whose ΔEi changes are the four largest among all bridges. Their removal will increase the travel time. It is not easy to interpret the factors based only on the distribution of destinations. Several bridges have more destinations in their vicinity, while others do not. However, focusing on the population distribution and the distribution of the road network helps to understand the reason for the travel time increase.
In case 28, the change in the ΔEi is larger. Figure 6-a shows that the bridge is located in a densely populated area in the city of S. Although bridge 15 and bridge #26 are also located in the same area, the
When comparing the removal impact of bridge #28 (Figure 6-a) and those affected by the removal of bridge #15 (Figure 6-b), Twenty-six cells were affected by the removal of bridge #28. In contrast, the number of cells affected by the removal of bridge #15 is 14, whose distribution is more spatially limited. Although the number of people affected by the removal of each bridge is almost the same (i.e., 1,344 people in the case of bridge #15 and 1,371 people in the case of bridge #28), the change in the increase in their total travel time is drastically different (i.e., 136.72 minutes in the case of bridge #15 and 2759.62 minutes in the case of bridge #28). From this, it can be concluded that because the impact of the removal of bridge #28 covered a wide area, and the time required to reach the destinations was greater than that of the removal of bridge #15, which was only locally affected.
The factors that caused this difference are the access to the main roads and the distribution of the population using each main road. Bridge #28 is located on the road that connects to the northern area. However, bridge #15 is located near the ring road that encircles the island. When Bridge #28 is removed, several people who used the main road are forced to make a detour. Conversely, even if bridge #15 is removed, the ring road will still take people to the facility with only a few detours.
Bridge removal that causes smaller increase in travel timeThe second group comprises bridges with the smallest increment in travel time. It is because the paths that include these bridges were not selected as the shortest path.
Let us consider case 10 as an example of the latter case: bridge #10 is located approximately 5 km west from the most densely populated area in the city of S, as shown in Figure 7.
From these figures, we can see that only one grid cell was affected by the removal of bridge #10, and only three people lived in the cell. The change in their total travel time is only 4.02 min. Therefore, we can conclude that the reason for this result is that most people in this area travel to the most populated area and do not use bridge #10.
Bridges where the ΔEi does not change can be further divided into two types: bridges not chosen as the shortest route and those on roads without grid cells or destinations on one side.
As an example of the first case, bridge #16 is located in the suburbs, approximately 2 km north of the city centre (Figure 8-a), and people in the vicinity of the bridge use the shortest route (red solid line) to reach their destination in the city centre. Since bridge #16 is on the way to the longer path (blue dashed line, shown in Figure 8-a) and not on the shortest path, there is no change in the ΔEi whether the bridge is removed or not.
In the latter case, the path from the city centre to bridge #24 is a complete dead end; there is neither a destination nor a starting point. Hence, no one uses this bridge. Therefore, there is no change in the ΔEi if the bridge is removed. Such a bridge is not considered in Analysis 2 because it is not suitable as a withdrawal plan in reality.
Bridge with negative change in ΔEiIn this group, there is only bridge #26. The removal of this bridge decreased the number of paths by 20 because an origin could not reach its destination due to the removal of the bridge. As shown in Figure 9, the inhabitants of the grid cell are fixed to the path marked with a red arrow. These individuals will no longer reach a destination when the bridge is removed. Although this phenomenon occurred because the centre of gravity of the network was used as a starting point, it was not considered here not only as a technical problem in the analysis but also as a case that happens. Such cases can also occur when an isolated village is separated from other areas by removing a bridge. When dismantling such bridges, it is necessary to discuss more various aspects other than travel time change and cost reduction. Therefore, bridge #26 was excluded as a candidate for removal in Analysis 2.
As aforementioned, when the analysis results are combined with numerical information and geographic factors, the importance of each bridge is appropriately assessed, and the evaluation of each bridge in Analysis 1 is justified as an analytical method.
In this analysis, the candidate bridges are 42, except for the eight bridges based on the discussion in Analysis 1 (i.e., one bridge whose ΔEi is minus and seven bridges ΔEi is 0).
If we include the bridges whose ΔEi is 0 and select the bridges to be removed in order to minimize the increase in travel time, these bridges are preferentially selected to reduce the maintenance budget. However, these bridges are often used for purposes other than vehicle travel for facility use. Therefore, the decision to remove such bridges should be made taking into account many other aspects. Removing these bridges because there are no residents or facilities ahead of them in the analysis is not consistent with the intent of this study, which emphasizes the impacts on the daily lives of citizens in a real urban area. Therefore, we excluded the candidate bridges whose ΔEi is 0.
In addition, the bridge with minus ΔEi should be excluded in this analysis. It is because such a case can occur when an isolated village is separated from other areas by dismantling a bridge. When removing such bridges that separate the origin or destination from various regions, examining factors other than travel time change and cost reduction is vital.
Bridge removal plans based on the scenariosIn this section, we proposed concrete bridge removal plans based on the scenario analysis. As mentioned in Section 2.2.2, we used four scenarios: 1) removal of the lowest impact bridge at long-term cost reduction (Scenario 1-1), 2) removal of the lowest impact bridge with very long-term cost reduction (Scenario 1-2), 3) removal of the highest cost bridge with long-term cost reduction (Scenario 2-1), and 4) removal of the highest cost bridge with very long-term cost reduction (Scenario 2-2). Table 2 summarizes the results of the analysis.
| Removal Policy |
Target year |
Target reduction cost (million yen) |
Achieved reduction cost (million yen) |
Number of removed bridges |
(sec) |
Number of meshes affected | ||
|---|---|---|---|---|---|---|---|---|
| Case 0 | - | - | - | - | 0 | 0.00 | - | |
| Scenario1-1 | Low ranking | 2030 | 10.60 | 13.99 | 23 | 0.54 | 153 | |
| Scenario1-2 | Low ranking | 2050 | 30.53 | 31.29 | 36 | 4.76 | 410 | |
| Scenario2-1 | Highest LCC | 2030 | 10.60 | 10.66 | 2 | 0.38 | 55 | |
| Scenario2-2 | Highest LCC | 2050 | 30.53 | 30.74 | 20 | 18.04 | 393 |
The table shows that 23 bridges in Scenario 1-1, 36 bridges in Scenario 1-2, 2 bridges in Scenario 2-1, and 20 bridges in Scenario 2-2 need to be removed to achieve the reduction target. Moreover, the result shows that the number of bridges to be removed is less when the bridges with the highest LCC are selected in order, as compared to when the bridges with the lowest LCC are selected. Figure 10 shows the locations of bridges removed in each scenario and the grid cells that are affected by the removal of the bridges. The number of detected paths did not change at 22,602, indicating that network fragmentation was prevented in the removal plans of Analysis 2 by proactively excluding the bridges whose shortest paths decreased.
Since multiple bridges are removed in each scenario, the changes in the ΔE are larger than in Analysis 1, where each bridge was removed individually. It can be seen that not only the cells around the removed bridge but also several distant cells are affected by the removal of the bridge. This tendency is particularly pronounced in Scenario 2-1. On the contrary, as for Scenario 1-2, the affected cells are essentially distributed around the removed bridges. Although many bridges were removed in this scenario, 720/1,130 cells and 31,051/43,918 individuals were not affected. These differences indicate that the selection of bridges to be removed based on the impact on residents’ travel time is more effective in preserving residents’ daily lives. This also means that the change in travel time of individuals is larger for an LCC-based approach than for a convenience-based approach. Regarding the impact on passenger travel, the number of meshes in Table 4-2 shows that the number of affected meshes is smaller in the LCC-based removal approach than in the convenience-based approach, but the changes in the ΔE are almost the same between these approaches. In implementing the removal plan, it is assumed that consensus can be efficiently formed by actively explaining the plan to the residents of the affected meshes.
In addition, when the target year is set at 2050, the number of bridges removed increases significantly in both scenarios. With regard to the removal plans (i.e., Scenarios 1 and 2), the ΔE for the lower impact bridges is 0.16 seconds lower for the target year 2030 and 13 seconds lower in 2050, compared to the highest cost removal plans.
Considering the advantages and disadvantages of the bridge removal plans as a whole, it can be seen that removing the two bridges with the highest LCC in 2030 (Scenario 2-1) has less impact on residents than selecting the 23 bridges with the lower LCC (Scenario 1-1). This suggests that choosing a bridge based on the traffic impact on residents does not necessarily have a lower impact on residents' lives. However, considering that the difference is only 0.16 seconds and that the cost reduction amount in Scenario 1-1 is much larger than in Scenario 2-1. Moreover, it is interesting to note that in 2050, the removal of 36 bridges from the lowest transportation impact order affects 13 seconds less in ΔE than the removal of 20 bridges from the highest LCC cost order. Furthermore, the cost reduction in Scenario 1-2 is almost 7 million yen more than the reduction target, while the cost reduction in Scenario 2-1 is approximately 2 million yen. In this case, it is more cost-effective to select the bridges to be removed which have little impact on their convenience.
In conclusion, even if the number of the bridge removal is minimal, the removal of bridges with the highest LCC can significantly influence the lives of residents. However, selecting bridges to minimize the impact on residents' lives may also result in cost savings. Occasionally, the cost savings can be higher than the LCC-oriented removal strategy. These findings indicate the feasibility of the bridge removal strategy proposed in this study, which considers the bridge’s influence on residents’ lives.
In this study, we calculated the contribution of convenience as one of the indices of bridge performance in community consensus building, after quantifying the impact of bridge removal on the road network through GIS network analysis. We also proposed a method of securing financial resources by reducing bridge layout and discussed the method. The results are summarized below.
(1) The impact of bridge removal on residents’ lives was objectively quantified by the contribution of bridges to convenience, taking into account the interrelationships between regions, population distribution, and the distribution of public facilities, which are difficult to understand using linear distance relationships alone. In addition, the validity of the analysis method was checked using numerical information and geographical factors.
(2) We proposed a rational method for selecting bridges to be removed to secure financial resources. As a result, the cost reduction target of the municipality was set at approximately 30 million yen, and the reduction of bridge locations based on the contribution to convenience could reduce the increase in travel time to only approximately 4.76 seconds per resident.
Bridges are important infrastructure for daily life, and the removal of bridges used in daily life directly leads to the decline in the living standards of citizens. However, conventional cases selected the bridge mainly from the maintenance budget constraints. That is a very top-down approach. As a result, it is often difficult to reach a consensus on the removal of such important infrastructure, and the government has difficulty making decisions. To address this problem, we have proposed a new method to help create a bridge removal plan. Our method takes into account not only budgetary constraints but also the impact on the time it takes for citizens to access important public facilities in the region. By considering the impact on citizens’ daily lives, this led to a more appropriate assessment of the value of the bridge from the residents' perspective. These bottom-up assessments would facilitate consensus-building and administrative decision-making.
In addition, we adapted the network theory analysis to the actual urban space and evaluated its impact. The results show the usefulness of our approach in creating an infrastructure management plan against the depopulation problems. The selection of the bridges to be demolished based on both the financial situation and the impact on residents’ lives can be a groundbreaking case study for the depopulation problem that will become more severe in the future.
LimitationsThe limitation of this study is that the real situation is more complicated because the cost of demolition is deducted from the reduction amount. It is necessary to carefully consider the selection of bridges to be removed if the number of bridges is actually reduced.
Owing to the original travel time and the diversity of options, residents in various locations will evaluate and feel the same amount of time reduction differently. This type of research is still insufficient. Therefore, we did not weigh the decreased amount by location in this study. Further research into this attribute is desirable to improve the fairness of the proposed method, which we expect will be applied to the method proposed in this study.
A network dataset should be developed that better reflects the actual road environment. Road networks consist of a complex interaction of various factors such as traffic volumes, intersections, temporary stops, and even freeways. In order to obtain reasonable and defensible results, the information to be reproduced must be selected based on the scale of the community, such as the area and the population.
Moreover, the fact that the destinations are limited public facilities does not seem to correspond to reality. In real life, residents do not use the road network only to access these facilities. Because willingness to use the facilities is skewed between destinations, we need to account for the weighting of each destination in the analysis results. We need to consider the impact of the bridge demolition on residents’ lives from different perspectives. In this study, we demonstrated our proposed method using specific instances. Consequently, the evaluation was constrained to the perspective of its necessity in residents’ daily lives. Therefore, travel demand was restricted to the resident population, and destinations were confined to facilities essential for daily living. Our proposed method can accommodate variances in demand based on different attributes, such as travellers. A more detailed analysis of these distinctions is planned for future research. An essential objective is determining generality by contrasting results from diverse transportation demand estimates.
It is also necessary to include the cost of demolishing the bridge in the cost reduction for maintaining the budget in Verification 2.
The results of this study are derived from only a small number of factors. Our future research plans are to construct a rational and defensible system that can more accurately match the results of the network analysis to the "impact on residents' lives" and to create an evaluation index that can more objectively analyse the data obtained.
Considering such limitations, the study’s results would enable us to take the better measure of a cost reduction plan for infrastructure that has little change on residents’ daily lives.
Conceptualization, S.Y., N.A.; methodology, S.Y., N.A., T.S.; investigation, N.A.; resources, S.Y.; data curation, N.A.; writing-original draft preparation, N.A.; writing—review and editing, N.A., T.S.; supervision, S.Y., T.S.
All authors have read and agreed to the published version of the manuscript.
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.
This study was supported by Institute of Disaster Mitigation for Urban Cultural Heritage, Ritsumeikan University. We wish to thank the Construction Division of S city for providing useful information and materials.
This research was funded by the Institute of Disaster Mitigation for Urban Cultural Heritage, Ritsumeikan University.
| Facility type | Major examples of facilities | Number |
|---|---|---|
| Medical institution | Hospitals, (General)clinics, Dental clinics (based on the Medical Services Act) | 233 |
| School | Elementary schools, junior high schools, secondary schools, colleges, junior colleges, universities, and special support schools (based on School education law). | 73 |
| Police station | Police facilities | 50 |
| Public meeting facility | City hall, branch offices, liaison offices / Community halls, Established meeting places, managed and operated by the city | 179 |
| Facilities for attracting customers | Facilities with spaces for holding events such as attractions and exhibitions, and facilities with “bleachers” for watching entertainment sports, etc. | 52 |
| Fire station | Firefighting facilities | 9 |
| City park | Urban parks as defined by the Urban park law | 14 |
|
Roadside station (Michi no Eki) |
Rest facilities with three functions: “Resting function” for road users, “Information transmission function” for road users and local residents, and “Regional cooperation function” for towns and cities to join hands and work together to create vibrant region | 3 |
| Welfare facility | Facilities for the care elderly and disabled people as well as children | 155 |
| Logistics base | Container terminals, air cargo terminals, rail cargo stations, bonded areas, truck terminals, wholesales markets | 15 |
| Cultural facility | Facilities that collect, preserve, and exhibit culturally valuable artworks and living things, and also conduct education, dissemination, and research related to these cultures | 79 |
| Post office | Offices of Japan Post Co. | 55 |
