Model for Estimating Limiting Access, PT and Egress Distances for Different Modes in a Public Transportation System (Part-1)

Abstract

Origin and destination nodes of a trip by a direct mode would get the labels of First Origin (FO) and Last Destination (LD) when the same trip is performed by public transport (PT) with PT nodes {stop(s)/station(s)} of PT leg of the trip at origin and destination ends being second origin (SO) and first destination (FD) respectively.  Geometry of hexadecant (22.5°) circular and annulus sectors formed by the cardinal and intercardinal directions suggests that the limits of access/egress distances for ring radial PT networks is about 8 kms at a distance of 20-km radius from the PT network center of a city or an urban area of 25-km radius. The 8-km limiting access/egress distances envelope the additional distance of 5 kms beyond the 20-km point bringing peripheral areas of an urban area within the influence of PT network.  Transport authorities of different cities across the globe have defined distance bands less than 1.6 kms or that distance, that can be traveled in about 10 minutes on either side, along public transport corridors as their influence area for walk and cycle as access/egress modes in the planning guidelines. However, the literature of access/egress distances reported greater than 1.6 kms for these modes, and; further higher (8+kms) for motorized modes (personalized/hired). And, the average travel times by access/egress modes are in the range of 12-18 minutes except for buses and cycles which have greater averages as reported in literature. The literature also reported longer access/egress distances for faster PT modes and terminal type stations and stops as compared to slower PT modes and other types of stations and stops respectively. A review of walk access distances in the literature reported mixed impacts (positive, negative, n-shaped, u-shaped and insignificant) for the same studied factors between studies except for the factors (operational frequency, use frequency and type) related to PT which exhibited positive, u-shaped and insignificant impacts. However, the PT trip length is reported to have mixed impacts (positive and negative) vis-à-vis access and egress distances between studies.  

The reported range of access and egress distances of a PT network are within the limits of distances governed by geometry of hexadecant circular and annulus sectors suggesting access/egress trips are not likely to significantly alter the direction of travel from that of base trip.

In the light of this and mixed impacts of PT distance on access and egress distances, it is felt appropriate to examine mathematically their relationship with base trip length, speed and time, and;  rate (monetary) and cost of travel associated with respective modes on each leg of travel, by base and PT modes.

That is, patronage to PT services depends on, whether the total travel time and total travel cost, considering both access and egress modes and PT mode, are respectively less than or equal to the base mode travel time (Eqns. 10) and base mode travel cost (Eqns. 12) or at least one of them.

It is determined mathematically (based on the definition of ICR (Inter Connectivity Ratio [33] - Eqn.14) that the sum of limiting travel times of access and egress modes is a maximum of half of the travel time by base mode and; each has a maximum of a quarter of the base travel time. That is, one of the two could be nil and the other a maximum (equal to quarter of the base travel time) suggesting PT availability at one of origin or destination nodes. These limiting values will be lower considering the updated definition of ICR [36] which includes ticketing time, waiting time and others in the PT system in addition to PT mode travel time (Eqn.40). Updated ICR may be termed as ICM (Index of Co-Modality [36]) based on comparison with ICR and another alternate (Eqn.46).

The limits of travel costs of access and egress modes (individually and sum) have similar characteristics in relation to base travel cost based on an ICR, defined in cost terms in this article, similar to the one in travel time terms.

Limiting access and egress distances by different access and egress modes are a fraction of base trip distance with fractions being one-fourth of the ratios of speeds [of access (or egress) mode to base mode] or rates [of base mode to access (or egress) mode]. Estimated limiting access and egress distances based on speeds would be different from those estimated based on rates and; the minimum of the two is the overall limiting distance. Desirable (minimum), acceptable (Arithmetic mean) and tolerable (maximum) access/egress distances for a base mode, base trip length pair are defined.

Limiting access and egress distances along with the relationship between minimum and maximum speeds (of access/egress modes) and their averages (arithmetic mean and harmonic mean) confirmed that access/egress distances for faster PT mode services (low density networks) would be higher compared to the slower PT mode services (high density networks) observed in literature. The access/egress distance models exhibited that access/egress distances would be lower with increasing speeds of base mode. This explains the reason for low patronage of public transport in peripheral areas (urban sprawls) where speeds are normally higher and as most of the trip generating units would be beyond the limits of access/egress distances.

The estimated access and egress geometrical distances (straight line) of different modes based on average speeds and rates in an urban area for base trips of lengths between 5 and 30 kms at 5-km intervals suggest that the direction of travel due to PT usage would not be significantly different from that of the base direction intended between an origin node and a destination node.

That is, access and egress nodes of PT service would be within hexadecant sectors on either side of the base direction of travel between the origin node and destination node and; within circles, at one or both ends, of radii equal to the arc length (39 percent of the base trip length) of the hexadecant sector of base trip length radius. The limiting length of travel by PT would vary between 0.2 to 1.8 times of the base trip length based on the ratios of rates [of base mode to access (or egress) mode]. And, limiting PT travel length based on corresponding speed ratios would be between 0.65 times the base trip length and it itself; indicating that these estimates are a subset of those based on rates.

The envelope in which base or access, egress and PT travel occurs, between origin and destination nodes of a trip, has a geometrical shape of a stadium (Fig-15a).

Introduction

Public transportation (PT) systems in urban areas comprise of both road and rail-based systems. Road-based systems are conventional bus systems (CBS) and bus rapid transit systems (BRT). Rail-based systems are suburban (or commuter) rail (SR), Metro services and light rail transit (LRT). CBS as it operates on roads on a more flexible network is directly or more easily accessible compared to other modes of public transport, commonly termed MRTS (mass rapid transit systems). However, CBS shares road space with other modes of transport and is prone to delays associated with congestion as a result of mixed traffic conditions discouraging its patronage. MRTS, on the other hand, operates on a dedicated network on roads or rail and; on surface, elevated or underground according to space constraints and hence is not easily accessible. Thus, MRTS patronage, in some systems or on certain corridors, is short of its full potential being a high (or higher) capacity system compared to CBS. Addressing the low accessibility, if any, of MRTS stations through appropriate measures would enhance the utilization of MRTS capacity and aids in reducing road congestion levels.

The difference in accessibility is primarily due to the location and spacing of the bus stops on CBS network and that of stations on MRTS networks. The location of bus stops could be closer to the traffic generating points on CBS network due to its flexibility. The location of stations on MRTS networks is governed by the system requirements and size of infrastructure required to serve the passenger demand. The efficacy of each of the modes of MRTS in urban transportation systems is not only dependent on the quality of its infrastructure and operational planning but also on other supporting transport systems. That is, integration of MRTS with other modes of transport, either private personal modes or intermediate public transport systems (IPT), commonly termed feeder services and last mile connectivity, as well as that of other forms of public transport systems.

Last Mile Connectivity

Last mile connectivity (LMC) is a phrase adopted from tele communications network terminology. A typical telephone network (Fig-01) comprises of exchanges (of different levels) and devices at end user premises in user zones. The connecting network from the last level of exchange to user zones is the last mile in the telephone network. LMC is a constraining factor for expansion of the telecommunication services. However, addressing the issues which are constraining LMC for a particular user zone and development of the LMC network provides communication facilities with all other user zones on the telephone network. That is, user at one end is automatically connected to users at other ends through the development of an LMC in a thus far unconnected zone. For example, if the north-western corner user zone (NWZ) is connected to the corresponding exchange through an LMC network NWZ gets connected with the entire network of user zones. That is end users in NWZ can communicate within themselves and all end users in other user zones.

FOLD

In transportation system a user moves from one end (origin) to another end (destination) and users from the same zone have different destinations. Thus, addressing the LMC issues at one end may encourage certain users to use MRTS whose destination end has appropriate LMC of their choice, and; not all likely users. Users desire appropriate connectivity at both origin and destination ends. LMC issues in transportation are not restricted to just the development of a network (infrastructure); in fact, it may exist as a plan or part of the zonal network, but also that of supporting infrastructural facilities for parking, drop and pickup points for personalized modes (owned or hired) and; terminal facilities for feeder bus system. 

That is, LMC in telecommunication is a system related constraint whereas, in transportation it’s a combination of system, operations and user’s preference.

Four different types of trips (Fig-02) between an Origin (O) and a Destination (D) involving the use of MRTS are performed requiring different supporting facilities for encouraging MRTS patronage. The origin outside MRTS network is termed First Origin (FO) and likewise the destination outside MRTS network is termed Last Destination (LD).

T1 – Trips with both origin and destination on the MRTS network. E.g. O2-D7 and O6-D3

T2 – Trips with origin on MRTS network and destination outside of it. E.g. O3-LD1 and O7-LD4

T3 – Trips with origin outside of MRTS network and destination on it. E.g. FO4-D2 and FO8-D6

T4 – Trips with both origin and destination outside of MRTS network. E.g. FO1-LD5 and FO4-LD8

The location of station, which feeds traffic to MRTS, is critical to attract passengers from generating zones. However, it is constrained by the system and space requirements. The station location results in it (or more specifically the MRTS) being attractive to some passengers to ride MRTS; and not, for some others. The attractiveness of MRTS is thus dependent on the mode one is likely to use to access a station from FO or to egress from a station to LD. The modes used for the two purposes of access and egress are walk; personal modes - bicycle, motorcycle or car; IPT – cycle rickshaw, auto rickshaw or taxi and; minibuses termed feeder buses. The set of modes used to serve accessing (incoming) or egressing (exiting) passenger demand at a particular MRTS station constitute the FOLD transportation system (TS) along with the supporting infrastructure needed for efficient operations at and in its influence area of that station.

Station Influence Area

People use any of the FOLD-TS modes to or from a station depending on how much spatial separation exists between the station and actual location of either the FO or LD of their trip. That is, the feeding mode selection depends on the distance one has to travel to or from a station.The influence area or the catchment area of a station (Fig-03) could be conceptualized as concentric circles with station as the center point and circumference of different circles as the range (limiting distance) of different modes. The actual size and shape of the influence area also depends on the network structure and category of stations, such as; intermediate (typical), central (transfer) or terminal.

Influence Area of Adjacent Stations

Inter station distance is longer in rapid rail and suburban rail systems and; is shorter in metro and BRT systems except limited stations that are apart. Thus, in the later systems, influence area of adjacent stations  is likely to overlap resulting in an exclusive elliptical (Fig-04) influence area  for each station. The gaps at the far end of the influence area of each station close-up as the inter station distance reduces and form a strip (Fig-05) of the influence area (sectional or line influence area) unless physical barriers (canals, rivers, etc.) exist at the edges.

Line Influence Area

The station influence area also depends on the parking and; pick and drop facilities availability at each of the stations or their absence for respective feeder modes. Usually terminal station (Fig-05) locations are chosen such that the requisite and adequate facilities could be provided there at. The absence or low level of facilities for certain category of feeder modes at some intermediate stations is mainly due to the lack of space at such stations. It results in one or more stations on the corridor, depending on the space availability at their location and attractiveness, being developed as central (or interchange or inter-modal or transfer) stations with facilities required for handling feeder services. The influence areas of the central stations overlap the influence areas of adjacent intermediate stations.

Network Influence Area

Network structure – a grid, a radial, a ring radial type or any other (Fig-06) – influences the size of the influence area of a station. Different lines in the network form the boundaries of zones (with each being a set of influence areas of stations) which may be termed network (MRTS) zones. That is, a network zone may be served by several stations. Also, a network zone may be a set of complete or partial traffic zones which are the basic aggregate areas of a transport plan. The overall influence area of a station between two parallel lines or two concentric rings of PT could be the entire space between the corresponding lines or adjacent inner space(s) of the respective lines; or excluding the center space, if the two lines are spatially separated by a wide intervening area. In the case of radiating lines, the intervening space of the two lines closer to the radiating point (or center) would be the influence area of the stations on both the lines and as the lines move out the space splits in to two halves, with one half being the influence area of one line and the other half being that of the other line. Further down the lines, the intervening space will expand leaving the center space, as in wide parallel and concentric rings, being not a part of the influence area of either lines.

Influence Area Size

The accessibility of stations on a network are governed by the network zone’s shape and location of its centroid. In a ring-radial network, (Fig-06 and Fig-07), the network zones are circular sectors (S_i) or annulus sectors (A_i). Cs₁ and Cs₂ are centroids of circular sectors  BOC and JOK, formed by the inner (radius r₁) and outer (radius r₂) ring metro corridors passing through arcs BC and JK. C_A1 and C_A2 are centroids of annulus sectors BCKJ and JKTS. The annulus sector BCKJ is enveloped by two ring metros and two radial metros. And, the annulus JKTS, by three metros, two radials and one ring on the inner side and by an outer (peripheral) ring (radius r₃).

The centroid of circular sectors, such as BOC, indicate that the maximum access distance to a metro line from any point in any subsector, such as, CS₁-O-C (bound by a particular metro line, or MRTS in general, and the lines joining the centroid to the end points of that metro line forming the sector) is the height of the respective subsectors. There are three subsectors in a circular sector (Fig-07) with heights of two being equal.

An inner concentric circle of radius approximating to or equal to centroid of the circular sector BOC defines the core influence area of central and neighbouring stations. The core area is delineated in to two pockets (1 and 2) and which are respectively served by their adjoining radials. These stations in core area can function as interchange stations for other lines that pass through this area reducing the load at central station.

The residual area of the circular sector BOC post core influence area delineation would be an annulus sector – which corresponds to annulus sector bdCB – forming the inner influence area of the inner ring metro line. The spatial separation (Bb) of the inner arc (bd) and outer arc (BC) of this annulus indicates the maximum access distance to the inner ring metro line from any point in the annulus.

In case, the circle of core influence area passes through the centroid of circular sector BOC, that is,   , then, from (Fig-06 and Fig-07) and Eqn. 3:

In case, the circle of core influence area does not pass through centroid of circular sector BOC, then, is assumed or predetermined indicating, value of Bb from Eqn. 7b is less than or greater than the value of Bb from Eqn. 7a.

The influence area of the metros along the radials of the annulus bdCB would be half of the annulus size; for, the delineator is the angle bisector of the radials. The maximum access distance to the metro lines along each of the radials from any point in the respective sub-annulus sectors varies between half of the arc lengths of arc bd and arc BC.

That is, the annulus sub-sector comprising pockets 3 and 6 is served the by both the ring and an adjoining radial metro, and; the annulus sub-sector comprising pockets 4 and 5 by the same ring and another radial metro. 

 The spatial separation of arcs of annulus BCKJ (Fig-06 and Fig-07) is the difference in the radii  of the circles which formed this annulus. That means, from any point in the annulus, one of the ring metros, inner or outer, is at a maximum accessible distance of half of this difference. However, in case the arc passing through the centroid is considered as the delineating line then maximum accessible distances to the two ring metro lines would differ.

The maximum average accessible distance to the metro lines along the radials would be half of the arc length of the central arc of the annulus. The annulus may be delineated in to eight pockets such that the common boundary, along the central arc, of each of the pockets (outer/inner) is a quarter of a length of the central arc. Such a delineation results in maximum average accessible distance from any point in the peripheral pockets (10/11 or 7/14) to adjoining radials would be quarter of the arc length of the central arc. The average accessible distance to the radials, from a point in the central pockets, 8/9 and 12/13, would be between quarter and half of the arc length of the central arc.

The annulus JKTS is outside of the outer metro ring and the pockets 15 to 18 are served by the outer metro ring. And, the pockets 19 and 22 are served by the radial metro lines. Pockets 20 and 21 may not be within the influence area of any of the metro lines serving the annulus depending on the radii of the two rings forming this annulus.

This distance gives the spatial separation between parallel corridors along a diameter and a parallel chord or between two parallel chords.

Influence Area - Metro

BRT and metro corridors have been planned and are under development in many cities across India. Metro network (Fig-08) in Delhi, the national capital, is in the final stage of the third phase of development and the fourth phase commenced. It comprises (not by lines of operation but as a network of corridors) of a north-south (shown in Yellow) and an east-west corridor (shown in Blue) along with a chord running east-west in the north (shown in Red) and a centre-south-east radial (shown in Yellow). It also comprises of a high-speed metro from centre to airport in the south-west (shown in Orange) with an average inter station spacing of 4.5 kms compared to an average inter station spacing of 1.3 kms on the regular network. It also has two corridors (shown in Cyan), one an almost circular one and the other a semi-circular one (combining with an arc section of the east-west corridor running in E-ESE sector) along the ring roads in the city. The E-W corridor also has an arc section in W-WSW sector. And a link is likely to connect WSW end of the two metro lines (Blue and Orange) to SW end of the N-S corridor (Yellow).

Circular transport corridors are by their nature of service have a common starting and ending point and are not a perfect circle as they deviate and weave to connect traffic generating centres.

The spatial separation distances between the centre of metro network at Rajiv Chowk and the stations on the inner circular metro corridor vary between 6 and 12 kms with an average spacing of 8 kms. The average spatial separation for outer circular metro corridor is 11 kms with spatial separation from centre and its stations varying between 9.5 and 14 kms.

The dimensions of core and annuli influence areas as defined earlier and other relevant dimensions for a mass transport network with an inner ring of radius 8 kms and outer 11 kms, and; influence rings with 3 kms spatial separation are presented in (Table-01) and (Fig-09).

The centroid of the circular sector, formed by the inner ring metro, 8-km radius, is at 4.80 kms for NW-C-NE like sectors or at 5.20 kms for E-C-NE like sectors, as centroid location varies with angle of the sector under consideration. It may, thus, for practical application be considered that the centroid of inner ring metro circular sectors is at 5 kms from the centre. That is, the core circular area is of 5 kms radius. This suggests that the inner ring metro is at a maximum accessible distance of 3 kms from any point in the core annuli formed between the edge of the core circular area and the inner ring metro.

The radial metros, from any point in the core circular area would be at a maximum accessible distance of about 1.9 kms in the case of E-C-NE radials and about 3.5 kms in the case of NW-C-NE radials (based on Eqn. 6). 

These distances are also applicable for any point in core annuli which are common to the core triangles (Example: Pockets 3 and 4 Fig-07) formed by the centroid of inner ring metro circular sectors.

The chord line in red weaves in a 2-km band, for, its inner and outer edges are at 3.5 and 5.5 kms (based on Eqn. 9) respectively from the central axis (or E-W line in blue). This indicates that a metro line is at an average access distance of 2.25 kms from any point in the band between chord metro corridor and E-W metro corridor.

The spatial separation of the two ring metros varies, with it being about 3 kms in the southern annuli (SW - SE) and, about 6 kms in the eastern (SE - E) and western (SW - W) annuli. That is, from any point, in southern annuli, a metro line is at an average  access distance of about 1.5 kms and; in eastern and western annuli, it is about 3 kms.

This indicates that a metro line is at an average access distance of 2.25 kms from any point in the annuli formed by the two metro rings.

In the cases of the annuli which are closer to the centre and smaller angles of annular formation, the centroids of the annuli will be closer to the intersection point of the line on which centroid is located and the midpoint of the central arc of the annulus. In such cases the difference could be ignored for all practical purposes of identifying the influence area(s) of metro lines enveloping an annulus.

Example: Annulus Sector_5-8 centroid is at 6.45 kms from the centre and the mid-point of the central arc is at 6.50 kms from the centre.

The average arc length of annulus subsectors or half-octant (hexadecant [37]) annuli of octants, such as, E-C-NE, of the ring metros, for example, that of Delhi, weaving mainly between 8 and 14 kms radius; is 4.32 kms, ranging between 3.14 and 5.50 kms. The maximum arc length of hexadecant annulus being 7.85 kms at a radius of 20 kms towards the periphery of a city / study area.

One of the metro corridors (Delhi Airport to Tughlakabad) under consideration in Phase-4 of Delhi Metro, which could be an arc of a third ring metro, weaves between 11.00 and 17.00 km radii circles. The arc length of hexadecant annulus with a radius of 17:00 kms is 6.68 kms.

It may be assumed that the influence area of metro extends, along a radial, in the outward direction, up to a distance equal to that of the arc length of hexadecant annulus at the location of a peripheral station on a radial metro or that of a station on peripheral ring metro.

That is, the influence area from the centre of the metro network, which is not geometrical centre; for example, in Delhi; would be about 24 kms considering the peripheral ring metro is abutting a 17-km radius circle.

Incidentally, this is the size of the radius of equivalent circles with which the geographical areas of Delhi and Kolkata are represented (Table 02).  Indian cities are classified as linear or circular (including semi-circular/ linear-circular). All Indian cities, except one, including the four large metropolitan cities, have an area (that includes municipal area as well as the surrounding planning area) less than 1900 sq. kms. That is, Indian cities have an area less than the area of an equivalent circle of radius 24 kms and all circular cities with 4 million plus population have a radius greater than 15 kms [01-02].

It may thus be assumed that the influence area of a metro network of a city or an urban area, in general, extends up to a maximum of about 8 kms across and ahead at and beyond the boundary of the urban area respectively in to the surrounding planning areas.

This indicates that the influence area of a metro network of circular cities / urban areas (15 to 24 km radius) in India could at maximum be a circular area of about 23 to 32 kms radius from the city centre or metro network centre. Unless, metro network extends in to surrounding planning areas.

Arc lengths, ranging between 3.14 and 7.85 kms, of hexadecant annuli (Fig-10) govern the access distance to radial metro corridors depending on the placement of the radial metro corridors within a hexadecant annulus. That is, weaving within hexadecant annuli or fairly moving along any of the radial bounds of the hexadecant annuli.

Bus services from/to points beyond 8.00 kms may be considered as intermodal services and not as feeder services for such routes may cater more to interzonal transport demand and not exclusively the feeder transport demand.

Varied access distances, either average or maximum, presented earlier, in different influence areas of circular, annular or triangular shapes and sizes, indicate the requirement of using different types of feeder transport modes for access/egress operations to/from a station.

Considering 8 kms as the upper limit for access distance, a range of distances (Table 03) are suggested for different feeder modes for purposes of planning based on the arc lengths of hexadecant annuli. Suggested ranges consider a distance range of up to 3.2 kms for pedestrians and non-motorised transport (NMT) modes and above 3.2 kms for motorised transport (personal/hired/mass).

Actual travel distances by feeder modes would be higher than those identified in distance ranges as they are geometrical distances to a metro corridor and not based on actual travel paths; which, depends on the road network in an urban area and also presence of physical barriers, if any, between a station and the point of origin or destination; that is, the detour factor. Thus, the infrastructure facilities need to be planned accordingly. In practice users in the lower range may opt for a mode suggested in the higher range and vice versa. That is, user characteristics.

Travel Characteristics

An Indian study, on traffic and transportation policies [01], based on a sample of 30 cities, established statistical relationships between average trip length (kms) of all modes in a city with relevant parameters. The parameters considered were, population, area (sq. kms.), and shape factor of city. A polynomial model was developed with the geographical area and; a multiple linear regression model with population and the shape factor, which is the ratio between minimum and maximum spread (in kms) of a city.

Eighteen (18) of the cities are classified as circular shaped with a dozen (12) of them having a population of over one million (Table 02) which, is the threshold for initiating the planning of metro (or in general high capacity) services for a city as per National Urban Transport Policy [03].

The average trip length of all modes is linearly related with the radius (Fig-11) of an equivalent circle of the same area as that of a city.

Average trip length of all transport modes, as a single group, is low as it includes active modes of transport which have smaller trip lengths. For example, Trip lengths of active modes – walk, cycle and cycle rickshaw – in Delhi (1994) were 0.83, 4.89 and 2.26 kms respectively. While, the motorised modes – car, motor cycle, auto rickshaw (6.14 kms), taxi and bus – had trip lengths ranging between 10.0 and 11.5 kms, a narrow band, and that of suburban rail is 18.17 kms [04].

Trip lengths of active modes – walk and cycle – increased to 2.2 and 6.9 kms respectively and the trip length range of motorised modes upped to a wider band of 10-15 kms as per census of India, 2011,  and; average trip length by metro is 18.9 kms in Delhi [05].

The trip length, in Delhi, by metro is similar to that of suburban rail in 1994. So also, that of motor cycles, increasing from 10.03 [04] to 10.5 [05] kms. The trip length of cars increased by 30 percent to 14.6 [05] from 11.28 [04] kms. And that of buses by 16 percent to 12.4 [05] from 10.66 [04] kms.

This suggests that, in Delhi, in 1994, prior to the introduction of metro services in 2002, the average trip lengths, of motorised modes, were around 10.75 kms. However, by the completion of phase-2 of metro development in 2011, the trip lengths of personalised modes – motor cycles, auto rickshaws and cars – had distinct trip lengths of around 11, 13[05] and 15 kms respectively. Likewise, the trip lengths of mass transport modes – buses and metros – had distinct trip lengths of 13 and 19 kms respectively. This distinctly indicates that the trip lengths by more convenient modes are longer than the modes that have lower convenience.

Average travel time of personal motorised modes in Delhi in 1994 was around 37 minutes and decreased to about 30 minutes in 2001 (Table 04). That is, during that period the speeds of motor cycles and cars increased by about 40 and 20 percent respectively and; that of autorickshaws and buses by 20 and 10 percent respectively.

The increase in speeds could, perhaps, be attributed to the fact that, that period coincided with the introduction of new technology models, in to the transport sector in India, beginning mid 1980s and galloping by mid 1990s to the end of the century. It also coincided with the liberalisation of Indian economy during the same period, beginning 1991, stimulating their faster introduction [07-11].

The increase in speeds are validated by the observation that Basic Desired Speed (BDS) of new technology car models (96.25 km/h) is higher than those of older models (71.40 km/h). It is also reported increase in BDS of all other modes on stretches with improved road infrastructure. So also, free speeds have increased, as indicated by the intercepts in speed-flow models [12]. The models also suggest speeds decrease with increasing flows.

As the same set of vehicles move on urban road networks, it may thus be fair to assume that the average speeds in cities are impacted by the pace of adoption of new technology vehicles by individuals and also the pace of infrastructure development to meet the increasing traffic flows as result of urban population growth. This fact is validated by the observed speeds in cities in 1994 and 2007 and; thereafter.

The average speeds, in 2007, across four categories of cities (based on population size – M – millions) – 1-2M, 2-4M, 4-8M and 8M+ – are 18, 22, 19 and 17 km/h respectively [01]. The draft [02] compared the average speeds in 1994 and in 2007. This comparison indicated that the average speeds in cities of one million plus population were in the range of 19 to 27 km/h (1-2M population cities) in 1994 and 10 to 29 km/h (4-8M population cities) in 2007 as the speed ranges of other categories are a subset of these ranges. The comparison suggests that speeds in certain cities are higher in 2007 and of others are lower than the speeds in 1994.

An e-cab service provider reported that the speeds in all the 4M+ population cities except Ahmedabad (Table 02) are in the range of 17 (Kolkata) to 23 km/h (Delhi/Pune) in 2015. All the others are in the range 18 to 21 km/h with Mumbai (20 km/h) having the average of the overall range. Speeds in the other three cities are 18 (Bangalore), 19 (Hyderabad) and 21 (Chennai) km/h [13]. It also reported speeds of 20 km/h in Bangalore in 2017 (an increase between 2015 and 2017) and 17 km/h in 2018, a value lower than that in 2015. That is a fall of 3 km/h since 2017 which is also reported for the other 4M+ population cities [14].

Considering Bangalore as a reference and its’ ups and downs in car speeds between 2015 and 2018, it may be appropriate to assume that car speeds in 4M+ cities vary between 15 and 25 km/h. And, the overall range in 1M+ population cities, car speeds vary between 10 and 30 km/h, enveloping the 4M+ range, depending on the population growth (travel/traffic demand) infrastructure development (traffic management) and rate of technology adoption in cities’ car fleet. That is, the average car speed in 1M+ population cities would be about 20 km/h.

Access/Egress Characteristics – National

Delhi (Metro)

Access and egress characteristics in in Delhi, in 2010, through a small sample study, indicate use of different modes at the two ends of a metro trip. Exception to this is, metro users walking to access metro also mostly use walk as a mode for egress trip. Walk and cycle rickshaw are the predominant access modes (73 percent of access trips) and walk the egress mode (65 percent of egress trips). Average total access and egress time is about 21 minutes, and; metro travel time is about 29 minutes with journey time being 50 minutes. Sample did not reveal the use of car or cycle as access and egress modes. Majority of non-metro users indicated willingness to use metro provided better feeder services to support metro patronage are available resulting in addressing the long journey time [15].  

Access and egress characteristics in Delhi, in 2011, also indicate maximum metro users walk (44 percent) to and from stations. And least use either a cycle or a motorcycle, together less than 5 percent. Auto rickshaw usage is about 21 percent and the rest of the modes – cycle rickshaw, car and bus - about 10 percent each [16]. Majority of the potential users (52 percent) of metro, in 1994, indicated walk as an access/egress mode [04].

The walk, in 2011, is the predominant access/egress mode although with a lower share of total metro users than that reported (~52 percent) in 1994. Bicycle and motorcycle share of access/egress modes, in 2011, is higher than that indicated (~3 percent) in 1994. Hired modes – cycle rickshaw and auto rickshaw – combined share of access/egress modes, in 2011, is about 4 times the indicated share (~8 percent) in 1994 while the car share is over 5 times that indicated (~1.5 percent) in 1994. The share of bus as an access/egress mode, in 2011, is about one-third of the share indicated (~34 percent) in 1994.  

The access/egress distance (Table 05) on an average is about 4 kms by motorized modes and a maximum of 5.5 kms by buses. The least is that of autorickshaws followed by motorcycles in the increasing order and both lower than the overall average while that of cars and buses higher than the overall average. The least access/egress distance is by walk (0.58 kms) which is about one-tenth of that of buses. Access/egress distance by cycle rickshaw is about 2 kms. Access/egress trip length and population density are significant factors in determining mode choice for access and egress trips. Access/egress distances of personalized modes (including walk) are higher for modes with high average speeds. Same is the case with fared modes, auto rickshaws and buses as well as cycle rickshaws with an average speed same as that of cycle or lower. The speeds of auto rickshaws are lower as they include both personalized and shared categories. Average inter connectivity ratio (ICR – ratio of access and egress time to total travel time by public transport and access/egress modes [33]) is 0.38 with it varying between 0.2 and 0.5 for approximately 88 percent of trips. PT (metro  and bus) trip length estimates (different for different ICR values and based on its definition), increase with increase in access and egress time, and, are distinctly different by a factor corresponding to the speeds of the respective PT modes and the gap between the two trip lengths also widening. Bus stop and station density of respective networks suggest lower access/egress distances to bus (6000 stops) than that to metro (132 stations) [16].

A six-station alighting passenger study in 2015, with a sample size of thirty at each station, on Delhi metro network indicated the use of NMT (non-motorized transport) modes by more than half the users at four stations for access and egress trips and less than half at two peripheral stations on the network, in fact about a quarter, at one. Maximum average egress distance at one of the peripheral stations, a terminal station, with a large parking space for private vehicles, is about 4.2 kms indicating that the influence area of that station is large. Users at this station have high average total access-egress distance with an access distance of about 1.2 kms, the least at the sampled stations. Total average access-egress distance at three of the stations, including the two peripherals, is about 5.1 to 5.4 kms. The third station with this characteristic serves a university and the adjoining areas with an almost equal split between access and egress distances; and access distance (2.7 kms) being the highest among sampled stations. The least egress distance of 0.7 km is at a station serving a tourist place with sparse development on one side and dense low-income residential development on another side which is perhaps the reason for low utilization (currently used by tourists only) of a large parking space. The station, located in a dense and congested area with primarily commercial and mixed land uses and narrow roads, in the walled city, also has an equal split access-egress distances of about 1.25 kms. Total travel time of access (15.8 mins) and egress (11.6 mins) modes of metro users at this station is more than the travel time by metro (22.8 mins). At other stations, the same (ranging between 11 and 21 mins) is lower than the travel time by metro (ranging between 26 and 41 mins). Mostly, trip length (access or egress) on one end of metro users is shorter than the other. The overall ratio (of distances) with the minimum of access or egress to the other is 1:2 varying between 1:4 and 1:1 at individual stations [18].

Another limited size random sample study of metro user in Delhi revealed predominant use of slow modes (walking – 60 percent and cycle rickshaw – 10 percent) for access and egress trips. It reported a threshold walking distance of 2.5 kms beyond which user percentage drops substantially. The fitted cumulative frequency distribution suggests access time of less than 12.5 minutes for ~78 percent of non-motorized access trips and ~58 percent of motorized access trips. The corresponding values for egress trips are ~88 and ~67 percent respectively. It also reported maximum access/egress time of ~30 minutes for both motorized and non-motorized access/egress modes. ICR for ~80 percent of trips is less than 0.42 with maximum being about 0.8 [19].

Delhi (Bus)

An on-board bus passenger survey in 2002 in Delhi on 20 percent sample of bus routes indicated that less than one percent of cycle owners use cycle as an access mode with most (96+ percent) walking to bus stop and the rest other modes (motorcycle, cycle rickshaw and auto rickshaw). Access trip length for cycle users varies between 3 and 8 kms with an average of 6.67 kms. Walking access trip length, for about 56 percent of cycle owners is over 500m, and; overall, for about 49 percent (cycle owners and non-owners); with maximum distance being over 2 kms and average being 0.75 kms.  Average access trip length by cycle rickshaw is 1.61 kms, by motorcycle 2.94 kms and by auto rickshaw 3.75 kms [20].

Access/egress characteristics of bus passengers at bus stops on two sampled routes in Delhi in 2014 indicated over 87 percent walked and the rest used cycle rickshaws or shared auto rickshaws. The average walking distance is 647m and the 85^th percentile is 1069m, varying between 637m (work) and 677m (shopping) with education (654m) and recreation (660m) in the middle. Average walking distance reduced with increase in age. Among males it decreased from 667m (15-25 years) to 638m (35+ years) and among females from ~644m (15-25 and 26-35 years) to 632m (35+ years). The maximum interpreted distance from fitted Lognormal curve is ~2.5 kms [21].

Mumbai (Sub-Urban Rail)

Household surveys around two stations on the sub-urban rail network in Mumbai indicated an average access walking distance range between 786m and 1245m for different categories of users (by occupation, income and other categories) with an overall average of 910m and maximum of 2500m. Acceptable average walking distance range for different categories is between 1050m and 1900m with desirable range (15^th percentile values) being in between 340m and 600m. Average access cycling distance range for different categories is between 1984m and 3066m, with acceptable range between 2900m and 4050m and; desirable range between 1020m and 2420m. On an average desirable access walking and cycling distances are 500m and 1600m respectively. About 86 percent of trips less than 1250m access distance are by walk and about 53 percent of trips above 1250m are by bus. Bus (up to 1800m), auto rickshaw (up to 2000m) and taxi (up to 3000m) are preferred for access distances between 1250m and 3000m due to fare advantage and; beyond this range private vehicles car and motorcycle substitute hired modes auto rickshaw and taxi. Most motorized personal or hired modes have access distances of less than 12 mins travel time and; most buses and cycles higher travel time duration. Walk has the second largest share in the two-time slabs with some walk trips lasting up to 20-30 minutes for distances of 2-3 kms. Motorized personal or hired modes have trips with travel time 5-20 minutes and; bus and cycle 10-25 minutes. Average access distance to access modes (excluding personalized modes) is 235m with it being less than 100m for about 48 percent users and less than 300m for another 32 percent. That is, for about 80 percent it is less than 300m. The average waiting time of users is 3.66 minutes for access modes, perhaps, due to lower frequency of these modes at residential land use end, and; 0.62 minutes for egress modes, again perhaps due to higher frequency of these modes at non-residential land use end resulting in a need for improving availability of access modes at residential land use end [22].

Kolkata (Proposed BRT)

Zone of Tolerance (ZOT) is a service quality measure on a user-bound scale between Desired Service Level and Acceptable Service Level varying between individuals, user groups and over time. Desired Service Level is user’s belief that it can be provided and should be provided, while; Acceptable Service Level is that which users believe is the threshold for acceptance [23]. It [23] estimated ZOT of nearness of bus stop (i.e. access/egress distance) on two bus corridors along a proposed BRT corridor in Kolkata, through a random on-board survey on nine bus routes. The two ZOTs, desirable and minimum acceptable, for bus stop nearness, respectively are 175m and 321m for one corridor, and; 278m and 573m for the other.

Access/Egress Characteristics – International

Walk influence area is, generally considered to be, an acceptable or maximum walking distance and as a thumb rule set at a radial distance of 500 to 1000m from a public transport line/corridor in general, or precisely a station or a stop. One cited study defined it as a circular area of a radial distance that is travelled in about 10 minutes. That is, about 800m at 1.33m/s average walking speed. This distance is different in different regions and is for example 1600m in Great Britain for one-way walk trips, 1200m in Toronto, Canada for rail passengers and about 750m (at 1.25 m/s) in six metropolitan areas in Korea for subway passengers [24].

The planning guidelines of transport authorities in Sydney, Perth, Vancouver and Helsinki specify an area up to 300m to 800m from a public transport (bus/rail) route as the service area to cater to about 90 percent of the households in public transport contract zones. Sydney specifies 400m for daytime and 800m for nighttime by scheduling services accordingly. Studies indicated public transport users walk longer than rules of thumb straight line distances [25].

Bangkok, Thailand (Metro)

A limited sample, 249 respondents across 48 transit stations, study; within an area of 1-km radius examined the access characteristics of TOD (Transit Oriented Development) residents and the possibility of users of motorcycle taxis (moving on sidewalks) shifting to walk as an access mode. It indicated majority (62 percent) walk to a station and 22 percent use motorcycle taxis with the rest using cars, vans, taxis, buses and songthaew (a pickup truck or shared two-rowed taxi). Average access distance by walk is 335m and by motorcycle taxi 582m. Overall average access distance is 428m and by other modes 632m. Two distances  – actual (based on shortest paths between station and FO / LD locations using Google Maps) and acceptable (respondent’s opinion of the farthest place to walk) – were compared. Actual average walking distance, of walkers, 335m (6.5 minutes) is less than the average acceptable walking distance 498m (9.1 minutes). Actual average walking (travel) distance for motorcycle taxi users, 544m (7.4 minutes, 582m, by motor- cycle taxi), is higher than the average acceptable distance of 508m (9.2 minutes). The share of walk and motorcycle taxi access trips are 76 and 17 percent respectively of all access trips within 500m; while, the same are 25 and 35 percent respectively of all access trips between 500 and 1000m. The share of car and taxi access trips are 14 and 8 percent respectively while that of mass transport modes, bus and songthaew are 6 and 8 percent respectively of all access trips in the upper distance slab (500 - 1000m) [24].

A study of 600 respondents at six stations in BMR (Bangkok Metropolitan Region) in a multidimensional analysis indicated a fair level of average accessibility (ranging between 0.54 to 0.63) based on predictors (between 0 – poor accessibility and 1 – good accessibility on a standard scale) identified through factor analysis.  The disaggregated analysis indicated accessibility concerns concerning women, elderly and disabled. Nearly 77 percent respondents live at a radial distance farther than 500m from a station with about 75 percent living in areas between 500m and the station district boundary and 2 percent beyond it and within BMR. The rest, about 23 percent, live within 500m radial distance from a station with a 5 to 10 minutes’ walk.  Bus is the major feeder mode (40 percent) followed by van (23 percent) and para transits (hired motor- cycles and taxis – preferred modes of elderly and disabled) including boat and personalized vehicles (cycle and car). Average travel time (includes access, waiting and in metro times and excluding egress time) between origin and destination is about 40 to 60 minutes with poor to fair accessibility level. Egress distance for majority (70 percent) metro users is less than 500m, especially, the elderly. Some have longer egress distances and use multiple modes (walk-paratransit/bus, walk-bus-paratransit, and walk-bus/van-boat) to the final destination as is to access a station [25]. 

Sydney, Australia (Bus/Rail)

Access characteristics in Sydney, to bus stop or rail station, based on Household Travel surveys (HTS), 2006 to 2008, indicate about 98 percent of walk access trips are less than 2 kms and only three are longer than 5 kms. The distance estimation is based on (x, y) coordinates of access trip origin and destination using ARCGIS software. Predominant access mode is walk with its share being 89 percent to bus stop and 50 percent to rail station. Rail passengers also use car – self driving or as passenger (17 percent each) – and bus (14 percent). Public transport share of all trips is equally split (approximately) between bus (5.2 percent) and rail (5.8 percent). Average walking distance to rail (805m) is longer than bus (461m) with overall (public transport) average being 573m varying between 235m (lower quartile - LQ) and 824m (upper quartile - UQ). Train and bus users have distinct access distances and; trip and demographic characteristics. The access distance by train users varies between 539m (LQ) and 1018m (UQ). Corresponding values for bus users are 162m (LQ) and 655m (UQ). An average walk access distance of 891m and 1167m (UQ) was reported for train users with a train travel time between 30 and 45 minutes (significantly different from that in less than 15 minutes train trip time). Average and UQ access distances for bus users are 517m and 860m respectively for a bus travel time of more than 45 minutes. The average travel time (/distance) by trains is 34 minutes (/19 kms) and by bus is 23 minutes (/6.4 kms). Average walk access distance to trains varied between 772m and 807m (>800m) during different peaks of the day and is 873m (>800m) in the evening. Corresponding values for the bus are 447m and 491m during the day and; 480m in the evening, with all greater than 400m. Walking access distance is not significantly related with different demographic and trip characteristics. Variations in walk access distances to bus and rail, once decided to use public transport, is perhaps, primarily due to the differences in the supply of the two public transport modes. That is, stop and station spacing (or their density on respective networks) considering the existence of 35000 bus stops in comparison to the 300 rail stations and; perhaps, due to greater level of non-residential land uses abutting rail corridors as against bus corridors [26].

An earlier study (2004) indicates walk as an access mode for 47 percent rail users and as egress mode for 84 percent. Average walk access distance is 700m (10 minutes) and egress distance is 600m (9 minutes). Average distance by bus is 5 kms for both access (16 percent users) and egress (12 percent users) trips with an average time of 15 and 18 minutes respectively. Average access distance by car for ‘park and ride’ users (16 percent) is 7 kms (13 minutes) and for ‘drop and ride’ users (19 percent) is 6 kms (11 minutes). Park and ride users walk to the station for about 300m (5 minutes) on an average. Users of bus as access mode, on an average, walk 300m to access it and 200m after alighting it to reach station. Access distances with long averages to some stations across different locations varied between 8 and 15 kms for cars and between 6 and 9 kms for buses [27].

Toronto, Canada (Surface Transit)

Access characteristics to Surface Transit (subways, streetcars, buses and new technology transit) in Toronto, based on 2001 TTS (Transportation Tomorrow Survey) dataset indicated distinct access distances (airline distance of trip origin to transit corridor, not stop or station) of users of bus (including streetcars) and subway systems. The access distance to later system being about 1.6 times that of the former. Average access distance of bus users in CBD and inner-city areas is 150m and 200m respectively, while; that of subway users is 230m and 350m respectively. In outer-city areas the subway access distances are about 375m to 450m. This is primarily due to denser transit network in core-city areas as compared to that in peripheral areas as also the difference in stop and station spacing of the respective modes. TTC (Toronto Transit Commission) defines service area of 300m radius from any point on the transit corridor not differentiating the different modes of the Surface Transit. The percentage of users within 300m service area varies between 56 and 69 percent in different, outer to inner, areas, while; it varies, between 74 and 86 percent within 500m radial area from the transit corridor; indicating a large share of (18/17 respective percentage) transit users are outside of the defined service area limit or in the buffer area. Socio-economic characteristics (for example, dwelling type and vehicle ownership among others) have a fair relationship and travel characteristics (except number of transfers) a weak relationship with access distance to transit services. Transit users prefer to walk more than transferring on one or more occasions [28].

Jinan, China (BRT)

Access and egress characteristics of an over 1200 respondents sample indicated that the walk distances (geo-measured on the basis of user markings of origin and destination and route on a map) varied across 3 station categories – Typical, Transfer and Terminal – at 19 BRT stations along 3 corridors. Average walking distances are the lowest at Typical Stations (549m) and the highest at Terminal Stations (1392m) with Transfer Stations (586m) in between suggesting walking distance at Terminal Stations is over double that of a non-Terminal Station. Over 80 percent of respondents had an access distance more than 600m (the defined buffer area at each station) to a terminal station. The corresponding maximum walking distances are 2738m, 5114m and 2067m respectively. Comparison of access/egress distances with airline distances (reverse estimated) indicated highest detour (1.59 = 475m/299m) along an arterial-edge BRT (with stations in mid-block and curbside) corridor. The detour values of two other types of corridors – integrated boulevard and below expressway – running along the arterial medians (or roadway center) with stations at intersections are almost equal with them being 1.36 (= 649m/477m) and 1.33 (= 580m/436m) respectively. This suggests that the influence area defined as a radial distance from a station is least (about 300m) along an arterial edge corridor in comparison to integrated boulevard (500m) and below expressway (450m).  Considering corresponding maximum access distances (1635m, 2023m and 2738m), the influence areas for each corridor type could be 1000m, 1500m and 2000m respectively. Integrated boulevard corridor has a significant positive relation with access distance indicating on an average, users walk about 160m more than that at the other two corridors. Statistical modeling suggested station related parameters had significant relationship with access distance and the same is not true with socio-economic and travel parameters. Results further indicated the need for flexibility in station influence area definitions, i.e., defining influence areas of different sizes considering station context and corridor types, and not a unitary for the entire city. The model-based influence area guidelines varied between 300m (lowest) for a non-terminal station on an arterial-edge corridor and 1500m (highest) for a terminal station on an integrated boulevard corridor. The influence area range for stations on below expressway type corridor is between 400m (non-terminal station) and 1200m (terminal station) [29].

Rio de Janeiro, Brazil (Metro)

Access mode to the subway stations is walk for about 65 percent of the users and cycle for a mere 0.2 percent users according to studies by the subway operator. A study at 2 terminal stations, with one representing median-high income and high-density neighborhood (MHIHDN) and the other median-low income and low-density neighborhood (MLILDN), on 2 subway lines indicated that the predominant mode (over 50 percent of subway users) being walk at one station and bus at the other. Walk (57 percent) is the predominant mode at the MHIHDN station due to proximity of the station followed by subway integrated bus system (36 percent) and bus (2 percent) and; nil cycles. At the other terminal (MLILDN), bus (53 percent) is the predominant mode followed by cycle (23 percent), walk (11 percent), and subway integrated bus system (4 percent). Automobile-based modes (car, van taxi or ride) share at these stations, is low, and; are 5 and 9 percent respectively at the two terminals (MHIHDN and MLILDN). Bus and subway integrated bus system access mode users at the two terminals (MHIHDN and MLILDN) travel from distances as far as 30 or 50 kms respectively. About 90 percent of the access trips at these two stations have an access time less than 40 minutes and the rest have access times of up to 60 minutes or more (4 percent at MHIHDN station and 1 percent at MLILDN station). At MHIHDN station most of access trips (73 percent) are in the access time range 4 to 15 minutes. About 83 percent of access trips at MLILDN station are in the access time range 11 to 40 minutes [30].

Madrid, Spain (Metro)

Average walk access distances from home and workplaces to metro stations are 516m and 471m respectively. The comparison of these distances on the current street network in the city with that of the estimated distances for four types of alternate street network scenarios that exist currently in other urban areas indicated that the estimated distances are lower than the current distances in all but those estimated for ILD (Irregular Low Density - typically observed in suburbs) scenario. The estimated corresponding distances in ILD scenario were 590m and 543m respectively indicating an additional travel distance of 73±1m. The other three  scenarios are Irregular High Density (IHD - typically observed in traditional and historical city centers), Orthogonal Grid (OG – typical small rectangular blocks of streets) and Station Oriented (SO – radial pattern street networks). The SO scenario had the least distances with the corresponding distances being 412m and 375m respectively indicating savings in travel distance of 100±4m. The corresponding distances in the other two, IHD and OG scenarios, being in the range 450m to 500m are similar to the distances in the current scenario consistent with types of networks currently prevalent in the city. This analysis was carried out by superimposing these network types on the current metro network and stations using GIS.  Population and employment cumulative curves by distance bands (100m) indicate that their respective differences between different types of street networks tend to decrease beyond 600m and are negligible at 1200m suggesting street network design and placement of population and employment activities in different distance bands has an impact on the access distances and station demand.  The study estimated decreasing detour factors (ratio of network access distance to Euclidean distance) across different alternate street networks, with the highest being for ILD (1.53/1.64) type and the least for SO (1.21/1.14) type and IHD (1.27/1.23) / OG (1.24/1.21)  in the middle in the same order based on access distances estimated for the 16 cardinal points around a station for the normally adopted station service area threshold boundaries (at 400m/800m respectively). The estimated average access distances for each type are greater than the magnitude of the radius (400m/800m) of the threshold boundaries [31].

Access/Egress Characteristics – Salient Observations

A.     National:

1.      Access and egress modes of most trips are usually different with some exceptions;

2.      Walk is the predominant mode for both access and egress trips;

3.      Walk access mode mainly results in it being the egress mode too;

4.      Access and egress distances differ by network and stop/station densities of bus and metro networks respectively;

5.      Access/egress distances to bus stops are lower than that to metro stations;

6.      Access/egress distances at peripheral stations/stops are higher than at other locations;

7.      Cycle usage is limited even among the cycle owners;

8.      Access/egress trip length determines the choice of access/egress modes;

9.      Average access distance to hired access/egress modes is 250m;

10.    Average wait time for hired access modes at residential land use end is higher, at about six times, than that for hired egress modes at non-residential land use end;

11.    Average of access and egress wait time for hired modes at respective ends is about 2 minutes;

12.    Average access/egress time is less than 12.5 minutes (except for cycle and bus – greater than 12.5 minutes) and maximum 30 minutes;

13.    Access-egress (or egress-access) distance ratio varies between 1:1 and 1:4;

14.    Interconnectivity ratio is greater than 0.5 for a small percentage of trips;

15.    Access or egress distances (of different forms of mass transport – bus / metro / sub-urban rail – modes) in different cities, approximately, ranged between:

a.      Walk – 0.5 to 2.5 kms;

b.      Cycle – 2.5 to 6.5 kms;

c.      Cycle rickshaw – 1.5 to 2.0 kms;

d.      Auto rickshaw and taxi – 2.0 to 4.0 kms;

e.      Car – 3.0 to 6.0 kms;

f.       Motorcycle – 3.0 to 4.5 kms;

g.      Bus (feeder mode) – 2.0 to 7.0 kms

B.     International:

1.      Public transport service area or influence area, by walk, by definition, by transport authorities, across different cities ranges between 300m to 1600m radial distance from a stop/station or the corridor;

2.      Service area is also defined as a radial distance that is travelled in n (e.g.10) minutes;

3.      Some cities define different sizes for daytime and nighttime based on frequency of service;

4.      Walk access/egress distances are longer than thumb rule based radial distances;

5.       Detour factors ranged between 1.14 and 1.36 (excluding 1.5+ values) in service areas based on either road network type in a service area or by public transport corridor type serving an area with latter category having higher detour factors;

6.      Average access/egress distances are governed by level of placement of population and employment activities in different bands of a defined size (e.g. 100m) within the limits of identified stop/station service areas and beyond;

7.      Access/egress distances differ by the land use – residential/non-residential – abutting the public transport corridor;

8.      Access/egress distances are distinctly different between different forms of public transport;

9.      Access/egress distances vary by stop/station spacing or their density on the respective networks;

10.    Access distances are higher in peripheral areas than in inner areas;

11.    Access/egress distances vary by station type – typical, transfer and terminal;

12.    Long access distances up to 15 kms and over are reported at some terminal stations;

13.    Average access distance to station from park and ride is 300m (5 minutes);

14.    Average access distance to feeder bus stop is 300m;

15.    Average access distance to station from feeder bus stop is 200m;

16.    Average access/egress time is less than 18.0 minutes and maximum 40 minutes (except for buses at terminal stations – 40-60 minutes);

17.    Access or egress distances (of different forms of mass transport – bus / metro / sub-urban rail – modes) in different cities, approximately, ranged between:

a.      Walk – 0.15 to 2.0/5.0 kms;

b.      Motorcycle – 0.6 kms;

c.      Car – 6.0 to 8.0/15.0 kms;

d.      Bus (feeder mode) – 5.0 to 9.0/50.0 kms

The trips associated with long access distances (beyond 8.0 kms radial distance) may be considered as intermodal trips.

An exploratory review, of over 40 studies, of walk access distances to public transport indicates walk distances vary across different regions by different factors (demographic, socio-economic and trip/travel related) with varied levels of significance.  Most studied factors, including density, in the reviewed set indicated one or more than one of; positive, negative, insignificant, n-shaped and u-shaped effects between studies for the same factor. Exceptions to this are for those related to public transport (PT-frequency, PT-type, PT-trip length and PT-use frequency) and employment for which the impacts are one of  positive, insignificant and u-shaped barring PT-trip length, as one study out of four showed a negative impact and the remaining three positive. Eight studies examined the impacts of employment with two reporting positive and the rest insignificant impact.  Frequency of use by PT was examined in four studies with one reporting positive, one u-shaped and the other two insignificant impacts. One of the eleven studies that examined the impact of public transport type reported insignificant with ten reporting positive indicating public transport users walk more for rail than metro/LRT and shortest for buses [32].

 [Click here for Part-2] [Click here for PDF] - Parts (1 to 3)