Framework for Evaluating Traffic Impact of a New Large Commercial Land use—A Case Study from Bengaluru, India

Abstract

Indian cities are witnessing rampant growth in terms of population and economic growth. The urban population percentage increased from 28.6% in 2001 to 37.7% in 2011. Being a leading IT exporter, Bengaluru sees an increasing number of commercial buildings burdening the existing road infrastructure, leading to heavy congestion in particular areas of Bengaluru. This calls for an understanding of any upcoming commercial development before it is thoroughly established to take adequate measures to reduce the negative impacts. However, the Traffic Impact Assessment (TIA) using the traditional four-stage modeling cannot be used in the current context due to the impact of COVID-19 on the existing traffic. This paper adopts an improved methodological framework that can be used to evaluate the impact of traffic in any situation. An average of 6.6% percentage increase (maximum 25% on some roads) in V/Cs is observed before and after the new development on the adjoining roads. To improve the level of service, several scenarios are considered and analyzed to provide the best mitigation strategies. This improved methodological framework can help to evaluate a new development’s impact for any given situation, and what mitigation strategies and types of transportation improvements may be essential to maintain a smooth moving traffic with a satisfactory level of service.

1 Introduction

Indian cities are witnessing rampant growth in terms of population and economic growth. The urban population percentage increased from 28.6% in 2001 to 37.7% in 2011. There is a substantial increase in India's GDP for most of the last two decades, leading to rising per capita incomes and a reduction in absolute poverty. This growth is physically manifested in increasing employment offices and the associated housing and retail development in the Indian cities. Land use is a fundamental determining factor for movement and activity. The nature of land use influences the source and nature of travel, and consequently, the traffic generation and distribution processes. Among the various land uses, commercial land use contributes more to the traffic flow in cities due to its trip-making nature (built-up area and occupancy are more for commercial [4]). Bengaluru, also known as Silicon Valley of India, is a leading IT exporter, where the number of commercial buildings is more. This rapid growth of employment offices in the city acts as a hub for trip generation and attraction, potentially impacting the existing road infrastructure. This is seen to reflect on the urbanization trends and the mobility choices of people which in turn have resulted in an exponential rise in motorization. This leads to heavy congestion in particular areas of Bengaluru. According to the ninth edition of the Tom Tom Traffic Index report, Bengaluru has the world's worst traffic congestion [13]. Therefore, it is imperative to understand the effect of any upcoming commercial development before it is thoroughly established, to take adequate measures to reduce the negative impacts, if any.

Traffic Impact Assessment (TIA) is a technical analysis of traffic problems and safety issues relating to a specific development. This is done to estimate the impact of a new development's site-generated traffic on the adjoining roads. Upcoming or new projects require a comprehensive TIA to help determine the forecasted traffic on the adjoining roads, identify the potential problems associated with the new development, provide suitable mitigation strategies to tone down the negative impacts, and provide suggestions/ recommendations for hassle-free movement of vehicles on roads. TIA, though recently being studied in India due to the need for developing sustainable solutions to ease out congestion, a few researchers across the world have studied this widely. Most of the studies have used the conventional four-stage modeling [3, 8, 9]. Wagner [17] has studied the impact of logistics-related land use and suggested a methodological improvement of integrating the trip generation and trip distribution estimation into the regional land use and transport models. In a similar study by Sharmeen et al. [11], a generic methodology for TIA is developed by forecasting the background and the developing traffic for a mixed land use in Dhaka city. Mihans et al. [6] did a TIA in the neighborhood of Skudai Town in the Johor Bahru region in which the entry mean trip rate was estimated using trip rate analysis, cross-classification analysis and regression analysis and did a comparison of values obtained from Highway Planning Unit (HPU) Malaysian manual.

Regidor and Teodoro [10] mentioned that detailed description of the traffic conditions around the study site’s vicinity is required. At least a few kilometers around the site should be studied for its impact. Also, the future projected year should be at least 10 years down the line [1] with base year ideally varying by location, extent and purpose of development. To ensure that predictions of traffic impacts are accurate, the TIA should include technical guidance about baseline conditions and key assumptions, with a full assessment of existing transport infrastructure [14]. Environment impacts should also be taken into consideration, as more sustainable solutions are required [1, 15]. The methodology used for predicting trips would have a significant effect on a TIA [7]. Therefore, a more comprehensive methodology which is inclusive of all situations is required.

However, the existing methodologies in the literature cannot be used for the present COVID-19 situation as the traffic is not the same as in the pre-COVID situation, which can overestimate the model values. This necessitates an improved methodological framework that can evaluate the impact of traffic in any situation. A new large commercial land use in Bengaluru is considered the case study to apply the proposed methodology and framework. Therefore, the objectives of this paper are (a) to develop an improved methodological framework to evaluate the traffic impact (b) to perform a traffic impact assessment of a new large commercial land use which is under development in Bengaluru (c) to assess the level of service (LOS) on roads adjoining the new commercial development based on the estimated traffic loads and suggest road improvement measures to maintain a given LOS, wherever possible, and (d) to propose possible ways to reduce traffic loads in the present and near future and ways to integrate with upcoming Bengaluru metro rail station near the new commercial development.

The remainder of the paper is organized as follows: In the next section, details of the case study and data collection are presented. Section 4 describes the methodological framework adopted, and Sect. 5 discusses the results. Finally, we conclude with the discussions of the study and possible future work.

2 Description of the Case Study

2.1 Study Site

ITC Green Centre is an ongoing commercial complex with predominantly office space (buildings 1–6) and a hotel (building 7) in which building 1 and 2 are already existing. It is located on a 13.7-hectare (33.98 acres) site and is approximately 5.5 km northeast of Bengaluru's central business district. The major land use near the campus is residential. The proposed development is planned to be built in 4 phases and will be replacing the existing old buildings on the site, phase-wise. Phase 1 of the campus includes building 1 and 2, which is existing with the ground and 11 floors, and four basement floors for parking in common. Phase 2 includes building 3 with the ground and 12 floors and three basement floors for parking. Phase 3 includes building 4 and 5 with ground and 12 floors each, 2 basements in each building for parking, and a Multi-Level Car Parking (MLCP) in building 4 to accommodate the overall campus's deficit car parks. Phase 4 includes building 6 with the ground and 11 floors, four basements for parking, and the ITC hotel. Two scenarios are considered for hotel building: (a) Scenario 1–150 rooms with 600 occupancies (includes staff and guest) and (b) Scenario 2–350 rooms with 1400 occupancies (includes staff and guest) The project site in Bengaluru and the layout is shown in Figs. 1 and 2 respectively, and the details of all phases are given in Table 1.

Fig. 1

Project site and the adjoining roads

Fig. 2

ITC green centre layout

Table 1 Details of ITC green centre buildings

2.2 Data Collection and Assumptions Made

Data collection included both primary and secondary data. The secondary data collected from ITC Green Centre were the entry/exit time of vehicles to ITC, pin codes of cars entering ITC for January 2020 (before lockdown) and the vehicle circulation plan inside the ITC campus. Pin codes represent the area of residence of the car users. The pin code details of the car users were collected to identify the shortest route from their location to ITC Green Centre. For validation, a peak hour volume count survey was done on the three roads adjoining the ITC campus.

Due to the non-availability of data, the following assumptions were made.

• i. Mode split for hotel employees was assumed to be the same as office buildings.

• ii. From the data gathered for hotel guests and visitors, the maximum entries and exits in one-hour interval were assumed as the entries and exits in peak hour as a worst-case scenario.

• iii. Directional split percentages of other modes (except two-wheelers) were assumed to be the same as car distribution.

• iv. Directional split percentage of two-wheelers from Gate 4 was assumed based on its connectivity to the main road.

• v. The occupancy value of a 13-seater tempo traveler was assumed to be 9.75 (75% occupancy).

3 Methodology

The conventional four-stage modeling involves collecting the volume count data from the adjoining roads and adding the additional volume induced due to the new development. However, in the current pandemic situation, collecting the existing volume count data would give wrong figures due to increasing percentage of people opting for work from home, change in travel patterns, etc. Therefore, a generic methodology using the forecasted values from the Regional Travel Demand Model (TDM) developed for Bengaluru using TransCAD [16] was adopted to assess the traffic impact of the new development. The flow chart depicting the steps involved in developing Travel Demand Model is shown in Fig. 3. The volume on the adjoining roads of the development center was linearly forecasted for the year 2020 for all 384 Traffic Analysis Zones considered in the TDM. The forecasting was done using the growth rates from the Comprehensive Traffic and Transportation Study for Bengaluru Metropolitan Region, which was determined from the primary data (sixteen different types of surveys pertaining to the network, users and operator) and secondary data (Revised Structure Plan-2031, Draft Master Plan-2015, Plan Bengaluru-2020, Census Data, Public transport, fare structure, etc.) (CTTS Report 2010). Using the volume obtained from the TDM model, the impact of the new development on the adjoining roads was assessed. The methodology adopted for the case study is depicted in Fig. 4.

Fig. 3

Process flow chart of the travel demand model [16]

Fig. 4

Methodology adopted for the case study

4 Results

4.1 Trip Rate Analysis

From the built-up area and occupancy details of ITC Green Centre and old campus buildings, average trip generation rates for the peak period of a typical weekday and the number of trips produced and attracted to the development were calculated using the formula

$$Trip rate=\frac{Occupancy}{Built up area}$$

This is presented in Table 2.

Table 2 Average trip rates

The following representations are to be noted:

• a. Scenario 1 a—Employees in scenario 1 of hotel

• b. Scenario 1 b—Guests in scenario 1 of hotel

• c. Scenario 2 a—Employees in scenario 2 of hotel

• d. Scenario 2 b—Guests in scenario 2 of hotel

4.2 Peak Hour Calculation

The morning peak was identified to be between 9 and 10 AM, and the evening peak was identified to be between 6 and 7 PM for both ITC Green Centre and the old campus. It is seen from Figs. 5 and 6 that 20 and 30% of the total trips in a day were attracted to ITC Green Centre and old campus, respectively, in the morning peak hour and 20% and 24% of the total trips in a day were produced from ITC Green Centre and old campus, respectively, in the evening peak hour. Similarly, 8% of the total trips in a day were produced from ITC Green Centre in the morning peak hour, and 11% of the total trips in a day were attracted to ITC Green Centre in the evening peak hour. Productions from old campus during morning peak and attraction during evening peak were almost 0%. From this and the total trips in a day, the trips produced and attracted to ITC Green Centre and old campus during the morning and evening peak period were calculated, which is summarized in Table 3. The morning trip production and evening trip attraction of ITC Green Centre are more when compared to the old campus trips. This could be due to a greater number of shift-based trips from ITC Green Centre than from the old campus buildings, which needs a further understanding.

Fig. 5

Peak hour calculation for ITC green centre

Fig. 6

Peak hour calculation for old campus

Table 3 Trips produced and attracted to ITC Green Centre and old campus during peak hour

4.3 Mode Split

The mode-wise trips were then calculated using the average occupancy values [5]. The mode split percentages considered are reported in Tables 4 and 5. Mode split for hotel employees was assumed to be the same as office buildings. Only two-wheeler and car were considered, and all other modes were distributed to two-wheeler and car according to their percentages.

Table 4 Mode split percentages for office buildings

Table 5 Mode split percentages for employees and guests of hotel

Using the mode split percentages, the peak hour trips and the average occupancy values, the vehicle trips by mode were calculated for AM and PM peaks. Using the Passenger Car Units (PCU) from the Indo-HCM manual, 2017, the vehicle trips were converted to their equivalent passenger car values.

4.4 Trip Assignment

The trips were assigned using User Equilibrium (UE) trip assignment method. The UE method is based on Wardrop’s first principle which assumes that “no user can unilaterally reduce their travel costs by shifting to another route”. For a given OD pair, the UE conditions can be represented as

$${f}_{k}\left({c}_{k}-u\right)=0: \forall k$$

(i)

$${c}_{k}-u\ge 0: \forall k$$

(ii)

where ${f}_{k}$ is the flow on path $k$, ${c}_{k}$ is the travel cost on path $k$, and u is the minimum cost. Equation (ii) can have two states:

• 1. If ${c}_{k}-u=0$, from Eq. (i), ${f}_{k}\ge 0$. This means that all used paths will have the same travel time.

• 2. If ${c}_{k}-u>0$, the from Eq. (i), ${f}_{k}=0$.

This means that all unused paths will have travel time greater than the minimum cost path [12].

The initial OD matrix used for assigning the trips included only the existing Phase 1 buildings of ITC and does not include the trips from the other phases. To add the additional trips from the new phase, about 23 links around ITC were considered which is shown in Fig. 7. The pin codes of cars were gathered, and their route to ITC from the origin was identified from google maps. It was observed that 60% of the cars entered from Wheeler's flyover side and the remaining 40% from Dodda Banaswadi roadside (opposite to Wheeler’s flyover side), which was assumed to be the same for all the vehicles entering through the main gate (Gate 1). Similarly, the directional split percentage for two-wheelers entering the campus from Jeevanahalli main road (Gate 4) was also assumed to be 80% from Banaswadi to Jeevanahalli main road and 20% from Ramakrishnappa and Charles Campbell Road, due to unavailability of data.

Fig. 7

Adjoining roads considered for assessing the impact of traffic

From the pin codes, it was observed that 60% of the trips came from Wheeler's flyover. Out of the 60%, 78% came straight from Wheeler's flyover and 22% came from pottery to Wheeler's flyover side. Out of the 78%, 67% again came straight from Wheeler's flyover side, and 33% came from Assaye road. Out of the 22%, 95% of the trips were from Lazar road, and 5% were from MM road. Similarly, out of the 40% trips coming from Dodda Banaswdi main road, 95% of the trips came from Indian oil flyover and 5% from Bayappanahalli main road.

For two-wheeler trips, it was assumed that 80% of the trips came from Banaswadi main road to Jeevanahalli main road and 20% from the railway line side. Out of the 80% trips from Banaswadi main road, 60% was from Wheeler's flyover side and 40% from Banaswadi roadside. The 20% from railway line side came from Ramkrishnappa road (one-way road) which diverged from Wheeler's road. The trips were assigned according to these directional split percentages except for one-way roads, where the respective trips were assigned to their alternate roads.

The Regional Travel Demand Model (TDM) was then used to forecast the existing travel demand for 2020 for peak hour, and the existing flow values were derived from the model for the existing condition. Then (Volume/ Capacity) V/C ratio for the adjoining roads was calculated for the existing condition (phase 1 of ITC Green Centre and existing buildings of old campus) to assess the current Level of Service (LOS) on the roads adjoining the campus. The capacity values were obtained from the Indian Highway Capacity Manual (Central Road Research Institute 2017). Similarly, the traffic volume on the adjoining roads was forecasted for the year 2030 using Regional Travel Demand Model (TDM), and the flow values for peak hour were derived from the model.

The peak period classified traffic volume estimates based on the TDM for Bengaluru for the adjoining roads and their corresponding V/C ratios and LOS, for an existing condition, were calculated. Then the incremental impact of traffic due to the development of each phase was also assessed. For this, both the ITC Green Centre and old campus buildings were considered. When a new building was included, the corresponding old buildings that will be replaced were excluded from the assessment. For the existing 2020, in most of the links, volume exceeds the capacity with V/C more than 1 representing LOS F. According to Indo-HCM, this symbolizes a zone of forced or breakdown flow which may result in long queues and delays. For the projected year 2030, the volume to capacity ratio was worse than the current 2020 year. This can lead to queues that are characterized by stop and go waves, which are extremely unstable. Long queues and delays are highly likely. Therefore, proper measures are to be taken to improve the level of service. An average of 6.6% (maximum 25% on some roads) increase in V/Cs of the adjoining roads was observed due to the development of ITC (before phase 1 and after phase 4).

4.5 Validation

A peak hour traffic volume count survey was done to validate the volumes derived from the Travel Demand Model. The survey was done on a typical working day (Wednesday, 30 December 2020) and under normal weather conditions. Six adjoining roads to ITC were chosen for the survey. The vehicles are counter mode-wise and converted to one unit using PCU values. The total trips are then compared with the volumes derived from TDM and it is seen from Table 6 that the predicted volumes from TDM are higher than the actual volumes on the day of count. Therefore, the calculations are done using TDM volumes which represents a worst-case scenario. However, the survey was done during COVID-19 situation, which is a limitation.

Table 6 Volumes from traffic volume count survey and TDM

4.6 Scenario Analysis

The following scenarios were analyzed to check the effect of those on the estimated traffic volumes, V/C ratios and LOS.

• a. Work from Home (30 and 50% of the employees).

• b. Staggered time

Morning peak hour trips (8-9 AM, 9-10 AM, 10-11 AM) and evening peak hour trips (5-6 PM, 6-7 PM, 7-8 PM).

• c. Mode shift

• i. Change in mode shares of ITC

(10% car, 10% TW, 37% office cabs, 29% office bus, 14% office tempo travelers).

• ii. Mode share used in bundle 3 of CLIMATRANS [16].

Based on the Intergovernmental Panel on Climate Change (IPCC) definition, inputs from multiple stakeholder meetings and a Delphi survey with various government officials of Bengaluru, 4 policy bundles were formulated in CLIMATRANS. Those mitigation policy bundles' main objective was to attain an optimum balance of push and pull strategy by developing policies that encourage public transportation and other sustainable modes. This can help reduce the vehicle kilometers traveled, which lead to a reduction in emissions and traffic congestion compared to the Business-as-usual scenario, thereby improving the quality of life of people in Bengaluru city.

Bundle 3 of CLIMATRANS is a mixture of Planning, Regulatory, and Economic instrument and includes the following policies:

• Increasing network coverage of Public Transit.

• Defining car restricted roads.

• Congestion Pricing.

• Park and Ride.

• Cycling and Walking infrastructure.

• Encouraging car-pooling and High Occupancy Lanes.

• High-density mix building use along main transport corridors.

Bundle 3 was identified as the best bundle in CLIMATRANS and so, the mode share percentages of bundle 3 were adopted for this scenario. The mode share percentages were 16% TW, 2% Car, 3% Auto, 40% Bus, 19% Metro, 6% Walk, and 16% Cycle.

• d. Combination of the above scenarios

The following combinations were analyzed.

• i. 30% of the employees WFH and staggered time.

• ii. 50% of the employees WFH and staggered time.

• iii. 30% of the employees WFH and change in the mode share of ITC.

• iv. 50% of the employees WFH and change in the mode share of ITC.

• v. Staggered time and change in the mode share of ITC.

• vi. 30% of the employees WFH, staggered time and change in the mode share of ITC with the existing volumes on adjoining roads.

• vii. 50% of the employees WFH, staggered time and change in the mode share of ITC with the existing volumes on adjoining roads.

• viii. 30% of the employees WFH, staggered time and change in the mode share of ITC with estimated volumes on adjoining roads, using the mode shares of bundle 3 of CLIMATRANS.

• ix. 50% of the employees WFH, staggered time and change in the mode share of ITC with the estimated volumes on adjoining roads, using the mode shares of bundle 3 of CLIMATRANS.

The combination of three scenarios provided better results out of which while comparing the BAU with the existing volume on the adjoining roads, the scenarios (30% WFH, staggering work time from 10 AM to 7 PM and mode shift in ITC trips) and (50% WFH, staggering work time from 10 AM to 7 PM and mode shift in ITC trips) were best for AM trips and the scenarios (30% WFH, staggering work time from 8 AM to 5 PM and mode shift in ITC trips) and (50% WFH, staggering work time from 8 AM to 5 PM and mode shift in ITC trips) were best for PM trips, for both 2020 and 2030. This is because the morning peak hour is 9 AM–10 AM, and so trips will be proportionately less from 10 AM than in 8–9 AM (as this also has a good number of trips). Similarly, the evening peak hour is 6 PM–7 PM and so the trips will be proportionately less around 5 PM than in 6–7 PM (as this also has a good number of trips). It was also seen that there was not much reduction in V/Cs during the peak hour time as 9 AM–6 PM is the general work time. Also, the percentage decrease in V/Cs due to 30% WFH and 50% WFH were almost similar. Therefore, a combination of staggering the work hours before and after the peak hours, with a proportion of employees working from home, and a proportion of employees shifting to shared modes provided by ITC can substantially reduce the V/Cs on the adjoining roads. Decrease in PM was more than AM could be due to external factors (such as school timings). Percentage decrease of V/C from BAU scenario's V/C and projected 2030s V/Cs due to the combination of three scenarios with the existing volumes on adjoining roads are summarized in Table 7.

Table 7 Percentage decrease of V/C from BAU scenario’s V/C and projected 2030s V/Cs due to the combination of three scenarios with the existing volumes on adjoining roads

Similarly, while comparing the BAU with the volumes obtained from bundle 3 scenario of CLIMATRANS, the scenarios (30% WFH, staggering 10 AM–7 PM and bundle 3 mode shift) and (50% WFH, staggering 10 AM–7 PM and bundle 3 mode shift) were best for AM trips, and the scenarios (30% WFH, staggering 8 AM–5 PM and ITC mode shift) and (50% WFH, staggering 8 AM–5 PM ITC mode shift) were best for PM trips, for both 2020 and 2030. It was also seen that there was not much reduction in V/Cs during the peak hour time. Also, the percentage decrease in V/Cs due to 30% WFH and 50% WFH were almost similar. As bundle 3 mode share values from CLIMATRANS were used for adjoining roads’ volumes, the percentage decrease in V/Cs were more (Table 8), as it was more toward using public transportation and sustainable modes. Therefore, a combination of staggering the work hours before and after the peak hours, with a proportion of employees working from home, and shift of ITC employees and the public to more sustainable modes can substantially reduce the V/Cs on the adjoining roads. Decrease in PM was more than AM could be due to external factors (such as school timings). Percentage decrease of V/C from BAU scenario's V/C and projected 2030s V/Cs due to combination of three scenarios with bundle 3 volumes on adjoining roads is summarized in Table 8.

Table 8 Percentage decrease of V/C from BAU scenario’s V/C and projected 2030’s V/Cs due to the combination of three scenarios with the existing volumes on adjoining roads

As the bundle 3 mode shift cannot be attained immediately, the best scenario would be the 50% work from home, staggering of work hours from 8–11 AM to 5–8 PM, and ITC mode shift.

4.7 Effect of Metro on the Best Scenario Results

Bengaluru metro has proposed Pottery town metro station, which is approximately 2.5Kms from ITC Green Centre. Therefore, 6% from two-wheelers trips and 4% from cars trips were shifted to the metro. This analysis was done for the best scenario (50% WFH, staggering 8–11 AM to 5–8 PM, and ITC mode shift), and the results are summarized in Table 9. It was seen that there was a 1%–2% reduction on an average in V/Cs from the best scenario V/Cs due to the shift of trips to the metro. If an exclusive skywalk is provided between ITC and the metro station, then there could be more shift of trips toward the metro.

Table 9 Percentage decrease of V/C from best scenario to V/Cs due to shift of trips to metro

5 Conclusion

This paper discusses the methodological framework adopted to assess the impact of an upcoming commercial development (ITC Green Centre) in Bengaluru. The Regional Travel Demand Model (TDM) that was developed for Bengaluru using TransCAD [16] was adopted to assess the traffic impact of the new development. The forecasted values from this model were considered for the assessment, which rules out the necessity for data collection in the conventional techniques, which cannot be done during the pandemic. The same can also be used in the post-pandemic situation by forecasting the volume of the background traffic.

An average of 6.6% percentage increase (maximum 25% on some roads) in V/Cs was observed before and after the new development on the adjoining roads. This might lead to heavy congestion on those roads with a poor level of service. Therefore, several scenarios were considered and analyzed to provide the best mitigation strategies to improve the level of service. The scenarios considered include work from home (30 and 50% employees), staggered time (ITC trips divided equally between 8–11 AM and 5–8 PM), mode share (shifting toward public transportation and other sustainable modes) and combinations of the three. The best scenario obtained from this was a combination of 50% work from home, staggering of work hours from 8–11 AM to 5–8 PM, and ITC mode shift has brought down the V/Cs to an average of 30% during AM and 50% during PM from BAU scenario. This scenario is suggested for implementation for more congestion and hassle-free movement of vehicles on the adjoining roads. Using the methodological framework developed, this work can help evaluate how appropriate a particular development is for a location, for any given situation, and what mitigation strategies and types of transportation improvements may be essential to maintain smooth moving traffic with a satisfactory level of service. The recommended scenarios have to be further investigated to see how feasible these are, post-COVID. Although work from home can continue to some extent, further study on the possibility of 30 and 50% work from home is required.