Kjeldahl vs Dumas: Two Ways to Measure Nitrogen, One Decision to Make

In short: the Dumas method combusts the sample at 900–1,000 °C and measures the released nitrogen gas by thermal conductivity — three to five minutes, no chemicals. The Kjeldahl method digests the sample in sulfuric acid, distills the ammonia into boric acid and titrates it — roughly an hour on a modern automatic analyzer, most of it unattended. Both finish at the same protein figure through a nitrogen-to-protein conversion factor. The real differences are the status each method holds in the standards, the matrices each one tolerates, and where the money goes over five years. If you need the Kjeldahl chemistry first, our step-by-step Kjeldahl method guide covers it.

The two principles in one minute

Kjeldahl is wet chemistry in three acts: digestion in concentrated sulfuric acid with a catalyst on a heated block, steam distillation of the liberated ammonia, and titration of the boric-acid receiver solution. It measures organic nitrogen plus ammonia. It does not fully recover nitrate, nitrite or nitrogen bound in N–N linkages unless a reduction step is added — a known boundary, written into the standards themselves.

Dumas is combustion analysis: the sample burns in pure oxygen at high temperature, the gas stream is scrubbed of water and CO2, and the remaining nitrogen is quantified as N2 by a thermal-conductivity detector. Because combustion converts essentially all nitrogen forms, Dumas reads slightly higher than Kjeldahl on the same sample — typically a small gap on protein-rich matrices, a noticeable one on nitrate-rich material such as silage or leafy vegetables. Neither result is 'wrong'; they measure slightly different definitions of nitrogen, and the standards account for it.

Side by side

Where Dumas is the right tool

Credit where due. If your laboratory runs hundreds of samples a day on uniform, well-ground matrices — flour, milk powder, finished feed — Dumas earns its price: no acid, no fume handling, minimal hands-on time, and results fast enough for at-line process control. The safety argument is real as well; a combustion furnace replaces litres of hot concentrated acid. For high-throughput commercial testing laboratories, Dumas is often the correct purchase.

Where Kjeldahl still wins

Reference-method status. When a protein figure is disputed — a rejected shipment, a label challenge, an accreditation audit — the method named in the standard decides. The reference method in ISO 1871 (food and feed), ISO 20483 (cereals and pulses), ISO 8968 (milk) and AOAC 981.10 (meat) is Kjeldahl. Combustion methods are recognized alongside it, but in many supply chains the Kjeldahl figure is the one with legal standing.

Difficult and inhomogeneous samples. Kjeldahl accepts gram-scale portions, which averages out the heterogeneity of real material — whole-grain feeds, meat products, soil, fertilizer, wastewater sludge. A 200 mg Dumas aliquot of the same sample puts far more pressure on grinding and subsampling discipline.

Budget and infrastructure. An automatic Kjeldahl line — digester, analyzer, fume handling — costs a fraction of a combustion analyzer, needs no high-purity gas supply chain, and its consumables are commodity reagents available anywhere. For laboratories in regions where specialty gases are expensive or unreliable, that logistics argument alone can decide the question.

Existing accreditation. Switching methods means revalidation and, often, rewriting the accreditation scope. A laboratory accredited around Kjeldahl keeps Kjeldahl — this is inertia, but rational inertia.

The standards question buyers forget

The same analyte can carry two official method numbers. Nitrogen in meat, for example: AOAC 981.10 is the Kjeldahl reference method, AOAC 990.03 is the combustion method — both official, both citable. Before choosing an instrument, check which method your customers, regulators and inter-laboratory schemes cite. That answer narrows the instrument choice faster than any specification sheet. Our standards library maps the major ISO, AOAC and GB methods to the instruments that run them.

The five-year cost reality

Purchase price is the visible difference; the operating pattern is the deciding one. Dumas concentrates cost in the instrument and a steady per-sample spend on capsules, catalysts and gases. Kjeldahl spreads it thin: inexpensive reagents, but more labour if the analyzer is not automated — which is precisely the problem automatic Kjeldahl analyzers exist to remove. As a rough rule drawn from customer conversations: below about 30 samples a day, a fully automatic Kjeldahl line usually carries the lower five-year cost; at sustained hundreds-per-day volumes, Dumas wins on throughput alone.

Our position, stated plainly

We manufacture Kjeldahl nitrogen analyzers and the block digesters and autosamplers around them, so weigh our position accordingly. We think the honest summary is this: Dumas is the better answer for a narrow profile — very high throughput, uniform samples, generous budget, no reference-method constraint. For everyone else, an automated Kjeldahl line covers the standards, the matrices and the budget at once. If you are weighing models, the Kjeldahl buyer's guide and the three-question model selector are the shortest path to a shortlist.