Bengaluru's tunnel road: A costly bet on toll, land and time

The contract amount considers only the toll amount and not the income from real estate and other sources that the bidder will get, say critics.

The state government has proposed an ambitious 16.74-km tunnel road between Hebbal and Central Silk Board in an attempt to increase the average speed of vehicles from 15 km/hour to 30 km/hour.

However, beneath the promise of providing a faster commute lies a complex financial structure in which the project depends heavily on public subsidies, long-term toll extraction, and the monetisation of the city’s five most valuable land parcels.

With an estimated cost of Rs 17,698 crore, the project is being executed under a modified build-own-operate-transfer (BOOT) model.

Initially, the government chose the hybrid annuity model (HAM), under which the government funds 40% of the cost upfront, and the rest in instalments, but the idea was dropped during the conceptual stage after only one company showed interest in the project.

On the other hand, half a dozen companies favoured the Engineering, Procurement and Construction (EPC) model, but the government was not keen to fund the project entirely.

Now, under a modified BOOT arrangement, the government will provide 40% of the project cost — approximately Rs 7,079 crore — as viability gap funding (VGF), while the remaining 60% will be borne by the private concessionaire. VGF is a grant designed to support capital-intensive public-private partnership (PPP) infrastructure projects that are economically essential but commercially unviable.

In addition to VGF, the government or the project implementing agency — Bengaluru Smart Infrastructure Ltd (B-SMILE) — will bear the costs and responsibility of acquiring land and shifting of utilities.

On its part, the concessionaire is expected to invest over Rs 10,600 crore over four years of the construction phase and recover the investment through a mix of revenues, including toll, commercial development rights, and other ancillary streams. The tender framework expects the concessionaire to raise up to 70% of Rs 10,600 crore as loans from the banks.

Selection of bidders

As per the tender documents, the selection of the concessionaire is based on the lowest Total Concession Value (TCV) quoted. TCV means the revenue the concessionaire expects to earn from tolls over the life of the project. The one who quotes the lowest wins.

To make the project attractive to bidders, the government has excluded lucrative ancillary revenues, such as real estate development at the city’s five landmark sites, parking fees, and advertising rights.

In July last year, B-SMILE floated the tenders for the project. After three extensions, the tender saw participation of four bidders.

During the technical evaluation, B-SMILE disqualified two bidders: One, Rail Vikas Nigam Ltd (RVNL), on the grounds that its joint venture (JV) partner, TLS Geetha Kameshwari, did not meet the technical criteria. Second, infrastructure major Dilip Buildcon was disqualified because tender clause 2.2.1(G) barred entities with a history of bridge or flyover collapses.

Lalbagh may see a huge construction at the intermediate station,
going by documents.

In the end, two companies made it to the finals.

Adani Enterprises — which had quoted a total concession value of Rs 22,267 crore for both packages — emerged as the lowest bidder. In the race, Vishwa Samudra Engineering Pvt Ltd emerged as the second-lowest, quoting a total of Rs 25,474 crore for both packages.

Critics say that this essentially means either B-SMILE underestimated the toll-collection capability or the bids were not truly competitive.

It can be noted that Vishwa Samudra’s technical qualification relied solely on the experience of Navayuga Engineering Company Limited (NECL), a company closely related to the former. However, NECL was the contractor of the Silkyara-Barkot tunnel in Uttarakhand, which saw a major accident in 2023. If B-SMILE had taken this into consideration, Adani would have been the sole bidder for the project.

Double trouble

Adani’s bid, although the lowest, stands several times higher than what the B-SMILE had anticipated. While B-SMILE’s share of investment will remain at 40% of the project cost, the concession period is likely to extend beyond 30 years, possibly up to 40 years, given the higher bidding price.

An extended concession period means the continued collection of toll charges and rights over five real estate parcels in prime areas, among other things.

While B-SMILE did not disclose its TCV estimates in the tender documents, a calculation of projected annual toll revenue — as mentioned in the concession framework presented in the Final DPR — indicates that the concessionaire could earn close to Rs 11,000 crore in toll revenue over thirty years.

For instance, the B-SMILE estimates toll revenue of Rs 716 crore for Package 1 and Rs 672 crore for Package 2 in 2031. These revenues are assumed to grow significantly over time, reaching nearly Rs 2,934 crore and Rs 2,722 crore per year by the late 2050s.

To account for inflation, the concession framework applies a discount rate of around 9.1% to future earnings. In effect, the toll revenues over the 30-year concession period — across two packages — amount to Rs 85,000 crore in absolute terms, and in TCV terms (i.e., at a discounted rate), they stand at around Rs 11,000 crore.

With Adani quoting Rs 22,267 crore, which is more than twice the B-SMILE’s estimate of Rs 11,000 crore (two packages combined), the concession period is expected to continue until the discounted toll revenue collected by the concessionnaire equals the quoted TCV. This also means the project ends in 30 years if toll revenue is strong, and the concession period can extend to 40 years if toll revenue falls short.

Key exclusion

One of the most consequential aspects of the modified BOOT model is what it leaves out. The total concession value explicitly excludes ancillary revenues, even though the concessionaire is entitled to them.

These include commercial development at five prime locations, advertising and naming rights, and parking and access fees. Official estimates indicate that the concessionaire will earn a total of Rs 300 crore every year from ancillary revenue, including Rs 80 crore from commercial development at intermediate stations. But this appears conservative.

With a Floor Space Index (FSI) of 5, each site — such as Palace Grounds, St John’s Hospital road, Lalbagh, Hebbal and Racecourse — could support a minimum of 15 floors, excluding five floors in the basement. Annual revenues could exceed Rs 500 crore, translating into nearly Rs 25,000 crore in Net Present Value (NPV) terms over the concession period. Yet, this entire revenue stream is excluded from TCV.

In Bengaluru, an FSI of 5 — which requires amendment to the revised master plan (RMP) — is generous. Now, a maximum of 3.25 FSI is provided on properties adjoining 100-foot roads. The last major attempt to raise the FSI to 4 along the Namma Metro corridor was withdrawn owing to public backlash against densifying the neighbourhoods.

Experts termed the plan “a perverse mobility solution”, stating that the TCV has taken into account partial revenue streams, but the more lucrative real estate revenue base — which generates predictable profits — is kept out.

“TCV does not capture the full financial picture. In effect, the metric used to select the concessionaire is misaligned with the project’s total economic value. The government stands to lose,” said independent mobility expert Satya Arikutharam.

Given that TCV excludes real estate revenues, Arikutharam said the government should have easily foreseen that the bidders would aim to maximise the concession period so as to retain control over the high-value real estate revenues. “The quoted TCV reflects this priority, i.e. the lowest bidder’s TCV cannot be recovered even in 40 years,” he said.

He also questioned the priorities. “By allowing real estate development in five places that are currently in the green zone, the government is adding additional traffic to the road network. Each site could host developments exceeding 2,50,000 sq ft. This may generate up to 2.5 lakh additional trips daily, compared to 1.25 lakh passenger car units expected to use the tunnel road,” he estimated. He added that the tunnel road is more of a real estate project than a mobility solution.