Pricing The Planet: India's Carbon Markets Go Live

In this edition, we unpack the two levers India has for cutting its carbon emissions while still growing as an economy, and the opportunity for startups as the country opens up its carbon market.

In January 2023, an investigation found that more than 90% of the rainforest carbon credits issued by Verra - the world’s largest carbon credit certifier - were “phantom credits” that didn’t represent real emissions reductions. Companies including Disney, Shell, and Gucci had spent years buying offsets for forest protection that, on closer inspection, mostly wasn’t happening. Confidence in a carbon market built over two decades cracked almost overnight.

Three years on, that crisis is still reshaping how the world prices a tonne of carbon. India launched its Carbon Market Portal at Prakriti 2026 in March, and has since been registering more than 700 obligated entities across nine heavy-emission sectors. The credits are expected to be issued from October, with formal trading opening on approved power exchanges in the months that follow.

In this dispatch, we walk through how the world has ended up here: why carbon is a problem unlike any other pollutant, the two levers the world has for dealing with it, where each of them stands today, and what India’s carbon market opening now means for founders building in this space.

Why This Is a Gigaton Problem, Not a Policy Footnote

Here’s the basic problem: every tonne of CO2 we put into the atmosphere stays there for centuries, trapping heat, and we’ve been putting in tens of billions of tonnes a year for decades. Left alone, there’s no natural process fast enough to undo it on a human timescale. So the question governments and companies have been wrestling with since the 1990s isn’t really whether to deal with carbon - it’s how.

Early attempts tried regulating carbon directly with emission standards, cleaner fuels, plant-level caps. That works for pollutants with few sources, but carbon comes from everything, so single-point regulation barely scratches the problem.

It was only through the 2010s that policy consensus converged on a different organizing idea: net zero. Net Zero focused on emitting no more carbon than you can remove, by some target date - for most countries, 2050; for India, 2070. Net zero is useful precisely because it doesn’t pretend there’s one fix. Global emissions run close to a little over 50 gigatons of CO2-equivalent a year, and getting to net zero means attacking that number from every direction at once: cleaning up the grid, electrifying transport, fixing food systems, protecting forests and oceans, decarbonizing heavy industry, and pulling carbon that’s already in the air back out of it. Each of those is its own multi-gigaton problem with its own cost curve and its own timeline.

As India scales up steel, cement, and aluminium production to meet demand, total emissions from these sectors keep rising, even as each tonne of output becomes more efficient. Steel and cement are the two largest emitters here: together they account for more than half of all industrial emissions, and around a fifth of India's total CO2 output. They are also growing fastest, with steel production up 8% and cement 10% last year alone, because they make the basic materials a country needs to build and urbanize. This is India's decarbonization problem in miniature: the biggest emitters are also the sectors it can least afford to slow down. India has committed to cutting emissions per unit of GDP by 47% by 2035, but that is an efficiency target, not an absolute one. The harder task is meeting this demand without building the kind of high-carbon plants and infrastructure that would keep emitting for decades to come.

Two levers exist for any tonne of industrial carbon: Capture (Don’t let carbon out at the source, before it reaches the atmosphere) or Carbon Credit (price it, and let someone else remove or avoid an equivalent tonne elsewhere). India is now moving on both at once.

Lever One: Capture

The first lever is to capture carbon at the source, before it ever reaches the atmosphere. Carbon Capture, Utilization, and Storage (CCUS) is the umbrella term.

Here we are concerned about industrial point-source capture, i.e, pulling CO2 out before it ever leaves the stack. Direct air capture, which pulls CO2 from ambient air rather than a smokestack, is mechanically similar but commercially behaves more like a removal pathway, so it shows up under credits in Lever Two instead.

Overall the value chain runs in four stages:

Capture: separating CO2 from flue gas or process emissions at the source such as a cement kiln, a steel furnace, a power plant.

Transport: pipeline or shipping to wherever it’s going next.

Usage: turning captured CO2 into products such as synthetic fuels, building materials, enhanced oil recovery, soda ash.

Storage: injecting it permanently into depleted reservoirs or saline aquifers.

Capture itself isn’t one technology. Multiple chemistries compete for the job, and which one makes sense depends on the economics of the process and the composition of the flue gas:

Post-combustion capture: scrubs CO2 from flue gas after fuel has already burned, usually with an amine solvent that absorbs CO2 and releases it under heat for storage. It’s the most retrofit-friendly option, which is why it’s the default for existing cement and steel plants.

Pre-combustion capture: converts fuel into a hydrogen-CO2 mix before burning, then strips the CO2 out, leaving hydrogen to burn cleanly. More efficient in principle, but it only works for new-build plants designed around it, not retrofits.

Oxy-fuel combustion: burns fuel in near-pure oxygen instead of air, producing a flue gas that’s almost entirely CO2 and water which is easier to separate, but the upfront cost of producing that much pure oxygen is high.

A fourth, nascent track is starting to show up alongside these three: biological or photonics-based conversion modules that skip separation altogether by converting CO2 straight into oxygen and biomass at the emission source. Even though we don’t have industrial scale yet, it’s a category worth watching as a handful of early-stage teams push toward pilot deployment.

Underlying all four is the same governing constraint: economics. Cost determines whether any of these can be deployed at scale. Capture, storage, and utilization each cost very different amounts and most CCUS estimates get this wrong by treating it as one number instead of three.

Capture which is pulling a dilute gas out of a much larger gas stream is cost intensive. In many industries, a capture system that can also turn by-products into usable inputs improves the economics meaningfully.

Storage is comparatively cheap once the capture problem is solved, but needs geology that can hold CO2 for centuries without leaking. Depleted oil and gas reservoirs and saline aquifers are the two viable options but sparsely available.

Utilisation is the part that can pay for itself, turning captured CO2 into synthetic fuels, building materials, or soda ash generates revenue that offsets the cost of capture.

Capture also needs continuous monitoring systems to keep count of the carbon captured. Continuous Emission Monitoring System (CEMS) units wired into capture systems, plus IoT and pressure sensors at storage sites are replacing the old approach of periodic manual stack testing. Real-time monitoring catches a leak immediately rather than at the next scheduled audit.

Globally, installed CCUS capacity is still only a little over 50 million tonnes a year today, though it could reach about 430 million tonnes by 2030. To play its role in net-zero pathways, CCUS would need to scale to several gigatonnes by 2050, so it is still very much an early-stage technology.

Lever Two: Carbon Credits

Policy can set emission limits, but not every industry can capture enough carbon at source to hit them. If you can’t capture a tonne, you can still account for it and that’s what carbon markets do.

Carbon credits make that possible. A company that can’t capture or reduce enough of its own emissions can pay another company that has avoided or removed carbon beyond its own requirement, by buying a credit representing that tonne of carbon. A carbon market is the mechanism for this at scale: government caps total emissions, issues a fixed pool of permits, and lets companies trade them. A firm that can cut a tonne for $10 sells its spare permit to one that would’ve spent $80 to cut the same tonne which results in the same outcome at a lower system-wide cost. That efficiency is why more than 70 compliance carbon markets now operate worldwide today, including the EU ETS, currently trading around €70-85 a tonne.

One credit equals one tonne of CO2 either avoided (a tonne that would’ve been emitted, but isn’t), reduced (efficiency gains, incremental), or removed (physically pulled out of the atmosphere and stored - trees, biochar, direct air capture). Removed is the hardest and most expensive to deliver, but the easiest to trust: you can measure carbon sitting in the ground rather than argue about what would have happened otherwise.

Compliance markets like the EU ETS are mandatory for covered entities and government-created. The voluntary market is opt-in: corporates buy credits to back net-zero claims, demand is discretionary, and quality varies enormously by credit type. Legacy renewable-energy and cookstove credits still trade for a few dollars a tonne while a durable removal bought on its own is much higher. That spread is the problem the market is now correcting for.

Around 2024, buyers, standards bodies, and rating agencies started asking whether these credits actually represented a real, additional, lasting benefit. Cookstove and older forest credits came under scrutiny over whether their claimed savings were real. Since then, the market has been moving toward durable removals, even though they cost ten to a hundred times more per tonne than the avoidance credits they're displacing.

💡 The price gap this creates is stark: a tonne of avoided emissions from an old cookstove project can sell for a few dollars. A tonne physically pulled from the air via direct air capture can sell for $500 or more all because one is provable and the other asks you to trust a counterfactual.

Removal splits into two families, and within each, durability rises with cost while readiness and scale fall.

Nature-based removal is cheaper and faster to deploy, but less permanent.

Afforestation and reforestation (ARR): plant or restore tree cover and run $5-50 a tonne but it’s reversible: fire, drought, pests, or a change in land use can release decades of stored carbon back into the atmosphere in a single event, which is exactly the failure mode that got older forest credits downgraded.

Soil carbon and regenerative agriculture: take a different path by cover cropping, reduced tillage, biochar amendments to farmland storing carbon in the ground while improving yields and farmer income. The historic weak point here was measurement: soil carbon is genuinely hard to verify at scale by sending people out with shovels, which kept costs high and credibility low.

Engineered removal is more durable but far more expensive.

Biochar: crop residue pyrolysed into stable char and tilled into soil. This trades at ~ $125-177 a tonne. Nature-based in feedstock but engineered in process, which gives it both a lower cost than pure engineered pathways and a durability story (centuries, not years). For India specifically, the feedstock economics are feasible since crop residue burning is a major air-quality problem already, so biochar turns a liability into a revenue stream.

Enhanced rock weathering (ERW): spreading crushed silicate rock on farmland, where it reacts with CO2 and locks it into stable carbonates over years to decades. This process sits further out at roughly $200-350.

BECCS (Bioenergy with Carbon Capture and Storage) and Direct Air Capture: These average above $500 a tonne today: durable over centuries, but capital-intensive and energy-hungry.

A credit is only as good as its proof, i.e, the evidence that the tonne it claims to represent was actually avoided, reduced, or removed. For years, that proof came from a person visiting a site once a year, sampling some soil or counting some trees, and filing a report months later. That’s slow enough that errors or double-counting could go undetected for years.

That’s now being replaced by always-on verification, known as digital Digital Measurement, Reporting, and Verification (dMRV): satellites tracking land use continuously, sensors feeding live data from the ground, and models filling in the gaps between physical measurements. A project monitored this way trades at a premium over one verified the old-fashioned way, because buyers can actually trust the numbers behind it.

India is moving on both levers

India's position is unusual: it has to cut emissions while still growing into a fully industrialized economy, which means it can't simply shrink its heavy industries the way a developed country might. So it is moving on both levers at once, capturing carbon at the source where it can, and building a market to price the carbon it can't yet capture.

This year, India introduced the Carbon Credit Trading Scheme (CCTS), its first compliance carbon market, aimed squarely at the heavy end of industry: nine sectors including steel, cement, aluminium, fertiliser, and petroleum refining, which together account for about 16% of India's total emissions. The mechanism is straightforward. Each company in these sectors gets a target for emissions per unit of output, set by the government and tightening each year. A company that does better than its target earns credits for the difference. A company that misses it must buy credits from the market to cover the gap.

What now remains is the final step: trading is expected to open in the coming months, once the floor and ceiling prices that will anchor the market are set.

On capture, the Union Budget 2026-27 announced a ₹20,000 crore scheme, spread over five years, to help heavy industry install carbon capture across power, steel, cement, refineries, and chemicals. Government's role is to lower that barrier through capital support, R&D funding, and shared pipeline and storage infrastructure between plants. It follows a national roadmap that runs pilots through 2030, builds industrial integration through 2035, and targets commercial scale by 2045. Some early pilots are already running: NTPC at Vindhyachal, Tata Steel at Jamshedpur, JSW Steel at Dolvi, ONGC's Gandhar project, and a Tamil Nadu plant converting captured CO2 into soda ash. India's projected installed capture capacity today is around 4 MtCO2 a year, against a global base of 430 MtCO2.

There are multiple startups that have already been working on the supply side. Equilibrium (a Kalaari portfolio company) runs eight projects across nine states - spanning 120,000 hectares and 150,000 smallholder farmers, with a pipeline that could deliver over 20 million tonnes of carbon removal, supported by a digital MRV stack. Varaha works with around 150,000 farmers across 20 projects in India, Nepal, and Bangladesh, and has signed biochar offtake deals with Google and Microsoft, among the largest of their kind globally. While these companies built their supply to sell into the global voluntary market first, this positions them well to become the catalysts for India’s own domestic trading once it opens.

Where Startups Can Build

At Kalaari, we’re optimistic about the whole carbon stack. Some opportunities lie on the capture frontier with tech-first platforms going after a real step-change in the cost of capturing carbon. The other is everything the market now needs to actually run: monitoring for industrial sites, tools to decide where captured carbon should go, compliance and pricing intelligence for companies newly under CCTS, and platforms that source high-integrity Indian removals for global buyers.

1. Tech-First Capture and Conversion Platforms

Any technologies and chemistries that can lower the unit economics of carbon capture while still scaling, is a significant startup opportunity. The bar is a real step-change in cost per tonne over incumbent CCS/CCU, not an incremental one.

2. MRV and Monitoring for Industrial Capture

Monitoring stacks for industrial capture sites, the CEMS/IoT layer described above, plus tooling that helps Indian industrial emitters navigate the international trade implications for affected industries.

3. Compliance and Market Intelligence for CCTS

CCTS-obligated entities need to track greenhouse gas reporting, intensity-versus-target position, and credit holdings under real penalty risk from mid-2026 - and once trading opens, they’ll need pricing and risk data, plus visibility into what any given tonne is allowed to claim across compliance, voluntary, and international schemes without double-counting. This leaves room for an India-first platform that covers all three: compliance-grade reporting, a “Bloomberg-lite for Indian carbon” once a real price exists, and the reconciliation layer underneath both.

4. High-Integrity Removal Origination

Platforms that source high-integrity Indian removals and sell them through international marketplaces sit on the most defensible end of the curve - software and MRV as the moat, not ownership of kilns or land.

The market just got its plumbing on both levers. Price discovery on one starts once credits are issued from October; capacity build-out on the other is just getting going. We’re keen to talk to Indian deeptech startups building anywhere in this stack - capture, MRV, registries, analytics, or origination. If that’s you, reach us at deeptech@kalaari.com.

---