Saturday, February 11, 2017

Modi government's solar policy - 2 :

Is the government’s overly aggressive solar thrust in public interest?

(This piece appeared in the Feb 11th issue of EPW ; reproduced below)

Shortly after coming to power the Modi government declared a fivefold increase in the 2022 target for solar generation capacity in the country to an eye popping 100 GW. Less than a year earlier, India’s electricity establishment had estimated 100GW to be India’s solar potential till 2032 (MoP 2014:22)! To see the numbers in perspective, India’s current solar capacity is less than 8 GW.

The target has been set without reference to the coal-fired capacity addition in progress and at a time when capacity utilization of existing thermal plants is very low and there is a large uncertainty on how electricity demand will develop in the next few years (Singh 2016, Tongia 2016:6).

The only argument the government has offered in favour of its aggressive solar thrust is that this would help India meet its international commitments on carbon emissions (GOI, 2015a). There have been questions raised about whether such a rapid build-up of non-fossil fuel capacity is indeed necessary to meet these commitments (Tongia 2016:17). These have remained unanswered.

The government estimates the investment requirement for 100 GW of solar generation to be of the order of Rs 6 lakh crores. Globally, RE is a favourite of investors and the government’s solar program has been enthusiastically received. Foreign investors such as SunEdison, SkyPower, Fortum India and SoftBank and Indian business houses including Adani, Tata and Mahindra have aggressively participated in the large solar tenders. Competition is fierce and the Ministry of New and Renewable Energy has had to hire large halls to accommodate all interested players during pre-bid meetings (Kenning 2015)!

How will such an aggressive solar program impact India’s electricity distribution companies? How will it affect the cost, availability and quality of electricity for consumers? Is the pace of solar adoption pushed by the government in public interest? These are some questions that this paper attempts to answer.

1.    Challenges of renewable energy on the grid

The thrust of the government is entirely on grid connected solar energy. A little background is useful to understand the challenge this poses for electricity distribution.

Electricity demand typically varies round the clock. For example, the all India average pattern shows a higher demand during the day than at night with a sharp late evening peak (PGCIL 2012: 57). It is a basic requirement of a stable electricity grid that demand and supply be “balanced”, or in other words, matched at all times and over different time scales.

Balancing demand and supply

There are several options for balancing. On the supply side, the output of power plants can be controlled to follow demand. On the demand side, the options can be to store energy when there is excess supply and to curtail demand forcibly or through economic disincentives when there is a deficit.

Conventional power plants – such as coal, gas-fired and reservoir based hydro power - are amenable to output control to varying extents. Their use in balancing is determined by their operational “flexibility” - the range over which their output can be changed and the rate at which the change can be made. The capacity available for flexible use is termed “balancing capacity”.

The output of gas-fired and hydro power plants with reservoirs can be changed rapidly and over a large range to handle changing load. These plants are high in flexibility. The old (“subcritical”) coal-fired plants were designed to provide a steady output. Output changes in these plants happen relatively slowly and over a smaller range and frequent output changes can lead to wear and tear with attendant costs. These plants are low on flexibility. Newer “supercritical” coal-fired plants are by design more flexible and resilient than the older subcritical plants (PGCIL 2012: 120-124).

Currently, demand is typically assessed from load profiles from the past (previous day, same day previous week or year) which can give an indication of the load variations to be expected. Conventional generators are scheduled to match the expected load.

The intra-day variation in demand is addressed mainly by varying output of reservoir based hydro plants. Coal plants provide the “base load” and their output is varied only in a small range (PGCIL 2012:125). In recent years, this range has been expanding steadily indicating need for increasing balancing capacity (MoP 2016b: 28). The use of gas-fired plants in balancing has been discouraged by non-availability of gas and high price.

When there is insufficient supply, “load shedding” is resorted to. The Indian grid has hardly any storage capacity available as the need for storage solutions has not been acutely felt in the past.

Implications of renewable energy for balancing

The presence of solar energy generators on the grid makes balancing more challenging for several reasons. One is that electricity regulation in India incentivizes solar energy by conferring a “must run” status on solar generators; their entire output must be accepted into the grid. This makes solar power plants “inflexible” from a balancing standpoint.

A second is that solar power is variable. Solar power plants produce power only in daylight hours and their output varies with the movement of the sun, peaking at midday. Balancing now needs to be carried out for load as well as supply variability.

A third reason is that solar output is dependent on weather. Cloudy or foggy conditions lower output and introduce intermittency into the variations. The expected output under such conditions, obtained from models using weather forecasting data, has to be available sufficiently in advance to enable scheduling of conventional generators for balancing. Since weather is not entirely predictable, actual generation will show deviations from forecasts and these have to be handled in real time.

Wind mills are the other major source of renewable energy (RE) in the Indian context. Together with solar, they account for over 90% (160 GW) of the RE target for 2022. These plants also have a “must run” status and produce output that is variable and influenced by weather conditions. From a balancing perspective, they have issues similar to solar.

Balancing areas in India’s federal electricity setup

There is another dimension to balancing that derives from India’s federal electricity setup - electricity provisioning is a state government responsibility. Each state has to maintain the supply-demand balance in its own grid which becomes the “balancing area”. Access to balancing capacity commensurate with the RE capacity planned is required in each balancing area, that is, at the level of every state.

The RE potential of a state depends on various factors like the level of solar irradiation and wind conditions. Seven states – Tamil Nadu, Karnataka, Andhra Pradesh, Maharashtra, Gujarat, Madhya Pradesh and Rajasthan – are suitable for both wind and solar generation and account for 70% of the aggregate wind and solar capacity planned across India (MNRE 2016). These have been termed “RE rich” states.

As to generation, historically, states have had their own dedicated power plants or shares in the capacity of central public sector power plants. State distribution utilities procure a bulk of their power requirements (89% in 2011-12) through long term power purchase agreements (PPA’s) with these state owned plants and some private plants (NTPC 2012). The remaining comes from generators with ‘untied’ capacity that are either recently commissioned private plants that have not found long term customers or private plants operating as merchant producers.

Long term PPA’s pretty much fix the generation resources and balancing capacity in the portfolio of a state. They also come in the way of states pooling their balancing resources. A state looking for additional balancing capacity outside of its fixed portfolio has to find it from the limited pool of ‘untied’ generators.
For these reasons, there can be a wide mismatch between the balancing capacity in different states and the RE capacity planned for them.  

The experience of Tamil Nadu:

Tamil Nadu currently has the highest RE capacity penetration among all states with RE (largely from wind mills) accounting for 56% of its overall generation capacity. Its balancing capacity is inadequate for this level of penetration (GIZ 2015: 54, 63-65). Use of its limited reservoir-based hydro capacity for balancing is restricted by irrigation release schedules and periods of high inflows into reservoirs when hydro power generation cannot be curtailed. Neighbouring Karnataka and Telangana, which are part of the Southern Electricity Region, are rich in hydro power resources, but these are not available to Tamil Nadu. The state has no flexible gas-fired plants and limited flexibility available in its old coal-fired plants (CEA 2013:13).

Till early 2016, in the absence of capability for wind power forecasting, short term power purchases were planned after making assumptions about wind generation. If wind power generation was greater than expected, after exhausting its limited balancing options, the state utility would have only two options - either back down power from private coal plants contracted for short term power or cut off wind power plants from the grid.

Either option has been problematic for the utility - violating contract provisions in one case and not respecting the “must-run” status accorded to wind generators in the other. The dispute involving the state utility, coal-fired plants and the wind power producers is now in the courts (Vaitheeswaran 2015). Legal issues aside, there are negative economic consequences either way. Varying power from coal plants means underutilization of capacity and higher costs related to wear and tear. Backing down wind power means wasted energy.

2.    Preparations for RE

The central government’s massive RE targets require a commensurate increase in balancing capability at least in the RE rich states. Balancing resources can be augmented by dedicated transmission corridors distributing RE across states, grid storage and additional flexible generation – all long gestation infrastructure (PGCIL 2012:116). Besides resources, accurate forecasting of RE generation is essential for balancing. What follows is an assessment of the central government’s preparatory work in each of these areas.

Grid Storage

Pumped storage is not only the most widely deployed grid level energy storage technology, it also the most flexible and competitive one (GIZ 2015: 80). Pumped storage hydro electric plants store and generate electricity by moving water between reservoirs at two different heights. While India has a very limited capacity of operational pumped storage, the electricity establishment has identified a number of hydropower projects that can be developed to support pumped storage (CEA 2013: 39-43). The government however has just woken up to the need to identify concrete projects and there is talk of setting up 10GW of pumped storage (ET Bureau 2016).

Grid level battery storage technologies are evolving and in one estimate 3-8 times more expensive than pumped storage (GIZ 2015:80). There are several vested interests active in promoting these technologies including the US – India business council and the government seems to have fallen for the hype created around them. The public sector Solar Energy Corporation of India has put out tenders for solar capacity with storage components potentially driving up the cost of solar electricity (Clover 2016). The storage component is miniscule as of now and nowhere near the scale needed to be practically useful to the distribution companies (DISCOMS).

It seems that storage can be safely discounted as an option for balancing in the run up to 2022.

Forecasting and Dispersing RE

Renewable energy management centres (REMC’s) are to be set up in at least all the RE rich states with the responsibility for state wide forecasting of RE. The costs incurred in managing the uncertainty in predicting renewable generation will not be part of its purchase cost; these costs are to be “socialized” among grid users (CERC 2015). Till mid 2015, there was no centralized forecasting for renewable generation anywhere in India (GIZ 2015:60). Tamil Nadu has inaugurated its REMC recently (Srikanth 2016).

Transmission corridors (termed “Green Energy Corridors”) providing RE clusters in RE rich states access to neighbouring states were a part of the 12th plan. The corridors are under implementation with an enlarged scope to include connectivity to the "ultra mega solar parks" and will enable RE generators to disperse electricity in a wider geography with more balancing resources than available in the RE rich states (MoP 2016b:42).

Both the forecasting and transmission infrastructure are early work in progress and there is no visibility into when they will be ready.

Flexible generation

There is little chance of capacity addition in gas-fired thermal plants in the 2022 time frame with existing gas-fired plants running at partial capacity because of the cost of gas which has to be imported. Hydro power projects totalling over 12 GW are under construction (CEA 2015). Possibly less than half of this capacity will be amenable to flexible use. Most projects are many years behind schedule because of environmental related standoffs and opposition from local populations.

Old coal-fired plants can be made more flexible through retro-fitting. This will require capital expenditure and there are no signs that governments (who own most of these plants) are seriously considering this option. A total of 73 GW of coal capacity is under construction of which supercritical plants account for 50 GW (CEA 2016, MoEFCC 2015:72).

One can conclude that coal-fired plants, in particular super-critical ones, will be the mainstay of RE balancing. With conventional capacity addition far lower than planned RE capacity addition (of 130 GW), India’s overall “balancing potential” – the ratio of balancing capacity to RE capacity – is set to decrease in the run up to 2022.

Market for balancing capacity

The mere existence of flexibility in generation will not translate to flexible operations as the later has negative financial implications for the operator. For instance, in the case of coal-fired plants, these are due to wear and tear reducing the life of the plant, higher maintenance costs and costs associated with capacity underutilization and lower efficiency. The government is therefore moving to incentivize flexible operations. There is already a regulation to compensate generators for holding capacity in reserve for responding to grid management requests in real time. A framework for market based pricing for balancing capacity is just down the line.

Will market based incentives solve the problem of making adequate balancing capacity available in the RE rich states?

There are some constraints. Firstly, the generation capacity available in the electricity market untied to PPA’s is currently limited, though it is slated to rise with the commissioning of new plants. Secondly, inter-regional transmission constraints can come in the way of RE rich states using flexible capacity from regions other than their own.

The later problem is illustrated by the Southern Electricity Region which has been facing a generation capacity deficit for several years. Coal-fired generators in the Western Electricity region are unable to provide power to the Southern Region because of transmission bottlenecks and their capacity lies underutilized. Market based pricing for electricity has not solved the problem of electricity deficit in the southern region in five years; electricity prices at the Indian Electricity Exchange have remained significantly higher for the southern region compared to the western region from 2011 onwards (Kasturi 2016:24).

Two years after announcing massive RE targets, the government still does not have an assessment of the actual balancing capacity available with the RE rich states or how this will grow in future! It appears to believe that the market for balancing capacity will somehow solve all problems.

3.    The real cost of solar

State utilities are generally strained financially and will not be keen to purchase RE as long as it is relatively expensive. To make RE more attractive, the central government has worked out ways of subsidizing it at the cost of public sector companies in the power or fuel sector. Inter-state transmission charges for solar electricity have been waived at the cost of the PGCIL.

NTPC contracts for solar power from producers and sells it to DISCOMS after subsidizing it in the following way. It “bundles” solar power with low cost power from its coal-fired plants and offers utilities power at a rate which is lower than its purchase price for solar electricity (Upadhyay 2015). This bundled price has to approach “grid parity” – the average price of electricity contracted by utilities - for NTPC to be able to find willing buyers.

Of course, even if solar prices reach grid parity it does not mean that solar has become cost effective compared to other sources of energy. The cost of balancing variability in generation through flexible capacity held in reserve must also be attributed to solar power. To this must also be added the cost of infrastructure for forecasting RE and the costs arising from errors in forecasting. The government has not even hazarded a guess at these costs yet.

As subsidies alone are not enough to make solar power attractive, the government has also taken recourse to coercion. The new tariff policy calls for high RE purchase obligations for DISCOMS with the target for solar alone being 8% of non-hydro power consumed by every utility by 2022 (MoP 2016a). To make sure that states comply with the RPO targets, such compliance has been made part of the conditions associated with the ‘Ujwal Discom Assurance Yojana’ (UDAY) that provides relief to indebted state DISCOMS (GOI, 2015b).

Negative consequences of force feeding RE

Forcing DISCOMS to absorb RE beyond their ability to handle it will have consequences for the health of the DISCOMS and the cost and quality of electricity supply. A key assumption behind UDAY is that power costs will come down with lower cost of coal and help DISCOM finances. Rapid solar penetration will push up the cost of power.

Utilities are already hard put to handle load variation even today. They lack accurate load forecasting, flexibility in conventional generation, balancing resources such as pumped storage and generation reserves to handle different eventualities on the grid (MoP 2016b: 11). For customers, this has meant a regime of poor quality and unscheduled power cuts. With high RE penetration and an expected further deterioration in balancing potential, this regime is bound to continue in to the future. The government is also preparing to use demand curtailment curtail for balancing by pushing for large scale installation of smart meters that will allow setting time-of-day tariff (MoP, 2016a).

Public interest will be better served if the pace of solar (and wind) capacity build up is compatible with the balancing capacity available with the states and their ability to manage RE variability. The government must pay as much attention to capacity building in inter-regional transmission, pumped storage and highly flexible generation as it is doing to solar generation.

Renewable energy targets based on these considerations rather than impetuous declarations will be sustainable and allow steady decrease of carbon emissions. A slower adoption of solar generation will be beneficial for yet another reason - solar power, as long term trends suggest, will only get cheaper with time.


CEA (2013): “Large scale grid integration of renewable energy sources - Way forward”, Central Electricity Authority, November,

-  (2015): “Hydro electric projects under execution”, Central Electricity Authority, November,

-  (2016): “Monthly report on broad status of thermal projects in the country”, Central Electricity Authority, July,

CERC (2015): “Framework on Forecasting, Scheduling and Imbalance Handling for Variable Renewable Energy Sources (Wind and Solar): Statement of Reasons”, Central Electricity Regulatory Commission,

Clover, Ian (2016): “India: storage to be included in 100 MW tranche of Andhra Pradesh 750 MW solar tender”, PV Magazine, 15 March,

ET Bureau (2016): “India readies plan to improve renewable power storage”, Economic Times, 22 August,

GIZ (2015): “Report on Forecasting, Concept of Renewable Energy Management Centres and Grid Balancing”, Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) 

GOI (2015a): “Revision of cumulative targets under National Solar Mission from 20,000 MW by 2021-22 to 100000 MW”, Government of India, 17 June,

GOI (2015b): “UDAY (Ujwal DISCOM Assurance Yojana) for financial turnaround of Power Distribution Companies”, Government of India, 5 November,

Kasturi, Kannan (2016): “Private Thermal Power in a Liberal Policy Regime”, Economic & Political weekly, Vol 51, No 10, pp 22-26

Kenning Tom (2015): “India’s cutthroat solar auctions – behind the hype”, PVTECH, 22 Dec,

MNRE (2016): “Tentative State wise break-up of Renewable Power target to be achieved by the year 2022 so that cumulative achievement is 175000 MW”, Ministry of New and Renewable Energy, (Sept 21, 2016)

MoEFCC (2015): “First Biennial Update Report to the United Nations Framework Convention on Climate Change”, Ministry of Environment, Forest and Climate Change, December,

MoP (2014): “Perspective Transmission Plan for twenty years (2014-2034)”, Ministry of Power, August,

MoP (2016a): “Resolution, Tariff Policy”, Ministry of Power, Gazette of India, 28 January

MoP (2016b): “Report of the Technical Committee on Large Scale Integration of Renewable energy, Need for Balancing, Deviation Settlement Mechanism and associated issues”, Ministry of Power, April,

NTPC (2012): “Annual Report, 2011-12”, National Thermal Power Corporation,

PGCIL (2012): “Transmission Plan for Envisaged Renewable Capacity, Vol 1”, Power Grid Corporation of India Limited, July,

Singh, Sarita (2016): “Power demand may be lower by 15% for five years starting FY18”, The Economic Times, 25 Apr,

Srikanth, R (2016): “Tangedco sets up centre to tap renewable energy”, The Hindu, 26 March,

Tongia, Rahul (2016): “India’s Updated (2016) Renewable Energy ‘Guidelines’: Bold targets, but can we meet them?”, Brookings India IMPACT Series, No. 082016-2.0

Upadhyay, Anindya (2015): “India's Modi Tells Coal Power Plants to Subsidize Solar”, Bloomberg, 7 September,

Vaitheesvaran, Bharani (2015): “Wind or conventional power? Tamil Nadu power producers battle it out in court”, The Economic Times, 26 August,

Tuesday, January 17, 2017

Financial Inclusion: Do the poor benefit from banking ?

The conditions attached to exchanging old notes that came with the notebandi from November 8th 2016, the subsequent scarcity of cash, and the current government campaign for cashless transactions have put intense pressure on the unbanked to open bank accounts and transact with banks.

Bringing the working poor into the fold of banking - presented as a welfare initiative with its terminology of “financial inclusion” - has been an ongoing project in India. In the current phase, it has become extremely coercive.

The standard narrative (see for example a recent Live Mint piece titled “How moving money to the bank helps the poor”) is that certain obstacles prevent the working poor from accessing banks, but once these are removed, they will be able to save more, derive additional income and avail of credit on favorable terms.

The problem of access

The biggest obstacle – the problem of access to a bank for rural customers - has been acknowledged by the government for many years. The Regional Rural Banks (RRB’s) – focused on rural geographies - were first thought to be the answer. With their limitations becoming apparent, the idea of banking correspondents (BC’s) was put forward in 2006. Including BC’s, there are still too few banking outlets in rural India, particularly in less developed districts.

Take Karnataka for example. While in urban Bengaluru, there are 24 bank branches and 77 ATM’s per lakh population, when it comes to districts such as Yadgir, Bidar and Koppal, this falls to 9-11 branches, 6-8 ATM’s and 5-8 banking correspondents per lakh population (Data culled from the RBI, a report in the The Hindu and the State Level Bankers Committee, Karnataka). Given that population density is much lower in these districts compared to Bengaluru, people have to travel much longer distances than their urban counterparts to avail of banking and the bank offices that service them deal with far too many customers.

Access to banking is not only about physical access.

The working poor in both rural and urban areas are often illiterate or versed in only the local language. They have to contend with the complexities of banking in the presence of unsympathetic bank officials. The recent incident of demonetized notes being deposited and then withdrawn the same day from the accounts of 250 MNREGA workers in Rajo Majra village of Punjab, all without their knowledge shows how the poor can be exploited. The workers apparently expected government transfers into their accounts and signed bank forms that they could not read.

A large number of the working poor are migrants who move from one work site to another. Accessing their accounts – which would be with a bank branch near home – presents many problems for these workers; opening multiple bank accounts also makes no sense.

Leaving aside the obstacles in accessing formal banking services, will the poor reap benefits once forced (as in the current situation) into the formal financial system?

Do bank deposits make sense?

Take the question of savings. A recent large sample national household survey reported in Live Mint shows that the difference between income and routine consumption expenditure of households in the bottom two quintiles of income is under Rs 2000. This is the monthly saving potential of 40% of households in India. In case of unforeseen expenses such as a medical emergency, this money will have to be accessed quickly. How would it make economic sense for these households to deposit these savings in a bank? The paltry interest that can be earned will not justify the costs of banking – the time and expenses on travel.

The lack of savings potential in the poor is ironically demonstrated by the “Jan Dhan Yojana” (JDY), a scheme promoted by the government to bring the unbanked into the fold of banking. Under this program 266 million accounts had been opened till the first week of Jan 2017 with the help of some inducements such as free insurance, free debit card and some overdraft facility.

About a quarter of these accounts, 65 million, still have zero balance. The average deposit in the accounts with money was less than Rs 2300 before notebandi. By the first week of January (after the expiry of the deadline for depositing old notes) this had gone up to Rs 3400. Most JDY account holders who have been brought recently into the banking system clearly did not have much in the nature of financial assets in cash.

Bank accounts then do not make sense as a savings instrument for the poor not only because of the difficulties of accessing a bank but also because the poor hardly have any savings.

The false promise of credit

What about access to credit? Banks with their commercial objectives can obviously provide credit only to those with predictable earnings or having assets such as land or housing that can serve as collateral. The reserve bank statistics on the credit extended by banks is revealing.

There were a total of 137 million individual credit accounts in the Indian banking system in 2015. Individuals could have more than one credit account – for example one for a vehicle loan and another for a housing loan. So the number of people with access to credit from the banking system is lower. For comparison, the number of eligible voters in 2014 was about 814 million.

These 137 million borrowers can be divided into petty borrowers whose credit limit is 2 lakhs and large borrowers with a higher credit limit. There were 29 million large borrowers and their average borrowing was nearly Rs 6 lakhs each. These borrowers would typically belong to the top quintile of household income earners and avail credit for housing, vehicles, education, etc.

On the other hand, the average borrowing of 108 million petty borrowers was only Rs 50,000. These would be typically farmers and self other employed persons. These numbers show up the limits of access to credit from the banking system for poor households.

Anecdotal evidence suggests that even the farmers who have access to credit borrow from informal channels as the institutional credit available does not meet their requirements. For the self employed without assets and for informal wage workers, credit from informal sources at high interest is the only recourse. For the typical vegetable cart vendor who requires credit on a daily basis, the wholesaler in the mandi serves the function as no bank would be able to fulfill his need.

The real motivation for “financial inclusion”

The poor when provided access to banking save little and can avail of only limited credit small; they cannot be important for the profits of the banks. In the words of the former RBI governor Raghuram Rajan, banking is not easily available to the poor “not just because the financial system is underdeveloped, but because they are hard to service profitably”.

The economic non-viability of making the poor transact with banks - the BC model too may have run its course as recent studies reported in The Wire indicate -  has pushed the government to promote ever newer models of service providers to carry forward the financial inclusion project, the latest being the Payments Bank. The latter can open accounts, hold small deposits and process payments but is barred from providing credit. The thinking behind the Payments Bank concept is that with relaxed regulatory requirements, it will be able to make profits from transactions of poor account holders where regular banks are unable to.

What accounts for the persistence of the government with financial inclusion when neither the banks nor the poor account holders benefit?

The answer could lie in the fact that a bank account for every worker is an essential requirement for moving India’s massive informal economy - that supports 90% of livelihoods - in the direction of formalism. The informal economy effectively shuts out corporate India from a large part of the domestic market and the government from potential revenue. Government policy has long favored the growth of the formal/corporate economy.

“Financial inclusion” then is not so much about benefiting the poor as it is about creating the infrastructure for formalizing the informal economy. In this project, if the poor have to bear some costs, that is just collateral damage.