Setting the right subsidy rates can be challenging, as the actual cost of electricity production by clean energy sources is usually not disclosed by producers (IRENA 2017).

Tariffs that are too high lead to higher costs than are necessary. This is particularly a severe problem in India, given that many state electricity distribution companies bear massive losses.

Tariffs that are too low will lead to a lack of interest from the suppliers and not enough capacity being built up.

Auction or regulator-designed tariffs can only work as well as the distribution company is able to guarantee the payments it must make to power suppliers.

Feed-in-tariffs are fixed payments by distribution companies to solar power producers under long-term contracts. A critical component of these tariffs is that they are decided by a regulatory authority either at the central or the state level.

Feed-in tariffs vary by state, but Gujarat continues to have among the lowest feed-in tariffs in the country (Ministry of New and Renewable Energy, 2013).

Located on the western coast of India, Gujarat has historically led the development of solar power in the country. In 2015, it had the largest installed solar power capacity, and it is the second leading state in solar production (Central Statistics Office 2017). A key component of its success has been its feed-in tariff mechanism that offers a guaranteed rate to solar power producers that supply energy to the grid.

The early adoption of solar power by Gujarat can be viewed as a response to the feed-in tariff structure adopted by the state. The tariff is equally applied to all projects, reflects the lifetime costs of solar production, and is levelized.^[1] Tariffs are typically put in place for many years and vary based on the type of fuel used. For solar, tariffs are typically fixed at 25 years.

In Gujarat, two different tariff levels were decided on. The much higher price would run for 12 years (from 2010 to 2022), while the much lower price would run for the succeeding 13 years (from 2023 to 2036). This design indicated that the Gujarat Electricity Regulatory Commission anticipated significant cost declines. In Order No. 3 in 2015, a sliding tariff was put in place. For every fiscal year, tariffs will be reduced by 7% from their previous level.

While different state regulatory authorities can decide their feed-in tariffs, the Central Electricity Regulatory Commission also arrives at its own estimates. Outside of Gujarat, however, tariffs remain fixed for the entirety of the period for which they are applicable.

Tariffs decided by the Gujarat Electricity Regulatory Commission and the Central Electricity Regulatory Commission have been significantly decreasing over the years, roughly 66% from 2010 to 2017. As of 2017, the Central Electricity Regulatory Commission has decided to determine tariffs on a project-by-project basis instead of a generic tariff for solar PV plants.

An alternative way of arriving at estimates of a sensible tariff is to conduct what are called as reverse auctions. In such auctions, rather than buyers bidding for a single supplier, multiple suppliers bid for a single project. This mechanism introduces competition and pushes prices down toward production costs (Azuela et al. 2014, and IRENA 2018). It is at least plausible that suppliers are better informed of their costs and are thus in a better position to estimate what a tariff should be (Shrimali et al. 2015).

An auction process has quickly become a popular way to discover prices and assign supply for renewable electricity generation (IRENA 2013).

Reverse auctions are also carried out at the central and state levels. Typically, the auctions were carried out using a closed-envelope, pay-as-bid scheme with fully disclosed ceiling prices, so suppliers were bidding on discounts to these ceiling prices.

The Jawaharlal Nehru National Solar Mission implemented reverse auctions for 25-year fixed price contracts beginning in 2010. The first phase of the auctions carried out between 2010 and 2011 were of a 25-year duration power purchase agreement.  The total quantity to be auctioned during this phase was fixed at 1,000 megawatts (MW). This has proved remarkably successful, with capacity growing from 30 MW in March 2011 (Government of India, Ministry of Power, Central Electricity Authority 2017) to 24,000 MW in June 2018, and prices are now at the lower end of the global range (Prateek 2018).

A unique element to these auctions in India is payment for electricity generated through the solar plant is bundled with cheaper thermal power to reduce the impact of costlier solar-generated electricity on consumers. The bundles can serve as a guarantee to solar power producers and keep retail prices manageable. In the second phase of the auctions (starting 2014), bundling was replaced by a viability gap funding mechanism, which is a subsidy paid to the solar plant for the first few years of its operations (Azuela et al. 2014). Given the poor financial condition of state energy distribution companies, which are often the contract counterparty, they can be perceived as bad risks and thus bundling is often not feasible.

The design of auctions varies by state. In four states— Andhra Pradesh, Odisha, Rajasthan, and Tamil Nadu—the auction process was switched over to a lowest-bid scheme. In such schemes, the winner of the auction pays the lowest bid offered by auction participants (Azuela et al. 2014).

Auctions typically result in much lower prices than estimates done by the regulatory commission. Auctions under Phase II of the National Solar Mission had tariffs below the regulator-decided rates of even Gujarat.

Both instruments involve a high level of knowledge. Feed-in tariffs require the regulator to calculate financial and operating costs, capacity utilization factors, and the duration of the tariff.

Well-designed tariffs can work well if backed by a trustable distribution company. It is important to note that Gujarat arrived at a fairly unique dynamically declining tariff. While this tariff turned out to be higher than the bid amounts in the auctions, it still got close. In addition, being profitable helps attract suppliers. Auctions have turned out be successful and have delivered low prices, subject to the caveat that these bids do not reflect winners’ curse.

Auction design must contend with establishing a ceiling price, the total quantity to be auctioned, the number of suppliers, and the process of selecting suppliers. The regulatory burden under the auction is comparably lighter at the contract award stage, while higher at the contract execution stage, as suppliers may not build what they committed.

Politically, the push toward renewables is gaining strength, but resistance from coal mining-dependent regions should be expected.

Experience has shown that auctions are much more cost-effective in the short run, resulting in prices that are roughly half of those under the tariff. However, auctions could be problematic; bidders may be too optimistic or aggressive in their bidding behavior, which may affect the ability of the suppliers to honor their commitments. In addition, the total capacity auctioned is often less than the capacity put up for sale.

Feed-in tariffs are also unlikely to be efficient, as they require regulators to estimate not just the level of prices, but also the change in the level over time.

Auctions are gaining in popularity across India, while feed-in tariffs have only worked in one state (Gujarat), thus suggesting auctions are better able to reach the environmental target. Possibly the biggest hurdle appears to be the financial health of the state distribution companies.

In general, auctions appear better than a regulator-designed tariff in discovering prices, subject to the caveat that a winner’s curse (where the winning bid is higher the true value) may operate. They avoid the lag induced by periodic tariff revisions under a regulator-designed tariff. They almost certainly imply a lower informational burden on regulators. However, the burden shifts over to auction design and the follow up after the auction is carried out.

Politically, auctions are increasingly popular worldwide, as they are viewed to be relatively free from corruption or nepotism. Economically, solar is competitive with coal and natural gas. These factors will help auctions scale. The only hurdle politically appears to be how entrenched the suppliers of traditional fuels are in the political system.

Some concern exists over their ability to perform over time, or rather, on the ability of an auction procedure to capture non-contractible elements of a project’s design. Some of the economic literature on auctions suggests that cost overruns may not be avoidable, particularly for complex procurement projects. In such cases, auctions simply are not a good instrument to allocate projects (Bajari, McMillan, and Tadelis 2008 and Bajari, Houghton, and Tadelis 2014). In addition, if auctions are not held at regular intervals, intermittent episodes of limited deployment can occur (IRENA 2013), which hurts the ability to reach a pre-defined environmental target.

Auctions can possibly be scaled at a higher rate than regulator-designed tariffs, subject to the ability to design them well. In addition, having a credible counterparty in the auction (the authority purchasing power) will be necessary to have an auction mechanism scale up. Regulator-designed tariffs can be difficult to scale. As these tariffs have to reflect local concerns, at least at the state level, a common set of tariffs cannot be decided. By contrast, an auction format removes this responsibility from the state regulator and places it in the hands of firms who are best placed to use it. Auction formats will probably not change significantly across states. Auctions can also arrive at a project-based bid. For a regulator to do the same would be far too costly at any reasonable level of scale. Auctions are not a perfect solution, but they do have the ability to scale.

^[1] A levelized tariff includes the present value of building and operating a plant across its lifetime.

G. A. Azuela et al. 2014. Performance of Renewable Energy Auctions: Experience in Brazil, [the People's Republic of] China and India.World Bank Policy Research Working Paper 7062. Washington, D.C.: World Bank.

G. Shrimali et al. 2015. Reaching India’s Renewable Energy Targets: Effective Project Allocation Mechanisms.Climate Policy Initiative Report.

Government of India, Central Electricity Authority. 2017. Executive Summary: Power Sector, April 2017. New Delhi.

Government of India, Central Statistics Office, Ministry of Statistics and Programme Implementation. 2017.Energy Statistics: Twenty Fourth Issue. New Delhi.

Government of India, Ministry of New and Renewable Energy. 2013. All India Renewable Energy Regulatory and Policy Database.

International Renewable Energy Agency (IRENA). 2013.Renewable Energy Auctions in Developing Countries. Abu Dhabi: IRENA.

IRENA. 2017.Renewable Energy Auctions: Analysing 2016. Abu Dhabi: IRENA.

IRENA. 2018.Renewable Power Generation Costs in 2017. Abu Dhabi: IRENA.

S. Prateek. 2018.Breaking: Winning Bids of Rps2.85-2.87/kWh Quoted in MSEDCL’s 500 MW Wind Auction.MercomIndia.com. 6 March.