AI-created editorial still life of lemongrass, a weighing balance, oil flask and field notebook.

One Tonne of Lemongrass: The Questions Behind the Oil Yield

October 25, 2026 Dilip Parab

In this article

First, agree on the denominator
Research gives conditions, not a universal promise
Three recoveries, three very different calculations
Quality needs its own evidence
The notebook I would build before scaling

“How much oil will one tonne of lemongrass give?” It sounds like the question that should settle the business case. It is only the beginning. We also need to know what was weighed, how it was processed, what quality emerged and what the buyer will actually pay.

Our work at Sugandha begins with growing and steam-distilling lemongrass in Konkan. That makes the question personally important. But experience with our own crop does not justify promising the same recovery to every grower.

First, agree on the denominator

Recovery by mass is the weight of separated oil divided by the weight of material charged, multiplied by 100. If 1,000 kg of fresh grass gives 3 kg of oil, that is 0.3% recovery on a fresh-weight basis. A percentage reported on dried material cannot be compared directly with that number.

Wilting changes the water carried by the grass. A higher percentage after wilting does not by itself prove more oil was produced per original harvest. Record both the starting harvest weight and the weight entering the vessel. Also keep kilograms and litres separate; converting between them requires density.

These are arithmetic and measurement distinctions. They can change the apparent result without changing the farmer’s income.

Research gives conditions, not a universal promise

A study of Cymbopogon flexuosus across three agro-ecological zones reported oil yields of 0.17–0.24% on a fresh-weight basis and variation in composition.^[1] Those are results from the study’s conditions, not a forecast for Tithavli.

A separate CSIR-CIMAP study published in 2024 examined harvest times. In its experiment, the time giving greater oil content differed from the time giving higher citral percentage in lemongrass.^[2] This is a useful reminder that “more oil” and “the desired composition” need not move together. It is not sufficient evidence to prescribe one clock time for every farm.

Three recoveries, three very different calculations

The following is an illustration, not our current price quotation, a field result or an earnings promise. Assume 1,000 kg of fresh biomass, an oil selling price of ₹1,000/kg and ₹2,500 of total allocated cost per batch.

Illustrative recovery	Oil	Oil sales	Sales less assumed cost
0.2%	2 kg	₹2,000	−₹500
0.3%	3 kg	₹3,000	₹500
0.5%	5 kg	₹5,000	₹2,500

At those assumptions, break-even recovery is 0.25%. Change the selling price or cost, and break-even changes. A real calculation needs cultivation, harvesting, transport, fuel, labour, water, maintenance, depreciation, finance and selling costs. An omitted cost makes the apparent surplus look better. A rejected batch or delayed payment changes the picture again.

Hydrosol and spent biomass may have uses. I would give them revenue in a forecast only when a usable specification, customer and realistic handling cost exist. A by-product is not income merely because it leaves the still.

Quality needs its own evidence

Smell matters, but it cannot establish every part of a buyer’s specification. Research uses instrumental analysis, including gas chromatography, to examine composition.^[1] A commercial discussion should identify the intended application, agreed tests, sampling method and acceptance limits. A percentage of one constituent alone is not a complete description of an oil.

Keep a labelled retained sample where appropriate. If two batches smell different, a record of their source and handling is more useful than trying to reconstruct events from memory.

The notebook I would build before scaling

For each batch, record the plant identity or cultivar where known, plot, harvest date, fresh weight, any wilting period, vessel charge weight, process duration, separated oil mass, sensory observations and test reference. Add labour hours, fuel use, other costs and the eventual sale and payment date.

Then compare like with like. If harvest maturity, weather and distillation conditions all change together, the result cannot tell us which change caused the difference. A useful trial changes one main factor, repeats the comparison and reports variation as well as the best result.

That is the standard I want our published field notes to move towards. This article sets out the method; it does not present a completed controlled trial from our farm.

The most valuable number may be the one that stops us scaling a weak assumption.

Cover image: AI-created editorial illustration.

Sources and further reading

Sources checked 25 October 2026. The viewpoint, proposed methods and illustrative examples are the author’s; externally sourced findings are identified in the text.

Dilip Parab · Co-founder, Sugandha · Tithavli, Sindhudurg

Have a field observation or a question that could improve this discussion? Share it with us.