Removing Hydrogen Sulphide from Compressed Biogas

Raw biogas from an anaerobic digester is mostly methane and carbon dioxide, and the number that causes the trouble is neither of them. Hydrogen sulphide, present in anything from the low hundreds to several thousand parts per million depending on the feedstock, is what stands between a digester and a saleable product.

For a compressed biogas plant it has to come out early and it has to come out reliably, and how that is done is one of the larger operating cost decisions the plant will make.

Why it has to go first

Three separate reasons, any one of which would be sufficient.

Corrosion. H2S with water gives an acidic condensate that attacks carbon steel, and it does so in the parts of the plant that are least convenient to replace: compressors, heat exchangers, and pipework. In an engine, the sulphur ends up in the oil as acid, which shortens the drain interval and then the engine.

Downstream poisoning. Sulphur is the classic catalyst poison, binding to active sites and not leaving. Any catalytic step after the digester, and any adsorbent used for upgrading, has a sulphur tolerance measured well below what raw biogas carries.

Specification. Compressed biogas sold as a fuel is bound by a specification that caps sulphur far below digester levels. It is a gate, not a target: the gas either meets it or it does not leave the site.

Removing H2S first also protects the carbon dioxide removal step that follows, which is the expensive part of the upgrading train and the one you least want to expose to an acid gas it was not designed for.

The scavengers

The simplest approach is a solid that reacts with H2S and is thrown away when it is full.

Iron sponge, iron oxide on a support, is the traditional version and still widely used. Impregnated activated carbon and proprietary media do the same job. The gas passes a bed, the sulphide is captured, and eventually the bed breaks through and is changed.

The economics are entirely governed by loading. Capital cost is low and the operation needs almost no attention, which makes it the right answer for a genuinely small H2S load. But consumption is stoichiometric: every kilogram of sulphur removed consumes media that must be bought, changed, and disposed of. Double the inlet concentration and you have doubled the running cost with nothing to offset it. On a plant with a high sulphur feedstock, a scavenger bed that looked cheap at commissioning becomes the largest consumable on site.

Chemical scrubbing

Caustic scrubbing absorbs H2S into an alkaline solution. It handles high loads, it responds quickly to a change in inlet concentration, and it is well understood.

What it produces is spent caustic containing sulphide, which is a disposal item with a cost and a compliance obligation attached. The plant has converted a gas-phase problem into a liquid-phase one. That is often the right trade, particularly where load is high and variable, but it should be entered into with the disposal route already identified rather than discovered later.

Biological scrubbing

A biological scrubber uses sulphide-oxidising bacteria, thiobacilli and relatives, to oxidise H2S to elemental sulphur or sulphate. The culture is held on packing in a vessel, gas passes through, and a controlled amount of air supplies the oxygen the organisms need.

The attraction is that the active agent reproduces. There is no stoichiometric reagent being consumed, no media to replace on a schedule, and no spent caustic stream to dispose of. On a plant with a meaningful sulphur load, that changes the operating cost structure rather than merely reducing it, which is why biological scrubbing has become the default on larger CBG installations.

The trade is that a biological system is a living one, with the obligations that implies. It needs to be kept in a temperature and pH band. It needs the right amount of oxygen, since too little starves the oxidation and too much drives it past sulphur to sulphate and acidifies the vessel. It responds to a step change in load over hours or days rather than immediately. And it needs feeding, because the culture requires nitrogen, phosphorus, and trace elements that the gas stream does not supply.

The failure that gets misdiagnosed

The characteristic biological scrubber problem is a slow rise in outlet H2S with nothing visibly wrong.

The instinctive reading is a fault in the vessel: channelling in the packing, a fan, a distribution problem. Occasionally it is. More often the culture has been underfed or has drifted out of balance, and the population that was doing the oxidation has declined. The vessel is fine. What is in it is not.

This is worth knowing because the two diagnoses lead to very different responses, and mechanical intervention on a nutrient problem costs a shutdown and fixes nothing. Before opening a vessel, the questions to answer are what the inlet concentration has been doing, whether the nutrient dosing has actually been maintained, what the pH and temperature have been over the past weeks, and whether the oxygen supply matches the current load.

Choosing between them

The load and its variability decide it, more than anything else.

Low and steady sulphur load favours a scavenger bed, because the media consumption stays small and nothing has to be attended to. High load favours biological scrubbing, because the consumable cost does not scale with the sulphur removed and there is no spent stream to dispose of. Highly variable or unpredictable load favours a chemical scrubber, or a hybrid arrangement where a biological stage takes the bulk and a polishing bed handles the peaks and guarantees the outlet.

The mistake worth avoiding is selecting on capital cost alone. The consumable and disposal costs over a year commonly exceed the difference in installed cost, and they are the numbers that determine whether the plant is worth running.

We supply gas purification technology for industrial streams, including H2S removal from compressed biogas, and we make the nutrient blend that keeps biological scrubber cultures active. If you are sizing a system, or you have one whose outlet has been creeping up, the operating figures are usually more informative than the drawings.