The Reform Indian Grid has been Waiting for : Capacity Market

What It Is, Why It Matters, and What's Still Missing

 CERC's Staff Paper on Capacity Market for Electricity in India, 29 April 2026

1.  What Is This About?

India's electricity grid has a problem that sounds strange: we have built a lot of power plants, but the lights still go out — or nearly go out — at certain times of day. How can there be a shortage when there is plenty of capacity on paper?

The answer is that the power plants we have are not always available when we need them most. A coal plant contracted to a state electricity board might be sitting idle in the afternoon while the grid is struggling during the evening peak. There is no market mechanism that says: 'Show up at 7 pm when demand is highest.' The plant just has to be technically available — not actually present at the critical moment.

In April 2026, India's electricity regulator — the Central Electricity Regulatory Commission (CERC) — released a Staff Paper proposing a new type of market called a Capacity Market. The idea is simple: pay power plants not just for the electricity they produce, but also for the promise that they will be available exactly when the grid is under pressure. This document explains what CERC is proposing, why it matters, and what important questions the proposal leaves unanswered.

2.  How the Current System Works — and Why It Is Failing

Right now, most electricity in India is bought through long-term contracts called Power Purchase Agreements, or PPAs. A state electricity company (called a DISCOM — Distribution Company) signs a contract with a power plant for 15 or 20 years. Under this contract:

•        The DISCOM pays a fixed monthly charge just for keeping the plant available — regardless of whether it uses the electricity.

•        When the DISCOM needs power, it tells the plant when to run and how much to generate. This is called the scheduling right.

•        The plant is paid separately for the actual electricity it generates.

This system worked well when the main challenge was simply building enough power plants. But it has serious weaknesses today:

•        The DISCOM controls when the plant runs — but the DISCOM's decision is based on its own billing needs, not on what the grid needs at that exact moment.

•        A plant contracted to one DISCOM cannot easily help another state that is short of power, even if it is sitting idle.

•        There is no extra reward for being available during the most difficult hours — say, a hot summer evening when solar power has switched off and everyone switches on their air conditioners.

•        India is chronically short of standby or reserve power — capacity that exists purely to handle sudden spikes or emergencies. When a large plant trips unexpectedly, there is often no backup.

The reserve problem is the most acute manifestation of this structural gap. India's ancillary services framework — Secondary Reserve Ancillary Services (SRAS) and Tertiary Reserve Ancillary Services (TRAS) — exists on paper but is chronically short of actual contracted reserve capacity. During solar hours, net load demand falls sharply, thermal units back down to minimum technical minima, and the grid is left with almost no spinning reserve. The Ministry of Power's annual Section 11 directions, contracting emergency gas capacity worth thousands of crores each year, are essentially an admission that the market structure is not providing what the grid needs.

3.  So, What Exactly Is a Capacity Market?

Think of it like hiring a plumber on a retainer. You pay them a monthly fee not because your pipes are leaking right now, but because you want the guarantee that if something goes wrong, they will show up. If your pipes do burst and they fix them, you pay extra for the actual work. But the retainer is the promise of availability.

A capacity market works the same way for electricity:

•        Power plants (or batteries, or factories willing to cut their demand) get paid a regular fee just for promising to be available when the grid needs them most.

•        When the grid actually needs them, they generate electricity and earn the market price for that electricity too.

•        If they promise to be available but fail to show up when called, they pay a penalty.

Why does this matter?  Without a capacity market, power plants only earn money when they actually sell electricity. If prices are too low or demand is uncertain, investors don't build new plants. The result is a future shortage. A capacity market solves this by giving investors a reliable income stream just for building and maintaining capacity — which encourages investment and keeps the lights on.

4.  How Other Countries Handle This

The CERC paper surveys a range of international capacity market models. The common thread is revealing: capacity is contracted on a Net-CONE basis (the cost of a new entrant adjusted downward by expected energy market revenues), load-serving entities bear the capacity charge but do not receive scheduling rights in return, and dispatch is indexed to real-time market conditions. These are precisely the features India's PPA-based system lacks.

India is not starting from scratch. Several countries have been running capacity markets for years, and CERC has studied their experiences. Here is a simplified comparison:

Germany's Strategic Reserve sits entirely outside energy markets and is activated only in extreme stress. Britain's Capacity Market runs descending-clock auctions four years ahead of delivery (T-4), paying capacity providers a monthly fee while they remain free to sell energy. PJM's Reliability Pricing Model anchors to a reliability standard of no more than one expected loss-of-load event in ten years, running three-year-forward Base Residual Auctions. California leaves RA contracting to individual LSEs but provides a centralized backstop when they fall short.

The key lesson from all these international examples: in a capacity market, the electricity company that pays for capacity does not get to decide when the power plant runs. Dispatch — the decision of when to generate — is handled by the market and the grid operator, not by the buyer. This is a big change from how India currently works.

5.  What CERC Is Actually Proposing

CERC's staff paper proposes three separate markets running alongside each other. Think of them as three tools addressing three different problems.

5.1  Track A: Long-Term Capacity Contracts (The Big Structural Change)

This is the most structurally ambitious track. It proposes decoupling the capacity contract from the energy dispatch arrangement — separating the 'right to capacity' from the 'right to dispatch.' Three options are offered:

•     Option 1 (most market-oriented): DISCOMs bid based solely on the capacity charge using a Net-CONE (the Cost of a New Entrant) demand curve. Both buyers and generators participate in Day-Ahead and Real-Time markets independently. Capacity providers must bid as price takers during NLDC-notified stress hours, with a penalty of 1.5x the discovered capacity charge for failure to clear.

•    Option 2: Retains the two-part tariff (capacity + variable charge) but modifies dispatch — DISCOMs retain scheduling rights only until the day-ahead market window opens. Settlement is based on a contract-for-difference structure between the MCP and the contracted price

•    Option 3 (most pragmatic near-term): A centralized Residual RA Capacity Market where a designated central agency procures capacity only for verified shortfalls after DISCOMs have met their own RA obligations. This preserves existing bilateral PPAs while layering market-based signals for new capacity

Three options are proposed for how this works in practice. Option 3 is the most likely starting point — it only creates a new market for the gap between what DISCOMs have already contracted and what they actually need. This is less disruptive and can start without dismantling existing PPAs.

 5.2  Track B: Reserve Market (The Most Urgent Fix)

This track tackles India's most immediate problem: the shortage of standby power for emergencies.

Right now, when the grid operator (NLDC) needs emergency capacity, it has to scramble — issuing special government orders, paying whatever it takes. Track B would change this by holding planned auctions every year to book standby capacity in advance, just like a hospital keeps blood banks stocked before an emergency, not after.

How it would work:

•        NLDC calculates how much standby capacity each state needs.

•        States that fall short have to join a central auction to buy the shortfall — and pay for it.

•        Power plants, batteries, and RE-plus-storage units bid to provide this standby capacity.

•        The winning plants earn a capacity fee just for being on standby. If actually called, they earn extra.

Example:  Imagine Uttar Pradesh needs 800 MW of standby reserve but only has 500 MW lined up through its own contracts. NLDC runs an auction. A gas plant bids to provide 300 MW for Rs. 50,000 per MW per year. It gets selected. For the next year, it earns Rs. 150 crore just for being available — even if it never actually generates a single unit. UP bears the cost because it failed to plan adequately.

 5.3  Track C: Short-Term Capacity Trading (The Quick Win)

This is the simplest and most immediately doable of the three tracks. It creates a marketplace where DISCOMs that have more contracted capacity than they need can sell the surplus to DISCOMs that are short.

Right now, a DISCOM in Tamil Nadu might have 500 MW of spare contracted capacity sitting idle, while a DISCOM in Rajasthan is scrambling to meet demand. There is no easy mechanism for Tamil Nadu to sell that spare capacity to Rajasthan. Track C creates quarterly auctions where exactly this kind of trade can happen — unlocking capacity that exists in the system but is currently invisible to it.

Critical Design Tensions

5.4  Partial Recovery and Investor Comfort

One of the deepest structural shifts is the move away from guaranteed full fixed-cost recovery through availability-based tariffs toward market-based capacity remuneration. The paper acknowledges that international experience suggests a robust energy market — specifically, the no-load-shedding guarantee — provides sufficient investor comfort. India's record on zero load-shedding is improving but not yet credible enough to rely on this mechanism alone, particularly for greenfield capacity requiring project finance.

5.5  The DISCOM Credit Problem

Any capacity market ultimately depends on DISCOMs as creditworthy obligated buyers. Many state DISCOMs carry accumulated losses, depend on state government subventions, and struggle to pay existing PPA obligations on time. A capacity market that sends a monetised signal to investors requires a solvent, commercially disciplined buyer on the other side. Without DISCOM financial reform running in parallel — or preceding — capacity market design, the market will be structurally weak from the demand side.

5.6  Gaming Risks in the Stress-Hour Design

The proposal that capacity providers must bid as price takers during NLDC-notified stress hours is elegant in theory but operationally demanding. Defining stress hours ex-ante is difficult without reliable short-term forecasting. The 1.5x penalty for failure to clear is intended as a strong deterrent, but enforcement requires real-time metering and telemetry at a scale India does not yet uniformly have.

5.7  Demand Curve Above the Price Cap

Drawing the demand curve slightly above the price cap — proposed to address the free-rider problem — risks inflating the Market Clearing Price. The paper itself acknowledges this tension. The counterbalance of mandatory price-taker behaviour during stress hours may not be sufficient without robust telemetry and enforcement infrastructure, and could become a source of sustained, above-efficient cost for consumers.

 6.  Key Terms Explained Simply

The staff paper uses several technical terms that are worth understanding before diving into the details.

Price taker: A price taker is someone who must accept whatever price the market sets — they cannot negotiate or hold out for a higher price. In the capacity market proposal, power plants would be required to bid as price takers during emergency hours. This means they cannot name their price and hold the grid to ransom during a shortage — they must offer their power at the going market rate.

Gaming: Gaming means finding clever ways to exploit the rules of a market to earn more money than you should, without actually providing more value. In a capacity market, a power plant might game the system by: deliberately staying offline during non-peak hours to create artificial scarcity, then earning high prices during emergencies. Or it might declare itself 'unavailable' on paper to avoid penalties while actually being available. The 1.5x penalty for not showing up during declared stress hours is designed to prevent this.

Net-CONE (Net Cost of New Entry): This is the benchmark price for capacity — the minimum amount a new power plant needs to earn from a capacity contract to justify being built. 'CONE' is the total cost of building and running a new plant. 'Net' means after subtracting the money it expects to earn selling electricity in the open market. If a new gas plant costs Rs. 80,000 per MW per year to build and run, but it expects to earn Rs. 30,000 per MW from electricity sales, the Net-CONE is Rs. 50,000 per MW. The capacity market uses this as the ceiling for capacity payments — the market will not pay more than what it would cost to build brand new capacity.

 Demand curve (in capacity markets): In a normal market, when price goes up, buyers buy less. The capacity market uses a 'demand curve' to capture this idea. It sets the maximum price it will pay for capacity (the cap) and a target volume it wants to buy. If there is more capacity on offer than needed, the price falls. If there is less, the price rises — but only up to the cap. This prevents both overpaying and getting too little.

 Derating: Not all power plants are equally reliable. An old coal plant might break down unexpectedly 20% of the time. A solar plant only generates during daylight. Derating is the process of giving a plant credit for less than its nameplate capacity based on its actual reliability track record. A 500 MW plant with 20% unreliability gets credited for only 400 MW. This keeps the market honest — a plant cannot promise 500 MW and then fail to deliver.

 Pay-as-cleared: In the proposed auctions, all winning bidders get paid the same price — the price of the last (most expensive) bid needed to fill the required volume. So if 10 plants bid, and the 10th plant bids Rs. 60,000 per MW, all 10 plants get Rs. 60,000 — even if some bid Rs. 40,000. This is different from 'pay-as-bid' where everyone gets only what they asked for. Pay-as-cleared is simpler and generally preferred because it encourages honest bidding.

 Stress hours: These are the specific hours declared by NLDC when the electricity grid is under maximum pressure — typically summer evenings between 6 pm and 10 pm when solar has faded and demand peaks. During stress hours, all plants that have signed capacity contracts must be available and must offer their power at the market rate (as price takers). Any plant that fails to show up during declared stress hours pays a heavy penalty.

 Missing money problem: Power plants earn money by selling electricity. But if electricity prices are kept low by regulation, or if there is too much capacity in the market, a plant may never earn enough to cover its construction and running costs — even though the grid needs it. This gap is called the 'missing money.' A capacity payment fills that gap by providing a separate, reliable income stream just for existing and being available.

 7.  What the Paper Gets Right — and What It Leaves Out

The CERC staff paper is an honest and thoughtful document. It presents options rather than final decisions, and it invites public comment. But there are several important questions it does not answer. These gaps need to be addressed before the capacity market can actually work.

7.1  The Old Power Plant Problem: The Biggest Gap

India has dozens of power plants that were built in the 1990s or early 2000s. Many of them have repaid their original loans, which means they no longer need high prices to survive. They are cheap to run from a fixed-cost perspective. But they are old — their equipment breaks down more often, they are less fuel-efficient, and they may face environmental or water-use restrictions.

Many of these plants are approaching the end of their official lifespan, but they still have physical life left. They are in a kind of no-man's-land: too old for long-term contracts, too useful to simply shut down.

The CERC paper completely ignores this category of plants. It does not say:

•    How their reliability will be measured for derating purposes (an old plant that breaks down more often should get less capacity credit).

•    Whether a plant that invests in refurbishment and modernisation can qualify for a new long-term capacity contract.

•    Whether there could be a special 'standby reserve' category for these plants — keeping them available for genuine emergencies without letting them compete in the main capacity market (which would suppress prices and discourage new investment).

•    How a plant transitions from being a full-time baseload supplier to a part-time peak-hour provider as newer, cheaper solar generation takes over daytime demand.

 Old plants with low fixed costs can bid very cheaply in capacity auctions. This sounds good — lower prices. But if old, unreliable plants fill capacity contracts, they may not actually show up during emergencies. And cheap old capacity discourages investment in new, cleaner, more reliable alternatives like battery storage. Getting the rules right for aging plants is critical for both reliability and the energy transition.

There is also the question of refurbishment. If an old plant spends money on major upgrades — new turbines, better pollution controls, improved fuel efficiency — should it be treated like a new plant for the purpose of capacity contracts? The current rules are silent. A clear answer would encourage owners to invest in improving old plants rather than just running them into the ground.

7.2  The Price Benchmark Has No Price

Every track in the proposed market uses Net-CONE as the upper limit for capacity payments. If the auction price tries to go above Net-CONE, the market stops buying more capacity. This is a crucial guardrail against overpaying.

But the paper never says how India should actually calculate Net-CONE. How much does it cost to build a new gas plant versus a new battery versus a new pumped hydro facility? How much electricity revenue should be assumed? At what interest rate? These are not simple calculations, and getting them wrong — even slightly — can result in either persistent shortages (if set too low) or billions in unnecessary consumer costs (if set too high).

The paper needs to commit to a transparent, independently reviewed Net-CONE study published before any auction is designed.

7.3  State Regulators Are Not On the Same Page

India has 28 states, each with its own electricity regulator (called a State Electricity Regulatory Commission, or SERC). The capacity market proposal requires each state to assess how much capacity it needs, file a plan, and demonstrate compliance.

But there is no protocol for what happens when a state files a plan that the national grid operator thinks is inadequate. Or when a state's regulator approves something that CERC disagrees with. Or when a state simply misses its deadlines. Without a clear coordination mechanism — with binding timelines and escalation provisions — the whole trigger system for the reserve capacity market could get stuck in bureaucratic disputes.

7.4  Large Companies Could Game the System

A few large generation companies in India own a significant share of the country's total capacity. In a capacity auction where few players have a lot of capacity, the big players could potentially manipulate their bids — offering less capacity than available to push up the clearing price, for example. This is called exercising market power, and it is one of the most litigated issues in every capacity market worldwide.

The CERC paper does not mention any safeguards against this: no limits on how much capacity any one company can clear, no review process for suspicious bids, no market monitoring framework. These are not minor details — they determine whether the market produces fair prices or becomes a mechanism for large companies to extract excess profits from consumers.

7.5  The Transition from Old Contracts is Unaddressed

India has thousands of existing PPAs, many running until 2035 or beyond. Under these contracts, DISCOMs currently tell power plants when to run. Under the new capacity market, DISCOMs would lose that right — the market would decide dispatch.

But the paper says nothing about how this transition happens. When does a DISCOM's scheduling right end? If a DISCOM is simultaneously holding an old PPA (with scheduling rights) and a new capacity contract (without scheduling rights), which one takes precedence? How are costs settled when the plant earns capacity market revenue on top of its PPA payment? These conflicts could generate years of legal disputes unless addressed clearly upfront.

 8.  What Should Happen, and in What Order

One of the most important insights in the CERC paper — though stated quietly — is that India cannot simply switch on a full capacity market overnight. The foundations need to be built first. Here is a sensible order:

•    First (now through 2027): Fix the basics. Each state must publish a credible plan for how much power it needs and how it will meet that need. The national grid operator must be able to calculate reserve shortfalls reliably. Energy markets must deepen so that plants have an actual market to sell into.

•    Second (2027–2029): Launch the Reserve Market first. This is the most urgent — and most tractable — problem. NLDC starts buying standby reserve annually. States that fail to plan adequately bear the cost. At the same time, start the short-term capacity trading market to unlock surplus contracted capacity.

•    Third (2028–2031): Introduce longer-term capacity auctions selectively, only for verified gaps that states cannot fill themselves. Use simple Option 3 (centralized residual procurement) rather than the more complex Option 1.

•     Finally (2030s): Gradually wind down old-style PPAs, converting them to the new framework as they expire. Plants transition from DISCOM-controlled dispatch to market-based dispatch. DISCOMs evolve from capacity owners to capacity buyers.

 The key principle:  Do not build a capacity market on top of broken foundations. If DISCOMs cannot pay their bills, if states will not file RA plans, if energy markets are too thin — the capacity market will fail regardless of how well it is designed. The design is only as good as the institutions implementing it.

 9.  What This Means for Different Stakeholders

For power plants

This has potentially significant implications for their business. Under the current system, they are paid for availability under a long-term PPA. In the new system, their revenue will have two components: a capacity payment (for committing to be available) and energy market revenue (for actual generation). The capacity payment may be lower than the current fixed charge, but if they are efficient and reliable, they can earn more from the energy market.

 If their plant is old, the rules are currently unclear. They should push strongly in the consultation process for explicit derating standards, refurbishment eligibility, and a transitional reserve category. These details will determine whether their asset has a viable future under the new framework.

For DISCOMs

They will lose the right to direct power plants on when to run. This may feel like a loss of control, but it could improve their financial position. At present, they pay fixed charges for capacity they do not always require. Under the new system, capacity payments could be lower and more targeted. However, this outcome depends on their ability to file credible resource adequacy plans and meet their obligations. DISCOMs that fail to plan will end up paying for centrally procured reserves—at prices determined by the market, not through negotiation.

For electricity consumers

A well-designed capacity market should deliver more reliable supply—fewer sudden outages during summer peaks and better management of the evening ramp when solar generation drops off. Over time, it should also encourage investment in storage and flexible generation, making the grid more resilient as renewable energy expands.

However, there is a risk. If the market is poorly designed—if old plants game the auctions, if Net CONE is set too high, or if large companies manipulate prices—capacity payments will increase their electricity bills without delivering reliability improvements. The extent to which gaps are addressed in the consultation process will directly determine whether consumers benefit or simply pay more..

For storage or clean energy investors

The capacity market proposal is broadly good news. India's current PPA system strongly favours established thermal plants. The new framework — especially the Reserve Market with multi-revenue stacking — creates a genuine revenue case for batteries, pumped hydro, and RE-plus-storage projects. For the first time, a storage project could earn from: a capacity payment for being available, ancillary service payments for providing grid balancing, and energy market revenue when it actually discharges.

The catch is that derating rules for storage and renewables have not been defined. Until CERC publishes technology-specific derating factors, it is impossible to know exactly what capacity credit a battery project will receive — and therefore impossible to fully model the investment case.

10. Where Should Capacity Be Traded?

The CERC paper proposes three distinct capacity tracks but says surprisingly little about where they should actually be traded — and the venue choice matters as much as the design itself. A one-size-fits-all approach will not work; each track has different needs:

• Short-term capacity trading (Track C) and the reserve market (Track B) belong on a power exchange — platforms like IEX or PXIL are well-suited because the products are standardised, contracts are short, and exchange settlement guarantees protect against DISCOM payment defaults, something bilateral contracts have repeatedly failed to prevent in India.

• Long-term RA capacity contracts (Track A) cannot work on a pure exchange — a fifteen-year capacity agreement involves lender due diligence, customised force majeure clauses, and payment security arrangements that no exchange can accommodate. The right model is what Britain uses: a centralised, transparent auction to discover the price, followed by bilateral contracts between a central agency and the winning generators.  

• Counterparty credit risk is the decisive factor in the Indian context — DISCOMs are notoriously poor payers, and exchange margining systems that automatically close out positions when a buyer defaults offer far stronger protection than chasing dues through regulators or courts.

• The paper's silence on trading venue is itself a gap — leaving the question of exchange versus OTC to be decided later creates regulatory uncertainty that will deter storage developers, aggregators, and flexible capacity providers from participating — precisely the new entrants the market most needs.

• In short, India's capacity market needs a hybrid approach: exchange infrastructure for the short end, auctioneer-plus-bilateral-OTC for the long end. This distinction should be resolved before the framework is finalised, not after the first auction is designed.

 11.  The Bottom Line

CERC's capacity market proposal is a serious, well-intentioned attempt to fix a structural problem in India's electricity sector. The problem is real: India's grid needs reliable, dispatchable capacity available at the moments of highest stress, and the current PPA system does not efficiently deliver this.

The three-track approach is sensible — addressing the most urgent problem (reserve shortage) first, unlocking existing capacity through short-term trading next, and only then creating new long-term capacity signals. The international experience has been studied carefully, and the technology-neutral stance is the right one for a country that needs both old thermal plant reliability and new storage scale-up.

But the paper has significant gaps that the consultation process must fill. The treatment of aging plants is the most glaring omission — a large, economically important asset class with no design clarity. The absence of a Net-CONE methodology, a CERC-SERC coordination protocol, market power safeguards, and a PPA transition framework are equally important.

This is an open consultation. Every stakeholder — power plant owners, state DISCOMs, consumer groups, renewable energy developers, and independent analysts — has an opportunity to push for better answers before the framework is finalised. The quality of those responses will determine whether India gets a capacity market that actually keeps the lights on, or a bureaucratic structure that adds cost without adding reliability.

Source: Staff Paper on Capacity Market for Electricity in India, April 2026. Central Electricity Regulatory Commission, New Delhi.

Disclaimer: This is an independent plain-language commentary on the CERC Staff Paper. It does not represent any regulatory or official position. Technical readers should refer to the original Staff Paper for precise definitions and regulatory language.