Karnataka’s government says independent laboratories found up to 31.2% ethanol in petrol sold as E20, a blend meant to contain 20%. The three public-sector oil marketing companies say the samples were not taken properly, and the Union Petroleum Ministry says all 893 of its own samples were within limits. Kerala has ordered checks at its pumps. The dispute is now a question of measurement: who sampled what, how, and with which test.
India’s switch to petrol blended with 20% ethanol, known as E20, has been one of the fastest fuel transitions anywhere. It is intended to cut crude-oil imports and support sugar-cane and grain farmers who supply the ethanol. Since last week, it has also become the subject of an unusual dispute between a state government and the Centre’s oil companies, conducted largely through competing laboratory results.
On Monday, Karnataka’s Food and Civil Supplies Minister, Rizwan Arshad, released the reports behind the state’s claim that petrol at some pumps contains far more ethanol than the standard allows. According to PTI, he said the state “stands by” its findings.
What Karnataka says it found
According to reports by PTI, The News Minute and Newslaundry, state officials collected 63 petrol samples from 21 September onwards. They were taken in seven districts, including Bengaluru, Kolar, Chikkaballapura, Ramanagara, Tumakuru and Mangaluru, from outlets of IOCL, BPCL and HPCL as well as Nayara, Shell and MRPL.
- First round: the samples were sent to the oil companies’ own laboratories. According to the state, none of the reports stated an ethanol percentage. “This made us suspicious, so we decided to send the samples to private NABL-accredited laboratories,” Arshad said, as quoted by The News Minute.
- Second round: the independent laboratories, which have not been named, reported ethanol contents including 31.2% in an IOCL sample and 30.5% in an HPCL sample. Other readings ranged downwards through 27%, 26.5% and 22.6% to a low of 13.5%. Arshad said nearly four out of five samples exceeded 20%.
Reports of the totals differ. The state minister refers to 63 samples. Chief Minister D. K. Shivakumar had referred to 67 locations, and Business Today cites 56 Karnataka samples in the Centre’s own testing. NSH has not seen the laboratory reports themselves.
How ethanol content is measured
The Bureau of Indian Standards specification for E20 petrol is IS 17021. According to PTI, its third amendment, issued in August 2025, lists accepted methods for measuring ethanol content:
- Fourier-transform infrared (FTIR) spectroscopy, which identifies ethanol by the infrared wavelengths its molecules absorb.
- Gas chromatography, which separates a fuel’s components and measures each.
- Water extraction, a simpler test. Ethanol mixes readily with water, so shaking a measured volume of petrol with water draws the ethanol into a lower aqueous layer, whose increase in volume indicates the ethanol content.
The water test is quick and cheap, and suits screening at the pump, but it is less precise than instrumental methods. The News Minute reports that Karnataka used a water-based separation test. Which methods the independent laboratories used for each sample is not stated in the reports NSH has read.
Sampling matters as much as the test. Ethanol absorbs moisture, and a sample exposed to water in a tank, a container or the air can give misleading results. Standard practice is to seal samples, record their chain of custody and keep retention samples for re-testing.
The other side
The oil marketing companies have challenged precisely these points. According to The News Minute, IOC, BPCL and HPCL say the sampling did not follow the prescribed procedure, and that some samples were collected in October without their representatives present. They say they have received no retention samples, outlet-wise details or laboratory reports. They maintain that their E20 conforms to IS 17021, and have asked for joint sampling and independent verification.
The Union Petroleum Ministry said on 7 November that all 893 samples it tested across 27 states and Union Territories contained between 19% and 21% ethanol, and that its Karnataka samples were within 19–20%.
Why the percentage matters
Ethanol contains about a third less energy per litre than petrol, so a higher blend delivers somewhat fewer kilometres per litre. Ethanol is also more corrosive to some older rubber, plastic and metal components, and attracts water.
Vehicles built for E20 are designed to cope; many older ones were designed for E10 or less. A 2021 NITI Aayog roadmap, cited by Newslaundry, estimated mileage losses with E20 of about 1–2% for vehicles designed for it. It put losses at about 6–7% for older cars and 3–4% for older two-wheelers.
Karnataka’s claims go much further. Arshad has spoken of mileage losses of 20–30% “in some cases” and warned that the blend “can cause serious damage to vehicle engines”. These figures are the state’s assertions and have not been independently tested. A blend of 30% instead of 20% would cut energy content by a few per cent, not by a quarter.
What Karnataka wants
The state’s demands go beyond re-testing. According to The News Minute and Newslaundry, Karnataka wants motorists to be able to choose between E10, E20, E30 and unblended petrol, the ethanol percentage displayed at the pump, independent testing and compensation for affected vehicle owners. Arshad has said he will convene a meeting with the oil companies. Some of these demands are matters of national fuel policy, which the Centre sets, and a state cannot implement them alone.
Kerala joins in
On Monday, Kerala’s minister Anoop Jacob ordered random inspections at pumps in Thiruvananthapuram, Ernakulam, Kozhikode and Kannur, according to Business Today. He said there was no confirmed case of excess blending in Kerala so far and that results were expected in about 10 days.
What would settle it
The scientific path to resolution is not complicated:
- Joint sampling: samples drawn in the presence of both sides, sealed and split.
- Accredited testing: analysis by an agreed laboratory using a gas-chromatography or FTIR method specified in the standard.
- Full disclosure: publication of results outlet by outlet.
Until that happens, consumers have two sets of numbers that cannot both be right, and no way to tell which is.
