Bold, audacious, gutsy. Doomed? Call it what you may but India’s plan to ramp up solar energy capacity more than 14-fold in six years – from 6.8 GW as of June 2016 to 100 GW in 2022 - is one of the world’s most ambitious solar programs yet. The target was announced at the G20 summit in November 2015, a few days before the climate summit in Paris. A year before this, the world’s biggest polluters, the US and China, signed a historic pact to cut back emissions. India declined to cut emissions, saying the developed world has had much longer to raise its standard of living by exploiting fossil fuels. However, it agreed to reduce the intensity of its emissions, that is emissions per GDP unit, by increasing its dependence on renewable energy (RE), particularly solar, considering India enjoys about 300 sunny days every year.
The government, thus, is an overdrive to meet its international commitment. Various state governments, particularly the power-deficient southern states, have embarked on a tender-issuing spree for utility-scale solar projects. Public sector banks, following the government’s cue, have committed 1.71 lakh crore to finance RE projects. As of now though private banks have surpassed PSBs in the amount sanctioned at 20,355 crore compared with the latter’s 17,237 crore.
To its credit, India has been a pioneer in using auctions as a means to grow a commercially viable renewable sector with minimal subsidies. Countries such as Germany, the US and China have grown their solar capacity mainly using tax incentives, surcharges, subsidies and generous feed-in tariffs, which have distorted the electricity market significantly, especially in Germany. Both Germany and US are now emulating India in adopting the tariff route for allocating projects.
According to Bridge to India, a RE consultancy, 35 tenders with an installation capacity of 15.5 GW were announced by India in 2015, a quantum leap from an annual capacity addition of 1GW for three consecutive years until 2014. Despite the leap, all tenders have seen a lot of traction, with aggressive bidding under 5/ kWh. ReNew Power, one of the biggest developers, recently won a NTPC-backed project in Telangana at a bid of 4.66/kWh. Indian corporates – Tata Power and Adani Power – too bagged projects in Karnataka and Uttar Pradesh at sub 5/kWh.
Basking in the sun
A solar park usually takes around three to six months to execute once the land has been acquired. ReNew Power, Tata, Adani and Azure Power, between them, operate around half of India’s total installed solar capacity of 6.8 GW. Geographically, more than half of the installed capacity in the country is concentrated in the states of Rajasthan, Gujarat and Tamil Nadu. Typically, the power generated by developers is sold to off-takers – state distribution companies (discoms) or PSUs such as NTPC which sell it to consumers – at a fixed price specified in a 25-year power purchase agreement (PPA).
Developers take a number of factors into account before quoting tariffs: capital cost incurred in developing the project including solar panels, photovoltaic (PV) modules, inverters, power control systems; credit worthiness of the off-taker; intensity of solar radiation and the capacity utilisation factor (CUF) i.e. ratio of actual generated power to installed capacity (18-20% in India); availability of land and availability of evacuation infrastructure connecting the project to the power grid.
In recent years, though, developers in India have been quoting record low tariffs, spurred by the plummeting prices of Chinese solar panels (~70% in five years), which constitute around 60% of the project cost (see: Cheap, Cheaper, Cheapest). Eyebrows were raised when Finland-based Fortum Energy quoted tariffs as low as 4.34/kWh for NTPC’s tender to set up a 70 MW solar project in Bhadla, Rajasthan in January 2016.
Inderpreet Wadhwa, founder, Azure Power, says 4.34/kWh was an aberration and that recent tenders have attracted higher bids of 4.6-4.8/kWh. And most feel tariffs will remain stable in the short term. Sumant Sinha, CEO, ReNew Power, echoes Wadhwa in saying that it’ll take a year and a half for tariffs to reduce further. He points out to the 10% reduction in tariff following the widespread adoption of solar trackers – devices that continuously align the panels towards the sun and at 2-3% of the project cost, increase the panel efficiency by 15-20%.
Vinay Rustagi, managing director, Bridge to India, however, pins the current low tariffs on a mix of growing volume, better manufacturing efficiencies and lower global commodity prices. “Technology improvement is only incremental. Module efficiency will, typically, increase from 16% to 16.5% annually. However, significant improvements in technology – say from 16% to 25% – will take 10-15 years,” he says.
Others feel the government aggressively wooing developers could bring down tariffs further. Kartik Shetty, analyst at India Ratings, says, “Tariffs could fall below 4/kWh over the medium term due to increasing competition.” Moreover, he adds, “Despite a tariff as low as 4.34/kWh, commercial viability would not be a concern, provided the developer is successful in executing the project at a lower capital cost and the off-takers honour the PPA in full.”
Going by the trend this far, solar tariffs have been falling, considering module prices usually follow the Swanson’s Law, named after the founder of American solar cell manufacturer SunPower, Richard Swanson. Swanson’s Law says that there is 20% reduction in the cost of photovoltaic (PV) panels with every doubling in global cell manufacturing capacity. And capacity is being added rapidly (see: Dragon on the Loose). According to an analysis by PVTech, Oct 2015 to March 2016 saw announcements of global photovoltaic manufacturing capacity addition reaching 49.4 GW, substantially higher than 16 GW the year before.
Here comes the eclipse
While the outlook for RE is promising, the sector in India will have to contend with pressing systemic issues that have the potential to put a spanner in the works. Solar projects are attractive to developers due to many reasons: relatively low land requirement compared with wind, low maintenance costs (1-2% of project cost) since they don’t have any moving parts, and long-term PPAs that assures off-take. To fund these projects, developers have been raising equity financing from domestic and, increasingly, international investors. A report by Bridge to India notes, “As the Indian solar market grows and project sizes increase, international utilities and IPPs with strong balance sheets and lower cost of capital are likely to play a greater role.”(see: Rush hour) This trend, it says, is already growing over the past one year. Debt financing, on the other hand, is being raised from private and public sector banks, NBFCs, issuances of corporate bonds, and in some cases, from international entities such as the Overseas Private Investment Corporation (OPIC), the US government’s development finance institution, and International Finance Corporation (IFC). ReNew raised about $250 million debt financing from the OPIC.
According to industry sources, the underlying assumption on which developers are bidding aggressively for projects is that they will be comfortably able to clear their project debts within 11-12 years (the typical loan tenure), and will be sitting on a cash cow thereafter.
But like many unbelievably good deals, is this too good to be true? Recently, NITI Aayog, the government’s planning body, warned that the RE plan may run into technical hurdles like grid stability and energy storage. India’s transmission infrastructure requires immediate upgradation, according to a 2013 report by FICCI and Booz titled Power Transmission: The Real Bottleneck, which observed that 30 of India’s key transmission lines are “severely overburdened”. A survey of solar CEOs by Bridge to India revealed that respondents were most concerned about transmission capacity and grid failure in achieving the target of 100 GW.
To address this, the government is attempting to upgrade infrastructure. It has rolled out power transmission contracts worth 108,000 crore in 2015. It has also mooted the idea of dedicated green energy corridors for inter-state transmission of RE-based electricity, but not much has been achieved this far. Recently, Tamil Nadu asked the central government to speed up construction of the corridor so it could sell excess RE-based electricity to other states.
Shetty of India Ratings says, “It is questionable if transmission and grid infrastructure will keep pace with the increasing capacity addition. While huge investment on this front has been committed, on the ground success has hardly been achieved.” He cites the example of Tamil Nadu, which saw large capacity addition in wind energy, however, the transmission infrastructure failed to keep pace with the increasing installation, resulting in wastage of power and losses for developers.
Of delays and discoms
Inadequate transmission infrastructure raises another major issue: counterparty risk. It, along with poor financial health, limits the discoms’ ability to off-take power. Most state discoms’ balance sheets are in deep red – thanks to poor collection rates, theft, heavy transmission losses and artificially low tariffs dictated by electoral politics.
Worryingly, state discoms are responsible for more than half the installed solar projects. According to Bridge to India, as of FY16, 3,506 MW out of the total installed capacity of 6,630 MW was under state policy i.e. 54%. Of the 12,558 MW of solar capacity that is still under construction, 60% of the capacity is under state policy.
Ketan Mehta, director, Rays Power Infra, an EPC contractor, says that states like Rajasthan are delaying payments by eight months while in Tamil Nadu’s case, the delay is up to a year. While, Mehta says, states often compensate developers with interest, it only adds to developers’ stress. Rustagi adds, “They (developers) build a working capital buffer to work with delayed payments. As for debt repayment, they assume in their projections that discoms will be late in payments by 3-6 months so even if there are delays, they can continue to meet their debt obligations.”
Little wonder then that tariffs for discom-backed projects tend to be significantly higher than ones backed by PSUs. Solar parks under the Jawaharlal Nehru National Solar Mission (JNNSM) attract lower bids since these already have land and transmission infrastructure in place, thereby reducing risk for the developer. As a result, bidding has been much fiercer for projects under the JNNSM, which is backed by NTPC. While successful bids for state projects have been in the range of around 5.1-5.5/kWh, those under the JNNSM have been in the range of 4.4-4.8/kWh.
According to Rahul Goswami, managing director, Greenstone Energy Advisors, investors are much more comfortable bidding for NTPC-backed projects, which have the central government’s implicit support. Banks also appear to charge higher interest rates for financing state-led projects. A FICCI report on solar energy financing notes, “Though the Centre has provided gross budgetary support of 486 crore towards payment security mechanism under JNNSM, giving some comfort to the lender, deteriorating financial health of the state utilities, with losses touching almost 82,000 crore make the lenders uncomfortable in financing state sponsored solar power projects. As a result, lenders tend to place solar projects in high-risk categories and that increases the cost of borrowing.” It, thus, moots the creation of a dedicated fund that supports any default by state utilities to mitigate the perceived risk of default. “This could be done through budgetary allocation from the National Clean Energy Fund,” it adds.
While Sinha doesn’t believe that banks charge different interest rates for national and state projects, Rustagi says public sector banks lack the skill sets to differentiate. “PSBs do not differentiate their rate based on whether it’s a NTPC or an Odisha project. The interest rate is primarily influenced by the developer’s reputation. If the off-taker is a particularly risky one, like the Tamil Nadu discom, they would rather not lend than at a higher rate,” says Rustagi. He adds, “Ideally, there are various factors that need to be taken into account: off-taker, viability, attractiveness of the tariff, sourcing of modules etc. but Indian PSBs are not sophisticated enough to differentiate their lending rates based on these parameters.”
The dismal state of discoms is already sparking some anxiety about retrospective renegotiation of PPAs. “The recent fall in solar tariff has put existing operational projects (especially in Rajasthan and Gujarat) at a renegotiation risk as these projects have signed 25-year power purchase agreements (PPA) at a higher tariff (10/kWh-12/kWh). A consistent decline in tariffs and the perilous position of the state electricity boards (SEBs) may lead to resistance from SEBs to honour off-take obligations, thereby resulting in renegotiation of PPAs,” notes the India Ratings report.
Fortunately, there is favourable precedent to thwart this possibility. Gujarat Urja Vikas Nigam (GUVN), in 2013, attempted to re-negotiate its PPA, saying that developers were reaping disproportionate profit from plummeting solar panel prices, after the contracts were signed. However, the Appellate Tribunal for Electricity rejected the appeal saying it had no merit. This has reassured investors who feared India would go the way of Spain and Greece in retrospectively amending feed-in tariffs or imposing retrospective taxes for RE.
However, there is an insidious and more dangerous possibility, which most industry reports or analysts don’t mention. “Utilities will not change tariffs. But they have the discretion to decline buying power in the name of stabilising the grid or maintaining the grid balance,” says Mehta about the occurence, technically referred to as curtailment.
“Curtailments can result when operators or utilities command wind and solar generators to reduce output to minimise transmission congestion or otherwise manage the system or achieve the optimal mix of resources. Curtailment of wind and solar resources typically occurs because of transmission congestion or lack of transmission access, but it can also occur for reasons such as excess generation during low load periods that could cause base load generators to reach minimum generation thresholds,” says a report by the US-based National RE Laboratory.
Curtailment is a very real risk that has international precedence in China, Germany and United States - countries that have much better contract enforcement as well as grid management. For instance, after a gale of wind energy projects were commissioned in Texas, utilities in the state declined to buy power as specified in the PPAs citing grid issues. As a result, RE PPAs in the US now have a clause specifying compensation to the developer in the event of curtailment.
In India, though, PPAs have no such clause and developers are completely exposed to curtailment risk, which is even more worrying since India’s transmission capacity is already overburdened. Rustagi of Bridge to India says, “Grid curtailment is one of the biggest operating risks for the solar sector and this risk will grow as the sector grows. In the current policy and legal framework, there is no protection for developers if the evacuation infrastructure is not sufficient. While developers are conscious of the risk, they are not pricing that in. They are bidding for projects assuming they can sell 100% of the power.”
Developers often conduct a load flow study of substations to figure out if the generated power will be off-taken and consumed. This gives them a better idea but curtailment will still be at the discoms’ discretion, which is a problem since the need for it often arises from the intermittency of solar power. For instance, the time of peak solar generation (12-2 pm) may not be the peak demand time for discoms since India is primarily an evening peak country. “There will inevitably be grid curtailment at such times,” says Rustagi. The curtailment risk is especially bigger for larger projects that are connected to the national grid, adds Mehta.
While wind has faced curtailment issues in the past, the solar industry experienced it’s first-ever curtailment in Tamil Nadu this month. The Solar Power Developers Association shot off a letter to discom Tamil Nadu Generation and Distribution Corporation citing a curtailment of up to 50-100% during peak generation periods. What makes matter worse is that the issue might not be just technical. Projects in Tamil Nadu typically attract higher tariffs because of the discom’s poor credit reputation. In such a scenario, discoms may create curtailment risk in their unwillingness to buy power at higher tariffs. “This is becoming a very important issue. So far we’ve had curtailment in wind in Rajasthan and Tamil Nadu. Now we have curtailment in solar in Tamil Nadu too. States are now backing down renewable energy. Sometimes they have genuine grid problems, sometimes they have payment issues. The government has to work very quickly to address this. There can be two ways of addressing this. Either 100% grid availability has to be ensured for renewable energy projects, or there has to be compensation for the power they’re not buying. But that isn’t happening,” says Sinha of ReNew Power.
It’s a vicious cycle, which if continues, could dent effective CUFs, thus posing a big risk to lenders as well. Shetty of India Ratings says that according to its sensitivity analysis, a state project at a tariff of 5.1/kWh (typical tariff quoted for state projects), a CUF of 19.25%, and a 75:25 debt to equity ratio, will yield an internal rate of return (IRR) of 12% and an average debt servicing coverage ratio (DSCR) of 1.33 throughout the tenure of the debt. However, if the Capacity Utilisation Factor CUF falls below 17%, the minimum DSCR (at any particular year) falls below 1, meaning that the developer won’t be able to service the debt fully.
Moreover, recent projects coming up in states like Telangana have a typical CUF of around 19-20%. If discoms curtail even 10% of power generated, developers will be unable to service debt. “If in one year, say there’s 2% curtailment; though it usually isn’t that low, it’s around 10-15%, it’s a straight loss of 10-15% revenue which can be very sizeable,” says Sinha.
Elsewhere, in the north-western Chinese provinces of Gansu and Xinjiang, solar projects have seen curtailment of 31% and 26% respectively. While the industry’s average PLF in India hovers around 18-19% compared with Germany’s 11-13%, with typical Return on Equity (RoE) at around 15-18% (14-15% at current tariffs), at such levels, Indian solar projects would undoubtedly become unviable. And if PSBs have an offtaker-agnostic lending policy, many of their loans have the potential to turn bad.
Mehta of Rays says, “At current below 5/kWh tariffs, things are very tight for developers.” Everything has to fall into place for them to reap around 15% RoE and to be able to service their debt: the project cost should be reasonable, land should be acquired close to the grid, and discoms will have to off-take power fully. Developers are currently walking on a razor’s edge, with wafer-thin margins, and it won’t take much to push them off the edge.
“Curtailment risk is not too far away in the future as many states are expected to achieve significant grid penetration in the next 1-2 years. Telangana is expected to add around 3 GW of solar power capacity by 2017 and is planning further allocations. With average estimated daytime demand of 6 GW in 2017, 50% of the state’s daytime demand is likely to be met from solar power. Karnataka and Andhra Pradesh are likely to face similar issues. But the most extreme case is Jharkhand where the average daytime power demand is less than 1 GW but the state has already tendered 1.2 GW of solar projects,” says Jasmeet Khurana, associate director, Bridge to India, in PVTech.
While the government is planning to upgrade transmission infrastructure through its green energy corridor program, such projects can typically take up to five years to become operational. “In comparison, solar projects become operational within 12-18 months. States with such high solar penetration are therefore expected to face significant grid curtailment,” adds Khurana.
So why are developers not demanding a curtailment clause in PPAs? Why are they not bidding more rationally, taking curtailment risk into account? Rustagi says it is sheer myopia on part of the developers who are falling over each other to win projects at any cost. “It is a race for winning projects rather than prudent long-term investing,” he adds. Sinha counters, “So far developers haven’t factored curtailment risk at all. How can you? It’s very hard to get an idea of how much curtailment will happen and where.”
Another explanation could be the financing pattern of the projects. Most of the projects have 30% equity and 70% debt. The equity has been funded by private investors and the debt by banks or development agencies. If the projections go haywire, it is the financier’s capital at risk, not the promoter, who has very little skin in the game.
It’s true that low tariffs are making investors a little queasy; having reached a level that could be deemed reckless. According to India Ratings, IRR for projects allocated in FY15 and FY16 will hover around 12-14%, much lower than historical returns of 20%. But Goswami says that while investors are not comfortable investing in projects at very low tariffs, they are comfortable with renewable as an asset class.
This comfort derived from herding might just change with the curtailment in Tamil Nadu. “Curtailment risk could have a substantial negative impact on the investment climate for the sector in the years to come. The government needs to prioritise investments into transmission infrastructure to pro-actively assuage concerns,” notes Khurana. If that doesn’t happen, the irrational exuberance the solar sector is displaying – in quoting rock-bottom tariffs without accounting for curtailment - might just throw up a spate of bank NPAs in the not-so-distant future, and sink investments made by international utilities and private equity investors. That is just the recipe for a sunstroke.