Month: October 2018

India’s solar sector is growing rapidly – total installed capacity has risen from just 1 GW in 2012 to over 25 GW at present and is expected to go up by over 10 GW every year. Fast growth and concentration in water-stressed areas exposes solar to growing water risks – scarcity, rising cost, poor quality, conflict with other social and economic uses and environmental degradation. We estimate that 94% of solar capacity in India is exposed to medium-high level of water risk.

Figure: Estimated water consumption in the solar sector by state

Source: BRIDGE TO INDIA research; India Water Tool, World Resources InstituteNote: Water consumption has been estimated based on installed solar capacity as on June 30, 2018.

Solar panels need to be cleaned regularly as soiling due to accumulation of dust, dirt, pollution, bird-droppings etc. can cause generation losses of 3-6%. These ‘soiling losses’ need to be reduced to about 1% – requiring an average of two cleaning cycles per month – to produce power at a competitive price. That requires water consumption of about 0.1 m3/ MWh, only about 5% of requirement for thermal power plants. But the challenge of procuring this water is acute as solar projects are usually located in dry areas. We also found in our discussions with developers and O&M contractors that there is high level of variation in water usage. Wasteful use is particularly common in southern regions where water availability is relatively better.

We estimate that 60% of the water used in solar sector is sourced from ground through borewells while the remaining 40% comes from surface water sources such as rivers, canals and lakes. Ground water is preferred by the industry as it is almost free and is operationally expedient. While it requires specific regulatory permissions, there are various reports of illegal extraction. If surface water is used, procurement responsibility is generally outsourced to a local vendor, who supplies water through tankers.

It is estimated that cost of cleaning, which primarily depends on cost of water and labour, ranges from INR 42,000-105,000/ MW per year. Higher costs are observed in the dry states of Rajasthan and Gujarat. These costs are rising rapidly. In parts of Rajasthan, water cost has almost doubled in the last 3-4 years due to sharp increase in demand. Karnataka recently hiked tariff for water for industrial use by 100 times. Such drastic, unforeseen increases can substantially affect project economics since cleaning accounts for up to 35% of total O&M cost.

While awareness of water risk is improving slowly, we believe that the solar industry is still managing it on a reactionary basis. Fortunately, there are technological solutions to manage this risk. The two main options include waterless cleaning and anti-soiling coating for panels. Both technologies are now technically and commercially proven. They can help not only in reducing water use by 50-100% but also in increasing power generation output. Payback periods for these solutions can be as low as 2-3 years.

Practices such as wasteful usage and illegal extraction have been going unchecked as the government has waived off requirement for most environmental approvals and permits for solar projects. In absence of such checks and balances, it is desirable that the government and the industry work together to adopt a water usage code for sustainable long-term growth

RE projects are facing increasing financing challenges. Liquidity in the Indian financial system has dried up considerably pushing up cost of debt finance by 1.0-1.5% over last year. To make matters worse, we understand that most private banks and non-banking finance companies (NBFCs) are unwilling to finance RE projects at present. Tough financing conditions are expected to pose a formidable challenge for about 10,000 MW of RE projects reaching financial closure stages.

Financing difficulties are expected to persist at least until H1 2019;

Small IPPs and developers are likely to be the worst affected as lenders shut down most new business;

Expected impact on projects is completion delay of up to 3-6 months and/ or reduction in equity returns by about 1-2%;

Debt financing environment for RE projects had been relatively benign in the last few years. Despite rapid growth in capacity addition, project developers were able to raise total estimated debt of INR 420 billion (USD 5.8 billion) for the 12,000 MW of new utility scale capacity addition in 2017-18 relatively easily. Precise numbers are hard to obtain, but we estimate that more than 75% of the total requirement was met from domestic sources including Indian banks, financial institutions, NBFCs and local capital markets. Balance came mainly from international financing agencies (World Bank, ADB, IFC, KFW) and international capital markets. Despite project viability concerns in the face of falling tariffs, interest rates fell steadily to about 9.5%. Other terms and conditions including tenor also became progressively more favourable.

Figure: Indian interest rates

Source: BRIDGE TO INDIA research, www.investing.com

Private NBFCs – L&T Infrastructure Finance, IL&FS, Tata Cleantech Capital and Reliance Capital, amongst others – have been a key source of finance for the sector. We estimate that these NBFCs provide up to 20% of total debt funding for the sector. They play a vital role in the value chain by sanctioning loans quickly and aggressively down-selling to banks and FIs. A spate of bad news including concerns about asset quality, tightening liquidity and falling INR have led to crisis like conditions for NBFCs. This has, in turn, led the financial system to freeze and forced lenders to suddenly question viability of 20-year debt financing for RE projects.

We expect the financing challenges to persist for another 6-8 months, until general elections in H1 2019, at the minimum. Small IPPs and developers are likely to be the worst affected as lenders restrict new business to top-tier players. The news comes at a highly inconvenient time with developers already struggling with GST, safeguard duty, falling INR and rising commodity price risks.

How can the government help? MNRE is believed to be pursuing relaxation of priority sector lending norms for the sector but we think that is highly unlikely to happen. Instead, it should seek to address increasing concerns around DISCOM payment risk and grid curtailment.

I attended one of the world’s leading solar module industry events – PV ModuleTech in Penang, Malaysia – this week. There were about 200 participants from across the world including mainly all leading module manufacturers, various testing agencies and independent engineers. There were also a few project developers and investors such as Brookfield, Shell, 8minuteenergy, Clara and Mahindra.

The event was mainly technology-focused with most discussions ranging around prospects of new technologies – n-type, PERC, half-cut cells, IBC, HJT, multi-busbars, frameless, glass-glass. Growing scale and rising cost pressures are forcing rapid improvements in the technology landscape, which is getting further compounded by multiple module form factors, an innovation leap in materials – back-sheets, adhesives, polymers – and multiple design considerations. There has been a common perception in India that solar industry is highly commoditized with multi-crystalline modules accounting for over 98% market share. But these modules are turning obsolete – worldwide share has already fallen from over 70% in 2015 to less than 50% now. As for the developers, there seemed to be a feeling that their job is becoming difficult in trying to evaluate different technologies and picking the right one. Some developers mentioned that they have to run as many as 30 different project design combinations before settling on a final plan.

Another theme that came up in many discussions was that quality and reliability of modules has been a frequent problem across the world. Within just 5-7 years of operations, developers are having to replace faulty modules (delamination, higher degradation, cracks) and retrofitting of many solar plants is going to become unavoidable in the coming years. Independent engineers play a critical role in this regard by testing and certifying all key aspects of project design.

With India becoming one of the top three international markets, it seems high on the priority list for most module makers. India came up frequently in various sessions and the recurrent theme unsurprisingly was the Indian developers’ unremitting focus on price. There were suggestions that Indian developers don’t care about quality. While that is too broad a statement, the underlying message is clear – poor quality is going to pose a formidable challenge for Indian developers, investors as well as policy makers. The fact that Indian lenders, particularly public sector banks and FIs, typically have lax technical oversight standards does not help this situation.

It was heartening to see SECI attending the event and trying to get a grasp of new technologies. Participation from India also included some module makers (Adani, Jakson, Vikram, Waaree) and Mahindra but disappointingly, most major developers were absent.

Future of manufacturing in India? It remains bleak in our view. I did not note any serious interest by Indian or international manufacturer to set up high-volume integrated facilities. Most likely, Indian participation would continue to be restricted to downstream assembly. Even if trade barriers and other policy measures are successful in bringing some new investments in manufacturing, India would remain reliant on international technology expertise for years to come.

A recent report by PI Berlin, supported by MNRE, NISE, SECI and KFW, has highlighted major risks facing Indian PV projects – sub-optimal design, inadequate EPC contracts, poor installation and lack of proper maintenance. The study confirms longstanding concerns about poor quality in the Indian solar sector.

There are multiple reasons that explain quality problems in the solar sector in India. Fierce competition in both utility scale and rooftop solar markets has led developers and EPC players to cut costs and compromise project quality. Many developers have a short-term view as they want to sell projects soon after construction and hence, are not committed to delivering long-term performance. Formulation of robust quality standards and their stringent implementation can solve these problems. But the Indian government has been slow to react. A set of quality standards was introduced for the first time on September 5, 2017 but deadline for certification has been extended multiple times from April 16, 2018 to January 1, 2019.

Currently BIS standards for solar equipment are almost identical to IEC standards, which were formulated keeping in mind temperate climates and hence, are not appropriate for harsh climatic conditions in India. We believe that India needs its independent standards suited for local operating conditions. Moreover, there is a need for devising location-specific standards especially with regards to humidity, sand and UV radiation that vary widely across the country. Lead can be taken from a few coastal states, which specify an additional corrosion test, over and above BIS standards, in their solar tenders.

The onus of developing standards and implementation lies collectively with BIS, MNRE and tendering authorities. Although BIS and NISE have already started devising India specific standards, progress has been extremely slow. Given that the BIS committee on solar meets only once in six months and discussion over changes or introduction of new standards takes multiple rounds of such meetings, formulation of new standards typically takes 2-3 years.  Certification and final approval for manufacturers can take another year. Apart from procedural delays, this process suffers from lack of adequate infrastructure and know-how amongst officials. Given the rapid upscaling of the sector, there is a dire need to accelerate the pace at which these standards are formulated and implemented.

The importance of developing quality standards has also become critical in view of new applications and technologies (floating solar, mono-PERC, bifacial modules half cut cells, glass-glass modules) becoming more popular. These applications have their unique requirements – for instance, floating solar plants have higher risk of moisture ingress and corrosion as compared to ground-mounted power plants. It is an opportune time for BIS to devise specific standards for these applications.

Uttar Pradesh conducted auction for its 500 MW utility scale solar tender on 10 October 2018. Winning tariffs came at between INR 3.17-3.23/ kWh (USD cents 4.4). Capacity was won by NTPC (160 MW), Adani (100 MW), Sukhbir Agro, Eden, Solar Arise and Avaada (50 MW each) and Maheshwari Mining (40 MW).

The tender marks the first instance of NTPC participating in and winning a project auction;

DISCOMs are not keen to buy solar power bilaterally from NTPC as tariffs offered by private developers in auctions are significantly lower;

NTPC’s aggressive bidding in this tender is an effort to stay relevant in the solar sector;

The last such auction by Uttar Pradesh – for a 1,000 MW tender in July 2018 – saw winning tariffs of between INR 3.48-3.55/ kWh. This tender was subsequently cancelled because the DISCOMs felt that the tariffs were too high. Tariffs have fallen in the latest tender because of clarity about safeguard duties and steep fall in module prices in the intervening months. But the new tariffs are still over 30% higher than tariffs in the recent Gujarat auction. Might Uttar Pradesh cancel this tender too? We do not believe so. The state has lower radiation, higher land prices and much worse DISCOM ratings in comparison to Gujarat, so the difference is justified. Moreover, public sector giant, NTPC, is a major winner this time and any cancellation talk would be unthinkable.

Indeed, the most interesting facet of this tender is NTPC’s foray into project auctions. To the best of our knowledge, it is the first time that NTPC has participated in and won a competitive auction. The move marks a radical departure from the company’s usual business – where PPAs are negotiated bilaterally with the offtakers and it earns a cosy regulated return of 15%.

We believe that NTPC has been forced to abandon its preferred formula because of DISCOMs’ unwillingness to pay higher tariffs to the company. Private developers are able to outcompete the company with their aggressive financial engineering and procurement tactics. As a result, NTPC has had to cancel most of the projects tendered by it on an EPC basis. The company has an ambitious target of developing 15,000 MW of solar capacity by March 2022. It has another target of getting 28.5% of its total energy from RE sources by 2032. But as on 30 June, 2018, total RE portfolio stood at only 930 MW (870 MW solar, 50 MW wind and about 10 MW small hydro), a mere 2% of its total capacity.

NTPC’s participation in the Uttar Pradesh auction is a bid to stay relevant in the rapidly transforming power sector. This formula, however, requires a different business paradigm. And despite a major advantage over private developers – it has direct recourse to state government in the event of a DISCOM default – we are not sure if NTPC could (or should) compete with private developers.

SECI has again extended bid submission date for the 10 GW integrated project development and module manufacturing tender. Revised date is 12 November, 2018. This is the fifth such extension after another round of poor response from the industry. We understand that many developers and 5-7 Chinese module manufacturers are keen to participate in the tender. But the conditions are punitive and they have been arguing for relaxation on several grounds.

The theory of combining two disparate parts of the value chain – manufacturing and project development – does not stack up;

Few companies want to be present in both manufacturing and project development;

The government is betting big on this tender to the detriment of other schemes and sector initiatives;

The two main changes being sought by the industry are increase in ceiling tariff (from INR 2.75/ kWh) and relaxation in cross-liability structure. There are some other areas of concern including high performance guarantees (combined INR 4.7 billion (USD 64 million) for manufacturing and project development) and stringent implementation period (two years) and minimum capacity utilisation (50-60% in the first two years) for the manufacturing component.

We continue to maintain that the integrated tender design is poorly conceived. In theory, developers can bid extra tariff for generation to cross-subsidise their investment in the challenging manufacturing business. But the low ceiling tariff does not leave much allowance for this. And most potential bidders – Softbank and Adani are the only two parties possibly interested in both components in our view – do not want an integrated play. The peculiar tender design is forcing them to consider complicated joint-venture arrangements, which could unravel down the line.

On its part, the government seems to have (wrongly) committed itself fully to the tender. SECI and MNRE are seriously looking into the industry concerns and it remains to be seen what concessions would be made. In fact, progress on all other tenders has been suspended while this tender is in the pipeline. But we see very low odds of a successful outcome. The developers are forced to be defensive in response to rising interest rates, falling Rupee and transmission capacity constraints. The DISCOMs want cheap power. And module manufacturing in India remains a tough proposition as highlighted again by the news of JSW reversing its decision to enter the business.

Central Electricity Regulatory Commission (CERC) has issued its final decision in response to appeals filed by Adani and Azure against NTPC and SECI respectively for ‘change in law’ compensation claims arising from GST implementation. It has opined that implementation of GST constitutes ‘change in law’ as defined in the respective PPAs. It has also ruled that project developers can claim compensation if they can provide precise details of all goods and services procured by them specifically for each project together with relevant invoices and an auditor’s certificate. Importantly, CERC has opined that the developers should be compensated for their extra financial outlay through an immediate one-time settlement by power procurers rather than through an upward adjustment in tariffs. But it has somewhat incoherently refused any compensation for increase in O&M costs as O&M outsourcing has been deemed a voluntary “commercial decision” made by the developers.

CERC’s decision is welcome but it still leaves a lot of ambiguity for developers;

The developers are appealing further to CERC to allow compensation through upward tariff adjustment;

A final decision on tax rates is still awaited from the Ministry of Finance in view of the inconsistent rulings by different state tax authorities;

CERC’s verdict comes after the Ministry of Power directed CERC to approve ‘pass through’ impact of all change in law events within a month of petition filing date. CERC decision is highly desirable. However, it still raises many doubts and a final solution may be many months away – state governments and DISCOMs may actually wait for a final decision from individual state regulators, who may take a different view. Moreover, as evident from the CERC ruling, the DISCOMs have taken the view that any decision in respect of PPAs is not binding on them if they are not party to it. So, a separate process may have to be undertaken for power sale agreements between DISCOMs and SECI or NTPC.

Expecting SECI and DISCOMs to make upfront payments is also unrealistic because of their precarious financial position. Although the amounts involved are relatively small – compensation is expected to be about INR 500,000 (USD 6,849) per MW – they are likely to argue inability to pay and/or delay payments. SECI would wait for corresponding payments from DISCOMs under its power sale agreements (PSAs) before paying out to the developers. The developers can anticipate these problems and we understand that they are appealing further to CERC to allow compensation through upward tariff adjustment, as also recommended by the Ministry of Power.

There is yet another major uncertainty associated with GST, which is applied at different rates – typically ranging between 5-28% – to various goods and services used in solar projects. A 5% concessional rate is applicable to ‘solar power generating systems,’ but many state tax authorities have taken a different and inconsistent view. Uttarakhand is the only state to endorse the 5% rate. Maharashtra, Rajasthan, Andhra Pradesh and Karnataka have taken the view that 18% rate is applicable to solar plants. The industry has been seeking urgent clarity from the Ministry of Finance in this regard.

GST has been a major pain point for the industry and it is unfortunate that we are still far from a solution despite more than fifteen months having elapsed since its implementation.

Total utility- scale solar and wind installed capacity reached 58 GW on September 30, 2018. Solar and wind capacity individually stood at 24 GW and 34 GW respectively. 

9874

Figure: Total solar and wind capacity on September 30, 2018

Last 12 months were notable for exponential surge in tender announcement. 46 GW of solar and wind tenders were announced in between October 2017- September 2018. 14 GW of utility scale solar and 8 GW wind projects were under various stages of execution on September 30, 2018. 

Figure: Top 10 players based on projects commissioned between October 2017 and September 2018

ReNew, Adani and Acme are the top three developers with installed capacity of 3.7 GW, 2.0 GW and 1.8 GW respectively, whereas Acme, ReNew and Azure take the lead in pipeline capacity (2.4 GW, 2.1 GW and 1.9 GW respectively). Canadian Solar (9.4% market share), JA Solar (6.6%) and Trina Solar (6.4%) were the top three module suppliers in the year. Domestic manufacturers’ combined market share increased marginally to 14.8% (12.2% last year). Chinese suppliers are beginning to dominate the inverter market also. China-based Huawei (16.3% market share) and Sungrow (14.9%) have emerged as the largest players replacing ABB and Japan’s TMEIC. Turbine market continues to be dominated by Gamesa (31.4%) and Suzlon (26.3%) constituting more than 50% of the total market share.

Although, Indian RE market has grown spectacularly over last four years, previous 12 months have seen several ups and downs. The tender activity for both solar and wind has gone up, but installation is slipping because of various operational and policy implementation issues. The sector growth is facing challenges from safeguard duty imposition, ceiling on bid tariffs, tender cancellation, falling rupee and higher interest rates.

BRIDGE TO INDIA’s latest edition of India RE Map – September 2018 contains other information on state-wise, policy-wise, player-wise progress of the Indian solar and wind market (download it here for free).

Spot prices for power on the energy exchange in India touched an all-time high of INR 17.61/ kWh (USD cents 24) last week. Average daily price on October 1, 2018 shot up to INR 7.64 (USD cents 11), about 150% higher than average for last year. Trading volumes have also simultaneously shot up by about 18% over last year.

The price spike owes largely to short-term supply and demand side issues;

States are reluctant to buy solar power despite it being amongst the cheapest sources of greenfield power;

Sustained solar demand beyond 10 GW per annum is dependent on finding a cheap solution to the intermittency problem;

Spot prices are ultimately a barometer of power demand-supply balance. Post monsoons, around September-October, wind and hydro power supply typically starts waning whereas demand sees a cyclical upturn. Prices therefore usually spike up in these months due to short-term aberrations rather than any fundamental long-term trends. This is an annual saga and thermal power should ideally be able to pick up the slack easily because of excess capacity. But the coal supply chain is highly constrained and rather unhelpfully, supplies run low at around this time of the year. Surge this year has been larger than expected as there is some discernible hardening in power demand in the recent months – 2018 YTD demand has increased y-o-y by 6%. Fall in wind power output also appears to be larger in comparison to last year.

As the factors driving increase in traded prices are primarily short-term, they do not, unfortunately, herald any likely increase in demand for solar power. In fact, the states are reluctant to buy solar power despite prices falling to all-time lows. We have maintained for some time now that one of the biggest challenges for solar power in India is relatively weak growth in power demand (and competition from abundant thermal power). DISCOMs have enough contracted supply and thermal power plants are still operating at low PLFs of about 62% only. With up to 60 GW of financially stressed thermal plants likely to see a debt resolution in the coming months, thermal power output may actually see a boost keeping solar demand subdued for another 2-3 years.

Variable RE penetration in some of the largest power consuming states (Tamil Nadu, Gujarat, Karnataka and Andhra Pradesh) is beginning to touch or even cross the critical 20% mark. DISCOMs need more firm power to meet incremental demand particularly for their evening peak loads. The key to creating sustainable and robust demand for solar is to solve the intermittency challenge.