SPV Applications for Villages

At the moment, rural energy and distributed generation is the best space for solar energy in India. A large part of the country receives good insolation for over 300 days. One of the chief advantages of solar energy is its ability to be customised for individual household needs – generation can be tailored to closely match the demand and financial constraints of the household. Installation does not require much time and maintenance, requirements are minimal. As aspirations and affordability rise, modules can be added, allowing greater loads and consumption. The range of applications, and the institutional mechanisms that have evolved to disseminate them, have made solar-based applications flexible as well as affordable.

As a stop-gap measure, solar lanterns are perfect substitutes for kerosene lanterns. If we consider the negative implications on health, the poor quality of light and the cost of subsidised kerosene over the lifetime of a solar lantern, we can very easily see how it fares better on a head-to-head basis. For the 28 million households that will remain unelectrified at the end of RGGVY, the total kerosene subsidy for lighting requirements is estimated

MNRE’s Solar Policies

It is worthwhile to briefly review the context in which JNNSM guidelines have been framed. Rural electrification comes primarily under the ambit of the MOP. As per Bhattacharyya (2006), immediately after Independence, the focus of rural electrification was on promoting irrigation and not domestic uses. In the ensuing decades, the MOP followed the route of extensive electrification, in terms of the number of villages, rather than an intensive approach to ensure all households were electrified. The f ocus on household electrification came only after the passing of the Electricity Act in 2003. There were some exceptions that p receded this Act, like the Kutir Jyoti Programme of 1989, which targeted a single-point connection to below poverty line (BPL) households.

With the RGGVY, launched in 2005, the MOP sought to achieve 100% intensive electrification by the year 2012, primarily by extending the grid. The mandate for the RGGVY was to provide at least 1 kilowatt hour (kWh) a day of electricity to every household.7 Under the distributed generation component of the RGGVY, solar is the source of electrification of last resort – that is, when extension of grid is unviable (even by MOP’s rather generous standards) and when other distributed sources of power are not available.8

What needs to be highlighted here before we study JNNSM’s off-grid guidelines is that rural electrification does not come

Economic & Political Weekly

EPW

d irectly under MNRE’s mandate and hence need not be the prime area of focus for the JNNSM either. The MNRE’s role in rural electrification has been to fill the gap left by the MOP. The Remote Village Electrification Programme is the largest such exercise and basically targets 18,000 villages that were too far from the grid.9 The MNRE’s primary role is to increase the use of renewable energy in the country. What we are arguing, however, is where solar energy fits best. We believe, as discussed earlier, that rural electrification is the best-fit at this moment. Hence, the JNNSM strategy towards off-grid rural electrification merits close examination.

For solar energy, MNRE’s two most relevant programmes were the Solar Lantern Programme and the Solar Photovoltaic or SPV Programme launched in 2009.10 Both lapsed with the implementation of JNNSM’s off-grid guidelines.

Overall, the JNNSM has incorporated several welcome changes over earlier programmes. For example, while the SPV Programme was restricted to five models eligible for government support, u nder JNNSM, the number of models has been increased to 11. The basic objective of the former was to substitute the use of k erosene lanterns in these beneficiaries’ homes. Its mandate was to provide one solar lantern to homes in unelectrified villages and to BPL families with a school-going girl child. Moreover, procurement and dissemination were restricted to state nodal agencies and Akshaya Urja shops.11 The new guidelines have relaxed this restriction.

The Role of Regional Rural Banks

We will slightly digress here to quickly review the role of financing institutions in promoting PV applications. The solar services industry in India is perhaps unique in leveraging the formal financial infrastructure. Sri Lanka and Bangladesh have relied on microfinance (World Bank 2008). In Latin America (Rogers 2006) and in certain regions in Africa (Gustavsson and Ellegard 2004), the leasing approach or micro-leasing, as it is called, has been prominent. Under this approach, the service provider has an energy service contract with the customer over the period of a few years. The customer pays a monthly fee while the service provider provides maintenance and continues to own the system. All our respondents were sceptical of such a model working in India and stressed the importance of end-user ownership. The main issue is one of the end-user’s incentive in maintaining the system and hence the associated risks for the service provider.

A United Nations Environment Programme (UNEP) solar loan programme in 2003 was one of the first instances of the Indian mainstream banks financing solar lighting systems.12 The UNEP support essentially worked as an interest subsidy. Syndicate Bank and Canara Bank in Karnataka were partners in the first phase of this programme. Since then, many other banks have taken up this initiative. A specific case in point is that of Aryavart Gramin Bank which worked with Tata BP Solar in financing solar home systems for a few villages in Uttar Pradesh (UP).13 The success of essentially market-based dissemination models, with occasional interest subsidies, orchestrated by the rural banking sector has prompted policymakers to make this an integral part of the JNNSM.

The target for off-grid applications is 200 megawatt (MW) by 2013 and eventually, 2,000 MW by 2022.14 This target includes solar lighting systems for rural electrification as well as community buildings, hospitals, etc, PV agricultural pumps, power plants for village-level micro-grids and street lights. Applications for commercial establishments like standby systems for diesel abatement and off-grid power plants for telecom towers also fall under this category. The target for rural lighting systems alone is 20 million units by 2022.15

problems which threaten to negate the positive steps outlined above. We will also discuss other issues with the JNNSM approach, as a result of which it falls short of its intended objective of supporting a sustainable p rogramme to reach poorer households.

Issues with the Off-grid Policy

As per the off-grid guidelines for the JNNSM, Rs 300 per watt peak (W) is considered the benchmark costs for systems with

p

battery storage, irrespective of size. The MNRE

will provide 30% of the benchmark costs as

The off-grid policy envisages the creation of a range of different channel partners such as system integrators, banks and microfinance institutions, financial intermediaries, renewable energy service companies and state nodal agencies, who will create off-grid solar projects and submit them to the MNRE’s Project Appraisal Committee for approval and financing. These channel partners will have to get themselves rated or accredited through an agency appointed for the purpose; based upon their rating they can put in project proposals. Each project proposal can aggregate different off-grid initiatives totalling 1 MW at a time.

One of the most positive steps in the new regime is the envisaged involvement of rural regional banks and non-commercial banking corporations as channel partners who can initiate projects and be given incentives to be proactive. Tied to this is the strategy of providing interest subsidy as opposed to high capital subsidies alone. While there is still a capital subsidy component to the tune of 30% of benchmark costs of the systems, this is much lower than in previous programmes. In addition, under the JNNSM (and in contrast to the previous SPV Programme, where a choice had to be made between the two), the beneficiary can a ccess the interest subsidy as well. While there is always a concern about whether banks will be selective in choosing the customers or prefer to avoid the risk of financing low-income households, interest subsidies are likely to be less market distorting.

To ensure efficiency in implementation, the financing is to be back-ended. Banks get 50% of the subsidy payment upfront and the rest on completion of the project. Currently, payments are channelled by MNRE through the National Bank for Agriculture and Rural Development (NABARD). Damian Miller, chief executive officer of Orb Energy, suggests that banks are a little wary of this clause at the moment as there is uncertainty on the route of disbursement.16 Possibly these are just teething problems and may get sorted out in the future. Anil Patni, deputy general manager of Tata BP Solar, feels that the critical success factors will be the speed and ease with which the channel partners and consumers can receive the subsidy from the government.17

All said and done, it is possible that banks have other issues with the government policy. Harish Hande, for instance, claims that SELCO has never had as much difficulty in getting loans as it has had since JNNSM.18 In the next section, we will study the

costs. By having the current regime in place, the intention is to be open to solar lighting systems in multiple sizes and to provide clear guidelines for banks.

However, on a per watt basis, the smaller systems are priced signi ficantly higher than the benchmark costs of Rs 300 per watt. The price does go down with an increase in size of the PV system. In the case of smaller systems, other systems will account for a higher proportion, such as the costs of internal wiring of homes, compact fluorescent lamp (CFL) bulbs, labour for installation and servicing in remote areas.

Further, in direct current (DC) systems that are better suited in the rural context, the product includes loads for lights and fans. These small systems, of less than 40W, form the bulk of the sys

p

tems purchased by rural households, and constitute over 50% of the unit sales for SELCO and Orb Energy. In fact for Orb Energy, they constitute almost 80 to 90% of the units sold.

In this context, the government’s prescribed benchmark costs seemed to be unfavourable to the very section of the population that could have greatly benefited from this programme. While all our respondents agreed that Rs 300/W is reasonable for alter

p

nating current (AC) systems at the kW level,20 the costs for rural systems on a per watt basis are higher. As per our respondents, the costs of smaller systems, of under 40 watt peak, is over Rs 600/W, while for 40 W to 100 W, this goes down to about

ppp

Rs 500/W. We can surmise that this perhaps asymptotically de

p

creases to the assumed benchmark costs of Rs 300/W for larger

p

systems, on the kW scale, for either rooftop models for larger buildings or for village-level power plants.

With the subsidised loan capped at Rs 150/W, to procure a

p

low wattage system, the user either has to pay the rest from his pocket or take an additional loan at market rate. We have plotted below the estimated costs for different size systems, based on feedback from SELCO and Orb Energy. We have also presented what these costs imply in terms of down-payment and monthly instalments for rural households. Here it is assumed that the c ustomer has to avail of two loans – one for Rs 150/W

p

at the subsidised interest rate and the other for the difference between the actual cost and the subsidised loan. We have assumed that the loan term is five years and while the subsidised rate of interest is 5%, the market rate is 12% to the customer. The down-payment is assumed to be equivalent to 20% of the actual costs.

Commoditisation of Systems

One of the announced objectives of the off

grid guidelines of the JNNSM is the commod

itisation of solar PV applications, i e, facili-

Effective Subsidy (in %)

20

versus service debate? The notion that a set

15

of models must be identified originates from

the belief that solar applications are prod

10

tating the availability of a range of products

ucts. The effect is on the quality of service

5

extended to the customer. Sahu estimates

p

Economic & Political Weekly

EPW

june 4, 2011 vol xlvi no 23

only requirement currently explicitly stated is a warranty of five years against manufacturing or installation defects for all components and at least five years for overall structure. Maintenance of installed systems has been repeatedly stressed as a critical requirement for the success of SPV dissemination (Nieuwenhaut et al 2001; van der Vleuten et al 2003). In fact, Wong (forthcoming) suggests that one of the main obstacles for dissemination programmes is the absence of “robust monitoring and efficient technical support”.27

One of the most often cited advantages of the SPV systems is its low overhead and maintenance costs. While this is true, simple issues like dust accumulation or the solar modules moving into the shade could adversely affect performance. Further, the battery is perhaps the most expensive component of the system; during its lifetime, it has to be replaced on average once every three to five years. There is reason to be worried that by overlooking requirements for maintenance and stipulating unnaturally low benchmark costs, this programme will con-tinue to support what has been described as “solar rubbish” (Schutzeichel 2010).

Alternative Institutional Models

As was pointed out earlier, despite loans, the costs of systems can still be prohibitively expensive. Different institutional models have been developed to help soften the financial burden of an SPV application. One alternative is a central charging station, where households can drop off lanterns (Chaurey and Kandpal 2009) or batteries to be charged. An entrepreneur trained by the service provider takes care of the central station and also maintains the batteries and systems. There have also been other experiments with the ownership of the system. The Energy and Resources Institute (TERI) model, for instance, involves a rental approach wherein households can rent the lantern on a daily basis for a small fee on the order of Rs two to five (Chaurey and Kandpal 2009).

The central charging station is a promising option. Despite prior experience with this option in the country (Chaurey and Kandpal 2009, 2010; Rao et al 2009), the guidelines do not explicitly mention this model. While the government may presumably be willing to consider such alternatives on a case-by-case basis, the absence of support for non-credit based institutional models could limit dissemination through smaller entrepreneurs.

One model that could potentially work would have a local intermediary non-governmental organisation (NGO) or a government body absorb the financing risks, instead of the service provider. Such an intermediary could bear the upfront costs and the loan advanced by the regional rural banks, in return for a monthly payment over a fixed period of time, say seven years or more. If at the end of this period, the user gets ownership of the system, it is essentially a credit model. Rao et al (2009) described a similar mechanism that was applied in an urban setting and with a central battery charging station. On the other hand, if the intermediary continues to own the system, it works like the leasing model. The advantages of such a scheme is that the burden of margin money can be avoided, making the system much more affordable for the

A Few Recommendations

Financing Requirements: There are a couple of specific requirements to avail of the loans that we believe are likely to become

o bstacles. One, there is a lock-in period of three years, that is, the user must avail the loan for a minimum period of three years (and a maximum of five years). While this stipulation is perhaps to ensure that banks do not demand repayment early, denying users the flexibility to pay when they can seems rather unhelpful. Second, down-payment requirements are known to be barriers for lowincome households who may not be able to afford Rs 1,500, for the most basic 10 watt system, in a single payment. Perhaps this requirement can be removed for systems less than 40 W.

p

Interest Subsidy and Removal of Benchmark Costs: The case for the removal of the benchmark costs should be clear from an earlier section. An alternative option could be to increase the benchmark costs, but it is not obvious what the representative costs should be. The rationale for the current flat costs per watt is to avoid the complexity of defining different costs and subsidies for systems of different sizes, and hence allowing a relatively larger number of sizes. Mandating product specifications and then suggesting costs would be a backward step. An option could be to treat systems smaller than, say, 40 W differently from

p

larger ones and hence, stipulate two sets of benchmark costs.

If the above issues with financing requirements are dealt with, an interest subsidy alone should be sufficient to make the systems more affordable. Again, this would not be able to cater to all s ections of the rural community. The poorest sections should be targeted separately.

Maintenance and Local Presence Requirements: Going back to the discussion on solar energy as a service, and not a product,

Notes remain unelectrified

the government must mandate an annual maintenance contract (AMC) of five years or so. While this is bound to increase the cost, the presence of a good servicing infrastructure is vital for good battery life and system performance. As a cost-cutting measure, service providers may continue to prefer developing local competence in servicing; this would be a very positive development for diffusion of the technology. Further, the hope is that with increased dissemination, auxiliary industries may develop around solar systems, thus building the requisite infrastructure. The government must, therefore, require an AMC and a local presence before extending support.

Specific Targeting?

As we mentioned previously, while the Solar Lantern Programme lapsed with the implementation of the off-grid guidelines, the current approach does not target BPL households separately anymore. Even the 10 W systems may be out of reach of these house

p

holds, despite the capital and interest subsidies. The government can even look to distribute solar lanterns with a capital subsidy of 90% or so to these families in order to replace kerosene lanterns. Such a subsidy does exist for special category states, viz, the north-east, Jammu and Kashmir and Uttarakhand. States like Bihar, Jharkhand, Orissa and Chhattisgarh have poor electrification levels and hence, may warrant special targeting. Another approach could be to have different rates of subsidies for systems of different sizes. Systems smaller than 20 W or maybe even 40 W

pp

can get higher rates of interest subsidy relative to larger sizes.

Conclusions

While solar power may often prove to be expensive relative to other alternatives for electrification, it stands out for its ability to cater to the diverse needs and circumstances of users. In designing support mechanisms, the government must keep in mind the flexibility SPV offers, and its associated technical and institutional merits. Despite positive measures in decentralising the disbursement of subsidy through the rural banking sector, the structure of this subsidy requires a review. The benchmark costs are considerably lower than the actual costs, leading to a distorted subsidy structure. In light of this, the recently announced decision to lower benchmark costs further to Rs 270/W seems ill ad

p

vised. As the capital subsidy is bound to be tied to a benchmark cost, the financial assistance must be restructured to involve an interest subsidy alone. Furthermore, attempts to enforce standardisation and an over-reliance on certain delivery routes, though well-intentioned, may do more harm than good to the goal of large-scale dissemination.