Well Foundations for Bridges are Obsolete

S.A.Reddi, Fellow Indian National Academy of Engineering


Bridge foundations are the most complicated and difficult to construct. Unexpected difficulties cause delays, extra costs and revision of designs due to altered situations. Loss of human lives was normal rather than exception. In the Sixties, more than 50 lives were lost due to the accident during pneumatic sinking operations for well foundations of Mahanadi bridge in Orissa. For Kali Bridge at Karwar in Karnataka, pneumatic sinking was required for inspection of the founding surface of the wells. Apart from delays and extra costs, large number of workers suffered caisson disease. Due to difficulties in well sinking, two contractors left the job and the third took more than 5 years to complete the well foundations. Alternative construction techniques and equipments have emerged. With the introduction of advanced piling equipment, large diameter piles up to 3.5 m dia are easily realized at a fast rate, with a significant reduction in cost and material quantities. Bridges elsewhere are now constructed with pile foundations.

Well Foundations for Bridges are Obsolete
Figure 1: Brahmaputra bridge foundations No. 17 & 18
Pre-cast RC bored piles of 2.5m diameter was first successfully realized in India for the old Thana Creek Bridge constructed in 1960s. For the 13 Km long Saudi–Baharain Causeway large diameter pre-cast pre-stressed bored piles were adopted. By 2005, piles of 2-3 m dia, upto 120m deep are extensively used in the rest of the world. These new techniques eliminate complicated weather-dependent and risky operations in water. They have reduced the delays considerably and minimized the technical and financial risks. The development of modern techniques has considerably reduced quantities of materials used for foundations as well as energy consumption and environmental impacts.

Well Foundation Problems at Brahmaputra Bridge at Tezpur

The bridge was more than 3 km long with 26 spans of 120 m and 2 shore spans of 70 m each, founded on 12m dia Wells. Sinking well No.2 to full depth was not possible due to bouldery strata Despite best efforts, the well only went down by 35.25 m after three seasons and at extra cost! The well was plugged at RL 32.075 m and 5 nos 1.5 m dia RCC bored piles (25-35m) were provided to anchor the well, one in the middle through the dredge hole and 4 outside at the four corners. Further a launching apron of crated boulder 3 m thick was laid making a circle of 60 m dia around the well, at RL 61.00. These extra works caused further delay in the comp- letion of the substructure works.

Well foundations on sloping rock: Brahmaputra Bridge at Jogighopa (2.28 km)

Well Foundations for Bridges are Obsolete
Figure 2: Pile foundations for Jamuna bridge
The wells of main span were 11 m x 17 m double ‘D’ type. Foundations 17 & 18 were resting on hard rock at steep incline of almost 1:1 slope. It was not possible to rest the foundations partially on two types of strata. Hence to found these wells, 1500 mm dia anchor piles, 12 nos for each foundation were provided through the body of the steining, extending to about 10 m below the cutting edge.

Due to the steep incline, part of the cutting edge was resting on the rock while the other parts were overhanging. In order to contain the bottom plug, two rows of jet grouted piles were introduced around the periphery of the well steining which acted as curtain wall (fig.1). 1500 dia piles also driven up to hard rock along the periphery through the steining. The completion of the project was extended by 3 years. Additional cost was several crores. Effective use of Pile Foundations - Jamuna River Bridge in Bangladesh (figs.2 &3).

Well Foundations for Bridges are Obsolete
Figure 3: Completed view of the Jamuna bridge

A 4.8 km long, four lane road bridge with 100 m spans was constructed in the 1990s on the Jamuna river (Brahmaputra in India). The foundation design was challenging. Very deep wells are extremely slow to construct, costly, increasing the total cost. Various alternatives were considered including caissons, driven precast piles and driven steel tubular piles. The only viable option was large diameter tubular steel piles driven at a rake (fig.2). The piles were fabricated in Korea, shipped to site and installed by hydraulic hammer. The diameter of the piles ranged between 2.5 and 3.15 m and the steel tubes were filled with concrete. Maximum pile length was 72 m below bed level. During one working season from October 1995 to June 1996 all the 121 main work piles plus two full scale trial piles were driven.

This optimization resulted in overall reduction in the bridge costs by more than 50%. This solution also reduced the use of resources (concrete and steel) considerably and was beneficial to environmental impact. The piles were installed in 8 months; the well foundations of three bridges across the same river constructed in India have taken 3-5 years each to complete.

Damages During Construction - Ganga Bridge at Bhagalpur (4.6 km)

The well foundations consist of single circular wells 11.6 m dia. The calculated maximum scour depth was 36 m below water level. The soil strata were sandy up to about 30 m followed by hard stiff clay. During construction, the wells started tilting and the problems continued right through the sinking. Despite extensive chiseling, the rate of sinking was painfully slow.
  • Well 2 - The founding level was 64.7 m below the water level. The rate of sinking through clay was about 1.5 - 2 cm/hr. 3500 crane hours were used to sink the well.
  • Well 17 - The well shifted by 1.86 m. The piers were to be retained at the original position; resulting in excessive moments in the well. To counter the moments counterweight was provided in the form of a dummy well sunk to a depth of 20 m in the adjacent area and connected to the main well through a common well cap.
  • Well 32 - The well shifted by 1150 mm. A similar solution as in well 17 was adopted.
  • Well 4 - During concreting of curb, sand leaked from the island and the entire curb tilted and sank by 4.5 m. A new sheet pile cofferdam had to be erected and a new well curb was cast. The total delay was one month.
  • Well 9 - the total height of the steining except last 2 m was completed with 7.5 m balance sinking. Due to presence of stiff clay, 8 m sump was made to facilitate sinking. After several weeks, the well suddenly jumped by about 9 m with the top of steining below water level. Work resumed after monsoon. A temporary RC cofferdam was constructed and the sunk well dewatered to expose and build up further steining. Time loss: about 6 months.
  • Wells 3 & 4 -Wells were sunk by about 44 and 37 m before the monsoon season in 1996. The wells were toppled due to scour and disappeared during the floods. Based on a number of trial bores well No.3 was found tilted along the bridge axis. Well No.4 was found on the upstream side along the direction of current. These wells weighing up to 9900 t could not be restored and were abandoned. New wells were cast and sunk by changing the span arrangement. Floating caissons were used.
  • Extra cost and time - The cost on completion was Rs.106 cr against accepted tender cost of Rs.55 cr! The time overrun was 5 years!

Tilts and Shifts in Well Foundations - Vasai Creek Bridges Near Mumbai

Bassein Creek road bridge near Mumbai (1970) faced problems of heavy tilting of the well foundations. Two of the foundations no.4 & 6 tilted very heavily and all attempt to correct the tile failed. The foundations were abandoned and the design of the bridge was changed to accommodate new foundations and longer spans. The project was delayed by six years with termination of the first contract, arbitration, litigation.

Despite previous histories of two bridges built across the same creek that faced problems with well foundations, the same were again adopted for another Vasai Creek Bridge. During construction, heavy tilting of wells was observed. The corrective measures for one well alone took almost two years delaying completion of the foundations; costing about Rs. 2 crores.

Sudden Jumping of Wells During Sinking

Sometimes the well sinks suddenly due to excessive sump or weak soil layer and the steining disappears below water level, making it difficult to continue further work on the well steining.

In one of the well foundations in a bridge across river Ganga, the total height of steining except last 2 m was concreted. The well was in the final stages of steining, with about 7.5 m to reach the founding level. As the well was stuck up in stiff clay, efforts were made to sink the well by creating a sump of about 8 m below the cutting edge. All of a sudden the well sank suddenly by about 9 m and the top of steining was below the water level by about 3.5 m. Rectification measures were very expensive and time consuming.

Ganga bridge at Varanasi: Very stiff clay was encountered at 25 m below and sinking of well foundations No. 3 and 5 was very difficult, did not move for three months. Then well No.3 jumped by several meters without any warning when two workers and one supervisor were taking sump sounding. The tragic accident killed all the three people. The well No.5 also jumped by about 5 m and was submerged in the water by 1m.

Artesian Conditions During Construction of Well Foundations
Nepal Bridge (Kohalpur / Mahakali Section )

Artesian conditions were encountered during soil investigations for the Shivganga bridge (8 spans of 32 m). At locations P-4 and P-5 artesian head of about 4.3 m was encountered at about 17 m below ground level. The well was redesigned with foundation terminating above the artesian layer, resulting in shallow foundations resting on clay. Due to founding the wells at shallow depth, it was necessary to provide adequate bed protection so as to prevent scour. The bed protection consisted of:
  • Upstream and downstream aprons
  • Cut-off walls, upstream & downstream
  • Concrete floor

An Expensive Solution Indeed !!

The completion was delayed by more than one season as the solution was based on an Expert Committee investigation and report. This led to delay in finalization of the designs and drawings for the foundation well and necessity of issuing variation orders to cover the items of cut-off walls and bed protection works which were not envisaged in the original contract.

Ganga Bridge, Patna

The 5.6 km long bridge comprises of 46 spans of 120 m each resting on 56 m deep well foundations (12m dia.). Two of the wells in the midstream (Nos. 41 and 45) encountered artesian conditions during the final stages of sinking There was continuous sand blowing filling the dredge hole to 5-6 m above the cutting edge. Months of efforts to sink the well proved futile. A technical advisory committee took about a year for arriving at a solution. Temporary steel cofferdam was built enveloping the well and an artificial head of about 6 m of water was created to counter act the sand bubbling. Delay: two years

Cracking of Well Steining During Construction

Cracking of well steining is one of the serious problems faced many times in the construction of well foundations, resulting in time and cost overruns. The causes are usually:
  1. Blasting, Dewatering
  2. Insufficient steining thickness
  3. Jumping due to excessive sump
  4. Sand blows
  5. Surcharge due to dumping dredge material close to well.
  6. Failure of cutting edges.
When such cracking occurs, at least one season is lost for the investigation, developing remedial measures, approvals of the same etc. In the last 45 years, the author is aware of more then 200 cases of bridges constructed by various construction agencies, where the dredge hole of well has to be filled with concrete due to cracks in steining.

Well Foundations for Bridges are Obsolete
Figure 4: Pasighat Bridge, AP
Figure 5: Pasighat Bridge, Boulder dredged during well sinking

During well sinking of Tapi Bridge, Maharashtra, hard strata was met. Due to blasting, extensive cracks developed in steinings New steinings had to be constructed inside the wells. The original contract period was four years. Attempts were made for five years to sink the wells. Work was suspended for five years for want of decision to revise the founding level. An expert committee ultimately recommended raising the foundation level of wells by more than 20m The bridge was completed after fourteen years. The contractor suffered losses due to the delays. The owner suffered substantial losses due to time overrun. Delay: 10 years.

Extraordinary Delays in Construction of Well Foundations Pasighat Bidge, Arunachal Pradesh, 703 m long

The project started in 1987 and the construction of well foundations continued for the next 20 years! As per the design, based on misleading soil data, six wells were to be sunk to about 50 m below bed. The actual strata met with during sinking were hard conglomerate with densely compacted and very large boulders (fig.6) were found right from the beginning of sinking. After 15 years of struggle to pneumatically sink the wells to RL – 50 m, the designed founding level was drastically raised by 22m in 2002.

Major Bridges (Worldwide) on Pile Foundations

Su Tong Bridge, China : The 6 km long Cable-stay bridge crosses Yangtze river near Shangai will carry a six lane highway with emergency lanes, with a record 1088 m main span and 300 m high concrete pylons. Each tower is supported on 131 cast-in-situ bored piles 120 m long and 2.7 m diameter. Due to strong currents, significant scour is expected around the foundations, and suitable scour protection is provided around the pylons. The central span has a clearance of 62 m for container ships to pass through. The bridge used 200,000 t of steel, 1 million cum of concrete.

Stonecutters Bridge, Hong Kong: The 1600 m long Stonecutters Bridge Hong Kong with cable-stay span of 1018 m is one of the longest in the world. The bridge is founded on 3.0 m dia piles, up to 90 m deep, socketted into rock. Bandra Worli Sea Link Mumbai: The sea link consists of 5.6 km long, 8-laned bridge with cable stayed portion of 600 m. The bridge is founded on 1.5 m diameter bored piles. Concrete for the piles is M50 grade and for pile caps is M60. Silica fume and fly ash are used for concrete preparation.

Analysis by the Federal Highway Administration (FHWA), USA

More than 100,000 bridges would be constructed during the next two decades. Foundations represent about 30% or more of the cost of the highway bridges. The predominant type of foundation system used for the highway bridges in the US is pile. Many bridges can tolerate significant magnitudes of a total and differential vertical settlement without becoming seriously over-stressed


The Indian Bridge Engineers are by and large mentally tuned to providing well foundations for bridges as a reflect action; whereas it is very necessary to analyze the comparative merits and demerits, construction time frame and cost of construction of bridges with well/pile foundations before finally choosing the type of foundation. An analysis of the history of well foundations during the last five decades indicates innumerable difficulties, delays and cost overruns in a majority of the cases. Realization of well foundations requires special skills and experience which are gradually dwindling.

Developments have taken place in respect of large diameter pile foundations as well as equipment for the same. The time and cost advantages of opting for pile foundations have been amply demonstrated worldwide and to a limited extent in India. Piles up to a diameter of 3m and depth of up to 120m have been realized for a number of major bridges worldwide, with cost saving of up to 40% when compared to well foundations.

There is currently no restriction in the IRC Code regarding use of pile foundation. However, many Owners impose restrictions in the tender documents, without any justification. The example cited above concerning the problems of well foundations amply justifies a second look on the choice of foundations. In fact, the use of well foundations for bridges should be an exception rather than the rule.

Chenab River Bridge at Akhnoor Near Jammu

The project was started in the early Seventies. A .scheme for a 231m long bridge with 5 spans (3x46+2x46.5) upstream of existing steel bridge was originally conceived. The scheme involved construction of five well foundations in the volatile Chenab River, to be sunk through difficult strata – hard conglomerate, in spite of insurmountable difficulties elsewhere under similar circumstances. Two successive contracts and 30 years later, the impossibility of sinking wells through such strata was realized.

Well Foundations for Bridges are Obsolete
Figure 6: Chenab Bridge at Akhnoor-Longest Span Cantilever PSC Bridge, eliminated wells

Based on lateral thinking, it was decided to abandon the partly sunk wells and go for a scheme with longer central span, eliminating the water foundations altogether. With a 160m central span, both the main pier foundations were located in the dry on the banks, resting on raft foundations. These foundations were completed in months instead of decades earlier in unsuccessful attempts to sink wells in water. The abutments consist of hollow box and piers consist of hollow rectangular section on raft foundations.

The superstructure was designed and constructed as a continuous cantilever of 280m length, with a central span of 160m (longest in India at the time). Two pairs of cantilever gantries were deployed. The bridge with the new layout eliminating well foundations was completed in 20 months

Other Records: The Chenab bridge deck was constructed with the shortest time cycle of 6 days consistently achieved for the construction of each pair of segments. This was made possible by an high early strength concrete which enabled pre-stressing at 60 hours after concreting. Fe500 steel reinforcement bars were used for the first time in India in a cantilever construction bridge. The huge Bearings with anchors were located among highly congested reinforcement; normal concrete placement, vibration was impossible. Special Conbextra Grout replaced normal High Strength Concrete below bearings. Self–compacting concrete (S.C.C) was used for the first time for concrete below the bearings.

Segments on both sides of the Pier were concreted simultaneously balancing the weights. When the 22nd segments were facing each other and the shuttering of the 23rd segment i.e. the linking segment was to have been placed there was no level difference and the levels matched on both tips to the nearest millimeter both in plan (centre line) as well as in elevation. This was possible because every day the levels were maintained by a team of surveyors with the help of total station. These levels were sent to the Design consultant who monitored these personally. In fact, after the concreting of each pair of segments the levels as actually measured and as envisaged by the designer fitted almost like a ‘T’. This proves that the parameters fixed by the Design consultant and the parameters as actually achieved during execution were complimenting each other. The cables were so placed that almost all the cables were straight and without any curve. Thus prestressing results were exactly as shown in approved drawings both in terms of extension and gauge pressure.

Well Foundations for Bridges are Obsolete
Figure 7: Chenab Bridge, Giant Bearings
Figure 8: Chenab Bridge Hydraulic Earthquake Dampers

The author was Value Engineering Consultant for the Fast Track Project.

NBMCW September 2010