JAN-2014

The Indian Roads CongressE-mail: [email protected]/[email protected]

Founded : December 1934IRC Website: www.irc.org.in

Jamnagar House, Shahjahan Road,New Delhi - 110 011Tel : Secretary General: +91 (11) 2338 6486Sectt. : (11) 2338 5395, 2338 7140, 2338 4543, 2338 6274Fax : +91 (11) 2338 1649

Kama Koti Marg, Sector 6, R.K. PuramNew Delhi - 110 022Tel : Secretary General : +91 (11) 2618 5303Sectt. : (11) 2618 5273, 2617 1548, 2671 6778,2618 5315, 2618 5319, Fax : +91 (11) 2618 3669

No part of this publication may be reproduced by any means without prior written permission from the Secretary General, IRC.

Edited and Published by Shri Vishnu Shankar Prasad on behalf of the Indian Roads Congress (IRC), New Delhi. The responsibility of the contents and the opinions expressed in Indian Highways is exclusively of the author/s concerned. IRC and the Editor disclaim responsibility and liability for any statement or opinion, originality of contents and of any copyright violations by the authors. The opinions expressed in the papers and contents published in the Indian Highways do not necessarily represent the views of the Editor or IRC.

Volume 42 NumbeR 1 JaNuaRy 2014 CoNTeNTs IssN 0376-7256

INDIaN HIGHWaysa ReVIeW oF RoaD aND RoaD TRaNsPoRT DeVeloPmeNT

Page

2-3 From the editor’s Desk - “Roads as a Means to Manage Food Inflation”

4 Important announcement-74th annual session to be Held at Guwahati from 18th to 22nd January 2014

5 Short Panelled Concrete Pavement in Built-Up Area Rajib Chattaraj and B.B. Pandey

13 Distresses in Concrete Pavement : A Case Study of Indore Byepass Vandana Tare and Akanksha Asati

20 Feasibility of providing A Skywalk for Pedestrian in Chandni Chowk, Delhi Purnima Parida, Jiten Shah and S. Gangopadhyay

30 Assessing Liquefaction Potential by CPT/Shear Wave Velocity Method and Evaluating Effect of Liquefied Soil on Foundation Design Bharat Katkar and Poonam Pendhari

44 Effect of Pedestrian Cross - Flow on Capacity of Urban Arterials Satish Chandra, G. Srinivasa Rao and Ashish Dhamaniya

52 Optimum Maintenance Predictions and Derivation of Transition Probability Matrices for Bridge Deterioration Modeling S.R. Katkar and Prashant P. Nagrale

66 Road Safety Audit and Safety Impact Assessment S.K. Chaudhary

76-80 Circular Issued by MORT&H

81 Tender Notice of NH Circle, Lucknow

82 Tender Notice of NH Circle, Kanpur


84 Tender Notice of NH Circle, Madurai

85 Tender Notice of NH Circle, Tirunelveli


87-88 IRC membership Form

2 INDIAN HIGHWAYS, JANUARY 2014

Dear Readers,

First of all, I wish “a Very Happy & Prosperous New year 2014” to all the esteem readers of Indian Highways.

Do the good roads and efficient road connectivity have a bearing on supply chain management? It may be a debatable issue for some and to some it may be more like a statement. One fact which remains undisputed is that the road is the basic transportation means and its efficiency to a larger extent influences the economic activities and development of the region.

Over the last few years, the country is facing the vagaries of high fluctuations in the prices of food commodity including the perishable items like fruits & vegetables. This year the impact has been remarkably visible. The food commodity is the basic essential need of the mankind and high inflation in this segment severely affects the majority of the population and destabilizes the economic growth, which means the other economic activities gets impacted-some to a major extent and some to a lesser extent. What role and to what extend the road sector can & should play its imperative role? Not much thought has been given to it but time has come when a serious thought is required to be given to this aspect.

The roads lead to spread of civilization. The spread of civilization means more production required for various segments of commodities. The demand & supply governs the pricing and simultaneously the economic viability of the producer/producing entity. However, one segment of the commodity i.e. food segment is witnessing a continuous increase in demand because of not only spread of civilization but also because of increasing aspirations & related demand creation. The road perhaps provides the easiest and the cheapest means of connecting the producing/production areas with the consuming/consumer areas. Some may argue that railway is the cheapest mode of transportation of commodities but it cannot compete with the flexibility and the better linkage connectivity of the roads. The cost factor to a greater degree can be accommodated if the road connectivity is made more efficient coupled with sustainability. The inclusiveness, which the road provides, cannot be provided by any other mode of transport. Therefore, it is preferable and also high time that the road development programmes are given special impetus and the due dedicated attention. The sector also requires conceptualization with due diligence. The delivery system requires a revisit to ensure that the projects are delivered not only in the prescribed time schedule but at the shortest possible time span. There have been some instances where on the one hand farmers of the producing areas are facing huge oversupply of the fruit/vegetable resulting in highly depressed prices and on the other hand at the same time in the consuming areas/cities, the consumers (the people) are suffering

From the editor’s Desk

RoaDs as a meaNs To maNaGe FooD INFlaTIoN

eDIToRIal

INDIAN HIGHWAYS, JANUARY 2014 3

from high prices of the same fruit/vegetable due to poor arrival of the produce. Don’t you think that better road connectivity and road transportation system would have helped in creating a win-win situation for both the farmers (the producers) and the people in the cities (the consumers)? It is not limited to one or two instances but if deeply evaluated the repeated instance of mismatch happening between demand and supply of the food articles which perceptibly can be addressed in a better manner if an efficient and dependable road transportation system is in place.

The world’s biggest rural road programme i.e. PMGSY is bearing its good result in the rural areas in the country and the high growth rate being experienced by consumers good industry is a testimony of the same. The rural areas are being provided with good roads but their proper integration and inter linkages with the higher category of roads need attention as the same may help to a greater extend the fastest movement of agricultural produce to the consumer areas. This all the more necessitates that the stalled/delayed road sector schemes/projects are completed at the earliest. No one may dispute that the slowness in the road sector has a cascading impact on other schemes & sector of economy. In addition to arresting the slowness in this sector, the changing scenario also demands for revisiting the inter linkages of the road sector development programmes with other sector development programmes especially the one related to creation of storage including cold storage facilities, etc. If the same are created along the arterial road routes, it may perhaps bring the cheers and smile on the faces of all and may also perhaps help in better management of food inflation.

“Strength does not come from physical capacity. It comes from an indomitable will”

Mahatama Gandhi Father of the Nation

Place: New Delhi Vishnu shankar Prasad Dated: 19th December, 2013 Secretary General


aTTeNTIoN INVITeD

For any enquiry about the 74th Annual Session like Registration, Membership etc. please address to Secretary General, (Kind Attn. Shri D. Sam Singh, Under Secretary) Indian Roads Congress, Kama Koti Marg, Sector-6, R.K. Puram, New Delhi-110 022. Phone + 91 11 26185273, 26185315, 26185319, E-mail: [email protected], or contact the following officers:

Registration membership Technical Presentation accommodation and Technical exhibition

Shri S.K. Chadha Under Secretary (I/C) Phone: + 91 11 2338 7140 E-mail: [email protected], [email protected]

Shri Mukesh Dubey Section Officer Phone: + 91 11 2338 7759 E-mail: [email protected]

Shri S.C. Pant Section Officer Phone: + 91 11 2618 5273 E-mail: [email protected]

Shri Suryya Kr. Baruah Local Organizing Secretary & SE, Building Circle -I, Highways, Guwahati- 641 018 (Assam) Phone: 0361-266 9873 M.: +91-98640 33268 E-mail: [email protected]

On the invitation of Government of Assam, the 74th Annual Session of the Indian Roads Congress will be held at Guwahati (Assam) from 18th to 22nd January, 2014. The Invitation Booklet containing the Tentative Programme, Registration Form, Accommodation Form etc. has already been sent to all the members and the same is also available in our website www.irc.org.in. Accommodation is available on first come first serve basis.

It is expected that more than 3000 Highway Engineers from all over the country and abroad will attend this Session. During the Annual Session of IRC, there has been a practice for various firms/organizations to make Technical Presentations on their products/technologies & case studies (with innovative construction methods or technologies or having special problems requiring out of the box thinking and special solutions). The presenters will get an opportunity to address a large gathering of highway professionals from Private Sector as well as decision makers in the Govt. Sector. These presentations evoke lively interactions among the participants.

A time slot of about 15 minutes is normally allocated for each Technical Presentation to be made through Power Point. Time is also given for floor interventions. Audio-visual equipment is made available at the venue for these Presentations. During such Technical Presentation session no other meetings will be held parallel so as to ensure maximum attendance during the Technical Presentation session. The stakeholders are, therefore, requested to participate in the event and book the slots at the earliest.

Interested Organizations may write to IRC conveying their willingness for participation and send the topic of their Technical Presentation by E-mail at [email protected] or through Speed Post alongwith a Demand Draft for Rs.50,000/- (Rupees Fifty Thousand only) drawn in favour of secretary General, Indian Roads Congress, New Delhi latest by 5th January, 2014 so that necessary arrangements can be made by IRC. Requests received after 5th January, 2014 will not be entertained. Since the time slot available is limited, the interested firms/organizations may reserve the slots at the earliest instead of waiting for the last date.

74th aNNual sessIoN To be HelD aT GuWaHaTI FRom 18TH To 22ND JaNuaRy 2014

ImPoRTaNT aNNouNCemeNT


sHoRT PaNelleD CoNCReTe PaVemeNT IN buIlT-uP aReaRajib ChattaRaj* and b.b. Pandey**

* Executive Engineer, P.W.D., West Bengal, E-mail: [email protected]** Advisor, Sponsored Research and Industrial Consultancy and Former Professor and Head, Civil Engineering Dept., IIT Kharagpur, E-mail: [email protected]

absTRaCTBituminous roads in built up areas are damaged in monsoon due to poor drainage. Cement concrete pavements become the obvious choice in such locations. Short panelled Concrete pavements develop much lower wheel load stresses than the conventional ones and require lower thickness. The paper describes construction of short panelled concrete pavements over stone set pavements in a built up area in Burdwan district, West Bengal. Stress analysis by Finite Element shows that such a pavement may have a long life due to much lower stresses. The cost of the pavement is a little higher than the bituminous pavements but far lower than the conventional concrete pavements whereas the durability is expected to be much higher than bituminous pavement, and can be same as that of conventional concrete pavement.

1 INTRoDuCTIoN

Bituminous pavements in built up areas are usually subjected to adverse moisture conditions due to inadequate and clogged drainage resulting in heavy damage during every monsoon though the proportion of traffic carrying heavy loads is not high. Concrete pavements constructed in different cities in India with inadequate drainage have performed well but the initial cost of the conventional concrete pavement is quite high because of higher thickness. A new type of thinner concrete pavement with shorter panel size similar to the white topping over bituminous pavements as per IRC:SP:76-2008(1) can be used in the construction of concrete pavements for village roads and city streets because of low flexural stresses caused by shorter panel sizes.

Such pavements may be termed as Panelled concrete pavement. Thinner concrete pavements in the form of short slabs of dimensions 0.5 m × 0.5 m to 2.0 m × 2.0 m formed by creating weakness by saw cutting to one third depths can be a low cost option for a durable pavement. White topping with short panels

has been used as overlay over damaged bituminous pavements after necessary repair at many locations in Delhi, Mumbai, Pune, Bangalore etc. and these measures have been a success. Indian Roads Congress has brought out a guideline IRC:SP:76-2008(1) for thickness design.

Using similar concept, concrete slab with short panels formed by saw cutting to one third the depth of slab thickness can be laid over the granular or cement treated sub base or any other non-conventional type like stone brick pavement for the construction of road pavements at a lower cost than a conventional concrete pavement. Type of subbase should depend upon the traffic they carry. Cemented subbases are necessary for heavy traffic condition. Cement concrete pavements constructed in cities over water bound macadam have performed well even under poor drainage condition.

The present paper describes the design aspect and construction of a short panelled concrete pavement over a road crust consisting of stone bricks as well as premix carpet over granular layers with poor drainage condition for a street in a small town in Burdwan district of West Bengal. Though heavy traffic is not heavy, the drainage condition is poor and bituminous surfacing was getting damaged frequently. It is expected that a pavement made up of short panelled concrete pavement would last much longer in spite of poor drainage condition due to low proportion of heavy vehicles. The new type of concrete pavement termed as Panelled Concrete Pavement is of size 1.0 m × 1.0 m similar to white topping(1) was used in the construction in the project. Construction details and checks for the safety of pavements against cracking due to wheel load have been examined. The

TeCHNICal PaPeRs


constructed pavement completed in March 2012, has been performing well during the last two years in spite of two heavy monsoons after the construction.

2 ReVIeW oF lITeRaTuRe

Examples of panelled pavements over granular or cementitious bases are limited. Most concrete pavements with short slabs are laid as overlays over milled bituminous pavements and the overlay is generally assumed(1) to be bonded to the bituminous base. A number of concrete pavements (2-9) with panels of size 0.5 m x 0.5 m to 1.5 m x 1.5 m with thickness from 50 mm to 150 mm have been constructed in USA including India. Such pavements laid as overlays over damaged bituminous pavements after necessary repair are known as white topping. Bonding of concrete slab with the milled bituminous surface was considered in the analysis. No analytical design method is available currently. An approximate design method is suggested in IRC:SP:76-2008(1). There is plenty of gap in the knowledge about the values of flexural stresses caused by wheel loads that may develop in the slab due to wheel load.

3 aNalysIs

3.1 If a pavement is made of small concrete blocks as shown in Fig. 3.1, there is only compressive stress at the bottom.

Fig. 3.1 Compressive Pressure at the Bottom

Fig. 3.2 Tensile Stresses in Slab due to Bending Moment in a Large Bottom

If the size of the slab is increased, tensile stresses are caused in the slab due to the bending moment from the overhanging part of the slab beyond the wheel contact area caused by the reaction from the soil/foundation. For wall or column, the foundation slab is thick and the deflection caused by loads is nearly constant implying nearly uniform reaction from the foundation soil. Since the deflection of the pavement slab is larger near the points of load application and decreases with distance from the loaded area, the reactive pressure that is proportional to the deflection is not constant. Three dimensional finite element method is the only method of stress analysis in such slabs of finite dimension. The foundation is considered as a Winkler foundation, also known as Dense Liquid foundation, and the pressure at a point is proportional to the deflection caused by the slab and is given as

p = k δ ... 1

Where,

p = pressure on the slab from the soil, MPa

k = modulus of subgrade reaction, MPa/m

δ = deflection of the slab, m

3.2 Fig. 3.3 shows a 4.0 m × 4.0 m pavement consisting of sixteen panels each a metre in length and width with one third depth from the surface saw cut to create a plane of weakness to induce full depth cracks to form interlocking panels due to zigzag cracks.

Fig. 3.3 4.0 m × 4.0 m pavement with 1.0 m × 1.0 m Panels formed by Saw Cutting to one third depth of the

Slab Saw Cut

TeCHNICal PaPeRs


Stress is analysed by placing a dual wheel carrying a load of 50 kN at a tyre pressure of 0.8 MPa tangential to an edge as shown in Fig. 3.3. The other dual wheel assembly of the concerned axle would be about 2.00 m away from the centre of the dual wheel causing little interference in stresses caused by the loaded panel shown in Fig. 3.3. The slabs may have load transfer at the weakened saw cut joints over a long period of time because of lower panel size. If there are plenty of overloaded vehicles, the interlocking behaviour of joints may decrease with time and load transfer may become negligible across a joint towards the terminal stage of the pavement life. It is, therefore, safer to compute stresses considering that there is no load transfer through the joints. If a pavement is laid over a cemented subbase, bonding may be weak, particularly when the concrete slab is cast after curing of the cemented subbase. ANSY’s Finite Element Software has been used for stress analysis. Each panel of size 1.0 m × 1.0 m and depth 150 mm is divided into fine mesh of size 80 mm size in each direction. The corner and edge springs of the Winkler foundation have their stiffnesses one fourth and half of those of interior ones respectively. Solid 45 element and combino 14 spring element of ANSY12 were used for the analysis.

4 sITe DeTaIls

4.1 The site is located in Mankar town, Burdwan district of West Bengal, 140 km westwards from Kolkata. The work of short panelled concrete pavement was done on a 600 meter long stretch of very bad conditioned road. The first 400 meter stretch of the road was constructed with 150 mm thick jhama brick consolidation, 150 mm thick base course (WBM) and 75 mm thick stone brick on top and the rest 200 meter length of road was topped with thin premix carpet and seal coat with the same under layers. It is located in a built up area which is subjected to water logging during the monsoon bringing about damage and depression on the stone brick surface and large ditches to the premix carpeted surface (Photo 1, 2 & 3). The road carries traffic of about 428 Commercial Vehicles per day and some heavily

loaded vehicles are also expected. Axle load data is not available. Cross sections of the two road sections are shown in Figs. 4.1 (a) and (b).

Photo 1 Dilapidated Condition of the road during Rainy Season.

Photo 2 Pre work Condition of the Pavement with Poor Drainage.

Photo 3 Stone Brick Set Pavement with Damaged Spots

TeCHNICal PaPeRs


Fig. 4.1 Cross Sections of two Stretches of Existing Road Overlaid with 100 mm of M 10 Lean Concrete and 150 mm of

Panelled M-40 Grade PQC

4.2 It was decided to construct a concrete pavement of 4.50 m wide with 150 mm depth PQC of M-40 grade with panel size of 1 m × 1 m formed by sawing to a depth of about 50 mm. There was no scope of greater width because of built up area on either side of the road. A lean concrete base of an average thickness of 100 mm with M-10 grade was done to provide a uniform support as a levelling course below the panelled concrete pavement.

5 CoNsTRuCTIoN oF CoNCReTe PaVemeNT5.1 both m-10 and m-40 Grade Concrete was

done as Ready mix Concrete. (Photo-4 & 5).

Photo 4 RMC Plant Site

Photo 5 RMC Plant and Transit Mixer

The source of coarse aggregate was Panchami variety, Dist: Birbhum (W.B). 50% of 20 mm down and 50% of 10 mm down stone aggregates were mixed to achieve 20 mm graded aggregate of nominal size as per Table-1 of IRC:44 and Table-2 of IS:383. The fine aggregate was Damodar river sand of Zone-III as per Table-2 of IRC:44 and Table-4 of IS:383. The cement used was OPC-43 grade. Super plasticizer was used as chemical admixture. The design mix of M-40 grade concrete was done as per IRC:44-2008 read along with IS:10262-2009. The mix proportion of M-40 concrete stood per m3 of concrete as 450 kg cement, 644 kg sand, 649 kg 20 mm down stone chips, 649 kg 10 mm down stone chips, 162 kg free water and 5.4 kg of super plasticizer. The mix proportion of M-10 has been designed as – per m3 of concrete – cement 280 kg, sand 777 kg, 20 mm down stone chips 637 kg, 10 mm down stone chips 637 kg, free water 168 kg and chemical admixture (super lasticizer) 3.36 kg. Since laying of concrete of even M-10 grade was done by pumping operation, chemical admixture was required to be added which resulted the finish product with higher strength than the requirement of M-10 grade of concrete. This was done with no extra cost than the rate of M-10 grade of concrete as per S.O.R.

5.2 Quality control measures for concreting work was taken both at RMC plant site (Photo-6 & 7) and the construction site (Photo-8 & 9). Concrete cubes from every batch were taken both for M-10 and M-40 concrete. Concrete cubes were taken at construction site also. Slump was set at plant site as 100-125 mm. The slump obtained at the construction site was in the range of 70-80 mm. Work was done during the month of March when the ambient temperature was around 38ºC. Thus, the drop in slump value was quite expected, in fact, due to this reason the initial slump at plant site was kept at a bit higher side. The characteristic compressive strength of concrete cubes were tested at 28 days, some were tested on 7 days.

TeCHNICal PaPeRs


On an average, compressive strength of lean concrete after 28 days was obtained as 23 MPa and that of PQC was 48 MPa.

Photo 6 Taking of Concrete Cubes and Slump at Plant Site

Photo 7 Testing of Characteristic Compressive Strength of Concrete Cubes

Photo 8 Slump Testing at Site

Photo 9 Taking of Concrete Cubes at Site

5.3 The laying of concrete was done manually. The undulation of the stone brick pavement was levelled with 100 mm average thick of lean concrete (M-10). Also the correction of camber had been made in this layer. Over the stretch on which M-10 lean concrete was laid in a day’s work, say 100 to 120 meter, on the very next day the PQC of 150 mm M-40 grade concrete was laid on the same stretch. Camber was maintained as 2%. (Photo-10 & 11).

5.4 Creating a discontinuity on the top one third depth of the concrete pavement became a challenging problem. Alternative to saw cutting was examined. Measures like putting metal and plastic strips and coated plywood (Photo-12) were also explored. After the initial hardening of the concrete, even the coated plywood was difficult to be detached to get a distinct groove of 3 mm. Since there was little time for experimentation with different alternatives, the conventional method of construction of joint cutting with a diamond saw to one third the depth of the slab was adopted (Photo-13). Grooving with diamond cutter up one third depth was done immediately on the next day morning of the previous day’s execution of M-40 grade concrete. The appropriate time for cutting the groove is very important in the sense of making a distinct as well as easily cut able groove with diamond cutter.

TeCHNICal PaPeRs


Photo 10 Laying of Concrete from Transit Mixer

Photo 11 Laying of Lean Concrete as Levelling Course

Photo 12 Making Discontinuity with Coated Plywood Strip

Photo 13 Making 3 mm Groove by Diamond Cutter

5.5 Immediately after making groove with saw cutting, curing was started. Curing was done with water ponding for at least 28 days (Photo-14 & 15). For the long period of curing as well as the concreting work, the road was blocked for 50 days. Local block administration and district administration extended their co-operation to arrange for a diverted route.

Phot 14 Curing of Concrete by Total Ponding

Photo 15 Curing of Concrete with Water Spray After Initial Setting

TeCHNICal PaPeRs


6 moDulus oF subGRaDe ReaCTIoN aND sTRess ComPuTaTIoN

6.1 The subgrade has a CBR of 5 and the corresponding k value is 42 MPa/m as per Table 2 of IRC:58-2011. Considering Tables 3 and 4 of IRC:58-2011, the effective k value over 100 mm lean concrete base is about 230 MPa/m. It may be mentioned that only fourth root of k value matters in flexural stress computation and a little variation in its value has negligible effect on flexural stresses. A value of 200 MPa/m is used in the analysis of stresses.

6.2 For the loading condition shown in Fig. 3.3 assuming no load transfer across the joints, the computed stress for the dual wheel load of 50 kN corresponding to the legal axle load limit of 100 kN (10.2 Tons) having a tyre pressure of 0.8 MPa is found as 1.70 MPa which is well below the 28 day modulus of rupture of 4.4 MPa for M40 concrete pavement. Curling stress due to temperature gradient is very low(10) and may be neglected all together. Since occasional overloading due to construction traffic is very common for most roads, the pavement will be safe even if the axle load is 200 kN since the computed stress value for this load is 3.40 MPa. Repeated action of traffic also will not be able to cause early damage due to low stresses. Load transfer at the joints and slight bonding with the lean concrete will impart additional strength. Temperature stress is ignored because they are negligible as clearly stated in the recently council approved draft of IRC:SP:62. Lean concrete was laid and got set before laying of PQC, hence full bonding between PQC and Lean concrete cannot be ensured. Therefore, long term bonding action between the two layers has not been taken into consideration. Concept of IRC:58-2011 for bonded layers can be used if the two layers are laid just one after another before the concrete of both the layers are set.

It may be noted that for the same dual wheel load of 50 kN, pavement width -3.75 m, transverse joint spacing - 4.0 m, the flexural stress due to load and temperature differential as per IRC:SP:62:200411 is found as 4.66 MPa for West Bengal which higher than the

28 day flexural strength of 4.4 MPa. A much higher thickness is needed if fatigue is also considered.

7 CosT ComPaRIsoN

7.1 Cost of this short panelled concrete pavement of 150 mm M-40 grade PQC along with 100 mm lean concrete with allied items had come to about Rs. 1500/- per sq. meter. Cost would have come even lower, had there been no lean concrete as levelling course. In comparison to this type of pavement, the other alternatives which are generally adopted on roads with poor and clogged drainage such as short lived 50 mm thick Bituminous Macadam and 25 mm Mastic Asphalt had been estimated as Rs. 1000/- per sq. meter and conventional rigid pavement of 250 mm PQC and allied items were estimated as Rs. 3300/- per sq. meter. Since composite action of the two layers is doubtful for long term performance, the method adopted in the paper for computation of stress gives a safer design.

8 CoNClusIoNs

1. Panelled concrete pavements can be a good alternative for reducing the cost of concrete pavements for built up areas, rural roads, bus bays etc.

2. Stresses are reduced drastically in concrete pavements with panels of size 1.0 m × 1.0 m.

3. If alternative route can be arranged, this type of pavement is very easy to construct with much higher durability than Mastic Asphalt surfaced bituminous pavement; and the serviceability is expected to be same as conventional rigid pavement at much lower cost.

4. This technology can emerge as a good long-term solution to the perpetual maintenance problem of the roads with poor drainage.

Gratitude : The authors gratefully acknowledge the permission granted by Sri Bibek Raha, the then Chief Engineer, PWD, Govt. of West Bengal, to execute this new technology for the first time on a problematic road of West Bengal.

TeCHNICal PaPeRs


ReFeReNCes1. IRC:SP:76-2008. “Tentative Guidelines for Conventional,

Thin and Ultra - Thin White topping”, Indian Roads Congress, New Delhi.

2. Jundhare, D.R. et al. (March 2011). “Edge Stresses and Deflections of Unbonded Conventional White Topping Overlay”, The Indian Concrete Journal, pp. 35-44.

3. Rasmussen, R.O. and Rozycki, D.K. (2004). “Thin and Ultra-Thin White Topping-A Synthesis of Highway Practice.” NCHRP Synthesis 338, Transportation research Board, Washington, D.C.

4. Speakman, J. and Scott III, H.N.(1996) “Ultra-Thin, Fiber- Reinforced Concrete Overlays for Urban Intersections,” Transportation Research Record 1532, Transportation Research Board, National Research Council, Washington, D.C., pp. 15–20.

5. Armaghani, J. and Tu, D.(1999) “Rehabilitation of Ellaville Weigh Station with Ultra-Thin Whitetopping,” Transportation Research Record 1654, Transportation Research Board, National Research Council, Washington, D.C., pp. 3–11.

6. Cole, L.W. and Mohsen, J.P.(1993) “Ultrathin Concrete Overlays on Asphalt,” Presented at the TAC Annual Conference, Ottawa, ON, Canada.

7. Lahue, S.P. and Risser, R.J. (Sep. 1992) Jr., “Ultra-Thin Concrete Overlay Supports Truck Loads,” Concrete Construction, Sep. 1992, pp. 664–667.

8. Cole, L.W.(1997), “Pavement Condition Surveys of Ultra-Thin White Topping Projects,” Proceedings of the 6th International Purdue Conference on Concrete Pavement Design and Materials for High Performance, Vol. 2, Indianapolis, Ind., pp. 175–187.

9. Wen, Haifang., Li, Xiaojun. and Martono, Wilfung. (2010) “Performance Assessment of Wisconsin’s White Topping and Ultra-thin White Topping Projects.” WHRP, Transportation Research Board.

10. IRC:SP:62-2013 (Revised Draft Approved by Council),’ Guidelines for Design and Construction of Cement Concrete Pavement for Rural Roads’, Indian Roads Congress, New Delhi.

11. IRC:SP:62-2004, ‘Guidelines for Design and Construction of Cement Concrete Pavement for Rural Roads’, Indian Roads Congress, New Delhi.


DIsTResses IN CoNCReTe PaVemeNT : a Case sTuDy oF INDoRe byePassVandana Tare* and Akanksha Asati**

* Professor, E-mail: [email protected]** M.E. Student, E-mail: [email protected]

1 INTRoDuCTIoN

India is a developing country and succeeding towards developed nation. To achieve this goal infrastructure development is necessary. Development of highways and expressways are the part of it. For construction of high performance pavement, modern equipments and technologies are needed. Since last decade, India also adopts these techniques for construction of concrete pavement in highways e.g. Delhi-Mathura road, Mumbai-Pune expressway and Indore byepass etc. Concrete pavements have a relatively long service life (30-40 year), provided these are properly designed, constructed and maintained. With mega projects like National Highway Development Project (NHDP) and Pradhan Mantri Gram Sadak Yojana (PMGSY) the pace of concrete pavement construction has increase recently. This is because concrete pavement known to perform better with minimum maintenance. For the better performance many precautions should be taken during construction of concrete pavement. If it is not done properly then the problem of uncontrolled cracking of concrete slab arises. In most of the panels various types of distresses are observed at Indore byepass. A distress survey was carried out over the entire byepass to assess type and extent of distresses.

Distresses observed at Indore byepass are longitudinal cracks, transverse cracks, corner cracks, spalling at cracks/joints, shoulder drop-off and shrinkage cracks. It is also observed that some of the cracks are extended upto partial and full depth of the pavement.

This paper includes the study of various types of cracks which occurred at Indore byepass, to find out their cause and to suggest remedial measures and techniques for crack repair.

2 abouT INDoRe byePass

Indore byepass is a part of National Highway-3. NH-3 is one of the most important arterial routes of India. It connects the historical, tourist town of Agra to Mumbai, the hub of India. The total length of NH-3 from Agra to Mumbai is 1208 km.

Indore byepass is a completely new alignment around Indore city constructed under World Bank aided project in the year 2001 and has rigid unreinforced concrete pavement in both carriageways from Manglia to Rau. The byepass is four lane carriageways with 10 m wide unpaved median and 1.5 m wide paved bituminous shoulder at outer side. Fig. 1 shows location of byepass and Fig. 2 shows the crust composition below the cement concrete pavement surface.

The salient features of Indore byepass are as under:

● Land width = 60 m

● Carriageways width = 2 nos. –8 m each

● Size of panels at outer lane =3.5 m x 5m at inner lane = 4.5 m x 5 m

● Formation width = 2 nos. –10.5 m each

● Median width = 10 m

● Road length = 32.20 km

● Adjacent land use: forms on both sides.

Crust Composition-

● Selected soil = 500 mm

● Granular subbase = 150 mm

● Cement concrete (DLC) M-10 =125 mm

● Cement concrete (PQC) M-40 = 340 mm

Transportation Engg., Department of Civil Engg. and Applied Mechanics, SGSITS, Indore (M.P.).

TeCHNICal PaPeRs


Fig. 1 Road Map of Indore City with Indore Byepass

Fig. 2 Crust Details of Indore Byepass

3 lITReTuRe ReVIeW

Binod Kumar et.al.(2011) described the typical distresses that have occurred on Delhi-Mathura NH-2 due to different causes and gave their remedial measures by full depth and partial depth repairs techniques. R.K. Jain and Sangal(2004) described different types of defects, such as minor cracks (width less than 0.5 mm), medium cracks (width 0.5 mm -

1.5 mm) and wide cracks (width more than 1.5 mm). These cracks should be repaired by re-sealing and full/partial depth repair and replacement of complete slabs. Satander Kumar(2004) described the types of defects in cement concrete pavement, in which it is found out that defects may occurred due to design deficiency, improper joint spacing, over loading or environment factors etc. Regarding the remedial measure, all major patches may preferably be full depths, skewed and dowelled and sometimes by repair material like epoxy, hot sulpher, high performance concrete, conventional concrete etc. Binod Kumar et.al.(2004) dealt with early cracks like random or uncontrolled cracks in concrete pavement. It affects the structural integrity and ride ability of the concrete pavement. To regain these problems of the concrete pavement, CPR (Concrete Pavement Restoration) techniques are used.

4 DIsTResses INDeNTIFICaTIoN

Distresses in concrete pavements are either structural or functional. Structural distresses primarily affect the pavement’s ability to carry traffic load. Functional distresses mainly affect the riding quality and safety of the traffic.

The following distresses have been observed at Indore byepass:–

4.1 longitudinal Cracks

Fig. 3 shows longitudinal crack in middle of the inner lane at Indore byepass. Longitudinal cracking is cracking that runs predominantly parallel to the centerline of the pavement near the longitudinal joint, at mid-slab locations, or in the wheel paths. At Indore byepass, most of the cracks are longitudinal, occurred near the median lane. Liquid limit and plastic limit tests were conducted on the different layers of soil sub-grade and median’s material. Results show that the plasticity index of those materials is greater than 6 affecting the drainage of underlying layers. Due to permeable 10 m wide median, water is percolated

TeCHNICal PaPeRs


under the layer and partial removal of the soil support occurred. Due to this bending stresses developed causing longitudinal cracks at the surface of the pavement. The infuriation of water, which is caused loss of support and damage, may be prevented by providing lining median with impervious material. Indore byepass is also cracked due to improperly treated black cotton swelling soils with high plasticity index and inadequate compaction of the sub-base. Swelling soils will cause upward movement of the pavement. This upward movement will create tensile bending stresses throughout the thickness of the slabs, therefore, the longitudinal cracks initiated on the top surface. Due to heavy traffic load these longitudinal cracks are stretched upto full depth of PQC like tearing of cloth. Moreover, longitudinal joints are provided at single lane that is 3.5 m as specified but here longitudinal joints are provided at 4.5 m which is more than specified, and added the problem of longitudinal cracking.

Fig. 3 Longitudinal Crack

There are various severity levels according to size that is width, length and depth of cracks. Narrow and medium cracks should be remedied by means of a stitched crack repair. Wide cracks should be remedied either by a longitudinal full depth repair or by means of a bay replacement repair.

Table 1 shows the severity level, of longitudinal cracks, number of cracked panels at Indore byepass and suggested repair techniques.

Table No. 1 Repair Techniques for Cracked Panels by longitudinal Cracks

severity level

Description No. of Cracked Panels

suggested Repair

Technique

Low Crack widths < 3 mm, no

spalling, and no measurable

faulting; or well sealed.

110 Seal with epoxy

Moderate Crack widths 3 mm -13 mm, or with spalling

< 75 mm, or faulting up to

13 mm.

300 Stitched crack repair

High Crack widths > = 13 mm, or with spalling > 75 mm, or faulting upto

13 mm

850 Full depth repair or bay replacement

repair

4.2 Transverse Cracks

Transverse cracks are cracks that run perpendicular to center line, resulting in a panel that is broken into two or more pieces. Transverse cracking is a common type of structural distress in concrete pavements.

Fig. 4 shows the transverse crack at outer lane of the byepass caused by poor joint load transfer, non-uniform base support, excessive slab curling, inadequate sawing and excessive spacing between joints (distance between joints is 5 m while specified is 4.5 m). Propagation and widening of transverse cracks was occurred due to traffic loading and environmental changes, leading sometimes to spalling and faulting.

There are also various severity levels according to size of cracks. Narrow and medium cracks should be remedied by joint or crack sealing repair. Wide cracks should be remedied either by a transverse full depth repair or by means of a bay replacement repair.

TeCHNICal PaPeRs


Fig. 4 Transverse Crack

Table 2 gives the severity level of transverse cracks, number of cracked panels and their remedial measures by which these cracks can be minimized.

Table No. 2 Repair Techniques for Cracked Panels by Transverse Cracks

severity level


suggested Repair

TechniqueLow Crack widths < 3 mm,

no spalling, and no measurable faulting; or well-sealed

40 Seal with epoxy

Moderate Crack widths 3 mm-6 mm, or with spalling < 75 mm, or faulting up to 6 mm

30 Crack and joint seal/fill or Partial/

full depth slab repair

High Crack widths > 6 mm, or with spalling > 75 mm, or faulting > 6 mm

50 Thin hot mix overlay or Full bay

reconstruction

4.3 Corner Cracks

Fig. 5 shows corner crack in a slab. Corner crack is a portion of the slab separated by a crack which intersects the adjacent transverse and longitudinal joints, describing approximately 45° angle with the direction of traffic. They range in length from 0.3 m to half the total slab length or width. They may begin as hairline cracks and some corner cracks extend upto the full depth of the slab. At site some corner cracks are also observed. The reason for this is due

to repeated loading on the slab corner, combined with a lack of sub-grade support. In this, water entered the lower layers through the transverse joint or the poorly maintained shoulder joint. Partial or full depth concrete patching or full depth joint replacement may be necessary when corner cracking is extensive.

Fig. 5 Corner Cracks

Table 3 shows details about the corner crack at Indore byepass.

Table No. 3 Repair Techniques for Cracked Panels by Corner Cracks.

severity level


suggested Repair

Technique

Low Crack is not spalled < 10% of length; no faulting; and corner piece is not broken into two and has no loss of material.

20 Seal with low viscosity

epoxy

Moderate Crack is spalled > 10% of its total length, faulting of crack or joint is < 13 mm and corner piece is not broken into two or more pieces.

50 Partial depth repair

High Crack is spalled at moderate to high severity for more than 10% of its total length, faulting of crack or joint is > = 13 mm, corner piece is broken into two or more pieces.

90 Full depth repair

TeCHNICal PaPeRs


4.4 spalling

Fig. 6 shows the spalling at joints and cracks. Joint spalling is the chipping of the concrete slab at the edges of longitudinal or transverse joints. Mostly, spalling of cracks and joints are occurred at the site along the whole length due to a combination of traffic action and thermal expansion. The riding quality of the pavement is reduced due to joint spalling.

Fig. 6 Spalling

Spalling less than 75 mm from the crack face should generally be repaired with a partial-depth patch. If it’s greater than 75 mm from the crack face should be repaired with a full-depth patch.

Table 4 shows number of cracked panels with spalling at joints and cracks and repair methods according to their severity levels.

Table No. 4 Repair Techniques for Cracked Panels by spalling

severity level


suggested Repair

TechniqueLow Spalls < 75 mm wide,

measured to the face of the joint, with no loss of material and no patching.

85 Partial-depth patch

Moderate Spalls 75 mm to 150 mm wide, measured to the face of the joint, with loss of material.

300 Full-depth patch

High Spalls > 150 mm wide measured to the face of the joint, with loss of material or are broken into two or more pieces.

200 Full-depth patch

4.5 shoulder Drop off

Fig. 7 shows the shoulder drop-off at the site. Shoulder drop-offs are normally associated with the use of a flexible shoulder on pavements where the main lanes are of concrete. Large differences in elevation of the main lanes versus the shoulder can be safety problems and should be addressed by maintenance forces. Frequently, the edge/shoulder joint has deteriorated and is now open. Water entered in this joint and it is the cause of erosion of the layers supporting the concrete. This is a gradual deterioration process.

To temporary repair, joint/crack sealing repair is used and patching with asphalt concrete can be done to level the edge shoulder drop off, if traffic is not driving over this joint. But in severe cases the existing shoulder will need to be completely removed, if traffic will drive over the shoulder joint.

Fig. 7 Shoulder Drop off

4.6 Plastic shrinkage Crack

Fig. 8 shows plastic shrinkage crack at Indore byepass. In some slabs, plastic shrinkage cracks are also found. These cracks are like a hairline cracks formed during PCC setting and curing. Usually, they do not extend through the entire depth of the slab. Shrinkage cracks are considered a distress if they occur in an uncontrolled manner. Reasons for these cracks are that construction might have taken place in hot weather and contraction joints sawed too late.

TeCHNICal PaPeRs


Fig. 8 Plastic Shrinkage Crack

To minimize the shrinkage cracking loss of water from the surface of the concrete must be minimized. One option is to moist cure the surface of the concrete immediately after placement, and to continue to do so for at least 24 hours. The most effective method is to keep the surface of the pavement wet. Other options are to erect wind barriers around the pavement or to erect sunshades to protect the surface from heat. In mild to moderate severity situations, the shrinkage cracks can be sealed and the slab should perform adequately. In severe situations, the entire slab may need replacement.

4.7 Joint seal Damage

Fig. 9 shows loss of joint sealant material from the joints. Joint seal damage enables incompressible material or a significant amount of water to infiltrate the joint from the surface. At Indore byepass, transverse as well as longitudinal seal damage is observed. This is occurred due to improper joint construction and high thermal expansion of the joint. Traffic actions pull the sealant from the joint. Due to this, cracks are reached upto different severity level more or less than 25% of total length of joint. These joints are repaired by resealing of joint which is short term repair technique to improve the service life of the pavement.

Fig. 9 Joint Seal Damage

5 CoNClusIoNs aND ReCommeNDaTIoNs

1) At Indore byepass longitudinal cracks, transverse cracks, corner cracks, spalling at joints/cracks, lane to shoulder drop-off and plastic shrinkage cracks are observed.

2) Causes of failure of Indore byepass are– ● Loss of sub-grade support. ● Layer’s material having high

plasticity index. ● Improper construction of joint. ● Width of panel and spacing between

joint is greater than specified. ● Poor joint load transfer. ● Slab curling. 3) There is a wide range of concrete repair

and restoration techniques which can be used as corrective, preventive and corrective and preventive measure. Although concrete repair and restoration does not necessarily increase structural capacity of a pavement, it only does extend the pavement’s service life. At Indore byepass maximum cracks are extended upto partial or full depth of pavement, so, they should be repaired only with partial/full depth repair techniques.

TeCHNICal PaPeRs


If longitudinal and transverse cracks of width upto 13mm occurred then stitching and joint sealing materials can be used to repair respectively.

ReFeReNCes1. Jain, R.K. and Sangal, M.M. (2004), “Maintenance and

Repair of Concrete Pavement.” Seminar on Design, Construction and Maintenance of Cement Concrete Pavements, 8-10 October, New Delhi, pp IV-87 to IV-99.

2. Binod Kumar, Gupta, Saroj. and Sood V.K. (2004), “Restoration Techniques for Distressed Concrete Pavement.” Seminar on Design, Construction and Maintenance of Cement Concrete Pavements, 8-10 October, New Delhi, pp IV-101 to IV-110.

3. Binod Kumar, Sharma, S.D. and Mathur Renu (2011), “Distresses in Concrete Pavement: A Case Study of Delhi-Mathura National Highway.” Indian Highways, Vol. 39, No. 10.

4. Satander Kumar (2004), “Causes of Distressed, Repair and Maintenance of Concrete Roads.” Seminar on Design, Construction and Maintenance of Cement Concrete Pavements, 8-10 October, New Delhi, pp IV-1 to IV-14.

5. IRC:SP:83-2008, “Guidelines for Maintenance, Repair and Rehabilitation of Cement Concrete Pavements”, Indian Roads Congress, New Delhi.

6. William Yeadon Bellinger and Richard Ben Rogers (1993), “Distress Identification Manual for the Long-Term Pavement Performance Project”, Strategic Highway Research Program, National Research Council, Washington, p-338.


FeasIbIlITy oF PRoVIDING a sKyWalK FoR PeDesTRIaN IN CHaNDNI CHoWK, DelHI

PuRnima PaRida*, jiten Shah** and S. GanGoPadhyay***

* Head, Transportation Planning Division, CSIR-Central Road Research Institute, New Delhi, E-mail: [email protected]** Research Scholar, Civil Engineering Deptt., SVNIT Surat, E-mail: [email protected]*** Director, CSIR - Central Road Research Institute, New Delhi, E-mail: [email protected]

absTRaCTIn urban areas a significant proportion of trips up to 1-2 kms in length are performed on foot. Moreover, every journey necessarily starts and ends as a walk trip. Little attention has been devoted to study the pedestrian behavior, flow characteristics, etc. and to use such data in the integrated design of transport infrastructure with due consideration to walking as a mode of transport. Provision of grade separated facilities will ensure the movement of pedestrian safe, comfortable and also reduces the travel time. A study was conducted in Chandni Chowk area of Delhi to assess the feasibility of providing a grade separated facility (skywalk) for a distance of 1305 meters. The study results are presented in this paper.

1 INTRoDuCTIoN

Walking is still a major mode of transport in cities of India. The pedestrian traffic has large share in metro as well as in mid size cities of India despite of fast growing number of vehicle. However there is a negligence towards study of pedestrian behavior, flow characteristics, capacity of pedestrian facilities etc. and use of such data in the integrated design of transport infrastructure with due consideration to walking as a mode of transport. Due to inadequate facilities provided for the pedestrian movement, there exists a constant conflict between pedestrian and motor vehicles in sharing the limited space of the road, resulting in pedestrian being involved or being the cause of most of the road accidents.

2 NeeD oF PeDesTRIaN FaCIlITIes

The transportation studies conducted in various cities like Delhi, Mumbai, Surat, Bangalore, Jamshedpur etc. have revealed that exclusive walk trips constitute 30-40 percent of the total trips. Further, walk trips are considerable for journeys to education and shopping

especially when the distance of the trip is less than 1 km. In urban areas, people should be able to walk with reasonable comfort and safety, as walking is an essential part of a wide variety of activities. The freedom with which a person can walk about and look around is a guide to the civilized quality of an urban area. It is distressing that the pedestrian traffic has not received adequate attention in providing the facilities such as walkways, sidewalks and pedestrian crossings.

Due to the absence of appropriate facilities, pedestrian are left to find space for themselves on urban streets, and Delhi is no exception to this. The share of walk mode is decreasing over a period of time and to arrest this declining share is the need of the hour. Table 1 presents the decrease in share of walk trip in some Indian cities.

Declining share of walk trips, increasing number of road accidents where about 50% involve pedestrians necessitate the demand to study pedestrian flow characteristics in a scientific manner and incorporate the results in planning, designing and upgrading the various pedestrian transport infrastructure/facilities.

Table 1 Decrease in share of Walk Trip in some Cities

City year share (%) year share(%)

Bangalore 1984 44.00 2007 8.33

Chennai 2002 47.00 2008 22.00

Delhi 2002 39.00 2008 21.00

source : ADB Sustainable Development Working Paper Series, No.17, February 2011

TeCHNICal PaPeRs


3 obJeCTIVes oF THe sTuDy

A study was undertaken with a view to understand the pedestrian flow characteristics and its travel pattern in study stretch of Chandni Chowk area of Delhi exhibiting heavy pedestrian traffic with varying trip length, purpose of trip and socio economic diversity. On the basis of study, skywalk facility is proposed by evaluating its economic viability. The objectives of the study are as ● To Study the traffic characteristics of

main stretch of Chandni Chowk from Jain Temple to Fateh Puri. (1.3 Km).

● To study socio-economic characteristic & travel pattern of pedestrians.

● To examine the feasibility of a skywalk facility on the basis of magnitude of pedestrian problems.

● Economic analysis of the proposed skywalk facility.

4 sTuDy aRea

Chandni Chowk in Delhi, built by Shahjahan (Moghul Emperor), was the widest avenue (22 meter carriageway) with 3 meter of side walk on either side. This formed the commercial axis during Moghul times. The glory of yesteryears has been now reduced to totally congested corridor of mixed traffic. The carriageway is not fully available to vehicles because of encroachment and parked vehicles. Sidewalks are occupied by informal sector making pedestrian journey difficult as well as unsafe, as they are forced to share the carriageway with vehicles.

For the purpose of understanding the pedestrian movement, entire stretch of Chandni Chowk was divided into four sections namely Lal Quila to Nai Sarak, Nai Sarak to Fateh Puri and fro movement. Fig. 1 shows the study area and the survey locations. The five points selected for conducting surveys are Lal Quila, Guru Dwara, Metro Station, Nai Sarak and Fateh Puri Chowk.

Fig. 1 Layout of Chandni Chowk and Location of Survey Points

TeCHNICal PaPeRs


Based on pilot survey, five survey locations were selected in such a way that, it’s having own attraction and different characteristics in surrounding areas. Pedestrian volume count and questionnaire based survey were carried out to determine the pedestrian flow characteristic. Total 315 pedestrians were

randomly selected for interview. The pedestrian were asked about their socio-economic profile, travel pattern and also willingness to use the proposed skywalk facility. A physical inventory was also done to ascertain the availability of space to construct a skywalk (Fig. 2).

Fig. 2 Existing Road Layout and Cross Section of Study Stretch

A pedestrian travel time survey was also conducted to observe the time taken in traversing the distance between origin and destination under free flow and congested state conditions. From the pilot survey, it was observed that, friction to pedestrian flow was significant throughout the entire stretch of Chandni Chawk i.e From Lal Quila to Fatehpuri Mosque, hence for the proposed skywalk facility, the entire stretch of 1.3 km was considered for design purpose.

5 suRVey ResulTs

5.1 VehicularTrafficVolume

The bidirectional traffic volume of 65,532 vehicles was observed for the duration of 10 hour (10:00 AM to 8:00 PM) on a normal working day and composition is as shown in Fig. 3. It can be seen from the figure that about 33% two wheelers, 25% cycle rickshaw (passenger and goods), 21% 3-wheelers, 19% cars and 2% minibus constitute the vehicle composition.

Fig. 3 Vehicle Composition

5.2 Hourly Pedestrian Flow

The hourly variation of pedestrian flow for same duration of 10 hour passing through study stretch is presented in Figs. 4 & 5. Total number of pedestrians from Lal Quila to Nai Sarak was 33,627 and from Nai Sarak to Lal Quila was 31,519. Evening peak flow is found more prominent than the morning peak flow in both directions, the highest peak flow of 3,962 was observed between 4:00 PM to 5:00 PM.

TeCHNICal PaPeRs


Fig. 4 Hourly Pedestrian Variations at Lal Quila

Fig. 5 Hourly Pedestrian Variations at Fatehpuri

From Fatehpuri Chowk pedestrian move towards Sadar Bazaar, Khari Baoli, and Katra Bariyan road and at this section movement of pedestrian is also high. The total number of pedestrian was observed to be 55,054 per day (10 hr) in both the direction, the pedestrian flow from Fatehpuri to Nai Sarak was 28,026 and from Nai Sarak to Fatehpuri was 27,028. Evening peak flow is more prominent in this section and observed highest peak hour volume of 3,395 during 6:00 PM to 7:00 PM (Fig. 5). The section wise pedestrian volume as well as peak hour volume is given in Table 2.

Table2PedestrianVolumeatIdentifiedSections

location Total Volume Peak Hour Volume

Lal Quila to Nai Sarak 33627 3746Nai Sarak to Fatehpuri 27028 3395Fatehpuri to Nai Sarak 28026 3379Nai Sarak to Lal Quila 31519 3962

5.3 CompositionofTraffic

The composition of traffic observed on this road as shown in Fig. 6 indicate that the share of walk mode is very high compared to other motorized and non motorized modes. The percentage contribution to walk mode in total modal composition remains highest throughout the day. The share of trips by walk is 47% followed by Cycle Rickshaws 13% , minibus 11%, 3-W 11%, two wheelers 10% and cars only 8%.

Fig. 6 Share of Person Trip by Various Modes

5.4 analysis of Pedestrian Response

The pedestrian walking on the study stretch were asked a set of questions to understand their socio economic characteristics, trip purpose, frequency and willingness to use the proposed skywalk facility. The needs and perception varies with age and gender, keeping this in view, care was taken to select respondents to represent these parameters. Out of total sample size 64 percent males and 36 percent females were interviewed with age ranges from 15 to >50 years. In that 36% of the respondents were from 31- 40 years, 33% of 21-30 years of age, 14% in 41-50 years of age, and 17% more than 50 years.

The profession wise distribution of pedestrian shows that 35% of total pedestrians are in private service, 24 percent students, 18% House wife, 15% in own business and other 8% from government sector and retired person. The Figs. 7 & 8 shows Age and Profession wise distribution of pedestrian.

TeCHNICal PaPeRs


Fig. 7 Age Wise Distribution

Fig. 8 Profession Wise Distribution

There can be many purposes for a walk trip. But for the sake of objectivity they are broadly classified in to work, education, shopping, home, recreation etc. Chandni Chowk being a wholesale market, the trip purpose for shopping is predominant. Distribution of the trip for shopping, work, recreation, home, educational, and other are 48%, 21%, 19%, 6%, 3%, and 3% respectively.

The query pertaining to frequency of walk trip revealed that occasional walk trips on the study stretch were 40% of the total trips, 31% of the trip were weekly and 21% daily, while 8% of the trips were either many times a day or alternate day. In Chandni Chowk area due to heavy vehicular traffic and encroachment on available sidewalk facility it becomes very difficult

to walk.Skywalk was a welcome alternative to 91 percent of pedestrians. Figs. 9 & 10 shows the trip purpose wise distribution and distribution regarding willingness to use proposed skywalk.

Fig. 9 Trip Purpose wise Distribution of Pedestrian

Fig. 10 Distribution of Pedestrian by Willingness to use Skywalk

5.5 origin and Destination Pattern

The Origin–Destination pattern of pedestrian formed the basis for estimating the number of pedestrian who will be using the proposed skywalk facility. The study area was divided into 12 zones for ascertaining the O-D pattern of pedestrian. Fig. 11 shows the travel pattern of pedestrian in study area. About 1,20,200 walking trips/day were performed on typical working day on study corridor.

● Majority of the trips were concentrated between zones 1, 2, 3, 5, 9, 11 and 12 which are in close proximity of the study area.

TeCHNICal PaPeRs


● Most of the pedestrian trips are in range of 4000 to 10000 as observed from Lal Quila to Fateh Puri and Fateh Puri to Metro station and vice versa .The demand to access grade separated structure at metro station is taken into account.

● The survey results showed that pedestrian were willing to use grade separated structure constructed parallel to the road, with descending and ascending facilities in between.

Fig. 11 Desired Line Diagram for Pedestrian Trips

5.6 analysis of Pedestrian Flow Parameters

Pedestrian travel time study conducted during peak and off-peak hours at the stretch starting from Lal Quila to Fatehpuri Chowk revealed that it takes 24.00 minutes to cover a distance of 1.3 kilometers during morning peak hour and 26.30 minutes during afternoon and 30.34 minutes during evening peak hours. It is apparent from the survey results that throughout the

day there is not much change in travel time. Pedestrian free speed was carried out at study corridor and other two identified locations for a length of 1.3 km equal to the study stretch distance namely Kalka Mode and Ashram Chowk having same characteristics, the purpose being to have average pedestrian walking speed under freeflow conditions. Table 3 and 4 shows travel time and speed of pedestrian at identified

TeCHNICal PaPeRs


locations and delays experienced by pedestrian on study corridor. In free flow conditions, a pedestrian take 1091 seconds to cover a distance of 1.3 km, but in congested state conditions it takes 1591 seconds to travel the same distance. In a distance of 1.3 km, a delay of 500 seconds (8.33 minutes) is experienced by the pedestrian, which is a time loss to them.

Table 3 Travel Time in Free Flow Conditions

location Time Taken (sec)

speed (m/sec)

Chandni Chowk 1098 1.18Kalka Mode 1109 1.17Ashram Chowk 1065 1.23average Walking Time & speed 1091 1.19

Table 4 Travel Time in Congested state Condition

Time length in Km.

starting Time

end Time (min)

Journey Time (sec)

Journey speed (m/sec)

10:00 – 1:00 1.3 0 .0 24:00 1440 0.9012:00 – 1:00 1.3 0.0 26:30 1590 0.823:00 – 4:00 1.3 0.0 25:00 1500 0.875:00 – 6:00 1.3 0.0 30:34 1834 0.71

average 0:00 26:31 1591 0.82

6 eCoNomIC eValuaTIoN oF sKyWalK

For the purpose of economic evaluation cost benefit analysis has been carried out and Economic Internal Rate of Return (EIRR) has been calculated in order to assess the economic viability of the project. In this analysis components of cost and benefits are; Cost of construction and maintenance of Skywalk and benefits accrued due to saving in time cost to pedestrians. Though there are other tangible benefits also like reduction in delays to vehicles (as pedestrians tend to share the main carriageway), which will save fuel loss and time loss to the occupants of the vehicles.

For estimating the construction cost, the width of the skywalk was ascertained on the basis of available codes. To estimate the width of skywalk, the peak hour pedestrian flow was taken. The average annual exponential growth rate of population of Delhi during 2001-2011 has been recorded as 1.92%. (Source: Statistical Abstract of Delhi, 2012: Directorate of Economics and Statistics) for a period of 20 years was taken for pedestrian. It was estimated that in the horizon year 2029 the peak hour flow of pedestrian will be 5887 (Table 5). As per the “Guidelines for Pedestrian Facilities” IRC:103-2012 and Highway

Capacity Manual (HCM) for high intensity commercial areas 4 meter width of side walk is required and same width for pedestrian volume in horizon year. This width corresponds to level of service C.

Table 5 Recommended Value for Various Width of sidewalk

year Peak hour Flow rate

Desirable width (in meter)IRC HCm

2009 3962 3.5 2.52014 4374 3.5 2.52019 4830 4.0 2.52024 5332 4.0 3.02029 5887 4.0 4.0

source : IRC:103-2012 & HCM (2010)

6.1 Project Cost

The capital cost of the project consists of cost incurred during the construction period. The total estimated construction cost will be Rs. 13.63 crores considering three intermediate stairways. Based on O-D survey the access from FOB is given at maximum distance of 300m at Guru Dwara, Metro station and Nai Sarak

TeCHNICal PaPeRs


and its ascending/descending through 1.8 m wide stairway (minimum width of stairway as per Ministry of Railways, India) on the sidewalk. The cost of sky walk was worked out as per cost of construction of Skywalk at Bandra Mumbai. It is assumed that the proposed Skywalk will be operational in the year 2011. From the year 2011 onwards, maintenance costs will be incurred. Maintenance costs are recurring costs, comprising of routine and periodic maintenance components. The routine maintenance, involving day to day repairs (1% of cost of construction) and periodic maintenance, aimed at maintenance of structure periodically (2.5% of cost of construction), condition of the structure components is proposed to be undertaken in years 2015, 2020 and 2025 i.e. after every five years of opening of traffic.

6.2 ProjectBenefits

The project benefit is worked out mainly from the travel time saving to the pedestrians and it will be reduced by means of provision of new pedestrian facility. For consideration of project benefit per capita income was used to convert time saving by pedestrian traversing a distance of 1.3 km into monetary terms. The time saving of user is worked out from the pedestrian travel time survey results. Average time

saved for pedestrian was worked out to be 8.33 minutes (0.139 hr). Per capita income of pedestrian is obtained from the Economic Survey of Delhi, 2007-2008 as Rs. 77,405/- per annum at constant prices, and in view of that, time saving per pedestrian per day is Rs. 2.94/- by considering per capita income with average 10 normal working hours per day, i.e. [{77405/ (365*10) =21.20}*0.139= Rs. 2.94 Rs./day].

6.3 economic analysis

The results of direct interview showed that about 91% of the respondents were willing to use the grade separated facility. But with a view not to exaggerate the benefits, three scenarios have been taken, considering variation in pedestrian volume using the proposed skywalk facility and also with variation in time saving i.e. Scenario-1 considered 70%, Scenario-2 with 50% and Scenaro-3 with 30% of the total pedestrian volume with average time saving of 8.33, 6 and 5 minutes respectively. These pedestrian are those who make direct trip to the destination having weekly, daily and frequent trip on this corridor. Table 6 presents annual time savings for pedestrians considering different scenario in view of variation in pedestrian volume and saving in travel time.

Table 6 annual Time saving by Pedestrian Considering Different scenario

scenarios Considered % of Total Pedestrian

Volume

average Time

saving in minutes

average Time saving

(hr/day)

Per Capita Income (Rs/

annum)

Working hrs. Per Day

Hourly Per Capita

Income (Rs)

Value of Time

saving/Ped/Day (Rs)

Total Time saving

Per Day (Rs)

annually Time saving

(Rs)

I 70% (84140) 8.33 0.139 77415 10 21 2.94 247646 90390957

II 50% (60100) 6 0.100 77405 10 21 2.12 127412 46505380

II 30% (36060) 5 0.083 77405 10 21 1.77 63706 23252690

In various scenarios, cost of construction, annual and periodic maintenance cost is same. In scenario-I and II the total time saving is more than the scenario-III. Considering most pessimistic approach, scenario-III having 30% of total volume with lowest time saving of 5 minutes (0.083hr) and also considering lower growth rate of 0.5%. As shown in Table 6, considering per capita income is Rs. 21/- for 10 hr. and value

of time saving per pedestrian per day is Rs. 1.77/-. From above consideration, total time saving per day is 63,706/- and annual time saving is estimated to be 2,32,52,690/-, which gives least benefits than other two.

In scenario I, II and III the EIRR value obtained 66%, 33% and 12% respectively, shown in Table 7. The result of scenario-III, the year wise projected

TeCHNICal PaPeRs


pedestrian traffic which is going to be benefited in term of saving in time is presented in Table 8. The EIRR value of 12% was obtained in this case; the benefits were squeezed both in terms of number of pedestrian using the skywalk and saving in travel time. This is equal to the required rate of 12% prescribed by the World Bank and other funding agencies. Thus with this consideration the provision of pedestrian Skywalk is economically justified. The EIRR is estimated taking into account the only the

time savings to the pedestrian to walk a known distance without considering other tangible and non tangible benefits; like increased availability of carriageway widths for vehicular movements culminating in fuel savings, in vehicle time saving of occupants and safety for both vehicular and pedestrian traffic at intersection as well as mid block. If these benefits are also included into the benefit string then the economic rate of return would have been much higher.

Table 7 Comparative values of eIRR for Different scenarios

scenarios Considered % of Total Pedestrian Volume

average Time saving in minutes

average Time saving (hr/day)

Growth Rate

eIRR

I 70% (84140) 8.33 0.139 0.5 66%II 50% (60100) 6 0.100 0.5 33%III 30% (36060) 5 0.083 0.5 12%

Table 8 economic Viability analysis

year Time Cost Capital + maintenance cost

Cumulative cost

Cumulative Benefit

Benefit

2009 232526902010 23368953 136300000 136300000 -1363000002011 23485798 1363000 137663000 23485798 -1141772022012 23603227 1363000 139026000 47089025 -919369752013 23721243 1363000 140389000 70810269 -695787312014 23839850 1363000 141752000 94650118 -471018822015 23959049 3407500 145159500 118609167 -265503332016 24078844 1363000 146522500 142688011 -38344892017 24199238 1363000 147885500 166887249 190017492018 24320234 1363000 149248500 191207484 419589842019 24441836 1363000 150611500 215649320 650378202020 24564045 3407500 154019000 240213364 861943642021 24564045 1363000 155382000 264777409 1093954092022 24564045 1363000 156745000 289341454 1325964542023 24564045 1363000 158108000 313905499 1557974992024 24564045 1363000 159471000 338469544 1789985442025 24564045 3407500 162878500 363033588 2001550882026 24564045 1363000 164241500 387597633 2233561332027 24564045 1363000 165604500 412161678 2465571782028 24564045 1363000 166967500 436725723 2697582232029 24564045 1363000 168330500 461289768 292959268

EIRR = 12%

TeCHNICal PaPeRs


7 CoNClusIoNs

This study takes into account the time savings to the pedestrian and has successfully demonstrated that the projects for pedestrian are also economically viable. Besides tangible benefits there are other non tangible benefits also, like comfort safety and security which also has to be considered while providing any facility for pedestrian. The provision of skywalk facilities in congested areas will provide a safe and comfortable journey to the pedestrian.

In this study only the benefits accrued to pedestrians in terms of time savings has been taken into account. Due to encroachment on sidewalks and heavy bidirectional movements of pedestrians, the pedestrians tend to share the carriageway and cause delays to vehicles as well as its a safety hazard to them. By providing a skywalk facility the vehicular delays will be reduced and the pedestrians will be able to move safely. If the benefits in terms of, fuel savings, time savings to vehicles and its occupants is also taken into account the EIRR will be much higher.

The transport planning should be aimed at moving people and not vehicles. The pedestrian are human beings therefore the facilities for movement across

and along the carriageway should not be planned and provided the way they are done for vehicles. Besides ‘Level of Service’ the ‘Quality of Service’ is also an important aspect which is totally ignored.

ReFeReNCes1. Directorate of Economics and Statistics, (2012):

“Statistical Abstract of Delhi”.

2. Economic Survey of Delhi, 2007-2008.

3. Fruin, J. J., (1971), “Designing for Pedestrian A Level of Service Concept”, Highway Research Record No. 355.

4. Highway Capacity Manual (2000), Transportation Research Board, USA.

5. “Guidelines for Pedestrian Facilities”, IRC:103-2012, Indian Roads Congress.

6. Jiten Shah (2010), “Evaluation of Feasibility of Skywalk for Pedestrian in Chandni Chowk Delhi” unpublished M.Tech thesis, M.S. University of Baroda.

7. Ministry of Railways (2009). “Manual for Standards and Specifications for Railway Stations”. Railway Board, Government of India, Volume 1 of 2, 1-260.

8. Tanaboriboon, Y., and Gayano, J.A. (1991), “Analysis of Pedestrian Movements in Bangkok”, Transportation Research Record 1252, 52-56.

9. Website: www.mmrdamumbai.org

10. Website: [email protected]

obITuaRy

The Indian Roads Congress express their profound sorrow on the sad demise of Shri S.N. Pradhan, resident of Pradhan’s Cottage, 77, Convent Road, Darjeeling; Shri K.L. Kapoor, resident of H.No. 294, Sector-10-A, Chandigarh; Shri Inder Mohan, resident of 4/15, Raghubir Kunj, behind Deewani, Agra and Shri J.P. Das, resident of Neel Kamal, Bariatu Road, Ranchi. They were very active members of the Indian Roads Congress.

May their soul rest in peace.


assessING lIqueFaCTIoN PoTeNTIal by CPT/sHeaR WaVe VeloCITy meTHoD aND eValuaTING eFFeCT oF

lIqueFIeD soIl oN FouNDaTIoN DesIGNbhaRat KatKaR* and Poonam PendhaRi**

* Assistant Geotechnical Engineer** Senior Design Engineer, E-mail: [email protected]

absTRaCTThe assessment of liquefaction potential based on SPT ‘N’ values has been extensively used over the years. However, reliability of SPT data depends on numerous factors that are difficult to control in practice. In this paper, a time-tested approach of assessment of liquefaction potential based on more reliable CPT and Shear wave velocity method are presented with an exemplary study. Estimation of lateral spread of overlaying soil layer due to liquefaction occurring in deeper sand deposit, in context of structures lying nearby shoreline is also presented. Estimate of land displacement and its significance on bridge design is covered. Also effect of cementation of sands due to ageing process on estimation of liquefaction potential is explained. Appropriate correction factor for aged sands is referred and presented. For assessment of structural stability in event of liquefaction, a method to estimate forces exerted by liquefied sand on foundation is described. From extensive review of various standards of practice, reduction in strength parameters of soil during a dynamic event have also been presented. Various checks and estimates for soil strength in event of liquefaction are provided. Inculcation of these checks in common engineering practice will be fruitful as well as help in avoiding misdemeanor caused due to negligence of minor but significant issues in foundation design.

1 INTRoDuCTIoN

Bridges are strategically categorized as buildings of national importance. Urban needs have created a requirement to establish foundations on sites previously avoided by consultants, such as sites with liquefiable sand deposits. Rejection of such sites, by realigning to different locations or adopting deep foundations is sometimes impractical as well leads to loss of time and money (IS1893)(6) To solve this problem, it is a common practice to apply expertise of Geotechnical consultant, who by implementing various methods of ground improvement provides a solution, which helps in adapting the site for the purpose intended. Limited amount of field tests prove unsatisfactory to provide a clear mapping of soil strata for application of

geotechnical solutions. Hence, development of semi-empirical solution had made their way in practice.

To provide a reliable and time-tested solution, many practitioners and researchers have relied on empirical formulas derived from monitored input data obtained from various field tests and case studies. The allegiance of such empirical equations though provide an overtly safe solution, experience based on these findings have proven economical in application compared to the conventional solutions. Assessment of liquefaction potential is also based on such empirical correlation extracted from observed data. Though Indian standard codes prohibit establishing any important project on liquefiable land, many foreign codes allow for remediation as well as include forces exerted on the structure by liquefied soil layer. The American code published by National Co-operative Highway Research Program (NCHRP 12-49)(3) provide comprehensive specifications for seismic design of bridges and refers to use simplified process for liquefaction assessment discussed in NCEER workshop, provided in Youd et al. 2001(13). Eurocode(5) 8 part 5 has also provided assessment procedure for liquefaction potential on similar lines. TRANSIT(10) New Zealand’s bridge manual for design also prescribes use of procedure given in Youd et al. 2001(13) for liquefaction assessment. The Japanese code for design specifications for highways JRA(4) has went further and added consideration for quantification of loading on pile foundations due to liquefied soil deposit. The details of which are provided in subsequent sections.

The purpose of this paper is to provide insight into aspects related to the liquefaction phenomenon and demonstrate methods to estimate various outcomes caused and resulting from the same. Simple equations

Spectrum Techno Solution Private Limited, Navi-Mumbai

TeCHNICal PaPeRs


and procedures to evaluate the liquefaction potential are explained lucidly with due addition of up-to-date literature on the subject. In continuation of efforts to increase awareness in liquefaction potential assessment initiated by Singh and Bhowmick, 2012(9), the present paper focuses on adding latest developments in the field, with reference to various international codes of practise. Firstly, assessment of liquefaction potential is explained based on Cone Penetration Test data (CPT) and by Shear wave velocity method. A sample problem is demonstrated to encourage and make aware the practitioner on input parameters required for the assessment. Though CPT and Shear wave velocity methods are scarcely used in arena of consultants in India and are generally put to selective use for very important structures such as strategic bridges, nuclear power plants etc. Indian standard code IS 4968(7) and IS 13372(8) provide the details of conducting and recording Cone Penetration Test (CPT) and Shear wave velocity test. The basic parameter of Sleeve resistance (fs) and tip resistance (qcA) are obtained from CPT, whereas Cross-Hole Seismic Test (CHST) provides with the parameter of Shear wave velocity (Vs). CPT test utilises a standard equipment for driving the CPT probe, the tip and sleeve resistance readings are recorded via electronic equipment’s. For CHST the dynamic pulse generating charge is lowered at the required location in one borehole, whereas the receptor sensors (geophones) are placed at variable depths depending upon soil stratification and site limits. The data is recorded through a series of sophisticated electronic circuits.

The data obtained from CPT and Shear wave velocity test is more reliable in terms of repeatability than conventional Standard Penetration Test (SPT), though exact profiling is difficult to obtain. Attaining a core log from site turns out to be very helpful in assessment of gradation of particles in soil layers. Empirical equations concerning lateral spread of soil present over a liquefied sand layer are then explained to have an estimate of horizontal soil movement. This estimate is useful to define the lateral movement occurring in the land mass near shore or on banks of rivers that might

affect the alignment of abutments of bridge located within the proximity.

The issue of additional resistance offered by soil due to ageing of sand layers on penetration resistance value is addressed consequently which is based on penetration resistance value. The correction to be applied to account for ageing of soil is also explained in latter sections of paper. Indian geological stratification indicates formation of soil layers due to deposition and weathering of rocks formed by magmatic spread in prehistoric age. Seismically stable Indian peninsular region has vast deposits of layered sands and sand lenses present below surficial deposit of alluvium.

From structural stability point of view, it is necessary to check stability of foundation against liquefaction occurring in shallow liquefiable soil deposit. The magnitude of force due to liquefied soil deposit acting on foundation can be estimated. This force estimation will help bridge designer to make structure earthquake resistant as well as liquefaction resistant.

2 lIqueFaCTIoN assessmeNT

The term liquefaction should not be confused with other phenomenon like sand boil or quicksand because the forces triggering these situations are not related to short-term dynamic loading. Though liquefaction may yield such situation. The mechanisms by which these phenomenon occur are different and are slightly related to each other. Mostly they are associated with the hydraulic conditions created in soil sub-surface.

Effects of liquefaction include occurrence of excessive vertical settlements and lateral displacements of overlaying soil mass. The simplified liquefaction assessment procedure based on SPT, CPT and Shear wave velocity method only stand correct for depths less than or equal to 15 m from ground level, for evaluation of liquefaction resistance in greater depths, the simplified procedure provided here does not yield proper results (Youd et al., 2001(13)), for such situations the evaluation should be done by more reliable method available at discretion of Geotechnical Engineer. To make simplified procedure applicable to

TeCHNICal PaPeRs


depths greater than 15 m, some correction factors are suggested, which are elaborated in section 2.3.

2.1 assessment of liquefaction Potential based on Cone Penetration Test (CPT) (youd et al., 2001(13))(Depth≤15m)

Cone penetration test provides a continuous profile of stratigraphy and is more consistent with greater repeatability of test. It is assumed that index properties of soil layers are known or are established empirically. Normalized value of tip resistance along-with clean sand correction and thin soil layer correction is applied to dimensionless value of resistance, which is then employed to assess liquefaction potential. Following procedure from Youd et al., 2001(13), one can assess liquefaction potential of soil. Resistance to liquefaction is provided by Cyclic Resistance Ratio (CRR) of soil, whereas potential of liquefaction is created by the Cyclic Stress Ratio (CSR). The procedure to evaluate CSR is independent of method of soil testing.

step I : Evaluation of CSR

CSR ag

rv

vod= ×

×

×0 65 0. max σ

σ ... (2.1)

where,

amax = Peak horizontal acceleration at ground surface due to an dynamic event (m/s2) (refer IS 1893-Part 1 (2002), Table 2)(6)

g = Acceleration due to gravity (m/s2) бV0 = Total overburden pressure (kN/m2) б’V0 = Effective overburden pressure

(kN/m2) rd = Stress reduction coefficient

(Calculated as given below)

)001210.0006205.005729.04177.01()001753.004052.04113.01(

25.15.0

5.15.0

zzzzzzzrd +−+−

++−= ... (2.2)

where,

z = Depth below ground surface (m)

For evaluation of Cyclic Resistance Ratio (CRR), a 4-step procedure is provided below.

step II : Correction for thin soil layer in cone penetration resistance

This correction is applied to modify resistance to penetration of cone into the ground due to presence of thin layer of soil with same properties as that of present above or below itself, but of reduced stiffness caused as a result of change in gradation of soil. This correction is necessary to avoid apparent reduction in penetration resistance due to presence of thin soil layer. The correction is to be applied as follows,

K HdH

c=

×

−

+0 25

171 77 1 0. . . ... (2.3)

where,

KH = Factor for thin layer

dc = Diameter of cone used for CPT testing (mm)

H = Thickness of interbedded (Thin) layer (mm)

The correction can be applied as,

qc = KH × qCA

where,

qCA = Measured cone resistance in field test (kN/m2)

qC = Corrected cone resistance (kN/m2)

step III : Normalize cone penetration resistance

It was observed that penetration resistance is affected by the changes in relative density of soil, overburden stress, lateral stress and prior stress-strain history. In order to account for this effect the normalization for cone penetration resistance is required. The procedure to evaluate it is as follows,

=

a

cQNc P

qCq 1 ... (2.4)

Where,

C PaQ

v=

≤

σ '......................... .

01 7 ... (2.5)

TeCHNICal PaPeRs


And,

qc1N = Normalized cone penetration resistance,

CQ = Normalizing factor for cone penetration resistance (kN/m2), Pa = Atmospheric pressure (kN/m2) (1 atm), n = Exponent that varies with the soil type, б’V0 = Effective overburden pressure (kN/m2)

To arrive at appropriate value of ‘n’ in absence of soil data, following equations should be used or the value can empirically assumed as 0.5 for clean sand and 1.0 for clayey soil.

( ) ( )0.52 23.47 log 1.22 logIc Q F = − + +

... (2.6)

where,

Qq

PPc v

a

a

v=

−( )

σσ

0

0' ... (2.7)

F fq

s

c v=

−( )

×

σ 0100% ... (2.8)

And,

Ic = Soil behaviour type index

qc = Cone tip resistance corrected for thin layer (kN/m2)

fs = Cone sleeve resistance (kN/m2)

бV0 = Total overburden pressure (kN/m2)

For ease of evaluating equations 2.6, 2.7 and 2.8, Fig. 1 should be referred. Here Ic serves as a limiting value, which can be used to decide an appropriate value of ‘n’. If value of Ic is greater than 2.6, the soil is considered as non-liquefiable and for verification purpose the soil gradation should be carried out for that particular soil layer. For smaller values of Ic the value of ‘n’ is appropriately chosen from the procedure described in Fig. 1.

Fig. 1 Flowchart for Calculation of Correction for Normalized Cone Resistance

step IV : Evaluate Clean-sand Equivalent Normalized Cone Penetration Resistance

This correction is required to correct penetration value obtained in silty soil for that observed in pure sand with no silts i.e. clean sand. This correction is required to be included as the data points used to obtain these empirical equations were based on sites having stratum with no silt content. Hence, to gain equivalence, the following equations were suggested by Youd et al(13).

(qc/1N)cs = Kc × qc1n ... (2.9)

where,

qc1N = Normalized cone penetration resistance (kN/m2), Kc = Correction factor for grain characteristics as follows,

For Ic > 1.64 use Kc = 1.0

For Ic < 1.64 use

Kc = –0.403I4c + 5.581I3

c – 21.63I2c + 33.5Ic – 17.88 ...(2.10)

* The value of Ic used in equation (2.10) is obtained from the analysis of flowchart given in Fig. 1.

TeCHNICal PaPeRs


step V : Calculation of Cyclic Resistance Ratio (CRR)

Now to evaluate liquefaction resisting component (CRR), following equations are used and the CRR value is corrected for appropriate magnitude of earthquake. Value of (qc1N)cs is as obtained from step IV above.

If (qc1N)cs < 50, then

( )17.5 0.833 0.05

1000c N csq

CRR

= +

... (2.11)

If 50 < (qc1N)cs < 160, then

( ) 31

7.5 93 0.081000c N csq

CRR

= +

... (2.12)

Calculated value of CRR is based on the moment magnitude of earthquake equal to 7.5, to convert it to earthquake of different moment magnitude values of CRR7.5 should be multiplied with Magnitude Scaling Factor (MSF) to be chosen from Table 1,

CRRM = CRR7.5 × MSF ... (2.13)

Table 1 Values of magnitude scaling Factor

moment magnitude of earthquake (m)

magnitude scaling Factor (msF)

5.5 1.436.0 1.326.5 1.197.0 1.087.5 1.008.0 0.948.5 0.89

step VI : Computation of Factor of Safety (FOS)

The Factor of Safety (FOS) against liquefaction can be calculated as,

CRRCSRFOS = ... (2.14)

If FOS > 1, then soil is non-liquefiable.

For FOS < 1, remedial measures for mitigation of liquefaction are suggested to be referred.

Table 2 Typical Computation for evaluation of liquefaction Potential by Cone Penetration Test

Depth (m) 0 8 12 15 25 35 45

qcA (kN/m2) 0 1000 3500 2500 4000 10000 20000

fs (kN/m2) 0 20 35 25 40 100 200

qc (kN/m2) 0 1782 6240 4457 7131 17829 35659

γb (kN/m3) 0 18.0 17.5 18.5 19.0 20.0 20.0

бv0 (kN/m2) 0 144.0 214.0 269.5 459.5 659.5 859.5

б'v0 (kN/m2) 0 94.0 124.0 149.5 239.5 339.5 439.5

F - 1.22 0.58 0.60 0.60 0.58 0.57

n - 1 1 1 1 1 1

Ic - 2.58 2.04 2.25 2.26 2.02 1.85

CQ - 1.06 0.81 0.67 0.42 0.29 0.23

qc1N - 18.9 50.3 29.8 29.7 52.5 81.1

Kc - 1.0 1.0 1.0 1.0 1.0 1.00

(qc1n)cs - 18.9 50.3 29.8 29.7 52.5 81.1

CRR7.5 - 0.07 0.09 0.07 0.07 0.09 0.13

rd - 0.94 0.86 0.76 0.54 0.47 0.44

CSR - 0.22 0.23 0.21 0.16 0.14 0.13

FOS - 3.40 2.51 2.85 2.16 1.54 1.03

Result NL* NL NL NL NL NL

* NL = Non-liquefied layer

Note : Values are assumed as follows:

H = 3000 mm, dc = 36 mm, Pa = 100 kN/m2,

Depth of water table = 3 m,

Peak horizontal acceleration at ground surface = 2.35 m/s2

2.2 assessment of liquefaction Potential based on shear Wave Velocity method (youd et al., 2001(13))(Depth≤15m)

Evaluation of liquefaction potential using shear wave velocity method is important for sites where, penetration tests are difficult to conduct because of adverse site conditions such as hard soils, gravelly soils or inaccessible sites where boring is not possible. Procedure for liquefaction assessment is simple and is completed in two steps.

TeCHNICal PaPeRs


step 1 : Calculate Cyclic Resistance Ratio (CRR)

−

−+

= *

11*1

15.7

118.2100

022.0sss

s

VVVVCRR ... (2.15)

where,

Vs1 = Measured shear wave velocity (m/s)

V*s1 = Limiting value of shear wave velocity (m/s) as obtained from interpolation between the following,

V*s1 = 200 for FC ≥ 35%

And V*s1 = 215 for FC ≤ 5%

where,

FC = Fines content (%)

It can be observed that as fines content (FC) increase, the shear wave velocity through that soil layer decreases, hence to limit the lower bound value of wave velocity, the value of V*s1 beyond FC of 35% is taken equal to 200 m/s. The estimation of fines content should be applied with appropriate field tests.

step 2 : Correct CRR7.5 for Design moment magnitude as per equation 2.13 and Table 1 for CPT test.

step 3 : Calculate Factor of Safety (FOS) as given in equation 2.14.

Hence for problem explained in Table 2, if the measured shear wave velocity depth 12 m with fines content of 18 % was 335 m/s, then calculated CRR value from equation 2.15 would be 0.211, giving a factor of safety of 1.09.

Limitation of liquefaction potential evaluation by Shear wave velocity method is that, liquefaction is associated with large-strain shear modulus, whereas small-strain shear modulus is evaluated by current method. This discrepancy may lead misrepresentation of liquefaction assessment. Advantage associated with this test is that it provides with an estimate to evaluate cementation caused because of ageing of sands, as described in section 4.

2.3 assessment of liquefaction Potential for soil Depth Greater than 15 m (youd et al., 2001(13))

For depth greater than 15 m, the effect of overburden and static shear stress condition are required to be incorporated in formulation. It was observed that the value of CRR increases non-linearly with increase in overburden stress. To make the simplified procedure (explained above) applicable for greater depth of soil, following correction in CRR values were suggested. For static shear stress observed in sloping ground, there is a absence of convergence in observed data, hence it is not recommended to use it in routine engineering practice and is not covered here.

FOS CSRCRR

K= × σ ... (2.16)

where,

KP

v

a

f

σσ

=

−0

1( )

... (2.17)

And, б’v0 = Effective overburden pressure

(kN/m2) Pa = Atmospheric pressure (kN/m2) f = Exponent based on site condition,

given as,

Relative density (Rd) f40% to 60% 0.7 to 0.860% to 80% 0.6 to 0.7

* Values beyond this range of relative density in sand are not covered in present literature

2.4 Design Charts for estimation of liquefaction Potential (youd et al., 2001(13))

The equations and calculation process explained above in sections 2.1 to 2.3 culminate into the visual representation in the charts below. The corrected values of in-situ test results are plotted on the abscissa and the CRR values are place on the ordinate. Numerous case histories of documented earthquake events at various locations worldwide were analysed through a rigours statistics and were plotted on the chart to define a commonly agreed boundary of liquefiable and non-liquefiable sites. These boundaries are

TeCHNICal PaPeRs


mathematically represented in equations 2.11, 2.12 for CPT and equation 2.15 for Shear wave velocity method. The curve represent in the Chart 1 for CPT is strictly applicable for earthquake moment magnitude of 7.5. As well as Chart 2 for Shear wave velocity test is applicable for earthquake moment magnitude range of 5.9 to 8.3 only. The readers should be cautious about direct use of these charts; a sound consideration should always be given to the earthquake moment magnitude while using the chart.

Chart 1 Recommended Curve for Calculation of CRR from CPT Data. (After Youd et al., 2001(13))

Chart 2 Recommended Curve for Calculation of CRR from Shear Wave Velocity Test Data. (After Youd et al., 2001(13))

3 CoRReCTIoN FoR lIqueFaCTIoN PoTeNTIal FRom aGeD saND DePosITs

Sand deposits existing at a site are prone to undergo cementation effect due to wetting and drying cycles. Cementation in soil particles is created due to deposition of bonding material like lime, silica etc., by virtue of groundwater flow, or by chemical degradation in the soil itself, this bonding can also be caused because of presence of clay or silt particles in the soil skeleton. Cementation apparently affects resistance to penetration for tests such as SPT and CPT. Cementation in sand increases with its age, if no disturbance has occurred in the deposit due to any dynamic loading, hence older sand deposits provide an apparent resistance, to penetration tests. Time period to be calculated for evaluating cementation correction factor depends on when an dynamic event have occurred area, time is calculated from the time of last disturbance that have occurred in soil. It is assumed that ‘ageing clock’ of cemented sands is reset when a dynamic event occurs. It is normally difficult to estimate age of sand deposit at a given site. Hence to evaluate the age of deposit and have a correct estimate of liquefaction potential, a procedure was devised by Andrus et al. 2009(1), which correlates data of SPT, CPT tests to the measured shear wave velocity (Vs) at site. These correlations are used to evaluate the age of the soil and are re-produced below for application purpose. The procedure used is simple and requires at least one of the test data either SPT or CPT to be available alongwith shear wave velocity test (Vs) data for the same site. Here for advantage of the reader, both procedures for SPT as well as CPT are provided. The step-by-step procedure is as follows,

step 1 : Correct the values of SPT, CPT and Shear wave velocity for clean sand

1. For SPT :

(N1)60CS = α + β(N1)60 ... (3.1)

Where, α and β are the coefficients determined from Table 3, depending on the Fines Content (FC) in the soil. (N1)60 = Corrected SPT ‘N’ value for energy correction.

TeCHNICal PaPeRs


2. For CPT :

It is same as step (IV) and equation 2.9 in procedure for evaluation of liquefaction by CPT test.

3. For Shear wave velocity (Vs)

V K V Ps cs cs s

a

r1

0 25

( ) =

σ

. ... (3.2)

where,

Kcs is determined from Table 3

Vs = Shear wave velocity measured in field (m/s)

б’v0 = Effective overburden pressure (kN/m2)

Pa = Atmospheric pressure (kN/m2)

Table3CoefficientsforCleanSandCorrection

FC ≤ 5% 5% < FC< 35% FC ≥ 35%

α 0.0 2(1.76 190 / )FCe − 5.0

β 1.01.5(0.99 )

1000FC+ 1.2

Kcs 1.0 1+(FC-5)T* 1+30T** T = 0.009 - 0.0109 (Vs/100) + 0.0038 (Vs/100)2

step 2 : Calculate the estimated shear wave velocity based on SPT/CPT test

1. For SPT :

(Vs1)cs = 87.8 {(N1)60cs}0.253 ... (3.3)

2. For CPT :

(Vs1)cs = 62.6[(qt1N)cs]0.231 ... (3.4)

step 3 : Calculate the measure to estimated velocity ratio (MEVR)

Divide equation (3.2) by (3.3) or (3.4) and from Table 4, lookup the value of ‘t’ and Kdr.

step 4 : Calculate the corrected CRR value

CRRK = CRR X KDR ... (3.5)

Where, CRR = Cyclic resistance ratio as obtained from equation 2.13 and 2.15.

Table 4 Determination of Deposit Resistance Correction Factor

Time since initial deposition

or critical disturbance, t

(years)

measured to estimated shear

wave velocity ratio, (meVR)

Deposit resistance correction

factor, KDR

0.1 0.85 0.661 0.94 0.83

10 1.02 1.00100 1.10 1.17

1000 1.18 1.3410,000 1.26 1.51

1,00,000 1.34 1.6810,00,000 1.43 1.85

1,00,00,000 1.51 2.02

The reason to use shear wave velocity method is that it is more sensitive to the cementation caused due to ageing of sands as compared to SPT or CPT tests. Without applying corrections to resistance value, the calculated CRR values may involve higher degree of error.

4 assessmeNT oF eFFeCT oN FouNDaTIoNs Due To lIqueFaCTIoN

One of the important aspects for Liquefaction mitigation for bridge foundations is estimation of effects caused by liquefaction. Two prominent effects are felt on foundation due to liquefaction in sub-soil layer. First being the loss of shear strength of soil, whereas the other is movement of soil in lateral direction i.e. lateral spread of soil. Both of these effects are harmful for structural stability as well as to maintain serviceability of structure. Estimation of these parameters is of utmost importance for proper liquefaction-resistant design of structure. The tolerable limits for these effects should be evaluated in order to make the structure liquefaction-proof. There are various empirical relations given in literature, off-which recent ones with their method of application are discussed below.

4.1 estimation of lateral spread occurring Due to liquefaction

During and after occurrence of liquefaction in sub-soil layer, the stable land present above liquefied zone

TeCHNICal PaPeRs


is subject to lateral movements. This phenomenon can be visualized equivalent to movement of tectonic plates over the earth’s mantle, only the process being momentary and rate of displacement being very high. Generally effect of lateral spread can be prominently seen at banks of river or protruding edges of land into sea. This condition is termed as free face condition of land. Whereas in general, the land in which no-such condition exists is considered as sloping ground condition. This concept can be visualized from Figs. 2a-2b.

(a) Sloping Ground Condition

(b) Free Face Condition

Fig. 2 Schematic Diagram for Estimation of Lateral Spread of Soil. (After Yi. F., 2010(11))

4.1.1 For Sloping Ground

Sloping ground condition is as shown in Fig. 2a. To estimate displacement of ground for this condition, the following formula is referred from Youd et al. 2002(12) and other researchers(2)(11).

15 15

15

log 16.213 1.532 1.406log * 0.0120.388log 0.540log 3.413log(100 )0.795log( 50 0.1 )

HD M R RS T FD mm

= − + − −+ + + −− +

... (4.1)

And,

R* = R0 + R, R0 = 10(0.89M-5.64)

where,

DH = Displacement of land (m)

M = Moment magnitude of earthquake

R* = Modified source distance value (km)

R = Mapped distance from site to seismic energy source (km), S = Ground slope (%), T15 = Cumulative thickness of saturated granular layer (m), F15 = Average fines content in T15 (%),

D5015 = Average mean grain size of soil in T15(mm)

4.1.2 For free face condition

Free face condition can be visualized in Figure 2b. Estimation for such condition might be required for structures present near bank of river or near shoreline or in other words vertical face of land beyond which resistance to its lateral movement cannot be offered, can be termed as free face. To include the free face component into the formulae, a dimensionless ratio (W) defined as height of free face (H) to distance from base of free face to the site (L) expressed in precent is included in formula.

15 15

15

log 16.713 1.532 1.406log * 0.0120.592log 0.540log 3.413log(100 )0.795log( 50 0.1 )

HD M R RW T FD mm

= − + − −+ + + −− +

... (4.2)

where,

W = Free face ratio as defined earlier

TeCHNICal PaPeRs


The displacement values obtained from these equations can be used for checking the displacement criteria of structures. For proper understanding of above formulae, an example with assumed values is provided below,

M S (%)

R (km)

T15 (m)

F15 (%)

D5015 (mm)

L (m)

H (m)

DH (mm)

7.5 2 100 6 12 1 - - 24.60

7.5 - 100 6 12 1 400 6 6.77

4.2 estimation of Force acting on Foundation Due to liquefaction

To consider load acting on foundation due to liquefaction occurring in sub-soil layers, the procedure provided by Japanese code, Design specification for bridges, Part 5, 2002(4) is followed. Japanese code divides the category of earthquake based on the ground motion parameters, obtained from seismic response spectra, which further depend on modification factors based on the ground conditions give in seismic map of Japan.

According to JRA(4), the design earthquake ground motion is classified into two parts,

1) level 1 earthquake ground motion : Earthquake with high probability of occurrence.

2) level 2 earthquake ground motion : Earthquake with less probability of occurrence, but strong enough to cause critical damage to structure. With is further classification as Type 1: Interplate earthquake with large magnitude and Type 2: Inland near field type earthquake.

Evaluation needed to decide design earthquake motion is cumbersome and requires number of parameters, hence for sake of simplicity, procedures to arrive at earthquake ground motions is not described here and should be dealt with great care and accuracy.

Parameters used for estimation of forces on foundation require assessment of liquefaction potential by Japanese code. This procedure though synonymous to the method provided in Youd et. al., 2001(13) has

some difference, which necessitates that liquefaction assessment procedure to be explained separately.

4.2.1 Assessment of Liquefaction Potential by Japanese Code

Before conducting an assessment for liquefaction potential, Japanese code stipulates that following conditions should be satisfied in order to classify a soil as liquefiable,

i) The ground water level should be present above 10 m depth from ground surface and the saturated soil layer should be present at less than 20 m depth.

ii) For liquefaction assessment the Fines content (FC) in the soil should be less than 35% or for soils with higher FC the plasticity index (Ip) should be less than 15.

iii) The soil layer to be assessed for liquefaction should have mean grain diameter of less than 10 mm (i.e. D50 < 10mm) and D10 < 1 mm.

When above conditions are satisfied completely, then soil is categorized as liquefiable and assessment for liquefaction potential can be carried out as given in following steps below,

step 1 : Calculate the modified SPT ‘N’ value

Correction for overburden pressure

1 '170

70V

NN =σ +

... (4.3)

where,

N = Penetration value obtained from SPT test

б’V = effective overburden at that level (kN/m2)

Correction for grain size

i) For sandy soils

Na = c1N1 + c2 ... (4.4)

where, the values of c1 and c2 are obtained from Table 5 and N1 as obtained in step 1.

TeCHNICal PaPeRs


Table 5 Correction for Grain size

0 %≤ FC* < 10%

10% ≤ FC < 60% 60% ≤ FC 10% ≤ FC

c1 1 4050

FC + 120FC

−-

c2 0 - - 1018

FC −

* FC = Fine Content (Percentage of soil passing through 75 μm, by weight)

ii) For gravelly soil

5010 11 0.36log

2aDN N = −

... (4.5)

where,

D50 = Mean grain diameter (mm)

step 2 : Evaluate dynamic shear strength ratio (R)

R = cwRL ... (4.6)

where,

cw = Modification factor depending on earthquake ground motion and RL.

1.0.........................................................( 0.1)3.5 0.67.........................................(0.1 0.4)2.0.........................................................( 0.

L

w L L

L

Rc R R

R

≤= + < ≤

< 4)

... (4.7)

RL = Cyclic triaxial shear stress ratio to be obtained from SPT test. The values of which can be determined as follows,

( )4.56

0.0882 ........................................( 14)1.7

0.0882 1.6 10 14 .....(14 )1.7

aa

La

a a

N NR

N N N−

<

= + × − ≤

... (4.8)

Where value of Na is as calculate in step 2.

step 3 : Evaluate seismic shear stress ratio (L)

'v

d hgV

L r k σ=

σ ... (4.9)

where,

rd = Reduction factor for seismic shear stress ratio in terms of depth, as provided in equation 4.10.

khg = Design horizontal seismic coefficient at ground surface level,

бV0 = Total overburden pressure (kN/m2),

б’V0 = Effective overburden pressure (kN/m2)

The reduction factor can be calculated out as,

rd = 1.0 – 0.015x ... (4.10)

where,

x = Depth from ground surface (m)

step 4 : Calculate liquefaction resistance factor (FL)

LRFl = ... (4.11)

where,

R = Dynamic shear strength ratio

L = Seismic shear stress ratio

If value of Fl is greater than one then soil is termed as non-liquefiable.

After evaluation of liquefaction potential, if the soil is found to be liquefiable, then loading on bridge foundation due to liquefaction induced ground flow can be obtained as given section 4.2.2.

4.2.2 Assessment of Force Acting on Foundation Due to Liquefaction (JRA(4))

The following formulation is only applicable for sites with free face, and the proximity of the structure is within 100 m of the free face with minimum thickness of liquefiable soil layer of 5 m and stratum is widely distributed in the area near the site. The direction on the force should be applied depending on existence of free face of land, as shown in Fig.3. The forces calculated are to be applied as uniformly varying loads with zero at top of layer and increasing with depth. Also lateral movement force and inertial force should not be considered simultaneously.

step 1 : Resisting force acted by the soil above the liquefied zone

Here pile length deeper below the liquefied zone is ignored to offer any resistance. Hence for calculation, the depth above the liquefiable zone is considered for lateral resistance.

TeCHNICal PaPeRs


qNL = Cs CNL KpγNLx ... (4.12)

where,

qNL = Lateral movement force acting on unit area of structural member in non- liquefied depth (kN/m2),

Cs = Modification factor based on the distance from water front (Free face) obtained from Table 6,

CNL = Modification factor for lateral movement force in non-liquefied layer obtained from Table 7 and PL (Given later in step),

Kp = Passive earth pressure coefficient of non-liquefied layer,

ϒNL = Mean unit weight of non-liquefying layer (kN/m3),

x = Depth to bottom of layer from ground level (m)

And,

20

0

(1 )(10 0.5 )L LP F x dx= − −∫ ... (4.13)

where,

PL = Liquefaction index (m2),

FL = Liquefaction resistant factor calculated in equation 4.3 (If, FL ≥ 1, take FL = 1)

x = Depth to bottom of layer from ground level (m).

Table6ModificationFactorforDistance from Water-Front

Distance (s) from Water front (m)

ModificationFactor (Cs)

s ≤ 50 1.050 < s ≤ 100 0.5

100 < s 0.0

Table7ModificationFactorforLateralMovementinNon-liquefying soil layer

liquefaction index (Pl) (m2) Modificationfactor(CNl)

PL ≤ 50 0.0

5 < PL ≤ 20

31 - P 0.2 L

20 < PL 1.0

step 2 : Force acting on foundation from liquefied layer

This force acts on foundation in liquefied zone, with minimum at the top of the liquefying layer and maximum at the bottom.

{ ( )}L s L N NL L NLLq C C H x H= γ + γ − ... (4.14)

where,

qL = Force acting on unit area of structural member in zone of liquefied soil (kN/m2),

CL = Modification factor for lateral movement in liquefied zone = 0.3 (Can be safely assumed)

ϒL = Mean unit weight of liquefying layer (kN/m3)

HNL = Thickness of non-liquefying layer (m)

To understand the usage of above procedure, an example is provided here. The parameters are assumed as given below.

Fl HNL (m) HL (m) ϒNL (kN/m3) ϒL (kN/m3) s (m)

0.8 5 7 18 17.5 78

Uniformly varying load acting on the pile can be evaluated from equations 4.11, 4.12 and 4.13 as in Table 8 and are to be applied to the structure as shown in Fig. 3.

TeCHNICal PaPeRs


Table 8 Forces acting on Pile Due to liquefaction in sub-surface layer

x = 0 x = HNL x = HL

qNL (kN/m2) 0 67.5 -qL (kN/m2) - 13.5 31.87

4.3 Reduction in strength of soil Due to liquefaction

In order to gain residual strength in soil geotechnical parameters such as strength and shear modulus can be reduced of a soil found to be liquefaction prone with (FL < 1). It will be appropriate to say that for soil stratum assessed to be prone for liquefaction, does not necessarily liquefy in every dynamic event. For phenomenon of liquefaction to take place, there are numerous factors which should be favourable. Hence to offset this probability of occurrence, the Japanese code(4) suggests considering reduced soil strength for design of structures on liquefiable stratum. Japanese code provides with Table 9 based on dynamic shear strength ratio and level of earthquake ground motion. The values obtained from Table 9 are to be multiplied with strength parameters, in order to obtain reduced strength.

Table 9 Reduction Factor for Geotechnical Parameters (JRa(4))

Distance from

ground surface (x)

R ≤ 0.3 0.3 < R

Earthquake ground motion

Earthquake ground motion

Level 1 Level 2 Level 1 Level 2

FL ≤ 0.333

0 ≤ x ≤ 10 0.166 0 0.333 0.166

10 < x ≤ 20 0.666 0.333 0.666 0.333

0.333< FL ≤

0.666

0 ≤ x ≤ 10 0.666 0.333 1 0.666

10 < x ≤ 20 1 0.666 1 0.666

0.666 < FL ≤ 1.0

0 ≤ x ≤ 10 1 0.666 1 1

10 < x ≤ 20 1 1 1 1

* Soil layer with reduced or zero geotechnical parameters shall be assumed to act as overburden in calculations.

where,

R = Dynamic shear strength ratio (From step 3 section 4.2.1),

FL = Liquefaction resistance factor (From step 1 section 4.2.1),

x = Depth from ground surface (m)

Fig. 3 Schematic Diagram for Considering Lateral Forces on the Piles. (JRA(4))

TeCHNICal PaPeRs


5 CoNClusIoNs

In this paper a summary of latest methods to evaluate various mechanisms associated with liquefaction phenomenon are explained. The procedures to evaluate them are provided along-with solved examples.

1. The assessment of liquefaction potential has been extended to cover Cone Penetration Test (CPT) and Shear wave velocity method.

2. To apply correction factor for cementation effect in sands, a procedure has been described based on shear wave velocity test. An accurate estimate of age of sand deposit can be evaluated by this procedure.

3. Quantification of lateral movement of soil occurring on land surface due to liquefaction occurring in sub-soil layer is elaborated.

4. Method to obtain force acting on pile foundation of bridge structure has been described as per Japanese code. The estimated forces can be used in current practice of structural design.

5. Consideration for reduction in strength parameters in soil deposit with possibility of liquefaction is reviewed from Japanese code.

7 aCKNoWleDGemeNT

The authors are thankful for the comments and suggestions provided by experts from IRC. Their reviews in improving the quality of paper are acknowledged. The authors also wish their gratitude to the management of Spectrum Techno-Consultants Pvt. Ltd. for providing an opportunity to work on the subject.

ReFeReNCes1. Andrus R. D., Hayati H. and Mohanan N. P. “Correcting

Liquefaction Resistance of Aged Sands Using Measured to Estimated Velocity Ratio” (2009). Journal of Geotechnical and Geoenvironmental Engineering, 135(6), pp. 735-744.

2. Barlett S. F. and Youd T. L. “Empirical Prediction of Liquefaction-Inducted Lateral Spread” (1995). Journal of Geotechnical Engineering, 121(4), pp. 316-331.

3. Comprehensive Specifications for the Seismic Design of Bridges (2001), NCHRP 12-49.

4. Design Specification for Highway Bridges, Part 5, Seismic Design (2002), Japan Road Association (JRA).

5. Eurocode 8: “Design of Structures for Earthquake Resistance - Part 5: Foundations, Retaining Structures and Geotechnical Aspects”

6. Indian Standard Code 1893 “Criteria for Earthquake Resistant Design of Structures”, Part 1 General Provisions and Buildings, 5th Revision (2002), Bureau of Indian Standards (BIS), New Delhi.

7. Indian Standard Code 4968, “Method for Subsurface Sounding for Soil”, Part 3 Static Cone Penetration Test, 1st Revision (1997), Bureau of Indian Standards (BIS), New Delhi.

8. Indian Standard Code 13372 “Seismic Testing of Rock Mass - Code of Practice”, Part 2 Between the Boreholes, (1996), Bureau of Indian Standards (BIS), New Delhi.

9. Singh H. and Bhowmick A. “Assessment of Liquefaction Potential for Bridge Design” (2012). Indian Highways, Indian Roads Congress, September 2012, pp. 17-28.

10. TRANSIT New Zealand, Part of New Zealand Transport Agency, “Bridge Manual”, ISBN 0-478-04132-2.

11. Yi F. “Procedure to Evaluate Liquefaction-Induced Lateral Spreading Based on Shear Wave Velocity” (2010). Proceedings of Fifth Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, San Diego, California. Paper no. 1.57a, pp. 1-10.

12. Youd T. L., Hansen C. M. and Bartlett S. F. “Revised Multilinear Regression Equations for Prediction of Lateral Spread Displacement” (2002). Journal of Geotechnical and Geoenvironmental Engineering, 128(12), pp. 1007-1017.

13. Youd T. L., Idriss I. M., Andrus R. D., Arango I., Castro G., Christian J. T., Sobry R., Finn W. D. L., Harder L.F. Jr., Hynes M. E., Ishihara K., Koester J. P., Liao S. S. C, Marcuson W. F. III., Martin G. R., Mitchell J. K., Moriwaki Y., Power M. S., Robertson P. K., Seed R. B. and Stokoe K. H II. “Liquefaction Resistance of Soils: Summary Report from the 1996 NCEER and 1998 NCEER/NSF Workshops on Evaluation of Liquefaction Resistance of Soils” (2001). Journal of Geotechnical and Geoenvironmental Engineering, 127(10), pp. 817-833.


eFFeCT oF PeDesTRIaN CRoss - FloW oN CaPaCITy oF uRbaN aRTeRIals

SatiSh ChandRa*, G. SRinivaSa Rao** and aShiSh dhamaniya***

* Professor, Department of Civil Engineering, Indian Institute of Technology Roorkee, Email: [email protected]** Assistant Professor, Department of Civil Engineering, Gitam Institute of Technology, Gitam University, Visakhapatnam.*** Research Scholar, Department of Civil Engineering, Indian Institute of Technology Roorkee & Asst. Prof. SVNIT, Surat. Email: [email protected]

absTRaCTPedestrians crossing a road at mid block section reduce the traffic stream speed and thus the capacity of the road. These crossings may be at designated places where pedestrian markings are made or at undesignated places where no such markings are present. The pedestrians crossing an urban road at undesignated places are not uncommon and they force the motor vehicles to provide suitable gaps for their crossing. The present study demonstrates the effect of such crossings on capacity of an urban arterial road. Data are collected on six sections of urban arterial roads in New Delhi, Jaipur and Chandigarh. Three sections were selected without any side friction to estimate the base value of capacity. Remaining three sections were with pedestrian flow across the road at undesignated crossing. Speed and flow were measured in field and these data were used to estimate the capacity of a section. The capacity of 6-lane divided urban road in New Delhi is estimated as 2065 pcu/hour/lane. It reduces to almost half when pedestrian cross-flow is 1360 ped/hr. A mathematical relation is suggested for reduction in road capacity with volume of pedestrian cross-flow.

1 INTRoDuCTIoN

In recent times, many cities have seen a large increase in road traffic and transport demand, which has consequently led to deterioration in capacity and inefficient performance of traffic systems. Capacity analysis is necessary to have periodic evaluation of the existing facility and to arrive at number of lanes to be provided in a new facility. The factors affecting the capacity of highway are physical conditions, traffic conditions, control conditions and environmental conditions. These conditions include geometry of the road, heterogeneity of traffic stream, time of the day, weather conditions and type of control exercised.

The main characteristic of traffic on Indian urban roads is its heterogeneous nature and loose structure

of regulatory system. It requires an elaborate analysis procedure to arrive at number of lanes at desired level of service by considering all factors affecting highway capacity. Activities at the road side and on the carriage way are always expected to affect the capacity of any road and the speed at which it operates. On urban roads, which are also subjected to heterogeneous traffic conditions, the amount of disturbance to traffic flow from side friction is often considerable and the number of sources of friction is also large. They include but not limited to pedestrians, non motorised vehicles, parked and stopping vehicles, bus stop, bus bays and commercial activities along the road. In many developing Asian countries, the range and intensity of such side friction is so great that these activities need to be incorporated explicitly into procedures for calculation of speed and capacity of road links (Bang, 1995)1.

The present study was taken up with an objective of evaluating the effect of pedestrians crossing on capacity of 6 – lane divided urban arterials in India.

2 baCKGRouND lITeRaTuRe

Literature on pedestrian cross-flow with reference to capacity of an urban road is extremely scant. Kiec and Bak (2012)9 studied the influence of various types of mid-block pedestrian crossings on road capacity. They built a simulation model in VISSIM and calibrated on the basis of research results on drivers and pedestrians behaviour. The results show a relevant impact of willingness to give a right of way on urban streets

TeCHNICal PaPeRs


with capacity reduction and delays. Some other studies on factors affecting capacity of urban roads are also summarized here. Farouki and Nixon (1976)7 studied effect of the carriageway width on speeds of cars in the special case of free flow conditions in sub-urban roads at Belfast. It was found that the mean free speed of cars in suburban area increases linearly with the carriageway width over certain range of width (5.2m to 11.3m). Yagar and Van Aerde (1983)15 found that speed changes exponentially with change in lane width. Chandra et al. (1995)2 made a comprehensive study on capacity of urban roads. They observed that the PCU for a vehicle type decreases with increase in its own proportion in the traffic stream; which in turn will reduce the capacity of highway. Van Aerde (1995)14 presented a generic speed-flow-density relationship, which was successfully applied and calibrated for both freeways and arterials in both the micro and the macro domains. Parker (1996)13 observed that traffic composition plays an important role in determining capacity. It was found that the percentage of Heavy Goods Vehicles (HGVs) within a traffic stream has a major effect on capacity due to length, limited manoeuvrability, lower desired speed and engine power to weight ratio. Lum et al. (1998)10 observed traffic volume and travel time data at a number of arterial roads in Singapore to analyse the speed-flow relationships for radial and ring arterial roads. Christopher and Mason (2000)6 presented a statistical approach to define the variables having significant effects on operating speed. Marwah and Singh (2000)12 presented level of service classification of urban heterogeneous traffic. They considered journey speeds of cars and motorised two wheelers, concentration, and road occupancy to define LOS. Fitzpatrick et al. (2001)8 investigated geometric, roadside, and traffic control device variables that may affect driver behaviour on four-lane suburban arterials. Maitra et al. (2003)11 observed that because of the encroachment of abutting land and/or non-availability of land, in many situations roads are

widened partially and these roads do not follow standard lane dimensions, which affect the capacity adversely. Chandra and Prasad (2004)4 estimated that capacity of an urban road section increases by approximately 9 percent for every 10 percent increase in the proportion of 2-wheeler. The capacity of a section with side friction is approximately 12 percent lower as compared to a section with no side friction. Chiguma (2007)5 adopted an empirical method to determine the effect of side friction factors on traffic performance measures on urban roads in Dar-es-salaam city in Tanzania. The results showed that side friction can have considerable effects on speed and capacity like other commonly used factors in capacity analysis.

3 DaTa ColleCTIoN

Data on six sections of 6-lane divided urban arterial roads were collected in New Delhi, Jaipur and Chandigarh. The present study is mainly focused on evaluating the effect of pedestrians cross flow on capacity of urban roads. In order to quantify this effect, it is required to estimate capacity of an urban arterial without any influencing factor. Therefore, 3 sections were chosen without any side friction. All sections were at midblock of urban arterial roads sufficiently away from the influence of intersection and any other side friction like parking, bus stop etc. Video recording technique was used for data collection. It gives real time information and provides facility to replay the cassette as many times as needed for data extraction. A trap of about 50 m length was made on the section using white self adhesive tape for measurement of speed. Video camera was mounted at a reasonable height so as to have the clear view of the section and field data were collected for 4 to 6 hours on a section. Pedestrians crossing the road were also covered in the video so that their numbers can also be counted later. Table 1 provides details of site conditions at each of the six sections selected for the present study and time of data collection.

TeCHNICal PaPeRs


Table 1 Name of Road and site Conditions

section Number

Name of the Road Time Duration of Field survey

site Condition

I Swaminagar, New Delhi 8 AM to 12 PM and 4 PM to 6 PM

No side friction

II Haus khaz, New Delhi 8 AM to 12 PM and 4 PM to 6 PM

No side friction

III JLN Marg, Jaipur 9 AM to 11 PM and 4 PM to 6 PM

No side friction

IV AIIMS, New Delhi 8 AM to 12 PM and 4 PM to 6 PM

Pedestrian cross - flow

V Cannught Place, New Delhi 8 AM to 12 PM and 4 PM to 6 PM


VI Sector-17, Chandigarh 9 AM to 11 PM and 4 PM to 6 PM


The video recorded data were used to extract the information on classified traffic volume count and speed of individual vehicles. All vehicles are divided in to 5 categories as shown in Table 2. The standard car in the present study is taken as a passenger car having length 3.98 m and width 1.54 m and engine

power of 1400 cc. Tata Indigo, Indica, Hyundai Getz, Toyota Corolla, Honda city etc. were considered as standard cars. The big car is taken as a passenger car having length 4.58 m, width 1.77 m and engine power 2500 cc. Toyota Innova, Tata Sumo, Ford Endeavour etc. were considered as big cars.

Table 2 Vehicle Categories and their average Dimensions

Class of Vehicle average Dimensions Projected Rectangular Plan area (m2)length (m) Width (m)

Standard Car 3.72 1.44 5.39

Big Car 4.58 1.77 8.11

Bus 10.10 2.43 24.74

Three-wheeler 3.20 1.40 4.48

Two-wheeler 1.87 0.64 1.20

The classified volume count of all vehicles passing through a specified point, say first line of the trap, during a fixed period of 5 minute provided the measurement for composition and flow. The average time taken by each category of vehicle to cover the trap length was recorded using a digital stop watch of 0.01 second accuracy. The precaution was taken to

consider only those periods during which there was continuous flow of the vehicles with representation of each category. For the purpose of speed measurement of a vehicle type, 6-8 vehicles of that category were randomly picked up during the flow and average of their speeds was taken for the analysis.

TeCHNICal PaPeRs


4 aNalysIs oF DaTaThe data on classified volume count and average speed of each vehicle category on the selected mid block were extracted from the video film in the manner as explained in the previous section. The data were

analyzed to obtain the composition of traffic stream, hourly traffic volume and speed (km/hr) of each type of vehicle on different mid block sections. The composition of traffic stream at different sections is given in Table 3.

Table3TrafficComposition(%)atDifferentSections

section standard Car big Car Heavy Vehicle 3-Wheeler 2-WheelerI 46.49 7.55 3.64 12.70 29.61II 46.54 7.11 1.65 16.63 28.06III 26.50 3.96 1.53 10.48 57.53IV 47.20 5.78 2.32 19.09 25.60V 47.55 6.40 0.40 20.41 25.23VI 29.00 3.47 2.23 13.26 52.04

4.1 estimation of PCu Values

Traffic on Indian roads is of heterogeneous character with wide variation in the static and dynamic characteristics of different types of vehicles. One class of vehicles cannot be considered equal to any other vehicle class as there is considerable difference in their physical and flow characteristics. One way of accounting this non-uniformity in static and dynamic characteristics of vehicles is to convert all vehicles in to a common unit and the most accepted unit for this purpose is passenger car unit (PCU). PCU is a complex parameter and depends upon all factors of geometry and traffic operation. Many researchers have developed methods to estimate PCU for a vehicle type and there exists a large variation in PCU values being adopted in different parts of world.

PCU is a measure of relative interaction caused by a vehicle to the traffic stream compared to passenger car under a specified set of roadway, traffic and other conditions. This interaction will depend on traffic, roadway and environmental conditions. For a given facility, roadway and environmental conditions remain almost unchanged during field observation time, and therefore, traffic characteristics like traffic composition, traffic volume, speed and size of each category of vehicle must be able to explain all variations in PCU values for a vehicle type. The

composition accounts for any change in the traffic and changing degree of damaging effect at different volume levels. The vehicular interaction and all other geometric influences culminate in the speed of the vehicle, and physical size of a vehicle is supposed to indicate maneuverability, acceleration or deceleration capability and space occupancy on the road which are crucial in the measurement of density. Considering all these factors, Chandra and Kumar (2003)3 proposed the following equation to determine PCU for a vehicle type.

ic

ici AA

VVPCU = ... (1)

where,

PCUi = Equivalent passenger car unit of vehicle i,

Vc = Speed of car (km/hr),

Vi = Speed of vehicle i (km/hr),

Ac = Projected rectangular plan area of car (m2),

Ai = Projected rectangular plan area of vehicle i (m2).

The PCU values calculated for different types of vehicles at different sections are given in Table 4.

TeCHNICal PaPeRs


Table 4 average PCu Values of Vehicles at Different sections

section standard Car big Car Heavy Vehicle 3-Wheeler 2-WheelerI 1.0 1.5 5.6 0.9 0.2II 1.0 1.6 6.3 1.0 0.2III 1.0 1.5 5.5 1.1 0.2IV 1.0 1.6 5.9 0.9 0.2V 1.0 1.6 5.2 1.1 0.2VI 1.0 1.5 6.2 1.1 0.3

4.2 speed - Volume Relationship

Three basic parameters of traffic flow, speed, volume and density are used for estimation of traffic carrying capacity of a road. Since the measurement of traffic density in mixed traffic situation is difficult, attempts have made always been to concentrate on speed volume relationship. For determination of speed-volume relationship in heterogeneous traffic condition, the volume calculated by total vehicles recorded for each counting period were converted into equivalent number of PCUs using values obtained in Section 4.1.

In a mixed traffic situation, large variation exists in speeds of slow moving and fast moving vehicles. Therefore, spot speed or space mean speed cannot be considered for mixed traffic. It needs to be modified to suit the heterogeneous traffic conditions. For this purpose many researchers suggested use of mean stream speed. Mean stream speed or weighted space mean speed as given by Equation 2 is used in the present study.

∑∑

=

== N

i i

N

i iim

n

VnV

1

1 ... (2)

where,

Vm = mean stream speed (km/hr),

ni = number of vehicles of category i,

Vi = Speed of vehicles of category i (km/hr),

N = Total number of categories of vehicles in traffic stream.

The average stream speed calculated by above equation is plotted against hourly traffic volume to estimate the capacity.

4.3 Capacity of 6-lane Divided urban arterials

The speed – volume data for section I is used to plot speed flow relationship. The complete shape of speed - flow diagram is difficult to obtain from field data as it would require flow to be observed from free flow to forced flow conditions. Therefore, speed-flow data were converted to speed-density data using the fundamental relation between the flow (Q), density (K) and speed (V) as given in Equation 3.

VQK = ... (3)

The speed – density curve for section I is shown in Fig. 1. The data points follow a straight line trend (Greenshield model) and it was used to develop theoretical speed – flow curve as shown in Fig. 2. The capacity of this section is obtained as 6075 pcu/hr for one direction of traffic flow.

Similar curves were drawn at sections II and III also and capacity values were estimated. FFS of standard car observed on mid block sections of six-lane urban arterial roads are used to fit a statistical distribution. The speed distribution curves at section I are shown in Figs. 3 and 4. This resembles the shape of a normal distribution. The calculated and critical values of chi-square were 0.65 and 7.81 respectively at 3 degree of freedom and 5 percent level of significance. The capacity values and 85th percentile FFS of passenger car, which is considered as the operating speed, for these sections are given in Table 5.

TeCHNICal PaPeRs


Fig. 1 Speed-Density Relationship at Section-I

Fig. 2 Speed-Flow Relationship at Section-I

Fig. 3 Frequency Distribution of FFS of Car at Section I

Fig. 4 Cumulative Frequency Distribution of FFS of Car at Section I

Table 5 Capacity and Free Flow speed at 6-lane Roads

section Total Capacity in one Direction

(pcu/hr)

Capacity (pcu/hr/lane)

85th Percentile

FFs (kmph)

I 6075 2025 80.80

II 6314 2104 86.60

III 4955 1652 70.41

The capacity of section III is less than that of other two sections. This section is located in Jaipur and influence of city size might have caused reduction in capacity. Further, the FFS in Jaipur is around 70 kmph, which is 80 percent of operating speed observed on Delhi roads. This is also a reason for low capacity value of section III.

4.4 effect of Pedestrian Cross-Flow

The speed–density curve for section IV is shown in Fig. 5. The data points follow a straight line trend and it was used to develop theoretical speed – flow curve as shown in Fig. 6. The capacity of this section is obtained as 4739 pcu/hr for one direction of traffic flow. Similar curves were drawn at sections V and VI also and capacity value was estimated. The capacity values and pedestrian flow as observed for these sections are given in Table 6.

TeCHNICal PaPeRs


Fig. 5 Speed-Density Relationship at Section- IV

Fig. 6 Speed-Flow Relationship at Section-IV

Table 6 Capacity and Pedestrian Flow at Different sections

section Total Capacity In one Direction

(pcu/hr)

Capacity (pcu/hr/

lane)

Pedestrian Flow (ped/

hr)IV 4739 1580 682V 5200 1733 508VI 3120 1040 1360

The effect of pedestrian crossing on the traffic stream is to reduce the stream speed and capacity of the section. Similar trend is observed in the present study also. Considering lane capacity of a 6 – lane urban arterial as 2065 (average of section I & II), the

presence of pedestrian cross-flow reduces the capacity of sections IV, V and VI by 23, 16 and 49.6 percent respectively. Fig. 7 shows the variation in percent reduction in capacity with pedestrian cross-flow. It is given by Equation 4.

% Reduction in Capacity = 0.035 * Pedestrian Volume. ... (4)

It suggests that a pedestrian volume of 100 ped/hr crossing the road will reduce its capacity by 3.52 percent.

Fig. 7 Percent Reduction in Capacity due to Pedestrian Cross-Flow

5 CoNClusIoN

IRC:106-1990 suggests capacity of a six lane urban arterial road as 3857 pcu/hr (1285 pcu/hr/lane). The capacity of 6-lane divided urban arterial roads in present study is estimated at 2065 pcu/hr/lane for no side friction condition. It is almost 60 percent higher than that reported in the IRC code, which was drafted more than 20 years back. The vehicle and road making technology in these years have improved considerably and these have resulted in higher capacity values. The pedestrians crossing a road will interact with vehicular traffic and thus reduce the capacity of the road. Reduction in capacity will however, depend on pedestrian cross flow. The present study shows that capacity of a 6 – lane urban road reduces to almost half

TeCHNICal PaPeRs


(49.6% reduction at section VI) when pedestrian cross-flow is 1360 peds/hr. This reduction is only 15% for a pedestrian cross-flow of 508 peds/hr. A mathematical equation is developed between the percent reduction in capacity and pedestrian cross-volume. It suggests that a pedestrian volume of 100 peds/hr crossing the road will reduce its capacity by 3.52 percent.

The present study is based on limited data and hence validity of straight line relation between reduction in capacity and pedestrian cross-flow needs to be checked in future studies by taking data on some more sections. Further, use of speed-density curve as obtained from speed-flow data has been made in this paper. It will be worthwhile to investigate any error caused by this calibration in estimation under mixed traffic condition.

ReFeReNCes

1. Bang, K.L. (1995), “Impact of Side Friction on Speed-Flow Relationships for Rural and Urban Highways”, HDM 4 Project Report, SWEROAD Indonesia, pp. 1-27.

2. Chandra S., Kumar, V., and Sikdar, P.K. (1995), “Dynamic PCU and Estimation of Capacity of Urban Roads”, Indian Highways, Indian Roads Congress, Vol. 23(4), pp. 17 – 28.

3. Chandra, S. and Kumar, U. (2003), “Effect of Lane Width on Capacity under Mixed Traffic Conditions in India”, Journal of Transportation Engineering, ASCE, Vol. 129(2), pp. 155-160.

4. Chandra, S. and Prasad, N.V. (2004), “Capacity of Multilane Urban Roads Under Mixed Traffic Conditions”, Highway Research Bulletin, Indian Roads Congress, No. 75, New Delhi, pp. 97-103.

5. Chiguma Masatu L.M. (2007), “Analysis of Side Friction Impacts on Urban Road Links”, Doctoral Thesis in Traffic and Transportation Planning, Royal Institute of Technology, Stockholm, Sweden.

6. Christopher M. P. and Mason, J.M. (2000), “Analyzing Influence of Geometric Design on Operating Speeds

Along Low-Speed Urban Streets”, Transportation Research Record 1737, TRB, National Research Council, Washington, D.C., pp. 18-25.

7. Farouki, O.T. and Nixon, W.J. (1976), “The Effect of Width of Sub-Urban Roads on the Mean Free Speeds of Cars”, Traffic Engineering Control, Vol. 17 (12), London, pp. 518-519.

8. Fitzpatrick, K., Carlson,P., Brewer, M. and Wooldridge, M. (2001), “Design Factors that Affect Driver Speed on Suburban Streets”, Transportation Research Record 1751, TRB, National Research Council, Washington, D.C., pp. 18-25.

9 Kiec, M. and Bak, R. (2012), “Influence of Various Types of Mid-Block Pedestrian Crossings on Urban Street Capacity”, Transportation Research Board Annual Meeting 2012, Paper No 12-3818.

10. Lum, K. M., Fan, H.S.L., Lam, S. H. and Olszewski, P. (1998), “Speed-Flow Modeling of Arterial Roads in Singapore”, Journal of Transportation Engineering, ASCE, Vol.124, pp. 213-222.

11. Maitra, B., Sikdar, P.K. and Dhingra, S.L. (2003), “Effect of Road Width on Traffic Congestion and Level of Service”, Highway Research Bulletin No. 68, Indian Roads Congress, New Delhi, pp. 123-132.

12. Marwah, B. R. and Singh, B. (2000). “Level of Service Classification for Urban Heterogeneous Traffic: A Case Study of Kanpur Metropolis.” Transportation Research Circular E- C018: 4th International Symposium on Highway Capacity, Maui, Hawaii, 271-286.

13. Parker, M.T. (1996), “The Effect of Heavy Goods Vehicles and Following Behavior on Capacity at Motorway Sites”, Traffic Engg. and Control, Vol. 37(9), London, pp. 524-532.

14. Van Aerde, M. (1995), “Single Regime Speed-Flow-Density Relationship for Congested and Uncongested Highways”, 74th TRB Annual Meeting, Washington, D.C., Paper No. 950802.

15. Yagar, S. and Van Aerde, M. (1983), “Geometric and Environmental Effects on Speeds of Two Lane Highways”, Transportation Research, Vol. 17A, No. 4, pp. 315-325.


oPTImum maINTeNaNCe PReDICTIoNs aND DeRIVaTIoN oF TRaNsITIoN PRobabIlITy maTRICes FoR bRIDGe

DeTeRIoRaTIoN moDelINGS. R. KatKaR* and PRaShant P. naGRale**

* Research Scholar, E-mail: [email protected]** Associate Professor, E-mail: [email protected]

absTRaCTThe ideal Bridge Maintenance Management System (BMMS) is the one which closely predicts the future condition of the structure. The traditional reactive management has now been taken over by proactive management. The study refereed to IRC:SP:35-1990, Appendix 4, Inspection Proforma and masonry register maintained by PWD Maharashtra. For this study, database of 20 number of RCC three girder type bridges for past consecutive years from their construction is quantified in 7 discrete condition states based on the inspection proforma comments. The data base is analyzed to test Markovian Property using Chi square test for inference testing. As available sample database is limited, the same is considered to develop database matrix. The model is generalized to construct one step Transition Probability Matrix (TPM) using Poisson’s distribution and is tested for goodness of fit using Chi square test. From one step TPM, successive TPMs are derived. Using deterministic approach, average condition state of bridge network under study is calculated. Considering 100 new bridges, probable number of bridges in each condition states in each year, up to 60 years, is predicted by using excel programming. Bridge deterioration process is presented in the form of curve. However, improved database collection methodology helps to build more accurate and complete Markov Decision Process (MDP) based model.

1 INTRoDuCTIoN

Goal of good BMMS provides an efficient method for estimating replacement time, budgeting and maintenance intervals. The ideal BMMS is the one which could very closely predict the future condition of the structure. If a proper and extensive record of past performance of a structure is available, a statistical model such Markovian model helps in future predictions.

Bridges are vital elements in road network and constitute a significant part of total construction cost of network. A survey carried out in Maharashtra State on State Roads has shown that there are 871 bridges with deficient load carrying capacity needing repairs

and strengthening (Panel Discussion on Bridge Management System, IRC, Dec. 1990). The BMMS needs a strong data base which has got to have an inventory of Bridge structures including the salient features of materials used, type of structures and locations, along with all the regards of past inspections, assessments and strengthening carried out to date. Now days, the bridge management is based on new concepts. The traditional reactive management has now been taken over by proactive management.

The BMMS in India is in its initial stage of implementation, IRC:SP:35-1990 “Guidelines for Inspection and Maintenance of Bridges” is the first organized step in this direction. Data collection is carried out based on the guidelines given in above mentioned IRC code and masonry register maintained by PWD. To design Maintenance Management System (MMS) it is required to understand the deterioration process of the bridge. To know extent of deterioration, inspection data and its quantification in terms of condition state is required.

Deterioration of any infrastructure facility system invariably changes with time. The transition from one condition state of system to another as a function of time or its corresponding transition probability depends on its prior condition states. However, if the transition probability depends only on the current state, the process of change may be modeled with the Markov process. If the condition state space is countable or finite set, then the process is called a Markov chain. Moreover, if change can occur only at discrete points of the parameters for example, at discrete instants of time, the process is a discrete parameter Markov chain. As mechanistic, empirical,

Sardar Patel College of Engineering, Andheri (w), Mumbai

TeCHNICal PaPeRs


expert etc. maintenance management systems are data centric, developing country like India where data collection and maintenance community are one and same which failed in inspection and collection of extensive data base. And hence probabilistic maintenance management systems are invariably useful in such situation. Here in this study, Markovian decision process (MDP) is used for probabilistic analysis of bridge maintenance predictions.

2 lITeRaTuRe ReVIeW

The maintenance is the strategy essentially helps in decision making and in prioritization of the defects and improvement of the bridge network. Madanat and Ibrahim (1995) have supported the applications of Markovian transition probabilities for infrastructure maintenance management system to provide forecast of facilities conditions. It is identified that existing approaches used to estimate Markovian transition probabilities from inspection data are mostly approximate and it has several statistical limitations. It is concluded that the Poisson’s model may be more effective model for prediction of infrastructure forecasts. Prapannachari, and Kaistha, (1994), in their paper have discussed the complete bridge maintenance management system and primary activities associated with BMMS in Indian scenario.

Allen et.al (1989) have highlighted the importance of generation of funding for repair and rehabilitation of road and bridges. It is further emphasized that the efforts towards repair and maintenance must be made for an effective allocation of the limited funds and hence prioritizing the maintenance of existing bridges to upgrade it for satisfactory level of service is essential.

Scherer and Glagola, (1994), tested the Markovian model for BMMS. The research indicates that the MDP is a powerful and useful technique of BMMS only when data collection for repairs and maintenance history can be improved in order to build more accurate and complete MDP based model. Abbas et.al (1994) described the pavement maintenance predictions

and developed a model to priorities the repair policy considering budget constraint. Yang (2004) reviewed different mathematical models showing deterioration behavior of various infrastructure facilities like roads and bridges.

Study of different types of deterministic models such as pure empirical model, mechanistic – empirical model, expert system model, probabilistic model etc., he concluded that the probabilistic model i.e. Markovian models are best suited for prediction of exact deterioration behavior of any infrastructure system. This fundamental finding is the base of the current research work.

The method used for the data collection and its quantification in terms of condition state is done by using IRC:SP:35-1990 “Guidelines for Inspection and Maintenance of Bridges”. Indian Highways (1990 and 1992) are referred to understand the historical development of BMMS in India. Further guidelines about inspection of bridges are referred from “Technical Circular” (1988) of Government of Maharashtra, PWD. In the present study use of Markov models for BMMS and its compliance with the Markovian property is explored.

3 sTuDy meTHoDoloGy

The typical infrastructure maintenance decision making environment involves multiple objectives and uncertainty, and is dynamic. One of the most commonly used infrastructure models is a Markov decision process (MDP).

3.1 markovian Property

Markov processes represent one of the best-known and most useful classes of stochastic process. A stochastic process is a Markov process if it satisfies the following condition- ‘given that the present (or most recent) state is known, the conditional probability of the next state is independent of states prior to the present (or most recent) state’.

TeCHNICal PaPeRs


P (Xt+1 = xt+1 | X0 = x0, X1 = x1, Xt = xt)

= P (Xt+1 = xt+1 | Xt = xt)

where,

t = 0, 1 , 2 …… t years; X0, X1, ……, Xt are random variables and x0, x1, …., xt are the corresponding condition

states.

Because of the Markovian property, these processes are often referring to as being ‘memoryless’.

3.2 Transition Probabilities

The conditional probability given by the right hand side of eq. (1) represents a transition probability. Thus a transition probability is defined as the conditional probability that the process will be in a specific future state given its most recent state. These probabilities are also termed as one-step transition probabilities, since they describe the system between t and t+1. Similarly, we refer to k-step transition probability as the conditional probability describing states in the system between t and t + k.

A convenient representation of the one step transition probability for the discrete state case is given by the following transition matrix:

where,

n = the number of exhaustive and mutually exclusive condition states and

Pij = the transition probability of going from the present (ith) state of the next (jth) state.

By definition, the elements on P must satisfy the following two properties,

0 ≤ Pij ≤ 1 for all i and j

P = 1, i = 1, 2,......., nijj=1

n

∑

3.3 markov Chains

A Markov chain is a stochastic process with the following properties:

1. Discrete state space,

2. Markovian property, and

3. One step transition probabilities that remains constant over time.

If additionally the discrete state space has finite number of states, then it is termed as finite state Markov chain. Markov chains constitute a prominent class of Markov processes that have desirable computational properties for real world implementation. Markov chain is completely determined once the transition matrix and the sets of unconditional probabilities for initial states are specified. Knowledge of these two sets of probabilities allows the Probabilistic prediction of specific states at future times.

If initial TPM at time t=1 (year) is known, i.e. P is known, then to determine all conditional transitional probabilities after 2 years, simply square the P as follows –

P [2] = P * P

P [3] = P[2] * P …..

P [k] = P[k-1] * P

where,

P is one step transition matrix;

P [k] is condition state TPM after kth time interval.

TeCHNICal PaPeRs


3.4 unconditional v/s Conditional Probabilities

In the preceding section, P[k] was defined as the k–step TPM between states i and j, just as in the case of the one step transition probability P, this probability is the conditional probability because the probability of state j after the Markov chain goes through transitions is statistically dependent on the initial state i. If the unconditional or absolute probability of state j after k transitions (uj [k]) is desired, then the following product must be determined:

u [k] = u [0] * P [k]

where,

u [k] = u1 [k], u2 [k], u3 [k], …. , un [k] is the row vector of unconditional probabilities for all n states after k transitions, u[0] is the row vector of initial unconditional probabilities, and P is the one step transition matrix.

From IRC:SP:35-1990, Appendix 4, Inspection Proforma is referred which includes the data to be collected about Bridge structures including the salient features of materials used, type of structures and locations, past inspections, assessments and strengthening carried out. to date. From this Routine Inspection Proforma and masonry register maintained by PWD divisions, one can understand whether bridge requires minor/major repair or minor/major rehabilitation. Accordingly available inspection Proforma comments, data of 20 number of RCC three girder type bridges for past consecutive years from their construction is quantified. Condition state description is defined according to Scherer and Glagola (1994) and is shown in the Table 1. In the absence of systematic and detailed data base collection, condition state definition step is the preliminary step towards efficient proactive maintenance management system.

Table 1 Condition state Descriptions

Condition state Description 7 New condition.6 Good condition: no repairs needed.5 Generally good condition: potential exists for minor maintenance.4 Fair condition: potential exists for major maintenance.3 Generally fair condition: potential exists for minor rehabilitation.2 Marginal condition: potential exists for major rehabilitation.1 Poor condition: rehabilitation requires immediately.

3.5 markov Chain Development

A critical component of utilizing the Markovian probabilities modeling approach is generating condition states that adhere to the Markovian property. Often, the Markovian assumption is made without verification, resulting in a model of questionable quality. The Markovian property states that the condition distribution of a transition from the present state to any future state given the past state is independent of any states and depends only on the present state.

The Markov chain development is the generation of transition probabilities that define the potential condition states of a bridge over time. Before the transition probability generation can occur, the classification hierarchy structure of the overall bridge management network system must be established. This involves classifying bridges according to its age, traffic loading, environment, etc. This classification structure reduces the overall bridge network to workable local network. Transition probabilities must be generated for each defined class. Once a one-step transition matrix is generated from observed data base

TeCHNICal PaPeRs


frequencies from which n-step transition matrices can be generated by matrix multiplication to give stochastic descriptions of future bridge states.

3.6 state Transition sequences

The seven bridge condition states described earlier represents seven possible condition ratings of a particular classification of bridges. It must follow that the deterioration probability of transition from the present to future state is not dependent on past states for the Markovian property to hold. Written in probabilistic notation:

P (j, m | i) = the probability of going from state j to state m, given state i occurred previously to j (i ≥ j ≥ m).

If the Markov property holds then P (j, m | i ) = Pjm. An exploratory analysis using available data is developed in the following sections to illustrate the verification of the Markovian property. Consider collected data on each bridge every year and maintains this information as a database. Therefore, the MDP discrete time intervals for the illustrations are one year:

Each of the cases described here involves an analysis of two different three-state transition sequences. A three-state transition sequence consists of three condition states: past, present and future. These three states correspond to three consecutive bridge condition rating occurring over one year period. Two possible transition sequences, with the same present and future states but different past states are tracked to determine if there is a difference in occurrence dependent of past state history. A simple frequency analysis of sequence occurrence is employed. If there is no significant difference in frequency between the sequences being tracked, this may indicate the Markovian property satisfied. An inference analysis using a Chi-square statistic is formulated to test the significance of the Markovian property. Transition sequences of the condition states most frequently occurring in the database are assumed sufficient to establish the Markovian property as it applies to the entire deterioration model.

The following terminologies are required for the informal analysis of state transition sequences based on frequency probabilities that we describe as follows: State transition sequence (STS): STS

refers to a particular three state sequence of concern. This sequence incorporates three consecutive condition ratings of a bridge to establish past- present-future identification.

State sequence occurrences (SSO): SSO refers to the number of times a specified STS appears in the available database.

Two-state occurrences (TSO): TSO refers to the number a specified two-state sequence appears in the available database. The two-state sequence involves the past state and the present state. Tracking these occurrences allow for the generation of frequency probabilities as will be described later for example two possible STSs are:

1. (6,6 | 7 ) past = 7, present = 6, future = 6

2. (6,6 | 6 ) past = 6, present = 6, future = 6

The two-state sequence of concern for (6, 6 | 7 ) is the transition from state 7 to state 6 while the two-state sequence of concern for (6, 6 | 6) state 6 remaining unchanged from the past to the present

Frequency probability : Frequency probability refers to the ratio of SSO over TSO:

( )TSOSSOi|,m,jP =

P j, m, | i =

j, m | i occurrencesi, j occurrences

( )( ) ( )

∑ ∑

This study is used to calculate future condition states of bridges using data based on inventory and quantification of inspection data sheet maintained in the form of masonry register by Maharashtra PWD & as per guidelines of IRC:SP:35-1990.

TeCHNICal PaPeRs


4 TesTING oF DaTa base

Three cases of state transition sequences are analyzed to identify Markovian compliance based on quantified database. Table 2, shows the question of interest is whether or not the transition from state 6 to state 6 (i.e., no change) is dependent on the previous state, (in this example state 6 or state 7). If the Markovian property holds, the probability of this transition should be independent of previous state. There were 20 total examples in the database of a bridges making the transition from state 7 (termed past) to state 6 (present), and in 14 of these cases the following transition was to state 6 (future). Therefore, the probability P (6, 6 | 7) = 0.70. Thus, there is a 70% chance of maintaining state 6 if the previous state was 7. Using 6 as the past state, we get P (6, 6 | 6) = 0.706. Without noise in the data and with perfect Markovian compliance, these numbers should be the same. However, given these difficulties, the probabilities are quite close, with a difference of 0.006.

Table 2 Frequency Probabilities to test markovian Properties

Case – I Comparison of P (6, 6 | 7) Versus P (6, 6 | 6 )

state Transition sequence

state sequence

occurrence

Two state occurrence TSO

SSO

(6, 6 | 7 ) 14 20 0.70

(6, 6 | 6) 36 51 0.706

--- ---- ---- 006.0=∆

Case II - Comparison of P (5, 5 | 6) Versus P (5, 5 | 5 )


state sequence

occurrence


SSO

(5, 5 | 6 ) 15 20 0.75(5, 5 | 5) 53 68 0.775

--- ---- ---- 025.0=∆

Case – III Comparison of P (5, 4 | 6) Versus P (5,4 | 5 )


state sequence

occurrence


SSO

(5, 4 | 6) 05 20 0.25(5, 4 | 5 ) 15 68 0.22

--- ---- ---- 030.0=∆

There is no significant difference in the frequency probabilities for all cases under study. A complete analysis would require testing many more cases for compliance. The results of this informal analysis suggest the independence of state-to–state deterioration transition from past state history. This independence is the basis of the Markovian property. To further verify the credibility of this informal analysis Chi-square test for inference is conducted.

4.1 Chi-Square(χ2) Test

It is defined by W. A. Spar and C. P. Bonini as the process by which the conclusion is drawn about some measure of a population based on a sample value. The measure might be a variable, such as the mean, S.D. etc. The purpose of sampling is to estimate some characteristics of the population from which the sample is selected.

Chi-square (χ2 test) distribution is mainly used to

i) test the goodness of fit,

ii) to test the independence of attributes and

iii) to test if the population has a specified value of the variance σ2.

The degree of freedom is the number of independent variants which make up the statistic χ2. The number of degrees of freedom is also defined as the total number of observations minus the number of independent constraints imposed on the observations. For a contingency table with “r” rows and “c” columns, the degrees of freedom (d.f.) = (r-1) (c-1). E.g. if a contingency Table 3 shows 2 rows and 2 columns, corresponding degree of freedom = (2-1)*(2-1) = 1.

TeCHNICal PaPeRs


Table 3 Contingency Table Illustrating Inference Testing of Data base

sequence state sequence

occurrenceSSO

Non sequence

occurrenceSSOTSO −

TSO

1 a b a+b2 c d c+d

Total a+c b+d a+b+c+dfor case – I for P( 6, 6 | 7) v/s P (6, 6 | 6)

1 14 6 202 36 15 51

Total 50 21 71

The informal analysis generates frequency probabilities for comparison purposes. However, the sample sizes of some of the generated frequencies vary drastically due to lack of data for some of the condition states. A closer examination of the frequency results by inference tests is utilized to minimize any discrepancies.

The general hypothesis in inference testing is that two distributions are different. In this scenario, the null hypothesis is that the generated frequency distributions for a specified case are from the same distribution. Verifying this null hypothesis based on Chi-squared Statistics gives insight into the concept of accepting the deterioration data shown in contingency Table 3. The Chi-squared value is defined as:

( ) ( )( )( )( )( )

22 ad bc a b c d

a b c d a c b d− + + +

χ =+ + + +

and is referenced to a Chi-square Table with one degree of freedom to determine the corresponding significance level (α). Creating the contingency tables for the three cases described previously, the following Chi-square values are generated.

Case 1 – Contingency table results for P( 6, 6 | 7) v/s P (6, 6 | 6)

( ) ( )( )( )( )( )

22 14*15 6*36 14 6 36 15

0.0023814 6 36 15 14 36 6 15

− + + +χ = =

+ + + + ,

from table α = 0.9335

Similarly, values arrived for case – II and case–III, χ2 = 0.0761,

From table α = 0.8450. From these results, there is no reason why the null hypothesis should be rejected. The inference tests for each case indicate that the frequencies are of the same distribution with high significance levels. This inference test helps further support the suggestion of independence of state deterioration transition on previous states.

These results indicate, for the limited data reasonably satisfies the Markovian property and can be used for development of a performance prediction model. If an MDP model is to be developed for a state BMMS, then a similar more exhaustive procedure must be used to verify the Markovian property. This data analysis is a very positive sign, indicating that the Markovian assumption may be appropriate for other BMMS.

5 FoRmulaTIoNs oF oNe sTeP TRaNsITIoN PRobabIlITy maTRICes

Once the Markovian property has been verified, various statistical analysis procedures described in the following can be utilized for the development of probability transition matrices. This analysis offers one technique for the generation of TPM for a given bridge classification. The procedure consists of two phases:

i) Data quantification of bridge condition data points;

ii) Deterioration probability distribution modeling.

The Bridge condition data from the available bridge database must be appropriately organized according to the defined condition states. One part of this data quantification process is the classification of the bridges. The second part involves the generation of data matrices for each classification of bridges. The data matrices display the actual number of bridge deteriorating from one state to another with respect to the initial condition states and class. These

TeCHNICal PaPeRs


database matrices are then used in the statistical analysis procedures for the generation of probability distributions for the deterioration prediction modeling of the condition states.

5.1 modeling for the Deterioration Prediction of the Condition states

The deterioration distribution modeling methodology consists of generating probability distribution for the different rows of a particular data matrix. One row of a data matrix indicates the possible future condition states given an initial condition state. Different deterioration modeling procedures are incorporated in the development of the various transition matrices. The procedure includes a formal probability distribution modeling procedure.

As the sample size for a data matrix row is greater 30, formal statistical analysis procedure is employed. One such procedure is the utilization of formal probability distributions. Since the data matrices once define seven distinct condition states, discrete distributions are used to model the probability that a bridge at a particular state at the next decision epoch, given the current bridge state. Such probability functions can be selected using goodness of fit tests, from numerous possible distributions. One discrete function that provides very useful distribution is the Poisson’s probability mass function. Table 4 shows the extract of available data base used in the analysis.

Table 4 extract of available Data base used in the analysis

To state →

7 6 5 4 3 2 1

From state

↓

7 44 20 00 00 00 00 006 00 51 20 00 00 00 005 00 00 68 20 00 00 004 00 00 00 95 20 00 003 00 00 00 00 158 20 002 00 00 00 00 00 522 201 00 00 00 00 00 00 1

Table 5 mean Calculation for First Row

xi fi fixi

0 44 00

1 20 20

2 00 00

3 00 00

4 00 00

5 00 00

6 00 00

N=64 Σ fxi = 20

Therefore, mean (m) = 0.3125fx

N=∑

Now, the above data checked for fitness of Poisson distribution, for Ist row of above data- matrix, if xi = condition state reduction per year. Table 5 shows the calculation procedure of mean for Ist row of above data- matrix and is found to be 0.3125.

The expected frequency of bridges by Poisson Law can be calculated using following equation,

0.312564* *0.3125

! !

m x xi

xi i

Ne m efx x

− −

= =

xi = 0, 1, 2, 3, 4, 5, 6, from above data, f (0) = 46.82 f (1) = 14.63 f (2) = 2.29, f (3) = 0.24, f (4) = 0.02, f (5)= 0.001 and f (6) = 0. Table 6 shows the test of goodness of fit between observed and Poisson’s distribution using Chi–square test for first row of the data matrix. The results shows that, the tabulated value of χ2 at 5% level with 1 d. f. is 3.84 and calculated value is found to be 0.6324. Since the calculated value of χ2 less than the tabulated value, we accept the Poisson distribution and conclude that it fit to observed data. Similarly, for other rows of data matrix, calculated expected frequency and calculated value of χ2 to test goodness of fit is evaluated and presented in Table 7.

TeCHNICal PaPeRs


Table6GoodnessofFitBetweenObservedandPoisson’sDistributionUsingΧ2 Test for 1st Row of Table 4

oi ei (oi – ei) (oi – ei)2 (oi – ei )

2 / ei

442000000

46.8214.632.290.24 0.020.0010

= 17.181

-2.82+2.82

7.95247.9467

0.16980.4624

64 64 Σ(Oi – Ei) = 0 -- χ2 = 0.6324 Table value at 5% = 3.84

Table 7 Testing of Goodness of Fit and evaluation of Transition Probabilities

First

Row

O. F. 44 20 00 00 00 00 00χ2 = 0.6324

E.F. 46.82 14.63 2.29 0.24 0.02 0.001 00T. P. 0.732 0.229 0.036 0.003 00 00 00

Second

Row

O. F. --- 51 20 00 00 00 00χ2 = 0.5050

E.F. ---- 53.57 15.09 2.12 0.20 0.01 0.0008T. P. --- 0.755 0.213 0.030 0.002 00 00

Third

Row

O. F. --- --- 68 20 00 00 00χ2 = 0.3129

E.F. --- --- 70.11 15.93 1.81 0.14 0.007T. P. --- --- 0.797 0.181 0.021 0.001 00

Fourth

Row

O. F. --- --- --- 95 20 00 00χ2 = 0.1745

E.F. --- --- --- 96.64 16.81 1.46 0.088T. P. --- --- --- 0.840 0.146 0.013 0.001

Fifth

Row

O. F. --- --- --- --- 158 20 00χ2 = 0.0695E.F. --- --- --- --- 159.082 17.874 1.004

T. P. --- --- --- --- 0.894 0.100 0.006Sixth

Row

O. F. --- --- --- --- --- 522 20χ2 = 0.0271E.F. --- --- --- --- --- 522.36 19.28

T. P. --- --- --- --- --- 0.964 0.036O. F. – Observer Frequency, E. F. – Expected Frequency, T. P. Transition Probability

From Table 6 it can be concluded that given data is fit for Poisson distribution. Similarly, testing of goodness of fit and evaluation of transition probabilities using Poisson’s distribution for all

rows of Table 4 is carried out. Results are tabulated in Table 7. Transition probabilities arrived from above calculations is arranged in matrix form as below.

TeCHNICal PaPeRs


Table 7

State 7 6 5 4 3 2 17 0.732 0.229 0.036 0.003 0 0 06 0 0.755 0.213 0.030 0.002 0 0

P =5 0 0 0.797 0.181 0.021 0.001 04 0 0 0 0.840 0.146 0.013 0.0013 0 0 0 0 0.894 0.100 0.0062 0 0 0 0 0 0.964 0.0361 0 0 0 0 0 0 1

Poisson’s distribution is found accurate mode of deterioration of bridge condition states. Once the transition matrices are generated, the application of matrix chains can be employed to predict deterioration over time. In this analysis the generated transition matrices represent the probability of an initial condition state “i” deteriorating to a future condition state “j” in a one-year period. Therefore, the time interval for the Markov chain is one year. The multiplication of the transition matrices allows for the probabilistic predictions of future bridge conditions. As expected, if no maintenance is performed, the bridge is eventually deteriorated to State 1.

The generated transition matrices of the deterioration model can, therefore, represent a probabilistic prediction of future bridge conditions. The probabilistic results indicate a range of possible condition states with an associated probability. A deterministic approach, compared to probabilistic results, yields only one potential future value for a condition state. Deterministic approaches may seem limited in the realistic modeling of bridge deteriorations however, they prove useful for comparisons.

The method of comparison is accomplished by generating deterministic decay curves. These decay curves represent expected condition values over time. The expected values are determined from the transition matrix P, and initial probability vector u0.

The initial probability vector u0, represents a priori probabilistic condition of a Bridge. If, for example, the bridge is new then it is in condition

state 7, yielding an initial probability vector {u0} = {1, 0, 0, 0, 0, 0, 0}. If the initial condition state is described stochastically by having 50% chance of being in state 7 and 50% chance of being in state 6 then, {u0} = {0.5, 0.5, 0, 0, 0, 0, 0}. This indicates that for a particular class of bridges, at the time of prediction, 50% of the bridges is in condition state 7 while remaining 50% satisfy condition state 6. The initial probability vector is multiplied by the one step transition matrix P, generating the vector uj. This vector represents the probability of deteriorating to condition state j, given initial probability vector u0, for each j, uj can then be multiplied by the condition state j, and summed over all j. This sum is the deterministic expected value of the condition state. Formally, this process proceeds as follows:

u j = {u0 } P

( ) { }jj

00 u*juE ∑=where,

E(u0) is the expected state of a bridge in one year given initial probability vector u0. In order to generate the expected state of a bridge in future years, the following generalization of Equation and Table 8 shows the expected state values for the calculated one step transition matrix-

( ) { }( )[ ]∑=j

jk

00 P*u*ju|kyear at stateE

Applying above formulae on one step transition matrix P, it is possible to calculate successive transition probability matrices. Results are tabulated in Table 8.

TeCHNICal PaPeRs


Table 8 successive TPm, average Condition Rating and Number of bridges in each Condition state for year 1, 5, 10, 30, 50 and 60 years

Condition State (C.S.) of Bridges average Condition

state 7.000 6.000 5.000 4.000 3.000 2.000 1.000 0.732 0.229 0.036 0.003 0.000 0.000 0.000 0.000 0.755 0.213 0.030 0.002 0.000 0.000 0.000 0.000 0.797 0.181 0.021 0.001 0.000

year 1 0.000 0.000 0.000 0.840 0.146 0.013 0.001 0.000 0.000 0.000 0.000 0.894 0.100 0.006 0.000 0.000 0.000 0.000 0.000 0.964 0.036 0.000 0.000 0.000 0.000 0.000 0.000 1.000

Weighted C.S. 5.124 1.374 0.180 0.012 0.000 0.000 0.000 6.690No. of bridges 73 23 4 0 0 0 0

Using deterministic approach, one can get average condition rating of corresponding bridge network of same class which is under study and illustration of

deterioration of 100 new bridges by using P is shown as below. Calculation explained above is carried out by using excel programming. Table 8 continued...


state 7.000 6.000 5.000 4.000 3.000 2.000 1.000 0.210 0.350 0.277 0.124 0.033 0.005 0.000 0.000 0.245 0.387 0.257 0.092 0.017 0.002 0.000 0.000 0.322 0.407 0.213 0.053 0.005

year 5 0.000 0.000 0.000 0.418 0.413 0.150 0.019 0.000 0.000 0.000 0.000 0.571 0.373 0.055 0.000 0.000 0.000 0.000 0.000 0.833 0.167 0.000 0.000 0.000 0.000 0.000 0.000 1.000



state 7.000 6.000 5.000 4.000 3.000 2.000 1.000 0.044 0.159 0.283 0.280 0.168 0.057 0.008 0.000 0.060 0.219 0.328 0.264 0.112 0.017 0.000 0.000 0.103 0.301 0.358 0.202 0.035

year 10 0.000 0.000 0.000 0.175 0.409 0.342 0.074 0.000 0.000 0.000 0.000 0.326 0.524 0.150 0.000 0.000 0.000 0.000 0.000 0.693 0.307 0.000 0.000 0.000 0.000 0.000 0.000 1.000

Weighted C.S. 0.309 0.957 1.414 1.122 0.505 0.114 0.008 4.429

No. of bridges 4 16 28 28 17 6 1

TeCHNICal PaPeRs



state 7.000 6.000 5.000 4.000 3.000 2.000 1.000 0.000 0.001 0.012 0.064 0.218 0.462 0.242 0.000 0.000 0.005 0.037 0.175 0.484 0.299 0.000 0.000 0.001 0.018 0.128 0.489 0.364

year 30 0.000 0.000 0.000 0.005 0.079 0.472 0.443 0.000 0.000 0.000 0.000 0.035 0.426 0.539 0.000 0.000 0.000 0.000 0.000 0.333 0.667 0.000 0.000 0.000 0.000 0.000 0.000 1.000

Weighted C.S. 0.001 0.008 0.061 0.255 0.654 0.924 0.242 2.145

No. of bridges 0 0 1 6 22 47 24 Condition State (C.S.) of Bridges average

Condition state

7.000 6.000 5.000 4.000 3.000 2.000 1.000 0.000 0.000 0.000 0.003 0.040 0.377 0.580 0.000 0.000 0.000 0.002 0.027 0.347 0.624 0.000 0.000 0.000 0.001 0.018 0.313 0.669

year 50 0.000 0.000 0.000 0.000 0.010 0.272 0.718 0.000 0.000 0.000 0.000 0.004 0.223 0.773 0.000 0.000 0.000 0.000 0.000 0.160 0.840 0.000 0.000 0.000 0.000 0.000 0.000 1.000



state 7.000 6.000 5.000 4.000 3.000 2.000 1.000 0.000 0.000 0.000 0.001 0.014 0.283 0.701 0.000 0.000 0.000 0.000 0.010 0.255 0.735 0.000 0.000 0.000 0.000 0.006 0.226 0.767

year 60 0.000 0.000 0.000 0.000 0.003 0.193 0.803 0.000 0.000 0.000 0.000 0.001 0.157 0.842 0.000 0.000 0.000 0.000 0.000 0.111 0.889 0.000 0.000 0.000 0.000 0.000 0.000 1.000

Weighted C.S. 0.000 0.000 0.000 0.002 0.043 0.567 0.701 1.314

No. of bridges 0 0 0 0 1 28 71

Predictions of condition ratings for bridge class under study arrived from one step TPM are

summarized in Table 9.

TeCHNICal PaPeRs


Table 9 year Wise expected Deterioration Values Calculated using Deterministic approach

Time (years)

Condition Rating

Time (years)

Condition Rating

0 7 35 1.905 5.56 40 1.72

10 4.43 45 1.5815 3.57 50 1.4720 2.94 55 1.3825 2.48 60 1.3130 2.15 65 1.26

Graphical presentation of the bridge deterioration over the age in years is shown in the Fig. 1.

Fig. 1 Bridge Deterioration Curve

In the absence of detailed data base and appropriate system required for comprehensive inspection by dedicated staff/resource the authentic source of database is the masonry register of PWD and proforma given by IRC:SP:35. On the basis of comments made by previously appointed field officials, one can’t describe condition state more than 7 as explained in Table 1. Bridge structure has large number of elements, its importance w.r.t. priority of repair and life varies and hence the model developed by authors considering 7 condition states can’t accommodate complexities involved in its deterioration process. In absence of proper maintenance management system,

this model helps to prioritize the repair/rehabilitation policy for the bridges. According to the life span of bridge milestone for repair/rehabilitation can be decided based on deterioration curve.

6 CoNClusIoN1. The Indian bridge construction industry is under

developed there is a need to develop the bridge maintenance prediction and rehabilitation methodologies in systems.

2. The MDP model described for BMMS have many critical issues such as development of transition matrices. The Markovian assumption is satisfactory for predicting the bridge deterioration.

3. The study shows that bridge data base satisfies Markovian property hence MDP model is appropriate to predict future condition states of deteriorating bridges.

4. It is observed that there is no significant difference between actual frequency of data base and Poisson’s distribution. It indicates Poisson’s distribution can be used to generalize transition probability matrices if there is lack of data base.

5. The Poisson’s distribution is found to be appropriate for generalization of deterioration of sample data base.

6. The Markovian model presented used as a tool for knowing aging process of bridge system.

7. By using deterministic approach average condition rating of a particular class of bridges under study is obtained by summation of weighted probabilities.

8. Using deterministic approach by knowing successive transition probability matrices, one can identify number of bridges in respective condition state for every year as shown in Table 8.

9. Condition rating of bridge system deteriorates gradually from 7.0 to 1.26 in the age span of 65 years.

TeCHNICal PaPeRs


10. Bridge deterioration curve is very useful to set milestones for repair/rehabilitation decisions according to predicted condition states. Also it is useful to prioritize bridge networks w.r.t. its type and age.

ReFeReNCes1. Allen, G. and Mc Keel, W. T. (1989), “Development of

Performance and Deterioration Curves as a Rational Basis for Structures Maintenance Management System”, Virginia Transportation Research Council, Charlottesville, Va.

2. Butt, A. A., Shahin, M. Y., Carpenter S. H. and Carnahen, J. V. (1994), “Application of Markov Processes to Pavement Management Systems at Network Level”, Third International Conference on Managing Pavements, 159 – 172.

3. IRC:SP:35-1990, “Guidelines For Inspection and Maintenance of Bridges,” Indian Roads Congress.

4. Madamat, S., and Ibrahim, W. (1995), ‘‘Poisson Regression Models of Infrastructure Transition Probabilities”, ASCE J. of Tran. Engg., 121 (3), 267- 272.

5. Panel Discussion (1990) on “Bridge Management System”, Indian Roads Congress, Journal Vol. 51 (2), 387-402.

6. Prapannachari, S. and Kaistha, S. K. (1994), “Comparative Study of Bridge Maintenance and Management System Adopted by Various Country Councils in UK with Special Reference to Their Applicability in India”, Indian Roads Congress, Journal Vol. – 55(3), 519 – 561.

7. Scherer, W. and Glagola, D. (1994),‘‘Markovian Models for Pavement Maintenance Management.” ASCE J. of Tran. Engg., 120 (1), 37-51.

8. Technical Circular for “Inspection of Bridges” (1988), Government of Maharashtra, PWD.

9. Text Book, “Probability Concepts in Engineering Planning and Design”, Vol. 2 Decision, risk and reliability revised Ed. John Willey & Sons, Singapore, By, Ang, A. and Tang, W. (1984).

10. Text Book, “Probability, Statistics and Decision for Civil Engineers”, McGraw-Hill Book Company, New York, By, Benjamin, J. and Cornell, A. (1970).

11. Yang, J. (2004), “Road Crack Condition Performance Modeling Using Recurrent Markov Chains and Artificial Neural Networks”, Thesis and Dissertations, Paper 1310.


RoaD saFeTy auDIT aND saFeTy ImPaCT assessmeNTS.K. ChaudhaRy*

* Assistant Engineer, Road Construction Deptt, Bihar Road Sub Division, Sakri, Darhanga, E-mail: [email protected]

1 INTRoDuCTIoN

Road safety audit is a formal procedure for independent assessment of the accident potential and likely safety performance of a specific design for a road or traffic scheme - whether new construction or an alteration to an existing road.

Road safety impact assessment is a formal procedure for independent assessment of the likely effects of proposed road or traffic schemes, or indeed other schemes that have substantial effects on road traffic, upon accident occurrence throughout the road network upon which traffic conditions may be affected by the schemes.

These two procedures enable the skills of road safety engineering and accident analysis to be used for the prevention of accidents on new or modified roads. They thus complement the use of these same skills to reduce the occurrence of accidents on existing roads by means of local safety schemes, in many cases in the form of low-cost measures.

This review aims to describe and illustrate the use of safety audits and safety impact assessment in helping to design and build safe road and traffic schemes, and at the planning stage in choosing which schemes to progress from among a range of possibilities.

Generally, roads are designed with a large number of criteria in mind, such as travel time, user comfort and convenience, fuel consumption, construction costs, environmental impact and objectives of urban or regional planning. Safety is one of the criteria, but is often implicitly assumed to be achieved by adhering to prescribed standards of alignment and layout for each element of the design. These standards are

indeed laid down with safety in mind, and some of these include explicit safety checklists (e.g. FGSV, 1988), but experience shows that adherence to them is not sufficient to ensure that a resulting design is free from avoidable hazardous features. Formal safety audit and safety impact assessment procedures ensure that independent expertise is used to make explicit the safety implications of an entire design and, in doing so, lead to safer designs of both new and modified roads.

Both procedures have strong contributions to make to rational and effective decision-making when considering alternative options, and safety audit is important to the achievement of a safe design for a chosen alternative. The two procedures are complementary - the aim is similar and the difference is in scope and timing.

The scope of safety audit is usually confined to an individual road scheme, which may be a new road or modification to an existing road. The basis for safety audit is the application of safety principles to the design of a new or a modified road section to prevent future accidents occurring or to reduce their severity. The procedure is usually carried out at one or all of five stages in carrying out a scheme: feasibility study, draft design, detailed design, pre-opening and a few months after opening. An essential element of the process is that it is carried out independently of the design team. It should be undertaken by a team of people who have experience and up-to-date expertise in road safety engineering and accident investigation.

The scope of safety impact assessment is dependent on the scale of the schemes being considered. For small-scale schemes, the impact of change can usually

TeCHNICal PaPeRs


be expected to be confined largely within the scheme itself. In this situation safety impact assessment and safety audit share many procedural characteristics. For larger schemes, the impact on accident occurrence can be expected to be felt over a larger part of the road network. In that case, the impact may be estimated using a scenario technique. By considering different road types, the corresponding values of relevant safety indicators and the forecast traffic volumes, the impact on accident occurrence can be estimated for different alternatives.

2 RoaD saFeTy auDITs

Road safety audit is a formal procedure for independent assessment of the accident potential and likely safety performance of a specific design for a road or traffic scheme, whether new construction or an alteration to an existing road [Eugene M. Wilson]. The procedures enable the skills of road safety engineering and accident analysis to be used for the prevention of accidents on new or modified roads. The internationally accepted definition of an RSA as used from The Canadian Road Safety Audit Guide [NCHRP] and is as follows: “An RSA is a formal and independent safety performance review of a road transportation project by an experienced team of safety specialists, addressing the safety of all road users.”

In safety audits “The main objective is to ensure that all new highway schemes operate as safely as is practicable. This means that safety should be considered throughout the whole preparation and construction of any project”.

2.1 MoreSpecificAimsAre

● To minimise the number and severity of accidents that will occur on the new or modified road;

● To avoid the possibility of the scheme giving rise to accidents elsewhere in the road network; and

● To enable all kinds of users of the new or modified road to perceive clearly how to use it safely.

Whatever the reason for the scheme, a safety audit always begins with a road design. An audit is intended to identify potential road safety problems by looking at the scheme as if through the eyes of the potential users of all kinds, and to make suggestions for solving these problems by applying the principles of road safety engineering. This means that an audit goes much farther than just assessing whether or not the relevant design standards are properly applied.

By minimizing at the design stage the risk of accidents during the lifetime of a road scheme, there is less likelihood of having to take accident remedial measures later, and the whole-life cost of the scheme can be reduced.

Road safety audit is an important means for paying explicit attention to road safety during the design of road schemes. This explicit attention should help everyone involved in making decisions regarding changes to road infrastructure to assess the safety implications of the many choices that arise during the design process, and thus increase the road safety awareness of infrastructure planners, designers and authorities.

3 Rsa meTHoDoloGy

For carrying out RSA in a systematic and impartial way, it is essential to follow a rigorous procedure [Prof. P.K. Sikdar and Dr. Nishi Mittal]. The four key elements which makes RSA most productive are:

I. Selections of projects for audit

II. Role of different organization in RSA

III. Team selection

IV. Audit organization

I. selections of Projects for audit

Road safety audits are applicable to all types of road projects, to all types of roads and to all existingroads, with the possible exception of

TeCHNICal PaPeRs


routine maintenance where the line marking is not being altered. A projectas small as a new school crossing or set of road humps, or as large as a major new freeway can benefit from aroad safety audit.

Road safety audit Can be Conducted on Road Projects as Diverse as

a) New freeway.

b) Major divided roads.

c) Pedestrian and bicycle routes.

d) Deviated local roads near major projects.

e) Local area traffic management schemes and their component parts.

f) Signal upgrading.

II. Role of Different organization in Rsa

RSA is based on the principal of an independent review. The process reveals that three parties will be involved in this process – Client, Designer and Auditor.

Role of Designer : Designer is responsible for planning/designing the project. Designer bears the responsibility for ensuring that a road safety audit is conducted and that the necessary measures are agreed on the basis of the auditor's recommendations and/or the client's decisions. The designer is also responsible for ensuring that the audit input information is unambiguously defined and that all circumstances are described in an easily understood manner.

Role of Client : Client is one who allots the project to the designer and owns the project. It is the task of the client to arbitrate in cases where the designer and

auditor disagree. The role of the client thus to:

● Select an appropriate audit team

● Provide all the relevant and necessary documents

● Hold a commencement meeting with auditor and designer

Role of auditor : Auditor’s responsibility is to carefully review the presented project material in its entirely, in the light best road safety expertise and from the viewpoints of all relevant road users. Auditor also indicates all circumstances that cause misgivings concerning road safety. Persons designated as road safety auditors shall have experience of road accident analysis. Auditors must be familiar with road planning, designing and construction work and must undertake to keep their expertise up to date. Auditors should work within the terms of reference. They should comment only on the safety implications of schemes and provide constructive recommendations as to how any potential difficulties can be resolved

III. Team selection

For large or significant projects, it is likely to have at least two members in the audit team, but not more than four members. For small projects ,single team member will be sufficient. One of the team member should be nominated as RSA manager. The one essential ingredient in RSA team is road safety engineering experience. It is also better to include local experienced people.

IV. audit organization

Practically twooptions are there for conducting a road safety audit:

● Audit by specialist auditors.

TeCHNICal PaPeRs


● Audit by those within the original design team or by any other road designers.

4 sTaGes IN Rsa

There are five stages at which a road safety audit can be conducted, regardless of the size and nature of a project. They are:

a) The feasibility stage.

b) The draft design stage.

c) The detailed design stage.

d) The pre-opening stage and

e) An audit of an existing road

5 oRGaNIzING aND CaRRyING ouT aN auDIT

The process of safety audit as applied to an individual road scheme can be seen as taking place at up to five stages, some of which can be combined for smaller schemes:

The feasibility stage - During this stage, the nature and extent of the scheme are assessed, and the starting points for the actual design are determined, such as route options, the relevant design standards, the relationship of the scheme to the existing road network, the number and type of intersections, and whether or not any new road is to be open to all kinds of traffic.

The draft design stage - Horizontal and vertical alignments and junction layout are broadly determined. At the completion of this stage, the design should be well enough established so that, if necessary, decisions can be made about land acquisition.

The detailed design stage - Layout, signing, marking, lighting, other roadside equipment and landscaping are determined.

The pre-opening stage - Immediately before the opening, a new or modified road should be driven, cycled and walked. It is advisable to do this under different conditions such as darkness and bad weather.

Monitoring of the road in use - When a new or improved road has been in operation for a few months, it is possible to assess whether it is being used as intended and whether any adjustments to the design are required in the light of the actual behaviour of the users.

Checklists have been designed for use during each stage of auditing. In practice, these checklists have proved very useful as reminders for the auditors, but there is also a risk that they are used too blindly as recipes without sufficient consideration for individual situations. What is required is a combination of judgement, skill and systematic working.

The essence of road safety audit is that it is carried out by auditors who are independent of the design team, have expertise in both highway design and road safety, and are properly trained and experienced in carrying out audits. This means that not only must they possess sufficient specialized professional knowledge and have the required experience, but they must also possess the communication skills necessary to present audit results constructively and encourage a positive response to them from the design team. Experience has shown that it is preferable to hire a small auditing team rather than a single auditor. The members of an auditing team can jointly offer more skills than an individual, and a team can operate its own system of checks and balances and thus be less susceptible to its assessments being swayed by personal preferences.

The results of audit should be documented and reported at each stage to the design team and in turn to the client for the scheme. They will usually include recommendations for improvements to the design. There is much to be said for linking a form of certification to the entire auditing process, and

TeCHNICal PaPeRs


having the audit results made public so that citizens, prospective users of the new or modified road, and other interested parties can make informed contributions to further decision-making. Whether this can be done or not depends greatly on the way in which the decision-making process relating to the scheme is organized. It is therefore, impossible to give a generally applicable rule in this regard.

The conduct of safety audits can sometimes lead to tensions between the audit team, the design team and the client for the scheme. What is necessary from the start, therefore, is to create a sufficiently solid, formal basis (whether or not anchored in law) that enables safety audits to be carried out successfully and the recommendations based on the audits to be implemented. There also needs to be commitment to the procedures on the part of the organizations involved. The procedures should include arrangements for dealing with situations in which the design team and the audit team are nevertheless at odds about carrying out the audit recommendations. What is required in these cases is a decision by the client for the scheme, and this may be assisted by some form of arbitration.

6 saFeTy auDIT aND exIsTING RoaDs

The development of safety audit for road and traffic schemes, and especially the fifth stage of monitoring the operation of such schemes after they have been open to traffic for some months, raises the question of the role of safety audit or analogous safety checking in respect of existing roads. There is a prima facie case that an independent assessment of conditions on an existing road would be likely to reveal deficiencies indicating scope for cost-effective measures for accident prevention additional to the accident remedial measures that are routinely identified by investigation of accident occurrence. Yet the task of checking all existing roads is demanding in terms of scarce resources of expertise.

This issue has been investigated in France by means of a pilot study covering nearly 2,000 km of roads ranging from motorways to local roads. The results

provide useful indications concerning complimentarily between safety checking and accident analysis, the range of deficiencies which it is practicable for the checking to cover, and ways of putting road sections of different kinds into an order of priority for checking during the many years it is likely to take to cover the whole network.

7 RoaD saFeTy auDIT IN INDIaIn India, at present there is no formal requirement for road safety audits to be undertaken. However, India has also started realizing the importance of road safety audits. It is because of Ministry of Road Transport and Highways sponsored the project on “Development of Safety Audit Methodology for Existing Roadway Sections” to Central Road Research Institute in April 2002. The National Highway Authority of India entrusted CRRI to carry out RSA of engineering design for construction packages under TNHP(8 packages) and GNTRIP (7 packages) on NH-2 . The total length of these 15 packages was about 900 km which was the longest road project for which RSA has been carried out in the world. Also, first RSA was carried out again by CRRI in 2000 on Indore Bypass. It is understood that the entire NHDP will be subjected to RSA as part of its implementation. However it is recognized that RSA are to be under taken all types of roads.

On 11 May 2011, the Decade of Action for Road Safety 2011-2020 was launched in more than 100 countries including India, with one goal: to prevent five million road traffic deaths globally by 2020. Moving from the Global Plan for the Decade to national action, many countries have taken measures towards improving road safety, either by developing national plans for the Decade (e.g. Australia, Mexico, the Philippines); introducing new laws (e.g. Chile, China, France, Honduras); or increasing enforcement of existing legislation (e.g. Brazil, Cambodia, the Russian Federation), among other concrete actions. The recent UN General Assembly resolution on global road safety sponsored by more than 80 countries gives further impetus to the Decade by calling on countries to implement road safety activities in each of the five pillars of the Global Plan.

TeCHNICal PaPeRs


8 CosTs oF CoNDuCTING RoaD saFeTy auDITs

In the safety audit manual published by TNZ (1993), the cost of audits was divided into three categories: consultant fees, the client’s time to manage the audit, and costs associated with implementing recommendations that are adopted. The client’s time on a project averaged about 1day per audit. It is important to note that additional costs may result from changes to a project’s scope and schedule. RTA indicated that a safety audit of a new facility cost approximately the same as a geotechnical survey (FHWA Study Tour, 1997). Recent experience places the average cost of a conventional audit for small to mid-sized projects between $1,000 and $5,000 (Sabey,1993, Jordan, 1994, Pieples, 1999). TNZ found that fees range from NZ $ 1000 to $ 8000 (US $ 700 to $ 6000) with most falling in the NZ $ 3000 to $ 5000 (US $ 2000 to $ 3600) range (1993). The actual cost depends greatly on the size and complexity of the project and composition of the required audit team. Hamilton Associates estimate that audits add approximately 5 to 10 percent to design costs, or less than one-half of 1 percent to construction expenses (1998). These estimates are slightly higher than costs experienced to date for the MRDC project. AUSTROADS approximates that audits will add 4 to 10 percent to the road design costs (1994).

As design costs are roughly 5 to 6 percent of the project sum, the increase in total cost is usually quite small. On smaller projects (traffic calming or retrofits), the costs may be a higher percentage of the overall capital cost. Costs of redesign/rectification should be considered which will vary on a project-to-project basis. The cost of rectifying deficiencies depends on how early in the design process the problem is identified as well as the amount of time required to redesign the area.

9 RoaD saFeTy ImPaCT assessmeNT

Being able to estimate explicitly the impact on road safety that results from building new roads or making substantial modifications to the existing road infrastructure that alter the capacity of the road network

in a certain geographic area is of crucial importance if road safety is not to suffer unintentionally from such changes. The same applies to other schemes and developments that have substantial effects on the pattern of road traffic. The procedure that has been designed for this purpose is known as road safety impact assessment. This procedure is intended to be applied at the planning stage, often proceeding to a definite design for the scheme. Safety impact assessment thus precedes and complements the eventual safety audit of any specific design for the scheme. A parallel to these two procedures can be seen in the Strategic Environmental Impact Assessment and the ordinary Environmental Impact Assessment. The two procedures together first provide an estimate of the impact of possible schemes on safety for an entire geographic area at the strategic level and then follow this with an audit of the safety of the specific design of the chosen scheme. For smaller schemes, the two procedures can be combined by extending the feasibility stage of the safety audit to include the likely effects of the scheme on accident occurrence in the surrounding network.

The results of safety impact assessment should be considered in the planning process alongside other information relevant to decision-making about which schemes should be implemented, and thus improve the quality of such decision making.

10 CaRRyING ouT a saFeTy ImPaCT assessmeNT

A scenario method is used to carry out a safety impact assessment. The starting point is the existing road network, the current pattern of traffic on that network, and the level of reported road accidents there. It is helpful, though not essential, to have the information in a digital form within a geographic information system (GIS), as in the German system Euska . This information relates to a road network which is made up of roads of a number of types that have different road safety characteristics. Each road consists of junctions and stretches of road between the junctions, with associated traffic volumes, and numbers of

TeCHNICal PaPeRs


accidents and casualties. Alternative scenarios to this current situation are the possible changes being studied in respect of the physical infrastructure and the associated traffic volumes in the road network in the future. If, for example, a new road is to be added to the existing network, the traffic and transport models can be used to estimate what this will mean for the traffic volumes throughout the network in the future.

The central step is to interpret these changes in terms of the impacts they will have on the numbers of accidents and casualties. To accomplish this, what are needed are quantitative indicators of risk (such as casualty rates per million vehicle-km) for each type of road, supplemented if possible by corresponding indicators for each main type of junction. One way of obtaining such indicators is to estimate them at a national level and adjust those if necessary using data for the area in question. In addition, thought should be given to any expected changes over time in the level of risk for each type of road or junction. These kinds of information enable safety impacts to be estimated.

If the various data are accessible from a computer, calculations of safety impacts for a range of scenarios and comparisons between impacts of different scenarios can be made quite readily. The procedure can be adapted in order to help to identify what changes are needed in a given scenario in order to bring its safety impact within some target range.

When implementing this scenario technique it is important to bear in mind the quality of the information being used. It is also important for the information to be accessible in such a way that calculations for a range of scenarios can be elaborated at relatively modest costs within a short period of time. For this purpose, the traffic and transport models should be set up in such a way that a road safety impact assessment module to apply the relevant indicators of risk for future years can be linked up with them readily.

11 CosT-eFFeCTIVeNess

The cost-effectiveness of road safety audits and safety impact assessments are at present difficult to quantify

rigorously. Both techniques are relatively recent, and it is difficult to find well documented cases in which both the benefits and the costs of the procedures have been established, but there is nevertheless useful evidence of the cost-effectiveness of safety audit. Whereas it is not too difficult to assess the costs of carrying out either procedure, estimating the benefits requires an estimate to be made of difference in the accident costs occurring on schemes which have been subject to impact assessment and/or audit, compared with the costs on similar schemes which have not.

The main immediate benefits of the procedures will be accident savings. In principle however, there are other longer term and more broadly based potential benefits; these include not just the immediate accident savings on the schemes subjected to the procedures, but more generally, improvements to the management of design and construction, reduced whole-life cost of road schemes, the development of good safety engineering practice, the explicit recognition of the safety needs of road users, and the improvement of design standards for safety.

12 INTeRNaTIoNal sCeNaRIo oF quaNTIFICaTIoN oF RoaD saFeTy beNeFITs

As regards the quantification of the immediate road safety benefits, there has been some experience in the UK, Denmark, Australia and New Zealand, which can give a broad indication of the value of road safety.

In 1994 a study was undertaken in an English county in which two groups of matched schemes, one group having been audited and the other not, were compared. This study estimated that the audited schemes showed a saving of about 1 accident per site per year compared with the schemes which were not audited - a saving which represents an accident cost saving per scheme well in excess of the cost of auditing the schemes.

Estimates have also been made of the benefits to a local highway authority of applying road safety audits to all of its road schemes. The Lothian Regional Council (a former local highway authority in Scotland) which

TeCHNICal PaPeRs


had about 3,000 injury accident per year, estimated that the consistent application of road safety audits would give a 1 per cent accident saving, and that such a saving would represent a benefit to cost ratio of about 14:1. In New Zealand a potential benefit to cost ratio of 20 has been estimated for the application of road safety audit procedures (Transit New Zealand, 1993).

One way of forming a judgement about the likely cost-effectiveness of road safety audits in the absence of objective accident savings data, is to compare the costs of carrying out an audit with the economic cost of a single injury accident. It then becomes apparent how large an accident saving would be needed to cover the audit costs. A review of road safety audit practice estimated that an average of 25 hours of the time of professional road safety engineers was required to complete an audit; 21 per cent of schemes took less that 10 hours and 7 per cent took more than 40 hours. Audit costs were estimated to be in the range of from £ 100 to £ 6,000 (at 1993 prices). In the UK, the 1994 value of preventing an injury accident was £55,650 , so the actual cost of carrying out a relatively extensive audit is a fraction of the value of preventing a single injury accident. In Australia, each stage of an audit of a scheme typically costs between AUS $ 1,000 to AUS $ 4,000 depending on the size of the scheme.

It has to be borne in mind however, that the actual costs of safety audit are not only the costs involved in completing the audit itself. Having audited the scheme, it is necessary in those cases where a design change is recommended, to make the appropriate design changes. The extent of such changes depends upon the quality of the original design. In view mentioned above, some redesign was required in about half of the schemes audited. Although the actual cost of redesign varied considerably from scheme to scheme, it was estimated that redesign costs ranged from about 0.5 per cent of the cost for the larger schemes to about 3 per cent of the cost for the smaller schemes. Australian and New Zealand experience suggests that safety audit adds about 4 per cent to road design costs.

Even including the costs of both the audit and any

subsequent redesign, it is clear from these figures that the saving of only one injury accident will more than repay the cost of the audit and its redesign consequences. Both the actual costs of the audit process and the redesign costs were included in a study conducted in Denmark in which the usefulness of safety audits was assessed in cost-benefit terms by a panel of experts. To assess the safety benefits of the audit process, the auditors estimated to the satisfaction of the panel the number of accidents which would be expected on the schemes with and without the changes in design recommended by the audit. The total reduction on the 13 schemes was estimated to be 34.5 accidents per year involving 21.3 casualties. The time costs involved for those carrying out the audits and for the resulting redesign amounted to about 0.5 per cent of the scheme costs – the proportion being rather larger for the small schemes and considerably smaller for the larger schemes. Construction costs were estimated to increase by about 1 per cent as a result of the audit. The study therefore concluded that safety audit is very effective in cost-benefit terms.

13 ComPaRaTIVe aNalysIs oF RoaD saFeTy aND RoaD saFeTy ImPaCT assessmeNT

Road safety is a quality aspect of road traffic and this aspect has to be balanced with aspects like: level of service, access for destinations, environmental impact, costs etc., when it comes to decisions in what infrastructures projects to invest. The scope of safety audit is usually confined to an individual road project, which may be a new road or rehabilitation of an existing road. The basis for safety audit is the application of safety principles to the design of a new or a rehabilitated road section, to prevent frequent occurrence of accidents or to reduce their severity. The procedure is usually carried out at some or all the five stages namely, feasibility study, preliminary design, detailed design, pre-opening and a few months after opening. An essential element of the safety audit is that it is carried out independently of the design team. It should be undertaken by a team of people who have

TeCHNICal PaPeRs


experience and up-to-date expertise in road safety engineering and accident investigation. It is, thus, a valuable process that gives an unbiased view of safety issues with support from safety experts.

The scope of safety impact assessment is dependent on the scale of the projects being considered. For small-scale projects, the impact of change can usually be expected to be confined largely within the project itself. In this situation, safety impact assessment and safety audit share many procedural characteristics. For larger projects, the impact on accident occurrence can be expected to be felt over a larger part of the road network. In that case, the impact may be estimated using a scenario technique. RIAs can be made on a more strategic level and on an individual project or scheme level. For both levels different tools are developed.

14 CoNClusIoNs

1. The road safety implications of planning decisions and infrastructure projects need to be taken explicitly into account in general policy-making at Community, national and local levels. The purpose is to avoid the cost of any unnecessary future accident and casualty problems.

2. The RSA has been recognized as an important preventive tool for improving road safety in many developed countries and its potential benefits.

3. Road safety impact assessment procedures are designed to assess the likely effects of the scheme or transport planning decision on accident occurrence, injury and damage over the whole of the road network which will be affected.

4. Safety audit of a specific design for a new or modified road assesses the accident potential and likely safety performance of the design with a view to enabling the scheme to operate

as safely as is practicable by identifying and recommending any necessary changes to the design.

5. For both safety impact assessment and safety audit, the application of safety principles is achieved through formal audit procedures carried out by expertise independent of the planning or road infrastructure project design team. Literature shows that audit work is best carried out as a team task with the team having specialist expertise in the road safety engineering and accident investigation and prevention fields.

6. Mandatory and cost-beneficial safety audit procedures programmed at well defined stages during the planning, design and construction of road schemes have been used in the UK, Denmark, Australia and New Zealand for several years and have contributed to identifiable improvements in road safety. Experience has shown that on most schemes it is necessary to prevent only one injury accident to more than repay the cost of the audit itself and any consequential design changes.

7 The benefits of safety audits and safety impact assessment are in:

● minimising the risk of accidents occurring in the future as a result of planning decisions on new transport infrastructure schemes;

● reducing the risk of accidents occurring in the future as a result of unintended effects of the design of road schemes;

● Reducing the long-term costs associated with a planning decision or a road scheme;

● Enhancing the awareness of road safety needs among policy-makers and scheme designers.

TeCHNICal PaPeRs


15 suGGesTIoNs

1. The RSA should be undertaken by an independent safety specialist and this could either be a suitably qualified individual from the road safety unit within the roads authority or a consultant with relevant expertise.

2. Opportunities exist for including of road safety interventions at various stages of the project cycle and the planning, design, and construction process.

3. Establish a road safety unit within the road and highway planning division to carry out traffic crash prevention and reduction activities.

4. Raise awareness of the socioeconomic costs of road accidents in country and sector operations reports and studies.

5. Include road safety in major documents prepared during the project preparation, processing, and implementation phases.

ReFeReNCes1. AUSTROADS (1994) Road Safety Audit. Sydney:

AUSTROADS National Office.

2. CEC (1997) Promoting Road Safety in the EU: The Programme for 1997-2001, COM(97)131 Final, Brussels: Commission of the European Communities.

3. CRAFER, A. (1995) Review of Road Safety Audit Procedures. Occasional Paper. London: Institution of Highways and Transportation.

4. Danish Road Directorate (1993) Safety Audit Handbook. Copenhagen: Danish Road Directorate.

5. EEC (1985) Council Directive on Assessment of the Effects of Certain Public Andprivate Projects on the Environment, (85/337/EEC), Official Journal of the European Communities , L 175,40.

6. ETSC (1996) Low-Cost Road and Traffic Engineering Measures for Casualty Reduction, Brussels: European Transport Safety Council.

7. European Parliament and Council of the European Union (1996) Community Guidelines for the Development of the Trans-European Transport Network. Decision No. 692/96/EC, Official Journal of the European Communities, L 228, 39.

8. FGSV (1988) Richtlinienfur Die Anlage Von Strassen, Teil Knotenpunkte, Abschnitt1 (RAS-K-1), Koln: For schungsgesellschaftfur Strassenunde Verkehrswesen.

9. GDV (1997) Euska, Zeitgemasse Verkehrssicherheitsarbeit. Bonn: Institutfur Strassenverkehr - Gesamtverband der Deutscher Versicherungswirtschafte.

10. IHT (1996) Guidelines for Road Safety Audit. London: Institution of Highways and Transportation.

11. ITE (1994) Informational Report: Road Safety Audit. Committee 4S-7, July 1994.Washington DC: Institute of Transportation Engineers.

12. JORDAN, P.W. (1994) Road Safety Audit: the AUSTROADS Approach. Road and Transport Research 3(1), 4-11.

13. MACHU, C. (1996) A New Approach to Improved Road Safety: Safety Checking of Road Infrastructure in Proceedings of the FERSI International Conference “Road Safety in Europe”, Birmingham, September 1996. VTI Konferens7A(4), 19-28.

14. OECD (1994) Environmental Impact Assessment of Roads. Paris: Organisation for Economic Co-operation and Development.

15. OGDEN, K.W., JORDAN, P.W. (1993) Road Safety Audit: an Overview. In the proceedings of the Pacific Rim Transport Technology Conference, Seattle, July1993.

16. SCHELLING, A. (1995) Road Safety Audit, the Danish Experience. In the Proceedings of the FERSI International Conference Road Safety in Europe and Strategic Highway Research Program, Prague, September 1995. VTI Konferens 4A(4), 1-8.

17. Surrey County Council (1994) Road Safety Audit: an Investigation Intocasualty Savings. Kingston upon Thames: Surrey County Council.

18. Transit New Zealand (1992) Accident Countermeasures: Literature Review.

19. TNZ Research Report Number 10. Wellington: Transit New Zealand.

20. Wegman, F.C.M., Roszbach, R., Mulder, J.A.G., Schoon, C.C., Poppe, F. (1994) Road Safety Impact Assessment: RIA. Report R-94-20. Leidschendam: SWOVInstitute for Road Safety Research.

21. Wrisberg, J., Nilsson, P.K. (1996) Safety Audit in Denmark - a Cost-Effective Activity. Copenhagen: Danish Road Directorate.













INDIaN RoaDs CoNGRess Hq office : Kama Koti Marg, Sector-6, R.K. Puram, New Delhi-110022 (India)

Phone. : Secy. Genl.:+91(11) 26185303, Sectt.:+91(11) 26185315, 26185319, 26185273, 26716778 Fax: +91(11) 26183669, E-mail : [email protected], [email protected] Website : www.irc.org.in

satellite office : Jamnagar House, Shahjahan Road, New Delhi-110011 (India) Phone: Secy.Genl.: +91(11) 23386486, Sectt: +91(11) 23384543, 23385395, 23387140, 23386274

Fax: +91(11) 23381649 E-mail: [email protected], Website : www.irc.org.in

FoRm a-1, aPPlICaTIoN FoR membeRsHIP

I desire to become a Member/Student Member/Life Member/Individual Associate Member of the Indian Roads Congress and agree that, if admitted, I will be subject to the Memorandum, Rules, Regulations and the Bye-Laws of the Society for the time being in force or as they may hereafter be altered or added to according to Law, and I undertake to promote the objects and interests of the Society so far as lie in my power, provided that whenever, I shall signify in writing to the Secretary General that I am desirous of withdrawing from the Society, I shall, after the payment of any arrears which may be due from me, be free from the obligations above mentioned.

I give below full and accurate statement of my academic qualifications, experience, etc.:

(To be FIlleD IN CaPITal leTTeRs) 1. Name______________________________mIDDle Name_____________________lasT Name___________________ (Dr./mr./mrs./ms.) 2. DaTe oF bIRTH (Day)_________________(moNTH)____________________(yeaR) (attach proof) 3. PRoFessIoNal /eDuCaTIoNal qualIFICaTIoNs (attach copy of Degree/Diploma certificates) (i) (ii) 4. DesIGNaTIoN_________________________________________________________________________________________ 5. oRGaNIsaTIoN________________________________________________________________________________________ 6. PReseNT aDDRess*

PIN CoDe______________________

PeRmaNeNT aDDRess*________________________________________________________________________________

________________________________________________________________________PIN CoDe______________________

(mobile No.) (Phone No. with sTD Code) (off.) (Res.)___________________

(e-mail)_________________________________________________________________________________________________

(* Kindly mention where you wish to receive the correspondence. member to intimate any change in address in future)

7. bRIeF oF ReleVaNT exPeRIeNCe Period Post Held Field of Work exp.1. exp.2. 8. aReas oF INTeResT __________________________________________________________________________________ 9. TyPe oF membeRsHIP souGHT (Please tick √) Member (M) Student Member (SM) Life Member (LM)

Individual Associate Member (IAM) Individual Associate Life Member (IALM) 10. WHeaTHeR seeKING CoNVeRsIoN oF sTuDeNT membeRsHIP To ReGulaR membeRsHIP

(aPPlICable oNly FoR m.TeCH/ReseaRCH sCHoloRs) (y/N) 11. membeRsHIP oF oTHeR PRoFessIoNal oRGaNIsaTIoN / INsTITuTe__________________________________ 12. amouNT To be RemITTeD____________________________________________________________________________

(a) DD/CHeque* ____________________________________ DaTeD_______________________________________ (*subject to realization)

(b) oNlINe PaymeNT (if applying online) ____________________________________________________________ DaTeD_________________________________________________sIGNaTuRe____________________________________

------------------------------------------------------------------------------------------------------------------------------------------------------------------------ FoR use IN THe IRC oFFICe Roll Number_____________________________________

Receipt No.______________________________________________________ Amount Received_________________________________

Date____________________________________________________________Collection No.____________________________________


1. qualIFICaTIoN RequIReD FoR membeR/sTuDeNT membeR/ lIFe membeR / INDIVIDual assoCIaTe membeR & INDIVIDual assoCIaTe lIFe membeR

1.1. Graduation in Engineering in relevant field. 1.2. Corporate Members of the Institution of Engineers (India) 1.3. Persons holding such qualifications as are acceptable by the Institution of Engineers (India) for election to Corporate

Membership. 1.4. Diploma holders with 20 years experience or 10 years experience and in the rank of SDO/AE. 1.5. Such qualified Engineers who are or have been connected with road sector and such other persons being Scientist of

eminence in or being engaged in a responsible capacity in the sciences allied to road sector including construction and maintenance, as the case may be, can be elected member of the Society after the approval of the Executive Committee.

1.6. Such Persons, who, while not eligible to be Member/Life Members, are or have been in the opinion of the Executive Committee, engaged in the administration of roads or road transport or so engaged in business connected with construction and maintenance of roads or with the manufacture or sale of materials used in connection with the making or repairing of roads or of road transport vehicles or the operation thereof for more than 5 years may be elected Individual Associate Members of the Society.

1.7 (a) Students of Final Year of BE/B.Tech or M.Tech/Ph.D in engineering may be elected as Student Members (SM). (b) Student Member can convert their Student Membership to Regular Membership by paying Rs.200/- after obtaining

B.E./B.Tech. degree (c) Students of M.Tech/Ph.D can also become Life Member of IRC by paying a sum of Rs.17000/-.

2. Fee To be PaID aloNGWITH aPPlICaTIoN FoRm a-1

Member Individual Associate Member _______________________

India & SAARC Student Foreign Countries Member Countries

Rs. Rs. us $ Rs. Entrance fee 400 400 100 1500 Annual Subscription 1300 900 150 5000 Total 1700 1300 250 6500 Less:Rebate on 1st Subscription (-) 200 (-) 200 Nil (-) 500 Net amount payable(mandatory) 1500* 1100 250 6000

NoTe: The rebate is also allowed on subsequent subscription if paid before the 30th June of every year.

members can become life members. The compounding fees are given in the age Table given below. Individual associate members elected under Rule 5-f whose subscription are not in arrears may also compound their annual subscriptions by a single payment depending upon their age at the time and which shall be 3.2 times of compounding fee for members as given in the age Table for compounding for life membership. age shall be reckoned as that on the birthday nearest 1st april of the year commuting annual subscriptions. such members shall be called as Individual associate life members. members from foreign countries (other than saaRC countries) whose subscriptions are not in arrears may compound their future annual subscriptions by a single payment of us $ 600.

applicants for life membership applying directly are required to add Rs.1500/-* in addition to the fee payable as per age Table given below.

aGe Table FoR ComPouNDING FoR lIFe membeRsHIP _____________________________________________________________________________________________

Age Compounding Fee Age Compounding fee Age Compounding Fee Years Rs. Years Rs. Years Rs.

_____________________________________________________________________________________________ 21 17000 35 14000 49 9600 22 17000 36 14000 50 9600 23 17000 37 13400 51 8800 24 17000 38 13400 52 8800 25 16000 39 12800 53 8000 26 16000 40 12800 54 8000 27 15600 41 12200 55 7200 28 15600 42 12200 56 7200 29 15200 43 11600 57 6600 30 15200 44 11600 58 6600 31 14800 45 11000 59 6000 32 14800 46 11000 60 6000 33 14400 47 10400 and over 34 14400 48 10400

_______________________________________________________________________________________________ Demand Draft/Cheque, may be drawn in favour of “Secretary General, Indian Roads Congress” payable at New Delhi.

Documents

JAN-2014