Power, Utilities & Renewables

Generation is the manufacturing step. A plant converts coal, gas, water, sunlight, wind or nuclear fuel into electricity. It is capital-heavy, and once built its economics depend almost entirely on two things — how much it costs to run and what price it can sell at.

Transmission is the highway. High-voltage lines carry bulk power from where it is generated to where it is consumed, often across states. A transmission company does not buy or sell electricity; it charges for the use of its network. That is a completely different business from generation — closer to a toll road than to a factory.

Distribution is the last mile. Distribution companies, commonly called discoms, buy power in bulk, run the local network, meter every household and business, and bill them. Most are owned by state governments, though there are private and franchised distribution businesses too. This is where consumer money enters the chain, and therefore where every rupee that flows back upstream has to pass.

Once you see it as three businesses, an obvious question follows for each: who decides the price?

This table is the whole sector in one screen. The price-setter determines the risk, and the risk determines what you should be watching.

Transmission and a large part of Indian generation operate on a regulated-return model, and it is worth being precise about what that means, because it is unlike almost anything else on the market.

The regulator approves the capital cost of an asset. It then allows the company to recover its operating cost and interest, plus a return on the equity invested in that approved asset base, provided the asset meets availability norms — broadly, that it is ready to run when called. Note what is missing: the market price of electricity. A regulated transmission line earns the same allowed return whether power prices doubled or halved.

So the investment case for a regulated business is not about electricity prices at all. It rests on two questions. First, how fast is the approved asset base growing — because profit grows when the company commissions more regulated assets, not when tariffs rise. Second, does the counterparty actually pay, because an allowed return that is billed and not collected is an accounting entry, not cash.

There is a third risk, less discussed: under-recovery. If the regulator disallows part of a claimed cost, or a project is delayed and the asset is commissioned later than assumed, the company earns less than the model implied. Regulated does not mean guaranteed; it means the price mechanism is administrative rather than commercial.

Merchant power is electricity sold without a long-term contract — into a power exchange, or through short-term bilateral deals, at whatever price the market clears at that hour.

That price is genuinely volatile. It moves with the weather, with how much hydro capacity is available, with how much wind and solar the grid received that day, with industrial demand, and with fuel costs. A hot, still afternoon with low hydro availability and high air-conditioning load is a very different market from a windy monsoon week.

A merchant plant is therefore a commodity business with a fixed cost base, and that combination produces enormous operating leverage in both directions. Fixed costs — depreciation, interest, fixed operations and maintenance — do not fall when the clearing price falls. When the price is well above the plant's variable cost, almost every extra unit sold drops through. When it is below, running the plant loses money.

Most listed generators are not purely one or the other. They have a contracted base and a merchant tail. The mix is the thing to find, because it tells you how much of the company's earnings is a bond and how much is a commodity.

Two identical power plants, same technology, same size, same location, can be completely different investments. What separates them is the contract, not the concrete. The power purchase agreement, or PPA, is the long-term contract under which a generator sells its output to a buyer — typically a distribution company or a large industrial consumer.

A well-structured PPA converts an uncertain commodity operation into a contracted cash flow. That is the single largest value transformation in this sector, and it happens on paper. So when you assess a generator, assess its contracts first.

Almost every recurring problem in Indian power traces back to distribution finances, so it is worth understanding the mechanism rather than just the headline.

A discom buys power in bulk and sells it to consumers at tariffs set by a state regulator. Those tariffs are set with social and political considerations, not only cost ones: agricultural and domestic consumers are often charged below the cost of supplying them, with the gap meant to be covered by higher industrial and commercial tariffs and by state government subsidy. When the subsidy arrives late or falls short, the gap sits on the discom.

On top of that sits AT&C loss — aggregate technical and commercial loss. It combines two different problems in one number: power physically lost in the network, and power that was delivered and billed but never collected. Both mean the discom paid for units it earns nothing on.

The result is a distribution business that is structurally short of cash. And because it is the only link that touches the consumer, its shortage becomes everyone else's. It delays payment to generators and transmission companies. Those companies' receivable days stretch. To keep operating they borrow working capital, which adds interest cost. A late payment surcharge exists in principle, but a counterparty that cannot pay the bill generally cannot pay the surcharge either.

This is why payment days upstream are a genuine investment metric in power and not a housekeeping detail. Central government schemes have repeatedly targeted this problem — through loss-reduction programmes, metering pushes and payment-security mechanisms requiring letters of credit before dispatch — and progress has been real but uneven across states.

The standard operating metric for a generating station is plant load factor (PLF). It answers a simple question: of everything the plant could physically have produced in the period, how much did it actually produce?

For a merchant plant, PLF is close to everything. It ran because someone bought the power at a price above its variable cost. High PLF means the plant was competitive; low PLF means it was not dispatched.

For a plant on a two-part PPA, PLF matters much less than most people assume. The capacity charge is earned for being available, not for running. A plant can have modest PLF and still collect its full fixed payment, because the buyer chose not to dispatch it. What would hurt is a forced outage that pulls declared availability below the contracted norm — that reduces the fixed payment directly.

For renewables the equivalent metric is the capacity utilisation factor, and it is structurally much lower than thermal PLF. A solar plant cannot generate at night. A wind farm generates when the wind blows. A low number here is physics, not underperformance, and comparing a solar plant's utilisation with a coal plant's PLF is a category error. Compare a solar asset against solar assets in comparable locations.

For a coal-fired plant, fuel is the largest running cost, and where the coal comes from changes the risk profile entirely.

Domestic coal supplied under a long-term linkage from a state-owned miner is the most stable source — administered pricing, though not always assured quantity. Coal bought through e-auction costs more and moves with domestic supply and demand. Imported coal is priced in international markets and paid for in foreign currency, so its cost moves with both the global coal price and the rupee.

Whether that variability actually reaches the profit line depends, again, on the contract. Under a PPA with full fuel cost pass-through, a coal price spike moves to the buyer. Under a fixed tariff won in a competitive bid without adequate pass-through, the generator absorbs it — and that mismatch, between a fixed selling price and a floating fuel cost, is the origin of a great deal of historical stress in Indian thermal generation.

Gas-based generation has its own version of the same problem. Where domestic gas is unavailable, plants run on imported liquefied natural gas, whose price makes generation expensive relative to alternatives, so those plants often run at low utilisation regardless of how much demand exists.

A renewable project inverts the cost structure of a thermal plant. Almost the entire cost is spent upfront, on modules or turbines, land, evacuation infrastructure and construction. Once built, the marginal cost of the next unit is close to zero, because there is no fuel to buy. That is why renewables are dispatched first wherever they are available and generally hold must-run status on the grid.

The difficulty is intermittency. Output follows the sun and the wind, not demand. Solar peaks in the middle of the day; consumption in India often peaks in the evening. So a grid with a lot of renewable capacity needs something to fill the gap — battery storage, hydro, flexible thermal capacity that can ramp up and down, and enough transmission to move power from a windy or sunny state to a demand centre elsewhere. This is why grid balancing, storage and transmission investment are not side topics in the renewable story; they are part of its cost.

There is also curtailment risk. If the grid cannot absorb the power at that moment, generation is cut back. Contracts usually address this, but the strength of that protection varies.

Most of these projects are sold through long-term auctions at a fixed tariff, which means the revenue line is locked at the moment of bidding, for decades. And they are financed like infrastructure — heavy long-tenure debt against a contracted cash flow. Put those two facts together and you get the sector's most under-appreciated sensitivity. The revenue is fixed by the auction, the cost of the debt is not. When interest rates rise, the same project bid at the same tariff produces a lower return on equity, and new capacity has to be bid at higher tariffs or not bid at all. This is exactly the sensitivity infrastructure carries, arriving through the financing rather than through demand.

Electricity demand data gets reported as though it were a sector earnings forecast. It is not, and there are at least five reasons the link breaks.

First, most contracted generation earns a tariff fixed by its PPA. Higher national demand does not raise that tariff. Second, the extra demand may have been served by renewables, which are dispatched first because their marginal cost is near zero — so the additional units went to somebody else's asset. Third, the marginal unit may have been supplied by a plant whose fuel cost sits above the clearing price, meaning it ran without earning. Fourth, the buyer may have taken the power and not paid for it, which converts extra demand into extra receivables. Fifth, and most awkward: if a discom is selling below its cost of supply to a consumer category, more demand from that category makes its finances worse, not better — and a weaker discom pays everyone upstream more slowly.

The correct chain of questions is therefore: which link, whose asset served the extra demand, at what price, under what contract, and did the buyer pay. Demand growth is the first link in that chain, not the conclusion.

Power is one of the better-instrumented sectors in India for a public investor. Generation and demand data are published frequently, exchange prices are visible, and regulatory orders are public documents. The routine below is mostly a matter of collecting the right series.

Everything above is a map of how the sector works. It does not describe any individual business, and in power the distance between the sector and the company is unusually large.

Two generators with the same installed capacity can have entirely different earnings quality depending on their contract mix, their counterparties, their fuel arrangements and their debt. A regulated utility with a growing asset base and a regulated utility with a stagnant one look identical in a sector description and behave nothing alike. A renewable developer's economics are set by decisions taken at the moment of bidding, years before the asset generates a single unit.

Use this framework to ask the right question of a company — which link, which contract, which counterparty, what debt — and then answer it from that company's own disclosures, its regulatory filings and its cash flow statement.