Rooftop Solar in India: Part 3: Policy Tools {Tariff Orders, Net / Gross Metering, Feed-In Tariff, Coal Cess, Accelerated Depreciation, MNRE Subsidy}

This is part of a set of articles: "Distributed / Rooftop Solar in India: A Gentle Introduction: Part 1","Rooftop Solar in India: Part 2 {Shadowing, Soiling, Diesel Offset}", "Rooftop Solar in India: Part 3: Policy Tools... Net Metering etc..." "Solar Economics 101: Introduction to LCOE and Grid Parity" , "Solar will get cheaper than coal power much faster than you think..", "Understanding Recent Solar Tariffs in India", "How Electric Scooters,... can spur adoption of Distributed Solar in India," "Solar + Ola! = Sola! ... The Coming Energy-Transportation Nexus in India", "UDAY: Quietly Disentangling India's Power Distribution Sector", "Understanding Solar Finance in India: Part 1", "Back to the Future: The Coming Internet of Energy Networks...", "Tesla Model 3: More than Yet-Another-Car: Ushering in the Energy-Transportation Nexus", "Understanding Solar Finance in India: Part 2 (Project Finance)", "Ola! e-Rickshaws: the dawn of electric mobility in India", "Understanding Solar Finance in India: Part 3 (Solar Business Models)" . 

This article will focus on solar policy tools in India, some references to how they relate to policies in Australia, USA, Germany, and how to conceptually understand them and their implications. {Disclaimer: I am not a policy expert. Please alert me in case you find any factual mistakes and I will fix them in the article.}

Before we get started, lets remember that "policy" does not necessarily mean "subsidy". It means that government is involved in setting the rules of the game, to either level the playing field and/or promote a push towards certain outcomes. In certain targeted cases, there can be subsidy or use of government procurement as a policy tool. We will clarify this point further in the article.

At the top level, the purpose of public policy is to encourage renewable (solar etc) power and push it to become a lower cost domestically produced energy source. India imports a significant amount of oil, coal and natural gas to drive our energy production in our transportation sector (cars, scooters, planes, trains), building sector (eg: diesel gen sets, thermal power). Fossil fuels impose a long run variable cost input, and are subject to market volatility, and an impact on India's long run current account deficit & exchange rates {i.e. impact of India's energy security}.

In contrast, India is blessed with an amazing solar resource that is well distributed across the country and that does not have a long run variable cost of input fuel. Further, solar PV, wind turbines and energy storage technologies are on a long run capital cost decline pathway (primarily due to learning curves of technology and manufacturing), and are being spurred further due to software/IOT/analytics advances. {Read more of this in: "Solar will get cheaper than coal power much faster than you think..","How Electric Scooters,... can spur adoption of Distributed Solar in India," } More specifically, if policy nudges the industry to invest more in these renewable energy technologies (especially solar and storage), the increased deployment of these technologies will drive down cost due to learning curves. If policy succeeds, it will get the proverbial boulder rolling down the hill (i.e. learning curves & more deployment) and the public investment in policy would pay dividends. 

#1 Retail Electricity Tariff & Clean Environment Cess (Coal Cess)

The first (indirect) policy tool is the retail electricity tariff, set by state electricity regulators. The particular tariff/kWh and the trend in prices, when compared to solar LCOE (levelized cost of energy) and its trend forms the basis of a "grid parity" statement (i.e. is my solar PV system approaching grid parity). Read more in my earlier article "Solar Economics 101: Introduction to LCOE and Grid Parity". For example, the Delhi government in its solar policy draft mentions: "While solar  energy tariffs have, on  average, fallen  6--‐8%  per year  since 1998  (solar  panel  prices have dropped 75%  in  the  last  six years),  conventional  energy  tariffs in  Delhi  have  risen  6.9%  per  year  on  average  since 2007".

This implies a crossover of solar vs grid energy costs, i.e. if the units (kWh) of solar energy production can be compensated at the retail electricity tariff rate, it would become economical to install solar (and achieve a return on the capital investment required. We will see that the methods such as FIT, net metering etc are methods to compensate solar production. Remember that to install solar, we have to finance the CAPEX today; and get returns over 25 years (without much opex). The term of financing, energy production yield, capex level, and interest rates matter: this has been discussed in my article "Understanding Recent Solar Tariffs in India".

What drives the levels of retail electricity tariffs & how will the coal cess affect it?

Firstly we observe that residential tariffs have some subsidy embedded in the tariff structure. In Karnataka, the first 30 units / month are priced at Rs. 2.70/unit; the next 70 units/month at Rs. 4/unit, the next 100 units/month at Rs. 5.40/unit, and beyond at Rs. 6.40/unit. If you are a small home (likely at lower income levels), your consumption is lower (perhaps no air conditioner etc), and you pay a lower average tariff. Think of this as similar to progressive taxation rates. Bijli Bachao web site tracks some of these retail residential tariff numbers and slabs.

Second, the commercial tariffs are about 20-25% higher than residential tariffs. In Karnataka, beyond the first 50 units (which is at Rs. 6.95/unit), the marginal LT commercial tariff rate is Rs. 7.95/unit which is 24% higher than the Rs. 6.4/unit residential top slab. Similarly in other states: Tamilnadu (Rs. 9.93 commercial, Rs. 8.23 industrial vs Rs. 6.2 residential, 60% higher for commercial), Andhra Pradesh (Rs. 9.33 commercial, Rs. 6.84 industrial vs Rs. 7.82 residential, 19% higher for commercial), Maharashtra (Rs. 10+ commercial vs Rs. 9.95 residential) etc. There is a larger-than-average "cross-subsidy surcharge" (CSS) embedded in commercial tariffs. This implies that within rooftop solar space, the "comparable" for solar looks better for commercial sector . We outlined this in: "Distributed / Rooftop Solar in India: A Gentle Introduction: Part 1".

Third, state regulators evaluate the average revenue requirement (ARR) of electricity discoms and utilities to set tariff orders, currently done annually. In brief, if you look at the 2015-16 tariff order by KERC, the agriculture sector (irrigation pumps IP use) is significantly cross-subsidized (24% of BESCOM's units consumed, 7% of revenue) primarily by commercial (19% units consumed, 31% revenue) and industrial (27% units consumed, 31% revenue). Residential sector is mildly subsidized (24% units consumed, 21% revenue), by "other" (eg: advertisement hoardings etc, 7% consumed, 10% revenue). In BESCOM's filing proposing an ARR of Rs. 12,389 crores, the power purchase costs (Rs. 9782 crores, i.e. 79%), transmission costs (Rs. 1053 crores, i.e. 8.5%) and O&M costs (Rs. 1025 crores, i.e. 8.27%) are key components. You can figure this out by looking at the state electricity regulator tariff order for 2015-16 (eg: google search "Telangana electricity tariff order 2015-16". The power purchase costs dominate your tariff (almost 80%, or 88% including transmission costs)!

As the cost of power purchase increases (including spot market / merchant power purchases) they are passed on to the consumer. The "clean environment cess" (previously the "clean energy cess", or coal cess) affects the tariff since it is a pass through item. In the 2016 budget, the government has raised the coal cess to Rs. 400/tonne and there is a stated incremental impact of 15 paise/unit. In my personal view the impact could be variable across sectors and states since the coal cess is a form of carbon tax, and affects both captive, merchant and wholesale/retail energy sectors where coal is being used the fossil fuel to drive energy production. Concentrated solar may now be considered by industries to augment coking coal.

The impact on lower-grade coal such as lignite (or brown coal) is likely to be higher since it is cheaper (Rs. 700-800/tonne vs Rs. 3000/tonne for imported/higher quality coal), its transport costs are higher (eg: railway wagon and rail costs), its gross calorific value GCV is lower (i.e. you need to burn >twice the amount of lignite to generate the same amount of electricity & deal with the ash). If a particular state has more of lignite coal inputs (which will see 25%+ input cost increase due to the cell), its average power purchase cost (APPC) could go up significantly, which will subsequently reflect in ARR and pass through to the retail tariffs. Even at the high end of Rs. 3000/tonne, this is a 10% cost (or 5% incremental cost increase which is passed through). Since power purchase cost is 80% of your retail bill, then, we can expect a 3-4% increase in your bill due to the cess unless it is blended with significant amounts of hydro or natural gas power. For example, at Rs. 6.4/unit marginal rate in Karnataka, I am guessing that it could lead to a 19-25 paisa/unit impact.

In summary, historically power purchase costs has increased at 3-5% nationwide, and given the subsidies to agriculture sector, and coal-cess trends, the rest of the retail tariff sectors are likely to see a trend tariff increase between 4-7% conservatively. Compare this to solar LCOE which is equivalent to a fixed tariff cost over a long period, such as 25 years. How can we implement this ? Lets go on to FIT, net metering etc next.

#2 Feed-in Tariff (FIT), Net Metering:

Feed in tariff (FIT) is the idea that as a solar plant owner, you are not subsidized for your capital investment directly (i.e. not a capital subsidy), but you are compensated for a productive asset, i.e. at a unit rate when you actually produce electricity and feed it into the grid. The amount of compensation may be fixed or variable, and can be implemented in different ways: net-metering, gross metering, generation-based incentives etc. 

Feed in tariffs were pioneered in Germany as part of their Energy transition (Energiewende) program. The feed in tariffs began at a high level (over 50 c/kWh and have now dropped below 3.3 c/kWh), i.e. for the solar PV systems installed early, those homes and businesses signed a long term (20+ year) power purchase agreement (PPA) with their utility. The programme was financed by the government with a surcharge on electricity users. Financial incentives for solar PV in several other countries are documented here.

"Net" metering is the idea where your electricity meter runs backward when you generate solar. A unit (kWh) of solar energy generated is offset against your consumption (which is priced at the marginal rate). If you are a low usage user (eg: 100 units /month), your relevant comparable to LCOE is your marginal (or average rate) which is Rs. 5.4 / kWh in Karnataka in residential, but Rs. 7.35/kWh in large commercial and Rs. 7.95 in small commercial. In some states the settlement is done month by month, and in some cases, the settlement is annually.

Since we saw that the tariff policy is different by segment (eg: commercial vs industrial vs residential) and embeds subsidies (eg: residential, agricultural irrigation pumps), when you net out solar production, what you are doing is either competing against a subsidized tariff (residential) or a loaded tariff (eg: commercial). Also the number of units of production matters. A 1 kWp system will produce 4-5 units/day or 120-150 units / month. Most states limit your system size to your connected load, often 3-10 kWp per electricity meter (note some homes have multiple meters for tenants/floors, even though there is one roof). 

So if you have a 3 kWp system and generate 450 units / month and have only 300 units of consumption, what happens?  The first 300 units of generation will wipe out your bill, i.e. your effective unit feed-in tariff will depend upon the tariff policy (eg: 2.7 / 4 / 5.4 / 6.4 Rs slab rates in Karnataka LT residential). Beyond 300 units, i.e. for the remaining 150 units, the solar tariff order will specify what the feed-in tariff will be. In Karnataka, the current excess generation FIT is Rs. 9.56/unit. This policy incentivizes the maximization of the size of the solar rooftop plant and maximization of production yield of the plant (eg: handling shading / soiling as discussed in Part 2 of this series). If you have a connected load of 10 kW and large roof in Karnataka, you should finance and install 10 kWp system asap provided you don't overpay for the EPC (i.e. around 75-80/Wp or 7.5L-8L for 10 kWp system as of early 2016).

However in many states (eg: Telangana), the excess generation FIT is only the wholesale price of electricity (or "APPC", i.e. average power purchase price) that may range from Rs. 3.5-4.5 / unit depending upon the state. For example, Telangana has a higher marginal residential tariff rate (Rs. 8.8/unit), but a lower FIT beyond the units consumed. In this policy, the optimal response is to size the system to offset consumption and no more, i.e. if you have 300 units/months consumption, install a 2 kWp system. If you have a lot of air conditioners etc and consume 500 units/month, go for 3-4 kWp system. If your seasonal consumption is highly variable (i.e. min of 300 units, and max of 600 units for 3 months in summer), then pick a number between the two values, i.e. about 3 kWp.

In countries like Germany, Australia where the FIT level has dropped below the retail price of electricity (which is about 25 c/kWh), and there is a net metering policy in force, there is an interesting "inversion" effect. My colleagues at IBM Research - Australia, and University of Melbourne studied the effect of this inversion effect on the adoption of energy storage (see paper here). Depending upon the spread between the FIT price and the retail price of electricity, it is now becoming economical for new solar PV installations to add in energy storage (eg: Tesla batteries; Australia has become an important market for Tesla Energy), and offset regular electricity use even when the sun is not shining (eg: evening peak prices). In some parts of Australia, either the utility imposes or offers a choice for time-of-use (ToU) pricing where the peak time prices can be quite high (eg: 50+ c/kWh).

Similarly in commercial buildings in Australia, there is a significant "demand charge", i.e. a component of tariff that is based upon the peak power level (i.e. kW and not kWh) consumed in ANY 15-min or 30-min interval that can add a significant penalty to the monthly bill. There are even proposals for such demand charges to be introduced in residential / small commercial space to optimize grid investments. So a smart software/analytics system that is aware of the particular tariff regime nuances (ToU, demand charge, availability of energy storage, forecasted demand/weather patterns) can optimize your monthly bill. Broadly this is a space for an emerging market for distributed energy resources (DER) and there is a lot of interest around it in New York's Renewable Energy Vision (REV) process.

#3 Gross Metering and Generation-based Incentive (GBI):

Karnataka is currently considering a gross-metering scheme for residential FIT. In such a scheme (similar to parts of Australia and Germany), the solar PV output is measured and considered fed-in to the grid first and compensated directly by a fixed FIT for all the solar units produced. This scheme is called "Gross Metering" , and is considered better at the earlier stages of market development as in India (i will explain why shortly). Note that this gross metering FIT policy can change over time: in Australia, the amount was initially set very high (eg: 50-60 c/kWh for early adopters, who presumably paid much higher capital costs), and then reduced with time as capex reduced (now about 5-6 c/kWh).

In Australia in some states, (eg: New South Wales, ACT) the feed in tariff started off as a gross-metering tariff and then later transitioned to net metering. Jigar Shah, the prominent solar entrepreneur, has often suggested a volumetric reduction in overall subsidies. It is also important to note that there is a significant tussle between utilities, regulatory bodies, government in setting the levels of FIT and even the continuation of policies like net metering. Spain had a significant back-and-forth between very solar-friendly policies and punitive arrangements against solar. Recently Nevada and Hawaii made significant changes in solar net metering policies; and California upheld net metering sparking a huge debate in the USA on the future of rate design. One article even posed this as an epic battle between Warren Buffett and Elon Musk, asking the rhetorical question "Who owns the sun?". At a more sober level, solar PV does pose externalities on the grid (subject of a future article), especially at higher penetration levels, but when controlled and managed along with storage, can be a outstanding distributed energy resource (DER).

Coming back, in the context of India, one benefit of gross-metering is that that gross metering receives fixed rates, and can be deposited into an escrow account unlike net metering where different units of solar production get different rates, which is netted off against an electricity bill. Assuming that the FIT level is good, this is very interesting from the point of view of financing: a financier (or third-party provider) can now bank on more predictable cash flows (which is a function of solar production alone), and can directly take a slice of the incoming flows into an escrow account (rather than being a function of "raw" credit risk of the home owner, with the caveat that the system is in good condition, is not damaged/stolen and produces/feeds-in electricity).

I personally believe that gross metering, along with escrow accounts to receive this cash flow, can enable new business models especially in India at this nascent stage of solar development, especially third-party ownership (like SolarCity in USA). I personally look forward to Karnataka pioneering this to include the startup capital of India, Bangalore. The reason third party ownership of solar PV systems hasn't taken off in India include credit/performance risk, and the solar net metering policies often explicitly do not allow the system to be in the name of an owner different from the meter. BESCOM recently allows escrow accounts even with net metering, but I believe that that will only receive compensation for excess units beyond the monthly bill.

One variant of gross metering is generation-based incentive (GBI). This is similar in idea to gross metering i.e. a feed in tariff based upon actual generation & fed-in amount is specified which is same over the number of units, and has been used for wind energy at the utility scale also in the past. However, a key difference is that it is currently used as a way to complement net metering schemes to provide some relief for the fact that there is subsidy embedded in existing residential tariff schemes that are offset by solar net metering. Delhi and Tamilnadu are mooting GBI schemes in addition to net metering. The GBI rates are of the order of Rs. 1-2 / kWh, which adds on to the benefit of net metering and/or FIT beyond the monthly bill amount. Often the GBI incentives are associated with limits (eg: for the first X megawatts of solar installations etc) similar to the volumetric policy phaseout suggested by Jigar Shah.

#4 Accelerated Depreciation (AD)

As I mentioned earlier, commercial tariffs are high. Profit-generating businesses (or third party owners of solar PV systems who have other profit/revenue flows) can take advantage of a tax scheme for solar PV called accelerated depreciation. {note: This scheme is not available for residential self-owned systems and there is some talk (but no action yet) about giving tax credits on income tax returns, similar to the ITC scheme in the USA.} In India, my understanding is that a business can depreciate 80% of the solar PV plant in 1 year and 20% in the 2nd year.

At a 30% tax rate, assuming you have EBTDA (i.e. business revenues AND earnings before depreciation and taxes, but AFTER paying interest for debt financing) AD translates into 24% effective subsidy in year 1 and 6% subsidy in year 2. Put in standard corporate finance terms, Accelerated depreciation (AD) can boost the project internal rate of return (P-IRR) for commercial LT tariffs (eg: Rs. 7.95/unit tariff) to the 18-20% range; and if you can arrange long term debt financing, the equity IRR (E-IRR) can jump up to 30%+ depending upon the term & rates of financing for a low-risk asset class (remember that the sun rises every day on average with ~100% probability!). In other words, this is an amazing return and a no-brainer for investment if you have a good profit earning enterprise (paying commercial or industrial tariffs) and a roof with decent sunlight exposure.

What if you do not have enough EBTDA? Then the depreciation benefit is lower, but it rolls over to subsequent years, i.e. you effectively depreciate over the period when you have enough post-interest earnings, to which depreciation can be applied to, and you remain with a positive (or zero) profit. For example the depreciation may effectively work out over 3-4 years instead of 2 years. This is still a "decent" deal.

ps: As per the latest 2016 budget, it is anticipated that the AD (accelerated depreciation) amount is likely to reduce from 80% to 40%, which means that the solar PV system will get depreciated over 3 years rather than 2 years. This is still valuable given the fact that it has significant residual value & robust production for 25 years+ when maintained well.

#5 Capital Subsidy for Rooftop Solar

The government has recently announced a 30% capital subsidy for rooftop solar  and funded it up to Rs. 5000 crores.  This is also targeted for residential solar. There appears to be a domestic content requirement (i.e. solar panels produced in India), and a process of working with empanelled installers and module makers. The details and processes etc for applying for subsidy appear to be still worked out by  the government.

Here are some extracts from the announcement:

"The capital subsidy of 30% will be provided for general category States/UTs and 70% for special category States i.e., North-Eastern States including Sikkim, Uttarakhand, Himachal Pradesh, Jammu & Kashmir and Lakshadweep, Andaman & Nicobar Islands, There will be no subsidy for commercial and industrial establishments In the private sector since they are eligible for other benefits such as accelerated depreciation, custom duty concessions, excise duty exemptions and tax holiday etc."

In the past, MNRE had a policy of between 15-30%, but given the limited amount of funds available, at the same time they encouraged you to go ahead without waiting for subsidy, since there was a priority list. Industry observers also have reported long wait times, and the default has been to go ahead without factoring in the subsidy. If you are considering foreign-made panels (eg: Trina, ReneSola, Solar Frontier, Canadian Solar etc), you should not wait for this subsidy (in the long term there makers will likely have a manufacturing facility in India).

MNRE also tracks the policies of different states and provides overall guidelines and policies. Net metering policies have been announced in several states in India. MNRE has a list of states and policy documents (Caveat: I haven't studied all of them - but this article should help you understand the impacts and implications!).

Miscellaneous and Conclusion

There a variety of other policy tools: RETs, community solar / virtual net metering etc. Renewable Energy Targets (RETs) for small scale or large-scale renewable deployments which tend to work when it has the enforcement backing (vs being a target without teeth). The idea here is to specify a quantity (of solar MW deployed for example by a time frame) instead of a price (eg: feed in tariff or a direct capital subsidy). In the USA, a combination of RETs, net metering policies in key states (esp california), and a more powerful direct subsidy, the federal investment tax credit (ITC) which can be transferred to tax equity investors (and other financial innovations to drive third party ownership) has led to the growth in solar. In Australia, the RET has been a politically sensitive topic and a consensus reached recently. The Australia small-scale RET scheme has in combination with the net metering / Feed-in tariff program, led to a huge growth of solar PV.

More recently a set of schemes for "community solar" (and one of the schemes is virtual net metering) has been proposed to allow homes and communities to invest in, or buy as-a-service solar PV plants installed in the communities or remotely. In the context of India, we have an open access policy for captive or group-captive ownership by large consumers (1 MW+ connected load) of ground mounted PV plants; combined with attractive policies in some states (eg: Karnataka waives the cross-subsidy surcharge for 10 years). If you are in a community such as a housing society with connected load less than 1MW (eg: 100-150kW), the common areas (eg: elevators, water treatment etc) can be offset with a solar PV canopy over the parking areas, and net metering can apply. 

To sum up, with the (a) excellent solar resource in India, (b) drop in capital cost of solar PV systems - India is the lowest in the world already - (Rs. 65-70,000/kWp for 50kWp and Rs. 75-80,000/kWp for 10kWp), (c) simple practices in operating/maintaining them (clean your panels well every week & watch out for shadowing!), (d) better financing availability (more innovation needed here to distribute the financing from banks, infra-dev sources), and (e) policy support mechanisms described above, rooftop solar PV is at the cusp of a huge ramp in deployment in India.

Twitter: @shivkuma_k

ps: If you like this post, consider reading some of the companion articles: "Distributed / Rooftop Solar in India: A Gentle Introduction: Part 1","Rooftop Solar in India: Part 2 {Shadowing, Soiling, Diesel Offset}", "Rooftop Solar in India: Part 3: Policy Tools... Net Metering etc..." "Solar Economics 101: Introduction to LCOE and Grid Parity" , "Solar will get cheaper than coal power much faster than you think..", "Understanding Recent Solar Tariffs in India", "How Electric Scooters,... can spur adoption of Distributed Solar in India," "Solar + Ola! = Sola! ... The Coming Energy-Transportation Nexus in India", "UDAY: Quietly Disentangling India's Power Distribution Sector", "Understanding Solar Finance in India: Part 1", "Back to the Future: The Coming Internet of Energy Networks...", "Tesla Model 3: More than Yet-Another-Car: Ushering in the Energy-Transportation Nexus", "Understanding Solar Finance in India: Part 2 (Project Finance)", "Ola! e-Rickshaws: the dawn of electric mobility in India", "Understanding Solar Finance in India: Part 3 (Solar Business Models)" . 

pps: List of all my LinkedIn Articles