Startup Deep Dive : Carbon Clean Solutions — Revenue fell 28.6% in 2025 while gross margin went from 2% to 46%

Carbon Clean Solutions booked $16.5 million (about ₹158 crore) of revenue in calendar 2025, a fall of 28.6% on the year before, and its Deloitte-audited accounts read like a company that had a good year. Both things are true at once: gross margin went from 2% of revenue in 2024 to 46% in 2025, the operating loss narrowed by a third, and cash burned in operations dropped from $44.8 million to $14.7 million (Companies House group accounts for the year ended 31 December 2025, filed 25 July 2026).

The company behind those numbers was started in 2009 by two IIT Kharagpur students who could not raise money in India, moved to London on a UK government grant in 2012, and went on to close what was then the largest equity round ever for point-source carbon capture: $150 million led by Chevron in May 2022. It now says its technology has captured more than 3 million tonnes of CO2 across 50-plus references, and it is, as of the July 2026 filing, out raising equity again. The figures below explain why a shrinking top line is the more interesting part of that story, and what still has to go right.

Quick facts

What they do

Carbon Clean sells equipment and chemistry that strip carbon dioxide out of industrial exhaust before it reaches the chimney. The buyers are cement kilns, steel plants, refineries, fertiliser producers, gas-fired power and energy-from-waste operators, and increasingly offshore oil and gas vessels; named clients in the company’s 2026 boilerplate include ADNOC, Aramco, NTPC Limited, Taiheiyo Cement and Tata Steel. The product line runs from a full-size CaptureX plant, through the containerised CycloneCC unit, to the solvent itself, and since 2025 the company also earns rental fees on units it owns and operates on a customer’s site. It does not transport or store the CO2 and it does not sell carbon credits; what it sells is the capture step, priced as equipment, engineering hours, solvent tonnes or, now, capacity by the month.

The origin

Aniruddha Sharma and Prateek Bumb met at IIT Kharagpur, where Sharma was doing a master’s in statistics and Bumb was studying engineering. They founded the company in 2009 while still students, at a time when, as Sharma told the Society of Petroleum Engineers’ TWA magazine in 2022, “it wasn’t common for students in India to start a business while still at university.” The founding bet was chemical rather than mechanical: conventional carbon capture used monoethanolamine (MEA) solvents that were energy hungry and corrosive, and the pair believed an amine-promoted buffer salt could do the same job with far less heat. The first patent filing, for Amine-Promoted Buffer Salts, came in 2010, according to the company’s own timeline. A University of Kentucky review cited by Scroll.in in January 2017 found the resulting CDRMax solvent captured more than 90% of CO2 while needing roughly 30% of the energy of a standard MEA plant. The Indian company was incorporated in Bengaluru on 12 October 2009 with a paid-up capital that Tofler still records at ₹30.92 lakh, a reminder of how small the starting point was.

The struggle years

The first business model failed inside six months. Sharma has described the original plan as a consulting firm advising power stations and steel plants, and the outcome in one sentence: “nobody wanted to get advice from 21-year-olds” (TWA, 2022). The pair pivoted to building technology with, by Sharma’s account, $26,000 from friends and family, enough to lease a small laboratory. Indian investors did not follow. Scroll.in reported in 2017 that the founders “couldn’t find investors for their company in India” and turned instead to the UK, which offered grants and entrepreneur visas; the company’s timeline dates the London relocation to 2012 and puts the first UK government grant at £3.5 million. The UK entity was incorporated on 25 June 2012, and a September 2015 raise of £3.4 million from Eldon Capital Management, reported by Scroll.in, was still the largest cheque the company had seen six years in. Revenue in those years came in part from an unglamorous side business: by June 2018 its Methpure solvent was running in 26 biogas-upgrading plants in Germany, processing 500,000 cubic metres of biogas a day (company release, June 2018).

The second hard stretch is more recent and is written into the audited accounts. Between the year ended 31 March 2023 and the year ended 31 December 2024, revenue rose from $6.0 million to $23.1 million, but gross margin collapsed from 14% to 2% as the mix shifted to fixed-price equipment supply contracts, and the loss after tax widened from $4.8 million to $33.3 million. Cash fell from $132.7 million at 31 March 2023 to $63.3 million at 31 December 2024. In 2024 the company also moved $117.9 million from its share premium account to distributable reserves through a capital reduction, a housekeeping step that makes future dividends legally possible but also signals how much of the Series C money had already been consumed as losses. Then came a customer failure: the 2025 accounts record that a customer with an outstanding receivable “has experienced significant financial difficulty,” a $568,000 trade receivable was fully provided for, and the company’s minority stake in the same entity, Liquid Wind AB of Gothenburg, carried at $2.8 million a year earlier, was written down to nil. Carbon Clean had won the full design-and-supply contract for Liquid Wind’s FlagshipONE e-methanol plant in 2023 (company release, 2023); the accounts also note that 2024 revenue included “a settlement agreement fee received in 2024 against Equipment Supply Sales revenue not repeated in 2025,” which explains part of the 2025 revenue fall.

The turning point

The event that changed the shape of the business was a 10-tonne-per-day box in the desert. In October 2023 ADNOC and Fertiglobe selected CycloneCC for a nitrogen fertiliser plant at the Al Ruwais industrial complex in Abu Dhabi (company release, 3 October 2023). On 3 April 2025 Carbon Clean announced the unit had completed roughly 4,000 operating hours over six months, was installed in under a week, delivered CO2 above the projected purity targets, and had taken the technology to Technology Readiness Level 7 (company release, April 2025). Six weeks later, on 15 May 2025, it announced that a commercial-scale CycloneCC C1 rotating packed bed, sized for up to 285 tonnes a day or about 100,000 tonnes a year, had passed factory acceptance tests at Thomas Broadbent and Sons in Huddersfield, a 20-fold scale-up on the Abu Dhabi unit (company release, May 2025).

The numbers either side of that demonstration are the clearest way to see what it did. In 2024, revenue was $23.1 million, of which $20.2 million was reported as equipment supply, and gross profit was $544,000 (2%). In 2025, revenue was $16.5 million, equipment supply fell to $8.6 million, equipment rental fees, carved out as a separate line for the first time in the 2025 accounts, rose from a restated $1.7 million to $4.75 million, engineering and consulting rose from $1.7 million to $2.7 million, and gross profit was $7.5 million (46%). Middle East revenue went from $1.7 million to $5.1 million. Cash used in operations fell 67.1%, from $44.8 million to $14.7 million (all figures from the audited group accounts for 2025 and 2024). The company had stopped selling one-off projects at cost and started earning from standard units it could rent.

The money behind it

Carbon Clean’s funding has come in three shapes: government grants, corporate venture money from the industries it wants to decarbonise, and one very large late round.

• 2012 to 2018, UK grants: £3.5 million on relocation in 2012, a second grant in 2015 for rotating packed bed absorber testing, and a third in 2018 to scale the RPB to 1 tonne per day (company timeline). The company’s July 2025 Global Innovation Centre release put cumulative UK government grant support at £5 million across 2012, 2015 and 2018.
• September 2015, £3.4 million: from Eldon Capital Management, reported by Scroll.in in January 2017.
• January 2019, $2.8 million: a US Department of Energy/NETL award shared with GTI for the ROTA-CAP rotating packed bed project, with field testing at the National Carbon Capture Center in Alabama (company release, January 2019).
• February 2020, $16 million: WAVE Equity Partners, Chevron Technology Ventures and Marubeni Corporation (company release, 17 February 2020).
• July 2020, $22 million Series B: Equinor Ventures and ICOS Capital as new investors, taking total raised to $38 million (company release, 2 July 2020); an $8 million extension with Cemex Ventures followed, closing the Series B at $30 million by August 2021 (company timeline; Series C release).
• May 2022, $150 million Series C: led by Chevron, with existing investors CEMEX Ventures, Marubeni and WAVE, and new investors AXA IM Alts, Samsung Ventures, Saudi Aramco Energy Ventures and TC Energy. The company put total private funding at $195 million (company release, 11 May 2022; TechCrunch, May 2022). Global Venturing reported Chevron’s share of the round as $100 million.

What the backers changed is visible in the register and the order book. Chevron is now the largest shareholder, with 78,616 Series C and 18,705 Series B shares, about 21.1% of the 461,513 shares in issue at the October 2025 confirmation statement, and Chevron’s Christopher Powers sits on the board; the relationship produced a CycloneCC pilot on a Chevron gas turbine in California’s San Joaquin Valley (company interview page). WAVE Equity, the 2020 lead, holds about 19.2% across its funds and its founder Praveen Sahay is a director. Cemex, Aramco and Samsung came as customers as much as investors: the Aramco and Samsung E&A demonstration agreement was announced in 2024, and Samsung E&A is the engineering partner on the MODEC offshore FEED signed on 27 February 2025. Not every backer stayed: the confirmation statement records TransCanada Energy Ltd’s 7,861 Series C shares transferred out on 15 September 2025. No round has disclosed a valuation, and none appears in the filings.

How it makes money

The revenue note in the 2025 accounts is the most honest description of the model, so it is worth walking through line by line.

• Equipment supply sales, $8.64 million in 2025 (2024: $18.49 million): fixed-price contracts to design and deliver capture plants, recognised on a percentage-of-completion basis. This was the line that broke the margin in 2024.
• Equipment rental fees, $4.75 million in 2025 (2024: $1.72 million, restated out of equipment supply): units the company owns and places on a customer’s site, the “carbon capture as a service” model its executives have pushed since late 2025. Technical equipment on the balance sheet jumped from $21,000 to $4.3 million during 2025 to support this.
• Engineering and consulting fees, $2.73 million (2024: $1.72 million): FEED studies such as the MODEC FPSO work and, from August 2026, a share of Uniper’s NorthStarH2 e-methanol FEED in Sweden, a package worth more than SEK 100 million across four engineering partners (Uniper/Carbon Clean release, 31 August 2026).
• Solvent supply, $347,000 (2024: $1.08 million): the original APBS-CDRMax and Methpure chemistry, sold at the point of delivery.
• Grant income, $23,000 (2024: $87,000): now negligible after a decade in which grants were the main lifeline.

Costs out are dominated by people: staff costs were $19.9 million in 2025 for an average of 169 employees (2024: $21.3 million for 175), split into 61 in customer delivery, 53 in innovation, 36 in corporate services, 11 in customer growth and 8 in management. Research and development expenditure was $4.1 million in 2025 against $8.9 million in 2024, and $2.0 million of it was capitalised onto a development asset that now stands at $16.4 million and represents the CycloneCC product line. Interest on cash deposits added $1.6 million. The part people get wrong is the cost-per-tonne headline. The $30 a tonne figure that has followed the company since the Tuticorin plant in 2016 is a company-stated target, not an audited number, and Sharma told Trellis in 2022 that current costs were “under $40 per metric ton.” The other misreading is that this is a carbon-credit business; it is a capital equipment and chemicals business whose customers’ economics depend on someone else providing the pipeline and the storage.

The numbers

Carbon Clean changed its year end from 31 March to 31 December in 2023, so the 2023 column below covers nine months. All figures are from Deloitte-audited consolidated accounts filed at Companies House and presented in US dollars, the group’s reporting currency; the rupee columns convert at $1 ≈ ₹96.0.

• Adjusted EBITDA: a loss of $22.4 million in 2025 against $34.6 million in 2024 (company-defined measure in the strategic report).
• Net assets: $71.7 million (about ₹688 crore) at 31 December 2025, down from $95.4 million a year earlier and $128.1 million at the end of 2023.
• Tax losses: $56.0 million of unutilised losses carried forward at 31 December 2025, with no deferred tax asset recognised because future profits are “not considered sufficiently probable.”
• Growth rankings: Deloitte verified three-year revenue growth of 1,426% for the 2024 EMEA Technology Fast 500 (rank 113) and 581% for the 2025 list (rank 300); the company also says the 2025 Sunday Times 100 Tech ranked it the seventh fastest-growing private hardware company in Britain.
• Legal-entity note: the standalone UK parent reported a loss of £17.75 million for 2025 (2024: £23.25 million) in sterling; the Indian company that started it all last disclosed operating revenue “under ₹1 crore” in March 2020 filings (Tofler) and is now a shareholder rather than an operating arm.

Where the money comes from

The geography split in the accounts is where the story gets uncomfortable for a company that describes itself as British with Indian roots.

• Europe excluding the UK: $9.20 million in 2025, 55.8% of revenue (2024: $20.67 million, 89.5%).
• Middle East: $5.05 million, 30.6% (2024: $1.72 million, 7.4%), the ADNOC and Fertiglobe work.
• Asia: $2.10 million, 12.7% (2024: $294,000, 1.3%), the first meaningful contribution from a region where it has BHP-JSW Steel, NTPC and Tata Steel relationships.
• North America: $110,000, 0.7% (2024: $321,000), despite a US headquarters opened in 2023 and Chevron as anchor shareholder.
• United Kingdom: $23,000, 0.1% (2024: $87,000), all of it grant income.

The surprise is the home market. The UK has supplied the grants, the visas, the manufacturing partner in Huddersfield and the registered office, and it delivered $23,000 of revenue in 2025. The company’s own December 2025 blog counts 27 UK carbon capture projects cancelled or paused since 2023, with delays averaging about two years. India is the mirror image: the Navi Mumbai Global Innovation Centre opened in July 2025 with more than 100 staff and space for 350, and the company said in June 2026 that Indian growth would support up to 250 UK manufacturing roles and £83 million for the UK economy over five years, yet Asia is still an eighth of revenue. The BHP-JSW joint study at Vijayanagar in Karnataka, which targets 100,000 tonnes of CO2 a day, is due to complete its feasibility phase in 2026; NTPC’s 20-tonne-a-day plant at Vindhyachal was awarded in August 2021; Tata Steel’s 5-tonne-a-day blast-furnace unit at Jamshedpur was commissioned in September 2021. These are demonstrations, not yet a market.

The risks

• Contract and counterparty risk, disclosed by the company. The strategic report lists contractual risk and delivery and supply chain risk first, including cost overruns on fixed-price contracts and tariff uncertainty. The mechanism showed up in 2025: contract assets, the unbilled work sitting on the balance sheet, swung from $14.75 million at end-2024 to $1.52 million a year later, and one customer’s failure cost a $568,000 receivable plus a $2.8 million investment write-off. With around fifty references and a handful of large contracts a year, a single delayed or cancelled project moves annual revenue by double-digit percentages, as 2025 demonstrated.
• Policy dependence. Point-source capture only pays when a customer faces a carbon price, a mandate or a subsidy and has somewhere to put the CO2. The company’s own December 2025 posts describe shortlisted projects withdrawing from the Danish CCS tender because “the financial risks are too great under the currently proposed commercial structure,” and 27 UK projects cancelled or paused since 2023. Its answer, a capture-as-a-service model that keeps the asset on Carbon Clean’s own balance sheet, transfers that policy risk onto the company: technical equipment rose from $21,000 to $4.3 million in 2025, and every rented unit is a bet that the customer’s incentive survives the lease.
• Cash and the open fundraise. Cash was $41.2 million at 31 December 2025 after operating outflows of $14.7 million in 2025 and $44.8 million in 2024, and the company has no financial debt. The directors state the business can fund itself for at least 12 months and that the equity raise underway is “not necessarily required” for working capital, while noting “positive investor engagement” and “some non-binding indications of interest.” The auditors raised no going-concern qualification. The dependency is elsewhere: the $16.4 million CycloneCC development asset was tested for impairment using the Teneo valuation prepared for that same fundraise, at a weighted cost of capital of 15% to 16.5%. If the raise prices poorly, the asset test gets harder in 2026.
• Technology and scale-up risk. CycloneCC reached TRL 7 in April 2025 on a 10-tonne-a-day unit. The 285-tonne-a-day C1 has passed a factory test but had not, as of the 2025 accounts, run on a live flue gas. The MODEC offshore pilot is scheduled for 2026 with a commercial unit of up to 100,000 tonnes a year to follow, and the strategic report names “potential competitor breakthroughs” as a principal risk. The company has 126 active patent assets across 30 countries (company boilerplate, 2026), but the gap between a factory-tested rotating packed bed and a fleet of them is where hardware companies most often stall.

The takeaway

The transferable lesson from Carbon Clean is that for a hardware company the unit of sale matters more than the size of the sale. For fifteen years the company sold projects: a plant at Tuticorin, a plant at Jamshedpur, a design-and-supply contract in Sweden. Projects brought revenue, and in 2024 they brought a record $23.1 million of it, at a 2% gross margin, with the customer’s own problems flowing straight back onto Carbon Clean’s balance sheet when one of them failed. The moment it had a standard box that could be installed in under a week and rented by the month, revenue fell and every other number improved: margin, operating loss, cash burn, the ratio of billed to unbilled work. Sharma’s founding pivot in 2009, from selling advice to selling a solvent, was the same move at a smaller scale. Founders building physical products for slow, regulated industries should notice that the buyer in those industries can rarely underwrite a bespoke project, but can usually sign a lease. Whether Carbon Clean can finance a fleet of leased units before its cash and its investors’ patience run out is the question the 2026 fundraise will answer, and it is a better question to be facing than the one it faced in 2024.

Frequently asked questions

Is Carbon Clean an Indian company?

It was founded in India in 2009 by two IIT Kharagpur graduates and the original Bengaluru company, Carbon Clean Solutions Private Limited, still exists and holds 10,000 shares in the group. Since 2012 the parent has been Carbon Clean Solutions Limited, a UK company, and the accounts are filed in London. Its largest single site is now the Global Innovation Centre in Navi Mumbai, opened in July 2025, and its Indian subsidiary is Carboncapture Technologies Private Limited.

How much money has Carbon Clean raised?

The company put total private funding at $195 million after the $150 million Series C in May 2022, on top of UK government grants it has variously described as £3.5 million initially and around £5 million cumulatively, plus a $2.8 million US Department of Energy award in 2019. Data providers such as Tracxn quote higher totals of about $243 million that include grants and prize money.

What is Carbon Clean’s valuation?

No valuation has been published for any round, and none appears in the Companies House filings. The 2025 accounts refer to a third-party valuation by Teneo prepared for an equity fundraise that was still open when the accounts were signed on 7 July 2026, but the number is not disclosed.

Is Carbon Clean profitable?

No. It reported a loss after tax of $24.2 million for 2025 and $33.3 million for 2024, and has unutilised tax losses of $56.0 million. Gross profit turned meaningfully positive for the first time in 2025 at $7.5 million, or 46% of revenue.

Who owns Carbon Clean?

By share count in the October 2025 confirmation statement, Chevron U.S.A. Inc. holds about 21%, WAVE Equity funds about 19%, the two founders about 4.5% each, Equinor and ICOS Capital about 4% each and AXA’s impact fund about 3.7%, with Cemex, Samsung, Japanese CCUS investment vehicles and dozens of angels making up the rest. Series B and Series C shares rank ahead of ordinary shares on a winding up.

Sources

Figures are as of September 2026. Currency converted at $1 ≈ ₹96.0 as of 18 September 2026 (Trading Economics).

• Companies House, Carbon Clean Solutions Limited (08116812), group accounts for the year ended 31 December 2025, filed July 2026
• Companies House, Carbon Clean Solutions Limited, group accounts for the year ended 31 December 2024, filed August 2025
• Companies House, Carbon Clean Solutions Limited, group accounts for the nine months ended 31 December 2023, filed July 2024
• Companies House, Carbon Clean Solutions Limited, confirmation statement dated 16 October 2025 (shareholder register), filed October 2025
• Companies House, Carbon Clean Solutions Limited, SH01 return of allotment and statement of capital, filed May 2026
• Companies House, company overview and filing history for 08116812, accessed September 2026
• Tofler, Carbon Clean Solutions Private Limited (U74900KA2009PTC051174) company profile, accessed September 2026
• Carbon Clean, press release: Carbon Clean raises $150m in record carbon capture funding round, May 2022
• Carbon Clean, press release: closes $22 million Series B round with Equinor and ICOS, July 2020
• Carbon Clean, press release: attracts $16 million growth capital, February 2020
• Carbon Clean, press release: successful completion of world’s first industrial deployment of CycloneCC (4,000 hours), April 2025
• Carbon Clean, press release: ADNOC selects CycloneCC for industrial project in the UAE, October 2023
• Carbon Clean, press release: successful factory test of CycloneCC C1 at commercial scale, May 2025
• Carbon Clean, press release: launch of Global Innovation Centre in India, July 2025
• Carbon Clean, press releases on MODEC: FEED contract with Samsung E&A, February 2025, and scale-up collaboration, June 2025
• Carbon Clean, press release: Carbon Clean’s growth in India boosts UK manufacturing, June 2026
• Carbon Clean, press release: appoints Guy Mansfield as Chief Financial Officer, May 2026
• Carbon Clean, press releases: TIME/Statista GreenTech ranking and Deloitte EMEA Technology Fast 500, June 2026; Deloitte EMEA Fast 500, May 2025
• Uniper and Carbon Clean, press release: Uniper advances NorthStarH2 into basic engineering, August 2026
• Carbon Clean, press release: NTPC selects Carbon Clean and GPIPL for 20 TPD capture plant, August 2021
• Carbon Clean, press release: FlagshipONE design contract with KBR and Liquid Wind, 2023; Liquid Wind partnership release, 2023
• Carbon Clean, press release: GTI and Carbon Clean receive $2.8m DOE funding, January 2019
• Carbon Clean, press release: solvents surpass 500k m3/day of biogas upgrading, June 2018
• Carbon Clean, Our journey (company timeline) and Chris Powers in conversation with Aniruddha Sharma (insights page), accessed September 2026
• Carbon Clean, insights: Chief Growth Officer blog on tender delays, December 2025; End of year reflections, December 2025
• Tata Steel, press release: commissions India’s first plant for CO2 capture from blast furnace gas at Jamshedpur, September 2021
• Hydrogen and Carbon Capture Technology World Expo news: BHP, Carbon Clean and JSW to explore CycloneCC, 2024
• TechCrunch, Carbon Clean raises $150M led by Chevron, May 2022
• Global Venturing, Analysis: Carbon Clean raises $150m, May 2022
• Trellis (formerly GreenBiz), Meet the startup shrinking industrial carbon capture, 2022
• Journal of Petroleum Technology / SPE TWA, interview with Carbon Clean CEO Aniruddha Sharma, 2022
• Scroll.in (republishing Quartz India), Two Indian engineers have crossed the biggest hurdle in making lower carbon emissions a reality, January 2017
• Yale Environment 360, Indian industrial plant converts captured CO2 into baking soda, January 2017
• Alfa Laval, customer story on the Tuticorin carbon capture plant, accessed September 2026
• Trading Economics, USD/INR rate, 18 September 2026

Found an error? Write to us and we’ll correct it in the open, dated, on the piece.